diff --git "a/checkpoint-1908000/trainer_state.json" "b/checkpoint-1908000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1908000/trainer_state.json"
@@ -0,0 +1,22581 @@
+{
+  "best_metric": 4.009130477905273,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/re-irr-sv-agr/lstm/4/checkpoints/checkpoint-1908000",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1908000,
+  "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": 10.82,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5473,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0522,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9852,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9385,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8823,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7148,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994131474623009e-05,
+      "loss": 6.6152,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993292879871958e-05,
+      "loss": 6.5142,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.4434,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.3732,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3151,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2482,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.1796,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.1237,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0564,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0163,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.975,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9339,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.8988,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8686,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8293,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.805,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7631,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7504,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7189,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7002,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6721,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6404,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6248,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5899,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5641,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5525,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.542,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 5.5196,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5083,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4655,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.462,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4422,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4265,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4222,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4158,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.3867,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3863,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3819,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3691,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3632,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3471,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958918684480691e-05,
+      "loss": 5.3251,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958080089729639e-05,
+      "loss": 5.3285,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957241494978587e-05,
+      "loss": 5.3084,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956402900227535e-05,
+      "loss": 5.3114,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955564305476483e-05,
+      "loss": 5.2796,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954725710725431e-05,
+      "loss": 5.2933,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953888753854752e-05,
+      "loss": 5.2666,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2844,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.256,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2434,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2383,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2334,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2186,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2216,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.1928,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.1905,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.2017,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1932,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1767,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1594,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1598,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413147462300904e-05,
+      "loss": 5.1606,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404761514790384e-05,
+      "loss": 5.153,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.939637556727987e-05,
+      "loss": 5.1494,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1453,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937962005106256e-05,
+      "loss": 5.145,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937123410355204e-05,
+      "loss": 5.1365,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1112,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.1063,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.104,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933770669231369e-05,
+      "loss": 5.0925,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932932074480317e-05,
+      "loss": 5.1014,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932093479729265e-05,
+      "loss": 5.0837,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0995,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0745,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.0608,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287407386054305e-05,
+      "loss": 5.0643,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279021438543785e-05,
+      "loss": 5.0646,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270635491033264e-05,
+      "loss": 5.0604,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0487,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.0434,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9245494027305433e-05,
+      "loss": 5.0357,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9237108079794913e-05,
+      "loss": 5.0342,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922872213228439e-05,
+      "loss": 5.0238,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0192,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 5.0186,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920358066855656e-05,
+      "loss": 5.0238,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919519472104604e-05,
+      "loss": 5.0035,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918680877353552e-05,
+      "loss": 4.9966,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917842282602501e-05,
+      "loss": 4.9899,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917003687851449e-05,
+      "loss": 4.988,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 4.9917,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 4.9787,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914489541478666e-05,
+      "loss": 4.9719,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913650946727614e-05,
+      "loss": 4.9635,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912812351976562e-05,
+      "loss": 4.9665,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.9394,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911136800354831e-05,
+      "loss": 4.9543,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910298205603779e-05,
+      "loss": 4.941,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909459610852727e-05,
+      "loss": 4.9473,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908621016101675e-05,
+      "loss": 4.9354,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907782421350623e-05,
+      "loss": 4.9431,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906945464479944e-05,
+      "loss": 4.9294,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906106869728892e-05,
+      "loss": 4.9222,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052682749778396e-05,
+      "loss": 4.9307,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9044296802267876e-05,
+      "loss": 4.9087,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9035910854757356e-05,
+      "loss": 4.925,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027541286050565e-05,
+      "loss": 4.9147,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.9125,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.9,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.8989,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.9007,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8932,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.892,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.881,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8836,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8642,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.8687,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.8521,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892694267353179e-05,
+      "loss": 4.8586,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.8453,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.8663,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.8714,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889341526229344e-05,
+      "loss": 4.8554,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888502931478292e-05,
+      "loss": 4.8347,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8453,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8437,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.8376,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851501903544574e-05,
+      "loss": 4.8323,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843115956034054e-05,
+      "loss": 4.8346,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834730008523534e-05,
+      "loss": 4.8301,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.8193,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.8205,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8198,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8089,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.8085,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.817,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.804,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876767518604683e-05,
+      "loss": 4.8023,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875928923853632e-05,
+      "loss": 4.8159,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87509032910258e-05,
+      "loss": 4.7916,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.78051233291626,
+      "eval_runtime": 327.7198,
+      "eval_samples_per_second": 1164.382,
+      "eval_steps_per_second": 36.388,
+      "step": 76320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874251734351528e-05,
+      "loss": 4.7776,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873413139600475e-05,
+      "loss": 4.7857,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.872574544849423e-05,
+      "loss": 4.7992,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.871735950098371e-05,
+      "loss": 4.7871,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870897355347319e-05,
+      "loss": 4.7927,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870058760596267e-05,
+      "loss": 4.768,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.869220165845215e-05,
+      "loss": 4.7849,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.868381571094163e-05,
+      "loss": 4.7561,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867542976343111e-05,
+      "loss": 4.7715,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866704381592059e-05,
+      "loss": 4.7592,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865865786841007e-05,
+      "loss": 4.7752,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865027192089955e-05,
+      "loss": 4.7705,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.864190235219277e-05,
+      "loss": 4.757,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.863351640468225e-05,
+      "loss": 4.7516,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.862513045717173e-05,
+      "loss": 4.7382,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.861674450966121e-05,
+      "loss": 4.7402,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860835856215069e-05,
+      "loss": 4.7511,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8599988993443896e-05,
+      "loss": 4.7458,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8591603045933376e-05,
+      "loss": 4.7392,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8583217098422856e-05,
+      "loss": 4.7553,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8574831150912336e-05,
+      "loss": 4.7487,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8566445203401816e-05,
+      "loss": 4.7406,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.855805925589129e-05,
+      "loss": 4.7285,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.854967330838077e-05,
+      "loss": 4.7484,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8541287360870256e-05,
+      "loss": 4.7236,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8532901413359736e-05,
+      "loss": 4.7327,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8524515465849216e-05,
+      "loss": 4.7289,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8516129518338696e-05,
+      "loss": 4.723,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8507759949631905e-05,
+      "loss": 4.7116,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8499374002121385e-05,
+      "loss": 4.7129,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8490988054610865e-05,
+      "loss": 4.7155,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8482602107100345e-05,
+      "loss": 4.7109,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8474216159589824e-05,
+      "loss": 4.7111,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8465846590883034e-05,
+      "loss": 4.7176,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8457460643372514e-05,
+      "loss": 4.7082,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8449074695861993e-05,
+      "loss": 4.707,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8440688748351473e-05,
+      "loss": 4.6852,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.843230280084095e-05,
+      "loss": 4.6927,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.842391685333043e-05,
+      "loss": 4.683,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.841554728462365e-05,
+      "loss": 4.6794,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840716133711313e-05,
+      "loss": 4.6897,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839877538960261e-05,
+      "loss": 4.692,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839038944209209e-05,
+      "loss": 4.6743,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.838200349458157e-05,
+      "loss": 4.6823,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.837361754707105e-05,
+      "loss": 4.6783,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836523159956053e-05,
+      "loss": 4.6822,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.835687840965747e-05,
+      "loss": 4.6852,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834849246214695e-05,
+      "loss": 4.676,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834010651463643e-05,
+      "loss": 4.6651,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833172056712591e-05,
+      "loss": 4.6721,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832333461961539e-05,
+      "loss": 4.6605,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8314948672104874e-05,
+      "loss": 4.6692,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8306562724594354e-05,
+      "loss": 4.6561,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.829819315588756e-05,
+      "loss": 4.6656,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828980720837704e-05,
+      "loss": 4.6456,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828142126086652e-05,
+      "loss": 4.6701,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8273035313356e-05,
+      "loss": 4.6512,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8264649365845476e-05,
+      "loss": 4.6475,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8256263418334956e-05,
+      "loss": 4.6545,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.824789384962817e-05,
+      "loss": 4.6456,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823950790211765e-05,
+      "loss": 4.6386,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8231121954607125e-05,
+      "loss": 4.6526,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.822273600709661e-05,
+      "loss": 4.6237,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.821435005958609e-05,
+      "loss": 4.6294,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.820596411207557e-05,
+      "loss": 4.6415,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.819757816456505e-05,
+      "loss": 4.641,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818919221705453e-05,
+      "loss": 4.6324,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818080626954401e-05,
+      "loss": 4.6229,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.817242032203349e-05,
+      "loss": 4.6226,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.816403437452297e-05,
+      "loss": 4.6271,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.815566480581618e-05,
+      "loss": 4.6328,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.814727885830566e-05,
+      "loss": 4.6267,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813889291079514e-05,
+      "loss": 4.627,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813052334208835e-05,
+      "loss": 4.6331,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.812213739457783e-05,
+      "loss": 4.63,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.811375144706731e-05,
+      "loss": 4.613,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8105365499556796e-05,
+      "loss": 4.6108,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8096979552046276e-05,
+      "loss": 4.6192,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8088609983339485e-05,
+      "loss": 4.6036,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8080224035828965e-05,
+      "loss": 4.616,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8071838088318445e-05,
+      "loss": 4.6031,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8063452140807925e-05,
+      "loss": 4.6259,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8055066193297405e-05,
+      "loss": 4.6062,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8046680245786885e-05,
+      "loss": 4.591,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8038294298276365e-05,
+      "loss": 4.6016,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.802990835076584e-05,
+      "loss": 4.6053,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.802152240325532e-05,
+      "loss": 4.6077,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8013152834548534e-05,
+      "loss": 4.5984,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8004766887038014e-05,
+      "loss": 4.5972,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7996380939527494e-05,
+      "loss": 4.6018,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7987994992016974e-05,
+      "loss": 4.5901,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797962542331019e-05,
+      "loss": 4.5895,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797123947579966e-05,
+      "loss": 4.5898,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796285352828914e-05,
+      "loss": 4.5948,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795446758077862e-05,
+      "loss": 4.5942,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.79460816332681e-05,
+      "loss": 4.5902,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793769568575758e-05,
+      "loss": 4.576,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792930973824706e-05,
+      "loss": 4.578,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792092379073654e-05,
+      "loss": 4.5817,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.791255422202975e-05,
+      "loss": 4.5812,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.790416827451923e-05,
+      "loss": 4.5773,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.789578232700871e-05,
+      "loss": 4.5739,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.788741275830193e-05,
+      "loss": 4.5662,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.787902681079141e-05,
+      "loss": 4.5773,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.787064086328089e-05,
+      "loss": 4.5465,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.786225491577037e-05,
+      "loss": 4.5706,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.785386896825985e-05,
+      "loss": 4.5619,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.784548302074933e-05,
+      "loss": 4.5695,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7837113452042536e-05,
+      "loss": 4.5577,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7828727504532016e-05,
+      "loss": 4.5702,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7820341557021496e-05,
+      "loss": 4.5579,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7811955609510976e-05,
+      "loss": 4.5548,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7803569662000456e-05,
+      "loss": 4.5673,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7795183714489936e-05,
+      "loss": 4.5496,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7786797766979416e-05,
+      "loss": 4.5632,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7778411819468896e-05,
+      "loss": 4.5632,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777005862956584e-05,
+      "loss": 4.557,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.776167268205532e-05,
+      "loss": 4.5481,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77532867345448e-05,
+      "loss": 4.5524,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.774490078703428e-05,
+      "loss": 4.5562,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773651483952376e-05,
+      "loss": 4.5412,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.772814527081697e-05,
+      "loss": 4.5578,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771975932330645e-05,
+      "loss": 4.5408,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771137337579593e-05,
+      "loss": 4.548,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.770298742828541e-05,
+      "loss": 4.5406,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.769461785957862e-05,
+      "loss": 4.5389,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7686231912068105e-05,
+      "loss": 4.5208,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7677845964557585e-05,
+      "loss": 4.5354,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7669460017047065e-05,
+      "loss": 4.5219,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7661090448340274e-05,
+      "loss": 4.5445,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7652704500829754e-05,
+      "loss": 4.5557,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7644318553319234e-05,
+      "loss": 4.5394,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7635932605808714e-05,
+      "loss": 4.5206,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7627546658298194e-05,
+      "loss": 4.535,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7619160710787674e-05,
+      "loss": 4.5288,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761077476327715e-05,
+      "loss": 4.532,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760240519457036e-05,
+      "loss": 4.5285,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759401924705984e-05,
+      "loss": 4.5286,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.758563329954932e-05,
+      "loss": 4.5242,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.75772473520388e-05,
+      "loss": 4.5249,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756887778333202e-05,
+      "loss": 4.5242,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.75604918358215e-05,
+      "loss": 4.5268,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755210588831097e-05,
+      "loss": 4.5184,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.754371994080045e-05,
+      "loss": 4.5138,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.753535037209367e-05,
+      "loss": 4.528,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.752696442458315e-05,
+      "loss": 4.5219,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751857847707262e-05,
+      "loss": 4.5113,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751020890836584e-05,
+      "loss": 4.5309,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.750182296085532e-05,
+      "loss": 4.5106,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.504174709320068,
+      "eval_runtime": 293.4935,
+      "eval_samples_per_second": 1300.168,
+      "eval_steps_per_second": 40.631,
+      "step": 152640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.74934370133448e-05,
+      "loss": 4.5019,
+      "step": 153088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.748505106583428e-05,
+      "loss": 4.504,
+      "step": 153600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.747666511832376e-05,
+      "loss": 4.5213,
+      "step": 154112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.746829554961697e-05,
+      "loss": 4.5127,
+      "step": 154624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7459909602106446e-05,
+      "loss": 4.5167,
+      "step": 155136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7451523654595926e-05,
+      "loss": 4.4965,
+      "step": 155648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7443137707085406e-05,
+      "loss": 4.515,
+      "step": 156160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7434751759574886e-05,
+      "loss": 4.4864,
+      "step": 156672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7426365812064366e-05,
+      "loss": 4.5106,
+      "step": 157184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7417996243357575e-05,
+      "loss": 4.494,
+      "step": 157696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7409610295847055e-05,
+      "loss": 4.5109,
+      "step": 158208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7401224348336535e-05,
+      "loss": 4.5093,
+      "step": 158720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7392838400826015e-05,
+      "loss": 4.4959,
+      "step": 159232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7384452453315495e-05,
+      "loss": 4.4926,
+      "step": 159744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7376066505804975e-05,
+      "loss": 4.4861,
+      "step": 160256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.736768055829446e-05,
+      "loss": 4.4848,
+      "step": 160768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.735929461078394e-05,
+      "loss": 4.4913,
+      "step": 161280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.735090866327342e-05,
+      "loss": 4.4923,
+      "step": 161792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.734253909456663e-05,
+      "loss": 4.4909,
+      "step": 162304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.733415314705611e-05,
+      "loss": 4.502,
+      "step": 162816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.732576719954559e-05,
+      "loss": 4.5011,
+      "step": 163328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.731738125203507e-05,
+      "loss": 4.4945,
+      "step": 163840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.730899530452455e-05,
+      "loss": 4.4859,
+      "step": 164352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.730062573581776e-05,
+      "loss": 4.4972,
+      "step": 164864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.729223978830724e-05,
+      "loss": 4.4831,
+      "step": 165376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.728385384079672e-05,
+      "loss": 4.4908,
+      "step": 165888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.72754678932862e-05,
+      "loss": 4.4874,
+      "step": 166400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.726708194577568e-05,
+      "loss": 4.4826,
+      "step": 166912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.725869599826516e-05,
+      "loss": 4.4736,
+      "step": 167424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.725031005075464e-05,
+      "loss": 4.4761,
+      "step": 167936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7241940482047855e-05,
+      "loss": 4.4784,
+      "step": 168448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7233554534537335e-05,
+      "loss": 4.4804,
+      "step": 168960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.722516858702681e-05,
+      "loss": 4.4839,
+      "step": 169472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.721678263951629e-05,
+      "loss": 4.4796,
+      "step": 169984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720839669200577e-05,
+      "loss": 4.4771,
+      "step": 170496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7200027123298984e-05,
+      "loss": 4.4758,
+      "step": 171008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.719164117578846e-05,
+      "loss": 4.4578,
+      "step": 171520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.718325522827794e-05,
+      "loss": 4.4687,
+      "step": 172032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.717486928076742e-05,
+      "loss": 4.4602,
+      "step": 172544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.71664833332569e-05,
+      "loss": 4.4589,
+      "step": 173056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7158097385746384e-05,
+      "loss": 4.464,
+      "step": 173568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.714972781703959e-05,
+      "loss": 4.4711,
+      "step": 174080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.714134186952907e-05,
+      "loss": 4.4581,
+      "step": 174592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.713295592201855e-05,
+      "loss": 4.4585,
+      "step": 175104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.712456997450803e-05,
+      "loss": 4.4662,
+      "step": 175616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.711618402699751e-05,
+      "loss": 4.46,
+      "step": 176128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.710781445829072e-05,
+      "loss": 4.4736,
+      "step": 176640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.70994285107802e-05,
+      "loss": 4.4581,
+      "step": 177152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.709104256326968e-05,
+      "loss": 4.4478,
+      "step": 177664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.708265661575916e-05,
+      "loss": 4.4604,
+      "step": 178176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.707427066824864e-05,
+      "loss": 4.4505,
+      "step": 178688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.706588472073812e-05,
+      "loss": 4.4595,
+      "step": 179200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.70574987732276e-05,
+      "loss": 4.4471,
+      "step": 179712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.704911282571708e-05,
+      "loss": 4.4516,
+      "step": 180224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.70407432570103e-05,
+      "loss": 4.4429,
+      "step": 180736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.703235730949978e-05,
+      "loss": 4.4565,
+      "step": 181248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.702397136198926e-05,
+      "loss": 4.4457,
+      "step": 181760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.701558541447874e-05,
+      "loss": 4.4454,
+      "step": 182272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.700719946696822e-05,
+      "loss": 4.4522,
+      "step": 182784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.69988135194577e-05,
+      "loss": 4.4395,
+      "step": 183296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6990443950750906e-05,
+      "loss": 4.4345,
+      "step": 183808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6982058003240386e-05,
+      "loss": 4.4566,
+      "step": 184320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6973672055729866e-05,
+      "loss": 4.4232,
+      "step": 184832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6965302487023075e-05,
+      "loss": 4.4286,
+      "step": 185344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6956916539512555e-05,
+      "loss": 4.4376,
+      "step": 185856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6948530592002035e-05,
+      "loss": 4.449,
+      "step": 186368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6940144644491515e-05,
+      "loss": 4.4353,
+      "step": 186880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6931758696980995e-05,
+      "loss": 4.4264,
+      "step": 187392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6923372749470475e-05,
+      "loss": 4.4313,
+      "step": 187904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6914986801959955e-05,
+      "loss": 4.4303,
+      "step": 188416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6906600854449435e-05,
+      "loss": 4.4409,
+      "step": 188928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6898231285742644e-05,
+      "loss": 4.4369,
+      "step": 189440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.688986171703586e-05,
+      "loss": 4.4348,
+      "step": 189952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.688147576952534e-05,
+      "loss": 4.4396,
+      "step": 190464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.687308982201481e-05,
+      "loss": 4.442,
+      "step": 190976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.686470387450429e-05,
+      "loss": 4.4226,
+      "step": 191488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.685631792699377e-05,
+      "loss": 4.4227,
+      "step": 192000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.684793197948325e-05,
+      "loss": 4.4332,
+      "step": 192512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.683954603197274e-05,
+      "loss": 4.4152,
+      "step": 193024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.683116008446222e-05,
+      "loss": 4.4318,
+      "step": 193536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.68227741369517e-05,
+      "loss": 4.4173,
+      "step": 194048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.681438818944118e-05,
+      "loss": 4.4375,
+      "step": 194560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.680600224193066e-05,
+      "loss": 4.4238,
+      "step": 195072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.679763267322387e-05,
+      "loss": 4.4116,
+      "step": 195584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.678924672571335e-05,
+      "loss": 4.416,
+      "step": 196096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.678086077820283e-05,
+      "loss": 4.4232,
+      "step": 196608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.677247483069231e-05,
+      "loss": 4.431,
+      "step": 197120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.676408888318179e-05,
+      "loss": 4.4172,
+      "step": 197632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6755719314475e-05,
+      "loss": 4.4225,
+      "step": 198144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.674733336696448e-05,
+      "loss": 4.4125,
+      "step": 198656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.673894741945396e-05,
+      "loss": 4.4215,
+      "step": 199168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.673057785074717e-05,
+      "loss": 4.4132,
+      "step": 199680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.672219190323665e-05,
+      "loss": 4.4096,
+      "step": 200192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.671380595572613e-05,
+      "loss": 4.4203,
+      "step": 200704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.670542000821561e-05,
+      "loss": 4.4184,
+      "step": 201216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.669703406070509e-05,
+      "loss": 4.4177,
+      "step": 201728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668864811319457e-05,
+      "loss": 4.4027,
+      "step": 202240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668026216568405e-05,
+      "loss": 4.4058,
+      "step": 202752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.667187621817353e-05,
+      "loss": 4.4103,
+      "step": 203264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6663490270663006e-05,
+      "loss": 4.4091,
+      "step": 203776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6655104323152486e-05,
+      "loss": 4.4089,
+      "step": 204288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.66467347544457e-05,
+      "loss": 4.4003,
+      "step": 204800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6638348806935175e-05,
+      "loss": 4.3996,
+      "step": 205312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.662996285942466e-05,
+      "loss": 4.4099,
+      "step": 205824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.662157691191414e-05,
+      "loss": 4.3817,
+      "step": 206336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.661320734320736e-05,
+      "loss": 4.4004,
+      "step": 206848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.660482139569683e-05,
+      "loss": 4.3916,
+      "step": 207360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.659643544818631e-05,
+      "loss": 4.4071,
+      "step": 207872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.658804950067579e-05,
+      "loss": 4.3957,
+      "step": 208384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6579679931969006e-05,
+      "loss": 4.4037,
+      "step": 208896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.657129398445848e-05,
+      "loss": 4.3946,
+      "step": 209408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.656290803694796e-05,
+      "loss": 4.3927,
+      "step": 209920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.655452208943744e-05,
+      "loss": 4.4037,
+      "step": 210432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.654613614192692e-05,
+      "loss": 4.3914,
+      "step": 210944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.65377501944164e-05,
+      "loss": 4.404,
+      "step": 211456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652936424690588e-05,
+      "loss": 4.3992,
+      "step": 211968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652097829939536e-05,
+      "loss": 4.3985,
+      "step": 212480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6512608730688575e-05,
+      "loss": 4.39,
+      "step": 212992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6504222783178055e-05,
+      "loss": 4.3856,
+      "step": 213504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6495836835667535e-05,
+      "loss": 4.4009,
+      "step": 214016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6487450888157015e-05,
+      "loss": 4.3843,
+      "step": 214528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6479064940646495e-05,
+      "loss": 4.3996,
+      "step": 215040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6470695371939704e-05,
+      "loss": 4.3843,
+      "step": 215552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6462309424429184e-05,
+      "loss": 4.3894,
+      "step": 216064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6453923476918664e-05,
+      "loss": 4.389,
+      "step": 216576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6445537529408144e-05,
+      "loss": 4.3852,
+      "step": 217088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6437151581897624e-05,
+      "loss": 4.3666,
+      "step": 217600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6428765634387104e-05,
+      "loss": 4.3846,
+      "step": 218112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6420379686876584e-05,
+      "loss": 4.3722,
+      "step": 218624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.64120101181698e-05,
+      "loss": 4.3871,
+      "step": 219136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.640362417065928e-05,
+      "loss": 4.4007,
+      "step": 219648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.639523822314876e-05,
+      "loss": 4.3949,
+      "step": 220160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.638685227563824e-05,
+      "loss": 4.3641,
+      "step": 220672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637848270693145e-05,
+      "loss": 4.3882,
+      "step": 221184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637009675942093e-05,
+      "loss": 4.3827,
+      "step": 221696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.636171081191041e-05,
+      "loss": 4.3809,
+      "step": 222208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.635332486439989e-05,
+      "loss": 4.3706,
+      "step": 222720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.634493891688936e-05,
+      "loss": 4.3836,
+      "step": 223232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.633655296937884e-05,
+      "loss": 4.3772,
+      "step": 223744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.632816702186832e-05,
+      "loss": 4.3758,
+      "step": 224256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.63197810743578e-05,
+      "loss": 4.3782,
+      "step": 224768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.631141150565102e-05,
+      "loss": 4.3797,
+      "step": 225280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.63030255581405e-05,
+      "loss": 4.3755,
+      "step": 225792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.629463961062998e-05,
+      "loss": 4.3728,
+      "step": 226304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.628625366311946e-05,
+      "loss": 4.3783,
+      "step": 226816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.627786771560894e-05,
+      "loss": 4.3778,
+      "step": 227328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.626948176809842e-05,
+      "loss": 4.3668,
+      "step": 227840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.62610958205879e-05,
+      "loss": 4.3842,
+      "step": 228352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.625270987307738e-05,
+      "loss": 4.3687,
+      "step": 228864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.369020462036133,
+      "eval_runtime": 293.4838,
+      "eval_samples_per_second": 1300.211,
+      "eval_steps_per_second": 40.633,
+      "step": 228960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6244340304370586e-05,
+      "loss": 4.3576,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6235954356860066e-05,
+      "loss": 4.3609,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6227568409349546e-05,
+      "loss": 4.3805,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6219182461839026e-05,
+      "loss": 4.3682,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6210796514328506e-05,
+      "loss": 4.3807,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6202410566817986e-05,
+      "loss": 4.3553,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6194024619307466e-05,
+      "loss": 4.3746,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6185638671796946e-05,
+      "loss": 4.3446,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.617725272428643e-05,
+      "loss": 4.3762,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6168866776775906e-05,
+      "loss": 4.3568,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6160480829265386e-05,
+      "loss": 4.3721,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6152094881754866e-05,
+      "loss": 4.3707,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614372531304808e-05,
+      "loss": 4.3542,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6135339365537555e-05,
+      "loss": 4.3602,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6126953418027035e-05,
+      "loss": 4.3531,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6118567470516515e-05,
+      "loss": 4.3465,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6110181523005995e-05,
+      "loss": 4.3617,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6101795575495475e-05,
+      "loss": 4.3574,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6093409627984955e-05,
+      "loss": 4.3582,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085023680474435e-05,
+      "loss": 4.366,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.607665411176765e-05,
+      "loss": 4.3698,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606826816425713e-05,
+      "loss": 4.3607,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605988221674661e-05,
+      "loss": 4.3538,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605151264803982e-05,
+      "loss": 4.3653,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.60431267005293e-05,
+      "loss": 4.3514,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.603474075301878e-05,
+      "loss": 4.3636,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.602635480550826e-05,
+      "loss": 4.3564,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601796885799774e-05,
+      "loss": 4.349,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600958291048722e-05,
+      "loss": 4.3438,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.60011969629767e-05,
+      "loss": 4.3505,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599281101546618e-05,
+      "loss": 4.3466,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598444144675939e-05,
+      "loss": 4.3545,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597605549924887e-05,
+      "loss": 4.3544,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596766955173835e-05,
+      "loss": 4.3518,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5959283604227835e-05,
+      "loss": 4.3519,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5950914035521044e-05,
+      "loss": 4.3489,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5942528088010524e-05,
+      "loss": 4.3332,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5934142140500004e-05,
+      "loss": 4.3418,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5925756192989484e-05,
+      "loss": 4.3342,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5917370245478964e-05,
+      "loss": 4.3396,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590898429796844e-05,
+      "loss": 4.3418,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590059835045792e-05,
+      "loss": 4.3453,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589222878175113e-05,
+      "loss": 4.3412,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.588384283424061e-05,
+      "loss": 4.3325,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.587545688673009e-05,
+      "loss": 4.345,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.586707093921957e-05,
+      "loss": 4.3431,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585868499170905e-05,
+      "loss": 4.3468,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585029904419853e-05,
+      "loss": 4.3382,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584192947549174e-05,
+      "loss": 4.326,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583354352798122e-05,
+      "loss": 4.3409,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582517395927444e-05,
+      "loss": 4.3303,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581678801176391e-05,
+      "loss": 4.3389,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580840206425339e-05,
+      "loss": 4.3287,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580001611674287e-05,
+      "loss": 4.3275,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.579163016923235e-05,
+      "loss": 4.3279,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.578324422172183e-05,
+      "loss": 4.3339,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577485827421131e-05,
+      "loss": 4.3302,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.576647232670079e-05,
+      "loss": 4.3315,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.575808637919027e-05,
+      "loss": 4.3302,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574970043167976e-05,
+      "loss": 4.322,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5741330862972966e-05,
+      "loss": 4.321,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5732944915462446e-05,
+      "loss": 4.3418,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5724558967951926e-05,
+      "loss": 4.3103,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5716173020441406e-05,
+      "loss": 4.3121,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5707787072930886e-05,
+      "loss": 4.3189,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5699401125420366e-05,
+      "loss": 4.3357,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5691015177909846e-05,
+      "loss": 4.3204,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5682629230399326e-05,
+      "loss": 4.3188,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5674259661692535e-05,
+      "loss": 4.3111,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5665873714182015e-05,
+      "loss": 4.3174,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.565752052427896e-05,
+      "loss": 4.3283,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564913457676844e-05,
+      "loss": 4.3217,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564074862925792e-05,
+      "loss": 4.3201,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56323626817474e-05,
+      "loss": 4.3314,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.562397673423688e-05,
+      "loss": 4.3293,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561559078672636e-05,
+      "loss": 4.3104,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560720483921584e-05,
+      "loss": 4.3133,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559881889170532e-05,
+      "loss": 4.3262,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55904329441948e-05,
+      "loss": 4.3008,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.558204699668427e-05,
+      "loss": 4.3211,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557367742797749e-05,
+      "loss": 4.3068,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556529148046697e-05,
+      "loss": 4.3291,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.555690553295645e-05,
+      "loss": 4.3139,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554851958544593e-05,
+      "loss": 4.3049,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554013363793541e-05,
+      "loss": 4.3034,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.553174769042489e-05,
+      "loss": 4.3183,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.552336174291437e-05,
+      "loss": 4.3232,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.551499217420758e-05,
+      "loss": 4.3123,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.550660622669706e-05,
+      "loss": 4.3112,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549822027918654e-05,
+      "loss": 4.3046,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548983433167602e-05,
+      "loss": 4.3163,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5481464762969226e-05,
+      "loss": 4.3062,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.547309519426244e-05,
+      "loss": 4.3058,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546470924675192e-05,
+      "loss": 4.3148,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54563232992414e-05,
+      "loss": 4.3134,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544793735173088e-05,
+      "loss": 4.3097,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543955140422036e-05,
+      "loss": 4.2963,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543116545670984e-05,
+      "loss": 4.3048,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542277950919932e-05,
+      "loss": 4.3037,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54143935616888e-05,
+      "loss": 4.3065,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.540602399298201e-05,
+      "loss": 4.3049,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539763804547149e-05,
+      "loss": 4.2962,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538925209796097e-05,
+      "loss": 4.3005,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538086615045045e-05,
+      "loss": 4.3052,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.537248020293993e-05,
+      "loss": 4.2811,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.536409425542941e-05,
+      "loss": 4.3005,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535572468672262e-05,
+      "loss": 4.2857,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.53473387392121e-05,
+      "loss": 4.3038,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533895279170158e-05,
+      "loss": 4.2959,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533056684419107e-05,
+      "loss": 4.2995,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5322197275484276e-05,
+      "loss": 4.2916,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5313811327973756e-05,
+      "loss": 4.2954,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5305425380463236e-05,
+      "loss": 4.3044,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5297039432952716e-05,
+      "loss": 4.288,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5288669864245925e-05,
+      "loss": 4.3001,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5280283916735405e-05,
+      "loss": 4.3,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5271914348028614e-05,
+      "loss": 4.2982,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5263528400518094e-05,
+      "loss": 4.2998,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5255142453007574e-05,
+      "loss": 4.283,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5246756505497054e-05,
+      "loss": 4.3028,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5238370557986534e-05,
+      "loss": 4.2891,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.522998461047602e-05,
+      "loss": 4.2976,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.52215986629655e-05,
+      "loss": 4.2863,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.521321271545498e-05,
+      "loss": 4.2921,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.520484314674819e-05,
+      "loss": 4.2905,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.519645719923767e-05,
+      "loss": 4.2857,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.518807125172715e-05,
+      "loss": 4.2732,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517968530421663e-05,
+      "loss": 4.2849,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517129935670611e-05,
+      "loss": 4.2796,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516291340919558e-05,
+      "loss": 4.2903,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515452746168506e-05,
+      "loss": 4.3007,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514615789297828e-05,
+      "loss": 4.2982,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513777194546776e-05,
+      "loss": 4.27,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512938599795724e-05,
+      "loss": 4.2952,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512100005044672e-05,
+      "loss": 4.286,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51126141029362e-05,
+      "loss": 4.2854,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.510422815542568e-05,
+      "loss": 4.2795,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509584220791516e-05,
+      "loss": 4.2886,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508745626040464e-05,
+      "loss": 4.2796,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507908669169785e-05,
+      "loss": 4.2879,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507070074418733e-05,
+      "loss": 4.2809,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5062331175480536e-05,
+      "loss": 4.2842,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5053945227970016e-05,
+      "loss": 4.2868,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5045559280459496e-05,
+      "loss": 4.2788,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5037173332948976e-05,
+      "loss": 4.2823,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502880376424219e-05,
+      "loss": 4.2913,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502041781673167e-05,
+      "loss": 4.2731,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.501203186922115e-05,
+      "loss": 4.2894,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.500364592171063e-05,
+      "loss": 4.282,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.2855424880981445,
+      "eval_runtime": 283.2935,
+      "eval_samples_per_second": 1346.981,
+      "eval_steps_per_second": 42.094,
+      "step": 305280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.499525997420011e-05,
+      "loss": 4.2726,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.498687402668959e-05,
+      "loss": 4.2692,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.497848807917907e-05,
+      "loss": 4.2847,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.497010213166855e-05,
+      "loss": 4.2804,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.496173256296176e-05,
+      "loss": 4.2921,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.495334661545124e-05,
+      "loss": 4.2643,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.494496066794072e-05,
+      "loss": 4.2806,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.49365747204302e-05,
+      "loss": 4.2628,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492820515172341e-05,
+      "loss": 4.2798,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4919835583016625e-05,
+      "loss": 4.2703,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4911449635506105e-05,
+      "loss": 4.2787,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4903063687995585e-05,
+      "loss": 4.2839,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4894694119288794e-05,
+      "loss": 4.2656,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4886308171778274e-05,
+      "loss": 4.2744,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4877922224267754e-05,
+      "loss": 4.2645,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4869536276757234e-05,
+      "loss": 4.2586,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4861150329246714e-05,
+      "loss": 4.2711,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4852764381736194e-05,
+      "loss": 4.2696,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4844378434225674e-05,
+      "loss": 4.2711,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4835992486715154e-05,
+      "loss": 4.2792,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.482762291800836e-05,
+      "loss": 4.2847,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481923697049785e-05,
+      "loss": 4.2749,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481085102298733e-05,
+      "loss": 4.2686,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.480246507547681e-05,
+      "loss": 4.2806,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.479407912796629e-05,
+      "loss": 4.2622,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.47857095592595e-05,
+      "loss": 4.2763,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.477732361174898e-05,
+      "loss": 4.2728,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476893766423846e-05,
+      "loss": 4.2618,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476055171672794e-05,
+      "loss": 4.2617,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.475216576921742e-05,
+      "loss": 4.2626,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.474377982170689e-05,
+      "loss": 4.2645,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.473539387419637e-05,
+      "loss": 4.2695,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.472702430548959e-05,
+      "loss": 4.2699,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471863835797907e-05,
+      "loss": 4.2675,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471025241046855e-05,
+      "loss": 4.2691,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470186646295803e-05,
+      "loss": 4.2668,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.469348051544751e-05,
+      "loss": 4.2482,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.468509456793699e-05,
+      "loss": 4.2612,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.467670862042647e-05,
+      "loss": 4.2499,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4668339051719676e-05,
+      "loss": 4.2561,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4659953104209156e-05,
+      "loss": 4.2573,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4651567156698636e-05,
+      "loss": 4.2687,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4643181209188116e-05,
+      "loss": 4.2579,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4634795261677596e-05,
+      "loss": 4.2491,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4626409314167076e-05,
+      "loss": 4.2664,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4618023366656556e-05,
+      "loss": 4.2574,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4609653797949765e-05,
+      "loss": 4.2613,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.460126785043925e-05,
+      "loss": 4.2603,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.459288190292873e-05,
+      "loss": 4.2441,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.458449595541821e-05,
+      "loss": 4.262,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.457611000790769e-05,
+      "loss": 4.2479,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.45677404392009e-05,
+      "loss": 4.2595,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.455935449169038e-05,
+      "loss": 4.2504,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.455096854417986e-05,
+      "loss": 4.2477,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.454258259666934e-05,
+      "loss": 4.2499,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.453419664915882e-05,
+      "loss": 4.2547,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.452582708045203e-05,
+      "loss": 4.2496,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.451744113294151e-05,
+      "loss": 4.2518,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450905518543099e-05,
+      "loss": 4.2534,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450066923792047e-05,
+      "loss": 4.2441,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4492299669213686e-05,
+      "loss": 4.2386,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4483913721703166e-05,
+      "loss": 4.2683,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4475527774192646e-05,
+      "loss": 4.2307,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4467141826682126e-05,
+      "loss": 4.2335,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4458772257975335e-05,
+      "loss": 4.2417,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4450386310464815e-05,
+      "loss": 4.2601,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4442000362954295e-05,
+      "loss": 4.2369,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4433614415443775e-05,
+      "loss": 4.2436,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4425244846736984e-05,
+      "loss": 4.2341,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4416858899226464e-05,
+      "loss": 4.2414,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4408472951715944e-05,
+      "loss": 4.2489,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4400087004205423e-05,
+      "loss": 4.2473,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4391701056694903e-05,
+      "loss": 4.2447,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.438331510918438e-05,
+      "loss": 4.2539,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.43749455404776e-05,
+      "loss": 4.2517,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.436655959296708e-05,
+      "loss": 4.2389,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435817364545655e-05,
+      "loss": 4.2368,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434978769794603e-05,
+      "loss": 4.2515,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434140175043551e-05,
+      "loss": 4.2261,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433301580292499e-05,
+      "loss": 4.2433,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.432462985541447e-05,
+      "loss": 4.2323,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431624390790395e-05,
+      "loss": 4.2544,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.430785796039343e-05,
+      "loss": 4.2413,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429948839168664e-05,
+      "loss": 4.2327,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429110244417613e-05,
+      "loss": 4.2246,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.428271649666561e-05,
+      "loss": 4.2471,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.427433054915509e-05,
+      "loss": 4.2464,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.426594460164457e-05,
+      "loss": 4.2422,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.425755865413405e-05,
+      "loss": 4.2354,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424917270662353e-05,
+      "loss": 4.2342,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424080313791674e-05,
+      "loss": 4.2407,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.423241719040622e-05,
+      "loss": 4.2314,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.42240312428957e-05,
+      "loss": 4.2319,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.421564529538518e-05,
+      "loss": 4.2425,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.420725934787466e-05,
+      "loss": 4.2387,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4198873400364137e-05,
+      "loss": 4.2365,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.419048745285361e-05,
+      "loss": 4.2263,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4182117884146826e-05,
+      "loss": 4.2328,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4173731936636306e-05,
+      "loss": 4.2273,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.416536236792952e-05,
+      "loss": 4.2337,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4156976420419e-05,
+      "loss": 4.2336,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.414860685171221e-05,
+      "loss": 4.2301,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.414022090420169e-05,
+      "loss": 4.227,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.413183495669117e-05,
+      "loss": 4.2286,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.412344900918065e-05,
+      "loss": 4.2188,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.411506306167013e-05,
+      "loss": 4.2265,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.410667711415961e-05,
+      "loss": 4.2149,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4098291166649083e-05,
+      "loss": 4.2342,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40899215979423e-05,
+      "loss": 4.2263,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.408153565043178e-05,
+      "loss": 4.2265,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.407314970292126e-05,
+      "loss": 4.2203,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.406476375541074e-05,
+      "loss": 4.2298,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.405637780790022e-05,
+      "loss": 4.2298,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40479918603897e-05,
+      "loss": 4.2191,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403960591287918e-05,
+      "loss": 4.2301,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403121996536866e-05,
+      "loss": 4.2339,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.402285039666187e-05,
+      "loss": 4.2276,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.401446444915135e-05,
+      "loss": 4.2307,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.400607850164083e-05,
+      "loss": 4.2142,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.399770893293404e-05,
+      "loss": 4.2291,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398932298542352e-05,
+      "loss": 4.2254,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3980937037913e-05,
+      "loss": 4.2311,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3972551090402484e-05,
+      "loss": 4.211,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3964165142891964e-05,
+      "loss": 4.2274,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3955779195381444e-05,
+      "loss": 4.2249,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3947393247870924e-05,
+      "loss": 4.2151,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3939007300360404e-05,
+      "loss": 4.2075,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3930621352849884e-05,
+      "loss": 4.2138,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3922235405339364e-05,
+      "loss": 4.2103,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3913849457828843e-05,
+      "loss": 4.2245,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3905463510318323e-05,
+      "loss": 4.2289,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.389709394161153e-05,
+      "loss": 4.2333,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388872437290474e-05,
+      "loss": 4.2028,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388033842539422e-05,
+      "loss": 4.2276,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38719524778837e-05,
+      "loss": 4.2227,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.386356653037318e-05,
+      "loss": 4.2164,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.385518058286267e-05,
+      "loss": 4.2119,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.384679463535215e-05,
+      "loss": 4.2283,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383840868784163e-05,
+      "loss": 4.2149,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38300227403311e-05,
+      "loss": 4.2181,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.382165317162432e-05,
+      "loss": 4.2103,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38132672241138e-05,
+      "loss": 4.2167,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.380488127660328e-05,
+      "loss": 4.2199,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.379649532909275e-05,
+      "loss": 4.2153,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3788125760385966e-05,
+      "loss": 4.2157,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3779739812875446e-05,
+      "loss": 4.2256,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.377135386536492e-05,
+      "loss": 4.2098,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3762984296658135e-05,
+      "loss": 4.2196,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.375459834914762e-05,
+      "loss": 4.2204,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.228407382965088,
+      "eval_runtime": 283.6023,
+      "eval_samples_per_second": 1345.515,
+      "eval_steps_per_second": 42.048,
+      "step": 381600
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.37462124016371e-05,
+      "loss": 4.2063,
+      "step": 381952
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.373784283293031e-05,
+      "loss": 4.2033,
+      "step": 382464
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.372945688541979e-05,
+      "loss": 4.2203,
+      "step": 382976
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.372107093790927e-05,
+      "loss": 4.2172,
+      "step": 383488
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.371268499039875e-05,
+      "loss": 4.2226,
+      "step": 384000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.370431542169196e-05,
+      "loss": 4.2081,
+      "step": 384512
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.369592947418144e-05,
+      "loss": 4.2134,
+      "step": 385024
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.368754352667092e-05,
+      "loss": 4.1994,
+      "step": 385536
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.367915757916039e-05,
+      "loss": 4.2128,
+      "step": 386048
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.367077163164987e-05,
+      "loss": 4.21,
+      "step": 386560
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.366238568413936e-05,
+      "loss": 4.214,
+      "step": 387072
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.365399973662884e-05,
+      "loss": 4.2174,
+      "step": 387584
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.364563016792205e-05,
+      "loss": 4.2055,
+      "step": 388096
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.363724422041153e-05,
+      "loss": 4.2082,
+      "step": 388608
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.362885827290101e-05,
+      "loss": 4.2042,
+      "step": 389120
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.362047232539049e-05,
+      "loss": 4.1946,
+      "step": 389632
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.361208637787997e-05,
+      "loss": 4.2027,
+      "step": 390144
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.360370043036945e-05,
+      "loss": 4.2092,
+      "step": 390656
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.359531448285893e-05,
+      "loss": 4.2087,
+      "step": 391168
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.358692853534841e-05,
+      "loss": 4.2185,
+      "step": 391680
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.357854258783789e-05,
+      "loss": 4.2219,
+      "step": 392192
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.357015664032737e-05,
+      "loss": 4.2139,
+      "step": 392704
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.356177069281685e-05,
+      "loss": 4.2053,
+      "step": 393216
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.355338474530633e-05,
+      "loss": 4.2167,
+      "step": 393728
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.354501517659954e-05,
+      "loss": 4.2034,
+      "step": 394240
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.3536629229089024e-05,
+      "loss": 4.2128,
+      "step": 394752
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.3528243281578504e-05,
+      "loss": 4.209,
+      "step": 395264
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.3519857334067984e-05,
+      "loss": 4.1977,
+      "step": 395776
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.351148776536119e-05,
+      "loss": 4.2001,
+      "step": 396288
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.350310181785067e-05,
+      "loss": 4.2059,
+      "step": 396800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.349473224914388e-05,
+      "loss": 4.2033,
+      "step": 397312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.348634630163336e-05,
+      "loss": 4.2084,
+      "step": 397824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.347796035412284e-05,
+      "loss": 4.2108,
+      "step": 398336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.346957440661232e-05,
+      "loss": 4.2054,
+      "step": 398848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.346120483790553e-05,
+      "loss": 4.2064,
+      "step": 399360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.345281889039501e-05,
+      "loss": 4.2081,
+      "step": 399872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.344443294288449e-05,
+      "loss": 4.1939,
+      "step": 400384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.343604699537398e-05,
+      "loss": 4.1925,
+      "step": 400896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.342766104786346e-05,
+      "loss": 4.1907,
+      "step": 401408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.341927510035294e-05,
+      "loss": 4.1952,
+      "step": 401920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.341088915284241e-05,
+      "loss": 4.1972,
+      "step": 402432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.340251958413563e-05,
+      "loss": 4.2095,
+      "step": 402944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.339413363662511e-05,
+      "loss": 4.1987,
+      "step": 403456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.338574768911458e-05,
+      "loss": 4.1904,
+      "step": 403968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.337736174160406e-05,
+      "loss": 4.2089,
+      "step": 404480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.336897579409354e-05,
+      "loss": 4.1941,
+      "step": 404992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.336058984658302e-05,
+      "loss": 4.2076,
+      "step": 405504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.33522038990725e-05,
+      "loss": 4.1984,
+      "step": 406016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3343834330365716e-05,
+      "loss": 4.1856,
+      "step": 406528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3335448382855196e-05,
+      "loss": 4.2027,
+      "step": 407040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3327062435344675e-05,
+      "loss": 4.1919,
+      "step": 407552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3318676487834155e-05,
+      "loss": 4.2042,
+      "step": 408064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3310290540323635e-05,
+      "loss": 4.1869,
+      "step": 408576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3301920971616844e-05,
+      "loss": 4.1917,
+      "step": 409088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3293535024106324e-05,
+      "loss": 4.1885,
+      "step": 409600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3285149076595804e-05,
+      "loss": 4.1957,
+      "step": 410112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3276763129085284e-05,
+      "loss": 4.1948,
+      "step": 410624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3268377181574764e-05,
+      "loss": 4.1879,
+      "step": 411136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3259991234064244e-05,
+      "loss": 4.1989,
+      "step": 411648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3251605286553724e-05,
+      "loss": 4.1893,
+      "step": 412160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3243219339043204e-05,
+      "loss": 4.1788,
+      "step": 412672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.323484977033641e-05,
+      "loss": 4.2118,
+      "step": 413184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.32264638228259e-05,
+      "loss": 4.1725,
+      "step": 413696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.321807787531538e-05,
+      "loss": 4.1706,
+      "step": 414208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.320969192780486e-05,
+      "loss": 4.186,
+      "step": 414720
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.320132235909807e-05,
+      "loss": 4.206,
+      "step": 415232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.319293641158755e-05,
+      "loss": 4.177,
+      "step": 415744
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.318455046407703e-05,
+      "loss": 4.191,
+      "step": 416256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.317616451656651e-05,
+      "loss": 4.1759,
+      "step": 416768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.316777856905599e-05,
+      "loss": 4.1835,
+      "step": 417280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.315942537915293e-05,
+      "loss": 4.1935,
+      "step": 417792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.315103943164241e-05,
+      "loss": 4.1923,
+      "step": 418304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.314265348413189e-05,
+      "loss": 4.1896,
+      "step": 418816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.313426753662137e-05,
+      "loss": 4.1918,
+      "step": 419328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3125881589110854e-05,
+      "loss": 4.1981,
+      "step": 419840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.3117495641600334e-05,
+      "loss": 4.1784,
+      "step": 420352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.310912607289354e-05,
+      "loss": 4.1849,
+      "step": 420864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.310074012538302e-05,
+      "loss": 4.1936,
+      "step": 421376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.30923541778725e-05,
+      "loss": 4.1727,
+      "step": 421888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.308396823036198e-05,
+      "loss": 4.1833,
+      "step": 422400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.307559866165519e-05,
+      "loss": 4.1815,
+      "step": 422912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.306721271414467e-05,
+      "loss": 4.1943,
+      "step": 423424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.305882676663415e-05,
+      "loss": 4.1893,
+      "step": 423936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.305044081912363e-05,
+      "loss": 4.1773,
+      "step": 424448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.304205487161311e-05,
+      "loss": 4.1709,
+      "step": 424960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.303366892410259e-05,
+      "loss": 4.1885,
+      "step": 425472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.302528297659207e-05,
+      "loss": 4.1898,
+      "step": 425984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.301689702908155e-05,
+      "loss": 4.1831,
+      "step": 426496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.300852746037477e-05,
+      "loss": 4.1855,
+      "step": 427008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3000157891667976e-05,
+      "loss": 4.1808,
+      "step": 427520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2991771944157456e-05,
+      "loss": 4.1843,
+      "step": 428032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2983385996646936e-05,
+      "loss": 4.1781,
+      "step": 428544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2975000049136416e-05,
+      "loss": 4.1749,
+      "step": 429056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.296661410162589e-05,
+      "loss": 4.1908,
+      "step": 429568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.295822815411537e-05,
+      "loss": 4.1859,
+      "step": 430080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.294984220660485e-05,
+      "loss": 4.1812,
+      "step": 430592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.294145625909433e-05,
+      "loss": 4.1707,
+      "step": 431104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.293308669038754e-05,
+      "loss": 4.1815,
+      "step": 431616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2924700742877025e-05,
+      "loss": 4.1711,
+      "step": 432128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2916314795366505e-05,
+      "loss": 4.1828,
+      "step": 432640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2907928847855985e-05,
+      "loss": 4.1799,
+      "step": 433152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2899542900345465e-05,
+      "loss": 4.1768,
+      "step": 433664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2891173331638674e-05,
+      "loss": 4.175,
+      "step": 434176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2882787384128154e-05,
+      "loss": 4.1671,
+      "step": 434688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2874401436617634e-05,
+      "loss": 4.1719,
+      "step": 435200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2866015489107114e-05,
+      "loss": 4.1689,
+      "step": 435712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2857629541596594e-05,
+      "loss": 4.1671,
+      "step": 436224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2849243594086074e-05,
+      "loss": 4.1764,
+      "step": 436736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2840857646575554e-05,
+      "loss": 4.1774,
+      "step": 437248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2832471699065034e-05,
+      "loss": 4.1731,
+      "step": 437760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.282410213035824e-05,
+      "loss": 4.1645,
+      "step": 438272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.281571618284772e-05,
+      "loss": 4.1745,
+      "step": 438784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.280733023533721e-05,
+      "loss": 4.1801,
+      "step": 439296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.279894428782669e-05,
+      "loss": 4.1692,
+      "step": 439808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.279055834031617e-05,
+      "loss": 4.1793,
+      "step": 440320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.278217239280565e-05,
+      "loss": 4.1773,
+      "step": 440832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.277378644529513e-05,
+      "loss": 4.1764,
+      "step": 441344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.276541687658834e-05,
+      "loss": 4.1774,
+      "step": 441856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.275703092907782e-05,
+      "loss": 4.1654,
+      "step": 442368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.27486449815673e-05,
+      "loss": 4.1734,
+      "step": 442880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.274025903405678e-05,
+      "loss": 4.1753,
+      "step": 443392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.273188946534999e-05,
+      "loss": 4.1788,
+      "step": 443904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2723519896643196e-05,
+      "loss": 4.1619,
+      "step": 444416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2715133949132676e-05,
+      "loss": 4.1749,
+      "step": 444928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.270674800162216e-05,
+      "loss": 4.1755,
+      "step": 445440
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.269836205411164e-05,
+      "loss": 4.1667,
+      "step": 445952
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.268997610660112e-05,
+      "loss": 4.1533,
+      "step": 446464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.268160653789433e-05,
+      "loss": 4.1678,
+      "step": 446976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.267322059038381e-05,
+      "loss": 4.1589,
+      "step": 447488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.266483464287329e-05,
+      "loss": 4.1733,
+      "step": 448000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.265644869536277e-05,
+      "loss": 4.1793,
+      "step": 448512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.264806274785225e-05,
+      "loss": 4.1808,
+      "step": 449024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2639676800341725e-05,
+      "loss": 4.1546,
+      "step": 449536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2631290852831205e-05,
+      "loss": 4.1772,
+      "step": 450048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2622904905320685e-05,
+      "loss": 4.173,
+      "step": 450560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2614518957810165e-05,
+      "loss": 4.1648,
+      "step": 451072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.2606133010299645e-05,
+      "loss": 4.1633,
+      "step": 451584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.259776344159286e-05,
+      "loss": 4.1785,
+      "step": 452096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.258937749408234e-05,
+      "loss": 4.1604,
+      "step": 452608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.258099154657182e-05,
+      "loss": 4.1716,
+      "step": 453120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.25726055990613e-05,
+      "loss": 4.1597,
+      "step": 453632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.256421965155078e-05,
+      "loss": 4.1704,
+      "step": 454144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.255583370404026e-05,
+      "loss": 4.1679,
+      "step": 454656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.254744775652974e-05,
+      "loss": 4.1658,
+      "step": 455168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.253906180901922e-05,
+      "loss": 4.1719,
+      "step": 455680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.253070861911616e-05,
+      "loss": 4.1695,
+      "step": 456192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.252232267160564e-05,
+      "loss": 4.1669,
+      "step": 456704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.251393672409512e-05,
+      "loss": 4.164,
+      "step": 457216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.25055507765846e-05,
+      "loss": 4.1737,
+      "step": 457728
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.187298774719238,
+      "eval_runtime": 287.5705,
+      "eval_samples_per_second": 1326.948,
+      "eval_steps_per_second": 41.468,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497164829074085e-05,
+      "loss": 4.1636,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2488778881563565e-05,
+      "loss": 4.1519,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480392934053045e-05,
+      "loss": 4.1677,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472006986542525e-05,
+      "loss": 4.1707,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2463621039032005e-05,
+      "loss": 4.1723,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455235091521485e-05,
+      "loss": 4.1582,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2446849144010965e-05,
+      "loss": 4.1624,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243846319650044e-05,
+      "loss": 4.1518,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243007724898992e-05,
+      "loss": 4.1665,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24216913014794e-05,
+      "loss": 4.1596,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241330535396888e-05,
+      "loss": 4.1671,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240491940645836e-05,
+      "loss": 4.1671,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239654983775157e-05,
+      "loss": 4.1572,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238816389024105e-05,
+      "loss": 4.164,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2379777942730534e-05,
+      "loss": 4.1495,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371391995220014e-05,
+      "loss": 4.1457,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363006047709494e-05,
+      "loss": 4.1574,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2354620100198974e-05,
+      "loss": 4.1644,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346234152688454e-05,
+      "loss": 4.1566,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2337848205177934e-05,
+      "loss": 4.1755,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329462257667414e-05,
+      "loss": 4.1716,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232109268896062e-05,
+      "loss": 4.1686,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23127067414501e-05,
+      "loss": 4.1589,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230432079393958e-05,
+      "loss": 4.1679,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229593484642906e-05,
+      "loss": 4.1521,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228756527772227e-05,
+      "loss": 4.17,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227917933021175e-05,
+      "loss": 4.1582,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227079338270123e-05,
+      "loss": 4.1508,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226240743519072e-05,
+      "loss": 4.157,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22540214876802e-05,
+      "loss": 4.1565,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224563554016968e-05,
+      "loss": 4.1564,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223726597146289e-05,
+      "loss": 4.1631,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222888002395237e-05,
+      "loss": 4.1612,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222049407644185e-05,
+      "loss": 4.1642,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212124507735056e-05,
+      "loss": 4.154,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2203738560224536e-05,
+      "loss": 4.1595,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195352612714016e-05,
+      "loss": 4.1552,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2186966665203496e-05,
+      "loss": 4.1427,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2178580717692976e-05,
+      "loss": 4.1437,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170211148986185e-05,
+      "loss": 4.1481,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216182520147567e-05,
+      "loss": 4.1516,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215343925396515e-05,
+      "loss": 4.1624,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145053306454625e-05,
+      "loss": 4.1537,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2136667358944105e-05,
+      "loss": 4.145,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128281411433585e-05,
+      "loss": 4.1645,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2119895463923065e-05,
+      "loss": 4.1477,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2111509516412545e-05,
+      "loss": 4.1617,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103139947705754e-05,
+      "loss": 4.1536,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2094754000195234e-05,
+      "loss": 4.1347,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2086368052684714e-05,
+      "loss": 4.1636,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2077982105174194e-05,
+      "loss": 4.142,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2069596157663674e-05,
+      "loss": 4.1604,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061210210153154e-05,
+      "loss": 4.1433,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205284064144637e-05,
+      "loss": 4.1465,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204445469393585e-05,
+      "loss": 4.1404,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203606874642533e-05,
+      "loss": 4.1477,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202768279891481e-05,
+      "loss": 4.1546,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201929685140429e-05,
+      "loss": 4.1399,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20109272826975e-05,
+      "loss": 4.1542,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200254133518698e-05,
+      "loss": 4.1473,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199415538767646e-05,
+      "loss": 4.1353,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198576944016594e-05,
+      "loss": 4.1631,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197738349265542e-05,
+      "loss": 4.1299,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196901392394863e-05,
+      "loss": 4.1262,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196062797643811e-05,
+      "loss": 4.145,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195224202892759e-05,
+      "loss": 4.1551,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1943856081417074e-05,
+      "loss": 4.1329,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1935470133906554e-05,
+      "loss": 4.1504,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927084186396034e-05,
+      "loss": 4.129,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1918698238885514e-05,
+      "loss": 4.1452,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191032867017872e-05,
+      "loss": 4.1489,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19019427226682e-05,
+      "loss": 4.1459,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189355677515768e-05,
+      "loss": 4.1449,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188517082764716e-05,
+      "loss": 4.1457,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1876784880136636e-05,
+      "loss": 4.156,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868398932626116e-05,
+      "loss": 4.1353,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860012985115596e-05,
+      "loss": 4.1424,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1851627037605076e-05,
+      "loss": 4.1443,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843241090094556e-05,
+      "loss": 4.1335,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183487152138777e-05,
+      "loss": 4.1398,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182648557387725e-05,
+      "loss": 4.1402,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181809962636673e-05,
+      "loss": 4.1505,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180973005765994e-05,
+      "loss": 4.1502,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180134411014942e-05,
+      "loss": 4.1278,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17929581626389e-05,
+      "loss": 4.1292,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178457221512838e-05,
+      "loss": 4.1444,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177618626761786e-05,
+      "loss": 4.1512,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176781669891107e-05,
+      "loss": 4.1409,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175943075140055e-05,
+      "loss": 4.136,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175104480389003e-05,
+      "loss": 4.1406,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174265885637951e-05,
+      "loss": 4.1393,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1734272908868996e-05,
+      "loss": 4.1367,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1725886961358476e-05,
+      "loss": 4.1318,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1717501013847956e-05,
+      "loss": 4.1463,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709115066337436e-05,
+      "loss": 4.1396,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1700745497630645e-05,
+      "loss": 4.1436,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1692359550120125e-05,
+      "loss": 4.1305,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1683973602609605e-05,
+      "loss": 4.1367,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1675587655099085e-05,
+      "loss": 4.133,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667201707588565e-05,
+      "loss": 4.1401,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1658815760078045e-05,
+      "loss": 4.1378,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1650429812567525e-05,
+      "loss": 4.13,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642043865057e-05,
+      "loss": 4.1375,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1633674296350214e-05,
+      "loss": 4.1262,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625288348839694e-05,
+      "loss": 4.1304,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1616902401329174e-05,
+      "loss": 4.1264,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1608516453818654e-05,
+      "loss": 4.1289,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160014688511187e-05,
+      "loss": 4.1283,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159176093760135e-05,
+      "loss": 4.1407,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158337499009082e-05,
+      "loss": 4.1236,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15749890425803e-05,
+      "loss": 4.1273,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156661947387352e-05,
+      "loss": 4.1302,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558233526363e-05,
+      "loss": 4.1366,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154984757885247e-05,
+      "loss": 4.1287,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154146163134195e-05,
+      "loss": 4.1406,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153307568383143e-05,
+      "loss": 4.1343,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152470611512465e-05,
+      "loss": 4.1363,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151632016761413e-05,
+      "loss": 4.1345,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150793422010361e-05,
+      "loss": 4.1281,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149954827259309e-05,
+      "loss": 4.1316,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149116232508257e-05,
+      "loss": 4.1367,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148277637757205e-05,
+      "loss": 4.1318,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147439043006153e-05,
+      "loss": 4.1241,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146600448255101e-05,
+      "loss": 4.1366,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1457634913844217e-05,
+      "loss": 4.1322,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1449248966333697e-05,
+      "loss": 4.1277,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1440863018823177e-05,
+      "loss": 4.1126,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1432477071312656e-05,
+      "loss": 4.1241,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424091123802136e-05,
+      "loss": 4.1174,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1415705176291616e-05,
+      "loss": 4.1326,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1407319228781096e-05,
+      "loss": 4.1396,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139893328127058e-05,
+      "loss": 4.1383,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139056371256379e-05,
+      "loss": 4.1152,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138217776505327e-05,
+      "loss": 4.1366,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137379181754275e-05,
+      "loss": 4.1307,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136540587003223e-05,
+      "loss": 4.1299,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135703630132544e-05,
+      "loss": 4.1224,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134865035381492e-05,
+      "loss": 4.1372,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134028078510813e-05,
+      "loss": 4.1197,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133189483759761e-05,
+      "loss": 4.1364,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132350889008709e-05,
+      "loss": 4.1171,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131512294257657e-05,
+      "loss": 4.13,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130673699506605e-05,
+      "loss": 4.1287,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129835104755554e-05,
+      "loss": 4.1256,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1289981478848746e-05,
+      "loss": 4.13,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1281595531338226e-05,
+      "loss": 4.1284,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273209583827706e-05,
+      "loss": 4.1279,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1264840015120915e-05,
+      "loss": 4.1245,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1256454067610395e-05,
+      "loss": 4.1366,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.155002593994141,
+      "eval_runtime": 300.0187,
+      "eval_samples_per_second": 1271.891,
+      "eval_steps_per_second": 39.748,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1248068120099875e-05,
+      "loss": 4.1207,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1239682172589355e-05,
+      "loss": 4.112,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123129622507883e-05,
+      "loss": 4.1292,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122291027756831e-05,
+      "loss": 4.1306,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121452433005779e-05,
+      "loss": 4.1307,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1206138382547275e-05,
+      "loss": 4.1214,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1197752435036754e-05,
+      "loss": 4.1237,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1189366487526234e-05,
+      "loss": 4.111,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1180980540015714e-05,
+      "loss": 4.125,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1172594592505194e-05,
+      "loss": 4.124,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1164208644994674e-05,
+      "loss": 4.1257,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1155822697484154e-05,
+      "loss": 4.1287,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114745312877736e-05,
+      "loss": 4.1222,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113906718126684e-05,
+      "loss": 4.1202,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113068123375632e-05,
+      "loss": 4.1136,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11222952862458e-05,
+      "loss": 4.1075,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111390933873528e-05,
+      "loss": 4.1201,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110552339122476e-05,
+      "loss": 4.1245,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109713744371424e-05,
+      "loss": 4.1188,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108876787500746e-05,
+      "loss": 4.1301,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108038192749694e-05,
+      "loss": 4.1357,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.107199597998642e-05,
+      "loss": 4.1335,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.10636100324759e-05,
+      "loss": 4.1209,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105522408496537e-05,
+      "loss": 4.1241,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104683813745485e-05,
+      "loss": 4.121,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103845218994433e-05,
+      "loss": 4.1266,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103006624243381e-05,
+      "loss": 4.1194,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102168029492329e-05,
+      "loss": 4.1173,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101329434741277e-05,
+      "loss": 4.1216,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100490839990225e-05,
+      "loss": 4.1147,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099652245239173e-05,
+      "loss": 4.1144,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098815288368494e-05,
+      "loss": 4.1242,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097976693617442e-05,
+      "loss": 4.1246,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097138098866391e-05,
+      "loss": 4.1265,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0963011419957117e-05,
+      "loss": 4.1116,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0954641851250326e-05,
+      "loss": 4.1253,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0946255903739806e-05,
+      "loss": 4.12,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0937869956229286e-05,
+      "loss": 4.1022,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0929484008718766e-05,
+      "loss": 4.1059,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0921098061208245e-05,
+      "loss": 4.1108,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0912728492501455e-05,
+      "loss": 4.117,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0904342544990935e-05,
+      "loss": 4.1202,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0895956597480414e-05,
+      "loss": 4.1167,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0887570649969894e-05,
+      "loss": 4.1052,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0879184702459374e-05,
+      "loss": 4.1296,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.087079875494886e-05,
+      "loss": 4.1118,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.086241280743834e-05,
+      "loss": 4.1251,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085402685992782e-05,
+      "loss": 4.1192,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.08456409124173e-05,
+      "loss": 4.09,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083725496490678e-05,
+      "loss": 4.1316,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082886901739626e-05,
+      "loss": 4.1043,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0820483069885734e-05,
+      "loss": 4.1205,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081212987998268e-05,
+      "loss": 4.1097,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080374393247216e-05,
+      "loss": 4.1079,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079535798496164e-05,
+      "loss": 4.1063,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078698841625485e-05,
+      "loss": 4.1059,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077860246874433e-05,
+      "loss": 4.1229,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0770216521233815e-05,
+      "loss": 4.0991,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0761830573723295e-05,
+      "loss": 4.1165,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0753444626212775e-05,
+      "loss": 4.1138,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0745058678702255e-05,
+      "loss": 4.0961,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0736672731191735e-05,
+      "loss": 4.1224,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.072828678368121e-05,
+      "loss": 4.0999,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071990083617069e-05,
+      "loss": 4.0926,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0711531267463904e-05,
+      "loss": 4.1056,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0703145319953384e-05,
+      "loss": 4.1165,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069475937244286e-05,
+      "loss": 4.0979,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.068637342493234e-05,
+      "loss": 4.1132,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067798747742182e-05,
+      "loss": 4.0938,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066961790871503e-05,
+      "loss": 4.1053,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066123196120451e-05,
+      "loss": 4.1126,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065284601369399e-05,
+      "loss": 4.1079,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064446006618347e-05,
+      "loss": 4.1129,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063607411867295e-05,
+      "loss": 4.107,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062768817116243e-05,
+      "loss": 4.1187,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061930222365191e-05,
+      "loss": 4.1014,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061091627614139e-05,
+      "loss": 4.108,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06025467074346e-05,
+      "loss": 4.1035,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059416075992408e-05,
+      "loss": 4.1003,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058577481241356e-05,
+      "loss": 4.1038,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.057740524370677e-05,
+      "loss": 4.1068,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056901929619625e-05,
+      "loss": 4.1129,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056063334868574e-05,
+      "loss": 4.1192,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055224740117522e-05,
+      "loss": 4.09,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05438614536647e-05,
+      "loss": 4.0951,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0535491884957906e-05,
+      "loss": 4.1085,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0527105937447386e-05,
+      "loss": 4.1146,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0518719989936866e-05,
+      "loss": 4.1059,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0510334042426346e-05,
+      "loss": 4.1011,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0501948094915826e-05,
+      "loss": 4.1101,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0493562147405306e-05,
+      "loss": 4.1035,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0485176199894786e-05,
+      "loss": 4.1037,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0476790252384266e-05,
+      "loss": 4.097,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0468420683677475e-05,
+      "loss": 4.1079,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0460034736166955e-05,
+      "loss": 4.109,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0451648788656435e-05,
+      "loss": 4.1088,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0443262841145915e-05,
+      "loss": 4.0967,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0434876893635395e-05,
+      "loss": 4.0977,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042650732492861e-05,
+      "loss": 4.0969,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041813775622182e-05,
+      "loss": 4.1065,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.04097518087113e-05,
+      "loss": 4.1047,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040136586120078e-05,
+      "loss": 4.0956,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039297991369026e-05,
+      "loss": 4.103,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038461034498347e-05,
+      "loss": 4.0937,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037622439747295e-05,
+      "loss": 4.0942,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036783844996243e-05,
+      "loss": 4.0897,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035945250245191e-05,
+      "loss": 4.0957,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035106655494139e-05,
+      "loss": 4.0956,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034268060743087e-05,
+      "loss": 4.1038,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033429465992035e-05,
+      "loss": 4.0957,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032590871240983e-05,
+      "loss": 4.0869,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031752276489931e-05,
+      "loss": 4.0993,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030915319619252e-05,
+      "loss": 4.1008,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0300767248682e-05,
+      "loss": 4.0976,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029239767997521e-05,
+      "loss": 4.1028,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0284011732464686e-05,
+      "loss": 4.099,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0275625784954166e-05,
+      "loss": 4.1095,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0267239837443646e-05,
+      "loss": 4.0986,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0258853889933126e-05,
+      "loss": 4.0962,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0250467942422606e-05,
+      "loss": 4.0941,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024208199491209e-05,
+      "loss": 4.1093,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.02337124262053e-05,
+      "loss": 4.0961,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.022532647869478e-05,
+      "loss": 4.0885,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021694053118426e-05,
+      "loss": 4.1015,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020855458367374e-05,
+      "loss": 4.1035,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020016863616322e-05,
+      "loss": 4.0909,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01917826886527e-05,
+      "loss": 4.0832,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.018341311994591e-05,
+      "loss": 4.0902,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.017502717243539e-05,
+      "loss": 4.0858,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.016664122492487e-05,
+      "loss": 4.0994,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015825527741435e-05,
+      "loss": 4.1036,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014986932990383e-05,
+      "loss": 4.1055,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014148338239331e-05,
+      "loss": 4.0789,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0133113813686527e-05,
+      "loss": 4.1056,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0124727866176006e-05,
+      "loss": 4.0969,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0116341918665486e-05,
+      "loss": 4.0954,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0107955971154966e-05,
+      "loss": 4.0908,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0099570023644446e-05,
+      "loss": 4.1016,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0091184076133926e-05,
+      "loss": 4.0877,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00827981286234e-05,
+      "loss": 4.104,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007441218111288e-05,
+      "loss": 4.0857,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0066042612406095e-05,
+      "loss": 4.0952,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0057673043699304e-05,
+      "loss": 4.097,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0049287096188784e-05,
+      "loss": 4.0942,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0040901148678264e-05,
+      "loss": 4.0951,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0032515201167744e-05,
+      "loss": 4.0949,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002412925365723e-05,
+      "loss": 4.0989,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0015743306146704e-05,
+      "loss": 4.0935,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0007357358636184e-05,
+      "loss": 4.1015,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.129909992218018,
+      "eval_runtime": 291.8189,
+      "eval_samples_per_second": 1307.63,
+      "eval_steps_per_second": 40.864,
+      "step": 610560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99989877899294e-05,
+      "loss": 4.086,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999060184241887e-05,
+      "loss": 4.0839,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.998221589490835e-05,
+      "loss": 4.0933,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997382994739783e-05,
+      "loss": 4.0967,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.996544399988731e-05,
+      "loss": 4.1025,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.995705805237679e-05,
+      "loss": 4.0894,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994867210486627e-05,
+      "loss": 4.0967,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994028615735575e-05,
+      "loss": 4.0821,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.99319329674527e-05,
+      "loss": 4.0887,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.992354701994218e-05,
+      "loss": 4.0919,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.991516107243166e-05,
+      "loss": 4.0951,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9906791503724874e-05,
+      "loss": 4.0955,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989840555621435e-05,
+      "loss": 4.095,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989001960870383e-05,
+      "loss": 4.0849,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.988163366119331e-05,
+      "loss": 4.0841,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987324771368279e-05,
+      "loss": 4.076,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.986486176617227e-05,
+      "loss": 4.0875,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.985647581866175e-05,
+      "loss": 4.0978,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984808987115123e-05,
+      "loss": 4.0857,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9839703923640707e-05,
+      "loss": 4.0984,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9831317976130187e-05,
+      "loss": 4.1038,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9822932028619666e-05,
+      "loss": 4.0983,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9814546081109146e-05,
+      "loss": 4.0949,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.980617651240236e-05,
+      "loss": 4.0883,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979779056489184e-05,
+      "loss": 4.0904,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.978942099618505e-05,
+      "loss": 4.0942,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.978103504867453e-05,
+      "loss": 4.0952,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.977264910116401e-05,
+      "loss": 4.0858,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.976426315365349e-05,
+      "loss": 4.0911,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.97558935849467e-05,
+      "loss": 4.0862,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974750763743618e-05,
+      "loss": 4.0815,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973912168992566e-05,
+      "loss": 4.0922,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973075212121887e-05,
+      "loss": 4.0913,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9722366173708356e-05,
+      "loss": 4.0957,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9713980226197836e-05,
+      "loss": 4.0837,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9705594278687316e-05,
+      "loss": 4.0944,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9697208331176796e-05,
+      "loss": 4.0913,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9688822383666276e-05,
+      "loss": 4.0695,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9680436436155756e-05,
+      "loss": 4.0807,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9672050488645236e-05,
+      "loss": 4.0777,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.966366454113471e-05,
+      "loss": 4.0833,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9655294972427925e-05,
+      "loss": 4.0956,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9646909024917405e-05,
+      "loss": 4.0859,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9638523077406885e-05,
+      "loss": 4.0783,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963013712989636e-05,
+      "loss": 4.0928,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.962175118238584e-05,
+      "loss": 4.083,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9613365234875325e-05,
+      "loss": 4.0966,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9604979287364805e-05,
+      "loss": 4.0886,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9596593339854285e-05,
+      "loss": 4.0602,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9588223771147494e-05,
+      "loss": 4.1005,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957985420244071e-05,
+      "loss": 4.0754,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957146825493018e-05,
+      "loss": 4.0904,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956308230741966e-05,
+      "loss": 4.0808,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.955469635990914e-05,
+      "loss": 4.0781,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.954631041239862e-05,
+      "loss": 4.0771,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953794084369183e-05,
+      "loss": 4.0761,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952955489618131e-05,
+      "loss": 4.096,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952116894867079e-05,
+      "loss": 4.0689,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.951278300116028e-05,
+      "loss": 4.0918,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950439705364976e-05,
+      "loss": 4.0835,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.949601110613924e-05,
+      "loss": 4.0639,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948764153743245e-05,
+      "loss": 4.0953,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947925558992193e-05,
+      "loss": 4.0736,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947086964241141e-05,
+      "loss": 4.0598,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.946248369490089e-05,
+      "loss": 4.0756,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945409774739037e-05,
+      "loss": 4.0861,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.944571179987985e-05,
+      "loss": 4.0718,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9437342231173056e-05,
+      "loss": 4.0839,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9428956283662536e-05,
+      "loss": 4.0673,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9420570336152016e-05,
+      "loss": 4.0715,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9412184388641496e-05,
+      "loss": 4.0833,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9403798441130976e-05,
+      "loss": 4.0811,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939542887242419e-05,
+      "loss": 4.0818,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.938704292491367e-05,
+      "loss": 4.0773,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937865697740315e-05,
+      "loss": 4.093,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937028740869636e-05,
+      "loss": 4.0719,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.936190146118584e-05,
+      "loss": 4.0795,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.935351551367532e-05,
+      "loss": 4.0732,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.93451295661648e-05,
+      "loss": 4.0708,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.933674361865428e-05,
+      "loss": 4.0782,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932835767114376e-05,
+      "loss": 4.0744,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.931997172363324e-05,
+      "loss": 4.0831,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.931158577612272e-05,
+      "loss": 4.0923,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.93031998286122e-05,
+      "loss": 4.0626,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.929481388110168e-05,
+      "loss": 4.0651,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928642793359116e-05,
+      "loss": 4.0805,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.927804198608064e-05,
+      "loss": 4.0817,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926967241737385e-05,
+      "loss": 4.0789,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9261302848667065e-05,
+      "loss": 4.0741,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9252916901156545e-05,
+      "loss": 4.0795,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.924453095364602e-05,
+      "loss": 4.074,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.92361450061355e-05,
+      "loss": 4.0735,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9227775437428714e-05,
+      "loss": 4.0725,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9219389489918194e-05,
+      "loss": 4.08,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.921100354240767e-05,
+      "loss": 4.0823,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.920261759489715e-05,
+      "loss": 4.0813,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9194231647386634e-05,
+      "loss": 4.0661,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9185845699876114e-05,
+      "loss": 4.0725,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9177459752365594e-05,
+      "loss": 4.0703,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9169073804855074e-05,
+      "loss": 4.0753,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916070423614828e-05,
+      "loss": 4.0774,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.915231828863776e-05,
+      "loss": 4.0644,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.914393234112724e-05,
+      "loss": 4.0755,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.913556277242045e-05,
+      "loss": 4.0619,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.912717682490993e-05,
+      "loss": 4.0678,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911879087739941e-05,
+      "loss": 4.0616,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911042130869262e-05,
+      "loss": 4.0697,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91020353611821e-05,
+      "loss": 4.069,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.909364941367159e-05,
+      "loss": 4.0751,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.908526346616107e-05,
+      "loss": 4.0644,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.907687751865055e-05,
+      "loss": 4.0648,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906849157114003e-05,
+      "loss": 4.067,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906010562362951e-05,
+      "loss": 4.0764,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.905171967611899e-05,
+      "loss": 4.0699,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904333372860847e-05,
+      "loss": 4.0733,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.903494778109795e-05,
+      "loss": 4.0699,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.902656183358743e-05,
+      "loss": 4.0837,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901817588607691e-05,
+      "loss": 4.0716,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9009806317370117e-05,
+      "loss": 4.0663,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9001420369859596e-05,
+      "loss": 4.0701,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8993034422349076e-05,
+      "loss": 4.0773,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8984648474838556e-05,
+      "loss": 4.0718,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8976262527328036e-05,
+      "loss": 4.0626,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8967876579817516e-05,
+      "loss": 4.0723,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8959490632306996e-05,
+      "loss": 4.0752,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8951121063600205e-05,
+      "loss": 4.0643,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8942735116089685e-05,
+      "loss": 4.056,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8934349168579165e-05,
+      "loss": 4.0628,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8925963221068645e-05,
+      "loss": 4.0574,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8917577273558125e-05,
+      "loss": 4.0744,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8909191326047605e-05,
+      "loss": 4.0744,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8900805378537085e-05,
+      "loss": 4.0784,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8892419431026565e-05,
+      "loss": 4.0535,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8884049862319774e-05,
+      "loss": 4.0781,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8875663914809254e-05,
+      "loss": 4.0709,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886729434610247e-05,
+      "loss": 4.0701,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885890839859195e-05,
+      "loss": 4.0657,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885052245108143e-05,
+      "loss": 4.0685,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.884213650357091e-05,
+      "loss": 4.0644,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.883375055606039e-05,
+      "loss": 4.0782,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882536460854987e-05,
+      "loss": 4.0551,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.881697866103935e-05,
+      "loss": 4.0701,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880859271352883e-05,
+      "loss": 4.0731,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880022314482204e-05,
+      "loss": 4.0702,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.879183719731152e-05,
+      "loss": 4.0629,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878346762860473e-05,
+      "loss": 4.0697,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8775098059897944e-05,
+      "loss": 4.0717,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8766712112387424e-05,
+      "loss": 4.0668,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8758326164876904e-05,
+      "loss": 4.0739,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.1106696128845215,
+      "eval_runtime": 292.3649,
+      "eval_samples_per_second": 1305.188,
+      "eval_steps_per_second": 40.788,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8749940217366384e-05,
+      "loss": 4.0649,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8741554269855863e-05,
+      "loss": 4.0545,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8733168322345343e-05,
+      "loss": 4.0656,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8724782374834823e-05,
+      "loss": 4.0727,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.87163964273243e-05,
+      "loss": 4.0736,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870801047981378e-05,
+      "loss": 4.0645,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869962453230326e-05,
+      "loss": 4.0689,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869123858479274e-05,
+      "loss": 4.0538,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8682852637282216e-05,
+      "loss": 4.064,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8674466689771696e-05,
+      "loss": 4.0643,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8666080742261176e-05,
+      "loss": 4.0678,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8657694794750656e-05,
+      "loss": 4.0693,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864932522604387e-05,
+      "loss": 4.0734,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864093927853335e-05,
+      "loss": 4.056,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.863255333102283e-05,
+      "loss": 4.0565,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.862416738351231e-05,
+      "loss": 4.0455,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.861578143600179e-05,
+      "loss": 4.0667,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.860739548849127e-05,
+      "loss": 4.0706,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859900954098075e-05,
+      "loss": 4.0562,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.859062359347023e-05,
+      "loss": 4.0697,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858223764595971e-05,
+      "loss": 4.0818,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857385169844919e-05,
+      "loss": 4.0723,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.856546575093867e-05,
+      "loss": 4.0674,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.855707980342815e-05,
+      "loss": 4.0632,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854871023472136e-05,
+      "loss": 4.0675,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854034066601458e-05,
+      "loss": 4.0643,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8531971097307786e-05,
+      "loss": 4.0675,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8523585149797266e-05,
+      "loss": 4.0682,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8515199202286746e-05,
+      "loss": 4.061,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8506813254776226e-05,
+      "loss": 4.0552,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8498427307265706e-05,
+      "loss": 4.0587,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8490041359755186e-05,
+      "loss": 4.0679,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8481655412244665e-05,
+      "loss": 4.0694,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8473269464734145e-05,
+      "loss": 4.0681,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8464883517223625e-05,
+      "loss": 4.0595,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8456513948516834e-05,
+      "loss": 4.0614,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8448128001006314e-05,
+      "loss": 4.0679,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8439742053495794e-05,
+      "loss": 4.047,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843135610598528e-05,
+      "loss": 4.0529,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8422970158474754e-05,
+      "loss": 4.0549,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8414584210964234e-05,
+      "loss": 4.0563,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840621464225745e-05,
+      "loss": 4.0687,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839782869474693e-05,
+      "loss": 4.0601,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83894427472364e-05,
+      "loss": 4.0549,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838105679972588e-05,
+      "loss": 4.0679,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837267085221536e-05,
+      "loss": 4.0554,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836430128350858e-05,
+      "loss": 4.0674,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835591533599805e-05,
+      "loss": 4.071,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834752938848753e-05,
+      "loss": 4.0361,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833914344097702e-05,
+      "loss": 4.0673,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83307574934665e-05,
+      "loss": 4.0547,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.832237154595598e-05,
+      "loss": 4.0648,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.831398559844546e-05,
+      "loss": 4.0544,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.830561602973867e-05,
+      "loss": 4.0508,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829723008222815e-05,
+      "loss": 4.0545,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828884413471763e-05,
+      "loss": 4.0508,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828045818720711e-05,
+      "loss": 4.0732,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827208861850032e-05,
+      "loss": 4.0417,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82637026709898e-05,
+      "loss": 4.0675,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825531672347928e-05,
+      "loss": 4.06,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824693077596876e-05,
+      "loss": 4.0396,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823854482845824e-05,
+      "loss": 4.0695,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8230158880947717e-05,
+      "loss": 4.0475,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8221772933437197e-05,
+      "loss": 4.0362,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821340336473041e-05,
+      "loss": 4.0475,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.820501741721989e-05,
+      "loss": 4.0645,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819663146970937e-05,
+      "loss": 4.0423,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818824552219885e-05,
+      "loss": 4.0613,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817985957468833e-05,
+      "loss": 4.0434,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817147362717781e-05,
+      "loss": 4.045,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.816308767966729e-05,
+      "loss": 4.0569,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8154701732156765e-05,
+      "loss": 4.0578,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8146315784646245e-05,
+      "loss": 4.0579,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8137929837135725e-05,
+      "loss": 4.0559,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8129560268428934e-05,
+      "loss": 4.0641,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812117432091842e-05,
+      "loss": 4.0529,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.81127883734079e-05,
+      "loss": 4.0534,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810441880470112e-05,
+      "loss": 4.0438,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.809603285719059e-05,
+      "loss": 4.0489,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8087663288483806e-05,
+      "loss": 4.0527,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8079277340973286e-05,
+      "loss": 4.0484,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8070891393462766e-05,
+      "loss": 4.0614,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806250544595224e-05,
+      "loss": 4.0652,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8054135877245455e-05,
+      "loss": 4.045,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8045749929734935e-05,
+      "loss": 4.0394,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803736398222441e-05,
+      "loss": 4.054,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802897803471389e-05,
+      "loss": 4.0574,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8020592087203375e-05,
+      "loss": 4.0542,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8012206139692855e-05,
+      "loss": 4.0478,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8003820192182335e-05,
+      "loss": 4.0565,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7995434244671815e-05,
+      "loss": 4.0492,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7987048297161295e-05,
+      "loss": 4.0494,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7978662349650775e-05,
+      "loss": 4.0487,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7970276402140254e-05,
+      "loss": 4.0554,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7961890454629734e-05,
+      "loss": 4.0554,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7953520885922944e-05,
+      "loss": 4.0584,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794515131721615e-05,
+      "loss": 4.0388,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793676536970563e-05,
+      "loss": 4.0524,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792837942219511e-05,
+      "loss": 4.043,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.792000985348833e-05,
+      "loss": 4.0526,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.791162390597781e-05,
+      "loss": 4.0539,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.790323795846729e-05,
+      "loss": 4.0445,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789485201095677e-05,
+      "loss": 4.0518,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788646606344625e-05,
+      "loss": 4.0374,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787808011593573e-05,
+      "loss": 4.0452,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786969416842521e-05,
+      "loss": 4.0379,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786130822091469e-05,
+      "loss": 4.0463,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.785292227340417e-05,
+      "loss": 4.0461,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784453632589365e-05,
+      "loss": 4.0477,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783615037838313e-05,
+      "loss": 4.0407,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78277644308726e-05,
+      "loss": 4.0419,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781939486216582e-05,
+      "loss": 4.0443,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78110089146553e-05,
+      "loss": 4.0508,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780262296714478e-05,
+      "loss": 4.0455,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779423701963426e-05,
+      "loss": 4.0484,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778586745092747e-05,
+      "loss": 4.0465,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777749788222068e-05,
+      "loss": 4.0574,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776911193471016e-05,
+      "loss": 4.0487,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776072598719964e-05,
+      "loss": 4.0444,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775234003968912e-05,
+      "loss": 4.0464,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.774397047098233e-05,
+      "loss": 4.0542,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.773558452347181e-05,
+      "loss": 4.0505,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.772719857596129e-05,
+      "loss": 4.0408,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771881262845077e-05,
+      "loss": 4.0437,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771042668094025e-05,
+      "loss": 4.0539,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770204073342973e-05,
+      "loss": 4.0409,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.769365478591921e-05,
+      "loss": 4.0336,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768526883840869e-05,
+      "loss": 4.0381,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76768992697019e-05,
+      "loss": 4.0319,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766851332219138e-05,
+      "loss": 4.0556,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766012737468086e-05,
+      "loss": 4.0446,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.765174142717034e-05,
+      "loss": 4.0576,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.764337185846355e-05,
+      "loss": 4.033,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763498591095303e-05,
+      "loss": 4.0506,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762659996344251e-05,
+      "loss": 4.0467,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761821401593199e-05,
+      "loss": 4.0485,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7609844447225204e-05,
+      "loss": 4.0438,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7601458499714684e-05,
+      "loss": 4.0405,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7593072552204164e-05,
+      "loss": 4.0387,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7584686604693644e-05,
+      "loss": 4.0584,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7576300657183124e-05,
+      "loss": 4.0316,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7567914709672604e-05,
+      "loss": 4.0485,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7559528762162084e-05,
+      "loss": 4.0519,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7551142814651564e-05,
+      "loss": 4.0436,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754277324594477e-05,
+      "loss": 4.0395,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753438729843425e-05,
+      "loss": 4.0481,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752600135092373e-05,
+      "loss": 4.0486,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751763178221694e-05,
+      "loss": 4.0461,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.750924583470642e-05,
+      "loss": 4.0515,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0947651863098145,
+      "eval_runtime": 295.4434,
+      "eval_samples_per_second": 1291.587,
+      "eval_steps_per_second": 40.363,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.75008598871959e-05,
+      "loss": 4.0339,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.749247393968538e-05,
+      "loss": 4.0322,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.748408799217487e-05,
+      "loss": 4.0454,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.747570204466435e-05,
+      "loss": 4.0494,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.746731609715383e-05,
+      "loss": 4.0529,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745894652844704e-05,
+      "loss": 4.0408,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745056058093652e-05,
+      "loss": 4.0454,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7442174633426e-05,
+      "loss": 4.0335,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.743378868591548e-05,
+      "loss": 4.0387,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742541911720869e-05,
+      "loss": 4.043,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.741703316969817e-05,
+      "loss": 4.0416,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7408647222187647e-05,
+      "loss": 4.0486,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7400277653480856e-05,
+      "loss": 4.0495,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7391891705970336e-05,
+      "loss": 4.0332,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.738350575845982e-05,
+      "loss": 4.0344,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.73751198109493e-05,
+      "loss": 4.0242,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736673386343878e-05,
+      "loss": 4.0467,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735834791592826e-05,
+      "loss": 4.0449,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7349961968417735e-05,
+      "loss": 4.0329,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7341576020907215e-05,
+      "loss": 4.0427,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7333190073396695e-05,
+      "loss": 4.0602,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7324804125886175e-05,
+      "loss": 4.0497,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7316418178375655e-05,
+      "loss": 4.0458,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7308048609668864e-05,
+      "loss": 4.0409,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7299662662158344e-05,
+      "loss": 4.047,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7291276714647824e-05,
+      "loss": 4.0406,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7282890767137304e-05,
+      "loss": 4.0414,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.727450481962679e-05,
+      "loss": 4.0445,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726611887211627e-05,
+      "loss": 4.0406,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.725773292460575e-05,
+      "loss": 4.0365,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724934697709523e-05,
+      "loss": 4.0368,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724096102958471e-05,
+      "loss": 4.0439,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.723257508207419e-05,
+      "loss": 4.0434,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722418913456367e-05,
+      "loss": 4.0489,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.721580318705315e-05,
+      "loss": 4.0393,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720743361834636e-05,
+      "loss": 4.0385,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719904767083584e-05,
+      "loss": 4.0452,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719066172332532e-05,
+      "loss": 4.0294,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718227577581479e-05,
+      "loss": 4.0308,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717388982830427e-05,
+      "loss": 4.0333,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.716550388079375e-05,
+      "loss": 4.0322,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715711793328324e-05,
+      "loss": 4.0428,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714874836457645e-05,
+      "loss": 4.0383,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714036241706593e-05,
+      "loss": 4.0391,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.713197646955541e-05,
+      "loss": 4.0403,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712359052204489e-05,
+      "loss": 4.0355,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.71152209533381e-05,
+      "loss": 4.0472,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.710683500582758e-05,
+      "loss": 4.0473,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709844905831706e-05,
+      "loss": 4.0135,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709006311080654e-05,
+      "loss": 4.0469,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7081693542099746e-05,
+      "loss": 4.0317,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7073307594589226e-05,
+      "loss": 4.0442,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7064921647078706e-05,
+      "loss": 4.0365,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.705653569956819e-05,
+      "loss": 4.0264,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704814975205767e-05,
+      "loss": 4.0346,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703976380454715e-05,
+      "loss": 4.0254,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703137785703663e-05,
+      "loss": 4.047,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702299190952611e-05,
+      "loss": 4.0236,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.701462234081932e-05,
+      "loss": 4.0426,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.70062363933088e-05,
+      "loss": 4.0378,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.699786682460201e-05,
+      "loss": 4.0222,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698948087709149e-05,
+      "loss": 4.0447,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698109492958097e-05,
+      "loss": 4.0245,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.697270898207045e-05,
+      "loss": 4.0189,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.696432303455993e-05,
+      "loss": 4.0239,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.695593708704941e-05,
+      "loss": 4.0446,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694755113953889e-05,
+      "loss": 4.0228,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693916519202838e-05,
+      "loss": 4.0419,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693079562332159e-05,
+      "loss": 4.0165,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6922426054614796e-05,
+      "loss": 4.0274,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6914040107104276e-05,
+      "loss": 4.0333,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6905654159593756e-05,
+      "loss": 4.0349,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6897268212083236e-05,
+      "loss": 4.037,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6888882264572716e-05,
+      "loss": 4.0365,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6880496317062196e-05,
+      "loss": 4.0408,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6872110369551675e-05,
+      "loss": 4.028,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6863724422041155e-05,
+      "loss": 4.0319,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685533847453063e-05,
+      "loss": 4.0262,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6846968905823844e-05,
+      "loss": 4.0292,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683858295831333e-05,
+      "loss": 4.0304,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683019701080281e-05,
+      "loss": 4.0225,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682182744209602e-05,
+      "loss": 4.0375,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68134414945855e-05,
+      "loss": 4.0435,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.680505554707498e-05,
+      "loss": 4.0311,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679666959956445e-05,
+      "loss": 4.0109,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.678828365205393e-05,
+      "loss": 4.034,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677989770454341e-05,
+      "loss": 4.0332,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677152813583663e-05,
+      "loss": 4.0382,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.676315856712984e-05,
+      "loss": 4.022,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.675477261961932e-05,
+      "loss": 4.0378,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.67463866721088e-05,
+      "loss": 4.028,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6738000724598285e-05,
+      "loss": 4.0275,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672961477708776e-05,
+      "loss": 4.0273,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672122882957724e-05,
+      "loss": 4.0336,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.671284288206672e-05,
+      "loss": 4.0326,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.67044569345562e-05,
+      "loss": 4.0384,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.669608736584941e-05,
+      "loss": 4.019,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668770141833889e-05,
+      "loss": 4.0292,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667931547082837e-05,
+      "loss": 4.0241,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667092952331785e-05,
+      "loss": 4.0316,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666254357580733e-05,
+      "loss": 4.0312,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.665415762829681e-05,
+      "loss": 4.026,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664577168078629e-05,
+      "loss": 4.0276,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66374021120795e-05,
+      "loss": 4.0161,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662901616456898e-05,
+      "loss": 4.0274,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662063021705846e-05,
+      "loss": 4.0138,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.661224426954794e-05,
+      "loss": 4.0249,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.660385832203742e-05,
+      "loss": 4.0212,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.65954723745269e-05,
+      "loss": 4.0308,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658708642701638e-05,
+      "loss": 4.0189,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657870047950586e-05,
+      "loss": 4.0205,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657033091079907e-05,
+      "loss": 4.0196,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.656194496328855e-05,
+      "loss": 4.0334,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.655355901577803e-05,
+      "loss": 4.0244,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.654518944707124e-05,
+      "loss": 4.0264,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.653680349956072e-05,
+      "loss": 4.0284,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6528433930853936e-05,
+      "loss": 4.0354,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6520047983343416e-05,
+      "loss": 4.026,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6511662035832896e-05,
+      "loss": 4.025,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6503276088322376e-05,
+      "loss": 4.0241,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6494890140811856e-05,
+      "loss": 4.0324,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6486520572105065e-05,
+      "loss": 4.0275,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6478134624594545e-05,
+      "loss": 4.0213,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6469748677084025e-05,
+      "loss": 4.0241,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6461362729573505e-05,
+      "loss": 4.0326,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6452976782062985e-05,
+      "loss": 4.0156,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6444590834552465e-05,
+      "loss": 4.0176,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.643620488704194e-05,
+      "loss": 4.0119,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6427835318335154e-05,
+      "loss": 4.0151,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641944937082464e-05,
+      "loss": 4.0301,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6411063423314114e-05,
+      "loss": 4.0293,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6402677475803594e-05,
+      "loss": 4.0391,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6394291528293074e-05,
+      "loss": 4.0115,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6385905580782554e-05,
+      "loss": 4.0307,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6377519633272034e-05,
+      "loss": 4.0238,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6369133685761514e-05,
+      "loss": 4.0297,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6360747738250994e-05,
+      "loss": 4.0209,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63523781695442e-05,
+      "loss": 4.0215,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.634399222203368e-05,
+      "loss": 4.0218,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.633560627452316e-05,
+      "loss": 4.0389,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.632722032701264e-05,
+      "loss": 4.0094,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631883437950212e-05,
+      "loss": 4.0284,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631046481079534e-05,
+      "loss": 4.0288,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.630207886328482e-05,
+      "loss": 4.0228,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.629370929457803e-05,
+      "loss": 4.0202,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.628532334706751e-05,
+      "loss": 4.0292,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.627693739955699e-05,
+      "loss": 4.024,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626855145204647e-05,
+      "loss": 4.0288,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626016550453595e-05,
+      "loss": 4.0275,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.080966949462891,
+      "eval_runtime": 297.0219,
+      "eval_samples_per_second": 1284.723,
+      "eval_steps_per_second": 40.149,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625177955702543e-05,
+      "loss": 4.0032,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624339360951491e-05,
+      "loss": 4.0143,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623500766200439e-05,
+      "loss": 4.0234,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622662171449387e-05,
+      "loss": 4.0275,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621823576698335e-05,
+      "loss": 4.0329,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620984981947283e-05,
+      "loss": 4.0204,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.620146387196231e-05,
+      "loss": 4.0271,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.619307792445179e-05,
+      "loss": 4.0181,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.618469197694127e-05,
+      "loss": 4.016,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.617630602943075e-05,
+      "loss": 4.0177,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.616792008192023e-05,
+      "loss": 4.0273,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.615953413440971e-05,
+      "loss": 4.0249,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6151164565702916e-05,
+      "loss": 4.0341,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6142778618192396e-05,
+      "loss": 4.0118,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6134392670681876e-05,
+      "loss": 4.0155,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6126006723171356e-05,
+      "loss": 4.0027,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6117620775660836e-05,
+      "loss": 4.0268,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6109234828150316e-05,
+      "loss": 4.0231,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6100848880639796e-05,
+      "loss": 4.0127,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6092462933129276e-05,
+      "loss": 4.0241,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6084076985618756e-05,
+      "loss": 4.0435,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607570741691197e-05,
+      "loss": 4.029,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606732146940145e-05,
+      "loss": 4.0239,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605895190069466e-05,
+      "loss": 4.0217,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605056595318414e-05,
+      "loss": 4.0267,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604218000567362e-05,
+      "loss": 4.0247,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.60337940581631e-05,
+      "loss": 4.0197,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.602540811065258e-05,
+      "loss": 4.0232,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.601702216314206e-05,
+      "loss": 4.0208,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.600863621563154e-05,
+      "loss": 4.015,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600025026812101e-05,
+      "loss": 4.0207,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.599186432061049e-05,
+      "loss": 4.0188,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.598347837309998e-05,
+      "loss": 4.0249,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.597510880439319e-05,
+      "loss": 4.0311,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596672285688267e-05,
+      "loss": 4.0214,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5958353288175885e-05,
+      "loss": 4.0161,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5949967340665365e-05,
+      "loss": 4.0274,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594158139315484e-05,
+      "loss": 4.0094,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.593319544564432e-05,
+      "loss": 4.0132,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59248094981338e-05,
+      "loss": 4.0099,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591642355062328e-05,
+      "loss": 4.0159,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590803760311276e-05,
+      "loss": 4.018,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589965165560224e-05,
+      "loss": 4.019,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589126570809172e-05,
+      "loss": 4.0214,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58828797605812e-05,
+      "loss": 4.0191,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.587449381307068e-05,
+      "loss": 4.0213,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5866107865560164e-05,
+      "loss": 4.0236,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5857738296853374e-05,
+      "loss": 4.0298,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5849352349342854e-05,
+      "loss": 3.9969,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5840966401832334e-05,
+      "loss": 4.0286,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583259683312554e-05,
+      "loss": 4.0118,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.582421088561502e-05,
+      "loss": 4.0224,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58158249381045e-05,
+      "loss": 4.0155,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580743899059398e-05,
+      "loss": 4.0122,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579905304308346e-05,
+      "loss": 4.0094,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579066709557294e-05,
+      "loss": 4.0063,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.578228114806242e-05,
+      "loss": 4.0331,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57738952005519e-05,
+      "loss": 4.0019,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.576550925304138e-05,
+      "loss": 4.0247,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.575715606313833e-05,
+      "loss": 4.0147,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574877011562781e-05,
+      "loss": 4.0063,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574038416811729e-05,
+      "loss": 4.025,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.573199822060677e-05,
+      "loss": 4.009,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.572361227309625e-05,
+      "loss": 3.9957,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.571522632558573e-05,
+      "loss": 4.0054,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57068403780752e-05,
+      "loss": 4.0252,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5698487188172145e-05,
+      "loss": 4.0045,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5690101240661625e-05,
+      "loss": 4.023,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5681715293151105e-05,
+      "loss": 3.9961,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5673329345640585e-05,
+      "loss": 4.0087,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.566494339813007e-05,
+      "loss": 4.0148,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.565655745061955e-05,
+      "loss": 4.0185,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5648171503109025e-05,
+      "loss": 4.0159,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5639785555598505e-05,
+      "loss": 4.0169,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5631399608087985e-05,
+      "loss": 4.0216,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5623013660577465e-05,
+      "loss": 4.0092,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5614644091870674e-05,
+      "loss": 4.0106,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.560627452316389e-05,
+      "loss": 4.0071,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559788857565337e-05,
+      "loss": 4.0145,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558950262814285e-05,
+      "loss": 4.0087,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558111668063232e-05,
+      "loss": 4.0072,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.557274711192554e-05,
+      "loss": 4.0187,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556436116441502e-05,
+      "loss": 4.0225,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55559752169045e-05,
+      "loss": 4.0121,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554758926939398e-05,
+      "loss": 3.9916,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553920332188346e-05,
+      "loss": 4.0148,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553081737437294e-05,
+      "loss": 4.0154,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552243142686242e-05,
+      "loss": 4.0221,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55140454793519e-05,
+      "loss": 4.0038,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.550565953184138e-05,
+      "loss": 4.0144,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549727358433086e-05,
+      "loss": 4.0159,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548890401562407e-05,
+      "loss": 4.0052,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5480534446917277e-05,
+      "loss": 4.0058,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5472148499406757e-05,
+      "loss": 4.0179,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.546376255189624e-05,
+      "loss": 4.0104,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.545537660438572e-05,
+      "loss": 4.0207,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54469906568752e-05,
+      "loss": 4.0042,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543860470936468e-05,
+      "loss": 4.0104,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543021876185416e-05,
+      "loss": 4.0027,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.542183281434364e-05,
+      "loss": 4.0139,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.541344686683312e-05,
+      "loss": 4.0151,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.54050609193226e-05,
+      "loss": 4.0048,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.539669135061581e-05,
+      "loss": 4.0102,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.538830540310529e-05,
+      "loss": 4.0018,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537991945559477e-05,
+      "loss": 4.0079,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.537153350808425e-05,
+      "loss": 3.993,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.536314756057373e-05,
+      "loss": 4.0071,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.535476161306321e-05,
+      "loss": 4.002,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.534637566555269e-05,
+      "loss": 4.0104,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533800609684591e-05,
+      "loss": 4.0035,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532962014933539e-05,
+      "loss": 4.0103,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532123420182486e-05,
+      "loss": 3.9994,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.531284825431434e-05,
+      "loss": 4.0153,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.530446230680382e-05,
+      "loss": 4.0051,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52960763592933e-05,
+      "loss": 4.0062,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528770679058651e-05,
+      "loss": 4.0079,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527932084307599e-05,
+      "loss": 4.0241,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527093489556547e-05,
+      "loss": 4.0051,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.526254894805495e-05,
+      "loss": 4.0044,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.525416300054443e-05,
+      "loss": 4.0101,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.524577705303391e-05,
+      "loss": 4.0144,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5237391105523396e-05,
+      "loss": 4.0069,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5229005158012876e-05,
+      "loss": 4.0083,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5220619210502356e-05,
+      "loss": 4.0021,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5212249641795565e-05,
+      "loss": 4.0118,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5203863694285045e-05,
+      "loss": 3.9983,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5195477746774525e-05,
+      "loss": 4.0002,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5187108178067734e-05,
+      "loss": 3.9947,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5178722230557214e-05,
+      "loss": 4.0023,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5170336283046694e-05,
+      "loss": 4.0118,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5161950335536174e-05,
+      "loss": 4.0091,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5153564388025654e-05,
+      "loss": 4.0182,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5145178440515134e-05,
+      "loss": 3.9988,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.513680887180835e-05,
+      "loss": 4.0099,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512842292429783e-05,
+      "loss": 4.0097,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512003697678731e-05,
+      "loss": 4.009,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.511165102927679e-05,
+      "loss": 4.0032,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510326508176627e-05,
+      "loss": 3.9999,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.509487913425575e-05,
+      "loss": 4.0077,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508649318674522e-05,
+      "loss": 4.0133,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.50781072392347e-05,
+      "loss": 3.9983,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506973767052792e-05,
+      "loss": 4.0096,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506136810182113e-05,
+      "loss": 4.0095,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505298215431061e-05,
+      "loss": 4.0056,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504459620680009e-05,
+      "loss": 4.0033,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.503621025928957e-05,
+      "loss": 4.008,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502782431177905e-05,
+      "loss": 4.0048,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5019454743072263e-05,
+      "loss": 4.0086,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5011068795561743e-05,
+      "loss": 4.0138,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.069571018218994,
+      "eval_runtime": 292.0143,
+      "eval_samples_per_second": 1306.754,
+      "eval_steps_per_second": 40.837,
+      "step": 915840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5002682848051223e-05,
+      "loss": 3.9997,
+      "step": 915968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4994296900540697e-05,
+      "loss": 3.991,
+      "step": 916480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4985910953030177e-05,
+      "loss": 4.0043,
+      "step": 916992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4977525005519656e-05,
+      "loss": 4.0115,
+      "step": 917504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.496915543681287e-05,
+      "loss": 4.0098,
+      "step": 918016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4960769489302346e-05,
+      "loss": 4.0077,
+      "step": 918528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.495239992059556e-05,
+      "loss": 4.0069,
+      "step": 919040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.494401397308504e-05,
+      "loss": 4.0026,
+      "step": 919552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.493562802557452e-05,
+      "loss": 3.996,
+      "step": 920064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4927242078064e-05,
+      "loss": 4.0005,
+      "step": 920576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491885613055348e-05,
+      "loss": 4.0108,
+      "step": 921088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.491047018304296e-05,
+      "loss": 4.0052,
+      "step": 921600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.490210061433617e-05,
+      "loss": 4.0159,
+      "step": 922112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.489371466682565e-05,
+      "loss": 3.9925,
+      "step": 922624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.488532871931513e-05,
+      "loss": 3.999,
+      "step": 923136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.487694277180461e-05,
+      "loss": 3.9869,
+      "step": 923648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486855682429409e-05,
+      "loss": 4.0073,
+      "step": 924160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.486017087678357e-05,
+      "loss": 4.0053,
+      "step": 924672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.485178492927305e-05,
+      "loss": 4.0005,
+      "step": 925184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.484339898176253e-05,
+      "loss": 4.0026,
+      "step": 925696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.483501303425201e-05,
+      "loss": 4.0228,
+      "step": 926208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.482664346554522e-05,
+      "loss": 4.0111,
+      "step": 926720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4818257518034706e-05,
+      "loss": 4.0062,
+      "step": 927232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4809871570524186e-05,
+      "loss": 4.0017,
+      "step": 927744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4801485623013666e-05,
+      "loss": 4.0144,
+      "step": 928256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4793116054306875e-05,
+      "loss": 4.0037,
+      "step": 928768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4784746485600084e-05,
+      "loss": 3.9995,
+      "step": 929280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4776360538089564e-05,
+      "loss": 4.01,
+      "step": 929792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4767974590579044e-05,
+      "loss": 3.9968,
+      "step": 930304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4759588643068524e-05,
+      "loss": 4.0021,
+      "step": 930816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4751202695558004e-05,
+      "loss": 4.0033,
+      "step": 931328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4742816748047484e-05,
+      "loss": 3.9999,
+      "step": 931840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4734430800536964e-05,
+      "loss": 4.0062,
+      "step": 932352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4726044853026444e-05,
+      "loss": 4.0148,
+      "step": 932864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4717658905515924e-05,
+      "loss": 4.0004,
+      "step": 933376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470928933680914e-05,
+      "loss": 4.0038,
+      "step": 933888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.470090338929862e-05,
+      "loss": 4.0108,
+      "step": 934400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46925174417881e-05,
+      "loss": 3.9899,
+      "step": 934912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.468413149427758e-05,
+      "loss": 3.9943,
+      "step": 935424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.467574554676706e-05,
+      "loss": 3.9965,
+      "step": 935936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.466737597806027e-05,
+      "loss": 3.9935,
+      "step": 936448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465899003054975e-05,
+      "loss": 3.9991,
+      "step": 936960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.465060408303923e-05,
+      "loss": 4.0036,
+      "step": 937472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.46422181355287e-05,
+      "loss": 4.0026,
+      "step": 937984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.463383218801818e-05,
+      "loss": 4.0027,
+      "step": 938496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462544624050766e-05,
+      "loss": 4.0027,
+      "step": 939008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.461706029299714e-05,
+      "loss": 4.0084,
+      "step": 939520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460867434548663e-05,
+      "loss": 4.0092,
+      "step": 940032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.460030477677984e-05,
+      "loss": 3.9841,
+      "step": 940544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.459191882926932e-05,
+      "loss": 4.0074,
+      "step": 941056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.45835328817588e-05,
+      "loss": 3.9948,
+      "step": 941568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.457514693424828e-05,
+      "loss": 4.0026,
+      "step": 942080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4566777365541486e-05,
+      "loss": 4.0016,
+      "step": 942592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4558391418030966e-05,
+      "loss": 3.9926,
+      "step": 943104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4550005470520446e-05,
+      "loss": 3.9925,
+      "step": 943616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4541619523009926e-05,
+      "loss": 3.9858,
+      "step": 944128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4533233575499406e-05,
+      "loss": 4.0204,
+      "step": 944640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4524847627988886e-05,
+      "loss": 3.987,
+      "step": 945152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4516461680478366e-05,
+      "loss": 4.0049,
+      "step": 945664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450809211177158e-05,
+      "loss": 3.9975,
+      "step": 946176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449970616426106e-05,
+      "loss": 3.9893,
+      "step": 946688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449132021675054e-05,
+      "loss": 4.0028,
+      "step": 947200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.448293426924002e-05,
+      "loss": 3.9958,
+      "step": 947712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44745483217295e-05,
+      "loss": 3.9835,
+      "step": 948224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.446616237421898e-05,
+      "loss": 3.9816,
+      "step": 948736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.445779280551219e-05,
+      "loss": 4.0136,
+      "step": 949248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.444940685800167e-05,
+      "loss": 3.9896,
+      "step": 949760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.444102091049115e-05,
+      "loss": 3.9983,
+      "step": 950272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.443263496298063e-05,
+      "loss": 3.982,
+      "step": 950784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.442424901547011e-05,
+      "loss": 3.9918,
+      "step": 951296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.441586306795959e-05,
+      "loss": 3.9947,
+      "step": 951808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4407477120449063e-05,
+      "loss": 4.0037,
+      "step": 952320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4399091172938543e-05,
+      "loss": 4.0012,
+      "step": 952832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.439070522542803e-05,
+      "loss": 3.9999,
+      "step": 953344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4382352035524975e-05,
+      "loss": 4.0046,
+      "step": 953856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4373966088014455e-05,
+      "loss": 3.992,
+      "step": 954368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4365580140503935e-05,
+      "loss": 3.9933,
+      "step": 954880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4357194192993415e-05,
+      "loss": 3.991,
+      "step": 955392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4348808245482895e-05,
+      "loss": 3.998,
+      "step": 955904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.434042229797237e-05,
+      "loss": 3.9936,
+      "step": 956416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.433203635046185e-05,
+      "loss": 3.9881,
+      "step": 956928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4323666781755064e-05,
+      "loss": 4.0051,
+      "step": 957440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.431528083424454e-05,
+      "loss": 4.0053,
+      "step": 957952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.430689488673402e-05,
+      "loss": 3.9967,
+      "step": 958464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.42985089392235e-05,
+      "loss": 3.9712,
+      "step": 958976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4290122991712984e-05,
+      "loss": 3.9963,
+      "step": 959488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4281737044202464e-05,
+      "loss": 4.0003,
+      "step": 960000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4273351096691944e-05,
+      "loss": 4.0031,
+      "step": 960512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4264965149181424e-05,
+      "loss": 3.9943,
+      "step": 961024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.425659558047463e-05,
+      "loss": 3.9949,
+      "step": 961536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.424822601176784e-05,
+      "loss": 3.9989,
+      "step": 962048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.423984006425732e-05,
+      "loss": 3.9884,
+      "step": 962560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.42314541167468e-05,
+      "loss": 3.9902,
+      "step": 963072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.422306816923628e-05,
+      "loss": 4.0002,
+      "step": 963584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.421468222172576e-05,
+      "loss": 3.9914,
+      "step": 964096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.420629627421524e-05,
+      "loss": 4.0081,
+      "step": 964608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.419791032670472e-05,
+      "loss": 3.9904,
+      "step": 965120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.418954075799794e-05,
+      "loss": 3.9891,
+      "step": 965632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.418115481048742e-05,
+      "loss": 3.9871,
+      "step": 966144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41727688629769e-05,
+      "loss": 3.9979,
+      "step": 966656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.416438291546638e-05,
+      "loss": 4.0003,
+      "step": 967168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.415599696795586e-05,
+      "loss": 3.9865,
+      "step": 967680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.414761102044534e-05,
+      "loss": 3.9931,
+      "step": 968192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.413922507293482e-05,
+      "loss": 3.9826,
+      "step": 968704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.41308391254243e-05,
+      "loss": 3.9957,
+      "step": 969216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4122469556717506e-05,
+      "loss": 3.9727,
+      "step": 969728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4114083609206986e-05,
+      "loss": 3.9884,
+      "step": 970240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4105697661696466e-05,
+      "loss": 3.9908,
+      "step": 970752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4097311714185946e-05,
+      "loss": 3.9955,
+      "step": 971264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4088925766675426e-05,
+      "loss": 3.983,
+      "step": 971776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4080539819164906e-05,
+      "loss": 3.9944,
+      "step": 972288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4072153871654386e-05,
+      "loss": 3.9843,
+      "step": 972800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4063767924143866e-05,
+      "loss": 3.9965,
+      "step": 973312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.405539835543708e-05,
+      "loss": 3.9905,
+      "step": 973824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4047012407926555e-05,
+      "loss": 3.9892,
+      "step": 974336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4038626460416035e-05,
+      "loss": 3.9898,
+      "step": 974848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.403025689170925e-05,
+      "loss": 4.0074,
+      "step": 975360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4021870944198724e-05,
+      "loss": 3.99,
+      "step": 975872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4013484996688204e-05,
+      "loss": 3.9893,
+      "step": 976384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4005099049177684e-05,
+      "loss": 3.995,
+      "step": 976896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3996713101667164e-05,
+      "loss": 3.9936,
+      "step": 977408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3988327154156644e-05,
+      "loss": 3.9914,
+      "step": 977920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.397995758544986e-05,
+      "loss": 3.9922,
+      "step": 978432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.397157163793934e-05,
+      "loss": 3.9867,
+      "step": 978944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.396318569042882e-05,
+      "loss": 3.9971,
+      "step": 979456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.39547997429183e-05,
+      "loss": 3.9786,
+      "step": 979968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.394643017421151e-05,
+      "loss": 3.9902,
+      "step": 980480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.393804422670099e-05,
+      "loss": 3.9802,
+      "step": 980992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392965827919047e-05,
+      "loss": 3.987,
+      "step": 981504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.392127233167995e-05,
+      "loss": 3.988,
+      "step": 982016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.391288638416943e-05,
+      "loss": 3.9971,
+      "step": 982528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3904533194266374e-05,
+      "loss": 4.0047,
+      "step": 983040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.389614724675585e-05,
+      "loss": 3.9836,
+      "step": 983552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.388776129924533e-05,
+      "loss": 3.9937,
+      "step": 984064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3879375351734813e-05,
+      "loss": 3.9926,
+      "step": 984576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.387098940422429e-05,
+      "loss": 3.9928,
+      "step": 985088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.386260345671377e-05,
+      "loss": 3.9921,
+      "step": 985600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.385421750920325e-05,
+      "loss": 3.9825,
+      "step": 986112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.384583156169273e-05,
+      "loss": 3.9947,
+      "step": 986624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.383744561418221e-05,
+      "loss": 3.999,
+      "step": 987136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382905966667169e-05,
+      "loss": 3.9814,
+      "step": 987648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.382070647676863e-05,
+      "loss": 3.9918,
+      "step": 988160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.381232052925811e-05,
+      "loss": 3.9992,
+      "step": 988672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.380393458174759e-05,
+      "loss": 3.987,
+      "step": 989184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.379554863423707e-05,
+      "loss": 3.9873,
+      "step": 989696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.378716268672655e-05,
+      "loss": 3.9953,
+      "step": 990208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377877673921603e-05,
+      "loss": 3.988,
+      "step": 990720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.377039079170551e-05,
+      "loss": 3.9919,
+      "step": 991232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.376200484419499e-05,
+      "loss": 4.0006,
+      "step": 991744
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.059726715087891,
+      "eval_runtime": 291.144,
+      "eval_samples_per_second": 1310.661,
+      "eval_steps_per_second": 40.959,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.375361889668448e-05,
+      "loss": 3.9936,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.374523294917396e-05,
+      "loss": 3.9737,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373684700166344e-05,
+      "loss": 3.9877,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372846105415292e-05,
+      "loss": 4.0004,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372007510664239e-05,
+      "loss": 3.994,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.371168915913187e-05,
+      "loss": 3.9923,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370330321162135e-05,
+      "loss": 3.9902,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.369491726411083e-05,
+      "loss": 3.9903,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.368653131660031e-05,
+      "loss": 3.9804,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.367814536908979e-05,
+      "loss": 3.9877,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366975942157927e-05,
+      "loss": 3.9933,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.366137347406875e-05,
+      "loss": 3.9903,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.365300390536196e-05,
+      "loss": 4.0028,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3644617957851446e-05,
+      "loss": 3.9776,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3636232010340926e-05,
+      "loss": 3.9801,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3627846062830406e-05,
+      "loss": 3.9727,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3619460115319886e-05,
+      "loss": 3.9934,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3611074167809366e-05,
+      "loss": 3.9909,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3602688220298846e-05,
+      "loss": 3.9867,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3594302272788326e-05,
+      "loss": 3.9865,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35859163252778e-05,
+      "loss": 4.008,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.357753037776728e-05,
+      "loss": 3.9949,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.356914443025676e-05,
+      "loss": 3.9916,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.356075848274624e-05,
+      "loss": 3.9861,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.355238891403945e-05,
+      "loss": 4.0035,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354400296652893e-05,
+      "loss": 3.9865,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3535633397822144e-05,
+      "loss": 3.9873,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3527247450311624e-05,
+      "loss": 3.9953,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3518861502801104e-05,
+      "loss": 3.9821,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3510475555290584e-05,
+      "loss": 3.986,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3502089607780064e-05,
+      "loss": 3.9878,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3493703660269544e-05,
+      "loss": 3.9833,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348535047036649e-05,
+      "loss": 3.9949,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347696452285597e-05,
+      "loss": 3.9991,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346857857534545e-05,
+      "loss": 3.9868,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346019262783492e-05,
+      "loss": 3.9877,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.34518066803244e-05,
+      "loss": 4.0006,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.344342073281388e-05,
+      "loss": 3.968,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.343503478530336e-05,
+      "loss": 3.982,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.342664883779285e-05,
+      "loss": 3.9836,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341826289028233e-05,
+      "loss": 3.976,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340987694277181e-05,
+      "loss": 3.9879,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340149099526129e-05,
+      "loss": 3.9916,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.339310504775077e-05,
+      "loss": 3.9873,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.338473547904398e-05,
+      "loss": 3.9877,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3376365910337187e-05,
+      "loss": 3.9855,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3367979962826667e-05,
+      "loss": 3.9899,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3359594015316146e-05,
+      "loss": 3.9974,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3351208067805626e-05,
+      "loss": 3.973,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3342822120295106e-05,
+      "loss": 3.9928,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3334452551588315e-05,
+      "loss": 3.9843,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.332608298288153e-05,
+      "loss": 3.9867,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331769703537101e-05,
+      "loss": 3.9888,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330931108786049e-05,
+      "loss": 3.9736,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330092514034997e-05,
+      "loss": 3.983,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.329253919283945e-05,
+      "loss": 3.9713,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.328415324532893e-05,
+      "loss": 4.0038,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.327576729781841e-05,
+      "loss": 3.972,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326738135030789e-05,
+      "loss": 3.9939,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325899540279737e-05,
+      "loss": 3.9786,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325062583409058e-05,
+      "loss": 3.9767,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.324223988658006e-05,
+      "loss": 3.9842,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.323385393906954e-05,
+      "loss": 3.9851,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.322546799155902e-05,
+      "loss": 3.9732,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.32170820440485e-05,
+      "loss": 3.9638,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3208712475341716e-05,
+      "loss": 3.9946,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3200326527831196e-05,
+      "loss": 3.9797,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3191940580320676e-05,
+      "loss": 3.9837,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3183554632810156e-05,
+      "loss": 3.9695,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3175185064103365e-05,
+      "loss": 3.9754,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3166799116592845e-05,
+      "loss": 3.9846,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3158413169082325e-05,
+      "loss": 3.9887,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3150027221571805e-05,
+      "loss": 3.9846,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3141641274061285e-05,
+      "loss": 3.9861,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3133271705354494e-05,
+      "loss": 3.9882,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3124885757843974e-05,
+      "loss": 3.9765,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3116499810333454e-05,
+      "loss": 3.9804,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310811386282294e-05,
+      "loss": 3.9787,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309974429411615e-05,
+      "loss": 3.9822,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309135834660563e-05,
+      "loss": 3.9814,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308297239909511e-05,
+      "loss": 3.9743,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307458645158458e-05,
+      "loss": 3.9906,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30662168828778e-05,
+      "loss": 3.9917,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305783093536728e-05,
+      "loss": 3.9798,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304944498785676e-05,
+      "loss": 3.964,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304105904034623e-05,
+      "loss": 3.9848,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303267309283571e-05,
+      "loss": 3.9867,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.302428714532519e-05,
+      "loss": 3.9887,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301590119781468e-05,
+      "loss": 3.9767,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300753162910789e-05,
+      "loss": 3.9817,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299914568159737e-05,
+      "loss": 3.9876,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299075973408685e-05,
+      "loss": 3.9749,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.298237378657633e-05,
+      "loss": 3.978,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.297398783906581e-05,
+      "loss": 3.9833,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296560189155529e-05,
+      "loss": 3.9785,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295721594404477e-05,
+      "loss": 3.9932,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294882999653425e-05,
+      "loss": 3.9809,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2940476806631185e-05,
+      "loss": 3.9755,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2932090859120665e-05,
+      "loss": 3.9762,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2923704911610145e-05,
+      "loss": 3.9836,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.291531896409963e-05,
+      "loss": 3.986,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.290693301658911e-05,
+      "loss": 3.9747,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.289856344788232e-05,
+      "loss": 3.977,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.28901775003718e-05,
+      "loss": 3.9693,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.288179155286128e-05,
+      "loss": 3.9843,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.287340560535076e-05,
+      "loss": 3.9604,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.286501965784024e-05,
+      "loss": 3.973,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.285663371032972e-05,
+      "loss": 3.9778,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.28482477628192e-05,
+      "loss": 3.982,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283986181530868e-05,
+      "loss": 3.9702,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283147586779816e-05,
+      "loss": 3.9795,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282310629909137e-05,
+      "loss": 3.9722,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2814736730384585e-05,
+      "loss": 3.9817,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2806350782874065e-05,
+      "loss": 3.9804,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2797964835363545e-05,
+      "loss": 3.9733,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2789578887853025e-05,
+      "loss": 3.9775,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2781192940342505e-05,
+      "loss": 3.9931,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2772806992831985e-05,
+      "loss": 3.9758,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2764437424125194e-05,
+      "loss": 3.9764,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2756051476614674e-05,
+      "loss": 3.978,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2747665529104154e-05,
+      "loss": 3.9797,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2739279581593634e-05,
+      "loss": 3.9788,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2730893634083114e-05,
+      "loss": 3.9782,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2722507686572594e-05,
+      "loss": 3.9751,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.271412173906207e-05,
+      "loss": 3.9829,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.270573579155155e-05,
+      "loss": 3.9697,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269736622284476e-05,
+      "loss": 3.9736,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268899665413798e-05,
+      "loss": 3.9698,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268061070662746e-05,
+      "loss": 3.9748,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.267222475911694e-05,
+      "loss": 3.9681,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266383881160642e-05,
+      "loss": 3.9889,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.265545286409589e-05,
+      "loss": 3.9882,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264706691658537e-05,
+      "loss": 3.9691,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263868096907485e-05,
+      "loss": 3.9793,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263031140036807e-05,
+      "loss": 3.9791,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262192545285754e-05,
+      "loss": 3.982,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261353950534702e-05,
+      "loss": 3.9729,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.26051535578365e-05,
+      "loss": 3.974,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259678398912972e-05,
+      "loss": 3.9816,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.25883980416192e-05,
+      "loss": 3.984,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258001209410868e-05,
+      "loss": 3.9688,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257162614659816e-05,
+      "loss": 3.9785,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2563240199087637e-05,
+      "loss": 3.9864,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2554854251577117e-05,
+      "loss": 3.9754,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2546468304066597e-05,
+      "loss": 3.9727,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2538098735359806e-05,
+      "loss": 3.9824,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2529712787849286e-05,
+      "loss": 3.9738,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2521326840338766e-05,
+      "loss": 3.9824,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2512940892828245e-05,
+      "loss": 3.9824,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0520243644714355,
+      "eval_runtime": 293.9171,
+      "eval_samples_per_second": 1298.295,
+      "eval_steps_per_second": 40.573,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2504554945317725e-05,
+      "loss": 3.9833,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2496168997807205e-05,
+      "loss": 3.9623,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2487783050296685e-05,
+      "loss": 3.9724,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247939710278617e-05,
+      "loss": 3.9898,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247101115527565e-05,
+      "loss": 3.9794,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.246262520776513e-05,
+      "loss": 3.9791,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2454239260254605e-05,
+      "loss": 3.975,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2445853312744085e-05,
+      "loss": 3.9749,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2437467365233565e-05,
+      "loss": 3.9673,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2429081417723045e-05,
+      "loss": 3.9766,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2420695470212525e-05,
+      "loss": 3.9729,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2412309522702005e-05,
+      "loss": 3.9838,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2403939953995214e-05,
+      "loss": 3.9852,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2395554006484694e-05,
+      "loss": 3.9658,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2387168058974174e-05,
+      "loss": 3.9689,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2378782111463654e-05,
+      "loss": 3.9572,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2370396163953134e-05,
+      "loss": 3.9784,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236201021644262e-05,
+      "loss": 3.9767,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235364064773583e-05,
+      "loss": 3.9737,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234525470022531e-05,
+      "loss": 3.9699,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.233686875271479e-05,
+      "loss": 3.9977,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232848280520427e-05,
+      "loss": 3.9841,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232009685769375e-05,
+      "loss": 3.9761,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.231171091018323e-05,
+      "loss": 3.9744,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230332496267271e-05,
+      "loss": 3.9906,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.229493901516219e-05,
+      "loss": 3.9752,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.228655306765167e-05,
+      "loss": 3.9721,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227816712014114e-05,
+      "loss": 3.9856,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226978117263062e-05,
+      "loss": 3.9668,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.22613952251201e-05,
+      "loss": 3.9704,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225300927760959e-05,
+      "loss": 3.9741,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.22446397089028e-05,
+      "loss": 3.9704,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223625376139228e-05,
+      "loss": 3.982,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2227884192685494e-05,
+      "loss": 3.9855,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221949824517497e-05,
+      "loss": 3.9723,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221112867646818e-05,
+      "loss": 3.9813,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220274272895766e-05,
+      "loss": 3.983,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219435678144714e-05,
+      "loss": 3.9594,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2185970833936616e-05,
+      "loss": 3.9682,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2177584886426096e-05,
+      "loss": 3.9748,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2169198938915576e-05,
+      "loss": 3.96,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2160812991405056e-05,
+      "loss": 3.9773,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215242704389454e-05,
+      "loss": 3.9778,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214404109638402e-05,
+      "loss": 3.9717,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.213567152767723e-05,
+      "loss": 3.9752,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212728558016671e-05,
+      "loss": 3.9763,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211889963265619e-05,
+      "loss": 3.9702,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.211051368514567e-05,
+      "loss": 3.9865,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.210214411643888e-05,
+      "loss": 3.9593,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.209375816892836e-05,
+      "loss": 3.9759,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208537222141784e-05,
+      "loss": 3.9768,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207698627390732e-05,
+      "loss": 3.9732,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20686003263968e-05,
+      "loss": 3.9756,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206021437888628e-05,
+      "loss": 3.9593,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205182843137576e-05,
+      "loss": 3.9738,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2043458862668976e-05,
+      "loss": 3.9548,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2035072915158456e-05,
+      "loss": 3.9909,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2026686967647936e-05,
+      "loss": 3.9616,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2018301020137416e-05,
+      "loss": 3.9791,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2009915072626896e-05,
+      "loss": 3.9709,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2001529125116376e-05,
+      "loss": 3.9651,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1993143177605856e-05,
+      "loss": 3.9693,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198475723009533e-05,
+      "loss": 3.9727,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1976387661388545e-05,
+      "loss": 3.9555,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1968001713878025e-05,
+      "loss": 3.9536,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1959615766367505e-05,
+      "loss": 3.9824,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.195122981885698e-05,
+      "loss": 3.9706,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1942843871346465e-05,
+      "loss": 3.9676,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.193447430263968e-05,
+      "loss": 3.954,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1926088355129154e-05,
+      "loss": 3.9651,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191771878642237e-05,
+      "loss": 3.9723,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190933283891185e-05,
+      "loss": 3.9734,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190094689140133e-05,
+      "loss": 3.9753,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18925609438908e-05,
+      "loss": 3.9735,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188417499638028e-05,
+      "loss": 3.9737,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.187578904886976e-05,
+      "loss": 3.9665,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186740310135924e-05,
+      "loss": 3.9658,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185901715384872e-05,
+      "loss": 3.9651,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185064758514193e-05,
+      "loss": 3.9719,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.184226163763142e-05,
+      "loss": 3.9688,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18338756901209e-05,
+      "loss": 3.963,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.182550612141411e-05,
+      "loss": 3.9739,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181712017390359e-05,
+      "loss": 3.9772,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180873422639307e-05,
+      "loss": 3.9688,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180034827888255e-05,
+      "loss": 3.9575,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179196233137203e-05,
+      "loss": 3.9645,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.178357638386151e-05,
+      "loss": 3.9724,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177519043635099e-05,
+      "loss": 3.9787,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.176680448884047e-05,
+      "loss": 3.9661,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1758434920133677e-05,
+      "loss": 3.9683,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1750048972623157e-05,
+      "loss": 3.9731,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1741679403916366e-05,
+      "loss": 3.9674,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173329345640585e-05,
+      "loss": 3.9632,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.172490750889533e-05,
+      "loss": 3.9693,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.171652156138481e-05,
+      "loss": 3.9682,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170813561387429e-05,
+      "loss": 3.9781,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169974966636377e-05,
+      "loss": 3.9722,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169136371885325e-05,
+      "loss": 3.9602,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.168299415014646e-05,
+      "loss": 3.9651,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.167460820263594e-05,
+      "loss": 3.9687,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.166622225512542e-05,
+      "loss": 3.9746,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16578363076149e-05,
+      "loss": 3.9629,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164946673890811e-05,
+      "loss": 3.9648,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164108079139759e-05,
+      "loss": 3.959,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163269484388707e-05,
+      "loss": 3.9697,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.162430889637655e-05,
+      "loss": 3.9483,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161592294886604e-05,
+      "loss": 3.9623,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160753700135552e-05,
+      "loss": 3.9668,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159915105384499e-05,
+      "loss": 3.9672,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159076510633447e-05,
+      "loss": 3.9578,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158237915882395e-05,
+      "loss": 3.9648,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157399321131343e-05,
+      "loss": 3.96,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156562364260664e-05,
+      "loss": 3.9659,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155723769509612e-05,
+      "loss": 3.9711,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15488517475856e-05,
+      "loss": 3.9606,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154046580007508e-05,
+      "loss": 3.9683,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153207985256456e-05,
+      "loss": 3.9773,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152369390505404e-05,
+      "loss": 3.9696,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1515324336347255e-05,
+      "loss": 3.959,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1506938388836734e-05,
+      "loss": 3.9706,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1498552441326214e-05,
+      "loss": 3.9682,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1490166493815694e-05,
+      "loss": 3.9639,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1481780546305174e-05,
+      "loss": 3.9689,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1473394598794654e-05,
+      "loss": 3.9633,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1465008651284134e-05,
+      "loss": 3.9692,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1456622703773614e-05,
+      "loss": 3.9587,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144825313506682e-05,
+      "loss": 3.9631,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14398671875563e-05,
+      "loss": 3.9559,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143149761884951e-05,
+      "loss": 3.9606,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142311167133899e-05,
+      "loss": 3.9541,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.141472572382847e-05,
+      "loss": 3.9749,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140633977631796e-05,
+      "loss": 3.9772,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.139795382880744e-05,
+      "loss": 3.9601,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138956788129692e-05,
+      "loss": 3.9649,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13811819337864e-05,
+      "loss": 3.9712,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.137279598627588e-05,
+      "loss": 3.9673,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.136442641756909e-05,
+      "loss": 3.9628,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135604047005857e-05,
+      "loss": 3.9599,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134765452254805e-05,
+      "loss": 3.9709,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133926857503753e-05,
+      "loss": 3.9712,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133089900633074e-05,
+      "loss": 3.96,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.132251305882022e-05,
+      "loss": 3.9642,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1314143490113426e-05,
+      "loss": 3.9762,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1305757542602906e-05,
+      "loss": 3.9606,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.129737159509239e-05,
+      "loss": 3.9619,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128898564758187e-05,
+      "loss": 3.972,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128061607887508e-05,
+      "loss": 3.9616,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.127223013136456e-05,
+      "loss": 3.97,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.126386056265777e-05,
+      "loss": 3.9684,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.045748710632324,
+      "eval_runtime": 295.2012,
+      "eval_samples_per_second": 1292.647,
+      "eval_steps_per_second": 40.396,
+      "step": 1144800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.125547461514725e-05,
+      "loss": 3.9407,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124708866763673e-05,
+      "loss": 3.9542,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123870272012621e-05,
+      "loss": 3.96,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123031677261569e-05,
+      "loss": 3.9759,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122193082510517e-05,
+      "loss": 3.9684,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.121354487759465e-05,
+      "loss": 3.9733,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120515893008413e-05,
+      "loss": 3.9592,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119677298257361e-05,
+      "loss": 3.9681,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118838703506309e-05,
+      "loss": 3.9525,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118000108755257e-05,
+      "loss": 3.966,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.117161514004205e-05,
+      "loss": 3.9617,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116322919253153e-05,
+      "loss": 3.9728,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.115485962382474e-05,
+      "loss": 3.9765,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.114647367631422e-05,
+      "loss": 3.9539,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11380877288037e-05,
+      "loss": 3.957,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112970178129318e-05,
+      "loss": 3.949,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112131583378266e-05,
+      "loss": 3.9591,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.111292988627214e-05,
+      "loss": 3.9657,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.110454393876162e-05,
+      "loss": 3.963,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.10961579912511e-05,
+      "loss": 3.9581,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108777204374058e-05,
+      "loss": 3.9887,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107938609623006e-05,
+      "loss": 3.9702,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1071016527523275e-05,
+      "loss": 3.9681,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1062630580012755e-05,
+      "loss": 3.9621,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1054244632502235e-05,
+      "loss": 3.9784,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1045858684991715e-05,
+      "loss": 3.9632,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103747273748119e-05,
+      "loss": 3.9629,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.102908678997067e-05,
+      "loss": 3.9719,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1020717221263884e-05,
+      "loss": 3.953,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.101233127375336e-05,
+      "loss": 3.9645,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.100394532624284e-05,
+      "loss": 3.9582,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.099555937873232e-05,
+      "loss": 3.9636,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09871734312218e-05,
+      "loss": 3.9681,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0978787483711283e-05,
+      "loss": 3.9733,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097041791500449e-05,
+      "loss": 3.9624,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.096203196749397e-05,
+      "loss": 3.9722,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.095364601998345e-05,
+      "loss": 3.9684,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.094526007247293e-05,
+      "loss": 3.9453,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.093689050376614e-05,
+      "loss": 3.9606,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092850455625562e-05,
+      "loss": 3.9592,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09201186087451e-05,
+      "loss": 3.9467,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.091173266123458e-05,
+      "loss": 3.9655,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.090334671372406e-05,
+      "loss": 3.9662,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.089496076621354e-05,
+      "loss": 3.9634,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.088657481870302e-05,
+      "loss": 3.957,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.08781888711925e-05,
+      "loss": 3.9708,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086980292368198e-05,
+      "loss": 3.9601,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086141697617146e-05,
+      "loss": 3.9774,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085303102866095e-05,
+      "loss": 3.9451,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.084464508115043e-05,
+      "loss": 3.9613,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.08362591336399e-05,
+      "loss": 3.9674,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0827905943736846e-05,
+      "loss": 3.9584,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0819519996226326e-05,
+      "loss": 3.9667,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0811134048715806e-05,
+      "loss": 3.9468,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0802748101205286e-05,
+      "loss": 3.9637,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0794362153694766e-05,
+      "loss": 3.9454,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0785992584987975e-05,
+      "loss": 3.9758,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0777606637477455e-05,
+      "loss": 3.953,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0769220689966935e-05,
+      "loss": 3.9651,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.076085112126015e-05,
+      "loss": 3.9623,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.075246517374963e-05,
+      "loss": 3.9544,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074407922623911e-05,
+      "loss": 3.9579,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.073569327872859e-05,
+      "loss": 3.9629,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072730733121807e-05,
+      "loss": 3.9459,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0718921383707544e-05,
+      "loss": 3.9438,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0710535436197024e-05,
+      "loss": 3.9669,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0702149488686504e-05,
+      "loss": 3.9594,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0693763541175984e-05,
+      "loss": 3.9585,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0685377593665463e-05,
+      "loss": 3.9433,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0676991646154943e-05,
+      "loss": 3.9513,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066860569864442e-05,
+      "loss": 3.9597,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066023612993764e-05,
+      "loss": 3.9624,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.065186656123085e-05,
+      "loss": 3.9662,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.064348061372033e-05,
+      "loss": 3.9594,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0635111045013544e-05,
+      "loss": 3.9646,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.062672509750302e-05,
+      "loss": 3.9552,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.06183391499925e-05,
+      "loss": 3.9526,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.060995320248198e-05,
+      "loss": 3.9546,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.060156725497146e-05,
+      "loss": 3.9573,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.059318130746094e-05,
+      "loss": 3.9595,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.058479535995042e-05,
+      "loss": 3.9545,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05764094124399e-05,
+      "loss": 3.9569,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.056802346492938e-05,
+      "loss": 3.9707,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055965389622259e-05,
+      "loss": 3.9558,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055126794871207e-05,
+      "loss": 3.9467,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.054288200120155e-05,
+      "loss": 3.9535,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.053449605369103e-05,
+      "loss": 3.9606,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052611010618051e-05,
+      "loss": 3.9667,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051772415866999e-05,
+      "loss": 3.9564,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0509354589963202e-05,
+      "loss": 3.9549,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0500968642452682e-05,
+      "loss": 3.9635,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0492582694942162e-05,
+      "loss": 3.9556,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0484213126235374e-05,
+      "loss": 3.9511,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0475827178724854e-05,
+      "loss": 3.957,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0467441231214334e-05,
+      "loss": 3.9571,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0459055283703814e-05,
+      "loss": 3.9669,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0450669336193294e-05,
+      "loss": 3.9602,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0442283388682774e-05,
+      "loss": 3.9486,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0433897441172254e-05,
+      "loss": 3.9508,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0425511493661734e-05,
+      "loss": 3.9568,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.041712554615121e-05,
+      "loss": 3.9651,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040873959864069e-05,
+      "loss": 3.9514,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0400370029933906e-05,
+      "loss": 3.9517,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.039198408242338e-05,
+      "loss": 3.9483,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0383614513716595e-05,
+      "loss": 3.9567,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0375228566206075e-05,
+      "loss": 3.9393,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0366842618695555e-05,
+      "loss": 3.9526,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0358456671185032e-05,
+      "loss": 3.9493,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0350070723674512e-05,
+      "loss": 3.9574,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0341684776163992e-05,
+      "loss": 3.9486,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.033329882865347e-05,
+      "loss": 3.9564,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.032491288114295e-05,
+      "loss": 3.9455,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.031652693363243e-05,
+      "loss": 3.9547,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0308140986121915e-05,
+      "loss": 3.9593,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0299755038611395e-05,
+      "loss": 3.9469,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0291369091100875e-05,
+      "loss": 3.9586,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0282999522394084e-05,
+      "loss": 3.9642,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0274629953687293e-05,
+      "loss": 3.9593,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0266244006176776e-05,
+      "loss": 3.9496,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0257858058666256e-05,
+      "loss": 3.9536,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0249472111155736e-05,
+      "loss": 3.9579,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0241086163645216e-05,
+      "loss": 3.9511,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0232700216134696e-05,
+      "loss": 3.961,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0224314268624176e-05,
+      "loss": 3.9483,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0215928321113656e-05,
+      "loss": 3.9607,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.020755875240687e-05,
+      "loss": 3.9457,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.019917280489635e-05,
+      "loss": 3.951,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.019078685738583e-05,
+      "loss": 3.9451,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0182417288679038e-05,
+      "loss": 3.9489,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0174047719972247e-05,
+      "loss": 3.9411,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.016566177246173e-05,
+      "loss": 3.9621,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.015727582495121e-05,
+      "loss": 3.9692,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.014888987744069e-05,
+      "loss": 3.9516,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.014050392993017e-05,
+      "loss": 3.9471,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.013211798241965e-05,
+      "loss": 3.9599,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.012373203490913e-05,
+      "loss": 3.9599,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.011534608739861e-05,
+      "loss": 3.954,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.010699289749555e-05,
+      "loss": 3.9462,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.009860694998503e-05,
+      "loss": 3.9611,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.009022100247451e-05,
+      "loss": 3.9557,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.008183505496399e-05,
+      "loss": 3.9517,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.007344910745347e-05,
+      "loss": 3.95,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.006506315994295e-05,
+      "loss": 3.9617,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.005667721243243e-05,
+      "loss": 3.9503,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0048291264921914e-05,
+      "loss": 3.9503,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0039905317411394e-05,
+      "loss": 3.9608,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0031519369900868e-05,
+      "loss": 3.951,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0023133422390348e-05,
+      "loss": 3.9555,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0014747474879828e-05,
+      "loss": 3.9587,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.000637790617304e-05,
+      "loss": 3.9512,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0393195152282715,
+      "eval_runtime": 313.451,
+      "eval_samples_per_second": 1217.386,
+      "eval_steps_per_second": 38.044,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9998008337466256e-05,
+      "loss": 3.9425,
+      "step": 1221632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9989622389955736e-05,
+      "loss": 3.9455,
+      "step": 1222144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9981236442445216e-05,
+      "loss": 3.9624,
+      "step": 1222656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.997285049493469e-05,
+      "loss": 3.9586,
+      "step": 1223168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.996446454742417e-05,
+      "loss": 3.96,
+      "step": 1223680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9956078599913652e-05,
+      "loss": 3.9461,
+      "step": 1224192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9947709031206865e-05,
+      "loss": 3.957,
+      "step": 1224704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.993932308369634e-05,
+      "loss": 3.9381,
+      "step": 1225216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.993093713618582e-05,
+      "loss": 3.9592,
+      "step": 1225728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.99225511886753e-05,
+      "loss": 3.9458,
+      "step": 1226240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.991416524116478e-05,
+      "loss": 3.9645,
+      "step": 1226752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9905795672457994e-05,
+      "loss": 3.9641,
+      "step": 1227264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.989742610375121e-05,
+      "loss": 3.945,
+      "step": 1227776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.988904015624069e-05,
+      "loss": 3.9452,
+      "step": 1228288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9880654208730163e-05,
+      "loss": 3.9385,
+      "step": 1228800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9872268261219643e-05,
+      "loss": 3.9435,
+      "step": 1229312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9863882313709123e-05,
+      "loss": 3.9568,
+      "step": 1229824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9855496366198603e-05,
+      "loss": 3.9511,
+      "step": 1230336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9847110418688086e-05,
+      "loss": 3.9501,
+      "step": 1230848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9838724471177566e-05,
+      "loss": 3.9758,
+      "step": 1231360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9830354902470775e-05,
+      "loss": 3.9587,
+      "step": 1231872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9821968954960255e-05,
+      "loss": 3.9588,
+      "step": 1232384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9813583007449735e-05,
+      "loss": 3.9482,
+      "step": 1232896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9805197059939215e-05,
+      "loss": 3.9643,
+      "step": 1233408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9796811112428695e-05,
+      "loss": 3.9477,
+      "step": 1233920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9788441543721907e-05,
+      "loss": 3.9566,
+      "step": 1234432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9780055596211387e-05,
+      "loss": 3.9555,
+      "step": 1234944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9771669648700867e-05,
+      "loss": 3.946,
+      "step": 1235456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9763283701190347e-05,
+      "loss": 3.9497,
+      "step": 1235968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9754897753679827e-05,
+      "loss": 3.9481,
+      "step": 1236480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9746511806169307e-05,
+      "loss": 3.9536,
+      "step": 1236992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9738125858658787e-05,
+      "loss": 3.9542,
+      "step": 1237504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9729756289952e-05,
+      "loss": 3.9603,
+      "step": 1238016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.972137034244148e-05,
+      "loss": 3.9543,
+      "step": 1238528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.971298439493096e-05,
+      "loss": 3.958,
+      "step": 1239040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.970459844742044e-05,
+      "loss": 3.9547,
+      "step": 1239552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.969621249990992e-05,
+      "loss": 3.938,
+      "step": 1240064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9687842931203132e-05,
+      "loss": 3.9496,
+      "step": 1240576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.967947336249634e-05,
+      "loss": 3.9452,
+      "step": 1241088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.967108741498582e-05,
+      "loss": 3.9374,
+      "step": 1241600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.96627014674753e-05,
+      "loss": 3.9527,
+      "step": 1242112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.965431551996478e-05,
+      "loss": 3.9571,
+      "step": 1242624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.964592957245426e-05,
+      "loss": 3.9483,
+      "step": 1243136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9637560003747473e-05,
+      "loss": 3.9483,
+      "step": 1243648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9629174056236953e-05,
+      "loss": 3.9565,
+      "step": 1244160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9620788108726433e-05,
+      "loss": 3.9537,
+      "step": 1244672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9612402161215913e-05,
+      "loss": 3.9587,
+      "step": 1245184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9604016213705393e-05,
+      "loss": 3.9414,
+      "step": 1245696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9595630266194873e-05,
+      "loss": 3.9466,
+      "step": 1246208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.958724431868435e-05,
+      "loss": 3.957,
+      "step": 1246720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.957885837117383e-05,
+      "loss": 3.9478,
+      "step": 1247232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.957047242366331e-05,
+      "loss": 3.9517,
+      "step": 1247744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.956208647615279e-05,
+      "loss": 3.9391,
+      "step": 1248256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.955370052864227e-05,
+      "loss": 3.9521,
+      "step": 1248768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.954531458113175e-05,
+      "loss": 3.9318,
+      "step": 1249280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9536945012424962e-05,
+      "loss": 3.9602,
+      "step": 1249792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9528559064914442e-05,
+      "loss": 3.9438,
+      "step": 1250304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9520173117403922e-05,
+      "loss": 3.9518,
+      "step": 1250816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.95117871698934e-05,
+      "loss": 3.9472,
+      "step": 1251328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.950341760118661e-05,
+      "loss": 3.946,
+      "step": 1251840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.949503165367609e-05,
+      "loss": 3.9438,
+      "step": 1252352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.948664570616557e-05,
+      "loss": 3.9524,
+      "step": 1252864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9478259758655054e-05,
+      "loss": 3.9352,
+      "step": 1253376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9469873811144534e-05,
+      "loss": 3.9328,
+      "step": 1253888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9461487863634014e-05,
+      "loss": 3.9528,
+      "step": 1254400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9453101916123494e-05,
+      "loss": 3.9515,
+      "step": 1254912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9444715968612974e-05,
+      "loss": 3.9448,
+      "step": 1255424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9436346399906183e-05,
+      "loss": 3.9325,
+      "step": 1255936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9427960452395663e-05,
+      "loss": 3.9396,
+      "step": 1256448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9419590883688875e-05,
+      "loss": 3.9483,
+      "step": 1256960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9411204936178355e-05,
+      "loss": 3.9517,
+      "step": 1257472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9402818988667835e-05,
+      "loss": 3.9502,
+      "step": 1257984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9394433041157315e-05,
+      "loss": 3.9497,
+      "step": 1258496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9386063472450524e-05,
+      "loss": 3.9527,
+      "step": 1259008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9377677524940008e-05,
+      "loss": 3.9497,
+      "step": 1259520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9369291577429488e-05,
+      "loss": 3.9351,
+      "step": 1260032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9360905629918968e-05,
+      "loss": 3.9427,
+      "step": 1260544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9352519682408448e-05,
+      "loss": 3.9513,
+      "step": 1261056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9344133734897928e-05,
+      "loss": 3.944,
+      "step": 1261568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9335764166191137e-05,
+      "loss": 3.9434,
+      "step": 1262080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9327378218680617e-05,
+      "loss": 3.9443,
+      "step": 1262592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9318992271170097e-05,
+      "loss": 3.9622,
+      "step": 1263104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.931060632365958e-05,
+      "loss": 3.943,
+      "step": 1263616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.930223675495279e-05,
+      "loss": 3.9348,
+      "step": 1264128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.929385080744227e-05,
+      "loss": 3.9442,
+      "step": 1264640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.928546485993175e-05,
+      "loss": 3.9486,
+      "step": 1265152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.927707891242123e-05,
+      "loss": 3.9552,
+      "step": 1265664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.926869296491071e-05,
+      "loss": 3.9419,
+      "step": 1266176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.926032339620392e-05,
+      "loss": 3.9464,
+      "step": 1266688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.92519374486934e-05,
+      "loss": 3.9476,
+      "step": 1267200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.924355150118288e-05,
+      "loss": 3.9466,
+      "step": 1267712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.923516555367236e-05,
+      "loss": 3.9369,
+      "step": 1268224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9226779606161834e-05,
+      "loss": 3.9438,
+      "step": 1268736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9218393658651318e-05,
+      "loss": 3.9503,
+      "step": 1269248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9210024089944534e-05,
+      "loss": 3.9602,
+      "step": 1269760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9201638142434007e-05,
+      "loss": 3.9432,
+      "step": 1270272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9193252194923487e-05,
+      "loss": 3.9402,
+      "step": 1270784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9184866247412967e-05,
+      "loss": 3.9435,
+      "step": 1271296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9176480299902447e-05,
+      "loss": 3.9435,
+      "step": 1271808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9168094352391927e-05,
+      "loss": 3.9544,
+      "step": 1272320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.915970840488141e-05,
+      "loss": 3.9417,
+      "step": 1272832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.915132245737089e-05,
+      "loss": 3.9412,
+      "step": 1273344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.91429528886641e-05,
+      "loss": 3.936,
+      "step": 1273856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.913456694115358e-05,
+      "loss": 3.9499,
+      "step": 1274368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.912618099364306e-05,
+      "loss": 3.9224,
+      "step": 1274880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.911781142493627e-05,
+      "loss": 3.9438,
+      "step": 1275392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.910942547742575e-05,
+      "loss": 3.9383,
+      "step": 1275904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.910103952991523e-05,
+      "loss": 3.9477,
+      "step": 1276416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.909265358240471e-05,
+      "loss": 3.9377,
+      "step": 1276928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.908426763489419e-05,
+      "loss": 3.9468,
+      "step": 1277440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.90758980661874e-05,
+      "loss": 3.9389,
+      "step": 1277952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.906751211867688e-05,
+      "loss": 3.9405,
+      "step": 1278464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9059126171166364e-05,
+      "loss": 3.9458,
+      "step": 1278976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9050740223655843e-05,
+      "loss": 3.935,
+      "step": 1279488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9042354276145323e-05,
+      "loss": 3.95,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9033968328634803e-05,
+      "loss": 3.9575,
+      "step": 1280512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9025598759928013e-05,
+      "loss": 3.9458,
+      "step": 1281024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9017212812417492e-05,
+      "loss": 3.9406,
+      "step": 1281536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9008826864906972e-05,
+      "loss": 3.9395,
+      "step": 1282048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9000440917396456e-05,
+      "loss": 3.9453,
+      "step": 1282560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8992054969885936e-05,
+      "loss": 3.9439,
+      "step": 1283072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8983685401179145e-05,
+      "loss": 3.9496,
+      "step": 1283584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8975299453668625e-05,
+      "loss": 3.941,
+      "step": 1284096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8966913506158105e-05,
+      "loss": 3.9486,
+      "step": 1284608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8958527558647585e-05,
+      "loss": 3.9375,
+      "step": 1285120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8950141611137065e-05,
+      "loss": 3.9408,
+      "step": 1285632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8941755663626548e-05,
+      "loss": 3.9338,
+      "step": 1286144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8933386094919757e-05,
+      "loss": 3.9361,
+      "step": 1286656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8925000147409237e-05,
+      "loss": 3.9305,
+      "step": 1287168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8916614199898717e-05,
+      "loss": 3.953,
+      "step": 1287680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8908228252388194e-05,
+      "loss": 3.958,
+      "step": 1288192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8899842304877674e-05,
+      "loss": 3.9484,
+      "step": 1288704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8891456357367153e-05,
+      "loss": 3.9295,
+      "step": 1289216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.88831031674641e-05,
+      "loss": 3.9511,
+      "step": 1289728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.887471721995358e-05,
+      "loss": 3.948,
+      "step": 1290240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.886633127244306e-05,
+      "loss": 3.9435,
+      "step": 1290752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.885794532493254e-05,
+      "loss": 3.9399,
+      "step": 1291264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8849559377422018e-05,
+      "loss": 3.9501,
+      "step": 1291776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8841173429911495e-05,
+      "loss": 3.9456,
+      "step": 1292288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8832787482400975e-05,
+      "loss": 3.9393,
+      "step": 1292800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8824401534890455e-05,
+      "loss": 3.9432,
+      "step": 1293312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8816015587379935e-05,
+      "loss": 3.9479,
+      "step": 1293824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8807629639869415e-05,
+      "loss": 3.9385,
+      "step": 1294336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8799260071162627e-05,
+      "loss": 3.9391,
+      "step": 1294848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8790874123652107e-05,
+      "loss": 3.9514,
+      "step": 1295360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8782488176141587e-05,
+      "loss": 3.9416,
+      "step": 1295872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8774102228631067e-05,
+      "loss": 3.9429,
+      "step": 1296384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8765716281120547e-05,
+      "loss": 3.953,
+      "step": 1296896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8757330333610027e-05,
+      "loss": 3.9369,
+      "step": 1297408
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.033699035644531,
+      "eval_runtime": 312.5228,
+      "eval_samples_per_second": 1221.002,
+      "eval_steps_per_second": 38.157,
+      "step": 1297440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8748944386099507e-05,
+      "loss": 3.9314,
+      "step": 1297920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8740558438588987e-05,
+      "loss": 3.9382,
+      "step": 1298432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8732172491078467e-05,
+      "loss": 3.9523,
+      "step": 1298944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.872378654356795e-05,
+      "loss": 3.9464,
+      "step": 1299456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.871540059605743e-05,
+      "loss": 3.9515,
+      "step": 1299968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.870701464854691e-05,
+      "loss": 3.9341,
+      "step": 1300480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8698628701036383e-05,
+      "loss": 3.9461,
+      "step": 1300992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8690242753525863e-05,
+      "loss": 3.9277,
+      "step": 1301504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8681856806015343e-05,
+      "loss": 3.9499,
+      "step": 1302016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8673470858504827e-05,
+      "loss": 3.9365,
+      "step": 1302528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8665084910994306e-05,
+      "loss": 3.9515,
+      "step": 1303040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8656698963483786e-05,
+      "loss": 3.9499,
+      "step": 1303552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8648329394776996e-05,
+      "loss": 3.9401,
+      "step": 1304064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8639943447266475e-05,
+      "loss": 3.9345,
+      "step": 1304576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8631557499755955e-05,
+      "loss": 3.9323,
+      "step": 1305088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8623171552245435e-05,
+      "loss": 3.9346,
+      "step": 1305600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.861478560473492e-05,
+      "loss": 3.9443,
+      "step": 1306112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.86063996572244e-05,
+      "loss": 3.9422,
+      "step": 1306624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.859801370971388e-05,
+      "loss": 3.939,
+      "step": 1307136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.858962776220336e-05,
+      "loss": 3.9633,
+      "step": 1307648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.858124181469284e-05,
+      "loss": 3.9514,
+      "step": 1308160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.857285586718232e-05,
+      "loss": 3.9458,
+      "step": 1308672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.85644699196718e-05,
+      "loss": 3.9401,
+      "step": 1309184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.855610035096501e-05,
+      "loss": 3.9576,
+      "step": 1309696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.854771440345449e-05,
+      "loss": 3.9354,
+      "step": 1310208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.853932845594397e-05,
+      "loss": 3.9476,
+      "step": 1310720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8530942508433444e-05,
+      "loss": 3.9456,
+      "step": 1311232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8522556560922924e-05,
+      "loss": 3.9414,
+      "step": 1311744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8514170613412404e-05,
+      "loss": 3.9349,
+      "step": 1312256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8505784665901887e-05,
+      "loss": 3.9375,
+      "step": 1312768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8497398718391367e-05,
+      "loss": 3.9428,
+      "step": 1313280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8489029149684576e-05,
+      "loss": 3.9447,
+      "step": 1313792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8480643202174056e-05,
+      "loss": 3.9473,
+      "step": 1314304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8472273633467272e-05,
+      "loss": 3.949,
+      "step": 1314816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.846388768595675e-05,
+      "loss": 3.9482,
+      "step": 1315328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.845550173844623e-05,
+      "loss": 3.9443,
+      "step": 1315840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.844711579093571e-05,
+      "loss": 3.9303,
+      "step": 1316352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.843872984342519e-05,
+      "loss": 3.9417,
+      "step": 1316864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8430360274718398e-05,
+      "loss": 3.9348,
+      "step": 1317376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8421974327207878e-05,
+      "loss": 3.9296,
+      "step": 1317888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8413588379697358e-05,
+      "loss": 3.9435,
+      "step": 1318400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.840520243218684e-05,
+      "loss": 3.947,
+      "step": 1318912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.839681648467632e-05,
+      "loss": 3.9372,
+      "step": 1319424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.83884305371658e-05,
+      "loss": 3.9383,
+      "step": 1319936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.838004458965528e-05,
+      "loss": 3.9432,
+      "step": 1320448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.837165864214476e-05,
+      "loss": 3.944,
+      "step": 1320960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.836328907343797e-05,
+      "loss": 3.9491,
+      "step": 1321472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.835490312592745e-05,
+      "loss": 3.9356,
+      "step": 1321984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8346533557220662e-05,
+      "loss": 3.9359,
+      "step": 1322496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8338147609710142e-05,
+      "loss": 3.9454,
+      "step": 1323008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8329761662199622e-05,
+      "loss": 3.9375,
+      "step": 1323520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8321375714689102e-05,
+      "loss": 3.9392,
+      "step": 1324032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8312989767178582e-05,
+      "loss": 3.9319,
+      "step": 1324544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8304603819668062e-05,
+      "loss": 3.944,
+      "step": 1325056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8296217872157542e-05,
+      "loss": 3.9226,
+      "step": 1325568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8287848303450755e-05,
+      "loss": 3.9475,
+      "step": 1326080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8279462355940234e-05,
+      "loss": 3.9375,
+      "step": 1326592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8271076408429714e-05,
+      "loss": 3.9391,
+      "step": 1327104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8262690460919194e-05,
+      "loss": 3.9438,
+      "step": 1327616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8254304513408674e-05,
+      "loss": 3.9353,
+      "step": 1328128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8245934944701883e-05,
+      "loss": 3.9321,
+      "step": 1328640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8237548997191367e-05,
+      "loss": 3.9475,
+      "step": 1329152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8229163049680847e-05,
+      "loss": 3.9213,
+      "step": 1329664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8220777102170327e-05,
+      "loss": 3.9249,
+      "step": 1330176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8212391154659807e-05,
+      "loss": 3.9422,
+      "step": 1330688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.820400520714928e-05,
+      "loss": 3.9408,
+      "step": 1331200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.819561925963876e-05,
+      "loss": 3.9369,
+      "step": 1331712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8187233312128243e-05,
+      "loss": 3.9244,
+      "step": 1332224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.817886374342146e-05,
+      "loss": 3.9269,
+      "step": 1332736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8170477795910932e-05,
+      "loss": 3.9342,
+      "step": 1333248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8162091848400412e-05,
+      "loss": 3.9455,
+      "step": 1333760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8153705900889892e-05,
+      "loss": 3.9387,
+      "step": 1334272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8145336332183105e-05,
+      "loss": 3.9429,
+      "step": 1334784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.813698314228005e-05,
+      "loss": 3.9427,
+      "step": 1335296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.812859719476953e-05,
+      "loss": 3.9426,
+      "step": 1335808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.812021124725901e-05,
+      "loss": 3.9267,
+      "step": 1336320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.811182529974849e-05,
+      "loss": 3.9322,
+      "step": 1336832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.810343935223797e-05,
+      "loss": 3.9462,
+      "step": 1337344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.809505340472745e-05,
+      "loss": 3.9311,
+      "step": 1337856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.808666745721693e-05,
+      "loss": 3.9365,
+      "step": 1338368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8078281509706406e-05,
+      "loss": 3.9313,
+      "step": 1338880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8069895562195886e-05,
+      "loss": 3.9523,
+      "step": 1339392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8061509614685366e-05,
+      "loss": 3.9341,
+      "step": 1339904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8053123667174846e-05,
+      "loss": 3.9247,
+      "step": 1340416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8044737719664326e-05,
+      "loss": 3.9374,
+      "step": 1340928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8036368150957538e-05,
+      "loss": 3.9398,
+      "step": 1341440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8027982203447018e-05,
+      "loss": 3.9447,
+      "step": 1341952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8019612634740227e-05,
+      "loss": 3.9341,
+      "step": 1342464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8011226687229707e-05,
+      "loss": 3.9352,
+      "step": 1342976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8002840739719187e-05,
+      "loss": 3.9413,
+      "step": 1343488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7994454792208667e-05,
+      "loss": 3.9342,
+      "step": 1344000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.798606884469815e-05,
+      "loss": 3.9292,
+      "step": 1344512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.797768289718763e-05,
+      "loss": 3.9336,
+      "step": 1345024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.796929694967711e-05,
+      "loss": 3.9446,
+      "step": 1345536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.796091100216659e-05,
+      "loss": 3.9454,
+      "step": 1346048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.79525414334598e-05,
+      "loss": 3.9413,
+      "step": 1346560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.794415548594928e-05,
+      "loss": 3.9262,
+      "step": 1347072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7935785917242492e-05,
+      "loss": 3.9348,
+      "step": 1347584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7927399969731972e-05,
+      "loss": 3.935,
+      "step": 1348096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7919014022221452e-05,
+      "loss": 3.9417,
+      "step": 1348608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7910628074710932e-05,
+      "loss": 3.9365,
+      "step": 1349120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7902242127200412e-05,
+      "loss": 3.9332,
+      "step": 1349632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.789385617968989e-05,
+      "loss": 3.9256,
+      "step": 1350144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7885486610983104e-05,
+      "loss": 3.9416,
+      "step": 1350656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7877100663472584e-05,
+      "loss": 3.9123,
+      "step": 1351168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7868714715962064e-05,
+      "loss": 3.9346,
+      "step": 1351680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7860328768451544e-05,
+      "loss": 3.9288,
+      "step": 1352192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7851959199744753e-05,
+      "loss": 3.9391,
+      "step": 1352704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7843573252234233e-05,
+      "loss": 3.9272,
+      "step": 1353216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7835187304723713e-05,
+      "loss": 3.9388,
+      "step": 1353728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7826801357213196e-05,
+      "loss": 3.9292,
+      "step": 1354240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7818415409702676e-05,
+      "loss": 3.931,
+      "step": 1354752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7810045840995885e-05,
+      "loss": 3.9401,
+      "step": 1355264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7801659893485365e-05,
+      "loss": 3.9275,
+      "step": 1355776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7793273945974845e-05,
+      "loss": 3.9387,
+      "step": 1356288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7784904377268058e-05,
+      "loss": 3.9464,
+      "step": 1356800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7776518429757538e-05,
+      "loss": 3.9371,
+      "step": 1357312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7768132482247018e-05,
+      "loss": 3.9303,
+      "step": 1357824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7759746534736498e-05,
+      "loss": 3.9335,
+      "step": 1358336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7751360587225978e-05,
+      "loss": 3.9387,
+      "step": 1358848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7742974639715458e-05,
+      "loss": 3.9261,
+      "step": 1359360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7734588692204938e-05,
+      "loss": 3.9442,
+      "step": 1359872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7726202744694414e-05,
+      "loss": 3.9286,
+      "step": 1360384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7717816797183894e-05,
+      "loss": 3.9389,
+      "step": 1360896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7709430849673374e-05,
+      "loss": 3.9335,
+      "step": 1361408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7701044902162854e-05,
+      "loss": 3.9306,
+      "step": 1361920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7692658954652334e-05,
+      "loss": 3.9209,
+      "step": 1362432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7684289385945543e-05,
+      "loss": 3.9261,
+      "step": 1362944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7675903438435023e-05,
+      "loss": 3.9212,
+      "step": 1363456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7667517490924506e-05,
+      "loss": 3.9424,
+      "step": 1363968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7659131543413986e-05,
+      "loss": 3.9518,
+      "step": 1364480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7650745595903466e-05,
+      "loss": 3.9373,
+      "step": 1364992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7642359648392946e-05,
+      "loss": 3.9227,
+      "step": 1365504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7633990079686155e-05,
+      "loss": 3.9401,
+      "step": 1366016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7625604132175635e-05,
+      "loss": 3.9342,
+      "step": 1366528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7617218184665115e-05,
+      "loss": 3.9358,
+      "step": 1367040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.76088322371546e-05,
+      "loss": 3.9327,
+      "step": 1367552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.760044628964408e-05,
+      "loss": 3.9376,
+      "step": 1368064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.759206034213356e-05,
+      "loss": 3.9324,
+      "step": 1368576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.758367439462304e-05,
+      "loss": 3.9343,
+      "step": 1369088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7575304825916248e-05,
+      "loss": 3.9348,
+      "step": 1369600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7566918878405727e-05,
+      "loss": 3.9399,
+      "step": 1370112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7558532930895207e-05,
+      "loss": 3.9354,
+      "step": 1370624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.755014698338469e-05,
+      "loss": 3.9261,
+      "step": 1371136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.754176103587417e-05,
+      "loss": 3.9424,
+      "step": 1371648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.753337508836365e-05,
+      "loss": 3.9371,
+      "step": 1372160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7524989140853124e-05,
+      "loss": 3.9299,
+      "step": 1372672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.751661957214634e-05,
+      "loss": 3.9456,
+      "step": 1373184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.750823362463582e-05,
+      "loss": 3.9286,
+      "step": 1373696
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.02972936630249,
+      "eval_runtime": 294.765,
+      "eval_samples_per_second": 1294.56,
+      "eval_steps_per_second": 40.456,
+      "step": 1373760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.74998476771253e-05,
+      "loss": 3.9244,
+      "step": 1374208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7491461729614776e-05,
+      "loss": 3.9275,
+      "step": 1374720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7483092160907992e-05,
+      "loss": 3.9436,
+      "step": 1375232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7474706213397472e-05,
+      "loss": 3.9386,
+      "step": 1375744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7466320265886952e-05,
+      "loss": 3.943,
+      "step": 1376256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.745793431837643e-05,
+      "loss": 3.9274,
+      "step": 1376768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.744954837086591e-05,
+      "loss": 3.9364,
+      "step": 1377280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.744116242335539e-05,
+      "loss": 3.9163,
+      "step": 1377792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.743277647584487e-05,
+      "loss": 3.9437,
+      "step": 1378304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.742439052833435e-05,
+      "loss": 3.9252,
+      "step": 1378816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7416020959627557e-05,
+      "loss": 3.9427,
+      "step": 1379328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7407635012117037e-05,
+      "loss": 3.939,
+      "step": 1379840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.739926544341025e-05,
+      "loss": 3.9334,
+      "step": 1380352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.739087949589973e-05,
+      "loss": 3.9251,
+      "step": 1380864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.738249354838921e-05,
+      "loss": 3.9237,
+      "step": 1381376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.737410760087869e-05,
+      "loss": 3.9266,
+      "step": 1381888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.736572165336817e-05,
+      "loss": 3.9332,
+      "step": 1382400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.735733570585765e-05,
+      "loss": 3.9326,
+      "step": 1382912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.734894975834713e-05,
+      "loss": 3.9332,
+      "step": 1383424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7340563810836613e-05,
+      "loss": 3.9498,
+      "step": 1383936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7332177863326093e-05,
+      "loss": 3.9441,
+      "step": 1384448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7323791915815573e-05,
+      "loss": 3.9386,
+      "step": 1384960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7315405968305053e-05,
+      "loss": 3.9321,
+      "step": 1385472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7307020020794533e-05,
+      "loss": 3.9414,
+      "step": 1385984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7298650452087742e-05,
+      "loss": 3.9293,
+      "step": 1386496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7290280883380954e-05,
+      "loss": 3.9377,
+      "step": 1387008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7281894935870434e-05,
+      "loss": 3.9358,
+      "step": 1387520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7273508988359914e-05,
+      "loss": 3.9332,
+      "step": 1388032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7265123040849394e-05,
+      "loss": 3.9249,
+      "step": 1388544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7256737093338874e-05,
+      "loss": 3.9286,
+      "step": 1389056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7248351145828354e-05,
+      "loss": 3.9335,
+      "step": 1389568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7239965198317834e-05,
+      "loss": 3.9383,
+      "step": 1390080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7231579250807314e-05,
+      "loss": 3.9428,
+      "step": 1390592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7223209682100527e-05,
+      "loss": 3.9361,
+      "step": 1391104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7214823734590007e-05,
+      "loss": 3.9351,
+      "step": 1391616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7206437787079486e-05,
+      "loss": 3.9344,
+      "step": 1392128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.719805183956896e-05,
+      "loss": 3.9213,
+      "step": 1392640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.718966589205844e-05,
+      "loss": 3.9338,
+      "step": 1393152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7181279944547923e-05,
+      "loss": 3.9243,
+      "step": 1393664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7172893997037403e-05,
+      "loss": 3.9236,
+      "step": 1394176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7164524428330612e-05,
+      "loss": 3.9301,
+      "step": 1394688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7156154859623828e-05,
+      "loss": 3.9392,
+      "step": 1395200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7147768912113308e-05,
+      "loss": 3.9293,
+      "step": 1395712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7139382964602784e-05,
+      "loss": 3.9274,
+      "step": 1396224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7130997017092264e-05,
+      "loss": 3.9396,
+      "step": 1396736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7122611069581744e-05,
+      "loss": 3.9317,
+      "step": 1397248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7114225122071224e-05,
+      "loss": 3.9443,
+      "step": 1397760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7105839174560704e-05,
+      "loss": 3.9253,
+      "step": 1398272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7097453227050184e-05,
+      "loss": 3.9228,
+      "step": 1398784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7089083658343393e-05,
+      "loss": 3.9389,
+      "step": 1399296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7080697710832877e-05,
+      "loss": 3.9286,
+      "step": 1399808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7072311763322357e-05,
+      "loss": 3.9335,
+      "step": 1400320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7063942194615566e-05,
+      "loss": 3.9258,
+      "step": 1400832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7055556247105046e-05,
+      "loss": 3.9294,
+      "step": 1401344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7047170299594526e-05,
+      "loss": 3.9208,
+      "step": 1401856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7038784352084006e-05,
+      "loss": 3.9339,
+      "step": 1402368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7030398404573485e-05,
+      "loss": 3.928,
+      "step": 1402880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.702201245706297e-05,
+      "loss": 3.9322,
+      "step": 1403392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.701362650955245e-05,
+      "loss": 3.9366,
+      "step": 1403904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.700524056204193e-05,
+      "loss": 3.9236,
+      "step": 1404416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6996870993335138e-05,
+      "loss": 3.9219,
+      "step": 1404928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6988485045824618e-05,
+      "loss": 3.9438,
+      "step": 1405440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6980099098314098e-05,
+      "loss": 3.9127,
+      "step": 1405952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.697172952960731e-05,
+      "loss": 3.9152,
+      "step": 1406464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.696335996090052e-05,
+      "loss": 3.9303,
+      "step": 1406976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.695497401339e-05,
+      "loss": 3.9368,
+      "step": 1407488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.694658806587948e-05,
+      "loss": 3.928,
+      "step": 1408000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.693820211836896e-05,
+      "loss": 3.9176,
+      "step": 1408512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.692983254966217e-05,
+      "loss": 3.922,
+      "step": 1409024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.692144660215165e-05,
+      "loss": 3.9212,
+      "step": 1409536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.691306065464113e-05,
+      "loss": 3.9361,
+      "step": 1410048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.690467470713061e-05,
+      "loss": 3.93,
+      "step": 1410560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.689628875962009e-05,
+      "loss": 3.9322,
+      "step": 1411072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.688790281210957e-05,
+      "loss": 3.9322,
+      "step": 1411584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.687951686459905e-05,
+      "loss": 3.9378,
+      "step": 1412096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.687113091708853e-05,
+      "loss": 3.9196,
+      "step": 1412608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6862744969578015e-05,
+      "loss": 3.924,
+      "step": 1413120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6854359022067495e-05,
+      "loss": 3.9357,
+      "step": 1413632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6845973074556975e-05,
+      "loss": 3.9194,
+      "step": 1414144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6837587127046448e-05,
+      "loss": 3.9308,
+      "step": 1414656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6829233937143393e-05,
+      "loss": 3.9226,
+      "step": 1415168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6820847989632876e-05,
+      "loss": 3.9413,
+      "step": 1415680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6812462042122356e-05,
+      "loss": 3.9277,
+      "step": 1416192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6804076094611836e-05,
+      "loss": 3.9171,
+      "step": 1416704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6795690147101316e-05,
+      "loss": 3.9233,
+      "step": 1417216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6787304199590796e-05,
+      "loss": 3.9301,
+      "step": 1417728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.677891825208027e-05,
+      "loss": 3.9402,
+      "step": 1418240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6770532304569753e-05,
+      "loss": 3.9251,
+      "step": 1418752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.676216273586297e-05,
+      "loss": 3.9285,
+      "step": 1419264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.675377678835245e-05,
+      "loss": 3.9231,
+      "step": 1419776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6745407219645657e-05,
+      "loss": 3.9341,
+      "step": 1420288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6737021272135137e-05,
+      "loss": 3.923,
+      "step": 1420800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6728635324624617e-05,
+      "loss": 3.9207,
+      "step": 1421312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6720249377114094e-05,
+      "loss": 3.935,
+      "step": 1421824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.671187980840731e-05,
+      "loss": 3.936,
+      "step": 1422336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.670349386089679e-05,
+      "loss": 3.9356,
+      "step": 1422848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.669510791338627e-05,
+      "loss": 3.9193,
+      "step": 1423360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6686721965875743e-05,
+      "loss": 3.926,
+      "step": 1423872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6678336018365223e-05,
+      "loss": 3.9258,
+      "step": 1424384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6669950070854706e-05,
+      "loss": 3.9312,
+      "step": 1424896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6661564123344186e-05,
+      "loss": 3.9288,
+      "step": 1425408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6653178175833666e-05,
+      "loss": 3.9223,
+      "step": 1425920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6644792228323146e-05,
+      "loss": 3.9188,
+      "step": 1426432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6636406280812626e-05,
+      "loss": 3.9327,
+      "step": 1426944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6628020333302106e-05,
+      "loss": 3.9058,
+      "step": 1427456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6619634385791586e-05,
+      "loss": 3.9233,
+      "step": 1427968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6611264817084795e-05,
+      "loss": 3.9169,
+      "step": 1428480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.660287886957428e-05,
+      "loss": 3.9322,
+      "step": 1428992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6594509300867487e-05,
+      "loss": 3.9227,
+      "step": 1429504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6586123353356967e-05,
+      "loss": 3.9281,
+      "step": 1430016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6577737405846447e-05,
+      "loss": 3.9215,
+      "step": 1430528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6569351458335927e-05,
+      "loss": 3.9235,
+      "step": 1431040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6560965510825407e-05,
+      "loss": 3.9306,
+      "step": 1431552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.655259594211862e-05,
+      "loss": 3.9227,
+      "step": 1432064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.65442099946081e-05,
+      "loss": 3.9317,
+      "step": 1432576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.653582404709758e-05,
+      "loss": 3.9344,
+      "step": 1433088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.652743809958706e-05,
+      "loss": 3.9319,
+      "step": 1433600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.651906853088027e-05,
+      "loss": 3.9223,
+      "step": 1434112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.651068258336975e-05,
+      "loss": 3.9188,
+      "step": 1434624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6502296635859232e-05,
+      "loss": 3.9341,
+      "step": 1435136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.649392706715244e-05,
+      "loss": 3.9188,
+      "step": 1435648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.648554111964192e-05,
+      "loss": 3.9339,
+      "step": 1436160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.64771551721314e-05,
+      "loss": 3.9203,
+      "step": 1436672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.646876922462088e-05,
+      "loss": 3.9293,
+      "step": 1437184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.646038327711036e-05,
+      "loss": 3.9257,
+      "step": 1437696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.645199732959984e-05,
+      "loss": 3.9239,
+      "step": 1438208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6443611382089324e-05,
+      "loss": 3.9116,
+      "step": 1438720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6435241813382533e-05,
+      "loss": 3.921,
+      "step": 1439232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6426855865872013e-05,
+      "loss": 3.9161,
+      "step": 1439744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6418469918361493e-05,
+      "loss": 3.93,
+      "step": 1440256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6410083970850973e-05,
+      "loss": 3.9425,
+      "step": 1440768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6401698023340453e-05,
+      "loss": 3.9329,
+      "step": 1441280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.639331207582993e-05,
+      "loss": 3.9096,
+      "step": 1441792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.638492612831941e-05,
+      "loss": 3.9365,
+      "step": 1442304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.637654018080889e-05,
+      "loss": 3.9257,
+      "step": 1442816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.636815423329837e-05,
+      "loss": 3.9272,
+      "step": 1443328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.635976828578785e-05,
+      "loss": 3.9164,
+      "step": 1443840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6351398717081062e-05,
+      "loss": 3.9315,
+      "step": 1444352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6343012769570542e-05,
+      "loss": 3.9238,
+      "step": 1444864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.633464320086375e-05,
+      "loss": 3.9259,
+      "step": 1445376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.632625725335323e-05,
+      "loss": 3.9264,
+      "step": 1445888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.631787130584271e-05,
+      "loss": 3.93,
+      "step": 1446400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.630948535833219e-05,
+      "loss": 3.9256,
+      "step": 1446912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.630109941082167e-05,
+      "loss": 3.9206,
+      "step": 1447424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6292713463311154e-05,
+      "loss": 3.9308,
+      "step": 1447936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6284327515800634e-05,
+      "loss": 3.9305,
+      "step": 1448448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6275941568290114e-05,
+      "loss": 3.9208,
+      "step": 1448960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6267571999583323e-05,
+      "loss": 3.9369,
+      "step": 1449472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6259186052072803e-05,
+      "loss": 3.9193,
+      "step": 1449984
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.025951385498047,
+      "eval_runtime": 295.864,
+      "eval_samples_per_second": 1289.751,
+      "eval_steps_per_second": 40.306,
+      "step": 1450080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6250800104562283e-05,
+      "loss": 3.9143,
+      "step": 1450496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6242414157051763e-05,
+      "loss": 3.9211,
+      "step": 1451008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6234028209541246e-05,
+      "loss": 3.9346,
+      "step": 1451520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6225642262030726e-05,
+      "loss": 3.9259,
+      "step": 1452032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6217256314520206e-05,
+      "loss": 3.9402,
+      "step": 1452544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6208870367009686e-05,
+      "loss": 3.9158,
+      "step": 1453056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6200484419499166e-05,
+      "loss": 3.9301,
+      "step": 1453568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.619209847198864e-05,
+      "loss": 3.9036,
+      "step": 1454080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6183712524478123e-05,
+      "loss": 3.9345,
+      "step": 1454592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6175326576967603e-05,
+      "loss": 3.9199,
+      "step": 1455104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6166940629457083e-05,
+      "loss": 3.9316,
+      "step": 1455616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6158554681946563e-05,
+      "loss": 3.9309,
+      "step": 1456128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6150185113239772e-05,
+      "loss": 3.9197,
+      "step": 1456640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6141799165729252e-05,
+      "loss": 3.9209,
+      "step": 1457152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.613341321821873e-05,
+      "loss": 3.9181,
+      "step": 1457664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.612502727070821e-05,
+      "loss": 3.9163,
+      "step": 1458176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6116641323197695e-05,
+      "loss": 3.9267,
+      "step": 1458688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6108255375687175e-05,
+      "loss": 3.9215,
+      "step": 1459200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6099869428176655e-05,
+      "loss": 3.9244,
+      "step": 1459712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6091499859469864e-05,
+      "loss": 3.938,
+      "step": 1460224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6083113911959344e-05,
+      "loss": 3.9399,
+      "step": 1460736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6074727964448824e-05,
+      "loss": 3.9285,
+      "step": 1461248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6066342016938304e-05,
+      "loss": 3.926,
+      "step": 1461760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6057956069427787e-05,
+      "loss": 3.932,
+      "step": 1462272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6049570121917267e-05,
+      "loss": 3.9191,
+      "step": 1462784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6041184174406747e-05,
+      "loss": 3.9347,
+      "step": 1463296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6032798226896227e-05,
+      "loss": 3.9267,
+      "step": 1463808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6024412279385707e-05,
+      "loss": 3.9214,
+      "step": 1464320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6016042710678916e-05,
+      "loss": 3.9174,
+      "step": 1464832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.600767314197213e-05,
+      "loss": 3.925,
+      "step": 1465344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.599928719446161e-05,
+      "loss": 3.9215,
+      "step": 1465856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.599090124695109e-05,
+      "loss": 3.9318,
+      "step": 1466368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.598251529944057e-05,
+      "loss": 3.9332,
+      "step": 1466880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.597412935193005e-05,
+      "loss": 3.9273,
+      "step": 1467392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.596574340441953e-05,
+      "loss": 3.9282,
+      "step": 1467904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5957357456909005e-05,
+      "loss": 3.9285,
+      "step": 1468416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.594898788820222e-05,
+      "loss": 3.9118,
+      "step": 1468928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.59406019406917e-05,
+      "loss": 3.9245,
+      "step": 1469440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.593221599318118e-05,
+      "loss": 3.9139,
+      "step": 1469952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5923830045670654e-05,
+      "loss": 3.9196,
+      "step": 1470464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5915444098160134e-05,
+      "loss": 3.9231,
+      "step": 1470976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5907058150649617e-05,
+      "loss": 3.9281,
+      "step": 1471488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5898672203139097e-05,
+      "loss": 3.9251,
+      "step": 1472000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5890286255628577e-05,
+      "loss": 3.9156,
+      "step": 1472512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5881900308118057e-05,
+      "loss": 3.9309,
+      "step": 1473024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5873514360607537e-05,
+      "loss": 3.9274,
+      "step": 1473536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5865144791900746e-05,
+      "loss": 3.9328,
+      "step": 1474048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5856758844390226e-05,
+      "loss": 3.9224,
+      "step": 1474560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.584837289687971e-05,
+      "loss": 3.911,
+      "step": 1475072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.583998694936919e-05,
+      "loss": 3.932,
+      "step": 1475584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.58316173806624e-05,
+      "loss": 3.9184,
+      "step": 1476096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.582323143315188e-05,
+      "loss": 3.9244,
+      "step": 1476608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.581484548564136e-05,
+      "loss": 3.9197,
+      "step": 1477120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.580645953813084e-05,
+      "loss": 3.9179,
+      "step": 1477632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.579808996942405e-05,
+      "loss": 3.9147,
+      "step": 1478144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.578970402191353e-05,
+      "loss": 3.9237,
+      "step": 1478656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.578131807440301e-05,
+      "loss": 3.9219,
+      "step": 1479168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.577293212689249e-05,
+      "loss": 3.9252,
+      "step": 1479680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.57645625581857e-05,
+      "loss": 3.9241,
+      "step": 1480192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.575617661067518e-05,
+      "loss": 3.9151,
+      "step": 1480704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5747790663164663e-05,
+      "loss": 3.9145,
+      "step": 1481216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5739404715654143e-05,
+      "loss": 3.9373,
+      "step": 1481728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5731018768143623e-05,
+      "loss": 3.906,
+      "step": 1482240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5722632820633103e-05,
+      "loss": 3.9059,
+      "step": 1482752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5714246873122583e-05,
+      "loss": 3.9178,
+      "step": 1483264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5705860925612063e-05,
+      "loss": 3.9291,
+      "step": 1483776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5697507735709005e-05,
+      "loss": 3.9184,
+      "step": 1484288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5689121788198484e-05,
+      "loss": 3.9128,
+      "step": 1484800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5680735840687964e-05,
+      "loss": 3.9095,
+      "step": 1485312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5672349893177444e-05,
+      "loss": 3.915,
+      "step": 1485824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5663963945666924e-05,
+      "loss": 3.9293,
+      "step": 1486336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5655577998156404e-05,
+      "loss": 3.9216,
+      "step": 1486848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5647208429449617e-05,
+      "loss": 3.9201,
+      "step": 1487360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5638822481939097e-05,
+      "loss": 3.9277,
+      "step": 1487872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5630436534428577e-05,
+      "loss": 3.9295,
+      "step": 1488384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5622050586918057e-05,
+      "loss": 3.9116,
+      "step": 1488896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5613664639407537e-05,
+      "loss": 3.9154,
+      "step": 1489408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5605278691897017e-05,
+      "loss": 3.9297,
+      "step": 1489920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5596892744386493e-05,
+      "loss": 3.9078,
+      "step": 1490432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5588523175679706e-05,
+      "loss": 3.9223,
+      "step": 1490944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.558013722816919e-05,
+      "loss": 3.9137,
+      "step": 1491456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5571767659462398e-05,
+      "loss": 3.9336,
+      "step": 1491968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5563381711951878e-05,
+      "loss": 3.9194,
+      "step": 1492480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5554995764441358e-05,
+      "loss": 3.9098,
+      "step": 1492992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5546609816930838e-05,
+      "loss": 3.9114,
+      "step": 1493504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5538223869420314e-05,
+      "loss": 3.9246,
+      "step": 1494016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5529837921909794e-05,
+      "loss": 3.9322,
+      "step": 1494528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.552146835320301e-05,
+      "loss": 3.9193,
+      "step": 1495040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.551308240569249e-05,
+      "loss": 3.9163,
+      "step": 1495552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5504696458181963e-05,
+      "loss": 3.9136,
+      "step": 1496064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5496310510671443e-05,
+      "loss": 3.9263,
+      "step": 1496576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5487924563160927e-05,
+      "loss": 3.9135,
+      "step": 1497088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5479538615650407e-05,
+      "loss": 3.9148,
+      "step": 1497600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5471152668139887e-05,
+      "loss": 3.9252,
+      "step": 1498112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5462783099433096e-05,
+      "loss": 3.9285,
+      "step": 1498624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5454397151922576e-05,
+      "loss": 3.9246,
+      "step": 1499136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5446011204412056e-05,
+      "loss": 3.9084,
+      "step": 1499648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5437625256901536e-05,
+      "loss": 3.9204,
+      "step": 1500160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.542923930939102e-05,
+      "loss": 3.915,
+      "step": 1500672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.54208533618805e-05,
+      "loss": 3.9218,
+      "step": 1501184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.541246741436998e-05,
+      "loss": 3.9211,
+      "step": 1501696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.540408146685946e-05,
+      "loss": 3.9136,
+      "step": 1502208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.539569551934894e-05,
+      "loss": 3.9129,
+      "step": 1502720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.538730957183842e-05,
+      "loss": 3.9226,
+      "step": 1503232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5378940003131628e-05,
+      "loss": 3.9001,
+      "step": 1503744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.537055405562111e-05,
+      "loss": 3.9164,
+      "step": 1504256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.536216810811059e-05,
+      "loss": 3.9037,
+      "step": 1504768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.53537985394038e-05,
+      "loss": 3.9223,
+      "step": 1505280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.534541259189328e-05,
+      "loss": 3.9136,
+      "step": 1505792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.533702664438276e-05,
+      "loss": 3.9172,
+      "step": 1506304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.532864069687224e-05,
+      "loss": 3.913,
+      "step": 1506816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.532025474936172e-05,
+      "loss": 3.9182,
+      "step": 1507328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5311868801851203e-05,
+      "loss": 3.924,
+      "step": 1507840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5303482854340677e-05,
+      "loss": 3.9109,
+      "step": 1508352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5295096906830157e-05,
+      "loss": 3.9205,
+      "step": 1508864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5286710959319636e-05,
+      "loss": 3.9245,
+      "step": 1509376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5278325011809116e-05,
+      "loss": 3.9211,
+      "step": 1509888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5269939064298596e-05,
+      "loss": 3.9219,
+      "step": 1510400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.526155311678808e-05,
+      "loss": 3.9074,
+      "step": 1510912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.525318354808129e-05,
+      "loss": 3.9277,
+      "step": 1511424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.524479760057077e-05,
+      "loss": 3.9148,
+      "step": 1511936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.523641165306025e-05,
+      "loss": 3.9213,
+      "step": 1512448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.522802570554973e-05,
+      "loss": 3.9122,
+      "step": 1512960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.521963975803921e-05,
+      "loss": 3.9199,
+      "step": 1513472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.521125381052869e-05,
+      "loss": 3.9181,
+      "step": 1513984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5202867863018172e-05,
+      "loss": 3.9145,
+      "step": 1514496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5194481915507652e-05,
+      "loss": 3.9041,
+      "step": 1515008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.518612872560459e-05,
+      "loss": 3.9094,
+      "step": 1515520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.517774277809407e-05,
+      "loss": 3.912,
+      "step": 1516032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.516935683058355e-05,
+      "loss": 3.9192,
+      "step": 1516544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5160987261876763e-05,
+      "loss": 3.9319,
+      "step": 1517056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5152601314366243e-05,
+      "loss": 3.9268,
+      "step": 1517568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5144215366855722e-05,
+      "loss": 3.9016,
+      "step": 1518080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5135829419345202e-05,
+      "loss": 3.928,
+      "step": 1518592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5127443471834682e-05,
+      "loss": 3.918,
+      "step": 1519104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5119057524324162e-05,
+      "loss": 3.9189,
+      "step": 1519616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5110671576813642e-05,
+      "loss": 3.9123,
+      "step": 1520128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5102302008106855e-05,
+      "loss": 3.922,
+      "step": 1520640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5093916060596335e-05,
+      "loss": 3.9152,
+      "step": 1521152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5085530113085815e-05,
+      "loss": 3.9227,
+      "step": 1521664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5077144165575295e-05,
+      "loss": 3.9159,
+      "step": 1522176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5068758218064775e-05,
+      "loss": 3.9196,
+      "step": 1522688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5060388649357984e-05,
+      "loss": 3.9218,
+      "step": 1523200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5052002701847467e-05,
+      "loss": 3.911,
+      "step": 1523712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5043616754336947e-05,
+      "loss": 3.9178,
+      "step": 1524224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5035230806826427e-05,
+      "loss": 3.9266,
+      "step": 1524736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5026844859315907e-05,
+      "loss": 3.9121,
+      "step": 1525248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5018458911805387e-05,
+      "loss": 3.9245,
+      "step": 1525760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.501007296429486e-05,
+      "loss": 3.917,
+      "step": 1526272
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.022130966186523,
+      "eval_runtime": 280.8016,
+      "eval_samples_per_second": 1358.934,
+      "eval_steps_per_second": 42.468,
+      "step": 1526400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5001687016784343e-05,
+      "loss": 3.911,
+      "step": 1526784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4993317448077556e-05,
+      "loss": 3.9112,
+      "step": 1527296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4984931500567036e-05,
+      "loss": 3.9227,
+      "step": 1527808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4976545553056516e-05,
+      "loss": 3.9204,
+      "step": 1528320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4968159605545996e-05,
+      "loss": 3.9325,
+      "step": 1528832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4959790036839205e-05,
+      "loss": 3.9064,
+      "step": 1529344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4951404089328688e-05,
+      "loss": 3.9198,
+      "step": 1529856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4943018141818168e-05,
+      "loss": 3.9037,
+      "step": 1530368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4934632194307648e-05,
+      "loss": 3.9226,
+      "step": 1530880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4926246246797128e-05,
+      "loss": 3.9126,
+      "step": 1531392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4917860299286605e-05,
+      "loss": 3.9185,
+      "step": 1531904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4909474351776085e-05,
+      "loss": 3.9268,
+      "step": 1532416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4901104783069297e-05,
+      "loss": 3.9087,
+      "step": 1532928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4892718835558777e-05,
+      "loss": 3.9159,
+      "step": 1533440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4884332888048257e-05,
+      "loss": 3.9102,
+      "step": 1533952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4875946940537737e-05,
+      "loss": 3.9081,
+      "step": 1534464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4867560993027217e-05,
+      "loss": 3.9134,
+      "step": 1534976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4859175045516697e-05,
+      "loss": 3.9161,
+      "step": 1535488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4850789098006177e-05,
+      "loss": 3.9161,
+      "step": 1536000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4842403150495657e-05,
+      "loss": 3.9279,
+      "step": 1536512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4834017202985137e-05,
+      "loss": 3.9339,
+      "step": 1537024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4825631255474617e-05,
+      "loss": 3.9228,
+      "step": 1537536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4817245307964097e-05,
+      "loss": 3.9161,
+      "step": 1538048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4808875739257306e-05,
+      "loss": 3.9263,
+      "step": 1538560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4800489791746786e-05,
+      "loss": 3.9097,
+      "step": 1539072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4792103844236266e-05,
+      "loss": 3.925,
+      "step": 1539584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4783717896725746e-05,
+      "loss": 3.9195,
+      "step": 1540096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.477533194921523e-05,
+      "loss": 3.9125,
+      "step": 1540608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4766962380508438e-05,
+      "loss": 3.9122,
+      "step": 1541120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4758576432997918e-05,
+      "loss": 3.914,
+      "step": 1541632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.475020686429113e-05,
+      "loss": 3.9164,
+      "step": 1542144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4741820916780607e-05,
+      "loss": 3.9216,
+      "step": 1542656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.473343496927009e-05,
+      "loss": 3.9245,
+      "step": 1543168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.472504902175957e-05,
+      "loss": 3.9211,
+      "step": 1543680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.471667945305278e-05,
+      "loss": 3.9205,
+      "step": 1544192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.470829350554226e-05,
+      "loss": 3.9206,
+      "step": 1544704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.469990755803174e-05,
+      "loss": 3.9019,
+      "step": 1545216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.469152161052122e-05,
+      "loss": 3.9169,
+      "step": 1545728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.46831356630107e-05,
+      "loss": 3.9037,
+      "step": 1546240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4674766094303912e-05,
+      "loss": 3.9109,
+      "step": 1546752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.466638014679339e-05,
+      "loss": 3.9122,
+      "step": 1547264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.465799419928287e-05,
+      "loss": 3.9253,
+      "step": 1547776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.464960825177235e-05,
+      "loss": 3.9161,
+      "step": 1548288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.464122230426183e-05,
+      "loss": 3.906,
+      "step": 1548800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.463283635675131e-05,
+      "loss": 3.9262,
+      "step": 1549312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.462445040924079e-05,
+      "loss": 3.9174,
+      "step": 1549824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.461606446173027e-05,
+      "loss": 3.9219,
+      "step": 1550336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4607694893023484e-05,
+      "loss": 3.9151,
+      "step": 1550848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.459930894551296e-05,
+      "loss": 3.9014,
+      "step": 1551360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4590939376806173e-05,
+      "loss": 3.9246,
+      "step": 1551872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4582553429295653e-05,
+      "loss": 3.9097,
+      "step": 1552384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4574167481785136e-05,
+      "loss": 3.9199,
+      "step": 1552896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4565781534274613e-05,
+      "loss": 3.9107,
+      "step": 1553408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4557395586764093e-05,
+      "loss": 3.9086,
+      "step": 1553920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4549009639253573e-05,
+      "loss": 3.9102,
+      "step": 1554432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4540623691743053e-05,
+      "loss": 3.9144,
+      "step": 1554944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4532237744232533e-05,
+      "loss": 3.9152,
+      "step": 1555456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4523868175525745e-05,
+      "loss": 3.9159,
+      "step": 1555968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4515482228015225e-05,
+      "loss": 3.9201,
+      "step": 1556480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4507096280504705e-05,
+      "loss": 3.9074,
+      "step": 1556992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4498710332994185e-05,
+      "loss": 3.904,
+      "step": 1557504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4490324385483665e-05,
+      "loss": 3.9321,
+      "step": 1558016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.448193843797314e-05,
+      "loss": 3.8962,
+      "step": 1558528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4473568869266354e-05,
+      "loss": 3.8968,
+      "step": 1559040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4465182921755837e-05,
+      "loss": 3.9104,
+      "step": 1559552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4456796974245317e-05,
+      "loss": 3.9275,
+      "step": 1560064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4448411026734794e-05,
+      "loss": 3.9051,
+      "step": 1560576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4440025079224274e-05,
+      "loss": 3.9113,
+      "step": 1561088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4431655510517486e-05,
+      "loss": 3.8997,
+      "step": 1561600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4423269563006966e-05,
+      "loss": 3.9063,
+      "step": 1562112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4414883615496446e-05,
+      "loss": 3.919,
+      "step": 1562624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4406497667985926e-05,
+      "loss": 3.9151,
+      "step": 1563136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4398111720475406e-05,
+      "loss": 3.9117,
+      "step": 1563648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4389742151768615e-05,
+      "loss": 3.9201,
+      "step": 1564160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4381356204258095e-05,
+      "loss": 3.9206,
+      "step": 1564672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4372970256747575e-05,
+      "loss": 3.9074,
+      "step": 1565184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.436458430923706e-05,
+      "loss": 3.9078,
+      "step": 1565696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.435619836172654e-05,
+      "loss": 3.9226,
+      "step": 1566208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.434781241421602e-05,
+      "loss": 3.898,
+      "step": 1566720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4339442845509227e-05,
+      "loss": 3.9153,
+      "step": 1567232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.433107327680244e-05,
+      "loss": 3.9064,
+      "step": 1567744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.432268732929192e-05,
+      "loss": 3.9255,
+      "step": 1568256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.43143013817814e-05,
+      "loss": 3.9124,
+      "step": 1568768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.430591543427088e-05,
+      "loss": 3.9043,
+      "step": 1569280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.429752948676036e-05,
+      "loss": 3.8979,
+      "step": 1569792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.428914353924984e-05,
+      "loss": 3.9178,
+      "step": 1570304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.428075759173932e-05,
+      "loss": 3.9202,
+      "step": 1570816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4272371644228796e-05,
+      "loss": 3.9155,
+      "step": 1571328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4263985696718276e-05,
+      "loss": 3.9074,
+      "step": 1571840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.425559974920776e-05,
+      "loss": 3.9091,
+      "step": 1572352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.424721380169724e-05,
+      "loss": 3.9177,
+      "step": 1572864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.423882785418672e-05,
+      "loss": 3.907,
+      "step": 1573376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.42304419066762e-05,
+      "loss": 3.9063,
+      "step": 1573888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4222105095576874e-05,
+      "loss": 3.9203,
+      "step": 1574400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4213719148066354e-05,
+      "loss": 3.9163,
+      "step": 1574912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4205333200555833e-05,
+      "loss": 3.9169,
+      "step": 1575424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4196947253045313e-05,
+      "loss": 3.9041,
+      "step": 1575936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4188577684338523e-05,
+      "loss": 3.9113,
+      "step": 1576448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4180191736828002e-05,
+      "loss": 3.9034,
+      "step": 1576960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4171805789317482e-05,
+      "loss": 3.9136,
+      "step": 1577472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4163419841806962e-05,
+      "loss": 3.9106,
+      "step": 1577984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4155033894296446e-05,
+      "loss": 3.9104,
+      "step": 1578496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4146647946785922e-05,
+      "loss": 3.9053,
+      "step": 1579008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4138261999275402e-05,
+      "loss": 3.9081,
+      "step": 1579520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4129876051764882e-05,
+      "loss": 3.9012,
+      "step": 1580032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4121490104254362e-05,
+      "loss": 3.9042,
+      "step": 1580544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4113120535547575e-05,
+      "loss": 3.8973,
+      "step": 1581056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4104734588037055e-05,
+      "loss": 3.9154,
+      "step": 1581568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4096348640526535e-05,
+      "loss": 3.9099,
+      "step": 1582080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4087962693016015e-05,
+      "loss": 3.9057,
+      "step": 1582592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4079576745505494e-05,
+      "loss": 3.9037,
+      "step": 1583104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4071190797994974e-05,
+      "loss": 3.914,
+      "step": 1583616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4062821229288184e-05,
+      "loss": 3.9138,
+      "step": 1584128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4054435281777667e-05,
+      "loss": 3.9036,
+      "step": 1584640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4046049334267147e-05,
+      "loss": 3.913,
+      "step": 1585152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4037663386756627e-05,
+      "loss": 3.9183,
+      "step": 1585664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4029293818049836e-05,
+      "loss": 3.9133,
+      "step": 1586176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4020907870539316e-05,
+      "loss": 3.9138,
+      "step": 1586688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4012521923028796e-05,
+      "loss": 3.9001,
+      "step": 1587200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4004152354322008e-05,
+      "loss": 3.9149,
+      "step": 1587712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3995766406811488e-05,
+      "loss": 3.9105,
+      "step": 1588224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3987380459300968e-05,
+      "loss": 3.9168,
+      "step": 1588736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3978994511790448e-05,
+      "loss": 3.8961,
+      "step": 1589248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3970608564279925e-05,
+      "loss": 3.9159,
+      "step": 1589760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3962222616769405e-05,
+      "loss": 3.9124,
+      "step": 1590272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3953836669258885e-05,
+      "loss": 3.9048,
+      "step": 1590784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.39454671005521e-05,
+      "loss": 3.8985,
+      "step": 1591296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3937081153041577e-05,
+      "loss": 3.9029,
+      "step": 1591808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3928695205531057e-05,
+      "loss": 3.9004,
+      "step": 1592320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3920309258020537e-05,
+      "loss": 3.9145,
+      "step": 1592832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3911923310510017e-05,
+      "loss": 3.9195,
+      "step": 1593344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3903537362999497e-05,
+      "loss": 3.9224,
+      "step": 1593856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3895151415488977e-05,
+      "loss": 3.8932,
+      "step": 1594368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3886765467978457e-05,
+      "loss": 3.9191,
+      "step": 1594880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3878379520467937e-05,
+      "loss": 3.9131,
+      "step": 1595392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3869993572957417e-05,
+      "loss": 3.9072,
+      "step": 1595904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3861607625446897e-05,
+      "loss": 3.9057,
+      "step": 1596416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3853221677936377e-05,
+      "loss": 3.9181,
+      "step": 1596928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3844852109229586e-05,
+      "loss": 3.9079,
+      "step": 1597440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.383646616171907e-05,
+      "loss": 3.9113,
+      "step": 1597952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.382808021420855e-05,
+      "loss": 3.9049,
+      "step": 1598464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.381969426669803e-05,
+      "loss": 3.9109,
+      "step": 1598976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3811324697991238e-05,
+      "loss": 3.9127,
+      "step": 1599488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3802938750480718e-05,
+      "loss": 3.9069,
+      "step": 1600000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3794552802970198e-05,
+      "loss": 3.91,
+      "step": 1600512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.378618323426341e-05,
+      "loss": 3.9195,
+      "step": 1601024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.377779728675289e-05,
+      "loss": 3.9047,
+      "step": 1601536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.376941133924237e-05,
+      "loss": 3.9132,
+      "step": 1602048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.376102539173185e-05,
+      "loss": 3.9122,
+      "step": 1602560
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.018652439117432,
+      "eval_runtime": 281.7295,
+      "eval_samples_per_second": 1354.459,
+      "eval_steps_per_second": 42.328,
+      "step": 1602720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.375263944422133e-05,
+      "loss": 3.9017,
+      "step": 1603072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.374425349671081e-05,
+      "loss": 3.9013,
+      "step": 1603584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.373586754920029e-05,
+      "loss": 3.9147,
+      "step": 1604096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.372748160168977e-05,
+      "loss": 3.9145,
+      "step": 1604608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.371909565417925e-05,
+      "loss": 3.919,
+      "step": 1605120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.371070970666873e-05,
+      "loss": 3.9045,
+      "step": 1605632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.370232375915821e-05,
+      "loss": 3.9108,
+      "step": 1606144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.369393781164769e-05,
+      "loss": 3.8979,
+      "step": 1606656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3685551864137167e-05,
+      "loss": 3.9113,
+      "step": 1607168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3677165916626647e-05,
+      "loss": 3.9091,
+      "step": 1607680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.366877996911613e-05,
+      "loss": 3.9121,
+      "step": 1608192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.366039402160561e-05,
+      "loss": 3.9149,
+      "step": 1608704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.365202445289882e-05,
+      "loss": 3.905,
+      "step": 1609216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.36436385053883e-05,
+      "loss": 3.9061,
+      "step": 1609728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.363525255787778e-05,
+      "loss": 3.9044,
+      "step": 1610240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.362686661036726e-05,
+      "loss": 3.8992,
+      "step": 1610752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.361848066285674e-05,
+      "loss": 3.9047,
+      "step": 1611264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3610094715346222e-05,
+      "loss": 3.9088,
+      "step": 1611776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3601708767835702e-05,
+      "loss": 3.9076,
+      "step": 1612288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.359332282032518e-05,
+      "loss": 3.9212,
+      "step": 1612800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.358493687281466e-05,
+      "loss": 3.924,
+      "step": 1613312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.357655092530414e-05,
+      "loss": 3.9151,
+      "step": 1613824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.356816497779362e-05,
+      "loss": 3.9099,
+      "step": 1614336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.35597790302831e-05,
+      "loss": 3.9159,
+      "step": 1614848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.355140946157631e-05,
+      "loss": 3.906,
+      "step": 1615360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.354302351406579e-05,
+      "loss": 3.9183,
+      "step": 1615872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3534653945359e-05,
+      "loss": 3.9108,
+      "step": 1616384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.352626799784848e-05,
+      "loss": 3.9015,
+      "step": 1616896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.351788205033796e-05,
+      "loss": 3.9047,
+      "step": 1617408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.350949610282744e-05,
+      "loss": 3.9109,
+      "step": 1617920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3501110155316923e-05,
+      "loss": 3.9079,
+      "step": 1618432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3492724207806403e-05,
+      "loss": 3.9131,
+      "step": 1618944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3484338260295883e-05,
+      "loss": 3.9185,
+      "step": 1619456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3475968691589092e-05,
+      "loss": 3.9131,
+      "step": 1619968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.34675991228823e-05,
+      "loss": 3.9118,
+      "step": 1620480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3459213175371785e-05,
+      "loss": 3.9157,
+      "step": 1620992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3450827227861265e-05,
+      "loss": 3.9004,
+      "step": 1621504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3442441280350745e-05,
+      "loss": 3.9031,
+      "step": 1622016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3434055332840224e-05,
+      "loss": 3.8973,
+      "step": 1622528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3425669385329704e-05,
+      "loss": 3.9026,
+      "step": 1623040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.341728343781918e-05,
+      "loss": 3.9048,
+      "step": 1623552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.340889749030866e-05,
+      "loss": 3.9184,
+      "step": 1624064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3400527921601877e-05,
+      "loss": 3.9086,
+      "step": 1624576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3392141974091357e-05,
+      "loss": 3.8991,
+      "step": 1625088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3383756026580833e-05,
+      "loss": 3.9201,
+      "step": 1625600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3375370079070313e-05,
+      "loss": 3.9057,
+      "step": 1626112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3367000510363526e-05,
+      "loss": 3.9185,
+      "step": 1626624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3358614562853002e-05,
+      "loss": 3.906,
+      "step": 1627136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3350228615342486e-05,
+      "loss": 3.8954,
+      "step": 1627648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3341842667831966e-05,
+      "loss": 3.9175,
+      "step": 1628160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3333456720321446e-05,
+      "loss": 3.9036,
+      "step": 1628672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3325070772810926e-05,
+      "loss": 3.9142,
+      "step": 1629184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3316684825300406e-05,
+      "loss": 3.8979,
+      "step": 1629696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3308315256593615e-05,
+      "loss": 3.9041,
+      "step": 1630208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3299929309083095e-05,
+      "loss": 3.8982,
+      "step": 1630720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3291543361572578e-05,
+      "loss": 3.9071,
+      "step": 1631232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3283157414062058e-05,
+      "loss": 3.9103,
+      "step": 1631744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3274787845355267e-05,
+      "loss": 3.9024,
+      "step": 1632256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3266401897844747e-05,
+      "loss": 3.9155,
+      "step": 1632768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3258015950334227e-05,
+      "loss": 3.902,
+      "step": 1633280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3249630002823707e-05,
+      "loss": 3.8951,
+      "step": 1633792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3241244055313187e-05,
+      "loss": 3.9255,
+      "step": 1634304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3232858107802667e-05,
+      "loss": 3.8878,
+      "step": 1634816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3224472160292147e-05,
+      "loss": 3.8842,
+      "step": 1635328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3216086212781627e-05,
+      "loss": 3.9051,
+      "step": 1635840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3207700265271107e-05,
+      "loss": 3.9207,
+      "step": 1636352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3199314317760587e-05,
+      "loss": 3.8948,
+      "step": 1636864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3190928370250067e-05,
+      "loss": 3.9058,
+      "step": 1637376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3182542422739546e-05,
+      "loss": 3.8932,
+      "step": 1637888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3174156475229026e-05,
+      "loss": 3.8997,
+      "step": 1638400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3165770527718506e-05,
+      "loss": 3.9123,
+      "step": 1638912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3157384580207986e-05,
+      "loss": 3.9089,
+      "step": 1639424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3148998632697466e-05,
+      "loss": 3.9056,
+      "step": 1639936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3140645442794408e-05,
+      "loss": 3.9077,
+      "step": 1640448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3132259495283888e-05,
+      "loss": 3.9147,
+      "step": 1640960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3123873547773368e-05,
+      "loss": 3.8962,
+      "step": 1641472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3115487600262848e-05,
+      "loss": 3.9039,
+      "step": 1641984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3107101652752328e-05,
+      "loss": 3.9112,
+      "step": 1642496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3098715705241808e-05,
+      "loss": 3.8938,
+      "step": 1643008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3090329757731288e-05,
+      "loss": 3.9053,
+      "step": 1643520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.30819601890245e-05,
+      "loss": 3.9033,
+      "step": 1644032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.307357424151398e-05,
+      "loss": 3.9142,
+      "step": 1644544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.306518829400346e-05,
+      "loss": 3.9075,
+      "step": 1645056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.305681872529667e-05,
+      "loss": 3.8967,
+      "step": 1645568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.304843277778615e-05,
+      "loss": 3.8909,
+      "step": 1646080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.304004683027563e-05,
+      "loss": 3.9086,
+      "step": 1646592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.303166088276511e-05,
+      "loss": 3.9122,
+      "step": 1647104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.302327493525459e-05,
+      "loss": 3.9046,
+      "step": 1647616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3014888987744072e-05,
+      "loss": 3.9058,
+      "step": 1648128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.300650304023355e-05,
+      "loss": 3.9013,
+      "step": 1648640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.299811709272303e-05,
+      "loss": 3.9069,
+      "step": 1649152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.298974752401624e-05,
+      "loss": 3.8998,
+      "step": 1649664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.298136157650572e-05,
+      "loss": 3.8973,
+      "step": 1650176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.29729756289952e-05,
+      "loss": 3.9143,
+      "step": 1650688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.296458968148468e-05,
+      "loss": 3.9091,
+      "step": 1651200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.295620373397416e-05,
+      "loss": 3.9078,
+      "step": 1651712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.294783416526737e-05,
+      "loss": 3.8944,
+      "step": 1652224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.293944821775685e-05,
+      "loss": 3.9072,
+      "step": 1652736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.293106227024633e-05,
+      "loss": 3.8956,
+      "step": 1653248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2922692701539543e-05,
+      "loss": 3.9088,
+      "step": 1653760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2914306754029023e-05,
+      "loss": 3.9045,
+      "step": 1654272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2905920806518503e-05,
+      "loss": 3.9033,
+      "step": 1654784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2897534859007983e-05,
+      "loss": 3.8983,
+      "step": 1655296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2889148911497462e-05,
+      "loss": 3.8962,
+      "step": 1655808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2880762963986942e-05,
+      "loss": 3.8994,
+      "step": 1656320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2872377016476422e-05,
+      "loss": 3.8936,
+      "step": 1656832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2863991068965902e-05,
+      "loss": 3.8945,
+      "step": 1657344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2855605121455382e-05,
+      "loss": 3.9039,
+      "step": 1657856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2847219173944862e-05,
+      "loss": 3.9051,
+      "step": 1658368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2838833226434342e-05,
+      "loss": 3.8997,
+      "step": 1658880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.283046365772755e-05,
+      "loss": 3.8939,
+      "step": 1659392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.282207771021703e-05,
+      "loss": 3.9031,
+      "step": 1659904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.281369176270651e-05,
+      "loss": 3.9075,
+      "step": 1660416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2805305815195995e-05,
+      "loss": 3.899,
+      "step": 1660928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2796919867685475e-05,
+      "loss": 3.9079,
+      "step": 1661440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2788533920174954e-05,
+      "loss": 3.9081,
+      "step": 1661952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2780164351468164e-05,
+      "loss": 3.9067,
+      "step": 1662464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2771778403957644e-05,
+      "loss": 3.9036,
+      "step": 1662976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2763392456447123e-05,
+      "loss": 3.8957,
+      "step": 1663488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2755006508936603e-05,
+      "loss": 3.9042,
+      "step": 1664000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2746620561426087e-05,
+      "loss": 3.9056,
+      "step": 1664512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2738234613915563e-05,
+      "loss": 3.9101,
+      "step": 1665024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2729848666405043e-05,
+      "loss": 3.8946,
+      "step": 1665536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2721462718894523e-05,
+      "loss": 3.9067,
+      "step": 1666048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2713093150187732e-05,
+      "loss": 3.9063,
+      "step": 1666560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2704707202677212e-05,
+      "loss": 3.8988,
+      "step": 1667072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2696321255166696e-05,
+      "loss": 3.8881,
+      "step": 1667584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2687935307656176e-05,
+      "loss": 3.8992,
+      "step": 1668096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2679565738949385e-05,
+      "loss": 3.8933,
+      "step": 1668608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2671196170242597e-05,
+      "loss": 3.9047,
+      "step": 1669120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2662810222732077e-05,
+      "loss": 3.9126,
+      "step": 1669632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2654424275221557e-05,
+      "loss": 3.9143,
+      "step": 1670144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2646038327711037e-05,
+      "loss": 3.8872,
+      "step": 1670656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2637652380200517e-05,
+      "loss": 3.9127,
+      "step": 1671168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2629266432689997e-05,
+      "loss": 3.9066,
+      "step": 1671680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2620880485179477e-05,
+      "loss": 3.9008,
+      "step": 1672192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2612494537668957e-05,
+      "loss": 3.8968,
+      "step": 1672704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2604108590158437e-05,
+      "loss": 3.912,
+      "step": 1673216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2595722642647917e-05,
+      "loss": 3.8967,
+      "step": 1673728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2587336695137397e-05,
+      "loss": 3.9085,
+      "step": 1674240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2578950747626877e-05,
+      "loss": 3.895,
+      "step": 1674752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.257058117892009e-05,
+      "loss": 3.9083,
+      "step": 1675264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2562195231409566e-05,
+      "loss": 3.9018,
+      "step": 1675776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2553825662702778e-05,
+      "loss": 3.9019,
+      "step": 1676288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2545439715192258e-05,
+      "loss": 3.9057,
+      "step": 1676800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2537053767681738e-05,
+      "loss": 3.9051,
+      "step": 1677312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2528667820171218e-05,
+      "loss": 3.9044,
+      "step": 1677824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2520281872660698e-05,
+      "loss": 3.8985,
+      "step": 1678336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2511895925150178e-05,
+      "loss": 3.9092,
+      "step": 1678848
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.015475749969482,
+      "eval_runtime": 292.4407,
+      "eval_samples_per_second": 1304.849,
+      "eval_steps_per_second": 40.777,
+      "step": 1679040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2503509977639658e-05,
+      "loss": 3.8997,
+      "step": 1679360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2495124030129138e-05,
+      "loss": 3.8926,
+      "step": 1679872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2486770840226083e-05,
+      "loss": 3.9032,
+      "step": 1680384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2478384892715563e-05,
+      "loss": 3.9105,
+      "step": 1680896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.246999894520504e-05,
+      "loss": 3.9109,
+      "step": 1681408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.246161299769452e-05,
+      "loss": 3.8966,
+      "step": 1681920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2453227050184e-05,
+      "loss": 3.9016,
+      "step": 1682432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2444857481477212e-05,
+      "loss": 3.8911,
+      "step": 1682944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2436471533966692e-05,
+      "loss": 3.9057,
+      "step": 1683456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2428085586456172e-05,
+      "loss": 3.8986,
+      "step": 1683968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2419699638945652e-05,
+      "loss": 3.9081,
+      "step": 1684480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.241133007023886e-05,
+      "loss": 3.9065,
+      "step": 1684992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2402960501532073e-05,
+      "loss": 3.8984,
+      "step": 1685504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2394574554021557e-05,
+      "loss": 3.9024,
+      "step": 1686016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2386188606511037e-05,
+      "loss": 3.8914,
+      "step": 1686528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2377802659000513e-05,
+      "loss": 3.8913,
+      "step": 1687040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2369416711489993e-05,
+      "loss": 3.9,
+      "step": 1687552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2361030763979473e-05,
+      "loss": 3.9055,
+      "step": 1688064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2352644816468953e-05,
+      "loss": 3.8988,
+      "step": 1688576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2344258868958433e-05,
+      "loss": 3.9196,
+      "step": 1689088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2335872921447913e-05,
+      "loss": 3.9124,
+      "step": 1689600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2327486973937393e-05,
+      "loss": 3.9103,
+      "step": 1690112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2319101026426873e-05,
+      "loss": 3.9009,
+      "step": 1690624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2310715078916353e-05,
+      "loss": 3.9087,
+      "step": 1691136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2302345510209565e-05,
+      "loss": 3.8961,
+      "step": 1691648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2293959562699042e-05,
+      "loss": 3.9131,
+      "step": 1692160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2285573615188525e-05,
+      "loss": 3.9018,
+      "step": 1692672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2277204046481738e-05,
+      "loss": 3.8929,
+      "step": 1693184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2268818098971218e-05,
+      "loss": 3.9024,
+      "step": 1693696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2260432151460694e-05,
+      "loss": 3.9005,
+      "step": 1694208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2252046203950174e-05,
+      "loss": 3.9016,
+      "step": 1694720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2243676635243387e-05,
+      "loss": 3.9078,
+      "step": 1695232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2235290687732867e-05,
+      "loss": 3.9078,
+      "step": 1695744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2226904740222347e-05,
+      "loss": 3.913,
+      "step": 1696256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.221853517151556e-05,
+      "loss": 3.8971,
+      "step": 1696768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.221014922400504e-05,
+      "loss": 3.9066,
+      "step": 1697280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2201763276494516e-05,
+      "loss": 3.9002,
+      "step": 1697792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2193377328983996e-05,
+      "loss": 3.8914,
+      "step": 1698304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.218499138147348e-05,
+      "loss": 3.8889,
+      "step": 1698816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.217662181276669e-05,
+      "loss": 3.8979,
+      "step": 1699328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2168235865256168e-05,
+      "loss": 3.8977,
+      "step": 1699840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2159849917745648e-05,
+      "loss": 3.9106,
+      "step": 1700352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2151463970235128e-05,
+      "loss": 3.9023,
+      "step": 1700864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2143078022724608e-05,
+      "loss": 3.8931,
+      "step": 1701376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.213470845401782e-05,
+      "loss": 3.9142,
+      "step": 1701888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.21263225065073e-05,
+      "loss": 3.8971,
+      "step": 1702400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.211793655899678e-05,
+      "loss": 3.9135,
+      "step": 1702912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.210955061148626e-05,
+      "loss": 3.9003,
+      "step": 1703424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.210116466397574e-05,
+      "loss": 3.8838,
+      "step": 1703936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.209277871646522e-05,
+      "loss": 3.9145,
+      "step": 1704448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2084392768954697e-05,
+      "loss": 3.8945,
+      "step": 1704960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.207600682144418e-05,
+      "loss": 3.9084,
+      "step": 1705472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.206762087393366e-05,
+      "loss": 3.8951,
+      "step": 1705984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.205923492642314e-05,
+      "loss": 3.8962,
+      "step": 1706496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.205084897891262e-05,
+      "loss": 3.8895,
+      "step": 1707008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.204247941020583e-05,
+      "loss": 3.8968,
+      "step": 1707520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.203409346269531e-05,
+      "loss": 3.9077,
+      "step": 1708032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.202570751518479e-05,
+      "loss": 3.8929,
+      "step": 1708544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2017321567674272e-05,
+      "loss": 3.9067,
+      "step": 1709056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2008935620163752e-05,
+      "loss": 3.9009,
+      "step": 1709568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.200056605145696e-05,
+      "loss": 3.8881,
+      "step": 1710080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.199218010394644e-05,
+      "loss": 3.9154,
+      "step": 1710592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.198379415643592e-05,
+      "loss": 3.8841,
+      "step": 1711104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.19754082089254e-05,
+      "loss": 3.8794,
+      "step": 1711616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.196702226141488e-05,
+      "loss": 3.9001,
+      "step": 1712128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.195863631390436e-05,
+      "loss": 3.9071,
+      "step": 1712640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.195025036639384e-05,
+      "loss": 3.8886,
+      "step": 1713152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.194186441888332e-05,
+      "loss": 3.9029,
+      "step": 1713664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.193349485017653e-05,
+      "loss": 3.8831,
+      "step": 1714176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1925125281469743e-05,
+      "loss": 3.8964,
+      "step": 1714688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1916739333959226e-05,
+      "loss": 3.9027,
+      "step": 1715200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1908353386448702e-05,
+      "loss": 3.8975,
+      "step": 1715712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1899967438938182e-05,
+      "loss": 3.9004,
+      "step": 1716224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1891581491427662e-05,
+      "loss": 3.8987,
+      "step": 1716736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1883211922720875e-05,
+      "loss": 3.909,
+      "step": 1717248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.187482597521035e-05,
+      "loss": 3.8893,
+      "step": 1717760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1866440027699835e-05,
+      "loss": 3.899,
+      "step": 1718272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1858054080189315e-05,
+      "loss": 3.8985,
+      "step": 1718784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1849700890286256e-05,
+      "loss": 3.8904,
+      "step": 1719296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1841314942775736e-05,
+      "loss": 3.8972,
+      "step": 1719808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1832928995265216e-05,
+      "loss": 3.8984,
+      "step": 1720320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1824543047754696e-05,
+      "loss": 3.9061,
+      "step": 1720832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1816157100244176e-05,
+      "loss": 3.905,
+      "step": 1721344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1807771152733656e-05,
+      "loss": 3.8841,
+      "step": 1721856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1799385205223136e-05,
+      "loss": 3.8878,
+      "step": 1722368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.179101563651635e-05,
+      "loss": 3.9008,
+      "step": 1722880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1782629689005825e-05,
+      "loss": 3.9071,
+      "step": 1723392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1774243741495305e-05,
+      "loss": 3.8986,
+      "step": 1723904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.176585779398479e-05,
+      "loss": 3.8939,
+      "step": 1724416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.175747184647427e-05,
+      "loss": 3.8979,
+      "step": 1724928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.174908589896375e-05,
+      "loss": 3.8978,
+      "step": 1725440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1740699951453228e-05,
+      "loss": 3.8969,
+      "step": 1725952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1732314003942705e-05,
+      "loss": 3.8901,
+      "step": 1726464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1723928056432185e-05,
+      "loss": 3.9057,
+      "step": 1726976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1715558487725397e-05,
+      "loss": 3.9013,
+      "step": 1727488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.170717254021488e-05,
+      "loss": 3.9069,
+      "step": 1728000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1698786592704357e-05,
+      "loss": 3.8892,
+      "step": 1728512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1690400645193837e-05,
+      "loss": 3.8969,
+      "step": 1729024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1682014697683317e-05,
+      "loss": 3.8929,
+      "step": 1729536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1673628750172797e-05,
+      "loss": 3.9018,
+      "step": 1730048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1665242802662277e-05,
+      "loss": 3.8959,
+      "step": 1730560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1656889612759222e-05,
+      "loss": 3.8955,
+      "step": 1731072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1648503665248702e-05,
+      "loss": 3.8966,
+      "step": 1731584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.164011771773818e-05,
+      "loss": 3.8876,
+      "step": 1732096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.163173177022766e-05,
+      "loss": 3.892,
+      "step": 1732608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.162334582271714e-05,
+      "loss": 3.8867,
+      "step": 1733120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.161495987520662e-05,
+      "loss": 3.8913,
+      "step": 1733632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.16065739276961e-05,
+      "loss": 3.8912,
+      "step": 1734144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1598187980185582e-05,
+      "loss": 3.905,
+      "step": 1734656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.158980203267506e-05,
+      "loss": 3.8866,
+      "step": 1735168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1581416085164538e-05,
+      "loss": 3.8901,
+      "step": 1735680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.157304651645775e-05,
+      "loss": 3.8935,
+      "step": 1736192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.156466056894723e-05,
+      "loss": 3.9005,
+      "step": 1736704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.155627462143671e-05,
+      "loss": 3.8928,
+      "step": 1737216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.154788867392619e-05,
+      "loss": 3.9052,
+      "step": 1737728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.153950272641567e-05,
+      "loss": 3.8997,
+      "step": 1738240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1531133157708883e-05,
+      "loss": 3.9007,
+      "step": 1738752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.152274721019836e-05,
+      "loss": 3.8964,
+      "step": 1739264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.151436126268784e-05,
+      "loss": 3.893,
+      "step": 1739776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.150597531517732e-05,
+      "loss": 3.8955,
+      "step": 1740288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1497605746470535e-05,
+      "loss": 3.9006,
+      "step": 1740800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1489219798960012e-05,
+      "loss": 3.8968,
+      "step": 1741312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1480833851449492e-05,
+      "loss": 3.8905,
+      "step": 1741824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1472447903938972e-05,
+      "loss": 3.9026,
+      "step": 1742336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1464061956428452e-05,
+      "loss": 3.8979,
+      "step": 1742848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1455676008917932e-05,
+      "loss": 3.8934,
+      "step": 1743360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1447290061407412e-05,
+      "loss": 3.8807,
+      "step": 1743872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.143890411389689e-05,
+      "loss": 3.8903,
+      "step": 1744384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.143051816638637e-05,
+      "loss": 3.8864,
+      "step": 1744896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.142213221887585e-05,
+      "loss": 3.9006,
+      "step": 1745408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.141374627136533e-05,
+      "loss": 3.907,
+      "step": 1745920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.140536032385481e-05,
+      "loss": 3.905,
+      "step": 1746432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.139697437634429e-05,
+      "loss": 3.8836,
+      "step": 1746944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1388604807637504e-05,
+      "loss": 3.9047,
+      "step": 1747456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1380218860126984e-05,
+      "loss": 3.8985,
+      "step": 1747968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1371832912616464e-05,
+      "loss": 3.8979,
+      "step": 1748480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1363446965105944e-05,
+      "loss": 3.8907,
+      "step": 1748992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1355061017595424e-05,
+      "loss": 3.9026,
+      "step": 1749504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1346691448888633e-05,
+      "loss": 3.8886,
+      "step": 1750016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1338305501378113e-05,
+      "loss": 3.9057,
+      "step": 1750528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1329919553867593e-05,
+      "loss": 3.8863,
+      "step": 1751040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1321533606357073e-05,
+      "loss": 3.9004,
+      "step": 1751552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1313147658846553e-05,
+      "loss": 3.8963,
+      "step": 1752064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1304778090139765e-05,
+      "loss": 3.8943,
+      "step": 1752576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1296392142629245e-05,
+      "loss": 3.8975,
+      "step": 1753088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.128803895272619e-05,
+      "loss": 3.8982,
+      "step": 1753600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1279653005215667e-05,
+      "loss": 3.8991,
+      "step": 1754112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1271267057705147e-05,
+      "loss": 3.8925,
+      "step": 1754624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.1262881110194627e-05,
+      "loss": 3.9045,
+      "step": 1755136
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.013008117675781,
+      "eval_runtime": 292.0175,
+      "eval_samples_per_second": 1306.74,
+      "eval_steps_per_second": 40.837,
+      "step": 1755360
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1254495162684107e-05,
+      "loss": 3.8908,
+      "step": 1755648
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1246109215173587e-05,
+      "loss": 3.8863,
+      "step": 1756160
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1237723267663066e-05,
+      "loss": 3.9012,
+      "step": 1756672
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1229337320152546e-05,
+      "loss": 3.9018,
+      "step": 1757184
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1220951372642026e-05,
+      "loss": 3.9026,
+      "step": 1757696
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1212565425131506e-05,
+      "loss": 3.8922,
+      "step": 1758208
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.120419585642472e-05,
+      "loss": 3.8964,
+      "step": 1758720
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1195809908914195e-05,
+      "loss": 3.8835,
+      "step": 1759232
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1187423961403675e-05,
+      "loss": 3.8989,
+      "step": 1759744
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.117903801389316e-05,
+      "loss": 3.8954,
+      "step": 1760256
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.117065206638264e-05,
+      "loss": 3.8969,
+      "step": 1760768
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.116226611887212e-05,
+      "loss": 3.9012,
+      "step": 1761280
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1153896550165328e-05,
+      "loss": 3.8966,
+      "step": 1761792
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1145510602654808e-05,
+      "loss": 3.8908,
+      "step": 1762304
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1137124655144288e-05,
+      "loss": 3.8873,
+      "step": 1762816
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1128738707633768e-05,
+      "loss": 3.883,
+      "step": 1763328
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.112036913892698e-05,
+      "loss": 3.8933,
+      "step": 1763840
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.111198319141646e-05,
+      "loss": 3.8976,
+      "step": 1764352
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.110359724390594e-05,
+      "loss": 3.8945,
+      "step": 1764864
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.109521129639542e-05,
+      "loss": 3.9057,
+      "step": 1765376
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.10868253488849e-05,
+      "loss": 3.9088,
+      "step": 1765888
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1078439401374376e-05,
+      "loss": 3.9075,
+      "step": 1766400
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1070069832667592e-05,
+      "loss": 3.8958,
+      "step": 1766912
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1061683885157072e-05,
+      "loss": 3.8963,
+      "step": 1767424
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.1053297937646552e-05,
+      "loss": 3.8956,
+      "step": 1767936
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.104491199013603e-05,
+      "loss": 3.9046,
+      "step": 1768448
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.103652604262551e-05,
+      "loss": 3.8945,
+      "step": 1768960
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.102814009511499e-05,
+      "loss": 3.8926,
+      "step": 1769472
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.101975414760447e-05,
+      "loss": 3.8978,
+      "step": 1769984
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.101138457889768e-05,
+      "loss": 3.8933,
+      "step": 1770496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.100299863138716e-05,
+      "loss": 3.8935,
+      "step": 1771008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.099461268387664e-05,
+      "loss": 3.9012,
+      "step": 1771520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.098622673636612e-05,
+      "loss": 3.9019,
+      "step": 1772032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.09778407888556e-05,
+      "loss": 3.9033,
+      "step": 1772544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0969487598952546e-05,
+      "loss": 3.8893,
+      "step": 1773056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0961101651442026e-05,
+      "loss": 3.9048,
+      "step": 1773568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0952715703931503e-05,
+      "loss": 3.8975,
+      "step": 1774080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0944329756420982e-05,
+      "loss": 3.8821,
+      "step": 1774592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0935943808910462e-05,
+      "loss": 3.8834,
+      "step": 1775104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0927557861399942e-05,
+      "loss": 3.8894,
+      "step": 1775616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0919171913889422e-05,
+      "loss": 3.8955,
+      "step": 1776128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0910785966378906e-05,
+      "loss": 3.8998,
+      "step": 1776640
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0902400018868382e-05,
+      "loss": 3.8961,
+      "step": 1777152
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0894014071357862e-05,
+      "loss": 3.8843,
+      "step": 1777664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0885628123847342e-05,
+      "loss": 3.9101,
+      "step": 1778176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0877258555140555e-05,
+      "loss": 3.892,
+      "step": 1778688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.086887260763003e-05,
+      "loss": 3.9063,
+      "step": 1779200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0860503038923247e-05,
+      "loss": 3.8979,
+      "step": 1779712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0852117091412727e-05,
+      "loss": 3.8685,
+      "step": 1780224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0843731143902207e-05,
+      "loss": 3.9121,
+      "step": 1780736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0835345196391684e-05,
+      "loss": 3.8873,
+      "step": 1781248
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0826959248881164e-05,
+      "loss": 3.9001,
+      "step": 1781760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0818589680174376e-05,
+      "loss": 3.8907,
+      "step": 1782272
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0810203732663856e-05,
+      "loss": 3.8887,
+      "step": 1782784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0801817785153336e-05,
+      "loss": 3.8863,
+      "step": 1783296
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0793431837642816e-05,
+      "loss": 3.8883,
+      "step": 1783808
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0785045890132296e-05,
+      "loss": 3.9055,
+      "step": 1784320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0776659942621776e-05,
+      "loss": 3.882,
+      "step": 1784832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0768273995111256e-05,
+      "loss": 3.8996,
+      "step": 1785344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0759904426404468e-05,
+      "loss": 3.896,
+      "step": 1785856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0751518478893948e-05,
+      "loss": 3.8784,
+      "step": 1786368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0743132531383428e-05,
+      "loss": 3.904,
+      "step": 1786880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0734746583872908e-05,
+      "loss": 3.8845,
+      "step": 1787392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0726360636362385e-05,
+      "loss": 3.8765,
+      "step": 1787904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0717974688851865e-05,
+      "loss": 3.8911,
+      "step": 1788416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0709588741341345e-05,
+      "loss": 3.8993,
+      "step": 1788928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0701202793830825e-05,
+      "loss": 3.8819,
+      "step": 1789440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0692816846320308e-05,
+      "loss": 3.8971,
+      "step": 1789952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0684430898809788e-05,
+      "loss": 3.8785,
+      "step": 1790464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0676044951299268e-05,
+      "loss": 3.8873,
+      "step": 1790976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0667659003788744e-05,
+      "loss": 3.8972,
+      "step": 1791488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0659273056278224e-05,
+      "loss": 3.8923,
+      "step": 1792000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.065091986637517e-05,
+      "loss": 3.8989,
+      "step": 1792512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.064253391886465e-05,
+      "loss": 3.8903,
+      "step": 1793024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.063414797135413e-05,
+      "loss": 3.8999,
+      "step": 1793536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.062576202384361e-05,
+      "loss": 3.8862,
+      "step": 1794048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.061737607633309e-05,
+      "loss": 3.8957,
+      "step": 1794560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0608990128822566e-05,
+      "loss": 3.8876,
+      "step": 1795072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0600604181312046e-05,
+      "loss": 3.8862,
+      "step": 1795584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.059221823380153e-05,
+      "loss": 3.8907,
+      "step": 1796096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.058384866509474e-05,
+      "loss": 3.8922,
+      "step": 1796608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0575462717584218e-05,
+      "loss": 3.8979,
+      "step": 1797120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0567076770073698e-05,
+      "loss": 3.9032,
+      "step": 1797632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0558690822563178e-05,
+      "loss": 3.8758,
+      "step": 1798144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.0550304875052658e-05,
+      "loss": 3.8801,
+      "step": 1798656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.054193530634587e-05,
+      "loss": 3.8946,
+      "step": 1799168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.053354935883535e-05,
+      "loss": 3.9012,
+      "step": 1799680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.052516341132483e-05,
+      "loss": 3.894,
+      "step": 1800192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.051677746381431e-05,
+      "loss": 3.8888,
+      "step": 1800704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.050839151630379e-05,
+      "loss": 3.897,
+      "step": 1801216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.050000556879327e-05,
+      "loss": 3.8901,
+      "step": 1801728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0491619621282747e-05,
+      "loss": 3.8901,
+      "step": 1802240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.048323367377223e-05,
+      "loss": 3.8842,
+      "step": 1802752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0474864105065443e-05,
+      "loss": 3.8964,
+      "step": 1803264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0466478157554923e-05,
+      "loss": 3.898,
+      "step": 1803776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.04580922100444e-05,
+      "loss": 3.9005,
+      "step": 1804288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.044970626253388e-05,
+      "loss": 3.8859,
+      "step": 1804800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.044133669382709e-05,
+      "loss": 3.8871,
+      "step": 1805312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.043295074631657e-05,
+      "loss": 3.8854,
+      "step": 1805824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.042456479880605e-05,
+      "loss": 3.8963,
+      "step": 1806336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0416195230099264e-05,
+      "loss": 3.8928,
+      "step": 1806848
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0407809282588744e-05,
+      "loss": 3.8876,
+      "step": 1807360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.039942333507822e-05,
+      "loss": 3.8904,
+      "step": 1807872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0391053766371433e-05,
+      "loss": 3.8848,
+      "step": 1808384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0382667818860916e-05,
+      "loss": 3.8855,
+      "step": 1808896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0374281871350396e-05,
+      "loss": 3.8814,
+      "step": 1809408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0365895923839873e-05,
+      "loss": 3.8889,
+      "step": 1809920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0357509976329353e-05,
+      "loss": 3.8868,
+      "step": 1810432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0349124028818833e-05,
+      "loss": 3.8952,
+      "step": 1810944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0340738081308313e-05,
+      "loss": 3.8848,
+      "step": 1811456
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0332352133797793e-05,
+      "loss": 3.8785,
+      "step": 1811968
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0323966186287276e-05,
+      "loss": 3.8897,
+      "step": 1812480
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0315596617580485e-05,
+      "loss": 3.8922,
+      "step": 1812992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0307210670069965e-05,
+      "loss": 3.8881,
+      "step": 1813504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0298824722559445e-05,
+      "loss": 3.8954,
+      "step": 1814016
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0290438775048925e-05,
+      "loss": 3.8907,
+      "step": 1814528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.02820528275384e-05,
+      "loss": 3.9,
+      "step": 1815040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0273666880027885e-05,
+      "loss": 3.8895,
+      "step": 1815552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0265297311321097e-05,
+      "loss": 3.8861,
+      "step": 1816064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0256911363810577e-05,
+      "loss": 3.8857,
+      "step": 1816576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0248525416300054e-05,
+      "loss": 3.9014,
+      "step": 1817088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0240139468789534e-05,
+      "loss": 3.8887,
+      "step": 1817600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0231753521279014e-05,
+      "loss": 3.8819,
+      "step": 1818112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.022338395257223e-05,
+      "loss": 3.8963,
+      "step": 1818624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0214998005061706e-05,
+      "loss": 3.8983,
+      "step": 1819136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0206612057551186e-05,
+      "loss": 3.8836,
+      "step": 1819648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0198226110040666e-05,
+      "loss": 3.8802,
+      "step": 1820160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0189840162530146e-05,
+      "loss": 3.8835,
+      "step": 1820672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.018148697262709e-05,
+      "loss": 3.8812,
+      "step": 1821184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.017310102511657e-05,
+      "loss": 3.8924,
+      "step": 1821696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.016471507760605e-05,
+      "loss": 3.9002,
+      "step": 1822208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0156329130095528e-05,
+      "loss": 3.8993,
+      "step": 1822720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0147943182585008e-05,
+      "loss": 3.8734,
+      "step": 1823232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0139557235074487e-05,
+      "loss": 3.9009,
+      "step": 1823744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0131171287563967e-05,
+      "loss": 3.89,
+      "step": 1824256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0122785340053447e-05,
+      "loss": 3.8903,
+      "step": 1824768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.011441577134666e-05,
+      "loss": 3.8851,
+      "step": 1825280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.010602982383614e-05,
+      "loss": 3.8957,
+      "step": 1825792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.009764387632562e-05,
+      "loss": 3.8856,
+      "step": 1826304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.00892579288151e-05,
+      "loss": 3.9016,
+      "step": 1826816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.008087198130458e-05,
+      "loss": 3.8833,
+      "step": 1827328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0072502412597792e-05,
+      "loss": 3.8917,
+      "step": 1827840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0064116465087272e-05,
+      "loss": 3.8909,
+      "step": 1828352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0055730517576752e-05,
+      "loss": 3.8907,
+      "step": 1828864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.0047344570066232e-05,
+      "loss": 3.8895,
+      "step": 1829376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.003897500135944e-05,
+      "loss": 3.8904,
+      "step": 1829888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.003058905384892e-05,
+      "loss": 3.896,
+      "step": 1830400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.00222031063384e-05,
+      "loss": 3.8883,
+      "step": 1830912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.001381715882788e-05,
+      "loss": 3.8976,
+      "step": 1831424
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.0110764503479,
+      "eval_runtime": 292.2673,
+      "eval_samples_per_second": 1305.623,
+      "eval_steps_per_second": 40.802,
+      "step": 1831680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.000543121131736e-05,
+      "loss": 3.8832,
+      "step": 1831936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.999704526380684e-05,
+      "loss": 3.8814,
+      "step": 1832448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.998865931629632e-05,
+      "loss": 3.8899,
+      "step": 1832960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.99802733687858e-05,
+      "loss": 3.896,
+      "step": 1833472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.997188742127528e-05,
+      "loss": 3.8998,
+      "step": 1833984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.996350147376476e-05,
+      "loss": 3.8859,
+      "step": 1834496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.995511552625424e-05,
+      "loss": 3.8946,
+      "step": 1835008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.994672957874372e-05,
+      "loss": 3.8806,
+      "step": 1835520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.99383436312332e-05,
+      "loss": 3.8864,
+      "step": 1836032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.992995768372268e-05,
+      "loss": 3.8893,
+      "step": 1836544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.992157173621216e-05,
+      "loss": 3.8921,
+      "step": 1837056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.991318578870164e-05,
+      "loss": 3.8937,
+      "step": 1837568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.990481621999485e-05,
+      "loss": 3.8932,
+      "step": 1838080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9896430272484333e-05,
+      "loss": 3.8821,
+      "step": 1838592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9888044324973813e-05,
+      "loss": 3.8831,
+      "step": 1839104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9879658377463293e-05,
+      "loss": 3.8781,
+      "step": 1839616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.987127242995277e-05,
+      "loss": 3.8893,
+      "step": 1840128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.986288648244225e-05,
+      "loss": 3.898,
+      "step": 1840640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.985450053493173e-05,
+      "loss": 3.8833,
+      "step": 1841152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.984611458742121e-05,
+      "loss": 3.9003,
+      "step": 1841664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9837761397518154e-05,
+      "loss": 3.9029,
+      "step": 1842176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9829375450007634e-05,
+      "loss": 3.8984,
+      "step": 1842688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9820989502497114e-05,
+      "loss": 3.8971,
+      "step": 1843200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9812603554986594e-05,
+      "loss": 3.8877,
+      "step": 1843712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.980421760747607e-05,
+      "loss": 3.89,
+      "step": 1844224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9795831659965554e-05,
+      "loss": 3.8971,
+      "step": 1844736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9787445712455034e-05,
+      "loss": 3.8941,
+      "step": 1845248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9779059764944514e-05,
+      "loss": 3.887,
+      "step": 1845760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9770673817433994e-05,
+      "loss": 3.8912,
+      "step": 1846272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9762287869923474e-05,
+      "loss": 3.8884,
+      "step": 1846784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.975390192241295e-05,
+      "loss": 3.8846,
+      "step": 1847296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.974551597490243e-05,
+      "loss": 3.8953,
+      "step": 1847808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9737162784999375e-05,
+      "loss": 3.8932,
+      "step": 1848320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9728776837488855e-05,
+      "loss": 3.8991,
+      "step": 1848832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9720390889978335e-05,
+      "loss": 3.8862,
+      "step": 1849344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9712021321271544e-05,
+      "loss": 3.897,
+      "step": 1849856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9703635373761024e-05,
+      "loss": 3.8917,
+      "step": 1850368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9695249426250508e-05,
+      "loss": 3.8741,
+      "step": 1850880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9686863478739988e-05,
+      "loss": 3.8835,
+      "step": 1851392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9678477531229468e-05,
+      "loss": 3.8819,
+      "step": 1851904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9670107962522677e-05,
+      "loss": 3.887,
+      "step": 1852416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9661722015012157e-05,
+      "loss": 3.8979,
+      "step": 1852928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9653336067501637e-05,
+      "loss": 3.8901,
+      "step": 1853440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9644950119991117e-05,
+      "loss": 3.881,
+      "step": 1853952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9636564172480597e-05,
+      "loss": 3.897,
+      "step": 1854464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9628178224970077e-05,
+      "loss": 3.8866,
+      "step": 1854976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9619792277459556e-05,
+      "loss": 3.9024,
+      "step": 1855488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9611406329949036e-05,
+      "loss": 3.8904,
+      "step": 1856000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9603020382438516e-05,
+      "loss": 3.8649,
+      "step": 1856512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9594634434927996e-05,
+      "loss": 3.9067,
+      "step": 1857024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9586248487417476e-05,
+      "loss": 3.8827,
+      "step": 1857536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.957787891871069e-05,
+      "loss": 3.8939,
+      "step": 1858048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.956949297120017e-05,
+      "loss": 3.8853,
+      "step": 1858560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9561123402493378e-05,
+      "loss": 3.8828,
+      "step": 1859072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9552737454982858e-05,
+      "loss": 3.8792,
+      "step": 1859584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9544351507472338e-05,
+      "loss": 3.8827,
+      "step": 1860096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9535965559961818e-05,
+      "loss": 3.9031,
+      "step": 1860608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.95275796124513e-05,
+      "loss": 3.8753,
+      "step": 1861120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.951921004374451e-05,
+      "loss": 3.8976,
+      "step": 1861632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.951082409623399e-05,
+      "loss": 3.8897,
+      "step": 1862144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.950243814872347e-05,
+      "loss": 3.8694,
+      "step": 1862656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.949405220121295e-05,
+      "loss": 3.9011,
+      "step": 1863168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9485666253702427e-05,
+      "loss": 3.8806,
+      "step": 1863680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.947728030619191e-05,
+      "loss": 3.8673,
+      "step": 1864192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9468910737485122e-05,
+      "loss": 3.8838,
+      "step": 1864704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9460524789974602e-05,
+      "loss": 3.8943,
+      "step": 1865216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.945213884246408e-05,
+      "loss": 3.8792,
+      "step": 1865728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.944375289495356e-05,
+      "loss": 3.8885,
+      "step": 1866240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.943536694744304e-05,
+      "loss": 3.8751,
+      "step": 1866752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.942698099993252e-05,
+      "loss": 3.879,
+      "step": 1867264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9418595052422002e-05,
+      "loss": 3.8904,
+      "step": 1867776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.941022548371521e-05,
+      "loss": 3.8904,
+      "step": 1868288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.940183953620469e-05,
+      "loss": 3.8898,
+      "step": 1868800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.939345358869417e-05,
+      "loss": 3.8851,
+      "step": 1869312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.938506764118365e-05,
+      "loss": 3.8991,
+      "step": 1869824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9376698072476864e-05,
+      "loss": 3.8807,
+      "step": 1870336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9368312124966344e-05,
+      "loss": 3.8877,
+      "step": 1870848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9359926177455823e-05,
+      "loss": 3.881,
+      "step": 1871360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9351540229945303e-05,
+      "loss": 3.881,
+      "step": 1871872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9343170661238513e-05,
+      "loss": 3.8878,
+      "step": 1872384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9334784713727992e-05,
+      "loss": 3.8825,
+      "step": 1872896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9326398766217472e-05,
+      "loss": 3.8949,
+      "step": 1873408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9318012818706956e-05,
+      "loss": 3.899,
+      "step": 1873920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9309626871196436e-05,
+      "loss": 3.8701,
+      "step": 1874432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9301240923685912e-05,
+      "loss": 3.8755,
+      "step": 1874944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9292854976175392e-05,
+      "loss": 3.8886,
+      "step": 1875456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9284469028664872e-05,
+      "loss": 3.8928,
+      "step": 1875968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9276083081154352e-05,
+      "loss": 3.8889,
+      "step": 1876480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9267697133643832e-05,
+      "loss": 3.8837,
+      "step": 1876992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9259311186133312e-05,
+      "loss": 3.8885,
+      "step": 1877504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9250941617426525e-05,
+      "loss": 3.8847,
+      "step": 1878016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9242555669916005e-05,
+      "loss": 3.8846,
+      "step": 1878528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9234169722405484e-05,
+      "loss": 3.8842,
+      "step": 1879040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9225783774894964e-05,
+      "loss": 3.8913,
+      "step": 1879552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.921739782738444e-05,
+      "loss": 3.8936,
+      "step": 1880064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9209011879873924e-05,
+      "loss": 3.8936,
+      "step": 1880576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9200625932363404e-05,
+      "loss": 3.8773,
+      "step": 1881088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9192239984852884e-05,
+      "loss": 3.8842,
+      "step": 1881600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9183870416146093e-05,
+      "loss": 3.8798,
+      "step": 1882112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9175484468635573e-05,
+      "loss": 3.888,
+      "step": 1882624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9167098521125053e-05,
+      "loss": 3.8891,
+      "step": 1883136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9158728952418266e-05,
+      "loss": 3.8782,
+      "step": 1883648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9150343004907746e-05,
+      "loss": 3.8866,
+      "step": 1884160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9141957057397226e-05,
+      "loss": 3.8748,
+      "step": 1884672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9133571109886706e-05,
+      "loss": 3.8823,
+      "step": 1885184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9125185162376186e-05,
+      "loss": 3.8739,
+      "step": 1885696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9116799214865666e-05,
+      "loss": 3.8836,
+      "step": 1886208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9108413267355145e-05,
+      "loss": 3.8809,
+      "step": 1886720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9100027319844625e-05,
+      "loss": 3.8893,
+      "step": 1887232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9091657751137838e-05,
+      "loss": 3.8769,
+      "step": 1887744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9083288182431047e-05,
+      "loss": 3.8764,
+      "step": 1888256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9074902234920527e-05,
+      "loss": 3.8814,
+      "step": 1888768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9066516287410007e-05,
+      "loss": 3.8898,
+      "step": 1889280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9058130339899487e-05,
+      "loss": 3.8831,
+      "step": 1889792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9049744392388967e-05,
+      "loss": 3.8877,
+      "step": 1890304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9041358444878447e-05,
+      "loss": 3.8857,
+      "step": 1890816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9032972497367927e-05,
+      "loss": 3.8987,
+      "step": 1891328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9024586549857407e-05,
+      "loss": 3.8846,
+      "step": 1891840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.901621698115062e-05,
+      "loss": 3.8833,
+      "step": 1892352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9007847412443828e-05,
+      "loss": 3.8843,
+      "step": 1892864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.899946146493331e-05,
+      "loss": 3.8922,
+      "step": 1893376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.899107551742279e-05,
+      "loss": 3.8839,
+      "step": 1893888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8982689569912268e-05,
+      "loss": 3.8794,
+      "step": 1894400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.897432000120548e-05,
+      "loss": 3.889,
+      "step": 1894912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.896593405369496e-05,
+      "loss": 3.8913,
+      "step": 1895424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.895754810618444e-05,
+      "loss": 3.8791,
+      "step": 1895936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.894916215867392e-05,
+      "loss": 3.8734,
+      "step": 1896448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.89407762111634e-05,
+      "loss": 3.8777,
+      "step": 1896960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8932406642456613e-05,
+      "loss": 3.8736,
+      "step": 1897472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8924020694946093e-05,
+      "loss": 3.89,
+      "step": 1897984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.891563474743557e-05,
+      "loss": 3.8915,
+      "step": 1898496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.890724879992505e-05,
+      "loss": 3.8949,
+      "step": 1899008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8898862852414533e-05,
+      "loss": 3.8708,
+      "step": 1899520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8890476904904013e-05,
+      "loss": 3.8954,
+      "step": 1900032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8882107336197222e-05,
+      "loss": 3.8867,
+      "step": 1900544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8873721388686702e-05,
+      "loss": 3.887,
+      "step": 1901056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8865335441176182e-05,
+      "loss": 3.8807,
+      "step": 1901568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8856949493665662e-05,
+      "loss": 3.8835,
+      "step": 1902080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.884856354615514e-05,
+      "loss": 3.8828,
+      "step": 1902592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.884017759864462e-05,
+      "loss": 3.8973,
+      "step": 1903104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.88317916511341e-05,
+      "loss": 3.8743,
+      "step": 1903616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.882340570362358e-05,
+      "loss": 3.8871,
+      "step": 1904128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8815036134916794e-05,
+      "loss": 3.8896,
+      "step": 1904640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8806650187406274e-05,
+      "loss": 3.8875,
+      "step": 1905152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8798280618699486e-05,
+      "loss": 3.8778,
+      "step": 1905664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8789894671188966e-05,
+      "loss": 3.8875,
+      "step": 1906176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8781508723678446e-05,
+      "loss": 3.891,
+      "step": 1906688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8773122776167923e-05,
+      "loss": 3.8844,
+      "step": 1907200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8764753207461135e-05,
+      "loss": 3.8924,
+      "step": 1907712
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.009130477905273,
+      "eval_runtime": 293.2046,
+      "eval_samples_per_second": 1301.45,
+      "eval_steps_per_second": 40.671,
+      "step": 1908000
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 7.639887324799606e+17,
+  "trial_name": null,
+  "trial_params": null
+}