diff --git "a/checkpoint-610560/trainer_state.json" "b/checkpoint-610560/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-610560/trainer_state.json"
@@ -0,0 +1,7241 @@
+{
+  "best_metric": 4.156889915466309,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/pp-mod-subj2/lstm/3/checkpoints/checkpoint-610560",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 610560,
+  "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": 10.8194,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5719,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.067,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9956,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9424,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8558,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.6935,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.5846,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.4975,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4297,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.3811,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3144,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.263,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.1904,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.136,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.067,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.035,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.9783,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9322,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.8985,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8769,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8304,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8075,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7724,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979875363855125e-05,
+      "loss": 5.7408,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7232,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.6853,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.673,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.648,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6195,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6054,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5823,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5682,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5513,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714926921053515e-05,
+      "loss": 5.5308,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706540973542995e-05,
+      "loss": 5.5215,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5139,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.4756,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9681399509815164e-05,
+      "loss": 5.4657,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9673013562304644e-05,
+      "loss": 5.4519,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9664627614794124e-05,
+      "loss": 5.4357,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9656241667283604e-05,
+      "loss": 5.4216,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4229,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.4026,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963110020355577e-05,
+      "loss": 5.387,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962271425604525e-05,
+      "loss": 5.3803,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3695,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3622,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3381,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.327,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3347,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3237,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3118,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.2875,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954727348605804e-05,
+      "loss": 5.2966,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953888753854752e-05,
+      "loss": 5.275,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530501591037e-05,
+      "loss": 5.2905,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2563,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2599,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2398,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9496974179798646e-05,
+      "loss": 5.2303,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948858823228813e-05,
+      "loss": 5.2318,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948020228477761e-05,
+      "loss": 5.2149,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.21,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.2032,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945506082104978e-05,
+      "loss": 5.2125,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9446674873539255e-05,
+      "loss": 5.1849,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9438288926028735e-05,
+      "loss": 5.1752,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9429902978518215e-05,
+      "loss": 5.1602,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1583,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413147462300904e-05,
+      "loss": 5.1737,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404761514790384e-05,
+      "loss": 5.1619,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.939637556727987e-05,
+      "loss": 5.1444,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938798961976935e-05,
+      "loss": 5.1404,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937960367225883e-05,
+      "loss": 5.143,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1359,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1131,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.118,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.1037,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933770669231369e-05,
+      "loss": 5.1055,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932932074480317e-05,
+      "loss": 5.1022,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932093479729265e-05,
+      "loss": 5.0905,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931254884978213e-05,
+      "loss": 5.0925,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930416290227161e-05,
+      "loss": 5.0832,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.0675,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287407386054305e-05,
+      "loss": 5.0665,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279021438543785e-05,
+      "loss": 5.0716,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270635491033264e-05,
+      "loss": 5.0642,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262249543522744e-05,
+      "loss": 5.0513,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253863596012224e-05,
+      "loss": 5.0386,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9245494027305433e-05,
+      "loss": 5.0431,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9237108079794913e-05,
+      "loss": 5.0302,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0416,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0192,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 5.0148,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920358066855656e-05,
+      "loss": 5.032,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919519472104604e-05,
+      "loss": 5.0004,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918680877353552e-05,
+      "loss": 4.9947,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917842282602501e-05,
+      "loss": 4.9991,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917003687851449e-05,
+      "loss": 4.9971,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916165093100397e-05,
+      "loss": 4.9973,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 4.9856,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914489541478666e-05,
+      "loss": 4.9793,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913650946727614e-05,
+      "loss": 4.9719,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912812351976562e-05,
+      "loss": 4.9706,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911973757225509e-05,
+      "loss": 4.9546,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911136800354831e-05,
+      "loss": 4.9543,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910298205603779e-05,
+      "loss": 4.9592,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909459610852727e-05,
+      "loss": 4.9467,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908621016101675e-05,
+      "loss": 4.9415,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907782421350623e-05,
+      "loss": 4.9358,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906943826599571e-05,
+      "loss": 4.9273,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906105231848519e-05,
+      "loss": 4.9364,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905266637097467e-05,
+      "loss": 4.922,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9044296802267876e-05,
+      "loss": 4.9191,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9196,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027541286050565e-05,
+      "loss": 4.9136,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019155338540045e-05,
+      "loss": 4.9097,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010769391029525e-05,
+      "loss": 4.9072,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002383443519005e-05,
+      "loss": 4.9064,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9048,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8852,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8965,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8864,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8876,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8696,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.8786,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.8627,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892694267353179e-05,
+      "loss": 4.8685,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891855672602127e-05,
+      "loss": 4.8463,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891017077851075e-05,
+      "loss": 4.869,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890178483100023e-05,
+      "loss": 4.874,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8545,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.8404,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8522,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8459,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.8439,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851501903544574e-05,
+      "loss": 4.8368,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843115956034054e-05,
+      "loss": 4.8399,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834730008523534e-05,
+      "loss": 4.8305,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.8263,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.8124,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.823,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8183,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.8102,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.8228,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.8007,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.8102,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875930561734005e-05,
+      "loss": 4.8111,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875091966982953e-05,
+      "loss": 4.7998,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.812014102935791,
+      "eval_runtime": 329.0718,
+      "eval_samples_per_second": 1159.598,
+      "eval_steps_per_second": 36.238,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874253372231901e-05,
+      "loss": 4.7881,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873414777480849e-05,
+      "loss": 4.794,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872576182729797e-05,
+      "loss": 4.7965,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.7873,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.788,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.7708,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.777,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7661,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7728,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.7638,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.7689,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865035381491821e-05,
+      "loss": 4.7706,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864196786740769e-05,
+      "loss": 4.7564,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863358191989717e-05,
+      "loss": 4.7541,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862519597238665e-05,
+      "loss": 4.7389,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861681002487613e-05,
+      "loss": 4.7494,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860842407736561e-05,
+      "loss": 4.7428,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.7369,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.7337,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.7617,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.7338,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.7344,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.7383,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854972244479197e-05,
+      "loss": 4.7276,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.7265,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.7141,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852458098106414e-05,
+      "loss": 4.7234,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851619503355362e-05,
+      "loss": 4.718,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85078090860431e-05,
+      "loss": 4.7051,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849942313853258e-05,
+      "loss": 4.7132,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849103719102206e-05,
+      "loss": 4.7046,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848266762231527e-05,
+      "loss": 4.7136,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847428167480475e-05,
+      "loss": 4.7108,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846589572729423e-05,
+      "loss": 4.6974,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845750977978371e-05,
+      "loss": 4.706,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449123832273194e-05,
+      "loss": 4.7134,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440754263566404e-05,
+      "loss": 4.6825,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432368316055884e-05,
+      "loss": 4.6939,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423982368545363e-05,
+      "loss": 4.6896,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415596421034843e-05,
+      "loss": 4.6815,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840721047352432e-05,
+      "loss": 4.6821,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83988245260138e-05,
+      "loss": 4.6906,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839045495730701e-05,
+      "loss": 4.6842,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838206900979649e-05,
+      "loss": 4.678,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.6824,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836531349357918e-05,
+      "loss": 4.6748,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835692754606866e-05,
+      "loss": 4.679,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834854159855814e-05,
+      "loss": 4.6607,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834015565104763e-05,
+      "loss": 4.6578,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.674,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.6727,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.6614,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.652,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829824229229876e-05,
+      "loss": 4.6634,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828985634478824e-05,
+      "loss": 4.6453,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828147039727771e-05,
+      "loss": 4.6713,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827308444976719e-05,
+      "loss": 4.6471,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826469850225667e-05,
+      "loss": 4.6591,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825631255474615e-05,
+      "loss": 4.645,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824792660723563e-05,
+      "loss": 4.6375,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823954065972511e-05,
+      "loss": 4.6416,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231171091018326e-05,
+      "loss": 4.6462,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222785143507806e-05,
+      "loss": 4.6295,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214399195997286e-05,
+      "loss": 4.6322,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206013248486766e-05,
+      "loss": 4.6472,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197627300976246e-05,
+      "loss": 4.6271,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189241353465726e-05,
+      "loss": 4.6174,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180871784758935e-05,
+      "loss": 4.6184,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172502216052144e-05,
+      "loss": 4.6168,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164116268541624e-05,
+      "loss": 4.6382,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155730321031104e-05,
+      "loss": 4.6323,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147344373520584e-05,
+      "loss": 4.6167,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138958426010064e-05,
+      "loss": 4.6206,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813057247849955e-05,
+      "loss": 4.6233,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812218653098903e-05,
+      "loss": 4.622,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811380058347851e-05,
+      "loss": 4.6052,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810543101477172e-05,
+      "loss": 4.6188,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80970450672612e-05,
+      "loss": 4.6063,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808865911975068e-05,
+      "loss": 4.6141,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808027317224016e-05,
+      "loss": 4.6061,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807188722472964e-05,
+      "loss": 4.6047,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806350127721912e-05,
+      "loss": 4.6082,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80551153297086e-05,
+      "loss": 4.6081,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804672938219807e-05,
+      "loss": 4.5911,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803835981349129e-05,
+      "loss": 4.5996,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802997386598077e-05,
+      "loss": 4.5972,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802158791847025e-05,
+      "loss": 4.6029,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801320197095973e-05,
+      "loss": 4.5948,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800481602344921e-05,
+      "loss": 4.5827,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996446454742424e-05,
+      "loss": 4.5923,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79880605072319e-05,
+      "loss": 4.5834,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797967455972138e-05,
+      "loss": 4.5957,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797128861221086e-05,
+      "loss": 4.577,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796290266470034e-05,
+      "loss": 4.5814,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954533095993546e-05,
+      "loss": 4.5918,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946147148483026e-05,
+      "loss": 4.5816,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937761200972506e-05,
+      "loss": 4.5627,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929375253461986e-05,
+      "loss": 4.5757,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792098930595147e-05,
+      "loss": 4.579,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791260335844095e-05,
+      "loss": 4.5809,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790421741093043e-05,
+      "loss": 4.5698,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789584784222364e-05,
+      "loss": 4.5701,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788746189471312e-05,
+      "loss": 4.5639,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78790759472026e-05,
+      "loss": 4.5699,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787068999969208e-05,
+      "loss": 4.5512,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786232043098529e-05,
+      "loss": 4.5595,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785393448347477e-05,
+      "loss": 4.566,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784554853596425e-05,
+      "loss": 4.5617,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783716258845373e-05,
+      "loss": 4.5464,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782877664094321e-05,
+      "loss": 4.555,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782040707223642e-05,
+      "loss": 4.5441,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812021124725906e-05,
+      "loss": 4.5602,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803635177215386e-05,
+      "loss": 4.5462,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795249229704866e-05,
+      "loss": 4.5378,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786863282194346e-05,
+      "loss": 4.5567,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778493713487555e-05,
+      "loss": 4.5497,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770107765977035e-05,
+      "loss": 4.5439,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761721818466515e-05,
+      "loss": 4.5462,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7753335870955995e-05,
+      "loss": 4.544,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7744966302249204e-05,
+      "loss": 4.549,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7736580354738684e-05,
+      "loss": 4.5282,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728194407228164e-05,
+      "loss": 4.5482,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719808459717644e-05,
+      "loss": 4.5363,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771143889101086e-05,
+      "loss": 4.5396,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770305294350034e-05,
+      "loss": 4.5301,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769466699598982e-05,
+      "loss": 4.5355,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76862810484793e-05,
+      "loss": 4.5243,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767791147977251e-05,
+      "loss": 4.5326,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766952553226199e-05,
+      "loss": 4.5091,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766113958475147e-05,
+      "loss": 4.5433,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765275363724095e-05,
+      "loss": 4.5358,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764436768973043e-05,
+      "loss": 4.5304,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763599812102364e-05,
+      "loss": 4.5148,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762761217351312e-05,
+      "loss": 4.5308,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76192262260026e-05,
+      "loss": 4.5304,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761084027849208e-05,
+      "loss": 4.5232,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760245433098156e-05,
+      "loss": 4.5159,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594084762274773e-05,
+      "loss": 4.5234,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758571519356798e-05,
+      "loss": 4.5201,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757732924605746e-05,
+      "loss": 4.5264,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756894329854694e-05,
+      "loss": 4.5028,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756055735103642e-05,
+      "loss": 4.5184,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75521714035259e-05,
+      "loss": 4.5155,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754380183481911e-05,
+      "loss": 4.5093,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753541588730859e-05,
+      "loss": 4.5219,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752702993979807e-05,
+      "loss": 4.5025,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751864399228755e-05,
+      "loss": 4.5115,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751027442358077e-05,
+      "loss": 4.5097,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750188847607025e-05,
+      "loss": 4.5096,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.525084018707275,
+      "eval_runtime": 315.6819,
+      "eval_samples_per_second": 1208.783,
+      "eval_steps_per_second": 37.775,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749350252855973e-05,
+      "loss": 4.4977,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748511658104921e-05,
+      "loss": 4.5038,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747673063353868e-05,
+      "loss": 4.5097,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746834468602816e-05,
+      "loss": 4.5055,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745995873851764e-05,
+      "loss": 4.5043,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745157279100712e-05,
+      "loss": 4.4878,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74431868434966e-05,
+      "loss": 4.4965,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743480089598608e-05,
+      "loss": 4.4882,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742641494847556e-05,
+      "loss": 4.4984,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741802900096504e-05,
+      "loss": 4.4904,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740964305345452e-05,
+      "loss": 4.4977,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401257105944e-05,
+      "loss": 4.5,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392887537237216e-05,
+      "loss": 4.4825,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384501589726696e-05,
+      "loss": 4.4876,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376115642216176e-05,
+      "loss": 4.4701,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367729694705656e-05,
+      "loss": 4.4828,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359343747195135e-05,
+      "loss": 4.4828,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350957799684615e-05,
+      "loss": 4.4747,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342571852174095e-05,
+      "loss": 4.4752,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733418590466357e-05,
+      "loss": 4.5013,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732579995715305e-05,
+      "loss": 4.4784,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731741400964253e-05,
+      "loss": 4.4767,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4837,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730065849342522e-05,
+      "loss": 4.4802,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292288924718433e-05,
+      "loss": 4.4735,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728390297720791e-05,
+      "loss": 4.4641,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727551702969739e-05,
+      "loss": 4.4752,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726713108218687e-05,
+      "loss": 4.4674,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725876151348009e-05,
+      "loss": 4.4622,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725037556596957e-05,
+      "loss": 4.4703,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724198961845904e-05,
+      "loss": 4.4605,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723360367094852e-05,
+      "loss": 4.4724,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722523410224174e-05,
+      "loss": 4.4672,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721684815473121e-05,
+      "loss": 4.4577,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720846220722069e-05,
+      "loss": 4.4683,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720007625971017e-05,
+      "loss": 4.4789,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719169031219965e-05,
+      "loss": 4.4465,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71833371222966e-05,
+      "loss": 4.463,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717495117478608e-05,
+      "loss": 4.4526,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716656522727556e-05,
+      "loss": 4.4549,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715817927976504e-05,
+      "loss": 4.4495,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149793332254516e-05,
+      "loss": 4.464,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141407384743996e-05,
+      "loss": 4.4593,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133021437233476e-05,
+      "loss": 4.4522,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124635489722956e-05,
+      "loss": 4.4556,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116249542212436e-05,
+      "loss": 4.4476,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7107863594701916e-05,
+      "loss": 4.4602,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099494025995125e-05,
+      "loss": 4.4401,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091108078484605e-05,
+      "loss": 4.4399,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708272213097409e-05,
+      "loss": 4.4537,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707433618346357e-05,
+      "loss": 4.4514,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706595023595305e-05,
+      "loss": 4.4435,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705756428844253e-05,
+      "loss": 4.4392,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704917834093201e-05,
+      "loss": 4.4477,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704079239342149e-05,
+      "loss": 4.4303,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70324228247147e-05,
+      "loss": 4.4551,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702403687720418e-05,
+      "loss": 4.439,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701565092969366e-05,
+      "loss": 4.4472,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700726498218314e-05,
+      "loss": 4.4412,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699889541347635e-05,
+      "loss": 4.4306,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699050946596583e-05,
+      "loss": 4.4325,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982139897259045e-05,
+      "loss": 4.4406,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973753949748525e-05,
+      "loss": 4.4217,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965368002238005e-05,
+      "loss": 4.4287,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956982054727485e-05,
+      "loss": 4.4388,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948612486020694e-05,
+      "loss": 4.4302,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940226538510174e-05,
+      "loss": 4.4158,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931840590999654e-05,
+      "loss": 4.4181,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6923454643489134e-05,
+      "loss": 4.4162,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915068695978614e-05,
+      "loss": 4.4424,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906682748468094e-05,
+      "loss": 4.4303,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68983131797613e-05,
+      "loss": 4.4207,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688992723225078e-05,
+      "loss": 4.4291,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688154128474026e-05,
+      "loss": 4.4251,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687315533722974e-05,
+      "loss": 4.4287,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686480214732669e-05,
+      "loss": 4.4129,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685641619981617e-05,
+      "loss": 4.4276,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684803025230565e-05,
+      "loss": 4.4204,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683964430479513e-05,
+      "loss": 4.4213,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683125835728461e-05,
+      "loss": 4.4146,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682287240977409e-05,
+      "loss": 4.4177,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681448646226357e-05,
+      "loss": 4.4211,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680610051475305e-05,
+      "loss": 4.4231,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679773094604626e-05,
+      "loss": 4.4079,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678934499853574e-05,
+      "loss": 4.4133,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678097542982895e-05,
+      "loss": 4.4108,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677258948231843e-05,
+      "loss": 4.4175,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676420353480791e-05,
+      "loss": 4.4158,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675581758729739e-05,
+      "loss": 4.397,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674743163978687e-05,
+      "loss": 4.4098,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739045692276346e-05,
+      "loss": 4.4053,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730659744765826e-05,
+      "loss": 4.4155,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722273797255305e-05,
+      "loss": 4.3952,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671392060735225e-05,
+      "loss": 4.4046,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670553465984173e-05,
+      "loss": 4.4156,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669714871233121e-05,
+      "loss": 4.4032,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668876276482069e-05,
+      "loss": 4.3873,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668037681731017e-05,
+      "loss": 4.4005,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667199086979965e-05,
+      "loss": 4.4062,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666360492228913e-05,
+      "loss": 4.4072,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665521897477861e-05,
+      "loss": 4.3998,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664684940607182e-05,
+      "loss": 4.3928,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66384634585613e-05,
+      "loss": 4.3914,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663007751105078e-05,
+      "loss": 4.3985,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662169156354026e-05,
+      "loss": 4.3828,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661332199483347e-05,
+      "loss": 4.3894,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660493604732295e-05,
+      "loss": 4.3959,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659655009981243e-05,
+      "loss": 4.3946,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658816415230191e-05,
+      "loss": 4.3771,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657977820479139e-05,
+      "loss": 4.3911,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571408636084604e-05,
+      "loss": 4.379,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563022688574084e-05,
+      "loss": 4.3901,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554636741063564e-05,
+      "loss": 4.3855,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546250793553044e-05,
+      "loss": 4.3776,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537864846042524e-05,
+      "loss": 4.3917,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652949527733573e-05,
+      "loss": 4.3914,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652110932982521e-05,
+      "loss": 4.3794,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651272338231469e-05,
+      "loss": 4.3851,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650433743480417e-05,
+      "loss": 4.3853,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649595148729365e-05,
+      "loss": 4.3855,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648756553978313e-05,
+      "loss": 4.3723,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647919597107634e-05,
+      "loss": 4.3868,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647081002356582e-05,
+      "loss": 4.3805,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646242407605531e-05,
+      "loss": 4.3813,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645403812854479e-05,
+      "loss": 4.3768,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644565218103427e-05,
+      "loss": 4.3836,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643728261232748e-05,
+      "loss": 4.3634,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642889666481696e-05,
+      "loss": 4.3777,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642051071730644e-05,
+      "loss": 4.358,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641212476979592e-05,
+      "loss": 4.3882,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64037388222854e-05,
+      "loss": 4.379,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639535287477488e-05,
+      "loss": 4.3786,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6386983306068086e-05,
+      "loss": 4.3635,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378597358557566e-05,
+      "loss": 4.3795,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370211411047046e-05,
+      "loss": 4.3821,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361825463536526e-05,
+      "loss": 4.3709,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353439516026006e-05,
+      "loss": 4.3677,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345053568515486e-05,
+      "loss": 4.376,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336667621004966e-05,
+      "loss": 4.3697,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328281673494446e-05,
+      "loss": 4.3769,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319912104787655e-05,
+      "loss": 4.3581,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311526157277135e-05,
+      "loss": 4.3689,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303140209766615e-05,
+      "loss": 4.367,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294754262256095e-05,
+      "loss": 4.36,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286368314745575e-05,
+      "loss": 4.379,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277998746038784e-05,
+      "loss": 4.3548,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269612798528264e-05,
+      "loss": 4.3656,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261226851017744e-05,
+      "loss": 4.3688,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625284090350723e-05,
+      "loss": 4.3616,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.390535354614258,
+      "eval_runtime": 309.1781,
+      "eval_samples_per_second": 1234.211,
+      "eval_steps_per_second": 38.57,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624445495599671e-05,
+      "loss": 4.3619,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623606900848619e-05,
+      "loss": 4.3586,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622768306097567e-05,
+      "loss": 4.3688,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621929711346515e-05,
+      "loss": 4.3645,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621094392356209e-05,
+      "loss": 4.3632,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620255797605157e-05,
+      "loss": 4.3512,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619417202854105e-05,
+      "loss": 4.3577,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618578608103053e-05,
+      "loss": 4.3412,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617740013352001e-05,
+      "loss": 4.3647,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616901418600949e-05,
+      "loss": 4.3509,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616062823849897e-05,
+      "loss": 4.3606,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615224229098845e-05,
+      "loss": 4.3617,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614390547988913e-05,
+      "loss": 4.3501,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61355195323786e-05,
+      "loss": 4.3451,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612713358486808e-05,
+      "loss": 4.3406,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611874763735756e-05,
+      "loss": 4.3426,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611036168984704e-05,
+      "loss": 4.3497,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610197574233652e-05,
+      "loss": 4.3393,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093589794826e-05,
+      "loss": 4.342,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608520384731548e-05,
+      "loss": 4.365,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607681789980496e-05,
+      "loss": 4.3468,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606843195229444e-05,
+      "loss": 4.3462,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606004600478392e-05,
+      "loss": 4.3521,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60516600572734e-05,
+      "loss": 4.3476,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604327410976288e-05,
+      "loss": 4.3424,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603488816225236e-05,
+      "loss": 4.332,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602650221474184e-05,
+      "loss": 4.3468,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601813264603506e-05,
+      "loss": 4.3322,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600974669852454e-05,
+      "loss": 4.3356,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600136075101402e-05,
+      "loss": 4.3365,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599297480350349e-05,
+      "loss": 4.3327,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598458885599297e-05,
+      "loss": 4.3469,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597620290848245e-05,
+      "loss": 4.3376,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596781696097193e-05,
+      "loss": 4.3297,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595944739226514e-05,
+      "loss": 4.3409,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595106144475462e-05,
+      "loss": 4.3491,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59426754972441e-05,
+      "loss": 4.327,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934289549733587e-05,
+      "loss": 4.3376,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5925903602223066e-05,
+      "loss": 4.3237,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917517654712546e-05,
+      "loss": 4.3376,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909131707202026e-05,
+      "loss": 4.3221,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900762138495235e-05,
+      "loss": 4.3409,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892376190984715e-05,
+      "loss": 4.3348,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5883990243474195e-05,
+      "loss": 4.3297,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5875604295963675e-05,
+      "loss": 4.3344,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867218348453155e-05,
+      "loss": 4.3246,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858832400942635e-05,
+      "loss": 4.3377,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5850446453432115e-05,
+      "loss": 4.3206,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5842076884725324e-05,
+      "loss": 4.3166,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833690937214804e-05,
+      "loss": 4.3336,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582532136850802e-05,
+      "loss": 4.3274,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58169354209975e-05,
+      "loss": 4.3286,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580854947348698e-05,
+      "loss": 4.3156,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580016352597646e-05,
+      "loss": 4.3279,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579177757846594e-05,
+      "loss": 4.3121,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578339163095542e-05,
+      "loss": 4.3341,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57750056834449e-05,
+      "loss": 4.3193,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576661973593438e-05,
+      "loss": 4.331,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575823378842385e-05,
+      "loss": 4.3242,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574984784091333e-05,
+      "loss": 4.3162,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574146189340281e-05,
+      "loss": 4.3129,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573309232469602e-05,
+      "loss": 4.3232,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572470637718551e-05,
+      "loss": 4.3068,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571632042967499e-05,
+      "loss": 4.3149,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570793448216447e-05,
+      "loss": 4.3203,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569954853465395e-05,
+      "loss": 4.3192,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569116258714343e-05,
+      "loss": 4.2983,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568277663963291e-05,
+      "loss": 4.3053,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567439069212239e-05,
+      "loss": 4.3011,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56660211234156e-05,
+      "loss": 4.3307,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565765155470881e-05,
+      "loss": 4.3155,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564928198600202e-05,
+      "loss": 4.3107,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5640896038491496e-05,
+      "loss": 4.3162,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632510090980976e-05,
+      "loss": 4.3112,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562412414347046e-05,
+      "loss": 4.3236,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561573819595994e-05,
+      "loss": 4.2978,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560735224844942e-05,
+      "loss": 4.3194,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55989663009389e-05,
+      "loss": 4.3067,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559058035342838e-05,
+      "loss": 4.3104,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558219440591786e-05,
+      "loss": 4.3086,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557380845840734e-05,
+      "loss": 4.3049,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556542251089682e-05,
+      "loss": 4.3138,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55570365633863e-05,
+      "loss": 4.3098,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554866699467951e-05,
+      "loss": 4.2979,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554028104716899e-05,
+      "loss": 4.3081,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553189509965847e-05,
+      "loss": 4.2969,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552350915214795e-05,
+      "loss": 4.3125,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551513958344116e-05,
+      "loss": 4.3055,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550675363593064e-05,
+      "loss": 4.2895,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498384067223856e-05,
+      "loss": 4.3056,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5489998119713336e-05,
+      "loss": 4.2966,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481612172202816e-05,
+      "loss": 4.3059,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473226224692296e-05,
+      "loss": 4.2914,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5464856655985505e-05,
+      "loss": 4.2989,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456470708474985e-05,
+      "loss": 4.3084,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448084760964465e-05,
+      "loss": 4.295,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439698813453945e-05,
+      "loss": 4.2869,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431312865943425e-05,
+      "loss": 4.2944,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422926918432905e-05,
+      "loss": 4.2938,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414540970922385e-05,
+      "loss": 4.3072,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406171402215594e-05,
+      "loss": 4.2941,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539778545470508e-05,
+      "loss": 4.2917,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538939950719456e-05,
+      "loss": 4.2891,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538101355968404e-05,
+      "loss": 4.2975,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372627612173514e-05,
+      "loss": 4.2756,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364241664662994e-05,
+      "loss": 4.2902,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535587209595621e-05,
+      "loss": 4.2913,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534748614844569e-05,
+      "loss": 4.2937,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533910020093516e-05,
+      "loss": 4.2715,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533071425342464e-05,
+      "loss": 4.2916,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532232830591412e-05,
+      "loss": 4.2846,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53139423584036e-05,
+      "loss": 4.2822,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530555641089308e-05,
+      "loss": 4.289,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52971868421863e-05,
+      "loss": 4.2737,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528880089467578e-05,
+      "loss": 4.296,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528043132596899e-05,
+      "loss": 4.29,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527204537845847e-05,
+      "loss": 4.2812,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526365943094795e-05,
+      "loss": 4.2888,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525527348343743e-05,
+      "loss": 4.2823,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524688753592691e-05,
+      "loss": 4.2887,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523850158841639e-05,
+      "loss": 4.271,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523011564090587e-05,
+      "loss": 4.2928,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522172969339535e-05,
+      "loss": 4.2794,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213376503492285e-05,
+      "loss": 4.2817,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520499055598177e-05,
+      "loss": 4.2806,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519660460847125e-05,
+      "loss": 4.2875,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518821866096073e-05,
+      "loss": 4.2726,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517983271345021e-05,
+      "loss": 4.2764,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517144676593969e-05,
+      "loss": 4.2649,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516306081842917e-05,
+      "loss": 4.2892,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515467487091865e-05,
+      "loss": 4.2832,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514632168101559e-05,
+      "loss": 4.2832,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513793573350507e-05,
+      "loss": 4.2689,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512954978599455e-05,
+      "loss": 4.2828,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512116383848403e-05,
+      "loss": 4.2855,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511277789097351e-05,
+      "loss": 4.2759,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510439194346299e-05,
+      "loss": 4.2721,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509600599595247e-05,
+      "loss": 4.2811,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087620048441956e-05,
+      "loss": 4.2792,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079234100931436e-05,
+      "loss": 4.2807,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070864532224645e-05,
+      "loss": 4.2649,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062478584714125e-05,
+      "loss": 4.2765,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054092637203605e-05,
+      "loss": 4.273,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5045706689693085e-05,
+      "loss": 4.2722,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5037337120986294e-05,
+      "loss": 4.2801,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5028967552279503e-05,
+      "loss": 4.266,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020581604768983e-05,
+      "loss": 4.2743,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501219565725846e-05,
+      "loss": 4.2713,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500380970974794e-05,
+      "loss": 4.2752,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.3080854415893555,
+      "eval_runtime": 304.2229,
+      "eval_samples_per_second": 1254.314,
+      "eval_steps_per_second": 39.198,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499542376223742e-05,
+      "loss": 4.2653,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498703781472691e-05,
+      "loss": 4.2706,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497865186721639e-05,
+      "loss": 4.2789,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497026591970587e-05,
+      "loss": 4.2699,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496187997219535e-05,
+      "loss": 4.2773,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495349402468482e-05,
+      "loss": 4.2591,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49451080771743e-05,
+      "loss": 4.2668,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493672212966378e-05,
+      "loss": 4.2546,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492833618215326e-05,
+      "loss": 4.2738,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491995023464274e-05,
+      "loss": 4.2671,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491156428713222e-05,
+      "loss": 4.2681,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49031783396217e-05,
+      "loss": 4.2699,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489479239211118e-05,
+      "loss": 4.2627,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488640644460066e-05,
+      "loss": 4.2572,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487802049709014e-05,
+      "loss": 4.2591,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486963454957962e-05,
+      "loss": 4.2483,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486126498087284e-05,
+      "loss": 4.2645,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485289541216605e-05,
+      "loss": 4.2522,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484450946465553e-05,
+      "loss": 4.2545,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483612351714501e-05,
+      "loss": 4.2834,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482773756963449e-05,
+      "loss": 4.2561,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481935162212397e-05,
+      "loss": 4.2623,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481096567461345e-05,
+      "loss": 4.2655,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480257972710293e-05,
+      "loss": 4.26,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479419377959241e-05,
+      "loss": 4.2588,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478580783208188e-05,
+      "loss": 4.2482,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4777438263375096e-05,
+      "loss": 4.2638,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4769052315864576e-05,
+      "loss": 4.2422,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4760666368354056e-05,
+      "loss": 4.2549,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752280420843536e-05,
+      "loss": 4.2478,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4743894473333016e-05,
+      "loss": 4.2493,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473552490462623e-05,
+      "loss": 4.2625,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472713895711571e-05,
+      "loss": 4.2554,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4718753009605185e-05,
+      "loss": 4.25,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710367062094665e-05,
+      "loss": 4.2574,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470199749338788e-05,
+      "loss": 4.2643,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693611545877354e-05,
+      "loss": 4.2468,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685225598366834e-05,
+      "loss": 4.2553,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676839650856314e-05,
+      "loss": 4.241,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668453703345794e-05,
+      "loss": 4.2555,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466006775583528e-05,
+      "loss": 4.2351,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465168180832476e-05,
+      "loss": 4.2617,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464329586081424e-05,
+      "loss": 4.2525,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463492629210745e-05,
+      "loss": 4.2485,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462654034459693e-05,
+      "loss": 4.2536,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461817077589014e-05,
+      "loss": 4.2452,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460978482837962e-05,
+      "loss": 4.2554,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46013988808691e-05,
+      "loss": 4.241,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459301293335858e-05,
+      "loss": 4.2359,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458464336465179e-05,
+      "loss": 4.2533,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576273795945004e-05,
+      "loss": 4.2476,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567887848434484e-05,
+      "loss": 4.2479,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559501900923964e-05,
+      "loss": 4.2372,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551115953413444e-05,
+      "loss": 4.245,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542730005902923e-05,
+      "loss": 4.2373,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534344058392403e-05,
+      "loss": 4.2514,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4525958110881883e-05,
+      "loss": 4.2385,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451757216337136e-05,
+      "loss": 4.2531,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450918621586084e-05,
+      "loss": 4.2454,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450080026835032e-05,
+      "loss": 4.2371,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44924143208398e-05,
+      "loss": 4.2371,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448402837332928e-05,
+      "loss": 4.2445,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447565880462249e-05,
+      "loss": 4.2269,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446727285711197e-05,
+      "loss": 4.2382,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445888690960145e-05,
+      "loss": 4.2408,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445050096209093e-05,
+      "loss": 4.245,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444213139338415e-05,
+      "loss": 4.2199,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443374544587363e-05,
+      "loss": 4.2303,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442535949836311e-05,
+      "loss": 4.2225,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441698992965632e-05,
+      "loss": 4.2495,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44086039821458e-05,
+      "loss": 4.2372,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440021803463528e-05,
+      "loss": 4.2372,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439183208712476e-05,
+      "loss": 4.2379,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4383462518417966e-05,
+      "loss": 4.2345,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4375076570907446e-05,
+      "loss": 4.2512,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4366690623396926e-05,
+      "loss": 4.2205,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4358304675886406e-05,
+      "loss": 4.2399,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4349918728375886e-05,
+      "loss": 4.2354,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341532780865366e-05,
+      "loss": 4.234,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333146833354846e-05,
+      "loss": 4.2357,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432477726464806e-05,
+      "loss": 4.231,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431639131713754e-05,
+      "loss": 4.2362,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308005369627015e-05,
+      "loss": 4.2368,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299619422116495e-05,
+      "loss": 4.2227,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291233474605975e-05,
+      "loss": 4.236,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428286390589919e-05,
+      "loss": 4.2235,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42744943371924e-05,
+      "loss": 4.2372,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426610838968188e-05,
+      "loss": 4.2295,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425772244217136e-05,
+      "loss": 4.2179,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424933649466084e-05,
+      "loss": 4.2311,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424095054715032e-05,
+      "loss": 4.2242,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42325645996398e-05,
+      "loss": 4.2338,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422417865212928e-05,
+      "loss": 4.2154,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421579270461876e-05,
+      "loss": 4.23,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420742313591197e-05,
+      "loss": 4.241,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419903718840145e-05,
+      "loss": 4.2208,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419065124089093e-05,
+      "loss": 4.2103,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418226529338041e-05,
+      "loss": 4.2213,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417387934586989e-05,
+      "loss": 4.2252,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416549339835937e-05,
+      "loss": 4.2317,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415710745084885e-05,
+      "loss": 4.2254,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414872150333833e-05,
+      "loss": 4.2197,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414033555582781e-05,
+      "loss": 4.2159,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4131965987121024e-05,
+      "loss": 4.2269,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123580039610504e-05,
+      "loss": 4.2033,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115194092099984e-05,
+      "loss": 4.2193,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410682452339319e-05,
+      "loss": 4.2197,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409843857588267e-05,
+      "loss": 4.2242,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409005262837215e-05,
+      "loss": 4.2015,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408166668086163e-05,
+      "loss": 4.221,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407329711215484e-05,
+      "loss": 4.2137,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406491116464432e-05,
+      "loss": 4.216,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40565252171338e-05,
+      "loss": 4.2153,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404813926962328e-05,
+      "loss": 4.2079,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403975332211276e-05,
+      "loss": 4.2216,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403138375340598e-05,
+      "loss": 4.2218,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402299780589546e-05,
+      "loss": 4.2105,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401461185838494e-05,
+      "loss": 4.2204,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400622591087442e-05,
+      "loss": 4.2106,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39978399633639e-05,
+      "loss": 4.2212,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398945401585338e-05,
+      "loss": 4.2046,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398106806834285e-05,
+      "loss": 4.2261,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972698499636066e-05,
+      "loss": 4.2055,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964312552125546e-05,
+      "loss": 4.2154,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3955926604615026e-05,
+      "loss": 4.2068,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39475406571045e-05,
+      "loss": 4.2202,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393915470959398e-05,
+      "loss": 4.208,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3930785140887195e-05,
+      "loss": 4.2106,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922399193376675e-05,
+      "loss": 4.1938,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914013245866155e-05,
+      "loss": 4.2255,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905627298355635e-05,
+      "loss": 4.2103,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897241350845115e-05,
+      "loss": 4.2129,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3888871782138324e-05,
+      "loss": 4.1996,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880485834627804e-05,
+      "loss": 4.217,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872099887117284e-05,
+      "loss": 4.2212,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863713939606764e-05,
+      "loss": 4.2061,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855327992096244e-05,
+      "loss": 4.2041,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384695842338945e-05,
+      "loss": 4.2137,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383857247587893e-05,
+      "loss": 4.2145,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383018652836842e-05,
+      "loss": 4.2155,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38218005808579e-05,
+      "loss": 4.1964,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381341463334738e-05,
+      "loss": 4.2098,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380502868583686e-05,
+      "loss": 4.2047,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379664273832634e-05,
+      "loss": 4.2081,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378825679081582e-05,
+      "loss": 4.2135,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377988722210903e-05,
+      "loss": 4.1974,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377150127459851e-05,
+      "loss": 4.2139,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376311532708799e-05,
+      "loss": 4.2026,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375472937957747e-05,
+      "loss": 4.2147,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.25103235244751,
+      "eval_runtime": 313.3193,
+      "eval_samples_per_second": 1217.898,
+      "eval_steps_per_second": 38.06,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374634343206695e-05,
+      "loss": 4.1995,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373795748455643e-05,
+      "loss": 4.2076,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372957153704591e-05,
+      "loss": 4.2101,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372120196833912e-05,
+      "loss": 4.2042,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37128160208286e-05,
+      "loss": 4.2105,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3704430073318084e-05,
+      "loss": 4.1967,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3696044125807564e-05,
+      "loss": 4.2013,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368765817829704e-05,
+      "loss": 4.1911,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367927223078652e-05,
+      "loss": 4.2092,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3670886283276e-05,
+      "loss": 4.2066,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366251671456921e-05,
+      "loss": 4.2014,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3654130767058686e-05,
+      "loss": 4.207,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36457611983519e-05,
+      "loss": 4.1957,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363737525084138e-05,
+      "loss": 4.1964,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362898930333086e-05,
+      "loss": 4.1954,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620603355820335e-05,
+      "loss": 4.1864,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361221740830982e-05,
+      "loss": 4.1983,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36038314607993e-05,
+      "loss": 4.1884,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359544551328878e-05,
+      "loss": 4.1919,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358705956577826e-05,
+      "loss": 4.2272,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357867361826774e-05,
+      "loss": 4.1905,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357030404956095e-05,
+      "loss": 4.2008,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356191810205043e-05,
+      "loss": 4.2017,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355353215453991e-05,
+      "loss": 4.1999,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354514620702939e-05,
+      "loss": 4.1919,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353676025951887e-05,
+      "loss": 4.1905,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352837431200835e-05,
+      "loss": 4.1992,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351998836449783e-05,
+      "loss": 4.1802,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351160241698731e-05,
+      "loss": 4.1901,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350323284828052e-05,
+      "loss": 4.1863,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3494846900770006e-05,
+      "loss": 4.191,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486477332063216e-05,
+      "loss": 4.1973,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478091384552696e-05,
+      "loss": 4.1977,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3469705437042175e-05,
+      "loss": 4.1899,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461319489531655e-05,
+      "loss": 4.1963,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3452949920824865e-05,
+      "loss": 4.2009,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3444563973314344e-05,
+      "loss": 4.1851,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436178025803824e-05,
+      "loss": 4.1988,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3427792078293304e-05,
+      "loss": 4.1765,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419406130782784e-05,
+      "loss": 4.1999,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411020183272264e-05,
+      "loss": 4.1714,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402634235761744e-05,
+      "loss": 4.2034,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394248288251224e-05,
+      "loss": 4.1913,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338587871954444e-05,
+      "loss": 4.1893,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337750915083765e-05,
+      "loss": 4.1917,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336912320332713e-05,
+      "loss": 4.1877,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336073725581661e-05,
+      "loss": 4.191,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335235130830609e-05,
+      "loss": 4.18,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334396536079557e-05,
+      "loss": 4.1706,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333557941328505e-05,
+      "loss": 4.1992,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332719346577452e-05,
+      "loss": 4.1889,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3318807518264e-05,
+      "loss": 4.1932,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331043794955722e-05,
+      "loss": 4.1783,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33020520020467e-05,
+      "loss": 4.1861,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329366605453618e-05,
+      "loss": 4.1793,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328528010702566e-05,
+      "loss": 4.1905,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327689415951514e-05,
+      "loss": 4.1805,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326850821200462e-05,
+      "loss": 4.1941,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32601222644941e-05,
+      "loss": 4.1922,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325175269578731e-05,
+      "loss": 4.1748,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324336674827679e-05,
+      "loss": 4.18,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323498080076627e-05,
+      "loss": 4.1884,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322659485325575e-05,
+      "loss": 4.1706,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3218225284548956e-05,
+      "loss": 4.1772,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3209839337038436e-05,
+      "loss": 4.184,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201453389527916e-05,
+      "loss": 4.1837,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193067442017396e-05,
+      "loss": 4.165,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3184681494506876e-05,
+      "loss": 4.172,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317631192580009e-05,
+      "loss": 4.1657,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316792597828957e-05,
+      "loss": 4.1912,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315954003077905e-05,
+      "loss": 4.181,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315115408326853e-05,
+      "loss": 4.18,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314276813575801e-05,
+      "loss": 4.1849,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313439856705122e-05,
+      "loss": 4.1757,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312602899834443e-05,
+      "loss": 4.1951,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311764305083391e-05,
+      "loss": 4.1662,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310925710332339e-05,
+      "loss": 4.1823,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310087115581287e-05,
+      "loss": 4.1803,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309248520830235e-05,
+      "loss": 4.1776,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308409926079183e-05,
+      "loss": 4.182,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3075713313281316e-05,
+      "loss": 4.1744,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3067327365770796e-05,
+      "loss": 4.1783,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3058941418260276e-05,
+      "loss": 4.1833,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050571849553485e-05,
+      "loss": 4.1647,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042185902042965e-05,
+      "loss": 4.1828,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3033799954532445e-05,
+      "loss": 4.1687,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025430385825654e-05,
+      "loss": 4.1831,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3017044438315134e-05,
+      "loss": 4.1713,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008658490804614e-05,
+      "loss": 4.1661,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000272543294094e-05,
+      "loss": 4.173,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991886595783574e-05,
+      "loss": 4.1741,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983500648273054e-05,
+      "loss": 4.1742,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975114700762534e-05,
+      "loss": 4.1605,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2966728753252014e-05,
+      "loss": 4.1764,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295835918454523e-05,
+      "loss": 4.1847,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294997323703471e-05,
+      "loss": 4.1699,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294158728952418e-05,
+      "loss": 4.1553,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293320134201366e-05,
+      "loss": 4.1688,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292483177330688e-05,
+      "loss": 4.1712,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291644582579636e-05,
+      "loss": 4.1743,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290805987828583e-05,
+      "loss": 4.1751,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289967393077531e-05,
+      "loss": 4.1671,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289128798326479e-05,
+      "loss": 4.1627,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288290203575427e-05,
+      "loss": 4.1742,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287453246704749e-05,
+      "loss": 4.1506,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286614651953697e-05,
+      "loss": 4.1662,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285776057202645e-05,
+      "loss": 4.1647,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284937462451593e-05,
+      "loss": 4.1716,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284098867700541e-05,
+      "loss": 4.1516,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283260272949489e-05,
+      "loss": 4.1636,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282421678198437e-05,
+      "loss": 4.1634,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281583083447385e-05,
+      "loss": 4.161,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807461265767056e-05,
+      "loss": 4.1641,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799075318256536e-05,
+      "loss": 4.1534,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2790689370746016e-05,
+      "loss": 4.172,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782319802039225e-05,
+      "loss": 4.1683,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2773933854528705e-05,
+      "loss": 4.1572,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276554790701819e-05,
+      "loss": 4.1715,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275716195950767e-05,
+      "loss": 4.1564,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274877601199715e-05,
+      "loss": 4.1716,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274039006448663e-05,
+      "loss": 4.1566,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273200411697611e-05,
+      "loss": 4.1695,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272361816946559e-05,
+      "loss": 4.154,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27152486007588e-05,
+      "loss": 4.1647,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270687903205201e-05,
+      "loss": 4.1573,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269849308454149e-05,
+      "loss": 4.1671,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269010713703097e-05,
+      "loss": 4.1541,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268172118952045e-05,
+      "loss": 4.1612,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267333524200993e-05,
+      "loss": 4.1405,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266494929449941e-05,
+      "loss": 4.1775,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265656334698889e-05,
+      "loss": 4.1565,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264817739947837e-05,
+      "loss": 4.164,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2639807830771585e-05,
+      "loss": 4.149,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631421883261065e-05,
+      "loss": 4.1679,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2623035935750545e-05,
+      "loss": 4.1687,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261464998824002e-05,
+      "loss": 4.1563,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26062640407295e-05,
+      "loss": 4.1567,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597894472022714e-05,
+      "loss": 4.1635,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258950852451219e-05,
+      "loss": 4.1647,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258112257700167e-05,
+      "loss": 4.173,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257273662949115e-05,
+      "loss": 4.1437,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256435068198063e-05,
+      "loss": 4.1655,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255596473447011e-05,
+      "loss": 4.1523,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2547578786959594e-05,
+      "loss": 4.1629,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539192839449074e-05,
+      "loss": 4.1598,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253082327074228e-05,
+      "loss": 4.1461,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252243732323176e-05,
+      "loss": 4.1657,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251405137572124e-05,
+      "loss": 4.1575,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250568180701445e-05,
+      "loss": 4.1606,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.2099761962890625,
+      "eval_runtime": 300.7093,
+      "eval_samples_per_second": 1268.97,
+      "eval_steps_per_second": 39.656,
+      "step": 457920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.249729585950393e-05,
+      "loss": 4.1547,
+      "step": 458240
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.248890991199341e-05,
+      "loss": 4.1545,
+      "step": 458752
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.248052396448289e-05,
+      "loss": 4.165,
+      "step": 459264
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.24721543957761e-05,
+      "loss": 4.1579,
+      "step": 459776
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.246376844826558e-05,
+      "loss": 4.1607,
+      "step": 460288
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.245538250075506e-05,
+      "loss": 4.1528,
+      "step": 460800
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.244699655324455e-05,
+      "loss": 4.1464,
+      "step": 461312
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.243861060573403e-05,
+      "loss": 4.1466,
+      "step": 461824
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.243022465822351e-05,
+      "loss": 4.1561,
+      "step": 462336
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.242183871071299e-05,
+      "loss": 4.1603,
+      "step": 462848
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.241345276320247e-05,
+      "loss": 4.1534,
+      "step": 463360
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.240508319449568e-05,
+      "loss": 4.1592,
+      "step": 463872
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2396713625788886e-05,
+      "loss": 4.1449,
+      "step": 464384
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2388327678278366e-05,
+      "loss": 4.1509,
+      "step": 464896
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2379941730767846e-05,
+      "loss": 4.1495,
+      "step": 465408
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2371555783257326e-05,
+      "loss": 4.1373,
+      "step": 465920
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2363169835746806e-05,
+      "loss": 4.1524,
+      "step": 466432
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2354783888236286e-05,
+      "loss": 4.1371,
+      "step": 466944
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2346397940725765e-05,
+      "loss": 4.1446,
+      "step": 467456
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.233802837201898e-05,
+      "loss": 4.1805,
+      "step": 467968
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.232964242450846e-05,
+      "loss": 4.1443,
+      "step": 468480
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.232125647699794e-05,
+      "loss": 4.1508,
+      "step": 468992
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.231287052948742e-05,
+      "loss": 4.1525,
+      "step": 469504
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.23044845819769e-05,
+      "loss": 4.1598,
+      "step": 470016
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2296098634466374e-05,
+      "loss": 4.1364,
+      "step": 470528
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2287712686955854e-05,
+      "loss": 4.1449,
+      "step": 471040
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2279326739445334e-05,
+      "loss": 4.1476,
+      "step": 471552
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2270940791934814e-05,
+      "loss": 4.139,
+      "step": 472064
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2262554844424294e-05,
+      "loss": 4.1447,
+      "step": 472576
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.2254168896913774e-05,
+      "loss": 4.1379,
+      "step": 473088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.2245782949403254e-05,
+      "loss": 4.1455,
+      "step": 473600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.223741338069647e-05,
+      "loss": 4.1518,
+      "step": 474112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.222902743318595e-05,
+      "loss": 4.1495,
+      "step": 474624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.222064148567543e-05,
+      "loss": 4.1425,
+      "step": 475136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.221227191696864e-05,
+      "loss": 4.1482,
+      "step": 475648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.220388596945812e-05,
+      "loss": 4.154,
+      "step": 476160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.219551640075133e-05,
+      "loss": 4.1384,
+      "step": 476672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.218713045324081e-05,
+      "loss": 4.1565,
+      "step": 477184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.217874450573029e-05,
+      "loss": 4.1297,
+      "step": 477696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.217035855821977e-05,
+      "loss": 4.151,
+      "step": 478208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.216197261070925e-05,
+      "loss": 4.1278,
+      "step": 478720
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.215358666319873e-05,
+      "loss": 4.1583,
+      "step": 479232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.214520071568821e-05,
+      "loss": 4.1427,
+      "step": 479744
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.213681476817769e-05,
+      "loss": 4.1473,
+      "step": 480256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.212842882066717e-05,
+      "loss": 4.1445,
+      "step": 480768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.212004287315665e-05,
+      "loss": 4.1453,
+      "step": 481280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.2111656925646134e-05,
+      "loss": 4.1478,
+      "step": 481792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.2103270978135614e-05,
+      "loss": 4.1384,
+      "step": 482304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.2094901409428823e-05,
+      "loss": 4.1214,
+      "step": 482816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.20865154619183e-05,
+      "loss": 4.1566,
+      "step": 483328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.207812951440778e-05,
+      "loss": 4.1413,
+      "step": 483840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.206974356689726e-05,
+      "loss": 4.1519,
+      "step": 484352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.206137399819047e-05,
+      "loss": 4.1315,
+      "step": 484864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.205298805067995e-05,
+      "loss": 4.1403,
+      "step": 485376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.204460210316943e-05,
+      "loss": 4.1316,
+      "step": 485888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.203621615565891e-05,
+      "loss": 4.1465,
+      "step": 486400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.2027830208148385e-05,
+      "loss": 4.1407,
+      "step": 486912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.201944426063787e-05,
+      "loss": 4.1492,
+      "step": 487424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.201107469193109e-05,
+      "loss": 4.1465,
+      "step": 487936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.200268874442057e-05,
+      "loss": 4.1293,
+      "step": 488448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.199430279691004e-05,
+      "loss": 4.1339,
+      "step": 488960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.198591684939952e-05,
+      "loss": 4.1505,
+      "step": 489472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.1977530901889e-05,
+      "loss": 4.1258,
+      "step": 489984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.196914495437848e-05,
+      "loss": 4.1333,
+      "step": 490496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.196075900686796e-05,
+      "loss": 4.1373,
+      "step": 491008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.195237305935744e-05,
+      "loss": 4.1463,
+      "step": 491520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.1944019869454386e-05,
+      "loss": 4.1203,
+      "step": 492032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.193563392194386e-05,
+      "loss": 4.1284,
+      "step": 492544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.192724797443334e-05,
+      "loss": 4.1277,
+      "step": 493056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.1918862026922826e-05,
+      "loss": 4.1361,
+      "step": 493568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.1910476079412306e-05,
+      "loss": 4.1413,
+      "step": 494080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.1902090131901786e-05,
+      "loss": 4.1373,
+      "step": 494592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.1893704184391266e-05,
+      "loss": 4.137,
+      "step": 495104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.1885318236880746e-05,
+      "loss": 4.1302,
+      "step": 495616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.1876948668173955e-05,
+      "loss": 4.1539,
+      "step": 496128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.1868579099467164e-05,
+      "loss": 4.1225,
+      "step": 496640
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.1860193151956644e-05,
+      "loss": 4.1372,
+      "step": 497152
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.185182358324986e-05,
+      "loss": 4.1379,
+      "step": 497664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.184343763573933e-05,
+      "loss": 4.1331,
+      "step": 498176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.183505168822881e-05,
+      "loss": 4.1388,
+      "step": 498688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.182666574071829e-05,
+      "loss": 4.1354,
+      "step": 499200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.181827979320778e-05,
+      "loss": 4.1299,
+      "step": 499712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.180989384569726e-05,
+      "loss": 4.1427,
+      "step": 500224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.180150789818674e-05,
+      "loss": 4.1229,
+      "step": 500736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.179312195067622e-05,
+      "loss": 4.1363,
+      "step": 501248
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.17847360031657e-05,
+      "loss": 4.13,
+      "step": 501760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.177635005565518e-05,
+      "loss": 4.1402,
+      "step": 502272
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.176796410814466e-05,
+      "loss": 4.1274,
+      "step": 502784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.175957816063414e-05,
+      "loss": 4.1228,
+      "step": 503296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.175120859192735e-05,
+      "loss": 4.1294,
+      "step": 503808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.174282264441683e-05,
+      "loss": 4.1335,
+      "step": 504320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.173445307571004e-05,
+      "loss": 4.1331,
+      "step": 504832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.172606712819952e-05,
+      "loss": 4.118,
+      "step": 505344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1717681180689e-05,
+      "loss": 4.1332,
+      "step": 505856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.170929523317848e-05,
+      "loss": 4.1402,
+      "step": 506368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1700909285667964e-05,
+      "loss": 4.1302,
+      "step": 506880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1692523338157444e-05,
+      "loss": 4.1142,
+      "step": 507392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.168415376945065e-05,
+      "loss": 4.1265,
+      "step": 507904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.167576782194013e-05,
+      "loss": 4.1243,
+      "step": 508416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.166738187442961e-05,
+      "loss": 4.1358,
+      "step": 508928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.165899592691909e-05,
+      "loss": 4.135,
+      "step": 509440
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.165060997940857e-05,
+      "loss": 4.1215,
+      "step": 509952
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1642224031898046e-05,
+      "loss": 4.1212,
+      "step": 510464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1633838084387526e-05,
+      "loss": 4.1325,
+      "step": 510976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1625452136877006e-05,
+      "loss": 4.1102,
+      "step": 511488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.161708256817022e-05,
+      "loss": 4.1274,
+      "step": 512000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.160871299946343e-05,
+      "loss": 4.1182,
+      "step": 512512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.160032705195292e-05,
+      "loss": 4.1349,
+      "step": 513024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.159195748324613e-05,
+      "loss": 4.1094,
+      "step": 513536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.158357153573561e-05,
+      "loss": 4.1213,
+      "step": 514048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1575185588225087e-05,
+      "loss": 4.1201,
+      "step": 514560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1566799640714567e-05,
+      "loss": 4.1267,
+      "step": 515072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1558413693204047e-05,
+      "loss": 4.1193,
+      "step": 515584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.155002774569352e-05,
+      "loss": 4.1147,
+      "step": 516096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1541641798183e-05,
+      "loss": 4.1271,
+      "step": 516608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.153325585067248e-05,
+      "loss": 4.129,
+      "step": 517120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1524902660769425e-05,
+      "loss": 4.1155,
+      "step": 517632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1516516713258905e-05,
+      "loss": 4.1322,
+      "step": 518144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1508130765748385e-05,
+      "loss": 4.1158,
+      "step": 518656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1499744818237864e-05,
+      "loss": 4.1247,
+      "step": 519168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1491358870727344e-05,
+      "loss": 4.1181,
+      "step": 519680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1482972923216824e-05,
+      "loss": 4.1256,
+      "step": 520192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1474586975706304e-05,
+      "loss": 4.1143,
+      "step": 520704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1466201028195784e-05,
+      "loss": 4.1229,
+      "step": 521216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1457831459488993e-05,
+      "loss": 4.1252,
+      "step": 521728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.144944551197847e-05,
+      "loss": 4.1215,
+      "step": 522240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.144105956446795e-05,
+      "loss": 4.1129,
+      "step": 522752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.143267361695743e-05,
+      "loss": 4.1227,
+      "step": 523264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.142428766944691e-05,
+      "loss": 4.1039,
+      "step": 523776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.141591810074012e-05,
+      "loss": 4.1336,
+      "step": 524288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.14075321532296e-05,
+      "loss": 4.1194,
+      "step": 524800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.139914620571909e-05,
+      "loss": 4.1225,
+      "step": 525312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.139076025820857e-05,
+      "loss": 4.1109,
+      "step": 525824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.138237431069805e-05,
+      "loss": 4.1321,
+      "step": 526336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.137398836318753e-05,
+      "loss": 4.1228,
+      "step": 526848
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.136560241567701e-05,
+      "loss": 4.1169,
+      "step": 527360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.135721646816649e-05,
+      "loss": 4.1162,
+      "step": 527872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.134883052065597e-05,
+      "loss": 4.1236,
+      "step": 528384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.134046095194918e-05,
+      "loss": 4.1225,
+      "step": 528896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.133207500443866e-05,
+      "loss": 4.1327,
+      "step": 529408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.132370543573187e-05,
+      "loss": 4.1032,
+      "step": 529920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.131531948822135e-05,
+      "loss": 4.1238,
+      "step": 530432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.130693354071083e-05,
+      "loss": 4.1139,
+      "step": 530944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.129854759320031e-05,
+      "loss": 4.1251,
+      "step": 531456
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.129016164568979e-05,
+      "loss": 4.1168,
+      "step": 531968
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1281775698179273e-05,
+      "loss": 4.1111,
+      "step": 532480
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.1273389750668753e-05,
+      "loss": 4.1251,
+      "step": 532992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.126500380315823e-05,
+      "loss": 4.1208,
+      "step": 533504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.125663423445144e-05,
+      "loss": 4.121,
+      "step": 534016
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.180113792419434,
+      "eval_runtime": 298.6947,
+      "eval_samples_per_second": 1277.529,
+      "eval_steps_per_second": 39.924,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124824828694092e-05,
+      "loss": 4.1185,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.12398623394304e-05,
+      "loss": 4.1143,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123147639191988e-05,
+      "loss": 4.1244,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1223090444409356e-05,
+      "loss": 4.1223,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1214704496898835e-05,
+      "loss": 4.1245,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1206318549388315e-05,
+      "loss": 4.1102,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1197932601877795e-05,
+      "loss": 4.1073,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1189546654367275e-05,
+      "loss": 4.1069,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1181160706856755e-05,
+      "loss": 4.1193,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117277475934624e-05,
+      "loss": 4.1254,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116438881183572e-05,
+      "loss": 4.1096,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11560028643252e-05,
+      "loss": 4.1215,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114763329561841e-05,
+      "loss": 4.1129,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113924734810789e-05,
+      "loss": 4.1112,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113086140059737e-05,
+      "loss": 4.1087,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.112247545308685e-05,
+      "loss": 4.1035,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111408950557633e-05,
+      "loss": 4.1117,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110570355806581e-05,
+      "loss": 4.1045,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109731761055529e-05,
+      "loss": 4.1035,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108893166304477e-05,
+      "loss": 4.1378,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1080545715534244e-05,
+      "loss": 4.1125,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1072159768023724e-05,
+      "loss": 4.1116,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106377382051321e-05,
+      "loss": 4.1145,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105538787300269e-05,
+      "loss": 4.1179,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10470183042959e-05,
+      "loss": 4.1009,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1038648735589115e-05,
+      "loss": 4.1095,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1030262788078595e-05,
+      "loss": 4.1121,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102187684056807e-05,
+      "loss": 4.1062,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101349089305755e-05,
+      "loss": 4.1026,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100510494554703e-05,
+      "loss": 4.1014,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099671899803651e-05,
+      "loss": 4.1088,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098833305052599e-05,
+      "loss": 4.1162,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09799634818192e-05,
+      "loss": 4.1119,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097157753430868e-05,
+      "loss": 4.1042,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096319158679816e-05,
+      "loss": 4.1115,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0954805639287644e-05,
+      "loss": 4.1148,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946419691777124e-05,
+      "loss": 4.1074,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0938033744266604e-05,
+      "loss": 4.1138,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0929647796756084e-05,
+      "loss": 4.0968,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0921261849245564e-05,
+      "loss": 4.1115,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.091289228053877e-05,
+      "loss": 4.0931,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.090450633302825e-05,
+      "loss": 4.117,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.089613676432146e-05,
+      "loss": 4.1102,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088775081681094e-05,
+      "loss": 4.11,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087936486930042e-05,
+      "loss": 4.1115,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08709789217899e-05,
+      "loss": 4.1049,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086259297427938e-05,
+      "loss": 4.1115,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08542234055726e-05,
+      "loss": 4.1018,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084583745806208e-05,
+      "loss": 4.0807,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083745151055156e-05,
+      "loss": 4.1199,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082906556304104e-05,
+      "loss": 4.1047,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082067961553052e-05,
+      "loss": 4.1171,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081229366802e-05,
+      "loss": 4.0986,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080390772050948e-05,
+      "loss": 4.103,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079553815180269e-05,
+      "loss": 4.0926,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078715220429217e-05,
+      "loss": 4.112,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0778766256781647e-05,
+      "loss": 4.1037,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0770380309271127e-05,
+      "loss": 4.1093,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0761994361760607e-05,
+      "loss": 4.1162,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0753624793053816e-05,
+      "loss": 4.0952,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0745238845543296e-05,
+      "loss": 4.0971,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073685289803278e-05,
+      "loss": 4.1155,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0728466950522255e-05,
+      "loss": 4.092,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0720081003011735e-05,
+      "loss": 4.0938,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0711695055501215e-05,
+      "loss": 4.1037,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070332548679443e-05,
+      "loss": 4.1082,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0694939539283904e-05,
+      "loss": 4.0886,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0686553591773384e-05,
+      "loss": 4.0937,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0678167644262864e-05,
+      "loss": 4.0875,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0669781696752344e-05,
+      "loss": 4.102,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0661395749241824e-05,
+      "loss": 4.1077,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0653009801731304e-05,
+      "loss": 4.1024,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064464023302452e-05,
+      "loss": 4.0992,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0636254285514e-05,
+      "loss": 4.0997,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062786833800348e-05,
+      "loss": 4.1174,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061948239049296e-05,
+      "loss": 4.0883,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061111282178617e-05,
+      "loss": 4.1054,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060272687427565e-05,
+      "loss": 4.102,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059434092676513e-05,
+      "loss": 4.0938,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058595497925461e-05,
+      "loss": 4.1079,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.057756903174409e-05,
+      "loss": 4.0995,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056918308423357e-05,
+      "loss": 4.0973,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056079713672305e-05,
+      "loss": 4.1071,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055241118921253e-05,
+      "loss": 4.0877,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054402524170201e-05,
+      "loss": 4.0999,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.053563929419149e-05,
+      "loss": 4.0958,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05272697254847e-05,
+      "loss": 4.1054,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0518883777974184e-05,
+      "loss": 4.0917,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0510497830463664e-05,
+      "loss": 4.0933,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0502111882953144e-05,
+      "loss": 4.0891,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049372593544262e-05,
+      "loss": 4.101,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0485356366735833e-05,
+      "loss": 4.0995,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0476970419225313e-05,
+      "loss": 4.0831,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.046858447171479e-05,
+      "loss": 4.1031,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0460198524204267e-05,
+      "loss": 4.1041,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.045182895549748e-05,
+      "loss": 4.0946,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.044344300798696e-05,
+      "loss": 4.0792,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.043505706047644e-05,
+      "loss": 4.0916,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042667111296592e-05,
+      "loss": 4.0889,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04182851654554e-05,
+      "loss": 4.104,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040989921794488e-05,
+      "loss": 4.1,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040152964923809e-05,
+      "loss": 4.0896,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039314370172757e-05,
+      "loss": 4.0896,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038475775421705e-05,
+      "loss": 4.0919,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037637180670653e-05,
+      "loss": 4.0836,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036798585919601e-05,
+      "loss": 4.0906,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035959991168549e-05,
+      "loss": 4.0851,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035121396417497e-05,
+      "loss": 4.1032,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034282801666445e-05,
+      "loss": 4.0766,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033444206915393e-05,
+      "loss": 4.0874,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032607250044714e-05,
+      "loss": 4.0857,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031768655293662e-05,
+      "loss": 4.0952,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030930060542611e-05,
+      "loss": 4.0834,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0300931036719316e-05,
+      "loss": 4.0825,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0292545089208796e-05,
+      "loss": 4.0965,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0284159141698276e-05,
+      "loss": 4.0959,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0275773194187756e-05,
+      "loss": 4.0805,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0267403625480965e-05,
+      "loss": 4.0966,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0259017677970445e-05,
+      "loss": 4.0865,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0250631730459925e-05,
+      "loss": 4.0925,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242245782949405e-05,
+      "loss": 4.0869,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0233859835438885e-05,
+      "loss": 4.0908,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0225473887928365e-05,
+      "loss": 4.0837,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0217087940417845e-05,
+      "loss": 4.092,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0208701992907324e-05,
+      "loss": 4.0942,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020033242420054e-05,
+      "loss": 4.0892,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019194647669002e-05,
+      "loss": 4.0822,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018357690798323e-05,
+      "loss": 4.0879,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.017519096047271e-05,
+      "loss": 4.072,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.016680501296219e-05,
+      "loss": 4.1044,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015841906545167e-05,
+      "loss": 4.0813,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015003311794115e-05,
+      "loss": 4.0914,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014164717043063e-05,
+      "loss": 4.0778,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013327760172384e-05,
+      "loss": 4.0983,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.012489165421332e-05,
+      "loss": 4.0923,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01165057067028e-05,
+      "loss": 4.0841,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010811975919228e-05,
+      "loss": 4.0878,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009973381168176e-05,
+      "loss": 4.0869,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009134786417124e-05,
+      "loss": 4.0895,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008296191666072e-05,
+      "loss": 4.1051,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00745759691502e-05,
+      "loss": 4.0674,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006620640044341e-05,
+      "loss": 4.0965,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.005782045293289e-05,
+      "loss": 4.0816,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004943450542237e-05,
+      "loss": 4.0916,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004104855791185e-05,
+      "loss": 4.0818,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003266261040133e-05,
+      "loss": 4.085,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0024293041694536e-05,
+      "loss": 4.0915,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001592347298775e-05,
+      "loss": 4.089,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000753752547723e-05,
+      "loss": 4.0874,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.156889915466309,
+      "eval_runtime": 280.1064,
+      "eval_samples_per_second": 1362.307,
+      "eval_steps_per_second": 42.573,
+      "step": 610560
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.351587732934069e+17,
+  "trial_name": null,
+  "trial_params": null
+}