diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,35443 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 20239,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7912962436676025,
+      "learning_rate": 4.9990118088838384e-05,
+      "loss": 10.3461,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3178839683532715,
+      "learning_rate": 4.9980236177676767e-05,
+      "loss": 9.7253,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.043405055999756,
+      "learning_rate": 4.997035426651515e-05,
+      "loss": 9.476,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0094900131225586,
+      "learning_rate": 4.9960472355353524e-05,
+      "loss": 9.2119,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7685112953186035,
+      "learning_rate": 4.9950590444191906e-05,
+      "loss": 9.0744,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.816589832305908,
+      "learning_rate": 4.994070853303029e-05,
+      "loss": 8.8707,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2649807929992676,
+      "learning_rate": 4.993082662186867e-05,
+      "loss": 8.677,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.543574571609497,
+      "learning_rate": 4.992094471070705e-05,
+      "loss": 8.5568,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.669053316116333,
+      "learning_rate": 4.9911062799545434e-05,
+      "loss": 8.2874,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4249765872955322,
+      "learning_rate": 4.9901180888383816e-05,
+      "loss": 8.2541,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1649515628814697,
+      "learning_rate": 4.98912989772222e-05,
+      "loss": 8.0488,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1681978702545166,
+      "learning_rate": 4.9881417066060574e-05,
+      "loss": 7.9646,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2630350589752197,
+      "learning_rate": 4.9871535154898956e-05,
+      "loss": 7.7305,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.286691904067993,
+      "learning_rate": 4.986165324373734e-05,
+      "loss": 7.5245,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9876737594604492,
+      "learning_rate": 4.9851771332575727e-05,
+      "loss": 7.4304,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7167022228240967,
+      "learning_rate": 4.984188942141411e-05,
+      "loss": 7.4031,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8193055391311646,
+      "learning_rate": 4.983200751025249e-05,
+      "loss": 7.3022,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.543209195137024,
+      "learning_rate": 4.9822125599090866e-05,
+      "loss": 7.2365,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.75989830493927,
+      "learning_rate": 4.981224368792925e-05,
+      "loss": 7.0086,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8119884729385376,
+      "learning_rate": 4.980236177676763e-05,
+      "loss": 7.1398,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4095029830932617,
+      "learning_rate": 4.979247986560601e-05,
+      "loss": 6.9534,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6824373006820679,
+      "learning_rate": 4.9782597954444394e-05,
+      "loss": 7.0708,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6116963624954224,
+      "learning_rate": 4.9772716043282776e-05,
+      "loss": 6.957,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2887672185897827,
+      "learning_rate": 4.976283413212116e-05,
+      "loss": 6.8803,
+      "step": 96
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2640583515167236,
+      "learning_rate": 4.9752952220959534e-05,
+      "loss": 6.6778,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3888378143310547,
+      "learning_rate": 4.9743070309797916e-05,
+      "loss": 6.7205,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4042987823486328,
+      "learning_rate": 4.97331883986363e-05,
+      "loss": 6.7003,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.531121015548706,
+      "learning_rate": 4.972330648747468e-05,
+      "loss": 6.7113,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5138936042785645,
+      "learning_rate": 4.971342457631306e-05,
+      "loss": 6.6541,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5610084533691406,
+      "learning_rate": 4.9703542665151444e-05,
+      "loss": 6.6966,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3988789319992065,
+      "learning_rate": 4.9693660753989826e-05,
+      "loss": 6.5739,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.511693000793457,
+      "learning_rate": 4.968377884282821e-05,
+      "loss": 6.6209,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4789738655090332,
+      "learning_rate": 4.967389693166658e-05,
+      "loss": 6.497,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.312217354774475,
+      "learning_rate": 4.9664015020504965e-05,
+      "loss": 6.4966,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2716199159622192,
+      "learning_rate": 4.965413310934335e-05,
+      "loss": 6.6663,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5896294116973877,
+      "learning_rate": 4.964425119818173e-05,
+      "loss": 6.7358,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.169236660003662,
+      "learning_rate": 4.963436928702011e-05,
+      "loss": 6.7237,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6182782649993896,
+      "learning_rate": 4.9624487375858494e-05,
+      "loss": 6.4832,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5328868627548218,
+      "learning_rate": 4.9614605464696876e-05,
+      "loss": 6.5984,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2593821287155151,
+      "learning_rate": 4.960472355353525e-05,
+      "loss": 6.5301,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7900025844573975,
+      "learning_rate": 4.959484164237363e-05,
+      "loss": 6.555,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1635518074035645,
+      "learning_rate": 4.9584959731212015e-05,
+      "loss": 6.5686,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4018573760986328,
+      "learning_rate": 4.95750778200504e-05,
+      "loss": 6.4757,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6255518198013306,
+      "learning_rate": 4.9565195908888786e-05,
+      "loss": 6.4482,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5604472160339355,
+      "learning_rate": 4.955531399772717e-05,
+      "loss": 6.448,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4451005458831787,
+      "learning_rate": 4.954543208656554e-05,
+      "loss": 6.5368,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.330167531967163,
+      "learning_rate": 4.9535550175403925e-05,
+      "loss": 6.473,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9696011543273926,
+      "learning_rate": 4.952566826424231e-05,
+      "loss": 6.4951,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7933244705200195,
+      "learning_rate": 4.951578635308069e-05,
+      "loss": 6.471,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6044100522994995,
+      "learning_rate": 4.950590444191907e-05,
+      "loss": 6.4612,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3888564109802246,
+      "learning_rate": 4.9496022530757454e-05,
+      "loss": 6.4879,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2492824792861938,
+      "learning_rate": 4.9486140619595836e-05,
+      "loss": 6.3364,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8178552389144897,
+      "learning_rate": 4.947625870843422e-05,
+      "loss": 6.5152,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.363503098487854,
+      "learning_rate": 4.946637679727259e-05,
+      "loss": 6.2865,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.319766640663147,
+      "learning_rate": 4.9456494886110975e-05,
+      "loss": 6.4775,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3917903900146484,
+      "learning_rate": 4.944661297494936e-05,
+      "loss": 6.4147,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.465641736984253,
+      "learning_rate": 4.943673106378774e-05,
+      "loss": 6.2991,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6081475019454956,
+      "learning_rate": 4.942684915262612e-05,
+      "loss": 6.442,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3391867876052856,
+      "learning_rate": 4.94169672414645e-05,
+      "loss": 6.3932,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4221361875534058,
+      "learning_rate": 4.9407085330302885e-05,
+      "loss": 6.5403,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.60570228099823,
+      "learning_rate": 4.939720341914126e-05,
+      "loss": 6.397,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3305834531784058,
+      "learning_rate": 4.938732150797964e-05,
+      "loss": 6.4258,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.617050290107727,
+      "learning_rate": 4.9377439596818025e-05,
+      "loss": 6.3044,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6071454286575317,
+      "learning_rate": 4.936755768565641e-05,
+      "loss": 6.3836,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0243899822235107,
+      "learning_rate": 4.935767577449479e-05,
+      "loss": 6.4785,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4023394584655762,
+      "learning_rate": 4.934779386333317e-05,
+      "loss": 6.4129,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6561346054077148,
+      "learning_rate": 4.933791195217155e-05,
+      "loss": 6.227,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4024381637573242,
+      "learning_rate": 4.932803004100993e-05,
+      "loss": 6.3957,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.576090693473816,
+      "learning_rate": 4.931814812984831e-05,
+      "loss": 6.359,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2421482801437378,
+      "learning_rate": 4.930826621868669e-05,
+      "loss": 6.3735,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.453856110572815,
+      "learning_rate": 4.9298384307525074e-05,
+      "loss": 6.4156,
+      "step": 284
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.426669955253601,
+      "learning_rate": 4.928850239636346e-05,
+      "loss": 6.3278,
+      "step": 288
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5175261497497559,
+      "learning_rate": 4.9278620485201845e-05,
+      "loss": 6.3645,
+      "step": 292
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3980427980422974,
+      "learning_rate": 4.926873857404023e-05,
+      "loss": 6.3977,
+      "step": 296
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6867117881774902,
+      "learning_rate": 4.92588566628786e-05,
+      "loss": 6.3506,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5795122385025024,
+      "learning_rate": 4.9248974751716985e-05,
+      "loss": 6.3062,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5149136781692505,
+      "learning_rate": 4.923909284055537e-05,
+      "loss": 6.302,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2091716527938843,
+      "learning_rate": 4.922921092939375e-05,
+      "loss": 6.2064,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.553070306777954,
+      "learning_rate": 4.921932901823213e-05,
+      "loss": 6.3236,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7350945472717285,
+      "learning_rate": 4.920944710707051e-05,
+      "loss": 6.3381,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2416383028030396,
+      "learning_rate": 4.9199565195908895e-05,
+      "loss": 6.4411,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4091325998306274,
+      "learning_rate": 4.918968328474727e-05,
+      "loss": 6.4273,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.261334776878357,
+      "learning_rate": 4.917980137358565e-05,
+      "loss": 6.2724,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6297985315322876,
+      "learning_rate": 4.9169919462424034e-05,
+      "loss": 6.2387,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.458993673324585,
+      "learning_rate": 4.9160037551262417e-05,
+      "loss": 6.3042,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.355808138847351,
+      "learning_rate": 4.91501556401008e-05,
+      "loss": 6.2788,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2890819311141968,
+      "learning_rate": 4.914027372893918e-05,
+      "loss": 6.3692,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5858672857284546,
+      "learning_rate": 4.913039181777756e-05,
+      "loss": 6.2324,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5306330919265747,
+      "learning_rate": 4.912050990661594e-05,
+      "loss": 6.4247,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.386789083480835,
+      "learning_rate": 4.911062799545432e-05,
+      "loss": 6.4219,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.288030743598938,
+      "learning_rate": 4.91007460842927e-05,
+      "loss": 6.2128,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5249669551849365,
+      "learning_rate": 4.9090864173131084e-05,
+      "loss": 6.2724,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.787862777709961,
+      "learning_rate": 4.9080982261969466e-05,
+      "loss": 6.3139,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7556533813476562,
+      "learning_rate": 4.907110035080785e-05,
+      "loss": 6.3006,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5250591039657593,
+      "learning_rate": 4.906121843964623e-05,
+      "loss": 6.3109,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.591663122177124,
+      "learning_rate": 4.905133652848461e-05,
+      "loss": 6.2026,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3261005878448486,
+      "learning_rate": 4.904145461732299e-05,
+      "loss": 6.2905,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.692723035812378,
+      "learning_rate": 4.903157270616137e-05,
+      "loss": 6.3296,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2827457189559937,
+      "learning_rate": 4.902169079499975e-05,
+      "loss": 6.229,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3109639883041382,
+      "learning_rate": 4.9011808883838134e-05,
+      "loss": 6.2625,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7134902477264404,
+      "learning_rate": 4.900192697267652e-05,
+      "loss": 6.1219,
+      "step": 404
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3591091632843018,
+      "learning_rate": 4.8992045061514905e-05,
+      "loss": 6.2059,
+      "step": 408
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4868909120559692,
+      "learning_rate": 4.898216315035328e-05,
+      "loss": 6.2244,
+      "step": 412
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4216508865356445,
+      "learning_rate": 4.897228123919166e-05,
+      "loss": 6.2348,
+      "step": 416
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6655827760696411,
+      "learning_rate": 4.8962399328030044e-05,
+      "loss": 6.2708,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2668567895889282,
+      "learning_rate": 4.8952517416868426e-05,
+      "loss": 6.1847,
+      "step": 424
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4796233177185059,
+      "learning_rate": 4.894263550570681e-05,
+      "loss": 6.2613,
+      "step": 428
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6667245626449585,
+      "learning_rate": 4.893275359454519e-05,
+      "loss": 6.3676,
+      "step": 432
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.780787467956543,
+      "learning_rate": 4.892287168338357e-05,
+      "loss": 6.341,
+      "step": 436
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.415461540222168,
+      "learning_rate": 4.891298977222195e-05,
+      "loss": 6.2441,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4248474836349487,
+      "learning_rate": 4.890310786106033e-05,
+      "loss": 6.282,
+      "step": 444
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0168416500091553,
+      "learning_rate": 4.889322594989871e-05,
+      "loss": 6.2457,
+      "step": 448
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.983055830001831,
+      "learning_rate": 4.8883344038737094e-05,
+      "loss": 6.1607,
+      "step": 452
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9059456586837769,
+      "learning_rate": 4.8873462127575476e-05,
+      "loss": 6.1443,
+      "step": 456
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5710817575454712,
+      "learning_rate": 4.886358021641386e-05,
+      "loss": 6.2631,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3912734985351562,
+      "learning_rate": 4.885369830525224e-05,
+      "loss": 6.2484,
+      "step": 464
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5551170110702515,
+      "learning_rate": 4.884381639409062e-05,
+      "loss": 6.3059,
+      "step": 468
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5288927555084229,
+      "learning_rate": 4.8833934482929e-05,
+      "loss": 6.0518,
+      "step": 472
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6663211584091187,
+      "learning_rate": 4.882405257176738e-05,
+      "loss": 6.2273,
+      "step": 476
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5548433065414429,
+      "learning_rate": 4.881417066060576e-05,
+      "loss": 6.1407,
+      "step": 480
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4325834512710571,
+      "learning_rate": 4.8804288749444144e-05,
+      "loss": 6.1346,
+      "step": 484
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.622158408164978,
+      "learning_rate": 4.8794406838282526e-05,
+      "loss": 6.1832,
+      "step": 488
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4777097702026367,
+      "learning_rate": 4.878452492712091e-05,
+      "loss": 6.2451,
+      "step": 492
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1623598337173462,
+      "learning_rate": 4.877464301595929e-05,
+      "loss": 6.1631,
+      "step": 496
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.598676323890686,
+      "learning_rate": 4.8764761104797665e-05,
+      "loss": 6.2348,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.610736608505249,
+      "learning_rate": 4.875487919363605e-05,
+      "loss": 6.3664,
+      "step": 504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4379361867904663,
+      "learning_rate": 4.874499728247443e-05,
+      "loss": 6.202,
+      "step": 508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3208808898925781,
+      "learning_rate": 4.873511537131281e-05,
+      "loss": 6.1286,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3087486028671265,
+      "learning_rate": 4.872523346015119e-05,
+      "loss": 6.2657,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3428252935409546,
+      "learning_rate": 4.871535154898958e-05,
+      "loss": 6.0752,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5278196334838867,
+      "learning_rate": 4.870546963782796e-05,
+      "loss": 6.267,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5938080549240112,
+      "learning_rate": 4.869558772666634e-05,
+      "loss": 6.177,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2772130966186523,
+      "learning_rate": 4.868570581550472e-05,
+      "loss": 6.2219,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4687914848327637,
+      "learning_rate": 4.8675823904343104e-05,
+      "loss": 6.2316,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4582056999206543,
+      "learning_rate": 4.8665941993181486e-05,
+      "loss": 6.0915,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3734614849090576,
+      "learning_rate": 4.865606008201987e-05,
+      "loss": 6.2694,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.426127314567566,
+      "learning_rate": 4.864617817085825e-05,
+      "loss": 6.1274,
+      "step": 548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3772287368774414,
+      "learning_rate": 4.863629625969663e-05,
+      "loss": 6.2704,
+      "step": 552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2581090927124023,
+      "learning_rate": 4.862641434853501e-05,
+      "loss": 6.1335,
+      "step": 556
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1419745683670044,
+      "learning_rate": 4.861653243737339e-05,
+      "loss": 6.1259,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2435179948806763,
+      "learning_rate": 4.860665052621177e-05,
+      "loss": 6.2289,
+      "step": 564
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5236430168151855,
+      "learning_rate": 4.859676861505015e-05,
+      "loss": 6.0294,
+      "step": 568
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.289353370666504,
+      "learning_rate": 4.8586886703888535e-05,
+      "loss": 6.1533,
+      "step": 572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5420417785644531,
+      "learning_rate": 4.857700479272692e-05,
+      "loss": 6.155,
+      "step": 576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8614426851272583,
+      "learning_rate": 4.85671228815653e-05,
+      "loss": 6.2381,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.405530571937561,
+      "learning_rate": 4.8557240970403675e-05,
+      "loss": 6.2005,
+      "step": 584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4058623313903809,
+      "learning_rate": 4.854735905924206e-05,
+      "loss": 6.1903,
+      "step": 588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7855048179626465,
+      "learning_rate": 4.853747714808044e-05,
+      "loss": 6.2939,
+      "step": 592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3504676818847656,
+      "learning_rate": 4.852759523691882e-05,
+      "loss": 6.0347,
+      "step": 596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3448097705841064,
+      "learning_rate": 4.85177133257572e-05,
+      "loss": 6.1058,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2908382415771484,
+      "learning_rate": 4.8507831414595585e-05,
+      "loss": 6.1323,
+      "step": 604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6946722269058228,
+      "learning_rate": 4.849794950343397e-05,
+      "loss": 6.0755,
+      "step": 608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0505646467208862,
+      "learning_rate": 4.848806759227235e-05,
+      "loss": 6.0742,
+      "step": 612
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.269295334815979,
+      "learning_rate": 4.8478185681110724e-05,
+      "loss": 6.2864,
+      "step": 616
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3444817066192627,
+      "learning_rate": 4.8468303769949107e-05,
+      "loss": 6.1599,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5646265745162964,
+      "learning_rate": 4.845842185878749e-05,
+      "loss": 6.1745,
+      "step": 624
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2452738285064697,
+      "learning_rate": 4.844853994762587e-05,
+      "loss": 6.1796,
+      "step": 628
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9167345762252808,
+      "learning_rate": 4.843865803646426e-05,
+      "loss": 6.1854,
+      "step": 632
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3464816808700562,
+      "learning_rate": 4.842877612530264e-05,
+      "loss": 6.1394,
+      "step": 636
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5352030992507935,
+      "learning_rate": 4.841889421414102e-05,
+      "loss": 6.1579,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2066044807434082,
+      "learning_rate": 4.84090123029794e-05,
+      "loss": 6.1284,
+      "step": 644
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3687278032302856,
+      "learning_rate": 4.839913039181778e-05,
+      "loss": 6.1768,
+      "step": 648
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4037103652954102,
+      "learning_rate": 4.838924848065616e-05,
+      "loss": 6.0552,
+      "step": 652
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8637155294418335,
+      "learning_rate": 4.8379366569494545e-05,
+      "loss": 6.126,
+      "step": 656
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8843746185302734,
+      "learning_rate": 4.836948465833293e-05,
+      "loss": 6.1839,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2985315322875977,
+      "learning_rate": 4.835960274717131e-05,
+      "loss": 6.0767,
+      "step": 664
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5230488777160645,
+      "learning_rate": 4.8349720836009684e-05,
+      "loss": 6.1762,
+      "step": 668
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8898823261260986,
+      "learning_rate": 4.8339838924848067e-05,
+      "loss": 6.0165,
+      "step": 672
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7203717231750488,
+      "learning_rate": 4.832995701368645e-05,
+      "loss": 6.1677,
+      "step": 676
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.478527307510376,
+      "learning_rate": 4.832007510252483e-05,
+      "loss": 6.2011,
+      "step": 680
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2331160306930542,
+      "learning_rate": 4.831019319136321e-05,
+      "loss": 6.1311,
+      "step": 684
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4134057760238647,
+      "learning_rate": 4.8300311280201595e-05,
+      "loss": 6.157,
+      "step": 688
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2819682359695435,
+      "learning_rate": 4.829042936903998e-05,
+      "loss": 6.0408,
+      "step": 692
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.646480917930603,
+      "learning_rate": 4.828054745787836e-05,
+      "loss": 6.0527,
+      "step": 696
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3179553747177124,
+      "learning_rate": 4.8270665546716734e-05,
+      "loss": 6.1771,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3149442672729492,
+      "learning_rate": 4.8260783635555116e-05,
+      "loss": 6.1106,
+      "step": 704
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.345340371131897,
+      "learning_rate": 4.82509017243935e-05,
+      "loss": 6.0963,
+      "step": 708
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3398826122283936,
+      "learning_rate": 4.824101981323188e-05,
+      "loss": 6.075,
+      "step": 712
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1765165328979492,
+      "learning_rate": 4.823113790207026e-05,
+      "loss": 6.1648,
+      "step": 716
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5758169889450073,
+      "learning_rate": 4.8221255990908645e-05,
+      "loss": 6.206,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3040266036987305,
+      "learning_rate": 4.8211374079747027e-05,
+      "loss": 6.0651,
+      "step": 724
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7494781017303467,
+      "learning_rate": 4.82014921685854e-05,
+      "loss": 6.0144,
+      "step": 728
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5416646003723145,
+      "learning_rate": 4.8191610257423784e-05,
+      "loss": 6.2011,
+      "step": 732
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3276399374008179,
+      "learning_rate": 4.8181728346262166e-05,
+      "loss": 6.2073,
+      "step": 736
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3996212482452393,
+      "learning_rate": 4.817184643510055e-05,
+      "loss": 6.0768,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3483880758285522,
+      "learning_rate": 4.816196452393893e-05,
+      "loss": 5.9309,
+      "step": 744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1830189228057861,
+      "learning_rate": 4.815208261277732e-05,
+      "loss": 6.0258,
+      "step": 748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3837305307388306,
+      "learning_rate": 4.8142200701615694e-05,
+      "loss": 6.0004,
+      "step": 752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1996792554855347,
+      "learning_rate": 4.8132318790454076e-05,
+      "loss": 6.067,
+      "step": 756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6578378677368164,
+      "learning_rate": 4.812243687929246e-05,
+      "loss": 6.1112,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.732629418373108,
+      "learning_rate": 4.811255496813084e-05,
+      "loss": 6.0361,
+      "step": 764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3690398931503296,
+      "learning_rate": 4.810267305696922e-05,
+      "loss": 6.0105,
+      "step": 768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.551343560218811,
+      "learning_rate": 4.8092791145807605e-05,
+      "loss": 5.9922,
+      "step": 772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.222483515739441,
+      "learning_rate": 4.8082909234645987e-05,
+      "loss": 6.1032,
+      "step": 776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4626158475875854,
+      "learning_rate": 4.807302732348437e-05,
+      "loss": 6.2455,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4593418836593628,
+      "learning_rate": 4.8063145412322744e-05,
+      "loss": 6.104,
+      "step": 784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4751644134521484,
+      "learning_rate": 4.8053263501161126e-05,
+      "loss": 6.0467,
+      "step": 788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.25166916847229,
+      "learning_rate": 4.804338158999951e-05,
+      "loss": 6.0525,
+      "step": 792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3588769435882568,
+      "learning_rate": 4.803349967883789e-05,
+      "loss": 6.1355,
+      "step": 796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2559925317764282,
+      "learning_rate": 4.802361776767627e-05,
+      "loss": 6.3282,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3895193338394165,
+      "learning_rate": 4.8013735856514654e-05,
+      "loss": 6.0936,
+      "step": 804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.691657543182373,
+      "learning_rate": 4.8003853945353036e-05,
+      "loss": 6.1079,
+      "step": 808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3020676374435425,
+      "learning_rate": 4.799397203419141e-05,
+      "loss": 6.0076,
+      "step": 812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.306980848312378,
+      "learning_rate": 4.7984090123029794e-05,
+      "loss": 6.1308,
+      "step": 816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3331334590911865,
+      "learning_rate": 4.7974208211868176e-05,
+      "loss": 6.0901,
+      "step": 820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2504887580871582,
+      "learning_rate": 4.796432630070656e-05,
+      "loss": 6.1079,
+      "step": 824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.462392807006836,
+      "learning_rate": 4.795444438954494e-05,
+      "loss": 6.0692,
+      "step": 828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.281559944152832,
+      "learning_rate": 4.794456247838332e-05,
+      "loss": 6.0399,
+      "step": 832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2926335334777832,
+      "learning_rate": 4.7934680567221704e-05,
+      "loss": 6.0005,
+      "step": 836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5699244737625122,
+      "learning_rate": 4.792479865606008e-05,
+      "loss": 5.9845,
+      "step": 840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.491740345954895,
+      "learning_rate": 4.791491674489846e-05,
+      "loss": 6.015,
+      "step": 844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3208236694335938,
+      "learning_rate": 4.790503483373684e-05,
+      "loss": 6.0501,
+      "step": 848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5497496128082275,
+      "learning_rate": 4.7895152922575225e-05,
+      "loss": 6.105,
+      "step": 852
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3851691484451294,
+      "learning_rate": 4.788527101141361e-05,
+      "loss": 6.1256,
+      "step": 856
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.506717324256897,
+      "learning_rate": 4.787538910025199e-05,
+      "loss": 5.9827,
+      "step": 860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2789814472198486,
+      "learning_rate": 4.786550718909038e-05,
+      "loss": 6.0632,
+      "step": 864
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2237069606781006,
+      "learning_rate": 4.7855625277928754e-05,
+      "loss": 6.0908,
+      "step": 868
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2778388261795044,
+      "learning_rate": 4.7845743366767136e-05,
+      "loss": 6.0442,
+      "step": 872
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.534972071647644,
+      "learning_rate": 4.783586145560552e-05,
+      "loss": 6.1035,
+      "step": 876
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4843637943267822,
+      "learning_rate": 4.78259795444439e-05,
+      "loss": 6.0233,
+      "step": 880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4751485586166382,
+      "learning_rate": 4.781609763328228e-05,
+      "loss": 6.109,
+      "step": 884
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.399499535560608,
+      "learning_rate": 4.7806215722120664e-05,
+      "loss": 6.0075,
+      "step": 888
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5269832611083984,
+      "learning_rate": 4.7796333810959046e-05,
+      "loss": 6.1387,
+      "step": 892
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.342552661895752,
+      "learning_rate": 4.778645189979742e-05,
+      "loss": 6.0763,
+      "step": 896
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2727460861206055,
+      "learning_rate": 4.77765699886358e-05,
+      "loss": 5.9279,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3465628623962402,
+      "learning_rate": 4.7766688077474185e-05,
+      "loss": 6.0226,
+      "step": 904
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.288557767868042,
+      "learning_rate": 4.775680616631257e-05,
+      "loss": 6.0799,
+      "step": 908
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.375345230102539,
+      "learning_rate": 4.774692425515095e-05,
+      "loss": 6.0868,
+      "step": 912
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.440955638885498,
+      "learning_rate": 4.773704234398933e-05,
+      "loss": 5.9835,
+      "step": 916
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3870395421981812,
+      "learning_rate": 4.7727160432827714e-05,
+      "loss": 6.0711,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4793189764022827,
+      "learning_rate": 4.771727852166609e-05,
+      "loss": 6.1001,
+      "step": 924
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2727391719818115,
+      "learning_rate": 4.770739661050447e-05,
+      "loss": 6.0782,
+      "step": 928
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3459304571151733,
+      "learning_rate": 4.769751469934285e-05,
+      "loss": 5.9659,
+      "step": 932
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6304296255111694,
+      "learning_rate": 4.7687632788181235e-05,
+      "loss": 6.0666,
+      "step": 936
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.721566081047058,
+      "learning_rate": 4.767775087701962e-05,
+      "loss": 6.0233,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.486752986907959,
+      "learning_rate": 4.7667868965858e-05,
+      "loss": 6.0544,
+      "step": 944
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.404415488243103,
+      "learning_rate": 4.765798705469638e-05,
+      "loss": 5.8913,
+      "step": 948
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.550802230834961,
+      "learning_rate": 4.764810514353476e-05,
+      "loss": 6.0449,
+      "step": 952
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5222808122634888,
+      "learning_rate": 4.763822323237314e-05,
+      "loss": 6.1524,
+      "step": 956
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2582489252090454,
+      "learning_rate": 4.762834132121152e-05,
+      "loss": 6.09,
+      "step": 960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5326069593429565,
+      "learning_rate": 4.76184594100499e-05,
+      "loss": 6.0265,
+      "step": 964
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5986106395721436,
+      "learning_rate": 4.7608577498888285e-05,
+      "loss": 6.1111,
+      "step": 968
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2689043283462524,
+      "learning_rate": 4.759869558772667e-05,
+      "loss": 6.0827,
+      "step": 972
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2669328451156616,
+      "learning_rate": 4.758881367656505e-05,
+      "loss": 6.1393,
+      "step": 976
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2476458549499512,
+      "learning_rate": 4.757893176540343e-05,
+      "loss": 6.0072,
+      "step": 980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4797428846359253,
+      "learning_rate": 4.756904985424181e-05,
+      "loss": 6.0025,
+      "step": 984
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9927531480789185,
+      "learning_rate": 4.7559167943080195e-05,
+      "loss": 6.087,
+      "step": 988
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6228762865066528,
+      "learning_rate": 4.754928603191858e-05,
+      "loss": 6.0355,
+      "step": 992
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5881098508834839,
+      "learning_rate": 4.753940412075696e-05,
+      "loss": 6.0462,
+      "step": 996
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.944541335105896,
+      "learning_rate": 4.752952220959534e-05,
+      "loss": 6.1294,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6126629114151,
+      "learning_rate": 4.751964029843372e-05,
+      "loss": 6.1586,
+      "step": 1004
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4057198762893677,
+      "learning_rate": 4.75097583872721e-05,
+      "loss": 6.24,
+      "step": 1008
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3622597455978394,
+      "learning_rate": 4.749987647611048e-05,
+      "loss": 5.9785,
+      "step": 1012
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3643813133239746,
+      "learning_rate": 4.748999456494886e-05,
+      "loss": 6.0392,
+      "step": 1016
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.793847918510437,
+      "learning_rate": 4.7480112653787245e-05,
+      "loss": 5.9712,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4792604446411133,
+      "learning_rate": 4.747023074262563e-05,
+      "loss": 6.0705,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.35064697265625,
+      "learning_rate": 4.746034883146401e-05,
+      "loss": 6.0608,
+      "step": 1028
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3092408180236816,
+      "learning_rate": 4.745046692030239e-05,
+      "loss": 5.885,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.560230016708374,
+      "learning_rate": 4.744058500914077e-05,
+      "loss": 6.1149,
+      "step": 1036
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5619248151779175,
+      "learning_rate": 4.743070309797915e-05,
+      "loss": 6.1887,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4519466161727905,
+      "learning_rate": 4.742082118681753e-05,
+      "loss": 6.0853,
+      "step": 1044
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4475151300430298,
+      "learning_rate": 4.741093927565591e-05,
+      "loss": 6.0921,
+      "step": 1048
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2523047924041748,
+      "learning_rate": 4.7401057364494295e-05,
+      "loss": 6.1132,
+      "step": 1052
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2426884174346924,
+      "learning_rate": 4.7391175453332677e-05,
+      "loss": 5.9939,
+      "step": 1056
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4118828773498535,
+      "learning_rate": 4.738129354217106e-05,
+      "loss": 6.082,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6150298118591309,
+      "learning_rate": 4.737141163100944e-05,
+      "loss": 6.0342,
+      "step": 1064
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3286446332931519,
+      "learning_rate": 4.7361529719847816e-05,
+      "loss": 5.9456,
+      "step": 1068
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4334193468093872,
+      "learning_rate": 4.73516478086862e-05,
+      "loss": 5.9225,
+      "step": 1072
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2666079998016357,
+      "learning_rate": 4.734176589752458e-05,
+      "loss": 6.0389,
+      "step": 1076
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2868504524230957,
+      "learning_rate": 4.733188398636296e-05,
+      "loss": 6.0719,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7467691898345947,
+      "learning_rate": 4.7322002075201344e-05,
+      "loss": 5.9916,
+      "step": 1084
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4924054145812988,
+      "learning_rate": 4.7312120164039726e-05,
+      "loss": 5.9197,
+      "step": 1088
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1824862957000732,
+      "learning_rate": 4.730223825287811e-05,
+      "loss": 5.9122,
+      "step": 1092
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.435604214668274,
+      "learning_rate": 4.729235634171649e-05,
+      "loss": 6.1722,
+      "step": 1096
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6537317037582397,
+      "learning_rate": 4.728247443055487e-05,
+      "loss": 6.0557,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.237634539604187,
+      "learning_rate": 4.7272592519393255e-05,
+      "loss": 6.1475,
+      "step": 1104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2861469984054565,
+      "learning_rate": 4.7262710608231637e-05,
+      "loss": 6.0261,
+      "step": 1108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3722236156463623,
+      "learning_rate": 4.725282869707002e-05,
+      "loss": 5.9994,
+      "step": 1112
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4114540815353394,
+      "learning_rate": 4.72429467859084e-05,
+      "loss": 5.923,
+      "step": 1116
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.572077751159668,
+      "learning_rate": 4.723306487474678e-05,
+      "loss": 5.9886,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.971572995185852,
+      "learning_rate": 4.722318296358516e-05,
+      "loss": 6.0632,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1920833587646484,
+      "learning_rate": 4.721330105242354e-05,
+      "loss": 6.066,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6061211824417114,
+      "learning_rate": 4.720341914126192e-05,
+      "loss": 5.9523,
+      "step": 1132
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.617601990699768,
+      "learning_rate": 4.7193537230100304e-05,
+      "loss": 6.0579,
+      "step": 1136
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.674851655960083,
+      "learning_rate": 4.7183655318938686e-05,
+      "loss": 5.869,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6465020179748535,
+      "learning_rate": 4.717377340777707e-05,
+      "loss": 5.9525,
+      "step": 1144
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4357322454452515,
+      "learning_rate": 4.716389149661545e-05,
+      "loss": 5.8798,
+      "step": 1148
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.319442868232727,
+      "learning_rate": 4.7154009585453826e-05,
+      "loss": 6.0924,
+      "step": 1152
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7044557332992554,
+      "learning_rate": 4.714412767429221e-05,
+      "loss": 6.0367,
+      "step": 1156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6891183853149414,
+      "learning_rate": 4.713424576313059e-05,
+      "loss": 5.9983,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.340993046760559,
+      "learning_rate": 4.712436385196897e-05,
+      "loss": 6.0353,
+      "step": 1164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3144937753677368,
+      "learning_rate": 4.7114481940807354e-05,
+      "loss": 5.9603,
+      "step": 1168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9154691696166992,
+      "learning_rate": 4.7104600029645736e-05,
+      "loss": 6.057,
+      "step": 1172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2976912260055542,
+      "learning_rate": 4.709471811848412e-05,
+      "loss": 5.992,
+      "step": 1176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4527958631515503,
+      "learning_rate": 4.70848362073225e-05,
+      "loss": 5.9253,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6319037675857544,
+      "learning_rate": 4.7074954296160875e-05,
+      "loss": 6.1028,
+      "step": 1184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7407358884811401,
+      "learning_rate": 4.706507238499926e-05,
+      "loss": 5.9943,
+      "step": 1188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.127054452896118,
+      "learning_rate": 4.705519047383764e-05,
+      "loss": 6.0707,
+      "step": 1192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.761857271194458,
+      "learning_rate": 4.704530856267602e-05,
+      "loss": 6.0144,
+      "step": 1196
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4145355224609375,
+      "learning_rate": 4.7035426651514404e-05,
+      "loss": 6.0829,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3968502283096313,
+      "learning_rate": 4.7025544740352786e-05,
+      "loss": 6.0907,
+      "step": 1204
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.406941294670105,
+      "learning_rate": 4.701566282919117e-05,
+      "loss": 5.9521,
+      "step": 1208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3803069591522217,
+      "learning_rate": 4.700578091802955e-05,
+      "loss": 5.9801,
+      "step": 1212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3069123029708862,
+      "learning_rate": 4.699589900686793e-05,
+      "loss": 5.9288,
+      "step": 1216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5246793031692505,
+      "learning_rate": 4.6986017095706314e-05,
+      "loss": 6.0128,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5078657865524292,
+      "learning_rate": 4.6976135184544696e-05,
+      "loss": 5.9037,
+      "step": 1224
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9630528688430786,
+      "learning_rate": 4.696625327338308e-05,
+      "loss": 5.9762,
+      "step": 1228
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4446896314620972,
+      "learning_rate": 4.695637136222146e-05,
+      "loss": 5.9254,
+      "step": 1232
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1959322690963745,
+      "learning_rate": 4.6946489451059835e-05,
+      "loss": 5.9263,
+      "step": 1236
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5779132843017578,
+      "learning_rate": 4.693660753989822e-05,
+      "loss": 5.9132,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3477551937103271,
+      "learning_rate": 4.69267256287366e-05,
+      "loss": 6.0072,
+      "step": 1244
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.950527310371399,
+      "learning_rate": 4.691684371757498e-05,
+      "loss": 5.9209,
+      "step": 1248
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0284175872802734,
+      "learning_rate": 4.6906961806413364e-05,
+      "loss": 6.0581,
+      "step": 1252
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3213205337524414,
+      "learning_rate": 4.6897079895251746e-05,
+      "loss": 5.9722,
+      "step": 1256
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2748960256576538,
+      "learning_rate": 4.688719798409013e-05,
+      "loss": 6.0285,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4361499547958374,
+      "learning_rate": 4.687731607292851e-05,
+      "loss": 5.9448,
+      "step": 1264
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5793545246124268,
+      "learning_rate": 4.6867434161766885e-05,
+      "loss": 6.0026,
+      "step": 1268
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7214800119400024,
+      "learning_rate": 4.685755225060527e-05,
+      "loss": 6.1118,
+      "step": 1272
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3730213642120361,
+      "learning_rate": 4.684767033944365e-05,
+      "loss": 5.9988,
+      "step": 1276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4589751958847046,
+      "learning_rate": 4.683778842828203e-05,
+      "loss": 6.0041,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3940002918243408,
+      "learning_rate": 4.682790651712041e-05,
+      "loss": 5.9818,
+      "step": 1284
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3168179988861084,
+      "learning_rate": 4.6818024605958795e-05,
+      "loss": 5.9819,
+      "step": 1288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.737804651260376,
+      "learning_rate": 4.680814269479718e-05,
+      "loss": 5.9301,
+      "step": 1292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2958004474639893,
+      "learning_rate": 4.679826078363555e-05,
+      "loss": 5.9922,
+      "step": 1296
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7017489671707153,
+      "learning_rate": 4.6788378872473935e-05,
+      "loss": 6.0218,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7935067415237427,
+      "learning_rate": 4.677849696131232e-05,
+      "loss": 5.8878,
+      "step": 1304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.293447256088257,
+      "learning_rate": 4.67686150501507e-05,
+      "loss": 6.1183,
+      "step": 1308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.470645785331726,
+      "learning_rate": 4.675873313898908e-05,
+      "loss": 5.8939,
+      "step": 1312
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3084948062896729,
+      "learning_rate": 4.674885122782746e-05,
+      "loss": 5.9267,
+      "step": 1316
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2316402196884155,
+      "learning_rate": 4.6738969316665845e-05,
+      "loss": 6.0664,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3546851873397827,
+      "learning_rate": 4.672908740550423e-05,
+      "loss": 5.9785,
+      "step": 1324
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1932753324508667,
+      "learning_rate": 4.671920549434261e-05,
+      "loss": 5.8807,
+      "step": 1328
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3659350872039795,
+      "learning_rate": 4.670932358318099e-05,
+      "loss": 6.0175,
+      "step": 1332
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.316653847694397,
+      "learning_rate": 4.669944167201937e-05,
+      "loss": 5.9124,
+      "step": 1336
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4429857730865479,
+      "learning_rate": 4.6689559760857755e-05,
+      "loss": 5.9828,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4865456819534302,
+      "learning_rate": 4.667967784969614e-05,
+      "loss": 6.0829,
+      "step": 1344
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.453019618988037,
+      "learning_rate": 4.666979593853452e-05,
+      "loss": 5.9101,
+      "step": 1348
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2942475080490112,
+      "learning_rate": 4.6659914027372895e-05,
+      "loss": 6.0094,
+      "step": 1352
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6731159687042236,
+      "learning_rate": 4.665003211621128e-05,
+      "loss": 5.9274,
+      "step": 1356
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5828598737716675,
+      "learning_rate": 4.664015020504966e-05,
+      "loss": 6.0568,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3310965299606323,
+      "learning_rate": 4.663026829388804e-05,
+      "loss": 5.9873,
+      "step": 1364
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5734776258468628,
+      "learning_rate": 4.662038638272642e-05,
+      "loss": 5.9194,
+      "step": 1368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8752501010894775,
+      "learning_rate": 4.6610504471564805e-05,
+      "loss": 5.945,
+      "step": 1372
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5967419147491455,
+      "learning_rate": 4.660062256040319e-05,
+      "loss": 5.8828,
+      "step": 1376
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3059850931167603,
+      "learning_rate": 4.659074064924156e-05,
+      "loss": 5.9692,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3276833295822144,
+      "learning_rate": 4.6580858738079945e-05,
+      "loss": 5.9998,
+      "step": 1384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.481294870376587,
+      "learning_rate": 4.6570976826918327e-05,
+      "loss": 6.0054,
+      "step": 1388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2492045164108276,
+      "learning_rate": 4.656109491575671e-05,
+      "loss": 5.7969,
+      "step": 1392
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3145158290863037,
+      "learning_rate": 4.655121300459509e-05,
+      "loss": 6.0625,
+      "step": 1396
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7083948850631714,
+      "learning_rate": 4.654133109343347e-05,
+      "loss": 6.0563,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5010440349578857,
+      "learning_rate": 4.6531449182271855e-05,
+      "loss": 5.9032,
+      "step": 1404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3582074642181396,
+      "learning_rate": 4.652156727111023e-05,
+      "loss": 5.9933,
+      "step": 1408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4667145013809204,
+      "learning_rate": 4.651168535994861e-05,
+      "loss": 5.8992,
+      "step": 1412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5343049764633179,
+      "learning_rate": 4.6501803448786994e-05,
+      "loss": 5.9115,
+      "step": 1416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1951204538345337,
+      "learning_rate": 4.6491921537625376e-05,
+      "loss": 5.8802,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.566027283668518,
+      "learning_rate": 4.648203962646376e-05,
+      "loss": 5.8612,
+      "step": 1424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2942713499069214,
+      "learning_rate": 4.647215771530214e-05,
+      "loss": 5.936,
+      "step": 1428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5581068992614746,
+      "learning_rate": 4.646227580414052e-05,
+      "loss": 5.8286,
+      "step": 1432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3115367889404297,
+      "learning_rate": 4.6452393892978905e-05,
+      "loss": 5.9605,
+      "step": 1436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3432234525680542,
+      "learning_rate": 4.6442511981817287e-05,
+      "loss": 5.8929,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.789021611213684,
+      "learning_rate": 4.643263007065567e-05,
+      "loss": 6.0817,
+      "step": 1444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8558330535888672,
+      "learning_rate": 4.642274815949405e-05,
+      "loss": 6.0362,
+      "step": 1448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2429139614105225,
+      "learning_rate": 4.641286624833243e-05,
+      "loss": 6.0426,
+      "step": 1452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4920710325241089,
+      "learning_rate": 4.6402984337170815e-05,
+      "loss": 5.8848,
+      "step": 1456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.456693172454834,
+      "learning_rate": 4.63931024260092e-05,
+      "loss": 6.0129,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2984966039657593,
+      "learning_rate": 4.638322051484757e-05,
+      "loss": 5.9807,
+      "step": 1464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3711044788360596,
+      "learning_rate": 4.6373338603685954e-05,
+      "loss": 5.9158,
+      "step": 1468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.432022213935852,
+      "learning_rate": 4.6363456692524336e-05,
+      "loss": 6.0205,
+      "step": 1472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.549072504043579,
+      "learning_rate": 4.635357478136272e-05,
+      "loss": 5.9811,
+      "step": 1476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.531638741493225,
+      "learning_rate": 4.63436928702011e-05,
+      "loss": 5.9643,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9268430471420288,
+      "learning_rate": 4.633381095903948e-05,
+      "loss": 6.022,
+      "step": 1484
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.491971731185913,
+      "learning_rate": 4.6323929047877865e-05,
+      "loss": 5.9582,
+      "step": 1488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.253057837486267,
+      "learning_rate": 4.631404713671624e-05,
+      "loss": 5.9783,
+      "step": 1492
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7390155792236328,
+      "learning_rate": 4.630416522555462e-05,
+      "loss": 5.9987,
+      "step": 1496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.605100154876709,
+      "learning_rate": 4.6294283314393004e-05,
+      "loss": 6.0706,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.30108642578125,
+      "learning_rate": 4.6284401403231386e-05,
+      "loss": 6.0712,
+      "step": 1504
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7230815887451172,
+      "learning_rate": 4.627451949206977e-05,
+      "loss": 6.0353,
+      "step": 1508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7896109819412231,
+      "learning_rate": 4.626463758090815e-05,
+      "loss": 5.9901,
+      "step": 1512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.710947871208191,
+      "learning_rate": 4.625475566974653e-05,
+      "loss": 5.9141,
+      "step": 1516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5355948209762573,
+      "learning_rate": 4.6244873758584914e-05,
+      "loss": 5.9802,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4352582693099976,
+      "learning_rate": 4.623499184742329e-05,
+      "loss": 6.0397,
+      "step": 1524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4116641283035278,
+      "learning_rate": 4.622510993626167e-05,
+      "loss": 6.0213,
+      "step": 1528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2762186527252197,
+      "learning_rate": 4.6215228025100054e-05,
+      "loss": 5.8323,
+      "step": 1532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8467053174972534,
+      "learning_rate": 4.6205346113938436e-05,
+      "loss": 5.8623,
+      "step": 1536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5691808462142944,
+      "learning_rate": 4.619546420277682e-05,
+      "loss": 6.0056,
+      "step": 1540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9999240636825562,
+      "learning_rate": 4.61855822916152e-05,
+      "loss": 6.1042,
+      "step": 1544
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9879581928253174,
+      "learning_rate": 4.617570038045358e-05,
+      "loss": 5.8972,
+      "step": 1548
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6930572986602783,
+      "learning_rate": 4.6165818469291964e-05,
+      "loss": 5.8287,
+      "step": 1552
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.064516544342041,
+      "learning_rate": 4.6155936558130346e-05,
+      "loss": 5.986,
+      "step": 1556
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5297229290008545,
+      "learning_rate": 4.614605464696873e-05,
+      "loss": 6.0793,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3446491956710815,
+      "learning_rate": 4.613617273580711e-05,
+      "loss": 5.971,
+      "step": 1564
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3736571073532104,
+      "learning_rate": 4.612629082464549e-05,
+      "loss": 6.049,
+      "step": 1568
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.439300298690796,
+      "learning_rate": 4.6116408913483874e-05,
+      "loss": 6.0757,
+      "step": 1572
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.766449213027954,
+      "learning_rate": 4.610652700232225e-05,
+      "loss": 5.8482,
+      "step": 1576
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3712018728256226,
+      "learning_rate": 4.609664509116063e-05,
+      "loss": 5.9108,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6353716850280762,
+      "learning_rate": 4.6086763179999014e-05,
+      "loss": 5.9461,
+      "step": 1584
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.549001932144165,
+      "learning_rate": 4.6076881268837396e-05,
+      "loss": 5.9503,
+      "step": 1588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8643873929977417,
+      "learning_rate": 4.606699935767578e-05,
+      "loss": 5.8876,
+      "step": 1592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1906590461730957,
+      "learning_rate": 4.605711744651416e-05,
+      "loss": 5.8155,
+      "step": 1596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4951763153076172,
+      "learning_rate": 4.604723553535254e-05,
+      "loss": 6.0645,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5761932134628296,
+      "learning_rate": 4.6037353624190924e-05,
+      "loss": 5.8113,
+      "step": 1604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.858544111251831,
+      "learning_rate": 4.60274717130293e-05,
+      "loss": 5.9492,
+      "step": 1608
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2514946460723877,
+      "learning_rate": 4.601758980186768e-05,
+      "loss": 6.0415,
+      "step": 1612
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4306968450546265,
+      "learning_rate": 4.600770789070606e-05,
+      "loss": 6.0258,
+      "step": 1616
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4032137393951416,
+      "learning_rate": 4.5997825979544445e-05,
+      "loss": 5.9323,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.424999713897705,
+      "learning_rate": 4.598794406838283e-05,
+      "loss": 5.937,
+      "step": 1624
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3444308042526245,
+      "learning_rate": 4.597806215722121e-05,
+      "loss": 5.8967,
+      "step": 1628
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4193394184112549,
+      "learning_rate": 4.596818024605959e-05,
+      "loss": 5.9547,
+      "step": 1632
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2748864889144897,
+      "learning_rate": 4.595829833489797e-05,
+      "loss": 5.9361,
+      "step": 1636
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.469801664352417,
+      "learning_rate": 4.594841642373635e-05,
+      "loss": 5.8623,
+      "step": 1640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.563642978668213,
+      "learning_rate": 4.593853451257473e-05,
+      "loss": 6.0724,
+      "step": 1644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.504905104637146,
+      "learning_rate": 4.592865260141311e-05,
+      "loss": 5.9072,
+      "step": 1648
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.543295979499817,
+      "learning_rate": 4.5918770690251495e-05,
+      "loss": 5.9527,
+      "step": 1652
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.244002103805542,
+      "learning_rate": 4.590888877908988e-05,
+      "loss": 6.0182,
+      "step": 1656
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3683278560638428,
+      "learning_rate": 4.589900686792826e-05,
+      "loss": 6.022,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5250694751739502,
+      "learning_rate": 4.588912495676664e-05,
+      "loss": 5.9414,
+      "step": 1664
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.940976619720459,
+      "learning_rate": 4.5879243045605023e-05,
+      "loss": 5.9301,
+      "step": 1668
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.503818154335022,
+      "learning_rate": 4.5869361134443405e-05,
+      "loss": 5.8361,
+      "step": 1672
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5099518299102783,
+      "learning_rate": 4.585947922328179e-05,
+      "loss": 5.8945,
+      "step": 1676
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6731679439544678,
+      "learning_rate": 4.584959731212017e-05,
+      "loss": 5.9091,
+      "step": 1680
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2417348623275757,
+      "learning_rate": 4.583971540095855e-05,
+      "loss": 5.9407,
+      "step": 1684
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4417905807495117,
+      "learning_rate": 4.5829833489796934e-05,
+      "loss": 6.0244,
+      "step": 1688
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8846923112869263,
+      "learning_rate": 4.581995157863531e-05,
+      "loss": 5.9297,
+      "step": 1692
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6629383563995361,
+      "learning_rate": 4.581006966747369e-05,
+      "loss": 5.845,
+      "step": 1696
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4970371723175049,
+      "learning_rate": 4.580018775631207e-05,
+      "loss": 5.9948,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.547971248626709,
+      "learning_rate": 4.5790305845150455e-05,
+      "loss": 5.9322,
+      "step": 1704
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2521891593933105,
+      "learning_rate": 4.578042393398884e-05,
+      "loss": 5.9448,
+      "step": 1708
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8124247789382935,
+      "learning_rate": 4.577054202282722e-05,
+      "loss": 5.9226,
+      "step": 1712
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.33345890045166,
+      "learning_rate": 4.57606601116656e-05,
+      "loss": 6.0771,
+      "step": 1716
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.122553825378418,
+      "learning_rate": 4.575077820050398e-05,
+      "loss": 5.8754,
+      "step": 1720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5283949375152588,
+      "learning_rate": 4.574089628934236e-05,
+      "loss": 5.8644,
+      "step": 1724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6564345359802246,
+      "learning_rate": 4.573101437818074e-05,
+      "loss": 5.8691,
+      "step": 1728
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6714503765106201,
+      "learning_rate": 4.572113246701912e-05,
+      "loss": 5.9227,
+      "step": 1732
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.543869972229004,
+      "learning_rate": 4.5711250555857505e-05,
+      "loss": 5.9588,
+      "step": 1736
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.248793601989746,
+      "learning_rate": 4.570136864469589e-05,
+      "loss": 5.9065,
+      "step": 1740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3636749982833862,
+      "learning_rate": 4.569148673353427e-05,
+      "loss": 5.8001,
+      "step": 1744
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2990843057632446,
+      "learning_rate": 4.568160482237265e-05,
+      "loss": 5.9017,
+      "step": 1748
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4160529375076294,
+      "learning_rate": 4.5671722911211026e-05,
+      "loss": 5.7955,
+      "step": 1752
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5701916217803955,
+      "learning_rate": 4.566184100004941e-05,
+      "loss": 5.9576,
+      "step": 1756
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5187253952026367,
+      "learning_rate": 4.565195908888779e-05,
+      "loss": 5.9054,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6530662775039673,
+      "learning_rate": 4.564207717772617e-05,
+      "loss": 5.9385,
+      "step": 1764
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.481063723564148,
+      "learning_rate": 4.5632195266564555e-05,
+      "loss": 5.8584,
+      "step": 1768
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2179416418075562,
+      "learning_rate": 4.562231335540294e-05,
+      "loss": 5.9652,
+      "step": 1772
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4156241416931152,
+      "learning_rate": 4.561243144424132e-05,
+      "loss": 5.7772,
+      "step": 1776
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5597620010375977,
+      "learning_rate": 4.56025495330797e-05,
+      "loss": 5.9441,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4729033708572388,
+      "learning_rate": 4.559266762191808e-05,
+      "loss": 5.9495,
+      "step": 1784
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4252592325210571,
+      "learning_rate": 4.5582785710756465e-05,
+      "loss": 5.8379,
+      "step": 1788
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3946348428726196,
+      "learning_rate": 4.557290379959485e-05,
+      "loss": 5.9435,
+      "step": 1792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.366830587387085,
+      "learning_rate": 4.556302188843323e-05,
+      "loss": 5.9755,
+      "step": 1796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2371442317962646,
+      "learning_rate": 4.555313997727161e-05,
+      "loss": 5.9288,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3472713232040405,
+      "learning_rate": 4.5543258066109986e-05,
+      "loss": 5.9537,
+      "step": 1804
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9578423500061035,
+      "learning_rate": 4.553337615494837e-05,
+      "loss": 6.0531,
+      "step": 1808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4482135772705078,
+      "learning_rate": 4.552349424378675e-05,
+      "loss": 5.8764,
+      "step": 1812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4886963367462158,
+      "learning_rate": 4.551361233262513e-05,
+      "loss": 6.0265,
+      "step": 1816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.89201819896698,
+      "learning_rate": 4.5503730421463515e-05,
+      "loss": 5.8338,
+      "step": 1820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.207160234451294,
+      "learning_rate": 4.54938485103019e-05,
+      "loss": 5.8131,
+      "step": 1824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4633376598358154,
+      "learning_rate": 4.548396659914028e-05,
+      "loss": 5.8665,
+      "step": 1828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4453964233398438,
+      "learning_rate": 4.547408468797866e-05,
+      "loss": 5.963,
+      "step": 1832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0406370162963867,
+      "learning_rate": 4.5464202776817036e-05,
+      "loss": 5.8779,
+      "step": 1836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6090114116668701,
+      "learning_rate": 4.545432086565542e-05,
+      "loss": 5.906,
+      "step": 1840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6481086015701294,
+      "learning_rate": 4.54444389544938e-05,
+      "loss": 6.0581,
+      "step": 1844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5279346704483032,
+      "learning_rate": 4.543455704333218e-05,
+      "loss": 5.937,
+      "step": 1848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.580449104309082,
+      "learning_rate": 4.5424675132170564e-05,
+      "loss": 5.8628,
+      "step": 1852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5227487087249756,
+      "learning_rate": 4.5414793221008946e-05,
+      "loss": 5.8597,
+      "step": 1856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3211569786071777,
+      "learning_rate": 4.540491130984733e-05,
+      "loss": 5.9294,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5065453052520752,
+      "learning_rate": 4.5395029398685704e-05,
+      "loss": 5.9311,
+      "step": 1864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6894686222076416,
+      "learning_rate": 4.5385147487524086e-05,
+      "loss": 5.9489,
+      "step": 1868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4861186742782593,
+      "learning_rate": 4.537526557636247e-05,
+      "loss": 5.9022,
+      "step": 1872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7374814748764038,
+      "learning_rate": 4.536538366520085e-05,
+      "loss": 5.9511,
+      "step": 1876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7111232280731201,
+      "learning_rate": 4.535550175403923e-05,
+      "loss": 5.9008,
+      "step": 1880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.887349009513855,
+      "learning_rate": 4.5345619842877614e-05,
+      "loss": 5.7785,
+      "step": 1884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.397018313407898,
+      "learning_rate": 4.5335737931715996e-05,
+      "loss": 5.8256,
+      "step": 1888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.312874674797058,
+      "learning_rate": 4.532585602055438e-05,
+      "loss": 5.7768,
+      "step": 1892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6667051315307617,
+      "learning_rate": 4.531597410939276e-05,
+      "loss": 5.9521,
+      "step": 1896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3993628025054932,
+      "learning_rate": 4.530609219823114e-05,
+      "loss": 5.8285,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5168507099151611,
+      "learning_rate": 4.5296210287069524e-05,
+      "loss": 5.9211,
+      "step": 1904
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9311803579330444,
+      "learning_rate": 4.5286328375907906e-05,
+      "loss": 6.015,
+      "step": 1908
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.33588707447052,
+      "learning_rate": 4.527644646474629e-05,
+      "loss": 5.8528,
+      "step": 1912
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.831453561782837,
+      "learning_rate": 4.526656455358467e-05,
+      "loss": 5.8254,
+      "step": 1916
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7024390697479248,
+      "learning_rate": 4.5256682642423046e-05,
+      "loss": 5.8458,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5297564268112183,
+      "learning_rate": 4.524680073126143e-05,
+      "loss": 5.8749,
+      "step": 1924
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4288631677627563,
+      "learning_rate": 4.523691882009981e-05,
+      "loss": 5.8288,
+      "step": 1928
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.723061203956604,
+      "learning_rate": 4.522703690893819e-05,
+      "loss": 5.8773,
+      "step": 1932
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3699047565460205,
+      "learning_rate": 4.5217154997776574e-05,
+      "loss": 5.94,
+      "step": 1936
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.883584976196289,
+      "learning_rate": 4.5207273086614956e-05,
+      "loss": 5.8437,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6371957063674927,
+      "learning_rate": 4.519739117545334e-05,
+      "loss": 5.9447,
+      "step": 1944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4362295866012573,
+      "learning_rate": 4.5187509264291713e-05,
+      "loss": 6.0218,
+      "step": 1948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9110528230667114,
+      "learning_rate": 4.5177627353130095e-05,
+      "loss": 5.9644,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.587083101272583,
+      "learning_rate": 4.516774544196848e-05,
+      "loss": 5.9193,
+      "step": 1956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5505852699279785,
+      "learning_rate": 4.515786353080686e-05,
+      "loss": 5.8284,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5493978261947632,
+      "learning_rate": 4.514798161964524e-05,
+      "loss": 6.0078,
+      "step": 1964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5519258975982666,
+      "learning_rate": 4.5138099708483624e-05,
+      "loss": 5.8343,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4279431104660034,
+      "learning_rate": 4.5128217797322006e-05,
+      "loss": 5.9317,
+      "step": 1972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8912272453308105,
+      "learning_rate": 4.511833588616038e-05,
+      "loss": 5.8355,
+      "step": 1976
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.440687656402588,
+      "learning_rate": 4.510845397499876e-05,
+      "loss": 5.9027,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8350549936294556,
+      "learning_rate": 4.5098572063837145e-05,
+      "loss": 5.9534,
+      "step": 1984
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4987351894378662,
+      "learning_rate": 4.508869015267553e-05,
+      "loss": 6.0618,
+      "step": 1988
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4456908702850342,
+      "learning_rate": 4.507880824151391e-05,
+      "loss": 5.9174,
+      "step": 1992
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5561717748641968,
+      "learning_rate": 4.506892633035229e-05,
+      "loss": 5.8602,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.976826548576355,
+      "learning_rate": 4.5059044419190673e-05,
+      "loss": 5.97,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.561604380607605,
+      "learning_rate": 4.5049162508029055e-05,
+      "loss": 5.9561,
+      "step": 2004
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.241724967956543,
+      "learning_rate": 4.503928059686743e-05,
+      "loss": 5.9314,
+      "step": 2008
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.012200355529785,
+      "learning_rate": 4.502939868570582e-05,
+      "loss": 5.8255,
+      "step": 2012
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6281402111053467,
+      "learning_rate": 4.50195167745442e-05,
+      "loss": 5.8459,
+      "step": 2016
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9645634889602661,
+      "learning_rate": 4.5009634863382584e-05,
+      "loss": 5.805,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7598261833190918,
+      "learning_rate": 4.4999752952220966e-05,
+      "loss": 5.7636,
+      "step": 2024
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8579994440078735,
+      "learning_rate": 4.498987104105935e-05,
+      "loss": 5.9409,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3588138818740845,
+      "learning_rate": 4.497998912989772e-05,
+      "loss": 5.8235,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.471695065498352,
+      "learning_rate": 4.4970107218736105e-05,
+      "loss": 5.9183,
+      "step": 2036
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.500862717628479,
+      "learning_rate": 4.496022530757449e-05,
+      "loss": 5.8246,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3388968706130981,
+      "learning_rate": 4.495034339641287e-05,
+      "loss": 5.9709,
+      "step": 2044
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5488966703414917,
+      "learning_rate": 4.494046148525125e-05,
+      "loss": 5.8914,
+      "step": 2048
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5115787982940674,
+      "learning_rate": 4.4930579574089633e-05,
+      "loss": 5.8664,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.632888674736023,
+      "learning_rate": 4.4920697662928015e-05,
+      "loss": 5.9276,
+      "step": 2056
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3317750692367554,
+      "learning_rate": 4.491081575176639e-05,
+      "loss": 5.8649,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7422010898590088,
+      "learning_rate": 4.490093384060477e-05,
+      "loss": 5.8982,
+      "step": 2064
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.052384614944458,
+      "learning_rate": 4.4891051929443155e-05,
+      "loss": 5.9476,
+      "step": 2068
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.492372989654541,
+      "learning_rate": 4.488117001828154e-05,
+      "loss": 5.8919,
+      "step": 2072
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3981868028640747,
+      "learning_rate": 4.487128810711992e-05,
+      "loss": 5.9484,
+      "step": 2076
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7076584100723267,
+      "learning_rate": 4.48614061959583e-05,
+      "loss": 5.7792,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4366710186004639,
+      "learning_rate": 4.485152428479668e-05,
+      "loss": 5.8862,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2299433946609497,
+      "learning_rate": 4.4841642373635065e-05,
+      "loss": 5.7836,
+      "step": 2088
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4261624813079834,
+      "learning_rate": 4.483176046247344e-05,
+      "loss": 5.9149,
+      "step": 2092
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5158166885375977,
+      "learning_rate": 4.482187855131182e-05,
+      "loss": 5.8517,
+      "step": 2096
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3522121906280518,
+      "learning_rate": 4.4811996640150205e-05,
+      "loss": 5.8754,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.343929648399353,
+      "learning_rate": 4.480211472898859e-05,
+      "loss": 5.771,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6901476383209229,
+      "learning_rate": 4.479223281782697e-05,
+      "loss": 5.8675,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.438178539276123,
+      "learning_rate": 4.478235090666535e-05,
+      "loss": 5.9775,
+      "step": 2112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2448184490203857,
+      "learning_rate": 4.477246899550373e-05,
+      "loss": 5.861,
+      "step": 2116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3768398761749268,
+      "learning_rate": 4.476258708434211e-05,
+      "loss": 5.8436,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5526635646820068,
+      "learning_rate": 4.475270517318049e-05,
+      "loss": 5.8113,
+      "step": 2124
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0470657348632812,
+      "learning_rate": 4.474282326201888e-05,
+      "loss": 5.8941,
+      "step": 2128
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.404012680053711,
+      "learning_rate": 4.473294135085726e-05,
+      "loss": 5.9352,
+      "step": 2132
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3800268173217773,
+      "learning_rate": 4.472305943969564e-05,
+      "loss": 5.9212,
+      "step": 2136
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8839443922042847,
+      "learning_rate": 4.4713177528534025e-05,
+      "loss": 5.8908,
+      "step": 2140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8185203075408936,
+      "learning_rate": 4.47032956173724e-05,
+      "loss": 5.7962,
+      "step": 2144
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6154285669326782,
+      "learning_rate": 4.469341370621078e-05,
+      "loss": 5.8918,
+      "step": 2148
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3525383472442627,
+      "learning_rate": 4.4683531795049165e-05,
+      "loss": 5.8528,
+      "step": 2152
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7467882633209229,
+      "learning_rate": 4.467364988388755e-05,
+      "loss": 5.8305,
+      "step": 2156
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7372554540634155,
+      "learning_rate": 4.466376797272593e-05,
+      "loss": 5.7785,
+      "step": 2160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7468005418777466,
+      "learning_rate": 4.465388606156431e-05,
+      "loss": 6.0067,
+      "step": 2164
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.874570608139038,
+      "learning_rate": 4.464400415040269e-05,
+      "loss": 5.9846,
+      "step": 2168
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5136998891830444,
+      "learning_rate": 4.4634122239241075e-05,
+      "loss": 5.8196,
+      "step": 2172
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5951347351074219,
+      "learning_rate": 4.462424032807945e-05,
+      "loss": 5.8991,
+      "step": 2176
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9109349250793457,
+      "learning_rate": 4.461435841691783e-05,
+      "loss": 5.8357,
+      "step": 2180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5938990116119385,
+      "learning_rate": 4.4604476505756214e-05,
+      "loss": 5.7961,
+      "step": 2184
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.457306981086731,
+      "learning_rate": 4.4594594594594596e-05,
+      "loss": 5.8715,
+      "step": 2188
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.478857398033142,
+      "learning_rate": 4.458471268343298e-05,
+      "loss": 5.9143,
+      "step": 2192
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.413000464439392,
+      "learning_rate": 4.457483077227136e-05,
+      "loss": 5.8109,
+      "step": 2196
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.348897099494934,
+      "learning_rate": 4.456494886110974e-05,
+      "loss": 5.8774,
+      "step": 2200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4481480121612549,
+      "learning_rate": 4.455506694994812e-05,
+      "loss": 5.7836,
+      "step": 2204
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.077380895614624,
+      "learning_rate": 4.45451850387865e-05,
+      "loss": 5.8215,
+      "step": 2208
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5154703855514526,
+      "learning_rate": 4.453530312762488e-05,
+      "loss": 5.8485,
+      "step": 2212
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3549476861953735,
+      "learning_rate": 4.4525421216463264e-05,
+      "loss": 5.9692,
+      "step": 2216
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3713443279266357,
+      "learning_rate": 4.4515539305301646e-05,
+      "loss": 5.8609,
+      "step": 2220
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3348569869995117,
+      "learning_rate": 4.450565739414003e-05,
+      "loss": 5.831,
+      "step": 2224
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4860191345214844,
+      "learning_rate": 4.449577548297841e-05,
+      "loss": 6.0256,
+      "step": 2228
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3161065578460693,
+      "learning_rate": 4.448589357181679e-05,
+      "loss": 5.9999,
+      "step": 2232
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8355900049209595,
+      "learning_rate": 4.447601166065517e-05,
+      "loss": 5.8118,
+      "step": 2236
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4929054975509644,
+      "learning_rate": 4.4466129749493556e-05,
+      "loss": 5.9227,
+      "step": 2240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6608703136444092,
+      "learning_rate": 4.445624783833194e-05,
+      "loss": 5.8468,
+      "step": 2244
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5917619466781616,
+      "learning_rate": 4.444636592717032e-05,
+      "loss": 5.9577,
+      "step": 2248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5820611715316772,
+      "learning_rate": 4.44364840160087e-05,
+      "loss": 5.8065,
+      "step": 2252
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6974551677703857,
+      "learning_rate": 4.4426602104847085e-05,
+      "loss": 5.963,
+      "step": 2256
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5884884595870972,
+      "learning_rate": 4.441672019368546e-05,
+      "loss": 5.7997,
+      "step": 2260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5706701278686523,
+      "learning_rate": 4.440683828252384e-05,
+      "loss": 5.8355,
+      "step": 2264
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5051394701004028,
+      "learning_rate": 4.4396956371362224e-05,
+      "loss": 5.9357,
+      "step": 2268
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6415704488754272,
+      "learning_rate": 4.4387074460200606e-05,
+      "loss": 5.7755,
+      "step": 2272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9928330183029175,
+      "learning_rate": 4.437719254903899e-05,
+      "loss": 6.0354,
+      "step": 2276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.568490982055664,
+      "learning_rate": 4.436731063787737e-05,
+      "loss": 5.9754,
+      "step": 2280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4449853897094727,
+      "learning_rate": 4.435742872671575e-05,
+      "loss": 5.939,
+      "step": 2284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.227518081665039,
+      "learning_rate": 4.434754681555413e-05,
+      "loss": 5.8591,
+      "step": 2288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7351478338241577,
+      "learning_rate": 4.433766490439251e-05,
+      "loss": 5.8574,
+      "step": 2292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8627262115478516,
+      "learning_rate": 4.432778299323089e-05,
+      "loss": 5.9735,
+      "step": 2296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.56145179271698,
+      "learning_rate": 4.4317901082069274e-05,
+      "loss": 5.9248,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5302497148513794,
+      "learning_rate": 4.4308019170907656e-05,
+      "loss": 5.8347,
+      "step": 2304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.597585678100586,
+      "learning_rate": 4.429813725974604e-05,
+      "loss": 5.7803,
+      "step": 2308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.529463768005371,
+      "learning_rate": 4.428825534858442e-05,
+      "loss": 5.8968,
+      "step": 2312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4933751821517944,
+      "learning_rate": 4.42783734374228e-05,
+      "loss": 5.9021,
+      "step": 2316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6076233386993408,
+      "learning_rate": 4.426849152626118e-05,
+      "loss": 5.8318,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7479904890060425,
+      "learning_rate": 4.425860961509956e-05,
+      "loss": 5.8686,
+      "step": 2324
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5784941911697388,
+      "learning_rate": 4.424872770393794e-05,
+      "loss": 5.8552,
+      "step": 2328
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.12491512298584,
+      "learning_rate": 4.4238845792776323e-05,
+      "loss": 5.9495,
+      "step": 2332
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6172250509262085,
+      "learning_rate": 4.4228963881614705e-05,
+      "loss": 5.7925,
+      "step": 2336
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4207360744476318,
+      "learning_rate": 4.421908197045309e-05,
+      "loss": 5.8818,
+      "step": 2340
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.528517246246338,
+      "learning_rate": 4.420920005929147e-05,
+      "loss": 5.848,
+      "step": 2344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4071799516677856,
+      "learning_rate": 4.4199318148129845e-05,
+      "loss": 5.7346,
+      "step": 2348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.543596863746643,
+      "learning_rate": 4.418943623696823e-05,
+      "loss": 5.87,
+      "step": 2352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.697926640510559,
+      "learning_rate": 4.4179554325806616e-05,
+      "loss": 5.8042,
+      "step": 2356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8649811744689941,
+      "learning_rate": 4.4169672414645e-05,
+      "loss": 5.8247,
+      "step": 2360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.496875286102295,
+      "learning_rate": 4.415979050348338e-05,
+      "loss": 5.9044,
+      "step": 2364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.469252586364746,
+      "learning_rate": 4.414990859232176e-05,
+      "loss": 5.8298,
+      "step": 2368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4524880647659302,
+      "learning_rate": 4.414002668116014e-05,
+      "loss": 5.8173,
+      "step": 2372
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2573550939559937,
+      "learning_rate": 4.413014476999852e-05,
+      "loss": 5.8825,
+      "step": 2376
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3610568046569824,
+      "learning_rate": 4.41202628588369e-05,
+      "loss": 5.9205,
+      "step": 2380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3914682865142822,
+      "learning_rate": 4.4110380947675283e-05,
+      "loss": 5.8465,
+      "step": 2384
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.628902792930603,
+      "learning_rate": 4.4100499036513665e-05,
+      "loss": 5.9335,
+      "step": 2388
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1272637844085693,
+      "learning_rate": 4.409061712535205e-05,
+      "loss": 5.8199,
+      "step": 2392
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.497086524963379,
+      "learning_rate": 4.408073521419043e-05,
+      "loss": 5.7442,
+      "step": 2396
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6062419414520264,
+      "learning_rate": 4.407085330302881e-05,
+      "loss": 5.7353,
+      "step": 2400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6324589252471924,
+      "learning_rate": 4.406097139186719e-05,
+      "loss": 5.8851,
+      "step": 2404
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7234266996383667,
+      "learning_rate": 4.405108948070557e-05,
+      "loss": 5.9936,
+      "step": 2408
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7293999195098877,
+      "learning_rate": 4.404120756954395e-05,
+      "loss": 5.8189,
+      "step": 2412
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.554434895515442,
+      "learning_rate": 4.403132565838233e-05,
+      "loss": 5.8536,
+      "step": 2416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.773876428604126,
+      "learning_rate": 4.4021443747220715e-05,
+      "loss": 5.808,
+      "step": 2420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.341245174407959,
+      "learning_rate": 4.40115618360591e-05,
+      "loss": 5.785,
+      "step": 2424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4888144731521606,
+      "learning_rate": 4.400167992489748e-05,
+      "loss": 5.7738,
+      "step": 2428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.269622564315796,
+      "learning_rate": 4.3991798013735855e-05,
+      "loss": 5.7985,
+      "step": 2432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2829160690307617,
+      "learning_rate": 4.398191610257424e-05,
+      "loss": 5.9229,
+      "step": 2436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5251401662826538,
+      "learning_rate": 4.397203419141262e-05,
+      "loss": 5.967,
+      "step": 2440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8288617134094238,
+      "learning_rate": 4.3962152280251e-05,
+      "loss": 5.8184,
+      "step": 2444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5593011379241943,
+      "learning_rate": 4.395227036908938e-05,
+      "loss": 5.9089,
+      "step": 2448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4598174095153809,
+      "learning_rate": 4.3942388457927765e-05,
+      "loss": 5.8489,
+      "step": 2452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.364039421081543,
+      "learning_rate": 4.393250654676615e-05,
+      "loss": 5.9755,
+      "step": 2456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.418669581413269,
+      "learning_rate": 4.392262463560453e-05,
+      "loss": 5.8468,
+      "step": 2460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7515896558761597,
+      "learning_rate": 4.3912742724442904e-05,
+      "loss": 6.0253,
+      "step": 2464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6947784423828125,
+      "learning_rate": 4.3902860813281286e-05,
+      "loss": 5.8654,
+      "step": 2468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.597745656967163,
+      "learning_rate": 4.3892978902119675e-05,
+      "loss": 5.78,
+      "step": 2472
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9116005897521973,
+      "learning_rate": 4.388309699095806e-05,
+      "loss": 5.871,
+      "step": 2476
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5638071298599243,
+      "learning_rate": 4.387321507979644e-05,
+      "loss": 5.8569,
+      "step": 2480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.321582794189453,
+      "learning_rate": 4.386333316863482e-05,
+      "loss": 5.7426,
+      "step": 2484
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4646013975143433,
+      "learning_rate": 4.38534512574732e-05,
+      "loss": 5.8682,
+      "step": 2488
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.777219533920288,
+      "learning_rate": 4.384356934631158e-05,
+      "loss": 5.778,
+      "step": 2492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7113094329833984,
+      "learning_rate": 4.383368743514996e-05,
+      "loss": 5.8171,
+      "step": 2496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6728547811508179,
+      "learning_rate": 4.382380552398834e-05,
+      "loss": 5.8537,
+      "step": 2500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5846984386444092,
+      "learning_rate": 4.3813923612826725e-05,
+      "loss": 5.9563,
+      "step": 2504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.277194023132324,
+      "learning_rate": 4.380404170166511e-05,
+      "loss": 5.9298,
+      "step": 2508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1459007263183594,
+      "learning_rate": 4.379415979050349e-05,
+      "loss": 5.9069,
+      "step": 2512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9131797552108765,
+      "learning_rate": 4.3784277879341864e-05,
+      "loss": 5.8698,
+      "step": 2516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5482133626937866,
+      "learning_rate": 4.3774395968180246e-05,
+      "loss": 5.8516,
+      "step": 2520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3043674230575562,
+      "learning_rate": 4.376451405701863e-05,
+      "loss": 5.8176,
+      "step": 2524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6391806602478027,
+      "learning_rate": 4.375463214585701e-05,
+      "loss": 5.7712,
+      "step": 2528
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7865711450576782,
+      "learning_rate": 4.374475023469539e-05,
+      "loss": 5.9933,
+      "step": 2532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6309571266174316,
+      "learning_rate": 4.3734868323533775e-05,
+      "loss": 5.7993,
+      "step": 2536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9978580474853516,
+      "learning_rate": 4.372498641237216e-05,
+      "loss": 5.8303,
+      "step": 2540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.507794976234436,
+      "learning_rate": 4.371510450121053e-05,
+      "loss": 5.7701,
+      "step": 2544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4123347997665405,
+      "learning_rate": 4.3705222590048914e-05,
+      "loss": 5.9114,
+      "step": 2548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.467590570449829,
+      "learning_rate": 4.3695340678887296e-05,
+      "loss": 5.893,
+      "step": 2552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4081734418869019,
+      "learning_rate": 4.368545876772568e-05,
+      "loss": 5.8997,
+      "step": 2556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4744434356689453,
+      "learning_rate": 4.367557685656406e-05,
+      "loss": 5.7802,
+      "step": 2560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.405860185623169,
+      "learning_rate": 4.366569494540244e-05,
+      "loss": 5.9679,
+      "step": 2564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8207783699035645,
+      "learning_rate": 4.3655813034240824e-05,
+      "loss": 5.9814,
+      "step": 2568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5685409307479858,
+      "learning_rate": 4.3645931123079206e-05,
+      "loss": 5.9399,
+      "step": 2572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.218668818473816,
+      "learning_rate": 4.363604921191758e-05,
+      "loss": 5.7926,
+      "step": 2576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3519160747528076,
+      "learning_rate": 4.3626167300755964e-05,
+      "loss": 5.7725,
+      "step": 2580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0880026817321777,
+      "learning_rate": 4.361628538959435e-05,
+      "loss": 5.9561,
+      "step": 2584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6406432390213013,
+      "learning_rate": 4.3606403478432735e-05,
+      "loss": 5.8689,
+      "step": 2588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4885075092315674,
+      "learning_rate": 4.359652156727112e-05,
+      "loss": 5.7307,
+      "step": 2592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5429643392562866,
+      "learning_rate": 4.35866396561095e-05,
+      "loss": 5.8006,
+      "step": 2596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7911728620529175,
+      "learning_rate": 4.3576757744947874e-05,
+      "loss": 5.8635,
+      "step": 2600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0220367908477783,
+      "learning_rate": 4.3566875833786256e-05,
+      "loss": 5.8923,
+      "step": 2604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5731260776519775,
+      "learning_rate": 4.355699392262464e-05,
+      "loss": 5.9229,
+      "step": 2608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0656206607818604,
+      "learning_rate": 4.354711201146302e-05,
+      "loss": 5.9055,
+      "step": 2612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5131778717041016,
+      "learning_rate": 4.35372301003014e-05,
+      "loss": 5.8296,
+      "step": 2616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.953755497932434,
+      "learning_rate": 4.3527348189139784e-05,
+      "loss": 5.7472,
+      "step": 2620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.496201515197754,
+      "learning_rate": 4.3517466277978166e-05,
+      "loss": 5.8703,
+      "step": 2624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6766773462295532,
+      "learning_rate": 4.350758436681654e-05,
+      "loss": 5.8507,
+      "step": 2628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5544074773788452,
+      "learning_rate": 4.3497702455654924e-05,
+      "loss": 5.7533,
+      "step": 2632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.169536828994751,
+      "learning_rate": 4.3487820544493306e-05,
+      "loss": 5.9255,
+      "step": 2636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.78980553150177,
+      "learning_rate": 4.347793863333169e-05,
+      "loss": 5.8978,
+      "step": 2640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4349106550216675,
+      "learning_rate": 4.346805672217007e-05,
+      "loss": 5.7736,
+      "step": 2644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5959937572479248,
+      "learning_rate": 4.345817481100845e-05,
+      "loss": 5.8393,
+      "step": 2648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2932758331298828,
+      "learning_rate": 4.3448292899846834e-05,
+      "loss": 6.0105,
+      "step": 2652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8422064781188965,
+      "learning_rate": 4.3438410988685216e-05,
+      "loss": 5.7793,
+      "step": 2656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.948499083518982,
+      "learning_rate": 4.342852907752359e-05,
+      "loss": 5.9639,
+      "step": 2660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4169591665267944,
+      "learning_rate": 4.3418647166361973e-05,
+      "loss": 5.8024,
+      "step": 2664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4500057697296143,
+      "learning_rate": 4.3408765255200356e-05,
+      "loss": 5.8555,
+      "step": 2668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5837527513504028,
+      "learning_rate": 4.339888334403874e-05,
+      "loss": 5.9194,
+      "step": 2672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8204997777938843,
+      "learning_rate": 4.338900143287712e-05,
+      "loss": 5.9046,
+      "step": 2676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7072649002075195,
+      "learning_rate": 4.33791195217155e-05,
+      "loss": 5.7805,
+      "step": 2680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0868546962738037,
+      "learning_rate": 4.3369237610553884e-05,
+      "loss": 5.8899,
+      "step": 2684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.489852786064148,
+      "learning_rate": 4.335935569939226e-05,
+      "loss": 5.8553,
+      "step": 2688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7851141691207886,
+      "learning_rate": 4.334947378823064e-05,
+      "loss": 5.7927,
+      "step": 2692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4655989408493042,
+      "learning_rate": 4.333959187706902e-05,
+      "loss": 5.9079,
+      "step": 2696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4320762157440186,
+      "learning_rate": 4.332970996590741e-05,
+      "loss": 5.9267,
+      "step": 2700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6071585416793823,
+      "learning_rate": 4.3319828054745794e-05,
+      "loss": 5.7468,
+      "step": 2704
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3550992012023926,
+      "learning_rate": 4.3309946143584176e-05,
+      "loss": 5.8188,
+      "step": 2708
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1684658527374268,
+      "learning_rate": 4.330006423242255e-05,
+      "loss": 5.8959,
+      "step": 2712
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.543056607246399,
+      "learning_rate": 4.3290182321260933e-05,
+      "loss": 5.9079,
+      "step": 2716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4535815715789795,
+      "learning_rate": 4.3280300410099316e-05,
+      "loss": 5.814,
+      "step": 2720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6012686491012573,
+      "learning_rate": 4.32704184989377e-05,
+      "loss": 5.9343,
+      "step": 2724
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.63141667842865,
+      "learning_rate": 4.326053658777608e-05,
+      "loss": 5.744,
+      "step": 2728
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.621268630027771,
+      "learning_rate": 4.325065467661446e-05,
+      "loss": 5.6779,
+      "step": 2732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.809212565422058,
+      "learning_rate": 4.3240772765452844e-05,
+      "loss": 5.8384,
+      "step": 2736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3587805032730103,
+      "learning_rate": 4.3230890854291226e-05,
+      "loss": 5.7724,
+      "step": 2740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7285997867584229,
+      "learning_rate": 4.32210089431296e-05,
+      "loss": 5.8804,
+      "step": 2744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9753797054290771,
+      "learning_rate": 4.321112703196798e-05,
+      "loss": 5.8618,
+      "step": 2748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.357673168182373,
+      "learning_rate": 4.3201245120806365e-05,
+      "loss": 5.6777,
+      "step": 2752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5860601663589478,
+      "learning_rate": 4.319136320964475e-05,
+      "loss": 5.7696,
+      "step": 2756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.830790638923645,
+      "learning_rate": 4.318148129848313e-05,
+      "loss": 5.7376,
+      "step": 2760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4875571727752686,
+      "learning_rate": 4.317159938732151e-05,
+      "loss": 5.8386,
+      "step": 2764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5012223720550537,
+      "learning_rate": 4.3161717476159893e-05,
+      "loss": 5.8406,
+      "step": 2768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.872456669807434,
+      "learning_rate": 4.315183556499827e-05,
+      "loss": 5.7977,
+      "step": 2772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4684159755706787,
+      "learning_rate": 4.314195365383665e-05,
+      "loss": 5.6846,
+      "step": 2776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8385950326919556,
+      "learning_rate": 4.313207174267503e-05,
+      "loss": 5.9036,
+      "step": 2780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6362018585205078,
+      "learning_rate": 4.3122189831513415e-05,
+      "loss": 5.8381,
+      "step": 2784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5058547258377075,
+      "learning_rate": 4.31123079203518e-05,
+      "loss": 5.8647,
+      "step": 2788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.060148239135742,
+      "learning_rate": 4.310242600919018e-05,
+      "loss": 5.7431,
+      "step": 2792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9664667844772339,
+      "learning_rate": 4.309254409802856e-05,
+      "loss": 5.8879,
+      "step": 2796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9889500141143799,
+      "learning_rate": 4.308266218686694e-05,
+      "loss": 5.8715,
+      "step": 2800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3932100534439087,
+      "learning_rate": 4.307278027570532e-05,
+      "loss": 5.6858,
+      "step": 2804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6131174564361572,
+      "learning_rate": 4.30628983645437e-05,
+      "loss": 5.9675,
+      "step": 2808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9435076713562012,
+      "learning_rate": 4.305301645338208e-05,
+      "loss": 5.8918,
+      "step": 2812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5480365753173828,
+      "learning_rate": 4.304313454222047e-05,
+      "loss": 5.7797,
+      "step": 2816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5039936304092407,
+      "learning_rate": 4.3033252631058853e-05,
+      "loss": 5.7075,
+      "step": 2820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4041931629180908,
+      "learning_rate": 4.3023370719897236e-05,
+      "loss": 5.6871,
+      "step": 2824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9822477102279663,
+      "learning_rate": 4.301348880873561e-05,
+      "loss": 5.8225,
+      "step": 2828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6121152639389038,
+      "learning_rate": 4.300360689757399e-05,
+      "loss": 5.7735,
+      "step": 2832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6461482048034668,
+      "learning_rate": 4.2993724986412375e-05,
+      "loss": 5.8867,
+      "step": 2836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8447121381759644,
+      "learning_rate": 4.298384307525076e-05,
+      "loss": 5.7835,
+      "step": 2840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4918212890625,
+      "learning_rate": 4.297396116408914e-05,
+      "loss": 5.6962,
+      "step": 2844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7599409818649292,
+      "learning_rate": 4.296407925292752e-05,
+      "loss": 5.8473,
+      "step": 2848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9731839895248413,
+      "learning_rate": 4.29541973417659e-05,
+      "loss": 5.904,
+      "step": 2852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9410591125488281,
+      "learning_rate": 4.294431543060428e-05,
+      "loss": 5.8863,
+      "step": 2856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.942104458808899,
+      "learning_rate": 4.293443351944266e-05,
+      "loss": 5.7006,
+      "step": 2860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6088027954101562,
+      "learning_rate": 4.292455160828104e-05,
+      "loss": 5.7491,
+      "step": 2864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5795434713363647,
+      "learning_rate": 4.2914669697119425e-05,
+      "loss": 5.8635,
+      "step": 2868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7501459121704102,
+      "learning_rate": 4.290478778595781e-05,
+      "loss": 5.8032,
+      "step": 2872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9903556108474731,
+      "learning_rate": 4.289490587479619e-05,
+      "loss": 5.885,
+      "step": 2876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.539027452468872,
+      "learning_rate": 4.288502396363457e-05,
+      "loss": 5.6591,
+      "step": 2880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6801403760910034,
+      "learning_rate": 4.287514205247295e-05,
+      "loss": 5.8648,
+      "step": 2884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8741987943649292,
+      "learning_rate": 4.286526014131133e-05,
+      "loss": 5.9236,
+      "step": 2888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.261087417602539,
+      "learning_rate": 4.285537823014971e-05,
+      "loss": 5.7465,
+      "step": 2892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9104219675064087,
+      "learning_rate": 4.284549631898809e-05,
+      "loss": 5.7858,
+      "step": 2896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8742128610610962,
+      "learning_rate": 4.2835614407826474e-05,
+      "loss": 5.8395,
+      "step": 2900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9993197917938232,
+      "learning_rate": 4.2825732496664856e-05,
+      "loss": 5.7205,
+      "step": 2904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.668923258781433,
+      "learning_rate": 4.281585058550324e-05,
+      "loss": 5.8264,
+      "step": 2908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.509131669998169,
+      "learning_rate": 4.280596867434162e-05,
+      "loss": 5.8287,
+      "step": 2912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6229509115219116,
+      "learning_rate": 4.2796086763179996e-05,
+      "loss": 5.8989,
+      "step": 2916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9508914947509766,
+      "learning_rate": 4.278620485201838e-05,
+      "loss": 5.7111,
+      "step": 2920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3820146322250366,
+      "learning_rate": 4.277632294085676e-05,
+      "loss": 5.8116,
+      "step": 2924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.623271107673645,
+      "learning_rate": 4.276644102969514e-05,
+      "loss": 5.7792,
+      "step": 2928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7664297819137573,
+      "learning_rate": 4.275655911853353e-05,
+      "loss": 5.8097,
+      "step": 2932
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8661329746246338,
+      "learning_rate": 4.274667720737191e-05,
+      "loss": 5.8191,
+      "step": 2936
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6165112257003784,
+      "learning_rate": 4.273679529621029e-05,
+      "loss": 5.8802,
+      "step": 2940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5878552198410034,
+      "learning_rate": 4.272691338504867e-05,
+      "loss": 5.8965,
+      "step": 2944
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6278650760650635,
+      "learning_rate": 4.271703147388705e-05,
+      "loss": 5.6925,
+      "step": 2948
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6811779737472534,
+      "learning_rate": 4.2707149562725434e-05,
+      "loss": 5.7483,
+      "step": 2952
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4222928285598755,
+      "learning_rate": 4.2697267651563816e-05,
+      "loss": 5.7322,
+      "step": 2956
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6275368928909302,
+      "learning_rate": 4.26873857404022e-05,
+      "loss": 5.803,
+      "step": 2960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4676285982131958,
+      "learning_rate": 4.267750382924058e-05,
+      "loss": 5.8874,
+      "step": 2964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3239818811416626,
+      "learning_rate": 4.266762191807896e-05,
+      "loss": 5.8061,
+      "step": 2968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5904514789581299,
+      "learning_rate": 4.265774000691734e-05,
+      "loss": 5.8226,
+      "step": 2972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6302812099456787,
+      "learning_rate": 4.264785809575572e-05,
+      "loss": 5.8038,
+      "step": 2976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.495665192604065,
+      "learning_rate": 4.26379761845941e-05,
+      "loss": 5.8282,
+      "step": 2980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.776206612586975,
+      "learning_rate": 4.2628094273432484e-05,
+      "loss": 5.8956,
+      "step": 2984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3062102794647217,
+      "learning_rate": 4.2618212362270866e-05,
+      "loss": 5.7843,
+      "step": 2988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.649765968322754,
+      "learning_rate": 4.260833045110925e-05,
+      "loss": 5.7521,
+      "step": 2992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4618674516677856,
+      "learning_rate": 4.259844853994763e-05,
+      "loss": 5.8289,
+      "step": 2996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6106091737747192,
+      "learning_rate": 4.2588566628786006e-05,
+      "loss": 5.8524,
+      "step": 3000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6770355701446533,
+      "learning_rate": 4.257868471762439e-05,
+      "loss": 5.8803,
+      "step": 3004
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5975035429000854,
+      "learning_rate": 4.256880280646277e-05,
+      "loss": 5.8038,
+      "step": 3008
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5957070589065552,
+      "learning_rate": 4.255892089530115e-05,
+      "loss": 5.7577,
+      "step": 3012
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.676400065422058,
+      "learning_rate": 4.2549038984139534e-05,
+      "loss": 5.7606,
+      "step": 3016
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4680049419403076,
+      "learning_rate": 4.2539157072977916e-05,
+      "loss": 5.8903,
+      "step": 3020
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5760340690612793,
+      "learning_rate": 4.25292751618163e-05,
+      "loss": 5.7209,
+      "step": 3024
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6681965589523315,
+      "learning_rate": 4.251939325065468e-05,
+      "loss": 5.7767,
+      "step": 3028
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4011784791946411,
+      "learning_rate": 4.2509511339493055e-05,
+      "loss": 5.7752,
+      "step": 3032
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6614855527877808,
+      "learning_rate": 4.249962942833144e-05,
+      "loss": 5.8396,
+      "step": 3036
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5414800643920898,
+      "learning_rate": 4.248974751716982e-05,
+      "loss": 5.8238,
+      "step": 3040
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9518810510635376,
+      "learning_rate": 4.247986560600821e-05,
+      "loss": 5.9185,
+      "step": 3044
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8648451566696167,
+      "learning_rate": 4.246998369484659e-05,
+      "loss": 5.9562,
+      "step": 3048
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.052186965942383,
+      "learning_rate": 4.246010178368497e-05,
+      "loss": 5.7948,
+      "step": 3052
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8737437725067139,
+      "learning_rate": 4.245021987252335e-05,
+      "loss": 5.7977,
+      "step": 3056
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7005298137664795,
+      "learning_rate": 4.244033796136173e-05,
+      "loss": 5.7203,
+      "step": 3060
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.91958487033844,
+      "learning_rate": 4.243045605020011e-05,
+      "loss": 5.798,
+      "step": 3064
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6794556379318237,
+      "learning_rate": 4.2420574139038494e-05,
+      "loss": 5.8465,
+      "step": 3068
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8262078762054443,
+      "learning_rate": 4.2410692227876876e-05,
+      "loss": 5.7919,
+      "step": 3072
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3758339881896973,
+      "learning_rate": 4.240081031671526e-05,
+      "loss": 5.8541,
+      "step": 3076
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6782643795013428,
+      "learning_rate": 4.239092840555364e-05,
+      "loss": 5.7396,
+      "step": 3080
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5164580345153809,
+      "learning_rate": 4.2381046494392015e-05,
+      "loss": 5.8555,
+      "step": 3084
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.664754033088684,
+      "learning_rate": 4.23711645832304e-05,
+      "loss": 5.7878,
+      "step": 3088
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.615749716758728,
+      "learning_rate": 4.236128267206878e-05,
+      "loss": 5.8452,
+      "step": 3092
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.699905514717102,
+      "learning_rate": 4.235140076090716e-05,
+      "loss": 5.6857,
+      "step": 3096
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5537385940551758,
+      "learning_rate": 4.2341518849745543e-05,
+      "loss": 5.7966,
+      "step": 3100
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6133285760879517,
+      "learning_rate": 4.2331636938583926e-05,
+      "loss": 5.7937,
+      "step": 3104
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6025242805480957,
+      "learning_rate": 4.232175502742231e-05,
+      "loss": 5.847,
+      "step": 3108
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.157740831375122,
+      "learning_rate": 4.231187311626069e-05,
+      "loss": 5.8554,
+      "step": 3112
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3504477739334106,
+      "learning_rate": 4.2301991205099065e-05,
+      "loss": 5.864,
+      "step": 3116
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.835391879081726,
+      "learning_rate": 4.229210929393745e-05,
+      "loss": 5.8793,
+      "step": 3120
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7732737064361572,
+      "learning_rate": 4.228222738277583e-05,
+      "loss": 5.8299,
+      "step": 3124
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3115538358688354,
+      "learning_rate": 4.227234547161421e-05,
+      "loss": 5.6581,
+      "step": 3128
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5173206329345703,
+      "learning_rate": 4.226246356045259e-05,
+      "loss": 5.6791,
+      "step": 3132
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.850712537765503,
+      "learning_rate": 4.2252581649290975e-05,
+      "loss": 5.7433,
+      "step": 3136
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6634035110473633,
+      "learning_rate": 4.224269973812936e-05,
+      "loss": 5.8331,
+      "step": 3140
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8786728382110596,
+      "learning_rate": 4.223281782696773e-05,
+      "loss": 5.6907,
+      "step": 3144
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4625319242477417,
+      "learning_rate": 4.2222935915806115e-05,
+      "loss": 5.9035,
+      "step": 3148
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6950637102127075,
+      "learning_rate": 4.22130540046445e-05,
+      "loss": 5.7991,
+      "step": 3152
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7324408292770386,
+      "learning_rate": 4.220317209348288e-05,
+      "loss": 5.7367,
+      "step": 3156
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2330594062805176,
+      "learning_rate": 4.219329018232127e-05,
+      "loss": 5.8253,
+      "step": 3160
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.658172607421875,
+      "learning_rate": 4.218340827115965e-05,
+      "loss": 5.8604,
+      "step": 3164
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7080429792404175,
+      "learning_rate": 4.2173526359998025e-05,
+      "loss": 5.8495,
+      "step": 3168
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8790498971939087,
+      "learning_rate": 4.216364444883641e-05,
+      "loss": 5.7018,
+      "step": 3172
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.222698211669922,
+      "learning_rate": 4.215376253767479e-05,
+      "loss": 5.841,
+      "step": 3176
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7230134010314941,
+      "learning_rate": 4.214388062651317e-05,
+      "loss": 5.805,
+      "step": 3180
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.234403133392334,
+      "learning_rate": 4.213399871535155e-05,
+      "loss": 5.9041,
+      "step": 3184
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1015374660491943,
+      "learning_rate": 4.2124116804189935e-05,
+      "loss": 5.9052,
+      "step": 3188
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6418051719665527,
+      "learning_rate": 4.211423489302832e-05,
+      "loss": 5.7498,
+      "step": 3192
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.524634599685669,
+      "learning_rate": 4.210435298186669e-05,
+      "loss": 5.8123,
+      "step": 3196
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4214069843292236,
+      "learning_rate": 4.2094471070705075e-05,
+      "loss": 5.7962,
+      "step": 3200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6622087955474854,
+      "learning_rate": 4.208458915954346e-05,
+      "loss": 5.8462,
+      "step": 3204
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3304983377456665,
+      "learning_rate": 4.207470724838184e-05,
+      "loss": 5.8222,
+      "step": 3208
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9604005813598633,
+      "learning_rate": 4.206482533722022e-05,
+      "loss": 5.8678,
+      "step": 3212
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4757952690124512,
+      "learning_rate": 4.20549434260586e-05,
+      "loss": 5.8205,
+      "step": 3216
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6784671545028687,
+      "learning_rate": 4.2045061514896985e-05,
+      "loss": 5.8077,
+      "step": 3220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.231874465942383,
+      "learning_rate": 4.203517960373537e-05,
+      "loss": 5.7779,
+      "step": 3224
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.200303316116333,
+      "learning_rate": 4.202529769257374e-05,
+      "loss": 5.8089,
+      "step": 3228
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9842939376831055,
+      "learning_rate": 4.2015415781412124e-05,
+      "loss": 5.819,
+      "step": 3232
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8117098808288574,
+      "learning_rate": 4.2005533870250506e-05,
+      "loss": 5.9165,
+      "step": 3236
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.559996247291565,
+      "learning_rate": 4.199565195908889e-05,
+      "loss": 5.7312,
+      "step": 3240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4291075468063354,
+      "learning_rate": 4.198577004792727e-05,
+      "loss": 5.8351,
+      "step": 3244
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1594057083129883,
+      "learning_rate": 4.197588813676565e-05,
+      "loss": 5.8667,
+      "step": 3248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4217649698257446,
+      "learning_rate": 4.1966006225604035e-05,
+      "loss": 5.7143,
+      "step": 3252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.619981288909912,
+      "learning_rate": 4.195612431444241e-05,
+      "loss": 5.8356,
+      "step": 3256
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.002779960632324,
+      "learning_rate": 4.194624240328079e-05,
+      "loss": 5.7442,
+      "step": 3260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.639796257019043,
+      "learning_rate": 4.1936360492119174e-05,
+      "loss": 5.8934,
+      "step": 3264
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6575497388839722,
+      "learning_rate": 4.1926478580957556e-05,
+      "loss": 5.7446,
+      "step": 3268
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.411887526512146,
+      "learning_rate": 4.191659666979594e-05,
+      "loss": 5.803,
+      "step": 3272
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6861897706985474,
+      "learning_rate": 4.190671475863433e-05,
+      "loss": 5.8827,
+      "step": 3276
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5666462182998657,
+      "learning_rate": 4.18968328474727e-05,
+      "loss": 5.6994,
+      "step": 3280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.06365704536438,
+      "learning_rate": 4.1886950936311084e-05,
+      "loss": 5.7583,
+      "step": 3284
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5888551473617554,
+      "learning_rate": 4.1877069025149466e-05,
+      "loss": 5.778,
+      "step": 3288
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8238952159881592,
+      "learning_rate": 4.186718711398785e-05,
+      "loss": 5.8059,
+      "step": 3292
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9061448574066162,
+      "learning_rate": 4.185730520282623e-05,
+      "loss": 5.7522,
+      "step": 3296
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1226706504821777,
+      "learning_rate": 4.184742329166461e-05,
+      "loss": 5.7591,
+      "step": 3300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4875596761703491,
+      "learning_rate": 4.1837541380502995e-05,
+      "loss": 5.8178,
+      "step": 3304
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5928765535354614,
+      "learning_rate": 4.182765946934138e-05,
+      "loss": 5.7657,
+      "step": 3308
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4694312810897827,
+      "learning_rate": 4.181777755817975e-05,
+      "loss": 5.8106,
+      "step": 3312
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9643967151641846,
+      "learning_rate": 4.1807895647018134e-05,
+      "loss": 5.8297,
+      "step": 3316
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6413182020187378,
+      "learning_rate": 4.1798013735856516e-05,
+      "loss": 5.8,
+      "step": 3320
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6926419734954834,
+      "learning_rate": 4.17881318246949e-05,
+      "loss": 5.7488,
+      "step": 3324
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3814029693603516,
+      "learning_rate": 4.177824991353328e-05,
+      "loss": 5.8541,
+      "step": 3328
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9140582084655762,
+      "learning_rate": 4.176836800237166e-05,
+      "loss": 5.7962,
+      "step": 3332
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.598402976989746,
+      "learning_rate": 4.1758486091210044e-05,
+      "loss": 5.6331,
+      "step": 3336
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7428086996078491,
+      "learning_rate": 4.174860418004842e-05,
+      "loss": 5.8097,
+      "step": 3340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6181442737579346,
+      "learning_rate": 4.17387222688868e-05,
+      "loss": 5.7108,
+      "step": 3344
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6270121335983276,
+      "learning_rate": 4.1728840357725184e-05,
+      "loss": 5.7934,
+      "step": 3348
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.677556037902832,
+      "learning_rate": 4.1718958446563566e-05,
+      "loss": 5.8193,
+      "step": 3352
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8996258974075317,
+      "learning_rate": 4.170907653540195e-05,
+      "loss": 5.8761,
+      "step": 3356
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5995745658874512,
+      "learning_rate": 4.169919462424033e-05,
+      "loss": 5.8213,
+      "step": 3360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6188803911209106,
+      "learning_rate": 4.168931271307871e-05,
+      "loss": 5.9193,
+      "step": 3364
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0405361652374268,
+      "learning_rate": 4.1679430801917094e-05,
+      "loss": 5.76,
+      "step": 3368
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8527336120605469,
+      "learning_rate": 4.166954889075547e-05,
+      "loss": 5.7301,
+      "step": 3372
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9228570461273193,
+      "learning_rate": 4.165966697959385e-05,
+      "loss": 5.6206,
+      "step": 3376
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8511258363723755,
+      "learning_rate": 4.1649785068432233e-05,
+      "loss": 5.6747,
+      "step": 3380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.077301263809204,
+      "learning_rate": 4.1639903157270616e-05,
+      "loss": 5.7165,
+      "step": 3384
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5718106031417847,
+      "learning_rate": 4.1630021246109004e-05,
+      "loss": 5.8808,
+      "step": 3388
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6767138242721558,
+      "learning_rate": 4.1620139334947386e-05,
+      "loss": 5.6554,
+      "step": 3392
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6757532358169556,
+      "learning_rate": 4.161025742378576e-05,
+      "loss": 5.8836,
+      "step": 3396
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7318501472473145,
+      "learning_rate": 4.1600375512624144e-05,
+      "loss": 5.7989,
+      "step": 3400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6216315031051636,
+      "learning_rate": 4.1590493601462526e-05,
+      "loss": 5.7528,
+      "step": 3404
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.916919231414795,
+      "learning_rate": 4.158061169030091e-05,
+      "loss": 5.7544,
+      "step": 3408
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1574885845184326,
+      "learning_rate": 4.157072977913929e-05,
+      "loss": 5.7093,
+      "step": 3412
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.754546880722046,
+      "learning_rate": 4.156084786797767e-05,
+      "loss": 5.6852,
+      "step": 3416
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5894404649734497,
+      "learning_rate": 4.1550965956816054e-05,
+      "loss": 5.8409,
+      "step": 3420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5836817026138306,
+      "learning_rate": 4.154108404565443e-05,
+      "loss": 5.8131,
+      "step": 3424
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2382681369781494,
+      "learning_rate": 4.153120213449281e-05,
+      "loss": 5.7432,
+      "step": 3428
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.515018105506897,
+      "learning_rate": 4.1521320223331194e-05,
+      "loss": 5.8258,
+      "step": 3432
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6593300104141235,
+      "learning_rate": 4.1511438312169576e-05,
+      "loss": 5.6845,
+      "step": 3436
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5710387229919434,
+      "learning_rate": 4.150155640100796e-05,
+      "loss": 5.6832,
+      "step": 3440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8487441539764404,
+      "learning_rate": 4.149167448984634e-05,
+      "loss": 5.9216,
+      "step": 3444
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8978421688079834,
+      "learning_rate": 4.148179257868472e-05,
+      "loss": 5.8274,
+      "step": 3448
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6334807872772217,
+      "learning_rate": 4.1471910667523104e-05,
+      "loss": 5.8029,
+      "step": 3452
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6049920320510864,
+      "learning_rate": 4.146202875636148e-05,
+      "loss": 5.898,
+      "step": 3456
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.874814510345459,
+      "learning_rate": 4.145214684519986e-05,
+      "loss": 5.7016,
+      "step": 3460
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4894344806671143,
+      "learning_rate": 4.144226493403824e-05,
+      "loss": 5.8736,
+      "step": 3464
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7190723419189453,
+      "learning_rate": 4.1432383022876625e-05,
+      "loss": 5.8398,
+      "step": 3468
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1003429889678955,
+      "learning_rate": 4.142250111171501e-05,
+      "loss": 5.725,
+      "step": 3472
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4103918075561523,
+      "learning_rate": 4.141261920055339e-05,
+      "loss": 5.8784,
+      "step": 3476
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6862280368804932,
+      "learning_rate": 4.140273728939177e-05,
+      "loss": 5.7469,
+      "step": 3480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.669593095779419,
+      "learning_rate": 4.139285537823015e-05,
+      "loss": 5.7476,
+      "step": 3484
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.798256754875183,
+      "learning_rate": 4.138297346706853e-05,
+      "loss": 5.729,
+      "step": 3488
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.827129602432251,
+      "learning_rate": 4.137309155590691e-05,
+      "loss": 5.7887,
+      "step": 3492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7160648107528687,
+      "learning_rate": 4.136320964474529e-05,
+      "loss": 5.7244,
+      "step": 3496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.105008363723755,
+      "learning_rate": 4.1353327733583675e-05,
+      "loss": 5.8301,
+      "step": 3500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4461700916290283,
+      "learning_rate": 4.1343445822422064e-05,
+      "loss": 5.8252,
+      "step": 3504
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.478606104850769,
+      "learning_rate": 4.133356391126044e-05,
+      "loss": 5.751,
+      "step": 3508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9824438095092773,
+      "learning_rate": 4.132368200009882e-05,
+      "loss": 5.7331,
+      "step": 3512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8545335531234741,
+      "learning_rate": 4.13138000889372e-05,
+      "loss": 5.8547,
+      "step": 3516
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.836466908454895,
+      "learning_rate": 4.1303918177775585e-05,
+      "loss": 5.7603,
+      "step": 3520
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3290674686431885,
+      "learning_rate": 4.129403626661397e-05,
+      "loss": 5.8594,
+      "step": 3524
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4753937721252441,
+      "learning_rate": 4.128415435545235e-05,
+      "loss": 5.8027,
+      "step": 3528
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.090026617050171,
+      "learning_rate": 4.127427244429073e-05,
+      "loss": 5.8225,
+      "step": 3532
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7022199630737305,
+      "learning_rate": 4.1264390533129114e-05,
+      "loss": 5.8274,
+      "step": 3536
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0701937675476074,
+      "learning_rate": 4.125450862196749e-05,
+      "loss": 5.8273,
+      "step": 3540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5415451526641846,
+      "learning_rate": 4.124462671080587e-05,
+      "loss": 5.7365,
+      "step": 3544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.871156096458435,
+      "learning_rate": 4.123474479964425e-05,
+      "loss": 5.8028,
+      "step": 3548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.788110613822937,
+      "learning_rate": 4.1224862888482635e-05,
+      "loss": 5.7425,
+      "step": 3552
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4837939739227295,
+      "learning_rate": 4.121498097732102e-05,
+      "loss": 5.7204,
+      "step": 3556
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9719536304473877,
+      "learning_rate": 4.12050990661594e-05,
+      "loss": 5.6699,
+      "step": 3560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.865614652633667,
+      "learning_rate": 4.119521715499778e-05,
+      "loss": 5.7139,
+      "step": 3564
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9451555013656616,
+      "learning_rate": 4.1185335243836156e-05,
+      "loss": 5.7739,
+      "step": 3568
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0122992992401123,
+      "learning_rate": 4.117545333267454e-05,
+      "loss": 5.9195,
+      "step": 3572
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.606345295906067,
+      "learning_rate": 4.116557142151292e-05,
+      "loss": 5.8827,
+      "step": 3576
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9131412506103516,
+      "learning_rate": 4.11556895103513e-05,
+      "loss": 5.8901,
+      "step": 3580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.780287742614746,
+      "learning_rate": 4.1145807599189685e-05,
+      "loss": 5.8709,
+      "step": 3584
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.840616226196289,
+      "learning_rate": 4.113592568802807e-05,
+      "loss": 5.8292,
+      "step": 3588
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.641686201095581,
+      "learning_rate": 4.112604377686645e-05,
+      "loss": 5.7915,
+      "step": 3592
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6334573030471802,
+      "learning_rate": 4.111616186570483e-05,
+      "loss": 5.8706,
+      "step": 3596
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.521855592727661,
+      "learning_rate": 4.1106279954543206e-05,
+      "loss": 5.7234,
+      "step": 3600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7530171871185303,
+      "learning_rate": 4.109639804338159e-05,
+      "loss": 5.9393,
+      "step": 3604
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5222742557525635,
+      "learning_rate": 4.108651613221997e-05,
+      "loss": 5.7107,
+      "step": 3608
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9538359642028809,
+      "learning_rate": 4.107663422105835e-05,
+      "loss": 5.837,
+      "step": 3612
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5958921909332275,
+      "learning_rate": 4.1066752309896734e-05,
+      "loss": 5.8572,
+      "step": 3616
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4122501611709595,
+      "learning_rate": 4.105687039873512e-05,
+      "loss": 5.8037,
+      "step": 3620
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.674170732498169,
+      "learning_rate": 4.10469884875735e-05,
+      "loss": 5.6393,
+      "step": 3624
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.989425778388977,
+      "learning_rate": 4.103710657641188e-05,
+      "loss": 5.7025,
+      "step": 3628
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4842572212219238,
+      "learning_rate": 4.102722466525026e-05,
+      "loss": 5.7045,
+      "step": 3632
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6306571960449219,
+      "learning_rate": 4.1017342754088645e-05,
+      "loss": 5.9058,
+      "step": 3636
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7082542181015015,
+      "learning_rate": 4.100746084292703e-05,
+      "loss": 5.7987,
+      "step": 3640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4293181896209717,
+      "learning_rate": 4.099757893176541e-05,
+      "loss": 5.7266,
+      "step": 3644
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4269033670425415,
+      "learning_rate": 4.098769702060379e-05,
+      "loss": 5.7295,
+      "step": 3648
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6524076461791992,
+      "learning_rate": 4.0977815109442166e-05,
+      "loss": 5.7705,
+      "step": 3652
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.667291522026062,
+      "learning_rate": 4.096793319828055e-05,
+      "loss": 5.631,
+      "step": 3656
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8146156072616577,
+      "learning_rate": 4.095805128711893e-05,
+      "loss": 5.8973,
+      "step": 3660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.884358286857605,
+      "learning_rate": 4.094816937595731e-05,
+      "loss": 5.7693,
+      "step": 3664
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5343087911605835,
+      "learning_rate": 4.0938287464795694e-05,
+      "loss": 5.7735,
+      "step": 3668
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6283289194107056,
+      "learning_rate": 4.0928405553634076e-05,
+      "loss": 5.788,
+      "step": 3672
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5237561464309692,
+      "learning_rate": 4.091852364247246e-05,
+      "loss": 5.6661,
+      "step": 3676
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8047362565994263,
+      "learning_rate": 4.090864173131084e-05,
+      "loss": 5.6598,
+      "step": 3680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6127629280090332,
+      "learning_rate": 4.0898759820149216e-05,
+      "loss": 5.6893,
+      "step": 3684
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2976841926574707,
+      "learning_rate": 4.08888779089876e-05,
+      "loss": 5.7673,
+      "step": 3688
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.025730609893799,
+      "learning_rate": 4.087899599782598e-05,
+      "loss": 5.7116,
+      "step": 3692
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5258173942565918,
+      "learning_rate": 4.086911408666436e-05,
+      "loss": 5.8727,
+      "step": 3696
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.983288049697876,
+      "learning_rate": 4.0859232175502744e-05,
+      "loss": 5.8148,
+      "step": 3700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6084184646606445,
+      "learning_rate": 4.0849350264341126e-05,
+      "loss": 5.816,
+      "step": 3704
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.739123821258545,
+      "learning_rate": 4.083946835317951e-05,
+      "loss": 5.8155,
+      "step": 3708
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8195106983184814,
+      "learning_rate": 4.0829586442017884e-05,
+      "loss": 5.7962,
+      "step": 3712
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5186692476272583,
+      "learning_rate": 4.0819704530856266e-05,
+      "loss": 5.8265,
+      "step": 3716
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1515772342681885,
+      "learning_rate": 4.080982261969465e-05,
+      "loss": 5.734,
+      "step": 3720
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9751808643341064,
+      "learning_rate": 4.079994070853303e-05,
+      "loss": 5.7605,
+      "step": 3724
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5598946809768677,
+      "learning_rate": 4.079005879737141e-05,
+      "loss": 5.6552,
+      "step": 3728
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.642751693725586,
+      "learning_rate": 4.0780176886209794e-05,
+      "loss": 5.7607,
+      "step": 3732
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0432446002960205,
+      "learning_rate": 4.0770294975048176e-05,
+      "loss": 5.7103,
+      "step": 3736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8500975370407104,
+      "learning_rate": 4.076041306388656e-05,
+      "loss": 5.7195,
+      "step": 3740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5115128755569458,
+      "learning_rate": 4.075053115272494e-05,
+      "loss": 5.8491,
+      "step": 3744
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.241079092025757,
+      "learning_rate": 4.074064924156332e-05,
+      "loss": 5.8183,
+      "step": 3748
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3165819644927979,
+      "learning_rate": 4.0730767330401704e-05,
+      "loss": 5.7614,
+      "step": 3752
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4908742904663086,
+      "learning_rate": 4.0720885419240086e-05,
+      "loss": 5.7343,
+      "step": 3756
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.761338710784912,
+      "learning_rate": 4.071100350807847e-05,
+      "loss": 5.7702,
+      "step": 3760
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7084866762161255,
+      "learning_rate": 4.0701121596916844e-05,
+      "loss": 5.7966,
+      "step": 3764
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0340840816497803,
+      "learning_rate": 4.0691239685755226e-05,
+      "loss": 5.7835,
+      "step": 3768
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2564029693603516,
+      "learning_rate": 4.068135777459361e-05,
+      "loss": 5.7911,
+      "step": 3772
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2431206703186035,
+      "learning_rate": 4.067147586343199e-05,
+      "loss": 5.7461,
+      "step": 3776
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6614986658096313,
+      "learning_rate": 4.066159395227037e-05,
+      "loss": 5.8013,
+      "step": 3780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8505353927612305,
+      "learning_rate": 4.0651712041108754e-05,
+      "loss": 5.7317,
+      "step": 3784
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.03542160987854,
+      "learning_rate": 4.0641830129947136e-05,
+      "loss": 5.822,
+      "step": 3788
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5652294158935547,
+      "learning_rate": 4.063194821878552e-05,
+      "loss": 5.8001,
+      "step": 3792
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7387696504592896,
+      "learning_rate": 4.062206630762389e-05,
+      "loss": 5.7575,
+      "step": 3796
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7505065202713013,
+      "learning_rate": 4.0612184396462275e-05,
+      "loss": 5.6909,
+      "step": 3800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7365105152130127,
+      "learning_rate": 4.060230248530066e-05,
+      "loss": 5.7282,
+      "step": 3804
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.965735912322998,
+      "learning_rate": 4.059242057413904e-05,
+      "loss": 5.7692,
+      "step": 3808
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.409680962562561,
+      "learning_rate": 4.058253866297742e-05,
+      "loss": 5.8235,
+      "step": 3812
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8310275077819824,
+      "learning_rate": 4.0572656751815804e-05,
+      "loss": 5.7516,
+      "step": 3816
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.537810206413269,
+      "learning_rate": 4.0562774840654186e-05,
+      "loss": 5.6683,
+      "step": 3820
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7618563175201416,
+      "learning_rate": 4.055289292949256e-05,
+      "loss": 5.7643,
+      "step": 3824
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.603681206703186,
+      "learning_rate": 4.054301101833094e-05,
+      "loss": 5.6578,
+      "step": 3828
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3885388374328613,
+      "learning_rate": 4.0533129107169325e-05,
+      "loss": 5.8278,
+      "step": 3832
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6048877239227295,
+      "learning_rate": 4.052324719600771e-05,
+      "loss": 5.8245,
+      "step": 3836
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.443018913269043,
+      "learning_rate": 4.051336528484609e-05,
+      "loss": 5.5956,
+      "step": 3840
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6389518976211548,
+      "learning_rate": 4.050348337368447e-05,
+      "loss": 5.7129,
+      "step": 3844
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4876172542572021,
+      "learning_rate": 4.049360146252285e-05,
+      "loss": 5.7235,
+      "step": 3848
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8396450281143188,
+      "learning_rate": 4.0483719551361235e-05,
+      "loss": 5.8035,
+      "step": 3852
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4481145143508911,
+      "learning_rate": 4.047383764019962e-05,
+      "loss": 5.793,
+      "step": 3856
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8424584865570068,
+      "learning_rate": 4.0463955729038e-05,
+      "loss": 5.7926,
+      "step": 3860
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1682097911834717,
+      "learning_rate": 4.045407381787638e-05,
+      "loss": 5.7651,
+      "step": 3864
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.371147871017456,
+      "learning_rate": 4.0444191906714764e-05,
+      "loss": 5.7553,
+      "step": 3868
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8529584407806396,
+      "learning_rate": 4.0434309995553146e-05,
+      "loss": 5.738,
+      "step": 3872
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.445864677429199,
+      "learning_rate": 4.042442808439153e-05,
+      "loss": 5.6886,
+      "step": 3876
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6438809633255005,
+      "learning_rate": 4.04145461732299e-05,
+      "loss": 5.7889,
+      "step": 3880
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3593320846557617,
+      "learning_rate": 4.0404664262068285e-05,
+      "loss": 5.7318,
+      "step": 3884
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7897164821624756,
+      "learning_rate": 4.039478235090667e-05,
+      "loss": 5.7785,
+      "step": 3888
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6130775213241577,
+      "learning_rate": 4.038490043974505e-05,
+      "loss": 5.7774,
+      "step": 3892
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5768321752548218,
+      "learning_rate": 4.037501852858343e-05,
+      "loss": 5.8277,
+      "step": 3896
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9530898332595825,
+      "learning_rate": 4.036513661742181e-05,
+      "loss": 5.8019,
+      "step": 3900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9102182388305664,
+      "learning_rate": 4.0355254706260195e-05,
+      "loss": 5.7975,
+      "step": 3904
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7103482484817505,
+      "learning_rate": 4.034537279509857e-05,
+      "loss": 5.7725,
+      "step": 3908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5741090774536133,
+      "learning_rate": 4.033549088393695e-05,
+      "loss": 5.7367,
+      "step": 3912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2126150131225586,
+      "learning_rate": 4.0325608972775335e-05,
+      "loss": 5.9088,
+      "step": 3916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7655351161956787,
+      "learning_rate": 4.031572706161372e-05,
+      "loss": 5.847,
+      "step": 3920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7152812480926514,
+      "learning_rate": 4.03058451504521e-05,
+      "loss": 5.7307,
+      "step": 3924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.767104983329773,
+      "learning_rate": 4.029596323929048e-05,
+      "loss": 5.8074,
+      "step": 3928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9230767488479614,
+      "learning_rate": 4.028608132812886e-05,
+      "loss": 5.7124,
+      "step": 3932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6367123126983643,
+      "learning_rate": 4.0276199416967245e-05,
+      "loss": 5.7474,
+      "step": 3936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0680441856384277,
+      "learning_rate": 4.026631750580562e-05,
+      "loss": 5.8713,
+      "step": 3940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7396044731140137,
+      "learning_rate": 4.0256435594644e-05,
+      "loss": 5.7464,
+      "step": 3944
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2416396141052246,
+      "learning_rate": 4.0246553683482384e-05,
+      "loss": 5.9287,
+      "step": 3948
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.192002534866333,
+      "learning_rate": 4.0236671772320766e-05,
+      "loss": 5.8617,
+      "step": 3952
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6008896827697754,
+      "learning_rate": 4.022678986115915e-05,
+      "loss": 5.7929,
+      "step": 3956
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7184627056121826,
+      "learning_rate": 4.021690794999753e-05,
+      "loss": 5.7198,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4607460498809814,
+      "learning_rate": 4.020702603883591e-05,
+      "loss": 5.6012,
+      "step": 3964
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.070399761199951,
+      "learning_rate": 4.0197144127674295e-05,
+      "loss": 5.6176,
+      "step": 3968
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7792775630950928,
+      "learning_rate": 4.018726221651268e-05,
+      "loss": 5.8348,
+      "step": 3972
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0791399478912354,
+      "learning_rate": 4.017738030535106e-05,
+      "loss": 5.8916,
+      "step": 3976
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7796858549118042,
+      "learning_rate": 4.016749839418944e-05,
+      "loss": 5.8424,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2471671104431152,
+      "learning_rate": 4.015761648302782e-05,
+      "loss": 5.832,
+      "step": 3984
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6249337196350098,
+      "learning_rate": 4.0147734571866205e-05,
+      "loss": 5.72,
+      "step": 3988
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6861313581466675,
+      "learning_rate": 4.013785266070458e-05,
+      "loss": 5.7362,
+      "step": 3992
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8561558723449707,
+      "learning_rate": 4.012797074954296e-05,
+      "loss": 5.7032,
+      "step": 3996
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6355276107788086,
+      "learning_rate": 4.0118088838381344e-05,
+      "loss": 5.7695,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0652620792388916,
+      "learning_rate": 4.0108206927219726e-05,
+      "loss": 5.8284,
+      "step": 4004
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5604512691497803,
+      "learning_rate": 4.009832501605811e-05,
+      "loss": 5.8084,
+      "step": 4008
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8975857496261597,
+      "learning_rate": 4.008844310489649e-05,
+      "loss": 5.7599,
+      "step": 4012
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7917417287826538,
+      "learning_rate": 4.007856119373487e-05,
+      "loss": 5.7057,
+      "step": 4016
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6700019836425781,
+      "learning_rate": 4.0068679282573255e-05,
+      "loss": 5.7124,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7579398155212402,
+      "learning_rate": 4.005879737141163e-05,
+      "loss": 5.7449,
+      "step": 4024
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5005486011505127,
+      "learning_rate": 4.004891546025001e-05,
+      "loss": 5.6679,
+      "step": 4028
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.033926010131836,
+      "learning_rate": 4.0039033549088394e-05,
+      "loss": 5.8639,
+      "step": 4032
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0436344146728516,
+      "learning_rate": 4.0029151637926776e-05,
+      "loss": 5.7561,
+      "step": 4036
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.592089295387268,
+      "learning_rate": 4.001926972676516e-05,
+      "loss": 5.7275,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.860810399055481,
+      "learning_rate": 4.000938781560354e-05,
+      "loss": 5.829,
+      "step": 4044
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.988085150718689,
+      "learning_rate": 3.999950590444192e-05,
+      "loss": 5.7726,
+      "step": 4048
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7819504737854004,
+      "learning_rate": 3.99896239932803e-05,
+      "loss": 5.6843,
+      "step": 4052
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2862792015075684,
+      "learning_rate": 3.997974208211868e-05,
+      "loss": 5.7483,
+      "step": 4056
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5969876050949097,
+      "learning_rate": 3.996986017095706e-05,
+      "loss": 5.7825,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5458773374557495,
+      "learning_rate": 3.9959978259795444e-05,
+      "loss": 5.7123,
+      "step": 4064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.516552209854126,
+      "learning_rate": 3.9950096348633826e-05,
+      "loss": 5.7107,
+      "step": 4068
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6216940879821777,
+      "learning_rate": 3.994021443747221e-05,
+      "loss": 5.7065,
+      "step": 4072
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.020914077758789,
+      "learning_rate": 3.993033252631059e-05,
+      "loss": 5.938,
+      "step": 4076
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9217082262039185,
+      "learning_rate": 3.992045061514897e-05,
+      "loss": 5.8464,
+      "step": 4080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8421107530593872,
+      "learning_rate": 3.9910568703987354e-05,
+      "loss": 5.672,
+      "step": 4084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.107970952987671,
+      "learning_rate": 3.9900686792825736e-05,
+      "loss": 5.787,
+      "step": 4088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8105857372283936,
+      "learning_rate": 3.989080488166412e-05,
+      "loss": 5.8617,
+      "step": 4092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9893261194229126,
+      "learning_rate": 3.98809229705025e-05,
+      "loss": 5.766,
+      "step": 4096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.125051498413086,
+      "learning_rate": 3.987104105934088e-05,
+      "loss": 5.8332,
+      "step": 4100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6055108308792114,
+      "learning_rate": 3.9861159148179264e-05,
+      "loss": 5.7197,
+      "step": 4104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.727008581161499,
+      "learning_rate": 3.985127723701764e-05,
+      "loss": 5.7569,
+      "step": 4108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6093586683273315,
+      "learning_rate": 3.984139532585602e-05,
+      "loss": 5.6682,
+      "step": 4112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0254461765289307,
+      "learning_rate": 3.9831513414694404e-05,
+      "loss": 5.7424,
+      "step": 4116
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8866037130355835,
+      "learning_rate": 3.9821631503532786e-05,
+      "loss": 5.8635,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5677746534347534,
+      "learning_rate": 3.981174959237117e-05,
+      "loss": 5.8245,
+      "step": 4124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.763741374015808,
+      "learning_rate": 3.980186768120955e-05,
+      "loss": 5.7215,
+      "step": 4128
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9562249183654785,
+      "learning_rate": 3.979198577004793e-05,
+      "loss": 5.7907,
+      "step": 4132
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8377013206481934,
+      "learning_rate": 3.978210385888631e-05,
+      "loss": 5.6803,
+      "step": 4136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.684601902961731,
+      "learning_rate": 3.977222194772469e-05,
+      "loss": 5.772,
+      "step": 4140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5692558288574219,
+      "learning_rate": 3.976234003656307e-05,
+      "loss": 5.8362,
+      "step": 4144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6108436584472656,
+      "learning_rate": 3.9752458125401454e-05,
+      "loss": 5.7818,
+      "step": 4148
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.75248384475708,
+      "learning_rate": 3.9742576214239836e-05,
+      "loss": 5.8399,
+      "step": 4152
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5307252407073975,
+      "learning_rate": 3.973269430307822e-05,
+      "loss": 5.7638,
+      "step": 4156
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4091581106185913,
+      "learning_rate": 3.97228123919166e-05,
+      "loss": 5.7551,
+      "step": 4160
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6635953187942505,
+      "learning_rate": 3.971293048075498e-05,
+      "loss": 5.7285,
+      "step": 4164
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8554000854492188,
+      "learning_rate": 3.970304856959336e-05,
+      "loss": 5.5641,
+      "step": 4168
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6862932443618774,
+      "learning_rate": 3.969316665843174e-05,
+      "loss": 5.8047,
+      "step": 4172
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9776607751846313,
+      "learning_rate": 3.968328474727012e-05,
+      "loss": 5.7278,
+      "step": 4176
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6065359115600586,
+      "learning_rate": 3.96734028361085e-05,
+      "loss": 5.7115,
+      "step": 4180
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3388211727142334,
+      "learning_rate": 3.9663520924946885e-05,
+      "loss": 5.6954,
+      "step": 4184
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7472275495529175,
+      "learning_rate": 3.965363901378527e-05,
+      "loss": 5.834,
+      "step": 4188
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7328163385391235,
+      "learning_rate": 3.964375710262365e-05,
+      "loss": 5.6081,
+      "step": 4192
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0004336833953857,
+      "learning_rate": 3.963387519146203e-05,
+      "loss": 5.7344,
+      "step": 4196
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.126734733581543,
+      "learning_rate": 3.9623993280300414e-05,
+      "loss": 5.7677,
+      "step": 4200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.693859338760376,
+      "learning_rate": 3.9614111369138796e-05,
+      "loss": 5.7142,
+      "step": 4204
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.758854627609253,
+      "learning_rate": 3.960422945797718e-05,
+      "loss": 5.7381,
+      "step": 4208
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1982223987579346,
+      "learning_rate": 3.959434754681556e-05,
+      "loss": 5.9032,
+      "step": 4212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.588758111000061,
+      "learning_rate": 3.958446563565394e-05,
+      "loss": 5.6668,
+      "step": 4216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8513824939727783,
+      "learning_rate": 3.957458372449232e-05,
+      "loss": 5.5942,
+      "step": 4220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1672956943511963,
+      "learning_rate": 3.95647018133307e-05,
+      "loss": 5.8085,
+      "step": 4224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9962481260299683,
+      "learning_rate": 3.955481990216908e-05,
+      "loss": 5.7762,
+      "step": 4228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5460338592529297,
+      "learning_rate": 3.954493799100746e-05,
+      "loss": 5.7283,
+      "step": 4232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5077719688415527,
+      "learning_rate": 3.9535056079845845e-05,
+      "loss": 5.7272,
+      "step": 4236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.141517400741577,
+      "learning_rate": 3.952517416868423e-05,
+      "loss": 5.7612,
+      "step": 4240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6528626680374146,
+      "learning_rate": 3.951529225752261e-05,
+      "loss": 5.6216,
+      "step": 4244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8636984825134277,
+      "learning_rate": 3.950541034636099e-05,
+      "loss": 5.8001,
+      "step": 4248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4790719747543335,
+      "learning_rate": 3.949552843519937e-05,
+      "loss": 5.7866,
+      "step": 4252
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.534711241722107,
+      "learning_rate": 3.948564652403775e-05,
+      "loss": 5.708,
+      "step": 4256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.083681344985962,
+      "learning_rate": 3.947576461287613e-05,
+      "loss": 5.6374,
+      "step": 4260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5164477825164795,
+      "learning_rate": 3.946588270171451e-05,
+      "loss": 5.7652,
+      "step": 4264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6196374893188477,
+      "learning_rate": 3.9456000790552895e-05,
+      "loss": 5.7647,
+      "step": 4268
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.348536491394043,
+      "learning_rate": 3.944611887939128e-05,
+      "loss": 5.6887,
+      "step": 4272
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8124507665634155,
+      "learning_rate": 3.943623696822966e-05,
+      "loss": 5.8996,
+      "step": 4276
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.919488787651062,
+      "learning_rate": 3.9426355057068034e-05,
+      "loss": 5.8877,
+      "step": 4280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.190864086151123,
+      "learning_rate": 3.9416473145906416e-05,
+      "loss": 5.6733,
+      "step": 4284
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.941540241241455,
+      "learning_rate": 3.94065912347448e-05,
+      "loss": 5.6711,
+      "step": 4288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.46822190284729,
+      "learning_rate": 3.939670932358318e-05,
+      "loss": 5.7721,
+      "step": 4292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7537801265716553,
+      "learning_rate": 3.938682741242156e-05,
+      "loss": 5.663,
+      "step": 4296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6122314929962158,
+      "learning_rate": 3.9376945501259945e-05,
+      "loss": 5.6916,
+      "step": 4300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9700013399124146,
+      "learning_rate": 3.936706359009833e-05,
+      "loss": 5.7515,
+      "step": 4304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7138198614120483,
+      "learning_rate": 3.935718167893671e-05,
+      "loss": 5.6894,
+      "step": 4308
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5922127962112427,
+      "learning_rate": 3.934729976777509e-05,
+      "loss": 5.6651,
+      "step": 4312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7139687538146973,
+      "learning_rate": 3.933741785661347e-05,
+      "loss": 5.8207,
+      "step": 4316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8785958290100098,
+      "learning_rate": 3.9327535945451855e-05,
+      "loss": 5.6558,
+      "step": 4320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6499451398849487,
+      "learning_rate": 3.931765403429024e-05,
+      "loss": 5.7528,
+      "step": 4324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9110082387924194,
+      "learning_rate": 3.930777212312862e-05,
+      "loss": 5.7717,
+      "step": 4328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.643752098083496,
+      "learning_rate": 3.9297890211966994e-05,
+      "loss": 5.7846,
+      "step": 4332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1606407165527344,
+      "learning_rate": 3.9288008300805377e-05,
+      "loss": 5.7187,
+      "step": 4336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6892307996749878,
+      "learning_rate": 3.927812638964376e-05,
+      "loss": 5.7547,
+      "step": 4340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.863633632659912,
+      "learning_rate": 3.926824447848214e-05,
+      "loss": 5.6712,
+      "step": 4344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9984108209609985,
+      "learning_rate": 3.925836256732052e-05,
+      "loss": 5.6589,
+      "step": 4348
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6699365377426147,
+      "learning_rate": 3.9248480656158905e-05,
+      "loss": 5.7613,
+      "step": 4352
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7803456783294678,
+      "learning_rate": 3.923859874499729e-05,
+      "loss": 5.7216,
+      "step": 4356
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8623511791229248,
+      "learning_rate": 3.922871683383567e-05,
+      "loss": 5.6227,
+      "step": 4360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5841137170791626,
+      "learning_rate": 3.9218834922674044e-05,
+      "loss": 5.6657,
+      "step": 4364
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5960050821304321,
+      "learning_rate": 3.9208953011512426e-05,
+      "loss": 5.6565,
+      "step": 4368
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5138053894042969,
+      "learning_rate": 3.919907110035081e-05,
+      "loss": 5.7149,
+      "step": 4372
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.044208288192749,
+      "learning_rate": 3.918918918918919e-05,
+      "loss": 5.659,
+      "step": 4376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7930463552474976,
+      "learning_rate": 3.917930727802757e-05,
+      "loss": 5.6993,
+      "step": 4380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9749689102172852,
+      "learning_rate": 3.9169425366865954e-05,
+      "loss": 5.6016,
+      "step": 4384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.737301230430603,
+      "learning_rate": 3.9159543455704337e-05,
+      "loss": 5.651,
+      "step": 4388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.506861925125122,
+      "learning_rate": 3.914966154454271e-05,
+      "loss": 5.7291,
+      "step": 4392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6121779680252075,
+      "learning_rate": 3.9139779633381094e-05,
+      "loss": 5.6529,
+      "step": 4396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6343719959259033,
+      "learning_rate": 3.9129897722219476e-05,
+      "loss": 5.7278,
+      "step": 4400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9690921306610107,
+      "learning_rate": 3.912001581105786e-05,
+      "loss": 5.636,
+      "step": 4404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6430257558822632,
+      "learning_rate": 3.911013389989624e-05,
+      "loss": 5.7014,
+      "step": 4408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.506526231765747,
+      "learning_rate": 3.910025198873462e-05,
+      "loss": 5.8489,
+      "step": 4412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.704164743423462,
+      "learning_rate": 3.9090370077573004e-05,
+      "loss": 5.8284,
+      "step": 4416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7875306606292725,
+      "learning_rate": 3.9080488166411386e-05,
+      "loss": 5.633,
+      "step": 4420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9106501340866089,
+      "learning_rate": 3.907060625524977e-05,
+      "loss": 5.775,
+      "step": 4424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9393559694290161,
+      "learning_rate": 3.906072434408815e-05,
+      "loss": 5.689,
+      "step": 4428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8752541542053223,
+      "learning_rate": 3.905084243292653e-05,
+      "loss": 5.7892,
+      "step": 4432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5603359937667847,
+      "learning_rate": 3.9040960521764914e-05,
+      "loss": 5.8323,
+      "step": 4436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6264493465423584,
+      "learning_rate": 3.9031078610603297e-05,
+      "loss": 5.7913,
+      "step": 4440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8390896320343018,
+      "learning_rate": 3.902119669944168e-05,
+      "loss": 5.7695,
+      "step": 4444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5323280096054077,
+      "learning_rate": 3.9011314788280054e-05,
+      "loss": 5.7638,
+      "step": 4448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8048474788665771,
+      "learning_rate": 3.9001432877118436e-05,
+      "loss": 5.7824,
+      "step": 4452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.657486915588379,
+      "learning_rate": 3.899155096595682e-05,
+      "loss": 5.7037,
+      "step": 4456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7981828451156616,
+      "learning_rate": 3.89816690547952e-05,
+      "loss": 5.6465,
+      "step": 4460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5225144624710083,
+      "learning_rate": 3.897178714363358e-05,
+      "loss": 5.8237,
+      "step": 4464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4845541715621948,
+      "learning_rate": 3.8961905232471964e-05,
+      "loss": 5.7549,
+      "step": 4468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5387517213821411,
+      "learning_rate": 3.8952023321310346e-05,
+      "loss": 5.7666,
+      "step": 4472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7411534786224365,
+      "learning_rate": 3.894214141014872e-05,
+      "loss": 5.7204,
+      "step": 4476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.622908592224121,
+      "learning_rate": 3.8932259498987104e-05,
+      "loss": 5.6133,
+      "step": 4480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8925143480300903,
+      "learning_rate": 3.8922377587825486e-05,
+      "loss": 5.7133,
+      "step": 4484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7361855506896973,
+      "learning_rate": 3.891249567666387e-05,
+      "loss": 5.7737,
+      "step": 4488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.588912010192871,
+      "learning_rate": 3.890261376550225e-05,
+      "loss": 5.7253,
+      "step": 4492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7690430879592896,
+      "learning_rate": 3.889273185434063e-05,
+      "loss": 5.7693,
+      "step": 4496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1007232666015625,
+      "learning_rate": 3.8882849943179014e-05,
+      "loss": 5.733,
+      "step": 4500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1150882244110107,
+      "learning_rate": 3.8872968032017396e-05,
+      "loss": 5.6447,
+      "step": 4504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.85843825340271,
+      "learning_rate": 3.886308612085577e-05,
+      "loss": 5.7652,
+      "step": 4508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6900382041931152,
+      "learning_rate": 3.885320420969415e-05,
+      "loss": 5.7937,
+      "step": 4512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7769362926483154,
+      "learning_rate": 3.8843322298532535e-05,
+      "loss": 5.7276,
+      "step": 4516
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.092283010482788,
+      "learning_rate": 3.883344038737092e-05,
+      "loss": 5.6712,
+      "step": 4520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7213283777236938,
+      "learning_rate": 3.88235584762093e-05,
+      "loss": 5.7979,
+      "step": 4524
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5792101621627808,
+      "learning_rate": 3.881367656504768e-05,
+      "loss": 5.7273,
+      "step": 4528
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8439412117004395,
+      "learning_rate": 3.8803794653886064e-05,
+      "loss": 5.8053,
+      "step": 4532
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0028135776519775,
+      "learning_rate": 3.8793912742724446e-05,
+      "loss": 5.7714,
+      "step": 4536
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7748719453811646,
+      "learning_rate": 3.878403083156283e-05,
+      "loss": 5.8676,
+      "step": 4540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7069579362869263,
+      "learning_rate": 3.877414892040121e-05,
+      "loss": 5.5816,
+      "step": 4544
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.695499062538147,
+      "learning_rate": 3.876426700923959e-05,
+      "loss": 5.8817,
+      "step": 4548
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.573487639427185,
+      "learning_rate": 3.8754385098077974e-05,
+      "loss": 5.792,
+      "step": 4552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.576565146446228,
+      "learning_rate": 3.8744503186916356e-05,
+      "loss": 5.6382,
+      "step": 4556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.081883668899536,
+      "learning_rate": 3.873462127575473e-05,
+      "loss": 5.7344,
+      "step": 4560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9463318586349487,
+      "learning_rate": 3.872473936459311e-05,
+      "loss": 5.6469,
+      "step": 4564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8385138511657715,
+      "learning_rate": 3.8714857453431495e-05,
+      "loss": 5.8192,
+      "step": 4568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7394839525222778,
+      "learning_rate": 3.870497554226988e-05,
+      "loss": 5.5372,
+      "step": 4572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.578692078590393,
+      "learning_rate": 3.869509363110826e-05,
+      "loss": 5.7304,
+      "step": 4576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.551023006439209,
+      "learning_rate": 3.868521171994664e-05,
+      "loss": 5.8348,
+      "step": 4580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7546929121017456,
+      "learning_rate": 3.8675329808785024e-05,
+      "loss": 5.7532,
+      "step": 4584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7034144401550293,
+      "learning_rate": 3.8665447897623406e-05,
+      "loss": 5.6995,
+      "step": 4588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4249943494796753,
+      "learning_rate": 3.865556598646178e-05,
+      "loss": 5.7219,
+      "step": 4592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5602226257324219,
+      "learning_rate": 3.864568407530016e-05,
+      "loss": 5.6544,
+      "step": 4596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4957911968231201,
+      "learning_rate": 3.8635802164138545e-05,
+      "loss": 5.7934,
+      "step": 4600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6986734867095947,
+      "learning_rate": 3.862592025297693e-05,
+      "loss": 5.7564,
+      "step": 4604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6847269535064697,
+      "learning_rate": 3.861603834181531e-05,
+      "loss": 5.583,
+      "step": 4608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.73738431930542,
+      "learning_rate": 3.860615643065369e-05,
+      "loss": 5.6771,
+      "step": 4612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5951757431030273,
+      "learning_rate": 3.859627451949207e-05,
+      "loss": 5.6885,
+      "step": 4616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6416071653366089,
+      "learning_rate": 3.858639260833045e-05,
+      "loss": 5.7239,
+      "step": 4620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0191540718078613,
+      "learning_rate": 3.857651069716883e-05,
+      "loss": 5.6335,
+      "step": 4624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.677107810974121,
+      "learning_rate": 3.856662878600721e-05,
+      "loss": 5.6885,
+      "step": 4628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.619813323020935,
+      "learning_rate": 3.8556746874845595e-05,
+      "loss": 5.6838,
+      "step": 4632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8103737831115723,
+      "learning_rate": 3.854686496368398e-05,
+      "loss": 5.6114,
+      "step": 4636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6023643016815186,
+      "learning_rate": 3.853698305252236e-05,
+      "loss": 5.718,
+      "step": 4640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3481965065002441,
+      "learning_rate": 3.852710114136074e-05,
+      "loss": 5.6867,
+      "step": 4644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6693611145019531,
+      "learning_rate": 3.851721923019912e-05,
+      "loss": 5.8371,
+      "step": 4648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7420861721038818,
+      "learning_rate": 3.8507337319037505e-05,
+      "loss": 5.7511,
+      "step": 4652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6563256978988647,
+      "learning_rate": 3.849745540787589e-05,
+      "loss": 5.64,
+      "step": 4656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7449144124984741,
+      "learning_rate": 3.848757349671427e-05,
+      "loss": 5.7528,
+      "step": 4660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9364055395126343,
+      "learning_rate": 3.847769158555265e-05,
+      "loss": 5.7674,
+      "step": 4664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8317818641662598,
+      "learning_rate": 3.846780967439103e-05,
+      "loss": 5.7468,
+      "step": 4668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.949170470237732,
+      "learning_rate": 3.8457927763229415e-05,
+      "loss": 5.7701,
+      "step": 4672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7268710136413574,
+      "learning_rate": 3.844804585206779e-05,
+      "loss": 5.7698,
+      "step": 4676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8129808902740479,
+      "learning_rate": 3.843816394090617e-05,
+      "loss": 5.7412,
+      "step": 4680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.815779209136963,
+      "learning_rate": 3.8428282029744555e-05,
+      "loss": 5.7174,
+      "step": 4684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8246920108795166,
+      "learning_rate": 3.841840011858294e-05,
+      "loss": 5.7602,
+      "step": 4688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7265750169754028,
+      "learning_rate": 3.840851820742132e-05,
+      "loss": 5.7137,
+      "step": 4692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4503543376922607,
+      "learning_rate": 3.83986362962597e-05,
+      "loss": 5.78,
+      "step": 4696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0876848697662354,
+      "learning_rate": 3.838875438509808e-05,
+      "loss": 5.67,
+      "step": 4700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5135231018066406,
+      "learning_rate": 3.837887247393646e-05,
+      "loss": 5.7372,
+      "step": 4704
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9444183111190796,
+      "learning_rate": 3.836899056277484e-05,
+      "loss": 5.6755,
+      "step": 4708
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8341116905212402,
+      "learning_rate": 3.835910865161322e-05,
+      "loss": 5.7714,
+      "step": 4712
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6584597826004028,
+      "learning_rate": 3.8349226740451604e-05,
+      "loss": 5.6811,
+      "step": 4716
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7305800914764404,
+      "learning_rate": 3.8339344829289987e-05,
+      "loss": 5.6905,
+      "step": 4720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.826827883720398,
+      "learning_rate": 3.832946291812837e-05,
+      "loss": 5.6126,
+      "step": 4724
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5062451362609863,
+      "learning_rate": 3.831958100696675e-05,
+      "loss": 5.8465,
+      "step": 4728
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7500250339508057,
+      "learning_rate": 3.830969909580513e-05,
+      "loss": 5.6855,
+      "step": 4732
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0079612731933594,
+      "learning_rate": 3.829981718464351e-05,
+      "loss": 5.729,
+      "step": 4736
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8388514518737793,
+      "learning_rate": 3.828993527348189e-05,
+      "loss": 5.7142,
+      "step": 4740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8375358581542969,
+      "learning_rate": 3.828005336232027e-05,
+      "loss": 5.6261,
+      "step": 4744
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.785122275352478,
+      "learning_rate": 3.8270171451158654e-05,
+      "loss": 5.7206,
+      "step": 4748
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7444349527359009,
+      "learning_rate": 3.8260289539997036e-05,
+      "loss": 5.711,
+      "step": 4752
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0970168113708496,
+      "learning_rate": 3.825040762883542e-05,
+      "loss": 5.7049,
+      "step": 4756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8633285760879517,
+      "learning_rate": 3.82405257176738e-05,
+      "loss": 5.7639,
+      "step": 4760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7935363054275513,
+      "learning_rate": 3.8230643806512176e-05,
+      "loss": 5.7429,
+      "step": 4764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.099822521209717,
+      "learning_rate": 3.8220761895350564e-05,
+      "loss": 5.6922,
+      "step": 4768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7735134363174438,
+      "learning_rate": 3.8210879984188947e-05,
+      "loss": 5.777,
+      "step": 4772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.62099027633667,
+      "learning_rate": 3.820099807302733e-05,
+      "loss": 5.7513,
+      "step": 4776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.413672924041748,
+      "learning_rate": 3.819111616186571e-05,
+      "loss": 5.7412,
+      "step": 4780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0373451709747314,
+      "learning_rate": 3.818123425070409e-05,
+      "loss": 5.8356,
+      "step": 4784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7266334295272827,
+      "learning_rate": 3.817135233954247e-05,
+      "loss": 5.7796,
+      "step": 4788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5502921342849731,
+      "learning_rate": 3.816147042838085e-05,
+      "loss": 5.6788,
+      "step": 4792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.559435248374939,
+      "learning_rate": 3.815158851721923e-05,
+      "loss": 5.6855,
+      "step": 4796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8041141033172607,
+      "learning_rate": 3.8141706606057614e-05,
+      "loss": 5.712,
+      "step": 4800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.526065468788147,
+      "learning_rate": 3.8131824694895996e-05,
+      "loss": 5.8179,
+      "step": 4804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.73811674118042,
+      "learning_rate": 3.812194278373438e-05,
+      "loss": 5.6526,
+      "step": 4808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.927445650100708,
+      "learning_rate": 3.811206087257276e-05,
+      "loss": 5.7575,
+      "step": 4812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8947980403900146,
+      "learning_rate": 3.810217896141114e-05,
+      "loss": 5.6713,
+      "step": 4816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.586437702178955,
+      "learning_rate": 3.809229705024952e-05,
+      "loss": 5.7355,
+      "step": 4820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.980283498764038,
+      "learning_rate": 3.80824151390879e-05,
+      "loss": 5.6553,
+      "step": 4824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.615174651145935,
+      "learning_rate": 3.807253322792628e-05,
+      "loss": 5.7573,
+      "step": 4828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1818222999572754,
+      "learning_rate": 3.8062651316764664e-05,
+      "loss": 5.6662,
+      "step": 4832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.164574146270752,
+      "learning_rate": 3.8052769405603046e-05,
+      "loss": 5.6212,
+      "step": 4836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.612293243408203,
+      "learning_rate": 3.804288749444143e-05,
+      "loss": 5.6983,
+      "step": 4840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.661539912223816,
+      "learning_rate": 3.803300558327981e-05,
+      "loss": 5.6968,
+      "step": 4844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7480438947677612,
+      "learning_rate": 3.8023123672118185e-05,
+      "loss": 5.6364,
+      "step": 4848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5679854154586792,
+      "learning_rate": 3.801324176095657e-05,
+      "loss": 5.7243,
+      "step": 4852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.688027024269104,
+      "learning_rate": 3.800335984979495e-05,
+      "loss": 5.7444,
+      "step": 4856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0255541801452637,
+      "learning_rate": 3.799347793863333e-05,
+      "loss": 5.7882,
+      "step": 4860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.664778232574463,
+      "learning_rate": 3.7983596027471714e-05,
+      "loss": 5.8742,
+      "step": 4864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7059557437896729,
+      "learning_rate": 3.7973714116310096e-05,
+      "loss": 5.6959,
+      "step": 4868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6631444692611694,
+      "learning_rate": 3.796383220514848e-05,
+      "loss": 5.6701,
+      "step": 4872
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2237651348114014,
+      "learning_rate": 3.795395029398685e-05,
+      "loss": 5.7526,
+      "step": 4876
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5631194114685059,
+      "learning_rate": 3.7944068382825235e-05,
+      "loss": 5.6439,
+      "step": 4880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7338947057724,
+      "learning_rate": 3.7934186471663624e-05,
+      "loss": 5.699,
+      "step": 4884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.682901382446289,
+      "learning_rate": 3.7924304560502006e-05,
+      "loss": 5.7413,
+      "step": 4888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9311175346374512,
+      "learning_rate": 3.791442264934039e-05,
+      "loss": 5.8328,
+      "step": 4892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4761751890182495,
+      "learning_rate": 3.790454073817877e-05,
+      "loss": 5.6389,
+      "step": 4896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6680078506469727,
+      "learning_rate": 3.7894658827017145e-05,
+      "loss": 5.81,
+      "step": 4900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0462136268615723,
+      "learning_rate": 3.788477691585553e-05,
+      "loss": 5.7153,
+      "step": 4904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5725479125976562,
+      "learning_rate": 3.787489500469391e-05,
+      "loss": 5.5877,
+      "step": 4908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8569021224975586,
+      "learning_rate": 3.786501309353229e-05,
+      "loss": 5.6218,
+      "step": 4912
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4288854598999023,
+      "learning_rate": 3.7855131182370674e-05,
+      "loss": 5.6984,
+      "step": 4916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.483189582824707,
+      "learning_rate": 3.7845249271209056e-05,
+      "loss": 5.7004,
+      "step": 4920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9790661334991455,
+      "learning_rate": 3.783536736004744e-05,
+      "loss": 5.6083,
+      "step": 4924
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6783969402313232,
+      "learning_rate": 3.782548544888582e-05,
+      "loss": 5.7083,
+      "step": 4928
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.763135552406311,
+      "learning_rate": 3.7815603537724195e-05,
+      "loss": 5.6475,
+      "step": 4932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8823894262313843,
+      "learning_rate": 3.780572162656258e-05,
+      "loss": 5.741,
+      "step": 4936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9821616411209106,
+      "learning_rate": 3.779583971540096e-05,
+      "loss": 5.7479,
+      "step": 4940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6714750528335571,
+      "learning_rate": 3.778595780423934e-05,
+      "loss": 5.7442,
+      "step": 4944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7158372402191162,
+      "learning_rate": 3.777607589307772e-05,
+      "loss": 5.7097,
+      "step": 4948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8387162685394287,
+      "learning_rate": 3.7766193981916105e-05,
+      "loss": 5.534,
+      "step": 4952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7854334115982056,
+      "learning_rate": 3.775631207075449e-05,
+      "loss": 5.7142,
+      "step": 4956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.736302137374878,
+      "learning_rate": 3.774643015959286e-05,
+      "loss": 5.7211,
+      "step": 4960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1947836875915527,
+      "learning_rate": 3.7736548248431245e-05,
+      "loss": 5.6862,
+      "step": 4964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9635826349258423,
+      "learning_rate": 3.772666633726963e-05,
+      "loss": 5.7067,
+      "step": 4968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4694422483444214,
+      "learning_rate": 3.771678442610801e-05,
+      "loss": 5.6205,
+      "step": 4972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0147790908813477,
+      "learning_rate": 3.770690251494639e-05,
+      "loss": 5.7691,
+      "step": 4976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6146892309188843,
+      "learning_rate": 3.769702060378477e-05,
+      "loss": 5.7103,
+      "step": 4980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.098802089691162,
+      "learning_rate": 3.7687138692623155e-05,
+      "loss": 5.759,
+      "step": 4984
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7809268236160278,
+      "learning_rate": 3.767725678146154e-05,
+      "loss": 5.7146,
+      "step": 4988
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8064768314361572,
+      "learning_rate": 3.766737487029991e-05,
+      "loss": 5.6203,
+      "step": 4992
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6928107738494873,
+      "learning_rate": 3.76574929591383e-05,
+      "loss": 5.8565,
+      "step": 4996
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.65586256980896,
+      "learning_rate": 3.764761104797668e-05,
+      "loss": 5.7671,
+      "step": 5000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0193347930908203,
+      "learning_rate": 3.7637729136815065e-05,
+      "loss": 5.8168,
+      "step": 5004
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.571082592010498,
+      "learning_rate": 3.762784722565345e-05,
+      "loss": 5.6741,
+      "step": 5008
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5919872522354126,
+      "learning_rate": 3.761796531449183e-05,
+      "loss": 5.7805,
+      "step": 5012
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8394559621810913,
+      "learning_rate": 3.7608083403330205e-05,
+      "loss": 5.7168,
+      "step": 5016
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.177563190460205,
+      "learning_rate": 3.759820149216859e-05,
+      "loss": 5.8527,
+      "step": 5020
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9234048128128052,
+      "learning_rate": 3.758831958100697e-05,
+      "loss": 5.6813,
+      "step": 5024
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7297863960266113,
+      "learning_rate": 3.757843766984535e-05,
+      "loss": 5.5673,
+      "step": 5028
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6289762258529663,
+      "learning_rate": 3.756855575868373e-05,
+      "loss": 5.6836,
+      "step": 5032
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9951887130737305,
+      "learning_rate": 3.7558673847522115e-05,
+      "loss": 5.7777,
+      "step": 5036
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6676596403121948,
+      "learning_rate": 3.75487919363605e-05,
+      "loss": 5.731,
+      "step": 5040
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.914063811302185,
+      "learning_rate": 3.753891002519887e-05,
+      "loss": 5.747,
+      "step": 5044
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8118635416030884,
+      "learning_rate": 3.7529028114037254e-05,
+      "loss": 5.5851,
+      "step": 5048
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7621952295303345,
+      "learning_rate": 3.7519146202875637e-05,
+      "loss": 5.8087,
+      "step": 5052
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0561869144439697,
+      "learning_rate": 3.750926429171402e-05,
+      "loss": 5.7153,
+      "step": 5056
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6286967992782593,
+      "learning_rate": 3.74993823805524e-05,
+      "loss": 5.7822,
+      "step": 5060
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.714860439300537,
+      "learning_rate": 3.748950046939078e-05,
+      "loss": 5.7273,
+      "step": 5064
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5165153741836548,
+      "learning_rate": 3.7479618558229165e-05,
+      "loss": 5.6865,
+      "step": 5068
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6314889192581177,
+      "learning_rate": 3.746973664706755e-05,
+      "loss": 5.91,
+      "step": 5072
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.141998529434204,
+      "learning_rate": 3.745985473590592e-05,
+      "loss": 5.796,
+      "step": 5076
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.727159857749939,
+      "learning_rate": 3.7449972824744304e-05,
+      "loss": 5.7056,
+      "step": 5080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6345195770263672,
+      "learning_rate": 3.7440090913582686e-05,
+      "loss": 5.6393,
+      "step": 5084
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8869349956512451,
+      "learning_rate": 3.743020900242107e-05,
+      "loss": 5.6482,
+      "step": 5088
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2009177207946777,
+      "learning_rate": 3.742032709125945e-05,
+      "loss": 5.762,
+      "step": 5092
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.64715576171875,
+      "learning_rate": 3.741044518009783e-05,
+      "loss": 5.6038,
+      "step": 5096
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.730360984802246,
+      "learning_rate": 3.7400563268936215e-05,
+      "loss": 5.6776,
+      "step": 5100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7169520854949951,
+      "learning_rate": 3.739068135777459e-05,
+      "loss": 5.7149,
+      "step": 5104
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6068758964538574,
+      "learning_rate": 3.738079944661297e-05,
+      "loss": 5.7772,
+      "step": 5108
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.215623378753662,
+      "learning_rate": 3.737091753545136e-05,
+      "loss": 5.697,
+      "step": 5112
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.080639123916626,
+      "learning_rate": 3.736103562428974e-05,
+      "loss": 5.6851,
+      "step": 5116
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.041546583175659,
+      "learning_rate": 3.7351153713128125e-05,
+      "loss": 5.7803,
+      "step": 5120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9273161888122559,
+      "learning_rate": 3.734127180196651e-05,
+      "loss": 5.6727,
+      "step": 5124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5295966863632202,
+      "learning_rate": 3.733138989080488e-05,
+      "loss": 5.7002,
+      "step": 5128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5543110370635986,
+      "learning_rate": 3.7321507979643264e-05,
+      "loss": 5.7539,
+      "step": 5132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.867628574371338,
+      "learning_rate": 3.7311626068481646e-05,
+      "loss": 5.6902,
+      "step": 5136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7172714471817017,
+      "learning_rate": 3.730174415732003e-05,
+      "loss": 5.7929,
+      "step": 5140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8065274953842163,
+      "learning_rate": 3.729186224615841e-05,
+      "loss": 5.7148,
+      "step": 5144
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6474413871765137,
+      "learning_rate": 3.728198033499679e-05,
+      "loss": 5.6105,
+      "step": 5148
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2249326705932617,
+      "learning_rate": 3.7272098423835175e-05,
+      "loss": 5.7129,
+      "step": 5152
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8622581958770752,
+      "learning_rate": 3.7262216512673557e-05,
+      "loss": 5.6788,
+      "step": 5156
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7544127702713013,
+      "learning_rate": 3.725233460151193e-05,
+      "loss": 5.6447,
+      "step": 5160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9091047048568726,
+      "learning_rate": 3.7242452690350314e-05,
+      "loss": 5.6228,
+      "step": 5164
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5641545057296753,
+      "learning_rate": 3.7232570779188696e-05,
+      "loss": 5.7202,
+      "step": 5168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8679791688919067,
+      "learning_rate": 3.722268886802708e-05,
+      "loss": 5.7362,
+      "step": 5172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9691587686538696,
+      "learning_rate": 3.721280695686546e-05,
+      "loss": 5.6811,
+      "step": 5176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0612921714782715,
+      "learning_rate": 3.720292504570384e-05,
+      "loss": 5.7,
+      "step": 5180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4250149726867676,
+      "learning_rate": 3.7193043134542224e-05,
+      "loss": 5.7469,
+      "step": 5184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.013101577758789,
+      "learning_rate": 3.71831612233806e-05,
+      "loss": 5.6312,
+      "step": 5188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8997466564178467,
+      "learning_rate": 3.717327931221898e-05,
+      "loss": 5.6744,
+      "step": 5192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4980552196502686,
+      "learning_rate": 3.7163397401057364e-05,
+      "loss": 5.5909,
+      "step": 5196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8174835443496704,
+      "learning_rate": 3.7153515489895746e-05,
+      "loss": 5.7095,
+      "step": 5200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4909565448760986,
+      "learning_rate": 3.714363357873413e-05,
+      "loss": 5.7022,
+      "step": 5204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5424959659576416,
+      "learning_rate": 3.713375166757251e-05,
+      "loss": 5.6704,
+      "step": 5208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5841034650802612,
+      "learning_rate": 3.712386975641089e-05,
+      "loss": 5.6546,
+      "step": 5212
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7105164527893066,
+      "learning_rate": 3.7113987845249274e-05,
+      "loss": 5.6281,
+      "step": 5216
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.422480821609497,
+      "learning_rate": 3.710410593408765e-05,
+      "loss": 5.7145,
+      "step": 5220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8925801515579224,
+      "learning_rate": 3.709422402292603e-05,
+      "loss": 5.6512,
+      "step": 5224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7350256443023682,
+      "learning_rate": 3.708434211176442e-05,
+      "loss": 5.7257,
+      "step": 5228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9060580730438232,
+      "learning_rate": 3.70744602006028e-05,
+      "loss": 5.7056,
+      "step": 5232
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6758359670639038,
+      "learning_rate": 3.7064578289441184e-05,
+      "loss": 5.795,
+      "step": 5236
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7977421283721924,
+      "learning_rate": 3.7054696378279566e-05,
+      "loss": 5.6278,
+      "step": 5240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7420555353164673,
+      "learning_rate": 3.704481446711794e-05,
+      "loss": 5.619,
+      "step": 5244
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7721624374389648,
+      "learning_rate": 3.7034932555956324e-05,
+      "loss": 5.5908,
+      "step": 5248
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9039374589920044,
+      "learning_rate": 3.7025050644794706e-05,
+      "loss": 5.7308,
+      "step": 5252
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9757115840911865,
+      "learning_rate": 3.701516873363309e-05,
+      "loss": 5.7986,
+      "step": 5256
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6613463163375854,
+      "learning_rate": 3.700528682247147e-05,
+      "loss": 5.6056,
+      "step": 5260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9101234674453735,
+      "learning_rate": 3.699540491130985e-05,
+      "loss": 5.7109,
+      "step": 5264
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.04067325592041,
+      "learning_rate": 3.6985523000148234e-05,
+      "loss": 5.7513,
+      "step": 5268
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8164728879928589,
+      "learning_rate": 3.697564108898661e-05,
+      "loss": 5.5739,
+      "step": 5272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9348748922348022,
+      "learning_rate": 3.696575917782499e-05,
+      "loss": 5.6177,
+      "step": 5276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.20129132270813,
+      "learning_rate": 3.695587726666337e-05,
+      "loss": 5.7036,
+      "step": 5280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0877761840820312,
+      "learning_rate": 3.6945995355501755e-05,
+      "loss": 5.6192,
+      "step": 5284
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6023980379104614,
+      "learning_rate": 3.693611344434014e-05,
+      "loss": 5.6557,
+      "step": 5288
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4598197937011719,
+      "learning_rate": 3.692623153317852e-05,
+      "loss": 5.6821,
+      "step": 5292
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9218881130218506,
+      "learning_rate": 3.69163496220169e-05,
+      "loss": 5.7097,
+      "step": 5296
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.91704261302948,
+      "learning_rate": 3.6906467710855284e-05,
+      "loss": 5.7038,
+      "step": 5300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.594611406326294,
+      "learning_rate": 3.689658579969366e-05,
+      "loss": 5.5904,
+      "step": 5304
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.570440649986267,
+      "learning_rate": 3.688670388853204e-05,
+      "loss": 5.5548,
+      "step": 5308
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8532248735427856,
+      "learning_rate": 3.687682197737042e-05,
+      "loss": 5.691,
+      "step": 5312
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8759549856185913,
+      "learning_rate": 3.6866940066208805e-05,
+      "loss": 5.783,
+      "step": 5316
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0743143558502197,
+      "learning_rate": 3.685705815504719e-05,
+      "loss": 5.8346,
+      "step": 5320
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7164572477340698,
+      "learning_rate": 3.684717624388557e-05,
+      "loss": 5.6427,
+      "step": 5324
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8269315958023071,
+      "learning_rate": 3.683729433272395e-05,
+      "loss": 5.7506,
+      "step": 5328
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8456995487213135,
+      "learning_rate": 3.6827412421562327e-05,
+      "loss": 5.6247,
+      "step": 5332
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7062360048294067,
+      "learning_rate": 3.681753051040071e-05,
+      "loss": 5.6777,
+      "step": 5336
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.861578106880188,
+      "learning_rate": 3.68076485992391e-05,
+      "loss": 5.7011,
+      "step": 5340
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5446622371673584,
+      "learning_rate": 3.679776668807748e-05,
+      "loss": 5.6793,
+      "step": 5344
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8392596244812012,
+      "learning_rate": 3.678788477691586e-05,
+      "loss": 5.7193,
+      "step": 5348
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0166003704071045,
+      "learning_rate": 3.6778002865754244e-05,
+      "loss": 5.6509,
+      "step": 5352
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.791098952293396,
+      "learning_rate": 3.676812095459262e-05,
+      "loss": 5.7176,
+      "step": 5356
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9136872291564941,
+      "learning_rate": 3.6758239043431e-05,
+      "loss": 5.8457,
+      "step": 5360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2667651176452637,
+      "learning_rate": 3.674835713226938e-05,
+      "loss": 5.6658,
+      "step": 5364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8071796894073486,
+      "learning_rate": 3.6738475221107765e-05,
+      "loss": 5.7392,
+      "step": 5368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8109067678451538,
+      "learning_rate": 3.672859330994615e-05,
+      "loss": 5.6318,
+      "step": 5372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8517425060272217,
+      "learning_rate": 3.671871139878453e-05,
+      "loss": 5.8122,
+      "step": 5376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8264243602752686,
+      "learning_rate": 3.670882948762291e-05,
+      "loss": 5.5755,
+      "step": 5380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.98093581199646,
+      "learning_rate": 3.669894757646129e-05,
+      "loss": 5.7236,
+      "step": 5384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6828770637512207,
+      "learning_rate": 3.668906566529967e-05,
+      "loss": 5.7345,
+      "step": 5388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.671663761138916,
+      "learning_rate": 3.667918375413805e-05,
+      "loss": 5.5266,
+      "step": 5392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.236020565032959,
+      "learning_rate": 3.666930184297643e-05,
+      "loss": 5.6844,
+      "step": 5396
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.147383213043213,
+      "learning_rate": 3.6659419931814815e-05,
+      "loss": 5.6782,
+      "step": 5400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0428617000579834,
+      "learning_rate": 3.66495380206532e-05,
+      "loss": 5.6924,
+      "step": 5404
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5944437980651855,
+      "learning_rate": 3.663965610949158e-05,
+      "loss": 5.5994,
+      "step": 5408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8434429168701172,
+      "learning_rate": 3.662977419832996e-05,
+      "loss": 5.6628,
+      "step": 5412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6614712476730347,
+      "learning_rate": 3.6619892287168336e-05,
+      "loss": 5.7582,
+      "step": 5416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9906529188156128,
+      "learning_rate": 3.661001037600672e-05,
+      "loss": 5.8577,
+      "step": 5420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6746339797973633,
+      "learning_rate": 3.66001284648451e-05,
+      "loss": 5.6674,
+      "step": 5424
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0132009983062744,
+      "learning_rate": 3.659024655368348e-05,
+      "loss": 5.75,
+      "step": 5428
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6676075458526611,
+      "learning_rate": 3.6580364642521865e-05,
+      "loss": 5.7463,
+      "step": 5432
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.644707441329956,
+      "learning_rate": 3.6570482731360247e-05,
+      "loss": 5.5246,
+      "step": 5436
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6207093000411987,
+      "learning_rate": 3.656060082019863e-05,
+      "loss": 5.6253,
+      "step": 5440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9052598476409912,
+      "learning_rate": 3.6550718909037004e-05,
+      "loss": 5.6262,
+      "step": 5444
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.976206660270691,
+      "learning_rate": 3.6540836997875386e-05,
+      "loss": 5.6091,
+      "step": 5448
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.858864665031433,
+      "learning_rate": 3.653095508671377e-05,
+      "loss": 5.6392,
+      "step": 5452
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7667744159698486,
+      "learning_rate": 3.652107317555216e-05,
+      "loss": 5.7207,
+      "step": 5456
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4944398403167725,
+      "learning_rate": 3.651119126439054e-05,
+      "loss": 5.8588,
+      "step": 5460
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0682997703552246,
+      "learning_rate": 3.650130935322892e-05,
+      "loss": 5.7068,
+      "step": 5464
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.81681489944458,
+      "learning_rate": 3.64914274420673e-05,
+      "loss": 5.6977,
+      "step": 5468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6798757314682007,
+      "learning_rate": 3.648154553090568e-05,
+      "loss": 5.7233,
+      "step": 5472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8230525255203247,
+      "learning_rate": 3.647166361974406e-05,
+      "loss": 5.7243,
+      "step": 5476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4133111238479614,
+      "learning_rate": 3.646178170858244e-05,
+      "loss": 5.598,
+      "step": 5480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9329841136932373,
+      "learning_rate": 3.6451899797420825e-05,
+      "loss": 5.5917,
+      "step": 5484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5919127464294434,
+      "learning_rate": 3.6442017886259207e-05,
+      "loss": 5.6524,
+      "step": 5488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8038116693496704,
+      "learning_rate": 3.643213597509759e-05,
+      "loss": 5.6005,
+      "step": 5492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.781088948249817,
+      "learning_rate": 3.642225406393597e-05,
+      "loss": 5.6194,
+      "step": 5496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1184487342834473,
+      "learning_rate": 3.6412372152774346e-05,
+      "loss": 5.7654,
+      "step": 5500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6980026960372925,
+      "learning_rate": 3.640249024161273e-05,
+      "loss": 5.6293,
+      "step": 5504
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9863158464431763,
+      "learning_rate": 3.639260833045111e-05,
+      "loss": 5.852,
+      "step": 5508
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.160196542739868,
+      "learning_rate": 3.638272641928949e-05,
+      "loss": 5.6083,
+      "step": 5512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8680537939071655,
+      "learning_rate": 3.6372844508127874e-05,
+      "loss": 5.6863,
+      "step": 5516
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.162466526031494,
+      "learning_rate": 3.6362962596966256e-05,
+      "loss": 5.7776,
+      "step": 5520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7698131799697876,
+      "learning_rate": 3.635308068580464e-05,
+      "loss": 5.5958,
+      "step": 5524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8697733879089355,
+      "learning_rate": 3.6343198774643014e-05,
+      "loss": 5.6393,
+      "step": 5528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4883641004562378,
+      "learning_rate": 3.6333316863481396e-05,
+      "loss": 5.6008,
+      "step": 5532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7563178539276123,
+      "learning_rate": 3.632343495231978e-05,
+      "loss": 5.7821,
+      "step": 5536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8179692029953003,
+      "learning_rate": 3.631355304115816e-05,
+      "loss": 5.6816,
+      "step": 5540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5550754070281982,
+      "learning_rate": 3.630367112999654e-05,
+      "loss": 5.6527,
+      "step": 5544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9609134197235107,
+      "learning_rate": 3.6293789218834924e-05,
+      "loss": 5.7238,
+      "step": 5548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6006996631622314,
+      "learning_rate": 3.6283907307673306e-05,
+      "loss": 5.6861,
+      "step": 5552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9726330041885376,
+      "learning_rate": 3.627402539651169e-05,
+      "loss": 5.7777,
+      "step": 5556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6247719526290894,
+      "learning_rate": 3.626414348535006e-05,
+      "loss": 5.7374,
+      "step": 5560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.60383141040802,
+      "learning_rate": 3.6254261574188445e-05,
+      "loss": 5.7191,
+      "step": 5564
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6703567504882812,
+      "learning_rate": 3.624437966302683e-05,
+      "loss": 5.5841,
+      "step": 5568
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7297996282577515,
+      "learning_rate": 3.6234497751865216e-05,
+      "loss": 5.7305,
+      "step": 5572
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7804654836654663,
+      "learning_rate": 3.62246158407036e-05,
+      "loss": 5.6864,
+      "step": 5576
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5437934398651123,
+      "learning_rate": 3.621473392954198e-05,
+      "loss": 5.7737,
+      "step": 5580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8865420818328857,
+      "learning_rate": 3.6204852018380356e-05,
+      "loss": 5.6338,
+      "step": 5584
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0955584049224854,
+      "learning_rate": 3.619497010721874e-05,
+      "loss": 5.7677,
+      "step": 5588
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8487645387649536,
+      "learning_rate": 3.618508819605712e-05,
+      "loss": 5.6992,
+      "step": 5592
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.652840495109558,
+      "learning_rate": 3.61752062848955e-05,
+      "loss": 5.5231,
+      "step": 5596
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.615386724472046,
+      "learning_rate": 3.6165324373733884e-05,
+      "loss": 5.721,
+      "step": 5600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.860023856163025,
+      "learning_rate": 3.6155442462572266e-05,
+      "loss": 5.7149,
+      "step": 5604
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7823337316513062,
+      "learning_rate": 3.614556055141065e-05,
+      "loss": 5.7699,
+      "step": 5608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9639289379119873,
+      "learning_rate": 3.613567864024902e-05,
+      "loss": 5.6564,
+      "step": 5612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6741266250610352,
+      "learning_rate": 3.6125796729087405e-05,
+      "loss": 5.6792,
+      "step": 5616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6273597478866577,
+      "learning_rate": 3.611591481792579e-05,
+      "loss": 5.6275,
+      "step": 5620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.663702368736267,
+      "learning_rate": 3.610603290676417e-05,
+      "loss": 5.6499,
+      "step": 5624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3998336791992188,
+      "learning_rate": 3.609615099560255e-05,
+      "loss": 5.5426,
+      "step": 5628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5529056787490845,
+      "learning_rate": 3.6086269084440934e-05,
+      "loss": 5.7016,
+      "step": 5632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6349724531173706,
+      "learning_rate": 3.6076387173279316e-05,
+      "loss": 5.6359,
+      "step": 5636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6268702745437622,
+      "learning_rate": 3.60665052621177e-05,
+      "loss": 5.68,
+      "step": 5640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.83133065700531,
+      "learning_rate": 3.605662335095607e-05,
+      "loss": 5.644,
+      "step": 5644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7102351188659668,
+      "learning_rate": 3.6046741439794455e-05,
+      "loss": 5.7675,
+      "step": 5648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.901228666305542,
+      "learning_rate": 3.603685952863284e-05,
+      "loss": 5.7194,
+      "step": 5652
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9542912244796753,
+      "learning_rate": 3.602697761747122e-05,
+      "loss": 5.7244,
+      "step": 5656
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9861336946487427,
+      "learning_rate": 3.60170957063096e-05,
+      "loss": 5.5597,
+      "step": 5660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6458784341812134,
+      "learning_rate": 3.600721379514798e-05,
+      "loss": 5.6276,
+      "step": 5664
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3147518634796143,
+      "learning_rate": 3.5997331883986365e-05,
+      "loss": 5.623,
+      "step": 5668
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6608690023422241,
+      "learning_rate": 3.598744997282474e-05,
+      "loss": 5.7577,
+      "step": 5672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4289793968200684,
+      "learning_rate": 3.597756806166312e-05,
+      "loss": 5.6755,
+      "step": 5676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8258533477783203,
+      "learning_rate": 3.5967686150501505e-05,
+      "loss": 5.6546,
+      "step": 5680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.589067816734314,
+      "learning_rate": 3.595780423933989e-05,
+      "loss": 5.6725,
+      "step": 5684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7325801849365234,
+      "learning_rate": 3.5947922328178276e-05,
+      "loss": 5.7298,
+      "step": 5688
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7161133289337158,
+      "learning_rate": 3.593804041701666e-05,
+      "loss": 5.5812,
+      "step": 5692
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6219887733459473,
+      "learning_rate": 3.592815850585503e-05,
+      "loss": 5.5354,
+      "step": 5696
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.099241256713867,
+      "learning_rate": 3.5918276594693415e-05,
+      "loss": 5.7611,
+      "step": 5700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.576979160308838,
+      "learning_rate": 3.59083946835318e-05,
+      "loss": 5.7276,
+      "step": 5704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8135298490524292,
+      "learning_rate": 3.589851277237018e-05,
+      "loss": 5.6826,
+      "step": 5708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8552438020706177,
+      "learning_rate": 3.588863086120856e-05,
+      "loss": 5.6452,
+      "step": 5712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7755975723266602,
+      "learning_rate": 3.587874895004694e-05,
+      "loss": 5.7432,
+      "step": 5716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.728273868560791,
+      "learning_rate": 3.5868867038885325e-05,
+      "loss": 5.8236,
+      "step": 5720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.304983139038086,
+      "learning_rate": 3.585898512772371e-05,
+      "loss": 5.6947,
+      "step": 5724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9431263208389282,
+      "learning_rate": 3.584910321656208e-05,
+      "loss": 5.6718,
+      "step": 5728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6405879259109497,
+      "learning_rate": 3.5839221305400465e-05,
+      "loss": 5.7134,
+      "step": 5732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6948162317276,
+      "learning_rate": 3.582933939423885e-05,
+      "loss": 5.6881,
+      "step": 5736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9962084293365479,
+      "learning_rate": 3.581945748307723e-05,
+      "loss": 5.6576,
+      "step": 5740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2845137119293213,
+      "learning_rate": 3.580957557191561e-05,
+      "loss": 5.6479,
+      "step": 5744
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9110279083251953,
+      "learning_rate": 3.579969366075399e-05,
+      "loss": 5.7157,
+      "step": 5748
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9124606847763062,
+      "learning_rate": 3.5789811749592375e-05,
+      "loss": 5.8124,
+      "step": 5752
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4642013311386108,
+      "learning_rate": 3.577992983843075e-05,
+      "loss": 5.7739,
+      "step": 5756
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.943257451057434,
+      "learning_rate": 3.577004792726913e-05,
+      "loss": 5.6817,
+      "step": 5760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9670331478118896,
+      "learning_rate": 3.5760166016107515e-05,
+      "loss": 5.5489,
+      "step": 5764
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.63418447971344,
+      "learning_rate": 3.5750284104945897e-05,
+      "loss": 5.5785,
+      "step": 5768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7041114568710327,
+      "learning_rate": 3.574040219378428e-05,
+      "loss": 5.7375,
+      "step": 5772
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0340771675109863,
+      "learning_rate": 3.573052028262266e-05,
+      "loss": 5.6921,
+      "step": 5776
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7758631706237793,
+      "learning_rate": 3.572063837146104e-05,
+      "loss": 5.688,
+      "step": 5780
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6245428323745728,
+      "learning_rate": 3.5710756460299425e-05,
+      "loss": 5.7504,
+      "step": 5784
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6857564449310303,
+      "learning_rate": 3.57008745491378e-05,
+      "loss": 5.5688,
+      "step": 5788
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.740413784980774,
+      "learning_rate": 3.569099263797618e-05,
+      "loss": 5.6504,
+      "step": 5792
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8830409049987793,
+      "learning_rate": 3.5681110726814564e-05,
+      "loss": 5.6287,
+      "step": 5796
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8001620769500732,
+      "learning_rate": 3.567122881565295e-05,
+      "loss": 5.8408,
+      "step": 5800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.518395185470581,
+      "learning_rate": 3.5661346904491335e-05,
+      "loss": 5.6564,
+      "step": 5804
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.641120433807373,
+      "learning_rate": 3.565146499332972e-05,
+      "loss": 5.6351,
+      "step": 5808
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.126620292663574,
+      "learning_rate": 3.564158308216809e-05,
+      "loss": 5.6653,
+      "step": 5812
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0950162410736084,
+      "learning_rate": 3.5631701171006475e-05,
+      "loss": 5.63,
+      "step": 5816
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8449667692184448,
+      "learning_rate": 3.5621819259844857e-05,
+      "loss": 5.6507,
+      "step": 5820
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8015673160552979,
+      "learning_rate": 3.561193734868324e-05,
+      "loss": 5.6993,
+      "step": 5824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9131306409835815,
+      "learning_rate": 3.560205543752162e-05,
+      "loss": 5.8073,
+      "step": 5828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5919278860092163,
+      "learning_rate": 3.559217352636e-05,
+      "loss": 5.5955,
+      "step": 5832
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8440625667572021,
+      "learning_rate": 3.5582291615198385e-05,
+      "loss": 5.7592,
+      "step": 5836
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8236738443374634,
+      "learning_rate": 3.557240970403676e-05,
+      "loss": 5.7716,
+      "step": 5840
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5875698328018188,
+      "learning_rate": 3.556252779287514e-05,
+      "loss": 5.6328,
+      "step": 5844
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.722981572151184,
+      "learning_rate": 3.5552645881713524e-05,
+      "loss": 5.576,
+      "step": 5848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9150844812393188,
+      "learning_rate": 3.5542763970551906e-05,
+      "loss": 5.6884,
+      "step": 5852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.071272373199463,
+      "learning_rate": 3.553288205939029e-05,
+      "loss": 5.6853,
+      "step": 5856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.585911512374878,
+      "learning_rate": 3.552300014822867e-05,
+      "loss": 5.8361,
+      "step": 5860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6303049325942993,
+      "learning_rate": 3.551311823706705e-05,
+      "loss": 5.6433,
+      "step": 5864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.914119839668274,
+      "learning_rate": 3.5503236325905435e-05,
+      "loss": 5.6434,
+      "step": 5868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6995179653167725,
+      "learning_rate": 3.549335441474381e-05,
+      "loss": 5.517,
+      "step": 5872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6506705284118652,
+      "learning_rate": 3.548347250358219e-05,
+      "loss": 5.7451,
+      "step": 5876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7888357639312744,
+      "learning_rate": 3.5473590592420574e-05,
+      "loss": 5.7428,
+      "step": 5880
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5513304471969604,
+      "learning_rate": 3.5463708681258956e-05,
+      "loss": 5.6993,
+      "step": 5884
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9307150840759277,
+      "learning_rate": 3.545382677009734e-05,
+      "loss": 5.7173,
+      "step": 5888
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4725440740585327,
+      "learning_rate": 3.544394485893572e-05,
+      "loss": 5.7675,
+      "step": 5892
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1401283740997314,
+      "learning_rate": 3.54340629477741e-05,
+      "loss": 5.6074,
+      "step": 5896
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7221059799194336,
+      "learning_rate": 3.542418103661248e-05,
+      "loss": 5.5691,
+      "step": 5900
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4511234760284424,
+      "learning_rate": 3.541429912545086e-05,
+      "loss": 5.8371,
+      "step": 5904
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6160163879394531,
+      "learning_rate": 3.540441721428924e-05,
+      "loss": 5.5454,
+      "step": 5908
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6199833154678345,
+      "learning_rate": 3.5394535303127624e-05,
+      "loss": 5.618,
+      "step": 5912
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7704286575317383,
+      "learning_rate": 3.538465339196601e-05,
+      "loss": 5.6579,
+      "step": 5916
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.315131425857544,
+      "learning_rate": 3.5374771480804395e-05,
+      "loss": 5.6309,
+      "step": 5920
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7877010107040405,
+      "learning_rate": 3.536488956964277e-05,
+      "loss": 5.7664,
+      "step": 5924
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.655348777770996,
+      "learning_rate": 3.535500765848115e-05,
+      "loss": 5.7179,
+      "step": 5928
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0273170471191406,
+      "learning_rate": 3.5345125747319534e-05,
+      "loss": 5.6274,
+      "step": 5932
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6776584386825562,
+      "learning_rate": 3.5335243836157916e-05,
+      "loss": 5.722,
+      "step": 5936
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.862793207168579,
+      "learning_rate": 3.53253619249963e-05,
+      "loss": 5.6436,
+      "step": 5940
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7735410928726196,
+      "learning_rate": 3.531548001383468e-05,
+      "loss": 5.5571,
+      "step": 5944
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8442201614379883,
+      "learning_rate": 3.530559810267306e-05,
+      "loss": 5.5682,
+      "step": 5948
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9948362112045288,
+      "learning_rate": 3.5295716191511444e-05,
+      "loss": 5.6108,
+      "step": 5952
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7667597532272339,
+      "learning_rate": 3.528583428034982e-05,
+      "loss": 5.6617,
+      "step": 5956
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.569806456565857,
+      "learning_rate": 3.52759523691882e-05,
+      "loss": 5.7098,
+      "step": 5960
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.666054129600525,
+      "learning_rate": 3.5266070458026584e-05,
+      "loss": 5.6734,
+      "step": 5964
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6367340087890625,
+      "learning_rate": 3.5256188546864966e-05,
+      "loss": 5.6654,
+      "step": 5968
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5181485414505005,
+      "learning_rate": 3.524630663570335e-05,
+      "loss": 5.6392,
+      "step": 5972
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.067699432373047,
+      "learning_rate": 3.523642472454173e-05,
+      "loss": 5.6753,
+      "step": 5976
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6809829473495483,
+      "learning_rate": 3.522654281338011e-05,
+      "loss": 5.6829,
+      "step": 5980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5863231420516968,
+      "learning_rate": 3.521666090221849e-05,
+      "loss": 5.5417,
+      "step": 5984
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9677083492279053,
+      "learning_rate": 3.520677899105687e-05,
+      "loss": 5.6524,
+      "step": 5988
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5074371099472046,
+      "learning_rate": 3.519689707989525e-05,
+      "loss": 5.7254,
+      "step": 5992
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.552756667137146,
+      "learning_rate": 3.518701516873363e-05,
+      "loss": 5.5925,
+      "step": 5996
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.689927101135254,
+      "learning_rate": 3.5177133257572015e-05,
+      "loss": 5.629,
+      "step": 6000
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6513864994049072,
+      "learning_rate": 3.51672513464104e-05,
+      "loss": 5.7166,
+      "step": 6004
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8767340183258057,
+      "learning_rate": 3.515736943524878e-05,
+      "loss": 5.7523,
+      "step": 6008
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.079411029815674,
+      "learning_rate": 3.5147487524087155e-05,
+      "loss": 5.7882,
+      "step": 6012
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.992756962776184,
+      "learning_rate": 3.513760561292554e-05,
+      "loss": 5.587,
+      "step": 6016
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.477186679840088,
+      "learning_rate": 3.512772370176392e-05,
+      "loss": 5.7238,
+      "step": 6020
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6832951307296753,
+      "learning_rate": 3.51178417906023e-05,
+      "loss": 5.627,
+      "step": 6024
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8202719688415527,
+      "learning_rate": 3.510795987944068e-05,
+      "loss": 5.6409,
+      "step": 6028
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7379246950149536,
+      "learning_rate": 3.509807796827907e-05,
+      "loss": 5.6045,
+      "step": 6032
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7630070447921753,
+      "learning_rate": 3.5088196057117454e-05,
+      "loss": 5.6474,
+      "step": 6036
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8432222604751587,
+      "learning_rate": 3.507831414595583e-05,
+      "loss": 5.694,
+      "step": 6040
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1126575469970703,
+      "learning_rate": 3.506843223479421e-05,
+      "loss": 5.5952,
+      "step": 6044
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8852741718292236,
+      "learning_rate": 3.5058550323632593e-05,
+      "loss": 5.4868,
+      "step": 6048
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.05767560005188,
+      "learning_rate": 3.5048668412470975e-05,
+      "loss": 5.6139,
+      "step": 6052
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8169785737991333,
+      "learning_rate": 3.503878650130936e-05,
+      "loss": 5.6821,
+      "step": 6056
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8394947052001953,
+      "learning_rate": 3.502890459014774e-05,
+      "loss": 5.7144,
+      "step": 6060
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.533893346786499,
+      "learning_rate": 3.501902267898612e-05,
+      "loss": 5.6372,
+      "step": 6064
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7650554180145264,
+      "learning_rate": 3.50091407678245e-05,
+      "loss": 5.7372,
+      "step": 6068
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7776150703430176,
+      "learning_rate": 3.499925885666288e-05,
+      "loss": 5.5772,
+      "step": 6072
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.804993987083435,
+      "learning_rate": 3.498937694550126e-05,
+      "loss": 5.6967,
+      "step": 6076
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0517046451568604,
+      "learning_rate": 3.497949503433964e-05,
+      "loss": 5.7905,
+      "step": 6080
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6444928646087646,
+      "learning_rate": 3.4969613123178025e-05,
+      "loss": 5.5937,
+      "step": 6084
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.079911231994629,
+      "learning_rate": 3.495973121201641e-05,
+      "loss": 5.6773,
+      "step": 6088
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.929654598236084,
+      "learning_rate": 3.494984930085479e-05,
+      "loss": 5.7453,
+      "step": 6092
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0268120765686035,
+      "learning_rate": 3.4939967389693165e-05,
+      "loss": 5.6351,
+      "step": 6096
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7415413856506348,
+      "learning_rate": 3.493008547853155e-05,
+      "loss": 5.5666,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8862464427947998,
+      "learning_rate": 3.492020356736993e-05,
+      "loss": 5.7066,
+      "step": 6104
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8561365604400635,
+      "learning_rate": 3.491032165620831e-05,
+      "loss": 5.6259,
+      "step": 6108
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.681982398033142,
+      "learning_rate": 3.490043974504669e-05,
+      "loss": 5.5725,
+      "step": 6112
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4180033206939697,
+      "learning_rate": 3.4890557833885075e-05,
+      "loss": 5.6196,
+      "step": 6116
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7663787603378296,
+      "learning_rate": 3.488067592272346e-05,
+      "loss": 5.617,
+      "step": 6120
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6531801223754883,
+      "learning_rate": 3.487079401156184e-05,
+      "loss": 5.5927,
+      "step": 6124
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.739673376083374,
+      "learning_rate": 3.4860912100400214e-05,
+      "loss": 5.6717,
+      "step": 6128
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0703117847442627,
+      "learning_rate": 3.4851030189238596e-05,
+      "loss": 5.6708,
+      "step": 6132
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5977106094360352,
+      "learning_rate": 3.484114827807698e-05,
+      "loss": 5.6917,
+      "step": 6136
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7111542224884033,
+      "learning_rate": 3.483126636691536e-05,
+      "loss": 5.5998,
+      "step": 6140
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.138233184814453,
+      "learning_rate": 3.482138445575375e-05,
+      "loss": 5.6539,
+      "step": 6144
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5501015186309814,
+      "learning_rate": 3.481150254459213e-05,
+      "loss": 5.6418,
+      "step": 6148
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7464877367019653,
+      "learning_rate": 3.480162063343051e-05,
+      "loss": 5.5607,
+      "step": 6152
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0313351154327393,
+      "learning_rate": 3.479173872226889e-05,
+      "loss": 5.7242,
+      "step": 6156
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1652917861938477,
+      "learning_rate": 3.478185681110727e-05,
+      "loss": 5.6722,
+      "step": 6160
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6227622032165527,
+      "learning_rate": 3.477197489994565e-05,
+      "loss": 5.5609,
+      "step": 6164
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.144803047180176,
+      "learning_rate": 3.4762092988784035e-05,
+      "loss": 5.6724,
+      "step": 6168
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0227320194244385,
+      "learning_rate": 3.475221107762242e-05,
+      "loss": 5.6686,
+      "step": 6172
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.794042944908142,
+      "learning_rate": 3.47423291664608e-05,
+      "loss": 5.6792,
+      "step": 6176
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6213692426681519,
+      "learning_rate": 3.4732447255299174e-05,
+      "loss": 5.6446,
+      "step": 6180
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9633342027664185,
+      "learning_rate": 3.4722565344137556e-05,
+      "loss": 5.6725,
+      "step": 6184
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4858800172805786,
+      "learning_rate": 3.471268343297594e-05,
+      "loss": 5.7152,
+      "step": 6188
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4575623273849487,
+      "learning_rate": 3.470280152181432e-05,
+      "loss": 5.6749,
+      "step": 6192
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1202147006988525,
+      "learning_rate": 3.46929196106527e-05,
+      "loss": 5.7791,
+      "step": 6196
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9096757173538208,
+      "learning_rate": 3.4683037699491085e-05,
+      "loss": 5.4813,
+      "step": 6200
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.282806873321533,
+      "learning_rate": 3.467315578832947e-05,
+      "loss": 5.6761,
+      "step": 6204
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6041266918182373,
+      "learning_rate": 3.466327387716785e-05,
+      "loss": 5.7067,
+      "step": 6208
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9719665050506592,
+      "learning_rate": 3.4653391966006224e-05,
+      "loss": 5.6565,
+      "step": 6212
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8574167490005493,
+      "learning_rate": 3.4643510054844606e-05,
+      "loss": 5.6941,
+      "step": 6216
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8125001192092896,
+      "learning_rate": 3.463362814368299e-05,
+      "loss": 5.6066,
+      "step": 6220
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7597132921218872,
+      "learning_rate": 3.462374623252137e-05,
+      "loss": 5.7592,
+      "step": 6224
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0065159797668457,
+      "learning_rate": 3.461386432135975e-05,
+      "loss": 5.7911,
+      "step": 6228
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.03913950920105,
+      "learning_rate": 3.4603982410198134e-05,
+      "loss": 5.6571,
+      "step": 6232
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.114107608795166,
+      "learning_rate": 3.4594100499036516e-05,
+      "loss": 5.7014,
+      "step": 6236
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7447437047958374,
+      "learning_rate": 3.458421858787489e-05,
+      "loss": 5.712,
+      "step": 6240
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8154895305633545,
+      "learning_rate": 3.4574336676713274e-05,
+      "loss": 5.7241,
+      "step": 6244
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1502370834350586,
+      "learning_rate": 3.4564454765551656e-05,
+      "loss": 5.6274,
+      "step": 6248
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6285412311553955,
+      "learning_rate": 3.455457285439004e-05,
+      "loss": 5.6201,
+      "step": 6252
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.825987696647644,
+      "learning_rate": 3.454469094322842e-05,
+      "loss": 5.5851,
+      "step": 6256
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.224214553833008,
+      "learning_rate": 3.453480903206681e-05,
+      "loss": 5.7344,
+      "step": 6260
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8334683179855347,
+      "learning_rate": 3.4524927120905184e-05,
+      "loss": 5.6966,
+      "step": 6264
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.599949598312378,
+      "learning_rate": 3.4515045209743566e-05,
+      "loss": 5.6679,
+      "step": 6268
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9811068773269653,
+      "learning_rate": 3.450516329858195e-05,
+      "loss": 5.7008,
+      "step": 6272
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3280673027038574,
+      "learning_rate": 3.449528138742033e-05,
+      "loss": 5.7588,
+      "step": 6276
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8196971416473389,
+      "learning_rate": 3.448539947625871e-05,
+      "loss": 5.6388,
+      "step": 6280
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4781452417373657,
+      "learning_rate": 3.4475517565097094e-05,
+      "loss": 5.6588,
+      "step": 6284
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2682340145111084,
+      "learning_rate": 3.4465635653935476e-05,
+      "loss": 5.6705,
+      "step": 6288
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7862980365753174,
+      "learning_rate": 3.445575374277386e-05,
+      "loss": 5.6001,
+      "step": 6292
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.870069980621338,
+      "learning_rate": 3.4445871831612234e-05,
+      "loss": 5.6938,
+      "step": 6296
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.120589017868042,
+      "learning_rate": 3.4435989920450616e-05,
+      "loss": 5.633,
+      "step": 6300
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6841241121292114,
+      "learning_rate": 3.4426108009289e-05,
+      "loss": 5.6739,
+      "step": 6304
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8620730638504028,
+      "learning_rate": 3.441622609812738e-05,
+      "loss": 5.6794,
+      "step": 6308
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8764095306396484,
+      "learning_rate": 3.440634418696576e-05,
+      "loss": 5.724,
+      "step": 6312
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4684795141220093,
+      "learning_rate": 3.4396462275804144e-05,
+      "loss": 5.6413,
+      "step": 6316
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5952019691467285,
+      "learning_rate": 3.4386580364642526e-05,
+      "loss": 5.6075,
+      "step": 6320
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8536378145217896,
+      "learning_rate": 3.43766984534809e-05,
+      "loss": 5.6342,
+      "step": 6324
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8524278402328491,
+      "learning_rate": 3.4366816542319283e-05,
+      "loss": 5.6754,
+      "step": 6328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.715560793876648,
+      "learning_rate": 3.4356934631157665e-05,
+      "loss": 5.7763,
+      "step": 6332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8335529565811157,
+      "learning_rate": 3.434705271999605e-05,
+      "loss": 5.6376,
+      "step": 6336
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4588648080825806,
+      "learning_rate": 3.433717080883443e-05,
+      "loss": 5.6589,
+      "step": 6340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.085669755935669,
+      "learning_rate": 3.432728889767281e-05,
+      "loss": 5.6664,
+      "step": 6344
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9969209432601929,
+      "learning_rate": 3.4317406986511194e-05,
+      "loss": 5.6533,
+      "step": 6348
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.169795513153076,
+      "learning_rate": 3.4307525075349576e-05,
+      "loss": 5.7483,
+      "step": 6352
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6342527866363525,
+      "learning_rate": 3.429764316418795e-05,
+      "loss": 5.6761,
+      "step": 6356
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7355843782424927,
+      "learning_rate": 3.428776125302633e-05,
+      "loss": 5.7214,
+      "step": 6360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6192421913146973,
+      "learning_rate": 3.4277879341864715e-05,
+      "loss": 5.629,
+      "step": 6364
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8149397373199463,
+      "learning_rate": 3.42679974307031e-05,
+      "loss": 5.6567,
+      "step": 6368
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9818474054336548,
+      "learning_rate": 3.425811551954148e-05,
+      "loss": 5.6311,
+      "step": 6372
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8166816234588623,
+      "learning_rate": 3.424823360837987e-05,
+      "loss": 5.6253,
+      "step": 6376
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.760532021522522,
+      "learning_rate": 3.4238351697218243e-05,
+      "loss": 5.5845,
+      "step": 6380
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.153517961502075,
+      "learning_rate": 3.4228469786056625e-05,
+      "loss": 5.5968,
+      "step": 6384
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7822462320327759,
+      "learning_rate": 3.421858787489501e-05,
+      "loss": 5.547,
+      "step": 6388
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9486993551254272,
+      "learning_rate": 3.420870596373339e-05,
+      "loss": 5.7527,
+      "step": 6392
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8310834169387817,
+      "learning_rate": 3.419882405257177e-05,
+      "loss": 5.6288,
+      "step": 6396
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9566413164138794,
+      "learning_rate": 3.4188942141410154e-05,
+      "loss": 5.5835,
+      "step": 6400
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8529417514801025,
+      "learning_rate": 3.4179060230248536e-05,
+      "loss": 5.6979,
+      "step": 6404
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4912470579147339,
+      "learning_rate": 3.416917831908691e-05,
+      "loss": 5.6043,
+      "step": 6408
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7634273767471313,
+      "learning_rate": 3.415929640792529e-05,
+      "loss": 5.7066,
+      "step": 6412
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8041954040527344,
+      "learning_rate": 3.4149414496763675e-05,
+      "loss": 5.6787,
+      "step": 6416
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9450712203979492,
+      "learning_rate": 3.413953258560206e-05,
+      "loss": 5.6414,
+      "step": 6420
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1971383094787598,
+      "learning_rate": 3.412965067444044e-05,
+      "loss": 5.6731,
+      "step": 6424
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7693517208099365,
+      "learning_rate": 3.411976876327882e-05,
+      "loss": 5.6599,
+      "step": 6428
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.282921075820923,
+      "learning_rate": 3.4109886852117203e-05,
+      "loss": 5.6397,
+      "step": 6432
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8509403467178345,
+      "learning_rate": 3.4100004940955585e-05,
+      "loss": 5.6985,
+      "step": 6436
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8916041851043701,
+      "learning_rate": 3.409012302979396e-05,
+      "loss": 5.6531,
+      "step": 6440
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.615857481956482,
+      "learning_rate": 3.408024111863234e-05,
+      "loss": 5.605,
+      "step": 6444
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6781501770019531,
+      "learning_rate": 3.4070359207470725e-05,
+      "loss": 5.6553,
+      "step": 6448
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7929623126983643,
+      "learning_rate": 3.406047729630911e-05,
+      "loss": 5.6905,
+      "step": 6452
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9167546033859253,
+      "learning_rate": 3.405059538514749e-05,
+      "loss": 5.614,
+      "step": 6456
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7820324897766113,
+      "learning_rate": 3.404071347398587e-05,
+      "loss": 5.6548,
+      "step": 6460
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.117344379425049,
+      "learning_rate": 3.403083156282425e-05,
+      "loss": 5.6407,
+      "step": 6464
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.829023838043213,
+      "learning_rate": 3.402094965166263e-05,
+      "loss": 5.5939,
+      "step": 6468
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7248526811599731,
+      "learning_rate": 3.401106774050101e-05,
+      "loss": 5.8424,
+      "step": 6472
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.576296329498291,
+      "learning_rate": 3.400118582933939e-05,
+      "loss": 5.5122,
+      "step": 6476
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5682412385940552,
+      "learning_rate": 3.3991303918177775e-05,
+      "loss": 5.6437,
+      "step": 6480
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5419074296951294,
+      "learning_rate": 3.398142200701616e-05,
+      "loss": 5.6232,
+      "step": 6484
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.059434652328491,
+      "learning_rate": 3.3971540095854545e-05,
+      "loss": 5.6276,
+      "step": 6488
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.67318856716156,
+      "learning_rate": 3.396165818469292e-05,
+      "loss": 5.6936,
+      "step": 6492
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6248462200164795,
+      "learning_rate": 3.39517762735313e-05,
+      "loss": 5.6305,
+      "step": 6496
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.914443016052246,
+      "learning_rate": 3.3941894362369685e-05,
+      "loss": 5.5594,
+      "step": 6500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6861499547958374,
+      "learning_rate": 3.393201245120807e-05,
+      "loss": 5.6522,
+      "step": 6504
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8825210332870483,
+      "learning_rate": 3.392213054004645e-05,
+      "loss": 5.7009,
+      "step": 6508
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8397966623306274,
+      "learning_rate": 3.391224862888483e-05,
+      "loss": 5.5723,
+      "step": 6512
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7531960010528564,
+      "learning_rate": 3.390236671772321e-05,
+      "loss": 5.7406,
+      "step": 6516
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8257700204849243,
+      "learning_rate": 3.3892484806561595e-05,
+      "loss": 5.6363,
+      "step": 6520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8056398630142212,
+      "learning_rate": 3.388260289539997e-05,
+      "loss": 5.6686,
+      "step": 6524
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9961930513381958,
+      "learning_rate": 3.387272098423835e-05,
+      "loss": 5.5821,
+      "step": 6528
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7457338571548462,
+      "learning_rate": 3.3862839073076735e-05,
+      "loss": 5.7391,
+      "step": 6532
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8142015933990479,
+      "learning_rate": 3.385295716191512e-05,
+      "loss": 5.6761,
+      "step": 6536
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6556874513626099,
+      "learning_rate": 3.38430752507535e-05,
+      "loss": 5.7818,
+      "step": 6540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9555789232254028,
+      "learning_rate": 3.383319333959188e-05,
+      "loss": 5.7287,
+      "step": 6544
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5307425260543823,
+      "learning_rate": 3.382331142843026e-05,
+      "loss": 5.6996,
+      "step": 6548
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0899040699005127,
+      "learning_rate": 3.381342951726864e-05,
+      "loss": 5.6487,
+      "step": 6552
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.913406491279602,
+      "learning_rate": 3.380354760610702e-05,
+      "loss": 5.5607,
+      "step": 6556
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7522472143173218,
+      "learning_rate": 3.37936656949454e-05,
+      "loss": 5.5921,
+      "step": 6560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8658267259597778,
+      "learning_rate": 3.3783783783783784e-05,
+      "loss": 5.6933,
+      "step": 6564
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.596774697303772,
+      "learning_rate": 3.3773901872622166e-05,
+      "loss": 5.6883,
+      "step": 6568
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.769968032836914,
+      "learning_rate": 3.376401996146055e-05,
+      "loss": 5.6984,
+      "step": 6572
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.154975175857544,
+      "learning_rate": 3.375413805029893e-05,
+      "loss": 5.654,
+      "step": 6576
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5026295185089111,
+      "learning_rate": 3.3744256139137306e-05,
+      "loss": 5.536,
+      "step": 6580
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9100069999694824,
+      "learning_rate": 3.373437422797569e-05,
+      "loss": 5.7633,
+      "step": 6584
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7092385292053223,
+      "learning_rate": 3.372449231681407e-05,
+      "loss": 5.6036,
+      "step": 6588
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4627431631088257,
+      "learning_rate": 3.371461040565245e-05,
+      "loss": 5.6126,
+      "step": 6592
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7545133829116821,
+      "learning_rate": 3.3704728494490834e-05,
+      "loss": 5.6248,
+      "step": 6596
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8125765323638916,
+      "learning_rate": 3.3694846583329216e-05,
+      "loss": 5.6628,
+      "step": 6600
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6120585203170776,
+      "learning_rate": 3.3684964672167605e-05,
+      "loss": 5.595,
+      "step": 6604
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8188247680664062,
+      "learning_rate": 3.367508276100598e-05,
+      "loss": 5.5898,
+      "step": 6608
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7864772081375122,
+      "learning_rate": 3.366520084984436e-05,
+      "loss": 5.6559,
+      "step": 6612
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.448668956756592,
+      "learning_rate": 3.3655318938682744e-05,
+      "loss": 5.7192,
+      "step": 6616
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7164603471755981,
+      "learning_rate": 3.3645437027521126e-05,
+      "loss": 5.613,
+      "step": 6620
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6809614896774292,
+      "learning_rate": 3.363555511635951e-05,
+      "loss": 5.7096,
+      "step": 6624
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6097911596298218,
+      "learning_rate": 3.362567320519789e-05,
+      "loss": 5.7016,
+      "step": 6628
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.04894757270813,
+      "learning_rate": 3.361579129403627e-05,
+      "loss": 5.6882,
+      "step": 6632
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7453163862228394,
+      "learning_rate": 3.360590938287465e-05,
+      "loss": 5.7116,
+      "step": 6636
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0974841117858887,
+      "learning_rate": 3.359602747171303e-05,
+      "loss": 5.685,
+      "step": 6640
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.907728672027588,
+      "learning_rate": 3.358614556055141e-05,
+      "loss": 5.6337,
+      "step": 6644
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.708357334136963,
+      "learning_rate": 3.3576263649389794e-05,
+      "loss": 5.7255,
+      "step": 6648
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9592260122299194,
+      "learning_rate": 3.3566381738228176e-05,
+      "loss": 5.7785,
+      "step": 6652
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7609105110168457,
+      "learning_rate": 3.355649982706656e-05,
+      "loss": 5.6758,
+      "step": 6656
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6743985414505005,
+      "learning_rate": 3.354661791590494e-05,
+      "loss": 5.5952,
+      "step": 6660
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8798311948776245,
+      "learning_rate": 3.3536736004743315e-05,
+      "loss": 5.7004,
+      "step": 6664
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2910819053649902,
+      "learning_rate": 3.35268540935817e-05,
+      "loss": 5.5182,
+      "step": 6668
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6029813289642334,
+      "learning_rate": 3.351697218242008e-05,
+      "loss": 5.7538,
+      "step": 6672
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7451902627944946,
+      "learning_rate": 3.350709027125846e-05,
+      "loss": 5.7265,
+      "step": 6676
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8161840438842773,
+      "learning_rate": 3.3497208360096844e-05,
+      "loss": 5.5846,
+      "step": 6680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.196413993835449,
+      "learning_rate": 3.3487326448935226e-05,
+      "loss": 5.7234,
+      "step": 6684
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.707971215248108,
+      "learning_rate": 3.347744453777361e-05,
+      "loss": 5.6536,
+      "step": 6688
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8692013025283813,
+      "learning_rate": 3.346756262661199e-05,
+      "loss": 5.6981,
+      "step": 6692
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7000290155410767,
+      "learning_rate": 3.3457680715450365e-05,
+      "loss": 5.5349,
+      "step": 6696
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7592395544052124,
+      "learning_rate": 3.344779880428875e-05,
+      "loss": 5.5299,
+      "step": 6700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9749822616577148,
+      "learning_rate": 3.343791689312713e-05,
+      "loss": 5.6149,
+      "step": 6704
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9164543151855469,
+      "learning_rate": 3.342803498196551e-05,
+      "loss": 5.6748,
+      "step": 6708
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7371500730514526,
+      "learning_rate": 3.3418153070803893e-05,
+      "loss": 5.632,
+      "step": 6712
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.007580041885376,
+      "learning_rate": 3.3408271159642275e-05,
+      "loss": 5.6878,
+      "step": 6716
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.049675703048706,
+      "learning_rate": 3.339838924848066e-05,
+      "loss": 5.6347,
+      "step": 6720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5192993879318237,
+      "learning_rate": 3.338850733731904e-05,
+      "loss": 5.5949,
+      "step": 6724
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7811031341552734,
+      "learning_rate": 3.337862542615742e-05,
+      "loss": 5.5796,
+      "step": 6728
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0031299591064453,
+      "learning_rate": 3.3368743514995804e-05,
+      "loss": 5.6231,
+      "step": 6732
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9391804933547974,
+      "learning_rate": 3.3358861603834186e-05,
+      "loss": 5.6029,
+      "step": 6736
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.847928762435913,
+      "learning_rate": 3.334897969267257e-05,
+      "loss": 5.6513,
+      "step": 6740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.11073637008667,
+      "learning_rate": 3.333909778151095e-05,
+      "loss": 5.7021,
+      "step": 6744
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9884241819381714,
+      "learning_rate": 3.3329215870349325e-05,
+      "loss": 5.6343,
+      "step": 6748
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0201947689056396,
+      "learning_rate": 3.331933395918771e-05,
+      "loss": 5.6233,
+      "step": 6752
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.783299207687378,
+      "learning_rate": 3.330945204802609e-05,
+      "loss": 5.6181,
+      "step": 6756
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9971978664398193,
+      "learning_rate": 3.329957013686447e-05,
+      "loss": 5.6573,
+      "step": 6760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.188537120819092,
+      "learning_rate": 3.3289688225702853e-05,
+      "loss": 5.5245,
+      "step": 6764
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8785967826843262,
+      "learning_rate": 3.3279806314541235e-05,
+      "loss": 5.6986,
+      "step": 6768
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.19769024848938,
+      "learning_rate": 3.326992440337962e-05,
+      "loss": 5.5032,
+      "step": 6772
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.074648857116699,
+      "learning_rate": 3.3260042492218e-05,
+      "loss": 5.7612,
+      "step": 6776
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9131178855895996,
+      "learning_rate": 3.3250160581056375e-05,
+      "loss": 5.6805,
+      "step": 6780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2468371391296387,
+      "learning_rate": 3.324027866989476e-05,
+      "loss": 5.6507,
+      "step": 6784
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0636277198791504,
+      "learning_rate": 3.323039675873314e-05,
+      "loss": 5.7031,
+      "step": 6788
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6496247053146362,
+      "learning_rate": 3.322051484757152e-05,
+      "loss": 5.5691,
+      "step": 6792
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7620829343795776,
+      "learning_rate": 3.32106329364099e-05,
+      "loss": 5.6188,
+      "step": 6796
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8972188234329224,
+      "learning_rate": 3.3200751025248285e-05,
+      "loss": 5.6746,
+      "step": 6800
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8980793952941895,
+      "learning_rate": 3.319086911408667e-05,
+      "loss": 5.696,
+      "step": 6804
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.681355357170105,
+      "learning_rate": 3.318098720292504e-05,
+      "loss": 5.629,
+      "step": 6808
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9827635288238525,
+      "learning_rate": 3.3171105291763425e-05,
+      "loss": 5.6046,
+      "step": 6812
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6929621696472168,
+      "learning_rate": 3.316122338060181e-05,
+      "loss": 5.7076,
+      "step": 6816
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.706438660621643,
+      "learning_rate": 3.315134146944019e-05,
+      "loss": 5.6982,
+      "step": 6820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9626370668411255,
+      "learning_rate": 3.314145955827857e-05,
+      "loss": 5.6642,
+      "step": 6824
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.926810383796692,
+      "learning_rate": 3.313157764711695e-05,
+      "loss": 5.6876,
+      "step": 6828
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1014394760131836,
+      "learning_rate": 3.3121695735955335e-05,
+      "loss": 5.6713,
+      "step": 6832
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9832298755645752,
+      "learning_rate": 3.311181382479372e-05,
+      "loss": 5.5695,
+      "step": 6836
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.867805004119873,
+      "learning_rate": 3.31019319136321e-05,
+      "loss": 5.598,
+      "step": 6840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7277621030807495,
+      "learning_rate": 3.309205000247048e-05,
+      "loss": 5.6008,
+      "step": 6844
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0435431003570557,
+      "learning_rate": 3.308216809130886e-05,
+      "loss": 5.5801,
+      "step": 6848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6969705820083618,
+      "learning_rate": 3.3072286180147245e-05,
+      "loss": 5.5302,
+      "step": 6852
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8054941892623901,
+      "learning_rate": 3.306240426898563e-05,
+      "loss": 5.6352,
+      "step": 6856
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7074542045593262,
+      "learning_rate": 3.305252235782401e-05,
+      "loss": 5.5968,
+      "step": 6860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8238881826400757,
+      "learning_rate": 3.3042640446662385e-05,
+      "loss": 5.6272,
+      "step": 6864
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9226677417755127,
+      "learning_rate": 3.303275853550077e-05,
+      "loss": 5.6381,
+      "step": 6868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5835906267166138,
+      "learning_rate": 3.302287662433915e-05,
+      "loss": 5.489,
+      "step": 6872
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7852261066436768,
+      "learning_rate": 3.301299471317753e-05,
+      "loss": 5.7686,
+      "step": 6876
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1643717288970947,
+      "learning_rate": 3.300311280201591e-05,
+      "loss": 5.7131,
+      "step": 6880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8683335781097412,
+      "learning_rate": 3.2993230890854295e-05,
+      "loss": 5.6893,
+      "step": 6884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8597195148468018,
+      "learning_rate": 3.298334897969268e-05,
+      "loss": 5.5943,
+      "step": 6888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9217149019241333,
+      "learning_rate": 3.297346706853105e-05,
+      "loss": 5.6866,
+      "step": 6892
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7497360706329346,
+      "learning_rate": 3.2963585157369434e-05,
+      "loss": 5.6797,
+      "step": 6896
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8368375301361084,
+      "learning_rate": 3.2953703246207816e-05,
+      "loss": 5.5535,
+      "step": 6900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6737070083618164,
+      "learning_rate": 3.29438213350462e-05,
+      "loss": 5.5282,
+      "step": 6904
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5021297931671143,
+      "learning_rate": 3.293393942388458e-05,
+      "loss": 5.6176,
+      "step": 6908
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.836517572402954,
+      "learning_rate": 3.292405751272296e-05,
+      "loss": 5.6304,
+      "step": 6912
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9427759647369385,
+      "learning_rate": 3.2914175601561345e-05,
+      "loss": 5.6843,
+      "step": 6916
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1088333129882812,
+      "learning_rate": 3.290429369039973e-05,
+      "loss": 5.6674,
+      "step": 6920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9784494638442993,
+      "learning_rate": 3.28944117792381e-05,
+      "loss": 5.6965,
+      "step": 6924
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.666114091873169,
+      "learning_rate": 3.2884529868076484e-05,
+      "loss": 5.5627,
+      "step": 6928
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7081410884857178,
+      "learning_rate": 3.2874647956914866e-05,
+      "loss": 5.7073,
+      "step": 6932
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.857577919960022,
+      "learning_rate": 3.286476604575325e-05,
+      "loss": 5.6804,
+      "step": 6936
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.952774167060852,
+      "learning_rate": 3.285488413459163e-05,
+      "loss": 5.6911,
+      "step": 6940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.695110559463501,
+      "learning_rate": 3.284500222343001e-05,
+      "loss": 5.5463,
+      "step": 6944
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7788817882537842,
+      "learning_rate": 3.2835120312268394e-05,
+      "loss": 5.7336,
+      "step": 6948
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.170694589614868,
+      "learning_rate": 3.2825238401106776e-05,
+      "loss": 5.6662,
+      "step": 6952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9008727073669434,
+      "learning_rate": 3.281535648994516e-05,
+      "loss": 5.6276,
+      "step": 6956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.079054355621338,
+      "learning_rate": 3.280547457878354e-05,
+      "loss": 5.6058,
+      "step": 6960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6386624574661255,
+      "learning_rate": 3.279559266762192e-05,
+      "loss": 5.5782,
+      "step": 6964
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9276303052902222,
+      "learning_rate": 3.2785710756460305e-05,
+      "loss": 5.6403,
+      "step": 6968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7339591979980469,
+      "learning_rate": 3.277582884529869e-05,
+      "loss": 5.6689,
+      "step": 6972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9973163604736328,
+      "learning_rate": 3.276594693413706e-05,
+      "loss": 5.6529,
+      "step": 6976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5551352500915527,
+      "learning_rate": 3.2756065022975444e-05,
+      "loss": 5.609,
+      "step": 6980
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8106021881103516,
+      "learning_rate": 3.2746183111813826e-05,
+      "loss": 5.5107,
+      "step": 6984
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6934738159179688,
+      "learning_rate": 3.273630120065221e-05,
+      "loss": 5.6064,
+      "step": 6988
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6493241786956787,
+      "learning_rate": 3.272641928949059e-05,
+      "loss": 5.6641,
+      "step": 6992
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.817825436592102,
+      "learning_rate": 3.271653737832897e-05,
+      "loss": 5.5306,
+      "step": 6996
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5873377323150635,
+      "learning_rate": 3.2706655467167354e-05,
+      "loss": 5.6717,
+      "step": 7000
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.584267497062683,
+      "learning_rate": 3.2696773556005736e-05,
+      "loss": 5.5343,
+      "step": 7004
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.52151620388031,
+      "learning_rate": 3.268689164484411e-05,
+      "loss": 5.612,
+      "step": 7008
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8319940567016602,
+      "learning_rate": 3.2677009733682494e-05,
+      "loss": 5.611,
+      "step": 7012
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.779436469078064,
+      "learning_rate": 3.2667127822520876e-05,
+      "loss": 5.5905,
+      "step": 7016
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5194777250289917,
+      "learning_rate": 3.265724591135926e-05,
+      "loss": 5.6275,
+      "step": 7020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6653060913085938,
+      "learning_rate": 3.264736400019764e-05,
+      "loss": 5.6553,
+      "step": 7024
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6240752935409546,
+      "learning_rate": 3.263748208903602e-05,
+      "loss": 5.6956,
+      "step": 7028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6351805925369263,
+      "learning_rate": 3.2627600177874404e-05,
+      "loss": 5.6165,
+      "step": 7032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8824491500854492,
+      "learning_rate": 3.261771826671278e-05,
+      "loss": 5.5225,
+      "step": 7036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.713706612586975,
+      "learning_rate": 3.260783635555116e-05,
+      "loss": 5.6428,
+      "step": 7040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0658748149871826,
+      "learning_rate": 3.2597954444389543e-05,
+      "loss": 5.6175,
+      "step": 7044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7719210386276245,
+      "learning_rate": 3.2588072533227926e-05,
+      "loss": 5.6819,
+      "step": 7048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1328206062316895,
+      "learning_rate": 3.257819062206631e-05,
+      "loss": 5.5427,
+      "step": 7052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5215253829956055,
+      "learning_rate": 3.256830871090469e-05,
+      "loss": 5.7679,
+      "step": 7056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.001835584640503,
+      "learning_rate": 3.255842679974307e-05,
+      "loss": 5.748,
+      "step": 7060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5775867700576782,
+      "learning_rate": 3.2548544888581454e-05,
+      "loss": 5.5609,
+      "step": 7064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7867364883422852,
+      "learning_rate": 3.2538662977419836e-05,
+      "loss": 5.6232,
+      "step": 7068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.010613203048706,
+      "learning_rate": 3.252878106625822e-05,
+      "loss": 5.6611,
+      "step": 7072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0896737575531006,
+      "learning_rate": 3.25188991550966e-05,
+      "loss": 5.6961,
+      "step": 7076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7435530424118042,
+      "learning_rate": 3.250901724393498e-05,
+      "loss": 5.816,
+      "step": 7080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.452756643295288,
+      "learning_rate": 3.2499135332773364e-05,
+      "loss": 5.6278,
+      "step": 7084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5467629432678223,
+      "learning_rate": 3.2489253421611746e-05,
+      "loss": 5.7114,
+      "step": 7088
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8225852251052856,
+      "learning_rate": 3.247937151045012e-05,
+      "loss": 5.5523,
+      "step": 7092
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9562902450561523,
+      "learning_rate": 3.2469489599288503e-05,
+      "loss": 5.7284,
+      "step": 7096
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.379361152648926,
+      "learning_rate": 3.2459607688126886e-05,
+      "loss": 5.5628,
+      "step": 7100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9397469758987427,
+      "learning_rate": 3.244972577696527e-05,
+      "loss": 5.6596,
+      "step": 7104
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7778300046920776,
+      "learning_rate": 3.243984386580365e-05,
+      "loss": 5.7102,
+      "step": 7108
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.671886920928955,
+      "learning_rate": 3.242996195464203e-05,
+      "loss": 5.7244,
+      "step": 7112
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7230006456375122,
+      "learning_rate": 3.2420080043480414e-05,
+      "loss": 5.7147,
+      "step": 7116
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.751182198524475,
+      "learning_rate": 3.241019813231879e-05,
+      "loss": 5.7059,
+      "step": 7120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7081093788146973,
+      "learning_rate": 3.240031622115717e-05,
+      "loss": 5.5171,
+      "step": 7124
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7573639154434204,
+      "learning_rate": 3.239043430999555e-05,
+      "loss": 5.6948,
+      "step": 7128
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7177603244781494,
+      "learning_rate": 3.2380552398833935e-05,
+      "loss": 5.5412,
+      "step": 7132
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8538342714309692,
+      "learning_rate": 3.237067048767232e-05,
+      "loss": 5.645,
+      "step": 7136
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.174427032470703,
+      "learning_rate": 3.23607885765107e-05,
+      "loss": 5.6742,
+      "step": 7140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.461571455001831,
+      "learning_rate": 3.235090666534908e-05,
+      "loss": 5.6642,
+      "step": 7144
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8825712203979492,
+      "learning_rate": 3.234102475418746e-05,
+      "loss": 5.6553,
+      "step": 7148
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5318057537078857,
+      "learning_rate": 3.233114284302584e-05,
+      "loss": 5.6551,
+      "step": 7152
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0473759174346924,
+      "learning_rate": 3.232126093186422e-05,
+      "loss": 5.6036,
+      "step": 7156
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.186314821243286,
+      "learning_rate": 3.23113790207026e-05,
+      "loss": 5.762,
+      "step": 7160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6777757406234741,
+      "learning_rate": 3.2301497109540985e-05,
+      "loss": 5.6549,
+      "step": 7164
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.951010823249817,
+      "learning_rate": 3.229161519837937e-05,
+      "loss": 5.7491,
+      "step": 7168
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.590847134590149,
+      "learning_rate": 3.228173328721775e-05,
+      "loss": 5.52,
+      "step": 7172
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1618316173553467,
+      "learning_rate": 3.227185137605613e-05,
+      "loss": 5.5362,
+      "step": 7176
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.874100685119629,
+      "learning_rate": 3.226196946489451e-05,
+      "loss": 5.5656,
+      "step": 7180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7269055843353271,
+      "learning_rate": 3.2252087553732895e-05,
+      "loss": 5.5793,
+      "step": 7184
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7733286619186401,
+      "learning_rate": 3.224220564257128e-05,
+      "loss": 5.6867,
+      "step": 7188
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6358729600906372,
+      "learning_rate": 3.223232373140966e-05,
+      "loss": 5.8119,
+      "step": 7192
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.663542628288269,
+      "learning_rate": 3.222244182024804e-05,
+      "loss": 5.585,
+      "step": 7196
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0216832160949707,
+      "learning_rate": 3.2212559909086423e-05,
+      "loss": 5.5331,
+      "step": 7200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9756247997283936,
+      "learning_rate": 3.22026779979248e-05,
+      "loss": 5.7117,
+      "step": 7204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0371248722076416,
+      "learning_rate": 3.219279608676318e-05,
+      "loss": 5.6412,
+      "step": 7208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8078463077545166,
+      "learning_rate": 3.218291417560156e-05,
+      "loss": 5.5826,
+      "step": 7212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6436288356781006,
+      "learning_rate": 3.2173032264439945e-05,
+      "loss": 5.5825,
+      "step": 7216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0301754474639893,
+      "learning_rate": 3.216315035327833e-05,
+      "loss": 5.7317,
+      "step": 7220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0184414386749268,
+      "learning_rate": 3.215326844211671e-05,
+      "loss": 5.5583,
+      "step": 7224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9053844213485718,
+      "learning_rate": 3.214338653095509e-05,
+      "loss": 5.6326,
+      "step": 7228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7837705612182617,
+      "learning_rate": 3.2133504619793466e-05,
+      "loss": 5.6659,
+      "step": 7232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1691172122955322,
+      "learning_rate": 3.212362270863185e-05,
+      "loss": 5.6534,
+      "step": 7236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6231029033660889,
+      "learning_rate": 3.211374079747023e-05,
+      "loss": 5.5752,
+      "step": 7240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6797888278961182,
+      "learning_rate": 3.210385888630861e-05,
+      "loss": 5.5594,
+      "step": 7244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0142295360565186,
+      "learning_rate": 3.2093976975146995e-05,
+      "loss": 5.5551,
+      "step": 7248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6823694705963135,
+      "learning_rate": 3.208409506398538e-05,
+      "loss": 5.5654,
+      "step": 7252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.035011053085327,
+      "learning_rate": 3.207421315282376e-05,
+      "loss": 5.753,
+      "step": 7256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0539419651031494,
+      "learning_rate": 3.206433124166214e-05,
+      "loss": 5.481,
+      "step": 7260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8874220848083496,
+      "learning_rate": 3.2054449330500516e-05,
+      "loss": 5.5776,
+      "step": 7264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.067695140838623,
+      "learning_rate": 3.20445674193389e-05,
+      "loss": 5.7252,
+      "step": 7268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9559959173202515,
+      "learning_rate": 3.203468550817728e-05,
+      "loss": 5.7496,
+      "step": 7272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7679195404052734,
+      "learning_rate": 3.202480359701566e-05,
+      "loss": 5.5574,
+      "step": 7276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7871371507644653,
+      "learning_rate": 3.2014921685854044e-05,
+      "loss": 5.611,
+      "step": 7280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.027100086212158,
+      "learning_rate": 3.2005039774692426e-05,
+      "loss": 5.6949,
+      "step": 7284
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3536198139190674,
+      "learning_rate": 3.199515786353081e-05,
+      "loss": 5.5937,
+      "step": 7288
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9089040756225586,
+      "learning_rate": 3.198527595236919e-05,
+      "loss": 5.742,
+      "step": 7292
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9175152778625488,
+      "learning_rate": 3.197539404120757e-05,
+      "loss": 5.7345,
+      "step": 7296
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3115925788879395,
+      "learning_rate": 3.1965512130045955e-05,
+      "loss": 5.5774,
+      "step": 7300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8105931282043457,
+      "learning_rate": 3.195563021888434e-05,
+      "loss": 5.5753,
+      "step": 7304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1161036491394043,
+      "learning_rate": 3.194574830772272e-05,
+      "loss": 5.6975,
+      "step": 7308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7692967653274536,
+      "learning_rate": 3.19358663965611e-05,
+      "loss": 5.6869,
+      "step": 7312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7916195392608643,
+      "learning_rate": 3.1925984485399476e-05,
+      "loss": 5.7502,
+      "step": 7316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5649669170379639,
+      "learning_rate": 3.191610257423786e-05,
+      "loss": 5.6049,
+      "step": 7320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5286930799484253,
+      "learning_rate": 3.190622066307624e-05,
+      "loss": 5.5598,
+      "step": 7324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7816998958587646,
+      "learning_rate": 3.189633875191462e-05,
+      "loss": 5.5383,
+      "step": 7328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7937431335449219,
+      "learning_rate": 3.1886456840753004e-05,
+      "loss": 5.6448,
+      "step": 7332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.801894187927246,
+      "learning_rate": 3.1876574929591386e-05,
+      "loss": 5.638,
+      "step": 7336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8095769882202148,
+      "learning_rate": 3.186669301842977e-05,
+      "loss": 5.7017,
+      "step": 7340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.600150465965271,
+      "learning_rate": 3.185681110726815e-05,
+      "loss": 5.639,
+      "step": 7344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.827073574066162,
+      "learning_rate": 3.1846929196106526e-05,
+      "loss": 5.6148,
+      "step": 7348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7427455186843872,
+      "learning_rate": 3.183704728494491e-05,
+      "loss": 5.6816,
+      "step": 7352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5865864753723145,
+      "learning_rate": 3.182716537378329e-05,
+      "loss": 5.6173,
+      "step": 7356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4280364513397217,
+      "learning_rate": 3.181728346262167e-05,
+      "loss": 5.7189,
+      "step": 7360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7864285707473755,
+      "learning_rate": 3.1807401551460054e-05,
+      "loss": 5.7009,
+      "step": 7364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9539587497711182,
+      "learning_rate": 3.1797519640298436e-05,
+      "loss": 5.6526,
+      "step": 7368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8817050457000732,
+      "learning_rate": 3.178763772913682e-05,
+      "loss": 5.5829,
+      "step": 7372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9259823560714722,
+      "learning_rate": 3.1777755817975193e-05,
+      "loss": 5.5862,
+      "step": 7376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8345321416854858,
+      "learning_rate": 3.1767873906813576e-05,
+      "loss": 5.689,
+      "step": 7380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9445204734802246,
+      "learning_rate": 3.175799199565196e-05,
+      "loss": 5.5377,
+      "step": 7384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9036486148834229,
+      "learning_rate": 3.174811008449034e-05,
+      "loss": 5.6994,
+      "step": 7388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7880631685256958,
+      "learning_rate": 3.173822817332872e-05,
+      "loss": 5.5478,
+      "step": 7392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7411099672317505,
+      "learning_rate": 3.1728346262167104e-05,
+      "loss": 5.6932,
+      "step": 7396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0032172203063965,
+      "learning_rate": 3.1718464351005486e-05,
+      "loss": 5.6742,
+      "step": 7400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.963858723640442,
+      "learning_rate": 3.170858243984387e-05,
+      "loss": 5.6343,
+      "step": 7404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7364962100982666,
+      "learning_rate": 3.169870052868225e-05,
+      "loss": 5.6262,
+      "step": 7408
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.885438084602356,
+      "learning_rate": 3.168881861752063e-05,
+      "loss": 5.6402,
+      "step": 7412
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7210749387741089,
+      "learning_rate": 3.1678936706359014e-05,
+      "loss": 5.6839,
+      "step": 7416
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7860082387924194,
+      "learning_rate": 3.1669054795197396e-05,
+      "loss": 5.5595,
+      "step": 7420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.878787636756897,
+      "learning_rate": 3.165917288403578e-05,
+      "loss": 5.6312,
+      "step": 7424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6183253526687622,
+      "learning_rate": 3.164929097287416e-05,
+      "loss": 5.6786,
+      "step": 7428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8205534219741821,
+      "learning_rate": 3.1639409061712536e-05,
+      "loss": 5.6006,
+      "step": 7432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9625846147537231,
+      "learning_rate": 3.162952715055092e-05,
+      "loss": 5.5243,
+      "step": 7436
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.967149257659912,
+      "learning_rate": 3.16196452393893e-05,
+      "loss": 5.599,
+      "step": 7440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0951106548309326,
+      "learning_rate": 3.160976332822768e-05,
+      "loss": 5.6613,
+      "step": 7444
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8848724365234375,
+      "learning_rate": 3.1599881417066064e-05,
+      "loss": 5.7463,
+      "step": 7448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.793150544166565,
+      "learning_rate": 3.1589999505904446e-05,
+      "loss": 5.7087,
+      "step": 7452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5212810039520264,
+      "learning_rate": 3.158011759474283e-05,
+      "loss": 5.6169,
+      "step": 7456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9898124933242798,
+      "learning_rate": 3.15702356835812e-05,
+      "loss": 5.5823,
+      "step": 7460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9081612825393677,
+      "learning_rate": 3.1560353772419585e-05,
+      "loss": 5.673,
+      "step": 7464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.049330234527588,
+      "learning_rate": 3.155047186125797e-05,
+      "loss": 5.6275,
+      "step": 7468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.874483346939087,
+      "learning_rate": 3.154058995009635e-05,
+      "loss": 5.7012,
+      "step": 7472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.674302101135254,
+      "learning_rate": 3.153070803893473e-05,
+      "loss": 5.6507,
+      "step": 7476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1794581413269043,
+      "learning_rate": 3.1520826127773113e-05,
+      "loss": 5.6312,
+      "step": 7480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8025480508804321,
+      "learning_rate": 3.1510944216611496e-05,
+      "loss": 5.5184,
+      "step": 7484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8189462423324585,
+      "learning_rate": 3.150106230544988e-05,
+      "loss": 5.558,
+      "step": 7488
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8474452495574951,
+      "learning_rate": 3.149118039428825e-05,
+      "loss": 5.5896,
+      "step": 7492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8259671926498413,
+      "learning_rate": 3.1481298483126635e-05,
+      "loss": 5.6092,
+      "step": 7496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.675440788269043,
+      "learning_rate": 3.147141657196502e-05,
+      "loss": 5.5084,
+      "step": 7500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.073160409927368,
+      "learning_rate": 3.14615346608034e-05,
+      "loss": 5.6257,
+      "step": 7504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9725701808929443,
+      "learning_rate": 3.145165274964178e-05,
+      "loss": 5.5622,
+      "step": 7508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.097548007965088,
+      "learning_rate": 3.144177083848016e-05,
+      "loss": 5.5948,
+      "step": 7512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8091487884521484,
+      "learning_rate": 3.1431888927318545e-05,
+      "loss": 5.6698,
+      "step": 7516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8227310180664062,
+      "learning_rate": 3.142200701615692e-05,
+      "loss": 5.5319,
+      "step": 7520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9433192014694214,
+      "learning_rate": 3.141212510499531e-05,
+      "loss": 5.6118,
+      "step": 7524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7683522701263428,
+      "learning_rate": 3.140224319383369e-05,
+      "loss": 5.5843,
+      "step": 7528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8720110654830933,
+      "learning_rate": 3.1392361282672073e-05,
+      "loss": 5.626,
+      "step": 7532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7826591730117798,
+      "learning_rate": 3.1382479371510456e-05,
+      "loss": 5.6833,
+      "step": 7536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.870161533355713,
+      "learning_rate": 3.137259746034884e-05,
+      "loss": 5.7177,
+      "step": 7540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.076082706451416,
+      "learning_rate": 3.136271554918721e-05,
+      "loss": 5.563,
+      "step": 7544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7204762697219849,
+      "learning_rate": 3.1352833638025595e-05,
+      "loss": 5.5931,
+      "step": 7548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.109579563140869,
+      "learning_rate": 3.134295172686398e-05,
+      "loss": 5.6927,
+      "step": 7552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7823752164840698,
+      "learning_rate": 3.133306981570236e-05,
+      "loss": 5.501,
+      "step": 7556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9205490350723267,
+      "learning_rate": 3.132318790454074e-05,
+      "loss": 5.5822,
+      "step": 7560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1453466415405273,
+      "learning_rate": 3.131330599337912e-05,
+      "loss": 5.6234,
+      "step": 7564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.079413652420044,
+      "learning_rate": 3.1303424082217505e-05,
+      "loss": 5.5851,
+      "step": 7568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7375316619873047,
+      "learning_rate": 3.129354217105589e-05,
+      "loss": 5.51,
+      "step": 7572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7538070678710938,
+      "learning_rate": 3.128366025989426e-05,
+      "loss": 5.5565,
+      "step": 7576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8971534967422485,
+      "learning_rate": 3.1273778348732645e-05,
+      "loss": 5.6077,
+      "step": 7580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.975240707397461,
+      "learning_rate": 3.126389643757103e-05,
+      "loss": 5.8038,
+      "step": 7584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9399492740631104,
+      "learning_rate": 3.125401452640941e-05,
+      "loss": 5.6554,
+      "step": 7588
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.004110336303711,
+      "learning_rate": 3.124413261524779e-05,
+      "loss": 5.6242,
+      "step": 7592
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8853540420532227,
+      "learning_rate": 3.123425070408617e-05,
+      "loss": 5.7207,
+      "step": 7596
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.765731930732727,
+      "learning_rate": 3.1224368792924555e-05,
+      "loss": 5.5813,
+      "step": 7600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8963857889175415,
+      "learning_rate": 3.121448688176293e-05,
+      "loss": 5.6913,
+      "step": 7604
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8706163167953491,
+      "learning_rate": 3.120460497060131e-05,
+      "loss": 5.5406,
+      "step": 7608
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6852182149887085,
+      "learning_rate": 3.1194723059439694e-05,
+      "loss": 5.582,
+      "step": 7612
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8383783102035522,
+      "learning_rate": 3.1184841148278076e-05,
+      "loss": 5.664,
+      "step": 7616
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8068432807922363,
+      "learning_rate": 3.117495923711646e-05,
+      "loss": 5.6969,
+      "step": 7620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.020507335662842,
+      "learning_rate": 3.116507732595484e-05,
+      "loss": 5.5544,
+      "step": 7624
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7080477476119995,
+      "learning_rate": 3.115519541479322e-05,
+      "loss": 5.7007,
+      "step": 7628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6896196603775024,
+      "learning_rate": 3.11453135036316e-05,
+      "loss": 5.5952,
+      "step": 7632
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0468990802764893,
+      "learning_rate": 3.113543159246998e-05,
+      "loss": 5.735,
+      "step": 7636
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9977657794952393,
+      "learning_rate": 3.112554968130837e-05,
+      "loss": 5.7005,
+      "step": 7640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3371498584747314,
+      "learning_rate": 3.111566777014675e-05,
+      "loss": 5.6963,
+      "step": 7644
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8989230394363403,
+      "learning_rate": 3.110578585898513e-05,
+      "loss": 5.5717,
+      "step": 7648
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1724061965942383,
+      "learning_rate": 3.1095903947823515e-05,
+      "loss": 5.5531,
+      "step": 7652
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7669955492019653,
+      "learning_rate": 3.10860220366619e-05,
+      "loss": 5.6975,
+      "step": 7656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1776323318481445,
+      "learning_rate": 3.107614012550027e-05,
+      "loss": 5.7217,
+      "step": 7660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5925564765930176,
+      "learning_rate": 3.1066258214338654e-05,
+      "loss": 5.6475,
+      "step": 7664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.862136721611023,
+      "learning_rate": 3.1056376303177036e-05,
+      "loss": 5.6022,
+      "step": 7668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.893431305885315,
+      "learning_rate": 3.104649439201542e-05,
+      "loss": 5.5991,
+      "step": 7672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9323973655700684,
+      "learning_rate": 3.10366124808538e-05,
+      "loss": 5.737,
+      "step": 7676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.630057454109192,
+      "learning_rate": 3.102673056969218e-05,
+      "loss": 5.5885,
+      "step": 7680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0866503715515137,
+      "learning_rate": 3.1016848658530565e-05,
+      "loss": 5.5873,
+      "step": 7684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6265943050384521,
+      "learning_rate": 3.100696674736894e-05,
+      "loss": 5.6085,
+      "step": 7688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.116450309753418,
+      "learning_rate": 3.099708483620732e-05,
+      "loss": 5.6814,
+      "step": 7692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8399298191070557,
+      "learning_rate": 3.0987202925045704e-05,
+      "loss": 5.6584,
+      "step": 7696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8756877183914185,
+      "learning_rate": 3.0977321013884086e-05,
+      "loss": 5.7231,
+      "step": 7700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9492945671081543,
+      "learning_rate": 3.096743910272247e-05,
+      "loss": 5.6126,
+      "step": 7704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8977458477020264,
+      "learning_rate": 3.095755719156085e-05,
+      "loss": 5.4845,
+      "step": 7708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.920137643814087,
+      "learning_rate": 3.094767528039923e-05,
+      "loss": 5.6911,
+      "step": 7712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9091439247131348,
+      "learning_rate": 3.093779336923761e-05,
+      "loss": 5.7215,
+      "step": 7716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9468705654144287,
+      "learning_rate": 3.092791145807599e-05,
+      "loss": 5.5602,
+      "step": 7720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.171674966812134,
+      "learning_rate": 3.091802954691437e-05,
+      "loss": 5.6432,
+      "step": 7724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7514827251434326,
+      "learning_rate": 3.0908147635752754e-05,
+      "loss": 5.7022,
+      "step": 7728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6173917055130005,
+      "learning_rate": 3.0898265724591136e-05,
+      "loss": 5.6705,
+      "step": 7732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7290568351745605,
+      "learning_rate": 3.088838381342952e-05,
+      "loss": 5.5759,
+      "step": 7736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6534110307693481,
+      "learning_rate": 3.08785019022679e-05,
+      "loss": 5.7897,
+      "step": 7740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1814112663269043,
+      "learning_rate": 3.086861999110628e-05,
+      "loss": 5.6196,
+      "step": 7744
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8534756898880005,
+      "learning_rate": 3.085873807994466e-05,
+      "loss": 5.5227,
+      "step": 7748
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9489648342132568,
+      "learning_rate": 3.0848856168783046e-05,
+      "loss": 5.647,
+      "step": 7752
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.976659893989563,
+      "learning_rate": 3.083897425762143e-05,
+      "loss": 5.5386,
+      "step": 7756
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.00985050201416,
+      "learning_rate": 3.082909234645981e-05,
+      "loss": 5.723,
+      "step": 7760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8619848489761353,
+      "learning_rate": 3.081921043529819e-05,
+      "loss": 5.7385,
+      "step": 7764
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6431546211242676,
+      "learning_rate": 3.0809328524136574e-05,
+      "loss": 5.5918,
+      "step": 7768
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8953239917755127,
+      "learning_rate": 3.079944661297495e-05,
+      "loss": 5.6693,
+      "step": 7772
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1630382537841797,
+      "learning_rate": 3.078956470181333e-05,
+      "loss": 5.5721,
+      "step": 7776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.921631932258606,
+      "learning_rate": 3.0779682790651714e-05,
+      "loss": 5.7353,
+      "step": 7780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9565848112106323,
+      "learning_rate": 3.0769800879490096e-05,
+      "loss": 5.6735,
+      "step": 7784
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.029181718826294,
+      "learning_rate": 3.075991896832848e-05,
+      "loss": 5.5432,
+      "step": 7788
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7901182174682617,
+      "learning_rate": 3.075003705716686e-05,
+      "loss": 5.6048,
+      "step": 7792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7717857360839844,
+      "learning_rate": 3.074015514600524e-05,
+      "loss": 5.5689,
+      "step": 7796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8414852619171143,
+      "learning_rate": 3.073027323484362e-05,
+      "loss": 5.5628,
+      "step": 7800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.446169376373291,
+      "learning_rate": 3.0720391323682e-05,
+      "loss": 5.6246,
+      "step": 7804
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7307089567184448,
+      "learning_rate": 3.071050941252038e-05,
+      "loss": 5.6322,
+      "step": 7808
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0720410346984863,
+      "learning_rate": 3.0700627501358764e-05,
+      "loss": 5.6957,
+      "step": 7812
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8278499841690063,
+      "learning_rate": 3.0690745590197146e-05,
+      "loss": 5.6488,
+      "step": 7816
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9019477367401123,
+      "learning_rate": 3.068086367903553e-05,
+      "loss": 5.671,
+      "step": 7820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.330875873565674,
+      "learning_rate": 3.067098176787391e-05,
+      "loss": 5.7233,
+      "step": 7824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.958566427230835,
+      "learning_rate": 3.066109985671229e-05,
+      "loss": 5.5485,
+      "step": 7828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9183526039123535,
+      "learning_rate": 3.065121794555067e-05,
+      "loss": 5.627,
+      "step": 7832
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0147671699523926,
+      "learning_rate": 3.064133603438905e-05,
+      "loss": 5.5493,
+      "step": 7836
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8255884647369385,
+      "learning_rate": 3.063145412322743e-05,
+      "loss": 5.618,
+      "step": 7840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.188514471054077,
+      "learning_rate": 3.062157221206581e-05,
+      "loss": 5.6623,
+      "step": 7844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7055175304412842,
+      "learning_rate": 3.0611690300904195e-05,
+      "loss": 5.5607,
+      "step": 7848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6150466203689575,
+      "learning_rate": 3.060180838974258e-05,
+      "loss": 5.5892,
+      "step": 7852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7597399950027466,
+      "learning_rate": 3.059192647858096e-05,
+      "loss": 5.6028,
+      "step": 7856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0006275177001953,
+      "learning_rate": 3.0582044567419335e-05,
+      "loss": 5.5031,
+      "step": 7860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8812612295150757,
+      "learning_rate": 3.057216265625772e-05,
+      "loss": 5.5484,
+      "step": 7864
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7753983736038208,
+      "learning_rate": 3.0562280745096106e-05,
+      "loss": 5.661,
+      "step": 7868
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.891542673110962,
+      "learning_rate": 3.055239883393449e-05,
+      "loss": 5.5273,
+      "step": 7872
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.026078939437866,
+      "learning_rate": 3.054251692277287e-05,
+      "loss": 5.5718,
+      "step": 7876
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0182738304138184,
+      "learning_rate": 3.053263501161125e-05,
+      "loss": 5.694,
+      "step": 7880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9878616333007812,
+      "learning_rate": 3.052275310044963e-05,
+      "loss": 5.6561,
+      "step": 7884
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9343925714492798,
+      "learning_rate": 3.0512871189288012e-05,
+      "loss": 5.6317,
+      "step": 7888
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8456355333328247,
+      "learning_rate": 3.050298927812639e-05,
+      "loss": 5.5619,
+      "step": 7892
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6567593812942505,
+      "learning_rate": 3.0493107366964773e-05,
+      "loss": 5.5108,
+      "step": 7896
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7682963609695435,
+      "learning_rate": 3.0483225455803155e-05,
+      "loss": 5.8091,
+      "step": 7900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.653235912322998,
+      "learning_rate": 3.0473343544641537e-05,
+      "loss": 5.5357,
+      "step": 7904
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7630318403244019,
+      "learning_rate": 3.0463461633479916e-05,
+      "loss": 5.6776,
+      "step": 7908
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0381624698638916,
+      "learning_rate": 3.0453579722318298e-05,
+      "loss": 5.6855,
+      "step": 7912
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6931108236312866,
+      "learning_rate": 3.044369781115668e-05,
+      "loss": 5.6519,
+      "step": 7916
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7193225622177124,
+      "learning_rate": 3.0433815899995062e-05,
+      "loss": 5.5361,
+      "step": 7920
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7203538417816162,
+      "learning_rate": 3.042393398883344e-05,
+      "loss": 5.5749,
+      "step": 7924
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0116119384765625,
+      "learning_rate": 3.0414052077671823e-05,
+      "loss": 5.478,
+      "step": 7928
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5771147012710571,
+      "learning_rate": 3.0404170166510205e-05,
+      "loss": 5.6352,
+      "step": 7932
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8036284446716309,
+      "learning_rate": 3.0394288255348584e-05,
+      "loss": 5.5748,
+      "step": 7936
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.650652289390564,
+      "learning_rate": 3.0384406344186966e-05,
+      "loss": 5.6925,
+      "step": 7940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.871989369392395,
+      "learning_rate": 3.0374524433025348e-05,
+      "loss": 5.4965,
+      "step": 7944
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7644288539886475,
+      "learning_rate": 3.036464252186373e-05,
+      "loss": 5.521,
+      "step": 7948
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7584574222564697,
+      "learning_rate": 3.035476061070211e-05,
+      "loss": 5.4997,
+      "step": 7952
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6933388710021973,
+      "learning_rate": 3.034487869954049e-05,
+      "loss": 5.6415,
+      "step": 7956
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6390060186386108,
+      "learning_rate": 3.0334996788378873e-05,
+      "loss": 5.6538,
+      "step": 7960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8131755590438843,
+      "learning_rate": 3.0325114877217255e-05,
+      "loss": 5.6138,
+      "step": 7964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0270628929138184,
+      "learning_rate": 3.0315232966055633e-05,
+      "loss": 5.5461,
+      "step": 7968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8326473236083984,
+      "learning_rate": 3.0305351054894015e-05,
+      "loss": 5.5499,
+      "step": 7972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9747962951660156,
+      "learning_rate": 3.0295469143732397e-05,
+      "loss": 5.5334,
+      "step": 7976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.95182204246521,
+      "learning_rate": 3.028558723257078e-05,
+      "loss": 5.5317,
+      "step": 7980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.177966356277466,
+      "learning_rate": 3.0275705321409165e-05,
+      "loss": 5.5488,
+      "step": 7984
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0180881023406982,
+      "learning_rate": 3.0265823410247547e-05,
+      "loss": 5.5304,
+      "step": 7988
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7427732944488525,
+      "learning_rate": 3.0255941499085926e-05,
+      "loss": 5.6641,
+      "step": 7992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7346999645233154,
+      "learning_rate": 3.0246059587924308e-05,
+      "loss": 5.6368,
+      "step": 7996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.121185302734375,
+      "learning_rate": 3.023617767676269e-05,
+      "loss": 5.5946,
+      "step": 8000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9402865171432495,
+      "learning_rate": 3.0226295765601072e-05,
+      "loss": 5.679,
+      "step": 8004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0164012908935547,
+      "learning_rate": 3.021641385443945e-05,
+      "loss": 5.5902,
+      "step": 8008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8733190298080444,
+      "learning_rate": 3.0209002421068237e-05,
+      "loss": 5.532,
+      "step": 8012
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7538553476333618,
+      "learning_rate": 3.019912050990662e-05,
+      "loss": 5.6426,
+      "step": 8016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.886910319328308,
+      "learning_rate": 3.0189238598745e-05,
+      "loss": 5.5693,
+      "step": 8020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9799367189407349,
+      "learning_rate": 3.0179356687583383e-05,
+      "loss": 5.7278,
+      "step": 8024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9877070188522339,
+      "learning_rate": 3.0169474776421762e-05,
+      "loss": 5.5553,
+      "step": 8028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2819747924804688,
+      "learning_rate": 3.0159592865260144e-05,
+      "loss": 5.5856,
+      "step": 8032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.671416997909546,
+      "learning_rate": 3.0149710954098526e-05,
+      "loss": 5.588,
+      "step": 8036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.724274754524231,
+      "learning_rate": 3.0139829042936908e-05,
+      "loss": 5.5928,
+      "step": 8040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6406031847000122,
+      "learning_rate": 3.0129947131775287e-05,
+      "loss": 5.6456,
+      "step": 8044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9587162733078003,
+      "learning_rate": 3.012006522061367e-05,
+      "loss": 5.6093,
+      "step": 8048
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.987953543663025,
+      "learning_rate": 3.011018330945205e-05,
+      "loss": 5.8002,
+      "step": 8052
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.800836443901062,
+      "learning_rate": 3.0100301398290433e-05,
+      "loss": 5.7093,
+      "step": 8056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9985570907592773,
+      "learning_rate": 3.009041948712881e-05,
+      "loss": 5.6896,
+      "step": 8060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0569660663604736,
+      "learning_rate": 3.0080537575967194e-05,
+      "loss": 5.7986,
+      "step": 8064
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8365398645401,
+      "learning_rate": 3.0070655664805576e-05,
+      "loss": 5.4791,
+      "step": 8068
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.890994668006897,
+      "learning_rate": 3.0060773753643954e-05,
+      "loss": 5.5919,
+      "step": 8072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8914673328399658,
+      "learning_rate": 3.0050891842482337e-05,
+      "loss": 5.6298,
+      "step": 8076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.059929132461548,
+      "learning_rate": 3.004100993132072e-05,
+      "loss": 5.589,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.552819013595581,
+      "learning_rate": 3.00311280201591e-05,
+      "loss": 5.6104,
+      "step": 8084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.081338882446289,
+      "learning_rate": 3.002124610899748e-05,
+      "loss": 5.6061,
+      "step": 8088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.625524640083313,
+      "learning_rate": 3.001136419783586e-05,
+      "loss": 5.6501,
+      "step": 8092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.818798303604126,
+      "learning_rate": 3.0001482286674243e-05,
+      "loss": 5.6345,
+      "step": 8096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9189039468765259,
+      "learning_rate": 2.9991600375512626e-05,
+      "loss": 5.6594,
+      "step": 8100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9323093891143799,
+      "learning_rate": 2.9981718464351004e-05,
+      "loss": 5.6704,
+      "step": 8104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9228005409240723,
+      "learning_rate": 2.9971836553189386e-05,
+      "loss": 5.7182,
+      "step": 8108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8709877729415894,
+      "learning_rate": 2.9961954642027768e-05,
+      "loss": 5.6675,
+      "step": 8112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.015131711959839,
+      "learning_rate": 2.9952072730866147e-05,
+      "loss": 5.5459,
+      "step": 8116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3515000343322754,
+      "learning_rate": 2.994219081970453e-05,
+      "loss": 5.6607,
+      "step": 8120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6969714164733887,
+      "learning_rate": 2.993230890854291e-05,
+      "loss": 5.5014,
+      "step": 8124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8755816221237183,
+      "learning_rate": 2.9922426997381297e-05,
+      "loss": 5.6539,
+      "step": 8128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.724385380744934,
+      "learning_rate": 2.991254508621968e-05,
+      "loss": 5.6172,
+      "step": 8132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9022701978683472,
+      "learning_rate": 2.990266317505806e-05,
+      "loss": 5.6263,
+      "step": 8136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7855650186538696,
+      "learning_rate": 2.9892781263896443e-05,
+      "loss": 5.558,
+      "step": 8140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6109340190887451,
+      "learning_rate": 2.988289935273482e-05,
+      "loss": 5.5173,
+      "step": 8144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7251406908035278,
+      "learning_rate": 2.9873017441573203e-05,
+      "loss": 5.5703,
+      "step": 8148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8107340335845947,
+      "learning_rate": 2.9863135530411586e-05,
+      "loss": 5.6168,
+      "step": 8152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5449600219726562,
+      "learning_rate": 2.9853253619249964e-05,
+      "loss": 5.4385,
+      "step": 8156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.950967788696289,
+      "learning_rate": 2.9843371708088346e-05,
+      "loss": 5.6621,
+      "step": 8160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0183985233306885,
+      "learning_rate": 2.983348979692673e-05,
+      "loss": 5.5622,
+      "step": 8164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.118438243865967,
+      "learning_rate": 2.982360788576511e-05,
+      "loss": 5.7116,
+      "step": 8168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.065899610519409,
+      "learning_rate": 2.981372597460349e-05,
+      "loss": 5.7335,
+      "step": 8172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.075225830078125,
+      "learning_rate": 2.980384406344187e-05,
+      "loss": 5.4978,
+      "step": 8176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7782163619995117,
+      "learning_rate": 2.9793962152280253e-05,
+      "loss": 5.6495,
+      "step": 8180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8186616897583008,
+      "learning_rate": 2.9784080241118635e-05,
+      "loss": 5.5176,
+      "step": 8184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.926069736480713,
+      "learning_rate": 2.9774198329957014e-05,
+      "loss": 5.7383,
+      "step": 8188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8002229928970337,
+      "learning_rate": 2.9764316418795396e-05,
+      "loss": 5.6145,
+      "step": 8192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5681378841400146,
+      "learning_rate": 2.9754434507633778e-05,
+      "loss": 5.634,
+      "step": 8196
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.045126438140869,
+      "learning_rate": 2.9744552596472157e-05,
+      "loss": 5.5229,
+      "step": 8200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.116232395172119,
+      "learning_rate": 2.973467068531054e-05,
+      "loss": 5.6309,
+      "step": 8204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5766971111297607,
+      "learning_rate": 2.972478877414892e-05,
+      "loss": 5.5052,
+      "step": 8208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6305807828903198,
+      "learning_rate": 2.9714906862987303e-05,
+      "loss": 5.5024,
+      "step": 8212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7745639085769653,
+      "learning_rate": 2.970502495182568e-05,
+      "loss": 5.6619,
+      "step": 8216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.708950400352478,
+      "learning_rate": 2.9695143040664064e-05,
+      "loss": 5.634,
+      "step": 8220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8508306741714478,
+      "learning_rate": 2.9685261129502446e-05,
+      "loss": 5.7091,
+      "step": 8224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9124394655227661,
+      "learning_rate": 2.9675379218340828e-05,
+      "loss": 5.6441,
+      "step": 8228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.735961675643921,
+      "learning_rate": 2.9665497307179206e-05,
+      "loss": 5.5941,
+      "step": 8232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0475118160247803,
+      "learning_rate": 2.965561539601759e-05,
+      "loss": 5.5674,
+      "step": 8236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8113858699798584,
+      "learning_rate": 2.964573348485597e-05,
+      "loss": 5.6697,
+      "step": 8240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9009729623794556,
+      "learning_rate": 2.9635851573694356e-05,
+      "loss": 5.5668,
+      "step": 8244
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8220778703689575,
+      "learning_rate": 2.9625969662532738e-05,
+      "loss": 5.6735,
+      "step": 8248
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7952344417572021,
+      "learning_rate": 2.961608775137112e-05,
+      "loss": 5.5436,
+      "step": 8252
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8756263256072998,
+      "learning_rate": 2.96062058402095e-05,
+      "loss": 5.5825,
+      "step": 8256
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8858979940414429,
+      "learning_rate": 2.959632392904788e-05,
+      "loss": 5.6039,
+      "step": 8260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.780319333076477,
+      "learning_rate": 2.9586442017886263e-05,
+      "loss": 5.6614,
+      "step": 8264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.561244487762451,
+      "learning_rate": 2.9576560106724645e-05,
+      "loss": 5.6054,
+      "step": 8268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9080662727355957,
+      "learning_rate": 2.9566678195563024e-05,
+      "loss": 5.5438,
+      "step": 8272
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8043930530548096,
+      "learning_rate": 2.9556796284401406e-05,
+      "loss": 5.5615,
+      "step": 8276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8665918111801147,
+      "learning_rate": 2.9546914373239788e-05,
+      "loss": 5.6224,
+      "step": 8280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7628755569458008,
+      "learning_rate": 2.9537032462078166e-05,
+      "loss": 5.648,
+      "step": 8284
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5806515216827393,
+      "learning_rate": 2.952715055091655e-05,
+      "loss": 5.6113,
+      "step": 8288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7797776460647583,
+      "learning_rate": 2.951726863975493e-05,
+      "loss": 5.6718,
+      "step": 8292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.85147225856781,
+      "learning_rate": 2.9507386728593313e-05,
+      "loss": 5.571,
+      "step": 8296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8725066184997559,
+      "learning_rate": 2.949750481743169e-05,
+      "loss": 5.6677,
+      "step": 8300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6381113529205322,
+      "learning_rate": 2.9487622906270073e-05,
+      "loss": 5.5314,
+      "step": 8304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7837311029434204,
+      "learning_rate": 2.9477740995108455e-05,
+      "loss": 5.6066,
+      "step": 8308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7460732460021973,
+      "learning_rate": 2.9467859083946837e-05,
+      "loss": 5.5304,
+      "step": 8312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7592207193374634,
+      "learning_rate": 2.9457977172785216e-05,
+      "loss": 5.5576,
+      "step": 8316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7295989990234375,
+      "learning_rate": 2.9448095261623598e-05,
+      "loss": 5.6019,
+      "step": 8320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4635212421417236,
+      "learning_rate": 2.943821335046198e-05,
+      "loss": 5.6384,
+      "step": 8324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.829713225364685,
+      "learning_rate": 2.9428331439300362e-05,
+      "loss": 5.6704,
+      "step": 8328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.121614694595337,
+      "learning_rate": 2.941844952813874e-05,
+      "loss": 5.6339,
+      "step": 8332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.783610224723816,
+      "learning_rate": 2.9408567616977123e-05,
+      "loss": 5.6805,
+      "step": 8336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0255215167999268,
+      "learning_rate": 2.9398685705815505e-05,
+      "loss": 5.667,
+      "step": 8340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8871128559112549,
+      "learning_rate": 2.9388803794653884e-05,
+      "loss": 5.6883,
+      "step": 8344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9648714065551758,
+      "learning_rate": 2.9378921883492266e-05,
+      "loss": 5.6178,
+      "step": 8348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8247488737106323,
+      "learning_rate": 2.9369039972330648e-05,
+      "loss": 5.675,
+      "step": 8352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8563957214355469,
+      "learning_rate": 2.9359158061169033e-05,
+      "loss": 5.6308,
+      "step": 8356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8070034980773926,
+      "learning_rate": 2.9349276150007415e-05,
+      "loss": 5.6328,
+      "step": 8360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8198412656784058,
+      "learning_rate": 2.9339394238845797e-05,
+      "loss": 5.6216,
+      "step": 8364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7104226350784302,
+      "learning_rate": 2.9329512327684176e-05,
+      "loss": 5.6402,
+      "step": 8368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6858009099960327,
+      "learning_rate": 2.9319630416522558e-05,
+      "loss": 5.6666,
+      "step": 8372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.978938102722168,
+      "learning_rate": 2.930974850536094e-05,
+      "loss": 5.4985,
+      "step": 8376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6124422550201416,
+      "learning_rate": 2.9299866594199322e-05,
+      "loss": 5.478,
+      "step": 8380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.06709623336792,
+      "learning_rate": 2.92899846830377e-05,
+      "loss": 5.5993,
+      "step": 8384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.138789415359497,
+      "learning_rate": 2.9280102771876083e-05,
+      "loss": 5.6381,
+      "step": 8388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7755143642425537,
+      "learning_rate": 2.9270220860714465e-05,
+      "loss": 5.6309,
+      "step": 8392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0836946964263916,
+      "learning_rate": 2.9260338949552847e-05,
+      "loss": 5.5439,
+      "step": 8396
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8185824155807495,
+      "learning_rate": 2.9250457038391226e-05,
+      "loss": 5.5819,
+      "step": 8400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9818052053451538,
+      "learning_rate": 2.9240575127229608e-05,
+      "loss": 5.6349,
+      "step": 8404
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.246522903442383,
+      "learning_rate": 2.923069321606799e-05,
+      "loss": 5.6007,
+      "step": 8408
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9953023195266724,
+      "learning_rate": 2.922081130490637e-05,
+      "loss": 5.643,
+      "step": 8412
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9882394075393677,
+      "learning_rate": 2.921092939374475e-05,
+      "loss": 5.6082,
+      "step": 8416
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7919508218765259,
+      "learning_rate": 2.9201047482583133e-05,
+      "loss": 5.6575,
+      "step": 8420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6816109418869019,
+      "learning_rate": 2.9191165571421515e-05,
+      "loss": 5.6128,
+      "step": 8424
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9026116132736206,
+      "learning_rate": 2.9181283660259893e-05,
+      "loss": 5.5993,
+      "step": 8428
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1336424350738525,
+      "learning_rate": 2.9171401749098276e-05,
+      "loss": 5.5573,
+      "step": 8432
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6572264432907104,
+      "learning_rate": 2.9161519837936658e-05,
+      "loss": 5.7065,
+      "step": 8436
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1240315437316895,
+      "learning_rate": 2.915163792677504e-05,
+      "loss": 5.5141,
+      "step": 8440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.728661298751831,
+      "learning_rate": 2.914175601561342e-05,
+      "loss": 5.621,
+      "step": 8444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.108468770980835,
+      "learning_rate": 2.91318741044518e-05,
+      "loss": 5.5439,
+      "step": 8448
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.912477731704712,
+      "learning_rate": 2.9121992193290182e-05,
+      "loss": 5.6222,
+      "step": 8452
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.753421425819397,
+      "learning_rate": 2.9112110282128565e-05,
+      "loss": 5.6209,
+      "step": 8456
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7683593034744263,
+      "learning_rate": 2.9102228370966943e-05,
+      "loss": 5.6178,
+      "step": 8460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8550388813018799,
+      "learning_rate": 2.9092346459805325e-05,
+      "loss": 5.6337,
+      "step": 8464
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6444224119186401,
+      "learning_rate": 2.9082464548643707e-05,
+      "loss": 5.7698,
+      "step": 8468
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.868353247642517,
+      "learning_rate": 2.9072582637482093e-05,
+      "loss": 5.8062,
+      "step": 8472
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8483880758285522,
+      "learning_rate": 2.9062700726320475e-05,
+      "loss": 5.5449,
+      "step": 8476
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.059861660003662,
+      "learning_rate": 2.9052818815158857e-05,
+      "loss": 5.5988,
+      "step": 8480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7102926969528198,
+      "learning_rate": 2.9042936903997236e-05,
+      "loss": 5.5771,
+      "step": 8484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8607144355773926,
+      "learning_rate": 2.9033054992835618e-05,
+      "loss": 5.4982,
+      "step": 8488
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9258161783218384,
+      "learning_rate": 2.9023173081674e-05,
+      "loss": 5.6881,
+      "step": 8492
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.688467264175415,
+      "learning_rate": 2.901329117051238e-05,
+      "loss": 5.6507,
+      "step": 8496
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7011233568191528,
+      "learning_rate": 2.900340925935076e-05,
+      "loss": 5.59,
+      "step": 8500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6969784498214722,
+      "learning_rate": 2.8993527348189142e-05,
+      "loss": 5.6226,
+      "step": 8504
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.669765830039978,
+      "learning_rate": 2.8983645437027525e-05,
+      "loss": 5.5824,
+      "step": 8508
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9257421493530273,
+      "learning_rate": 2.8973763525865903e-05,
+      "loss": 5.5598,
+      "step": 8512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1687793731689453,
+      "learning_rate": 2.8963881614704285e-05,
+      "loss": 5.7037,
+      "step": 8516
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5989969968795776,
+      "learning_rate": 2.8953999703542667e-05,
+      "loss": 5.5294,
+      "step": 8520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8674973249435425,
+      "learning_rate": 2.894411779238105e-05,
+      "loss": 5.4714,
+      "step": 8524
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.83799409866333,
+      "learning_rate": 2.8934235881219428e-05,
+      "loss": 5.6442,
+      "step": 8528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0756680965423584,
+      "learning_rate": 2.892435397005781e-05,
+      "loss": 5.6252,
+      "step": 8532
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1030049324035645,
+      "learning_rate": 2.8914472058896192e-05,
+      "loss": 5.5707,
+      "step": 8536
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.740062952041626,
+      "learning_rate": 2.8904590147734574e-05,
+      "loss": 5.6083,
+      "step": 8540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9821360111236572,
+      "learning_rate": 2.8894708236572953e-05,
+      "loss": 5.5575,
+      "step": 8544
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0628159046173096,
+      "learning_rate": 2.8884826325411335e-05,
+      "loss": 5.5519,
+      "step": 8548
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3252205848693848,
+      "learning_rate": 2.8874944414249717e-05,
+      "loss": 5.6204,
+      "step": 8552
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7173527479171753,
+      "learning_rate": 2.8865062503088096e-05,
+      "loss": 5.5685,
+      "step": 8556
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7327337265014648,
+      "learning_rate": 2.8855180591926478e-05,
+      "loss": 5.6208,
+      "step": 8560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7945717573165894,
+      "learning_rate": 2.884529868076486e-05,
+      "loss": 5.6034,
+      "step": 8564
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8818703889846802,
+      "learning_rate": 2.8835416769603242e-05,
+      "loss": 5.5966,
+      "step": 8568
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.780110478401184,
+      "learning_rate": 2.882553485844162e-05,
+      "loss": 5.6438,
+      "step": 8572
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9500248432159424,
+      "learning_rate": 2.8815652947280003e-05,
+      "loss": 5.6569,
+      "step": 8576
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0697216987609863,
+      "learning_rate": 2.8805771036118385e-05,
+      "loss": 5.5631,
+      "step": 8580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8026626110076904,
+      "learning_rate": 2.8795889124956767e-05,
+      "loss": 5.6308,
+      "step": 8584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6089578866958618,
+      "learning_rate": 2.8786007213795152e-05,
+      "loss": 5.5297,
+      "step": 8588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8671869039535522,
+      "learning_rate": 2.8776125302633534e-05,
+      "loss": 5.5388,
+      "step": 8592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7763539552688599,
+      "learning_rate": 2.8766243391471913e-05,
+      "loss": 5.6698,
+      "step": 8596
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6824060678482056,
+      "learning_rate": 2.8756361480310295e-05,
+      "loss": 5.6613,
+      "step": 8600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9802032709121704,
+      "learning_rate": 2.8746479569148677e-05,
+      "loss": 5.6713,
+      "step": 8604
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8129801750183105,
+      "learning_rate": 2.873659765798706e-05,
+      "loss": 5.5339,
+      "step": 8608
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0134878158569336,
+      "learning_rate": 2.8726715746825438e-05,
+      "loss": 5.5781,
+      "step": 8612
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7483782768249512,
+      "learning_rate": 2.871683383566382e-05,
+      "loss": 5.5785,
+      "step": 8616
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1585028171539307,
+      "learning_rate": 2.8706951924502202e-05,
+      "loss": 5.5063,
+      "step": 8620
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0951032638549805,
+      "learning_rate": 2.8697070013340584e-05,
+      "loss": 5.5559,
+      "step": 8624
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7235571146011353,
+      "learning_rate": 2.8687188102178963e-05,
+      "loss": 5.6181,
+      "step": 8628
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0943381786346436,
+      "learning_rate": 2.8677306191017345e-05,
+      "loss": 5.5952,
+      "step": 8632
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9712262153625488,
+      "learning_rate": 2.8667424279855727e-05,
+      "loss": 5.6672,
+      "step": 8636
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.837053894996643,
+      "learning_rate": 2.8657542368694105e-05,
+      "loss": 5.6023,
+      "step": 8640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9455965757369995,
+      "learning_rate": 2.8647660457532487e-05,
+      "loss": 5.6813,
+      "step": 8644
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9085052013397217,
+      "learning_rate": 2.863777854637087e-05,
+      "loss": 5.5879,
+      "step": 8648
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.043121814727783,
+      "learning_rate": 2.862789663520925e-05,
+      "loss": 5.484,
+      "step": 8652
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.043996572494507,
+      "learning_rate": 2.861801472404763e-05,
+      "loss": 5.5774,
+      "step": 8656
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.140770673751831,
+      "learning_rate": 2.8608132812886012e-05,
+      "loss": 5.5304,
+      "step": 8660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.25034499168396,
+      "learning_rate": 2.8598250901724394e-05,
+      "loss": 5.4963,
+      "step": 8664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.623706340789795,
+      "learning_rate": 2.8588368990562776e-05,
+      "loss": 5.5669,
+      "step": 8668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8031169176101685,
+      "learning_rate": 2.8578487079401155e-05,
+      "loss": 5.5762,
+      "step": 8672
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0662543773651123,
+      "learning_rate": 2.8568605168239537e-05,
+      "loss": 5.4806,
+      "step": 8676
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1524598598480225,
+      "learning_rate": 2.855872325707792e-05,
+      "loss": 5.6503,
+      "step": 8680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9902993440628052,
+      "learning_rate": 2.8548841345916298e-05,
+      "loss": 5.4978,
+      "step": 8684
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8020424842834473,
+      "learning_rate": 2.853895943475468e-05,
+      "loss": 5.6205,
+      "step": 8688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.732035517692566,
+      "learning_rate": 2.8529077523593062e-05,
+      "loss": 5.4912,
+      "step": 8692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8738574981689453,
+      "learning_rate": 2.8519195612431444e-05,
+      "loss": 5.5621,
+      "step": 8696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9714980125427246,
+      "learning_rate": 2.850931370126983e-05,
+      "loss": 5.6087,
+      "step": 8700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0756115913391113,
+      "learning_rate": 2.849943179010821e-05,
+      "loss": 5.65,
+      "step": 8704
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.104881763458252,
+      "learning_rate": 2.8489549878946594e-05,
+      "loss": 5.6053,
+      "step": 8708
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7442312240600586,
+      "learning_rate": 2.8479667967784972e-05,
+      "loss": 5.6199,
+      "step": 8712
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0308890342712402,
+      "learning_rate": 2.8469786056623354e-05,
+      "loss": 5.7298,
+      "step": 8716
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7256563901901245,
+      "learning_rate": 2.8459904145461736e-05,
+      "loss": 5.4541,
+      "step": 8720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.112795352935791,
+      "learning_rate": 2.8450022234300115e-05,
+      "loss": 5.5393,
+      "step": 8724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.886513113975525,
+      "learning_rate": 2.8440140323138497e-05,
+      "loss": 5.6014,
+      "step": 8728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9806932210922241,
+      "learning_rate": 2.843025841197688e-05,
+      "loss": 5.6828,
+      "step": 8732
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.23341965675354,
+      "learning_rate": 2.842037650081526e-05,
+      "loss": 5.7338,
+      "step": 8736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7140839099884033,
+      "learning_rate": 2.841049458965364e-05,
+      "loss": 5.6041,
+      "step": 8740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7378863096237183,
+      "learning_rate": 2.8400612678492022e-05,
+      "loss": 5.5925,
+      "step": 8744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7445762157440186,
+      "learning_rate": 2.8390730767330404e-05,
+      "loss": 5.5118,
+      "step": 8748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9341727495193481,
+      "learning_rate": 2.8380848856168786e-05,
+      "loss": 5.6059,
+      "step": 8752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1164705753326416,
+      "learning_rate": 2.8370966945007165e-05,
+      "loss": 5.7005,
+      "step": 8756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7062251567840576,
+      "learning_rate": 2.8361085033845547e-05,
+      "loss": 5.5896,
+      "step": 8760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8553556203842163,
+      "learning_rate": 2.835120312268393e-05,
+      "loss": 5.6167,
+      "step": 8764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7515733242034912,
+      "learning_rate": 2.8341321211522308e-05,
+      "loss": 5.5941,
+      "step": 8768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8316551446914673,
+      "learning_rate": 2.833143930036069e-05,
+      "loss": 5.4873,
+      "step": 8772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9083267450332642,
+      "learning_rate": 2.8321557389199072e-05,
+      "loss": 5.6866,
+      "step": 8776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4909923076629639,
+      "learning_rate": 2.8311675478037454e-05,
+      "loss": 5.5345,
+      "step": 8780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7641175985336304,
+      "learning_rate": 2.8301793566875832e-05,
+      "loss": 5.6514,
+      "step": 8784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8059598207473755,
+      "learning_rate": 2.8291911655714215e-05,
+      "loss": 5.4931,
+      "step": 8788
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8962539434432983,
+      "learning_rate": 2.8282029744552597e-05,
+      "loss": 5.5765,
+      "step": 8792
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.188370943069458,
+      "learning_rate": 2.827214783339098e-05,
+      "loss": 5.6586,
+      "step": 8796
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9519965648651123,
+      "learning_rate": 2.8262265922229357e-05,
+      "loss": 5.6691,
+      "step": 8800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9207426309585571,
+      "learning_rate": 2.825238401106774e-05,
+      "loss": 5.7145,
+      "step": 8804
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8392800092697144,
+      "learning_rate": 2.824250209990612e-05,
+      "loss": 5.5793,
+      "step": 8808
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.228625774383545,
+      "learning_rate": 2.8232620188744503e-05,
+      "loss": 5.6521,
+      "step": 8812
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.18316650390625,
+      "learning_rate": 2.822273827758289e-05,
+      "loss": 5.7501,
+      "step": 8816
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.011134147644043,
+      "learning_rate": 2.821285636642127e-05,
+      "loss": 5.6124,
+      "step": 8820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9527933597564697,
+      "learning_rate": 2.820297445525965e-05,
+      "loss": 5.4269,
+      "step": 8824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6352803707122803,
+      "learning_rate": 2.8193092544098032e-05,
+      "loss": 5.7361,
+      "step": 8828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.889510154724121,
+      "learning_rate": 2.8183210632936414e-05,
+      "loss": 5.5663,
+      "step": 8832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8308645486831665,
+      "learning_rate": 2.8173328721774796e-05,
+      "loss": 5.6488,
+      "step": 8836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8273881673812866,
+      "learning_rate": 2.8163446810613175e-05,
+      "loss": 5.6201,
+      "step": 8840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8046854734420776,
+      "learning_rate": 2.8153564899451557e-05,
+      "loss": 5.6918,
+      "step": 8844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0153470039367676,
+      "learning_rate": 2.814368298828994e-05,
+      "loss": 5.567,
+      "step": 8848
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6771526336669922,
+      "learning_rate": 2.8133801077128317e-05,
+      "loss": 5.5549,
+      "step": 8852
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.917395830154419,
+      "learning_rate": 2.81239191659667e-05,
+      "loss": 5.564,
+      "step": 8856
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7779862880706787,
+      "learning_rate": 2.811403725480508e-05,
+      "loss": 5.6795,
+      "step": 8860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9387528896331787,
+      "learning_rate": 2.8104155343643464e-05,
+      "loss": 5.5145,
+      "step": 8864
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0087499618530273,
+      "learning_rate": 2.8094273432481842e-05,
+      "loss": 5.548,
+      "step": 8868
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.05503249168396,
+      "learning_rate": 2.8084391521320224e-05,
+      "loss": 5.6076,
+      "step": 8872
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6153981685638428,
+      "learning_rate": 2.8074509610158606e-05,
+      "loss": 5.6318,
+      "step": 8876
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6110368967056274,
+      "learning_rate": 2.806462769899699e-05,
+      "loss": 5.596,
+      "step": 8880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7750228643417358,
+      "learning_rate": 2.8054745787835367e-05,
+      "loss": 5.528,
+      "step": 8884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0455288887023926,
+      "learning_rate": 2.804486387667375e-05,
+      "loss": 5.4797,
+      "step": 8888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8901207447052002,
+      "learning_rate": 2.803498196551213e-05,
+      "loss": 5.6263,
+      "step": 8892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.78607177734375,
+      "learning_rate": 2.8025100054350513e-05,
+      "loss": 5.6771,
+      "step": 8896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7866969108581543,
+      "learning_rate": 2.8015218143188892e-05,
+      "loss": 5.6625,
+      "step": 8900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.713319182395935,
+      "learning_rate": 2.8005336232027274e-05,
+      "loss": 5.5529,
+      "step": 8904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7033851146697998,
+      "learning_rate": 2.7995454320865656e-05,
+      "loss": 5.6241,
+      "step": 8908
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.626865029335022,
+      "learning_rate": 2.7985572409704035e-05,
+      "loss": 5.623,
+      "step": 8912
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8610364198684692,
+      "learning_rate": 2.7975690498542417e-05,
+      "loss": 5.5305,
+      "step": 8916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.474336624145508,
+      "learning_rate": 2.79658085873808e-05,
+      "loss": 5.5765,
+      "step": 8920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0331220626831055,
+      "learning_rate": 2.795592667621918e-05,
+      "loss": 5.5884,
+      "step": 8924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7288097143173218,
+      "learning_rate": 2.794604476505756e-05,
+      "loss": 5.5913,
+      "step": 8928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5526161193847656,
+      "learning_rate": 2.793616285389595e-05,
+      "loss": 5.5849,
+      "step": 8932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.636829137802124,
+      "learning_rate": 2.7926280942734327e-05,
+      "loss": 5.5373,
+      "step": 8936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8936233520507812,
+      "learning_rate": 2.791639903157271e-05,
+      "loss": 5.7408,
+      "step": 8940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8193585872650146,
+      "learning_rate": 2.790651712041109e-05,
+      "loss": 5.6536,
+      "step": 8944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0485775470733643,
+      "learning_rate": 2.7896635209249473e-05,
+      "loss": 5.6758,
+      "step": 8948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8670713901519775,
+      "learning_rate": 2.7886753298087852e-05,
+      "loss": 5.5534,
+      "step": 8952
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8386154174804688,
+      "learning_rate": 2.7876871386926234e-05,
+      "loss": 5.6744,
+      "step": 8956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.060434103012085,
+      "learning_rate": 2.7866989475764616e-05,
+      "loss": 5.6095,
+      "step": 8960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5825614929199219,
+      "learning_rate": 2.7857107564602998e-05,
+      "loss": 5.5813,
+      "step": 8964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6022027730941772,
+      "learning_rate": 2.7847225653441377e-05,
+      "loss": 5.5509,
+      "step": 8968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.892376184463501,
+      "learning_rate": 2.783734374227976e-05,
+      "loss": 5.5502,
+      "step": 8972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8178150653839111,
+      "learning_rate": 2.782746183111814e-05,
+      "loss": 5.5913,
+      "step": 8976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.899789571762085,
+      "learning_rate": 2.7817579919956523e-05,
+      "loss": 5.6232,
+      "step": 8980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5999733209609985,
+      "learning_rate": 2.78076980087949e-05,
+      "loss": 5.5976,
+      "step": 8984
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.885884404182434,
+      "learning_rate": 2.7797816097633284e-05,
+      "loss": 5.5364,
+      "step": 8988
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7063078880310059,
+      "learning_rate": 2.7787934186471666e-05,
+      "loss": 5.5669,
+      "step": 8992
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8398665189743042,
+      "learning_rate": 2.7778052275310044e-05,
+      "loss": 5.5482,
+      "step": 8996
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7803176641464233,
+      "learning_rate": 2.7768170364148426e-05,
+      "loss": 5.6153,
+      "step": 9000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9626309871673584,
+      "learning_rate": 2.775828845298681e-05,
+      "loss": 5.5867,
+      "step": 9004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0748071670532227,
+      "learning_rate": 2.774840654182519e-05,
+      "loss": 5.5583,
+      "step": 9008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8239414691925049,
+      "learning_rate": 2.773852463066357e-05,
+      "loss": 5.5866,
+      "step": 9012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6811037063598633,
+      "learning_rate": 2.772864271950195e-05,
+      "loss": 5.4177,
+      "step": 9016
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1083476543426514,
+      "learning_rate": 2.7718760808340333e-05,
+      "loss": 5.5789,
+      "step": 9020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.621522068977356,
+      "learning_rate": 2.7708878897178715e-05,
+      "loss": 5.5616,
+      "step": 9024
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0952842235565186,
+      "learning_rate": 2.7698996986017094e-05,
+      "loss": 5.5983,
+      "step": 9028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8348654508590698,
+      "learning_rate": 2.7689115074855476e-05,
+      "loss": 5.4559,
+      "step": 9032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1072843074798584,
+      "learning_rate": 2.7679233163693858e-05,
+      "loss": 5.6188,
+      "step": 9036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5936877727508545,
+      "learning_rate": 2.7669351252532237e-05,
+      "loss": 5.5372,
+      "step": 9040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8489831686019897,
+      "learning_rate": 2.7659469341370626e-05,
+      "loss": 5.6375,
+      "step": 9044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1420514583587646,
+      "learning_rate": 2.7649587430209008e-05,
+      "loss": 5.5546,
+      "step": 9048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9371378421783447,
+      "learning_rate": 2.7639705519047386e-05,
+      "loss": 5.4121,
+      "step": 9052
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1076085567474365,
+      "learning_rate": 2.762982360788577e-05,
+      "loss": 5.6349,
+      "step": 9056
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1002743244171143,
+      "learning_rate": 2.761994169672415e-05,
+      "loss": 5.7135,
+      "step": 9060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5791290998458862,
+      "learning_rate": 2.761005978556253e-05,
+      "loss": 5.5199,
+      "step": 9064
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9685577154159546,
+      "learning_rate": 2.760017787440091e-05,
+      "loss": 5.6298,
+      "step": 9068
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6860204935073853,
+      "learning_rate": 2.7590295963239293e-05,
+      "loss": 5.67,
+      "step": 9072
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2700107097625732,
+      "learning_rate": 2.7580414052077675e-05,
+      "loss": 5.661,
+      "step": 9076
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9949781894683838,
+      "learning_rate": 2.7570532140916054e-05,
+      "loss": 5.6058,
+      "step": 9080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8357362747192383,
+      "learning_rate": 2.7560650229754436e-05,
+      "loss": 5.6563,
+      "step": 9084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.157716751098633,
+      "learning_rate": 2.7550768318592818e-05,
+      "loss": 5.5362,
+      "step": 9088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0846590995788574,
+      "learning_rate": 2.75408864074312e-05,
+      "loss": 5.4562,
+      "step": 9092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8942608833312988,
+      "learning_rate": 2.753100449626958e-05,
+      "loss": 5.6858,
+      "step": 9096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.882866621017456,
+      "learning_rate": 2.752112258510796e-05,
+      "loss": 5.5599,
+      "step": 9100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1065609455108643,
+      "learning_rate": 2.7511240673946343e-05,
+      "loss": 5.6257,
+      "step": 9104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8394254446029663,
+      "learning_rate": 2.7501358762784725e-05,
+      "loss": 5.548,
+      "step": 9108
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.759247899055481,
+      "learning_rate": 2.7491476851623104e-05,
+      "loss": 5.5753,
+      "step": 9112
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.773542881011963,
+      "learning_rate": 2.7481594940461486e-05,
+      "loss": 5.6327,
+      "step": 9116
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0030620098114014,
+      "learning_rate": 2.7471713029299868e-05,
+      "loss": 5.5694,
+      "step": 9120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7507848739624023,
+      "learning_rate": 2.7461831118138247e-05,
+      "loss": 5.6386,
+      "step": 9124
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7345106601715088,
+      "learning_rate": 2.745194920697663e-05,
+      "loss": 5.5398,
+      "step": 9128
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1354246139526367,
+      "learning_rate": 2.744206729581501e-05,
+      "loss": 5.5165,
+      "step": 9132
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6289561986923218,
+      "learning_rate": 2.7432185384653393e-05,
+      "loss": 5.6551,
+      "step": 9136
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.850968599319458,
+      "learning_rate": 2.742230347349177e-05,
+      "loss": 5.5557,
+      "step": 9140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6815087795257568,
+      "learning_rate": 2.7412421562330154e-05,
+      "loss": 5.5728,
+      "step": 9144
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.868223786354065,
+      "learning_rate": 2.7402539651168536e-05,
+      "loss": 5.5449,
+      "step": 9148
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6773638725280762,
+      "learning_rate": 2.7392657740006918e-05,
+      "loss": 5.588,
+      "step": 9152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8860238790512085,
+      "learning_rate": 2.7382775828845296e-05,
+      "loss": 5.537,
+      "step": 9156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8454649448394775,
+      "learning_rate": 2.7372893917683685e-05,
+      "loss": 5.5351,
+      "step": 9160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8175948858261108,
+      "learning_rate": 2.7363012006522064e-05,
+      "loss": 5.5808,
+      "step": 9164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8350155353546143,
+      "learning_rate": 2.7353130095360446e-05,
+      "loss": 5.5217,
+      "step": 9168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6785465478897095,
+      "learning_rate": 2.7343248184198828e-05,
+      "loss": 5.5363,
+      "step": 9172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.043757200241089,
+      "learning_rate": 2.733336627303721e-05,
+      "loss": 5.3924,
+      "step": 9176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7272677421569824,
+      "learning_rate": 2.732348436187559e-05,
+      "loss": 5.5548,
+      "step": 9180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7841123342514038,
+      "learning_rate": 2.731360245071397e-05,
+      "loss": 5.5787,
+      "step": 9184
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2284867763519287,
+      "learning_rate": 2.7303720539552353e-05,
+      "loss": 5.6589,
+      "step": 9188
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8492909669876099,
+      "learning_rate": 2.7293838628390735e-05,
+      "loss": 5.6947,
+      "step": 9192
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.870833396911621,
+      "learning_rate": 2.7283956717229114e-05,
+      "loss": 5.5846,
+      "step": 9196
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.007213592529297,
+      "learning_rate": 2.7274074806067496e-05,
+      "loss": 5.4946,
+      "step": 9200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.178415536880493,
+      "learning_rate": 2.7264192894905878e-05,
+      "loss": 5.6511,
+      "step": 9204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.777798056602478,
+      "learning_rate": 2.7254310983744256e-05,
+      "loss": 5.5872,
+      "step": 9208
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0451252460479736,
+      "learning_rate": 2.724442907258264e-05,
+      "loss": 5.5361,
+      "step": 9212
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.672156572341919,
+      "learning_rate": 2.723454716142102e-05,
+      "loss": 5.5812,
+      "step": 9216
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7018766403198242,
+      "learning_rate": 2.7224665250259402e-05,
+      "loss": 5.5854,
+      "step": 9220
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8152602910995483,
+      "learning_rate": 2.721478333909778e-05,
+      "loss": 5.5784,
+      "step": 9224
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.722531795501709,
+      "learning_rate": 2.7204901427936163e-05,
+      "loss": 5.5214,
+      "step": 9228
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9717679023742676,
+      "learning_rate": 2.7195019516774545e-05,
+      "loss": 5.5436,
+      "step": 9232
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6937377452850342,
+      "learning_rate": 2.7185137605612927e-05,
+      "loss": 5.5609,
+      "step": 9236
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6418997049331665,
+      "learning_rate": 2.7175255694451306e-05,
+      "loss": 5.5576,
+      "step": 9240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9024310111999512,
+      "learning_rate": 2.7165373783289688e-05,
+      "loss": 5.6251,
+      "step": 9244
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9837607145309448,
+      "learning_rate": 2.715549187212807e-05,
+      "loss": 5.6149,
+      "step": 9248
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5828512907028198,
+      "learning_rate": 2.714560996096645e-05,
+      "loss": 5.4367,
+      "step": 9252
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0914130210876465,
+      "learning_rate": 2.713572804980483e-05,
+      "loss": 5.632,
+      "step": 9256
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2636313438415527,
+      "learning_rate": 2.7125846138643213e-05,
+      "loss": 5.6487,
+      "step": 9260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0330276489257812,
+      "learning_rate": 2.7115964227481595e-05,
+      "loss": 5.5618,
+      "step": 9264
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7501007318496704,
+      "learning_rate": 2.7106082316319974e-05,
+      "loss": 5.4962,
+      "step": 9268
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0401079654693604,
+      "learning_rate": 2.7096200405158356e-05,
+      "loss": 5.5273,
+      "step": 9272
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.885672688484192,
+      "learning_rate": 2.7086318493996745e-05,
+      "loss": 5.5909,
+      "step": 9276
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8399213552474976,
+      "learning_rate": 2.7076436582835123e-05,
+      "loss": 5.5688,
+      "step": 9280
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6890549659729004,
+      "learning_rate": 2.7066554671673505e-05,
+      "loss": 5.5704,
+      "step": 9284
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0120296478271484,
+      "learning_rate": 2.7056672760511887e-05,
+      "loss": 5.6876,
+      "step": 9288
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8612735271453857,
+      "learning_rate": 2.7046790849350266e-05,
+      "loss": 5.5703,
+      "step": 9292
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8374227285385132,
+      "learning_rate": 2.7036908938188648e-05,
+      "loss": 5.4759,
+      "step": 9296
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8448069095611572,
+      "learning_rate": 2.702702702702703e-05,
+      "loss": 5.5785,
+      "step": 9300
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6031235456466675,
+      "learning_rate": 2.7017145115865412e-05,
+      "loss": 5.5566,
+      "step": 9304
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8467074632644653,
+      "learning_rate": 2.700726320470379e-05,
+      "loss": 5.6011,
+      "step": 9308
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.152393102645874,
+      "learning_rate": 2.6997381293542173e-05,
+      "loss": 5.4838,
+      "step": 9312
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6905425786972046,
+      "learning_rate": 2.6987499382380555e-05,
+      "loss": 5.6356,
+      "step": 9316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.052292823791504,
+      "learning_rate": 2.6977617471218937e-05,
+      "loss": 5.5832,
+      "step": 9320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9417749643325806,
+      "learning_rate": 2.6967735560057316e-05,
+      "loss": 5.6308,
+      "step": 9324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7844290733337402,
+      "learning_rate": 2.6957853648895698e-05,
+      "loss": 5.5116,
+      "step": 9328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.945826768875122,
+      "learning_rate": 2.694797173773408e-05,
+      "loss": 5.5016,
+      "step": 9332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.079559087753296,
+      "learning_rate": 2.693808982657246e-05,
+      "loss": 5.631,
+      "step": 9336
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7609906196594238,
+      "learning_rate": 2.692820791541084e-05,
+      "loss": 5.6115,
+      "step": 9340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.221853733062744,
+      "learning_rate": 2.6918326004249223e-05,
+      "loss": 5.6196,
+      "step": 9344
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8510643243789673,
+      "learning_rate": 2.6908444093087605e-05,
+      "loss": 5.4623,
+      "step": 9348
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1202962398529053,
+      "learning_rate": 2.6898562181925983e-05,
+      "loss": 5.6011,
+      "step": 9352
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9215221405029297,
+      "learning_rate": 2.6888680270764365e-05,
+      "loss": 5.6439,
+      "step": 9356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8922703266143799,
+      "learning_rate": 2.6878798359602748e-05,
+      "loss": 5.6081,
+      "step": 9360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.219496726989746,
+      "learning_rate": 2.686891644844113e-05,
+      "loss": 5.7311,
+      "step": 9364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.926804542541504,
+      "learning_rate": 2.6859034537279508e-05,
+      "loss": 5.5689,
+      "step": 9368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2841477394104004,
+      "learning_rate": 2.684915262611789e-05,
+      "loss": 5.6086,
+      "step": 9372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9708247184753418,
+      "learning_rate": 2.6839270714956272e-05,
+      "loss": 5.6008,
+      "step": 9376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7928872108459473,
+      "learning_rate": 2.6829388803794654e-05,
+      "loss": 5.6374,
+      "step": 9380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7741127014160156,
+      "learning_rate": 2.6819506892633033e-05,
+      "loss": 5.611,
+      "step": 9384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.63377845287323,
+      "learning_rate": 2.6809624981471415e-05,
+      "loss": 5.6611,
+      "step": 9388
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8570857048034668,
+      "learning_rate": 2.67997430703098e-05,
+      "loss": 5.5389,
+      "step": 9392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8265759944915771,
+      "learning_rate": 2.6789861159148183e-05,
+      "loss": 5.6445,
+      "step": 9396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8934123516082764,
+      "learning_rate": 2.6779979247986565e-05,
+      "loss": 5.6222,
+      "step": 9400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8162897825241089,
+      "learning_rate": 2.6770097336824947e-05,
+      "loss": 5.474,
+      "step": 9404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1060986518859863,
+      "learning_rate": 2.6760215425663325e-05,
+      "loss": 5.6903,
+      "step": 9408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9316869974136353,
+      "learning_rate": 2.6750333514501708e-05,
+      "loss": 5.4954,
+      "step": 9412
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7672759294509888,
+      "learning_rate": 2.674045160334009e-05,
+      "loss": 5.604,
+      "step": 9416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6622782945632935,
+      "learning_rate": 2.6730569692178468e-05,
+      "loss": 5.5769,
+      "step": 9420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6604558229446411,
+      "learning_rate": 2.672068778101685e-05,
+      "loss": 5.5286,
+      "step": 9424
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.779937744140625,
+      "learning_rate": 2.6710805869855232e-05,
+      "loss": 5.6161,
+      "step": 9428
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9902960062026978,
+      "learning_rate": 2.6700923958693614e-05,
+      "loss": 5.6306,
+      "step": 9432
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.084911823272705,
+      "learning_rate": 2.6691042047531993e-05,
+      "loss": 5.5941,
+      "step": 9436
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6512478590011597,
+      "learning_rate": 2.6681160136370375e-05,
+      "loss": 5.5051,
+      "step": 9440
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.100574016571045,
+      "learning_rate": 2.6671278225208757e-05,
+      "loss": 5.5,
+      "step": 9444
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.922301173210144,
+      "learning_rate": 2.666139631404714e-05,
+      "loss": 5.7458,
+      "step": 9448
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0800986289978027,
+      "learning_rate": 2.6651514402885518e-05,
+      "loss": 5.4861,
+      "step": 9452
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9604395627975464,
+      "learning_rate": 2.66416324917239e-05,
+      "loss": 5.6116,
+      "step": 9456
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9513726234436035,
+      "learning_rate": 2.6631750580562282e-05,
+      "loss": 5.6,
+      "step": 9460
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6587291955947876,
+      "learning_rate": 2.6621868669400664e-05,
+      "loss": 5.5058,
+      "step": 9464
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.724055290222168,
+      "learning_rate": 2.6611986758239043e-05,
+      "loss": 5.6968,
+      "step": 9468
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8488603830337524,
+      "learning_rate": 2.6602104847077425e-05,
+      "loss": 5.4925,
+      "step": 9472
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9515416622161865,
+      "learning_rate": 2.6592222935915807e-05,
+      "loss": 5.5739,
+      "step": 9476
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8610438108444214,
+      "learning_rate": 2.6582341024754186e-05,
+      "loss": 5.6003,
+      "step": 9480
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7559446096420288,
+      "learning_rate": 2.6572459113592568e-05,
+      "loss": 5.6684,
+      "step": 9484
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.95654296875,
+      "learning_rate": 2.656257720243095e-05,
+      "loss": 5.5024,
+      "step": 9488
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9304943084716797,
+      "learning_rate": 2.6552695291269332e-05,
+      "loss": 5.5273,
+      "step": 9492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9279637336730957,
+      "learning_rate": 2.654281338010771e-05,
+      "loss": 5.503,
+      "step": 9496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.975297212600708,
+      "learning_rate": 2.6532931468946093e-05,
+      "loss": 5.4805,
+      "step": 9500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8157209157943726,
+      "learning_rate": 2.6523049557784478e-05,
+      "loss": 5.5161,
+      "step": 9504
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9559147357940674,
+      "learning_rate": 2.651316764662286e-05,
+      "loss": 5.5609,
+      "step": 9508
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0093133449554443,
+      "learning_rate": 2.6503285735461242e-05,
+      "loss": 5.5939,
+      "step": 9512
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8078559637069702,
+      "learning_rate": 2.6493403824299624e-05,
+      "loss": 5.4802,
+      "step": 9516
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8105264902114868,
+      "learning_rate": 2.6483521913138003e-05,
+      "loss": 5.6028,
+      "step": 9520
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1344854831695557,
+      "learning_rate": 2.6473640001976385e-05,
+      "loss": 5.6429,
+      "step": 9524
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7309706211090088,
+      "learning_rate": 2.6463758090814767e-05,
+      "loss": 5.3914,
+      "step": 9528
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0924670696258545,
+      "learning_rate": 2.645387617965315e-05,
+      "loss": 5.5516,
+      "step": 9532
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.076195001602173,
+      "learning_rate": 2.6443994268491528e-05,
+      "loss": 5.5816,
+      "step": 9536
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.286782741546631,
+      "learning_rate": 2.643411235732991e-05,
+      "loss": 5.633,
+      "step": 9540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0521233081817627,
+      "learning_rate": 2.6424230446168292e-05,
+      "loss": 5.5289,
+      "step": 9544
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7153042554855347,
+      "learning_rate": 2.6414348535006674e-05,
+      "loss": 5.5659,
+      "step": 9548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8992023468017578,
+      "learning_rate": 2.6404466623845053e-05,
+      "loss": 5.5774,
+      "step": 9552
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.049445390701294,
+      "learning_rate": 2.6394584712683435e-05,
+      "loss": 5.5893,
+      "step": 9556
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7689001560211182,
+      "learning_rate": 2.6384702801521817e-05,
+      "loss": 5.4599,
+      "step": 9560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7153732776641846,
+      "learning_rate": 2.6374820890360195e-05,
+      "loss": 5.4882,
+      "step": 9564
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8958677053451538,
+      "learning_rate": 2.6364938979198577e-05,
+      "loss": 5.4936,
+      "step": 9568
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7627476453781128,
+      "learning_rate": 2.635505706803696e-05,
+      "loss": 5.4682,
+      "step": 9572
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.768936276435852,
+      "learning_rate": 2.634517515687534e-05,
+      "loss": 5.5129,
+      "step": 9576
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9146132469177246,
+      "learning_rate": 2.633529324571372e-05,
+      "loss": 5.7007,
+      "step": 9580
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9375848770141602,
+      "learning_rate": 2.6325411334552102e-05,
+      "loss": 5.6438,
+      "step": 9584
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8867487907409668,
+      "learning_rate": 2.6315529423390484e-05,
+      "loss": 5.5024,
+      "step": 9588
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8224083185195923,
+      "learning_rate": 2.6305647512228866e-05,
+      "loss": 5.4406,
+      "step": 9592
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8230053186416626,
+      "learning_rate": 2.6295765601067245e-05,
+      "loss": 5.6339,
+      "step": 9596
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9611897468566895,
+      "learning_rate": 2.6285883689905627e-05,
+      "loss": 5.377,
+      "step": 9600
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.842227816581726,
+      "learning_rate": 2.627600177874401e-05,
+      "loss": 5.61,
+      "step": 9604
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8939987421035767,
+      "learning_rate": 2.6266119867582388e-05,
+      "loss": 5.6035,
+      "step": 9608
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.800186276435852,
+      "learning_rate": 2.625623795642077e-05,
+      "loss": 5.6362,
+      "step": 9612
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1012604236602783,
+      "learning_rate": 2.6246356045259152e-05,
+      "loss": 5.7597,
+      "step": 9616
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9128572940826416,
+      "learning_rate": 2.6236474134097537e-05,
+      "loss": 5.5882,
+      "step": 9620
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9831055402755737,
+      "learning_rate": 2.622659222293592e-05,
+      "loss": 5.6463,
+      "step": 9624
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7044267654418945,
+      "learning_rate": 2.62167103117743e-05,
+      "loss": 5.4153,
+      "step": 9628
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.859174132347107,
+      "learning_rate": 2.620682840061268e-05,
+      "loss": 5.4704,
+      "step": 9632
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8365576267242432,
+      "learning_rate": 2.6196946489451062e-05,
+      "loss": 5.6485,
+      "step": 9636
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8253902196884155,
+      "learning_rate": 2.6187064578289444e-05,
+      "loss": 5.6777,
+      "step": 9640
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8828972578048706,
+      "learning_rate": 2.6177182667127826e-05,
+      "loss": 5.4909,
+      "step": 9644
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7956063747406006,
+      "learning_rate": 2.6167300755966205e-05,
+      "loss": 5.5966,
+      "step": 9648
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1321167945861816,
+      "learning_rate": 2.6157418844804587e-05,
+      "loss": 5.6254,
+      "step": 9652
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7192585468292236,
+      "learning_rate": 2.614753693364297e-05,
+      "loss": 5.4451,
+      "step": 9656
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9337345361709595,
+      "learning_rate": 2.613765502248135e-05,
+      "loss": 5.6618,
+      "step": 9660
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0022127628326416,
+      "learning_rate": 2.612777311131973e-05,
+      "loss": 5.5478,
+      "step": 9664
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6713944673538208,
+      "learning_rate": 2.6117891200158112e-05,
+      "loss": 5.5078,
+      "step": 9668
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.902187466621399,
+      "learning_rate": 2.6108009288996494e-05,
+      "loss": 5.6425,
+      "step": 9672
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1489810943603516,
+      "learning_rate": 2.6098127377834876e-05,
+      "loss": 5.6363,
+      "step": 9676
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7200430631637573,
+      "learning_rate": 2.6088245466673255e-05,
+      "loss": 5.6002,
+      "step": 9680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9988877773284912,
+      "learning_rate": 2.6078363555511637e-05,
+      "loss": 5.6169,
+      "step": 9684
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.99684476852417,
+      "learning_rate": 2.606848164435002e-05,
+      "loss": 5.4686,
+      "step": 9688
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6895607709884644,
+      "learning_rate": 2.6058599733188398e-05,
+      "loss": 5.566,
+      "step": 9692
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6524453163146973,
+      "learning_rate": 2.604871782202678e-05,
+      "loss": 5.5477,
+      "step": 9696
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7908339500427246,
+      "learning_rate": 2.603883591086516e-05,
+      "loss": 5.7,
+      "step": 9700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8683209419250488,
+      "learning_rate": 2.6028953999703544e-05,
+      "loss": 5.4894,
+      "step": 9704
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.883863925933838,
+      "learning_rate": 2.6019072088541922e-05,
+      "loss": 5.6895,
+      "step": 9708
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8453763723373413,
+      "learning_rate": 2.6009190177380304e-05,
+      "loss": 5.4554,
+      "step": 9712
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7885375022888184,
+      "learning_rate": 2.5999308266218686e-05,
+      "loss": 5.6365,
+      "step": 9716
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9146047830581665,
+      "learning_rate": 2.598942635505707e-05,
+      "loss": 5.5714,
+      "step": 9720
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9622520208358765,
+      "learning_rate": 2.5979544443895447e-05,
+      "loss": 5.6337,
+      "step": 9724
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.236665725708008,
+      "learning_rate": 2.596966253273383e-05,
+      "loss": 5.5048,
+      "step": 9728
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8453449010849,
+      "learning_rate": 2.595978062157221e-05,
+      "loss": 5.6236,
+      "step": 9732
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9595757722854614,
+      "learning_rate": 2.5949898710410597e-05,
+      "loss": 5.6084,
+      "step": 9736
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1944825649261475,
+      "learning_rate": 2.594001679924898e-05,
+      "loss": 5.4383,
+      "step": 9740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.821465015411377,
+      "learning_rate": 2.593013488808736e-05,
+      "loss": 5.4977,
+      "step": 9744
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7616369724273682,
+      "learning_rate": 2.592025297692574e-05,
+      "loss": 5.5552,
+      "step": 9748
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8618212938308716,
+      "learning_rate": 2.591037106576412e-05,
+      "loss": 5.6036,
+      "step": 9752
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.120798110961914,
+      "learning_rate": 2.5900489154602504e-05,
+      "loss": 5.4679,
+      "step": 9756
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2172162532806396,
+      "learning_rate": 2.5890607243440886e-05,
+      "loss": 5.583,
+      "step": 9760
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9342153072357178,
+      "learning_rate": 2.5880725332279264e-05,
+      "loss": 5.5783,
+      "step": 9764
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6747381687164307,
+      "learning_rate": 2.5870843421117647e-05,
+      "loss": 5.6047,
+      "step": 9768
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0254557132720947,
+      "learning_rate": 2.586096150995603e-05,
+      "loss": 5.6368,
+      "step": 9772
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9324589967727661,
+      "learning_rate": 2.5851079598794407e-05,
+      "loss": 5.6284,
+      "step": 9776
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6605185270309448,
+      "learning_rate": 2.584119768763279e-05,
+      "loss": 5.5681,
+      "step": 9780
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6846225261688232,
+      "learning_rate": 2.583131577647117e-05,
+      "loss": 5.5159,
+      "step": 9784
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0729458332061768,
+      "learning_rate": 2.5821433865309553e-05,
+      "loss": 5.5701,
+      "step": 9788
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.831186056137085,
+      "learning_rate": 2.5811551954147932e-05,
+      "loss": 5.5996,
+      "step": 9792
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.744837999343872,
+      "learning_rate": 2.5801670042986314e-05,
+      "loss": 5.5798,
+      "step": 9796
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7464189529418945,
+      "learning_rate": 2.5791788131824696e-05,
+      "loss": 5.4147,
+      "step": 9800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.083294630050659,
+      "learning_rate": 2.5781906220663078e-05,
+      "loss": 5.5884,
+      "step": 9804
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.848488211631775,
+      "learning_rate": 2.5772024309501457e-05,
+      "loss": 5.4192,
+      "step": 9808
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8457578420639038,
+      "learning_rate": 2.576214239833984e-05,
+      "loss": 5.6189,
+      "step": 9812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9450173377990723,
+      "learning_rate": 2.575226048717822e-05,
+      "loss": 5.5665,
+      "step": 9816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7913298606872559,
+      "learning_rate": 2.57423785760166e-05,
+      "loss": 5.5333,
+      "step": 9820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8808374404907227,
+      "learning_rate": 2.5732496664854982e-05,
+      "loss": 5.4615,
+      "step": 9824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1443545818328857,
+      "learning_rate": 2.5722614753693364e-05,
+      "loss": 5.599,
+      "step": 9828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.867958903312683,
+      "learning_rate": 2.5712732842531746e-05,
+      "loss": 5.6383,
+      "step": 9832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9032071828842163,
+      "learning_rate": 2.5702850931370125e-05,
+      "loss": 5.6756,
+      "step": 9836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0405712127685547,
+      "learning_rate": 2.5692969020208507e-05,
+      "loss": 5.5741,
+      "step": 9840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9483815431594849,
+      "learning_rate": 2.568308710904689e-05,
+      "loss": 5.5603,
+      "step": 9844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6286249160766602,
+      "learning_rate": 2.5673205197885274e-05,
+      "loss": 5.5213,
+      "step": 9848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.009946346282959,
+      "learning_rate": 2.5663323286723656e-05,
+      "loss": 5.6207,
+      "step": 9852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7582067251205444,
+      "learning_rate": 2.5653441375562038e-05,
+      "loss": 5.6453,
+      "step": 9856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7018730640411377,
+      "learning_rate": 2.5643559464400417e-05,
+      "loss": 5.6713,
+      "step": 9860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.920355200767517,
+      "learning_rate": 2.56336775532388e-05,
+      "loss": 5.6109,
+      "step": 9864
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.79157555103302,
+      "learning_rate": 2.562379564207718e-05,
+      "loss": 5.5102,
+      "step": 9868
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7163673639297485,
+      "learning_rate": 2.5613913730915563e-05,
+      "loss": 5.6472,
+      "step": 9872
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8523849248886108,
+      "learning_rate": 2.5604031819753942e-05,
+      "loss": 5.6162,
+      "step": 9876
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.006044626235962,
+      "learning_rate": 2.5594149908592324e-05,
+      "loss": 5.6793,
+      "step": 9880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.745415449142456,
+      "learning_rate": 2.558673847522111e-05,
+      "loss": 5.5417,
+      "step": 9884
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9379210472106934,
+      "learning_rate": 2.5576856564059492e-05,
+      "loss": 5.6047,
+      "step": 9888
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8513517379760742,
+      "learning_rate": 2.5566974652897875e-05,
+      "loss": 5.6515,
+      "step": 9892
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8587697744369507,
+      "learning_rate": 2.5557092741736253e-05,
+      "loss": 5.5598,
+      "step": 9896
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9905014038085938,
+      "learning_rate": 2.5547210830574635e-05,
+      "loss": 5.6137,
+      "step": 9900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.83003568649292,
+      "learning_rate": 2.5537328919413017e-05,
+      "loss": 5.6503,
+      "step": 9904
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.71113121509552,
+      "learning_rate": 2.55274470082514e-05,
+      "loss": 5.561,
+      "step": 9908
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6225950717926025,
+      "learning_rate": 2.5517565097089778e-05,
+      "loss": 5.4564,
+      "step": 9912
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8159841299057007,
+      "learning_rate": 2.550768318592816e-05,
+      "loss": 5.5984,
+      "step": 9916
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.003652572631836,
+      "learning_rate": 2.5497801274766542e-05,
+      "loss": 5.4597,
+      "step": 9920
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8188750743865967,
+      "learning_rate": 2.5487919363604924e-05,
+      "loss": 5.46,
+      "step": 9924
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.832154631614685,
+      "learning_rate": 2.5478037452443303e-05,
+      "loss": 5.5748,
+      "step": 9928
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9353375434875488,
+      "learning_rate": 2.5468155541281685e-05,
+      "loss": 5.5672,
+      "step": 9932
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9940686225891113,
+      "learning_rate": 2.5458273630120067e-05,
+      "loss": 5.5623,
+      "step": 9936
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8885998725891113,
+      "learning_rate": 2.544839171895845e-05,
+      "loss": 5.6176,
+      "step": 9940
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8526244163513184,
+      "learning_rate": 2.5438509807796828e-05,
+      "loss": 5.6358,
+      "step": 9944
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8423300981521606,
+      "learning_rate": 2.542862789663521e-05,
+      "loss": 5.5085,
+      "step": 9948
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6946587562561035,
+      "learning_rate": 2.5418745985473592e-05,
+      "loss": 5.452,
+      "step": 9952
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8842428922653198,
+      "learning_rate": 2.540886407431197e-05,
+      "loss": 5.4547,
+      "step": 9956
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8062175512313843,
+      "learning_rate": 2.5398982163150353e-05,
+      "loss": 5.5232,
+      "step": 9960
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.968763828277588,
+      "learning_rate": 2.5389100251988735e-05,
+      "loss": 5.6634,
+      "step": 9964
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.782151460647583,
+      "learning_rate": 2.5379218340827117e-05,
+      "loss": 5.4746,
+      "step": 9968
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0645010471343994,
+      "learning_rate": 2.5369336429665495e-05,
+      "loss": 5.5755,
+      "step": 9972
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.081439733505249,
+      "learning_rate": 2.5359454518503877e-05,
+      "loss": 5.5264,
+      "step": 9976
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9738670587539673,
+      "learning_rate": 2.534957260734226e-05,
+      "loss": 5.5369,
+      "step": 9980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.08581805229187,
+      "learning_rate": 2.533969069618064e-05,
+      "loss": 5.6849,
+      "step": 9984
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.998647928237915,
+      "learning_rate": 2.532980878501902e-05,
+      "loss": 5.628,
+      "step": 9988
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9414081573486328,
+      "learning_rate": 2.5319926873857402e-05,
+      "loss": 5.5865,
+      "step": 9992
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7478708028793335,
+      "learning_rate": 2.5310044962695788e-05,
+      "loss": 5.4073,
+      "step": 9996
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.031409740447998,
+      "learning_rate": 2.530016305153417e-05,
+      "loss": 5.6781,
+      "step": 10000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8171534538269043,
+      "learning_rate": 2.5290281140372552e-05,
+      "loss": 5.6865,
+      "step": 10004
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.128012180328369,
+      "learning_rate": 2.5280399229210934e-05,
+      "loss": 5.6194,
+      "step": 10008
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1675913333892822,
+      "learning_rate": 2.5270517318049313e-05,
+      "loss": 5.4726,
+      "step": 10012
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.229329824447632,
+      "learning_rate": 2.5260635406887695e-05,
+      "loss": 5.564,
+      "step": 10016
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8510156869888306,
+      "learning_rate": 2.5250753495726077e-05,
+      "loss": 5.4116,
+      "step": 10020
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9130499362945557,
+      "learning_rate": 2.524087158456446e-05,
+      "loss": 5.5758,
+      "step": 10024
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.073378562927246,
+      "learning_rate": 2.5230989673402837e-05,
+      "loss": 5.5724,
+      "step": 10028
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.248762845993042,
+      "learning_rate": 2.522110776224122e-05,
+      "loss": 5.6893,
+      "step": 10032
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0180504322052,
+      "learning_rate": 2.52112258510796e-05,
+      "loss": 5.642,
+      "step": 10036
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.786154866218567,
+      "learning_rate": 2.520134393991798e-05,
+      "loss": 5.6359,
+      "step": 10040
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8328882455825806,
+      "learning_rate": 2.5191462028756362e-05,
+      "loss": 5.5998,
+      "step": 10044
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1407294273376465,
+      "learning_rate": 2.5181580117594744e-05,
+      "loss": 5.7289,
+      "step": 10048
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6847882270812988,
+      "learning_rate": 2.5171698206433126e-05,
+      "loss": 5.6441,
+      "step": 10052
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7295290231704712,
+      "learning_rate": 2.5161816295271505e-05,
+      "loss": 5.5917,
+      "step": 10056
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.069204807281494,
+      "learning_rate": 2.5151934384109887e-05,
+      "loss": 5.3987,
+      "step": 10060
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8884251117706299,
+      "learning_rate": 2.514205247294827e-05,
+      "loss": 5.5554,
+      "step": 10064
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.767871618270874,
+      "learning_rate": 2.513217056178665e-05,
+      "loss": 5.6294,
+      "step": 10068
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8906821012496948,
+      "learning_rate": 2.512228865062503e-05,
+      "loss": 5.5464,
+      "step": 10072
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6515437364578247,
+      "learning_rate": 2.5112406739463412e-05,
+      "loss": 5.4848,
+      "step": 10076
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0448694229125977,
+      "learning_rate": 2.5102524828301794e-05,
+      "loss": 5.5193,
+      "step": 10080
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.867287278175354,
+      "learning_rate": 2.5092642917140173e-05,
+      "loss": 5.5843,
+      "step": 10084
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.833117961883545,
+      "learning_rate": 2.5082761005978555e-05,
+      "loss": 5.4566,
+      "step": 10088
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.842322826385498,
+      "learning_rate": 2.5072879094816937e-05,
+      "loss": 5.5783,
+      "step": 10092
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7041923999786377,
+      "learning_rate": 2.506299718365532e-05,
+      "loss": 5.6291,
+      "step": 10096
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9583604335784912,
+      "learning_rate": 2.5053115272493698e-05,
+      "loss": 5.5434,
+      "step": 10100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8357032537460327,
+      "learning_rate": 2.504323336133208e-05,
+      "loss": 5.5751,
+      "step": 10104
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9538915157318115,
+      "learning_rate": 2.503335145017047e-05,
+      "loss": 5.7138,
+      "step": 10108
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0937674045562744,
+      "learning_rate": 2.5023469539008847e-05,
+      "loss": 5.5237,
+      "step": 10112
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9279645681381226,
+      "learning_rate": 2.501358762784723e-05,
+      "loss": 5.6076,
+      "step": 10116
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8550045490264893,
+      "learning_rate": 2.500370571668561e-05,
+      "loss": 5.5574,
+      "step": 10120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0821774005889893,
+      "learning_rate": 2.499382380552399e-05,
+      "loss": 5.532,
+      "step": 10124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.088787078857422,
+      "learning_rate": 2.498394189436237e-05,
+      "loss": 5.6185,
+      "step": 10128
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7520018815994263,
+      "learning_rate": 2.497405998320075e-05,
+      "loss": 5.5648,
+      "step": 10132
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.963576078414917,
+      "learning_rate": 2.4964178072039136e-05,
+      "loss": 5.6652,
+      "step": 10136
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9875547885894775,
+      "learning_rate": 2.4954296160877515e-05,
+      "loss": 5.5179,
+      "step": 10140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9994404315948486,
+      "learning_rate": 2.4944414249715897e-05,
+      "loss": 5.6485,
+      "step": 10144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0803356170654297,
+      "learning_rate": 2.493453233855428e-05,
+      "loss": 5.4617,
+      "step": 10148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6625994443893433,
+      "learning_rate": 2.492465042739266e-05,
+      "loss": 5.5363,
+      "step": 10152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8753883838653564,
+      "learning_rate": 2.491476851623104e-05,
+      "loss": 5.3869,
+      "step": 10156
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.644220232963562,
+      "learning_rate": 2.4904886605069422e-05,
+      "loss": 5.5971,
+      "step": 10160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.017249584197998,
+      "learning_rate": 2.4895004693907804e-05,
+      "loss": 5.506,
+      "step": 10164
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.240140914916992,
+      "learning_rate": 2.4885122782746182e-05,
+      "loss": 5.6691,
+      "step": 10168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.842436671257019,
+      "learning_rate": 2.4875240871584565e-05,
+      "loss": 5.5673,
+      "step": 10172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8828648328781128,
+      "learning_rate": 2.4865358960422947e-05,
+      "loss": 5.5275,
+      "step": 10176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1830368041992188,
+      "learning_rate": 2.485547704926133e-05,
+      "loss": 5.6258,
+      "step": 10180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.012813091278076,
+      "learning_rate": 2.4845595138099707e-05,
+      "loss": 5.5947,
+      "step": 10184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7969881296157837,
+      "learning_rate": 2.483571322693809e-05,
+      "loss": 5.564,
+      "step": 10188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8949358463287354,
+      "learning_rate": 2.482583131577647e-05,
+      "loss": 5.6213,
+      "step": 10192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8343627452850342,
+      "learning_rate": 2.4815949404614854e-05,
+      "loss": 5.5869,
+      "step": 10196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8885095119476318,
+      "learning_rate": 2.4806067493453236e-05,
+      "loss": 5.586,
+      "step": 10200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0102386474609375,
+      "learning_rate": 2.4796185582291618e-05,
+      "loss": 5.5219,
+      "step": 10204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.127244472503662,
+      "learning_rate": 2.478630367113e-05,
+      "loss": 5.6196,
+      "step": 10208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7955610752105713,
+      "learning_rate": 2.477642175996838e-05,
+      "loss": 5.5848,
+      "step": 10212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9604214429855347,
+      "learning_rate": 2.476653984880676e-05,
+      "loss": 5.6014,
+      "step": 10216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.11737322807312,
+      "learning_rate": 2.4756657937645142e-05,
+      "loss": 5.5377,
+      "step": 10220
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0900325775146484,
+      "learning_rate": 2.4746776026483525e-05,
+      "loss": 5.5901,
+      "step": 10224
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7925187349319458,
+      "learning_rate": 2.4736894115321903e-05,
+      "loss": 5.4441,
+      "step": 10228
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.107675790786743,
+      "learning_rate": 2.4727012204160285e-05,
+      "loss": 5.5537,
+      "step": 10232
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0693657398223877,
+      "learning_rate": 2.4717130292998667e-05,
+      "loss": 5.5913,
+      "step": 10236
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.158755302429199,
+      "learning_rate": 2.4707248381837046e-05,
+      "loss": 5.7313,
+      "step": 10240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9274017810821533,
+      "learning_rate": 2.4697366470675428e-05,
+      "loss": 5.5108,
+      "step": 10244
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8750487565994263,
+      "learning_rate": 2.468748455951381e-05,
+      "loss": 5.5994,
+      "step": 10248
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6671373844146729,
+      "learning_rate": 2.4677602648352192e-05,
+      "loss": 5.6319,
+      "step": 10252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.721057415008545,
+      "learning_rate": 2.4667720737190574e-05,
+      "loss": 5.5567,
+      "step": 10256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7472023963928223,
+      "learning_rate": 2.4657838826028956e-05,
+      "loss": 5.5035,
+      "step": 10260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.897161602973938,
+      "learning_rate": 2.464795691486734e-05,
+      "loss": 5.5944,
+      "step": 10264
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9217979907989502,
+      "learning_rate": 2.4638075003705717e-05,
+      "loss": 5.5253,
+      "step": 10268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.010005474090576,
+      "learning_rate": 2.46281930925441e-05,
+      "loss": 5.7227,
+      "step": 10272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7467890977859497,
+      "learning_rate": 2.461831118138248e-05,
+      "loss": 5.5426,
+      "step": 10276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9532781839370728,
+      "learning_rate": 2.4608429270220863e-05,
+      "loss": 5.5081,
+      "step": 10280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.979059100151062,
+      "learning_rate": 2.4598547359059242e-05,
+      "loss": 5.5655,
+      "step": 10284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7593183517456055,
+      "learning_rate": 2.4588665447897624e-05,
+      "loss": 5.5235,
+      "step": 10288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8140133619308472,
+      "learning_rate": 2.4578783536736006e-05,
+      "loss": 5.5346,
+      "step": 10292
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9852521419525146,
+      "learning_rate": 2.4568901625574388e-05,
+      "loss": 5.5365,
+      "step": 10296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.00048828125,
+      "learning_rate": 2.4559019714412767e-05,
+      "loss": 5.5772,
+      "step": 10300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9229919910430908,
+      "learning_rate": 2.454913780325115e-05,
+      "loss": 5.4598,
+      "step": 10304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9021358489990234,
+      "learning_rate": 2.453925589208953e-05,
+      "loss": 5.4991,
+      "step": 10308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9008359909057617,
+      "learning_rate": 2.4529373980927913e-05,
+      "loss": 5.5531,
+      "step": 10312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8930054903030396,
+      "learning_rate": 2.4519492069766295e-05,
+      "loss": 5.5356,
+      "step": 10316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8766874074935913,
+      "learning_rate": 2.4509610158604677e-05,
+      "loss": 5.6213,
+      "step": 10320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8489048480987549,
+      "learning_rate": 2.4499728247443056e-05,
+      "loss": 5.5647,
+      "step": 10324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0344486236572266,
+      "learning_rate": 2.4489846336281438e-05,
+      "loss": 5.5734,
+      "step": 10328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.83133864402771,
+      "learning_rate": 2.447996442511982e-05,
+      "loss": 5.4536,
+      "step": 10332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6086121797561646,
+      "learning_rate": 2.4470082513958202e-05,
+      "loss": 5.5275,
+      "step": 10336
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0024595260620117,
+      "learning_rate": 2.446020060279658e-05,
+      "loss": 5.5924,
+      "step": 10340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9737377166748047,
+      "learning_rate": 2.4450318691634963e-05,
+      "loss": 5.5466,
+      "step": 10344
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9475603103637695,
+      "learning_rate": 2.4440436780473345e-05,
+      "loss": 5.5415,
+      "step": 10348
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1475892066955566,
+      "learning_rate": 2.4430554869311727e-05,
+      "loss": 5.6749,
+      "step": 10352
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7354161739349365,
+      "learning_rate": 2.4420672958150105e-05,
+      "loss": 5.4942,
+      "step": 10356
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9726040363311768,
+      "learning_rate": 2.4410791046988487e-05,
+      "loss": 5.5258,
+      "step": 10360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7719279527664185,
+      "learning_rate": 2.440090913582687e-05,
+      "loss": 5.5169,
+      "step": 10364
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8121774196624756,
+      "learning_rate": 2.439102722466525e-05,
+      "loss": 5.5811,
+      "step": 10368
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.656872034072876,
+      "learning_rate": 2.4381145313503634e-05,
+      "loss": 5.6121,
+      "step": 10372
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9030474424362183,
+      "learning_rate": 2.4371263402342016e-05,
+      "loss": 5.5099,
+      "step": 10376
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8425421714782715,
+      "learning_rate": 2.4361381491180398e-05,
+      "loss": 5.5167,
+      "step": 10380
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0104122161865234,
+      "learning_rate": 2.4351499580018776e-05,
+      "loss": 5.629,
+      "step": 10384
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7906516790390015,
+      "learning_rate": 2.434161766885716e-05,
+      "loss": 5.632,
+      "step": 10388
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.966325044631958,
+      "learning_rate": 2.433173575769554e-05,
+      "loss": 5.6271,
+      "step": 10392
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9197744131088257,
+      "learning_rate": 2.432185384653392e-05,
+      "loss": 5.5729,
+      "step": 10396
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9084253311157227,
+      "learning_rate": 2.43119719353723e-05,
+      "loss": 5.5213,
+      "step": 10400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.743037223815918,
+      "learning_rate": 2.4302090024210683e-05,
+      "loss": 5.6202,
+      "step": 10404
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6884185075759888,
+      "learning_rate": 2.4292208113049065e-05,
+      "loss": 5.52,
+      "step": 10408
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9082211256027222,
+      "learning_rate": 2.4282326201887444e-05,
+      "loss": 5.6913,
+      "step": 10412
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7120208740234375,
+      "learning_rate": 2.4272444290725826e-05,
+      "loss": 5.536,
+      "step": 10416
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0013391971588135,
+      "learning_rate": 2.4262562379564208e-05,
+      "loss": 5.6444,
+      "step": 10420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.751233458518982,
+      "learning_rate": 2.425268046840259e-05,
+      "loss": 5.4514,
+      "step": 10424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7701613903045654,
+      "learning_rate": 2.4242798557240972e-05,
+      "loss": 5.5492,
+      "step": 10428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.672849416732788,
+      "learning_rate": 2.4232916646079354e-05,
+      "loss": 5.5202,
+      "step": 10432
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9248921871185303,
+      "learning_rate": 2.4223034734917736e-05,
+      "loss": 5.5876,
+      "step": 10436
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7727632522583008,
+      "learning_rate": 2.4213152823756115e-05,
+      "loss": 5.5529,
+      "step": 10440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9755526781082153,
+      "learning_rate": 2.4203270912594497e-05,
+      "loss": 5.6183,
+      "step": 10444
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7922430038452148,
+      "learning_rate": 2.419338900143288e-05,
+      "loss": 5.5369,
+      "step": 10448
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.185535430908203,
+      "learning_rate": 2.4183507090271258e-05,
+      "loss": 5.6702,
+      "step": 10452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.659045934677124,
+      "learning_rate": 2.417362517910964e-05,
+      "loss": 5.4956,
+      "step": 10456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8726792335510254,
+      "learning_rate": 2.4163743267948022e-05,
+      "loss": 5.5971,
+      "step": 10460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8656306266784668,
+      "learning_rate": 2.4153861356786404e-05,
+      "loss": 5.6505,
+      "step": 10464
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1114795207977295,
+      "learning_rate": 2.4143979445624783e-05,
+      "loss": 5.6579,
+      "step": 10468
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7665212154388428,
+      "learning_rate": 2.4134097534463165e-05,
+      "loss": 5.5182,
+      "step": 10472
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0020570755004883,
+      "learning_rate": 2.4124215623301547e-05,
+      "loss": 5.6146,
+      "step": 10476
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.81464684009552,
+      "learning_rate": 2.411433371213993e-05,
+      "loss": 5.6334,
+      "step": 10480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0478081703186035,
+      "learning_rate": 2.410445180097831e-05,
+      "loss": 5.6295,
+      "step": 10484
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9547245502471924,
+      "learning_rate": 2.4094569889816693e-05,
+      "loss": 5.604,
+      "step": 10488
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.018515110015869,
+      "learning_rate": 2.4084687978655075e-05,
+      "loss": 5.498,
+      "step": 10492
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7030364274978638,
+      "learning_rate": 2.4074806067493454e-05,
+      "loss": 5.654,
+      "step": 10496
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0866682529449463,
+      "learning_rate": 2.4064924156331836e-05,
+      "loss": 5.5155,
+      "step": 10500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.075389862060547,
+      "learning_rate": 2.4055042245170218e-05,
+      "loss": 5.5227,
+      "step": 10504
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9765968322753906,
+      "learning_rate": 2.40451603340086e-05,
+      "loss": 5.6732,
+      "step": 10508
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0013983249664307,
+      "learning_rate": 2.403527842284698e-05,
+      "loss": 5.6551,
+      "step": 10512
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.044642448425293,
+      "learning_rate": 2.402539651168536e-05,
+      "loss": 5.5395,
+      "step": 10516
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6758370399475098,
+      "learning_rate": 2.4015514600523743e-05,
+      "loss": 5.5563,
+      "step": 10520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8717275857925415,
+      "learning_rate": 2.400563268936212e-05,
+      "loss": 5.5039,
+      "step": 10524
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9705373048782349,
+      "learning_rate": 2.3995750778200504e-05,
+      "loss": 5.6172,
+      "step": 10528
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7623977661132812,
+      "learning_rate": 2.3985868867038886e-05,
+      "loss": 5.4925,
+      "step": 10532
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9180505275726318,
+      "learning_rate": 2.3975986955877268e-05,
+      "loss": 5.6497,
+      "step": 10536
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.940900206565857,
+      "learning_rate": 2.396610504471565e-05,
+      "loss": 5.5216,
+      "step": 10540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7929134368896484,
+      "learning_rate": 2.3956223133554032e-05,
+      "loss": 5.483,
+      "step": 10544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8497039079666138,
+      "learning_rate": 2.3946341222392414e-05,
+      "loss": 5.5895,
+      "step": 10548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8685697317123413,
+      "learning_rate": 2.3936459311230793e-05,
+      "loss": 5.5565,
+      "step": 10552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2542967796325684,
+      "learning_rate": 2.3926577400069175e-05,
+      "loss": 5.5895,
+      "step": 10556
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.043692111968994,
+      "learning_rate": 2.3916695488907557e-05,
+      "loss": 5.4708,
+      "step": 10560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1556544303894043,
+      "learning_rate": 2.390681357774594e-05,
+      "loss": 5.5257,
+      "step": 10564
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0351712703704834,
+      "learning_rate": 2.3896931666584317e-05,
+      "loss": 5.5247,
+      "step": 10568
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7244350910186768,
+      "learning_rate": 2.38870497554227e-05,
+      "loss": 5.4927,
+      "step": 10572
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0928471088409424,
+      "learning_rate": 2.387716784426108e-05,
+      "loss": 5.4176,
+      "step": 10576
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0530073642730713,
+      "learning_rate": 2.3867285933099464e-05,
+      "loss": 5.6525,
+      "step": 10580
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8700250387191772,
+      "learning_rate": 2.3857404021937842e-05,
+      "loss": 5.5591,
+      "step": 10584
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6774561405181885,
+      "learning_rate": 2.3847522110776224e-05,
+      "loss": 5.5925,
+      "step": 10588
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1128244400024414,
+      "learning_rate": 2.3837640199614606e-05,
+      "loss": 5.5118,
+      "step": 10592
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9600874185562134,
+      "learning_rate": 2.382775828845299e-05,
+      "loss": 5.563,
+      "step": 10596
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9983313083648682,
+      "learning_rate": 2.381787637729137e-05,
+      "loss": 5.4649,
+      "step": 10600
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9052096605300903,
+      "learning_rate": 2.3807994466129753e-05,
+      "loss": 5.5366,
+      "step": 10604
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9954668283462524,
+      "learning_rate": 2.379811255496813e-05,
+      "loss": 5.5479,
+      "step": 10608
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9341228008270264,
+      "learning_rate": 2.3788230643806513e-05,
+      "loss": 5.4649,
+      "step": 10612
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9938849210739136,
+      "learning_rate": 2.3778348732644895e-05,
+      "loss": 5.5791,
+      "step": 10616
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6775513887405396,
+      "learning_rate": 2.3768466821483277e-05,
+      "loss": 5.61,
+      "step": 10620
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9239006042480469,
+      "learning_rate": 2.3758584910321656e-05,
+      "loss": 5.5689,
+      "step": 10624
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.060796022415161,
+      "learning_rate": 2.3748702999160038e-05,
+      "loss": 5.5823,
+      "step": 10628
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.95215904712677,
+      "learning_rate": 2.373882108799842e-05,
+      "loss": 5.4581,
+      "step": 10632
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9908874034881592,
+      "learning_rate": 2.3728939176836802e-05,
+      "loss": 5.5968,
+      "step": 10636
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6556943655014038,
+      "learning_rate": 2.371905726567518e-05,
+      "loss": 5.5399,
+      "step": 10640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9152753353118896,
+      "learning_rate": 2.3709175354513563e-05,
+      "loss": 5.6018,
+      "step": 10644
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3656210899353027,
+      "learning_rate": 2.3699293443351945e-05,
+      "loss": 5.4332,
+      "step": 10648
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9801435470581055,
+      "learning_rate": 2.3689411532190324e-05,
+      "loss": 5.5684,
+      "step": 10652
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1441659927368164,
+      "learning_rate": 2.367952962102871e-05,
+      "loss": 5.3962,
+      "step": 10656
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2612667083740234,
+      "learning_rate": 2.366964770986709e-05,
+      "loss": 5.5301,
+      "step": 10660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1340534687042236,
+      "learning_rate": 2.3659765798705473e-05,
+      "loss": 5.6125,
+      "step": 10664
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0850844383239746,
+      "learning_rate": 2.3649883887543852e-05,
+      "loss": 5.4433,
+      "step": 10668
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7150609493255615,
+      "learning_rate": 2.3640001976382234e-05,
+      "loss": 5.5438,
+      "step": 10672
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9754505157470703,
+      "learning_rate": 2.3630120065220616e-05,
+      "loss": 5.6475,
+      "step": 10676
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9348641633987427,
+      "learning_rate": 2.3620238154058995e-05,
+      "loss": 5.6183,
+      "step": 10680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9712179899215698,
+      "learning_rate": 2.3610356242897377e-05,
+      "loss": 5.5499,
+      "step": 10684
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9912450313568115,
+      "learning_rate": 2.360047433173576e-05,
+      "loss": 5.534,
+      "step": 10688
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8730937242507935,
+      "learning_rate": 2.359059242057414e-05,
+      "loss": 5.6078,
+      "step": 10692
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.082500696182251,
+      "learning_rate": 2.358071050941252e-05,
+      "loss": 5.5638,
+      "step": 10696
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.990276575088501,
+      "learning_rate": 2.35708285982509e-05,
+      "loss": 5.4854,
+      "step": 10700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8696178197860718,
+      "learning_rate": 2.3560946687089284e-05,
+      "loss": 5.5041,
+      "step": 10704
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1446194648742676,
+      "learning_rate": 2.3551064775927666e-05,
+      "loss": 5.6358,
+      "step": 10708
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7713229656219482,
+      "learning_rate": 2.3541182864766048e-05,
+      "loss": 5.5635,
+      "step": 10712
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.876599907875061,
+      "learning_rate": 2.353130095360443e-05,
+      "loss": 5.4398,
+      "step": 10716
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7540398836135864,
+      "learning_rate": 2.3521419042442812e-05,
+      "loss": 5.4983,
+      "step": 10720
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0557737350463867,
+      "learning_rate": 2.351153713128119e-05,
+      "loss": 5.4517,
+      "step": 10724
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7068967819213867,
+      "learning_rate": 2.3501655220119573e-05,
+      "loss": 5.4742,
+      "step": 10728
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9416307210922241,
+      "learning_rate": 2.3491773308957955e-05,
+      "loss": 5.6298,
+      "step": 10732
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8731729984283447,
+      "learning_rate": 2.3481891397796333e-05,
+      "loss": 5.5531,
+      "step": 10736
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.060716390609741,
+      "learning_rate": 2.3472009486634715e-05,
+      "loss": 5.6034,
+      "step": 10740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.003988742828369,
+      "learning_rate": 2.3462127575473098e-05,
+      "loss": 5.5876,
+      "step": 10744
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0887482166290283,
+      "learning_rate": 2.345224566431148e-05,
+      "loss": 5.5417,
+      "step": 10748
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.017455577850342,
+      "learning_rate": 2.3442363753149858e-05,
+      "loss": 5.546,
+      "step": 10752
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8705204725265503,
+      "learning_rate": 2.343248184198824e-05,
+      "loss": 5.6274,
+      "step": 10756
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9942293167114258,
+      "learning_rate": 2.3422599930826622e-05,
+      "loss": 5.4287,
+      "step": 10760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.894716739654541,
+      "learning_rate": 2.3412718019665004e-05,
+      "loss": 5.5343,
+      "step": 10764
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.020045042037964,
+      "learning_rate": 2.3402836108503387e-05,
+      "loss": 5.6098,
+      "step": 10768
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.01450252532959,
+      "learning_rate": 2.339295419734177e-05,
+      "loss": 5.5875,
+      "step": 10772
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9330328702926636,
+      "learning_rate": 2.338307228618015e-05,
+      "loss": 5.5679,
+      "step": 10776
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.871109962463379,
+      "learning_rate": 2.337319037501853e-05,
+      "loss": 5.5312,
+      "step": 10780
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.02812123298645,
+      "learning_rate": 2.336330846385691e-05,
+      "loss": 5.3512,
+      "step": 10784
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9817901849746704,
+      "learning_rate": 2.3353426552695293e-05,
+      "loss": 5.4152,
+      "step": 10788
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9336769580841064,
+      "learning_rate": 2.3343544641533675e-05,
+      "loss": 5.5362,
+      "step": 10792
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8532609939575195,
+      "learning_rate": 2.3333662730372054e-05,
+      "loss": 5.4603,
+      "step": 10796
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8806451559066772,
+      "learning_rate": 2.3323780819210436e-05,
+      "loss": 5.3589,
+      "step": 10800
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8947298526763916,
+      "learning_rate": 2.3313898908048818e-05,
+      "loss": 5.5674,
+      "step": 10804
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9279778003692627,
+      "learning_rate": 2.3304016996887197e-05,
+      "loss": 5.5096,
+      "step": 10808
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0926554203033447,
+      "learning_rate": 2.329413508572558e-05,
+      "loss": 5.6147,
+      "step": 10812
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0202767848968506,
+      "learning_rate": 2.328425317456396e-05,
+      "loss": 5.6283,
+      "step": 10816
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9681588411331177,
+      "learning_rate": 2.3274371263402343e-05,
+      "loss": 5.711,
+      "step": 10820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.059591054916382,
+      "learning_rate": 2.3264489352240722e-05,
+      "loss": 5.5387,
+      "step": 10824
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7772566080093384,
+      "learning_rate": 2.3254607441079107e-05,
+      "loss": 5.5673,
+      "step": 10828
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.169037103652954,
+      "learning_rate": 2.324472552991749e-05,
+      "loss": 5.4922,
+      "step": 10832
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7867958545684814,
+      "learning_rate": 2.3234843618755868e-05,
+      "loss": 5.4829,
+      "step": 10836
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0761024951934814,
+      "learning_rate": 2.322496170759425e-05,
+      "loss": 5.57,
+      "step": 10840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8901103734970093,
+      "learning_rate": 2.3215079796432632e-05,
+      "loss": 5.6031,
+      "step": 10844
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8197063207626343,
+      "learning_rate": 2.3205197885271014e-05,
+      "loss": 5.6221,
+      "step": 10848
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8209093809127808,
+      "learning_rate": 2.3195315974109393e-05,
+      "loss": 5.4747,
+      "step": 10852
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1222522258758545,
+      "learning_rate": 2.3185434062947775e-05,
+      "loss": 5.5274,
+      "step": 10856
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7312264442443848,
+      "learning_rate": 2.3175552151786157e-05,
+      "loss": 5.5181,
+      "step": 10860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9499653577804565,
+      "learning_rate": 2.316567024062454e-05,
+      "loss": 5.5653,
+      "step": 10864
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7755954265594482,
+      "learning_rate": 2.3155788329462918e-05,
+      "loss": 5.5213,
+      "step": 10868
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6512964963912964,
+      "learning_rate": 2.31459064183013e-05,
+      "loss": 5.4163,
+      "step": 10872
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9016536474227905,
+      "learning_rate": 2.3136024507139682e-05,
+      "loss": 5.4877,
+      "step": 10876
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9362504482269287,
+      "learning_rate": 2.312614259597806e-05,
+      "loss": 5.5776,
+      "step": 10880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0821099281311035,
+      "learning_rate": 2.3116260684816446e-05,
+      "loss": 5.4961,
+      "step": 10884
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6689753532409668,
+      "learning_rate": 2.3106378773654828e-05,
+      "loss": 5.5967,
+      "step": 10888
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8034151792526245,
+      "learning_rate": 2.3096496862493207e-05,
+      "loss": 5.5646,
+      "step": 10892
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8747068643569946,
+      "learning_rate": 2.308661495133159e-05,
+      "loss": 5.599,
+      "step": 10896
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.044126033782959,
+      "learning_rate": 2.307673304016997e-05,
+      "loss": 5.4739,
+      "step": 10900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0283472537994385,
+      "learning_rate": 2.3066851129008353e-05,
+      "loss": 5.5447,
+      "step": 10904
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0088658332824707,
+      "learning_rate": 2.305696921784673e-05,
+      "loss": 5.6111,
+      "step": 10908
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8219811916351318,
+      "learning_rate": 2.3047087306685114e-05,
+      "loss": 5.6571,
+      "step": 10912
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.010409116744995,
+      "learning_rate": 2.3037205395523496e-05,
+      "loss": 5.5405,
+      "step": 10916
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03324031829834,
+      "learning_rate": 2.3027323484361878e-05,
+      "loss": 5.633,
+      "step": 10920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9077966213226318,
+      "learning_rate": 2.3017441573200256e-05,
+      "loss": 5.5292,
+      "step": 10924
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7984883785247803,
+      "learning_rate": 2.300755966203864e-05,
+      "loss": 5.5423,
+      "step": 10928
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.038520336151123,
+      "learning_rate": 2.299767775087702e-05,
+      "loss": 5.5662,
+      "step": 10932
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0444748401641846,
+      "learning_rate": 2.2987795839715403e-05,
+      "loss": 5.5273,
+      "step": 10936
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0163471698760986,
+      "learning_rate": 2.2977913928553785e-05,
+      "loss": 5.5836,
+      "step": 10940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.246277093887329,
+      "learning_rate": 2.2968032017392167e-05,
+      "loss": 5.717,
+      "step": 10944
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.040771484375,
+      "learning_rate": 2.295815010623055e-05,
+      "loss": 5.5738,
+      "step": 10948
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9808095693588257,
+      "learning_rate": 2.2948268195068927e-05,
+      "loss": 5.6387,
+      "step": 10952
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1175899505615234,
+      "learning_rate": 2.293838628390731e-05,
+      "loss": 5.5871,
+      "step": 10956
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.99501633644104,
+      "learning_rate": 2.292850437274569e-05,
+      "loss": 5.5695,
+      "step": 10960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9630918502807617,
+      "learning_rate": 2.291862246158407e-05,
+      "loss": 5.5771,
+      "step": 10964
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.638632893562317,
+      "learning_rate": 2.2908740550422452e-05,
+      "loss": 5.551,
+      "step": 10968
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9305760860443115,
+      "learning_rate": 2.2898858639260834e-05,
+      "loss": 5.5803,
+      "step": 10972
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7646735906600952,
+      "learning_rate": 2.2888976728099216e-05,
+      "loss": 5.5042,
+      "step": 10976
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2676877975463867,
+      "learning_rate": 2.2879094816937595e-05,
+      "loss": 5.5913,
+      "step": 10980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.974547028541565,
+      "learning_rate": 2.2869212905775977e-05,
+      "loss": 5.5148,
+      "step": 10984
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.830438494682312,
+      "learning_rate": 2.285933099461436e-05,
+      "loss": 5.5472,
+      "step": 10988
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.166577100753784,
+      "learning_rate": 2.284944908345274e-05,
+      "loss": 5.4449,
+      "step": 10992
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0472869873046875,
+      "learning_rate": 2.283956717229112e-05,
+      "loss": 5.6287,
+      "step": 10996
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8216519355773926,
+      "learning_rate": 2.2829685261129505e-05,
+      "loss": 5.5427,
+      "step": 11000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0221810340881348,
+      "learning_rate": 2.2819803349967887e-05,
+      "loss": 5.6037,
+      "step": 11004
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8711715936660767,
+      "learning_rate": 2.2809921438806266e-05,
+      "loss": 5.5891,
+      "step": 11008
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8653993606567383,
+      "learning_rate": 2.2800039527644648e-05,
+      "loss": 5.4277,
+      "step": 11012
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9809545278549194,
+      "learning_rate": 2.279015761648303e-05,
+      "loss": 5.4548,
+      "step": 11016
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9581102132797241,
+      "learning_rate": 2.278027570532141e-05,
+      "loss": 5.5037,
+      "step": 11020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.251185417175293,
+      "learning_rate": 2.277039379415979e-05,
+      "loss": 5.5988,
+      "step": 11024
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7871466875076294,
+      "learning_rate": 2.2760511882998173e-05,
+      "loss": 5.3987,
+      "step": 11028
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8406583070755005,
+      "learning_rate": 2.2750629971836555e-05,
+      "loss": 5.6034,
+      "step": 11032
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1242032051086426,
+      "learning_rate": 2.2740748060674934e-05,
+      "loss": 5.512,
+      "step": 11036
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.062807083129883,
+      "learning_rate": 2.2730866149513316e-05,
+      "loss": 5.5253,
+      "step": 11040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1405367851257324,
+      "learning_rate": 2.2720984238351698e-05,
+      "loss": 5.579,
+      "step": 11044
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.989423155784607,
+      "learning_rate": 2.271110232719008e-05,
+      "loss": 5.5023,
+      "step": 11048
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7983574867248535,
+      "learning_rate": 2.270122041602846e-05,
+      "loss": 5.6174,
+      "step": 11052
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9673739671707153,
+      "learning_rate": 2.2691338504866844e-05,
+      "loss": 5.4921,
+      "step": 11056
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0545785427093506,
+      "learning_rate": 2.2681456593705226e-05,
+      "loss": 5.5915,
+      "step": 11060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.947089433670044,
+      "learning_rate": 2.2671574682543605e-05,
+      "loss": 5.6147,
+      "step": 11064
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6919525861740112,
+      "learning_rate": 2.2661692771381987e-05,
+      "loss": 5.5268,
+      "step": 11068
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9140796661376953,
+      "learning_rate": 2.265181086022037e-05,
+      "loss": 5.6551,
+      "step": 11072
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.034980297088623,
+      "learning_rate": 2.264192894905875e-05,
+      "loss": 5.5495,
+      "step": 11076
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2609176635742188,
+      "learning_rate": 2.263204703789713e-05,
+      "loss": 5.52,
+      "step": 11080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.154770612716675,
+      "learning_rate": 2.262216512673551e-05,
+      "loss": 5.5406,
+      "step": 11084
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8771696090698242,
+      "learning_rate": 2.2612283215573894e-05,
+      "loss": 5.4876,
+      "step": 11088
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8587368726730347,
+      "learning_rate": 2.2602401304412272e-05,
+      "loss": 5.6025,
+      "step": 11092
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8150922060012817,
+      "learning_rate": 2.2592519393250654e-05,
+      "loss": 5.4589,
+      "step": 11096
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.94694185256958,
+      "learning_rate": 2.2582637482089037e-05,
+      "loss": 5.4608,
+      "step": 11100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.919480800628662,
+      "learning_rate": 2.257275557092742e-05,
+      "loss": 5.4808,
+      "step": 11104
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9377970695495605,
+      "learning_rate": 2.2562873659765797e-05,
+      "loss": 5.5321,
+      "step": 11108
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8229963779449463,
+      "learning_rate": 2.2552991748604183e-05,
+      "loss": 5.5011,
+      "step": 11112
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7934753894805908,
+      "learning_rate": 2.2543109837442565e-05,
+      "loss": 5.4912,
+      "step": 11116
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.294386386871338,
+      "learning_rate": 2.2533227926280943e-05,
+      "loss": 5.5593,
+      "step": 11120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9534809589385986,
+      "learning_rate": 2.2523346015119325e-05,
+      "loss": 5.5784,
+      "step": 11124
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.121579885482788,
+      "learning_rate": 2.2513464103957708e-05,
+      "loss": 5.6106,
+      "step": 11128
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8341236114501953,
+      "learning_rate": 2.250358219279609e-05,
+      "loss": 5.5102,
+      "step": 11132
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8656764030456543,
+      "learning_rate": 2.2493700281634468e-05,
+      "loss": 5.5725,
+      "step": 11136
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.026597738265991,
+      "learning_rate": 2.248381837047285e-05,
+      "loss": 5.5864,
+      "step": 11140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7924002408981323,
+      "learning_rate": 2.2473936459311232e-05,
+      "loss": 5.5194,
+      "step": 11144
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9699978828430176,
+      "learning_rate": 2.2464054548149614e-05,
+      "loss": 5.5491,
+      "step": 11148
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0057380199432373,
+      "learning_rate": 2.2454172636987993e-05,
+      "loss": 5.5209,
+      "step": 11152
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9208364486694336,
+      "learning_rate": 2.2444290725826375e-05,
+      "loss": 5.6376,
+      "step": 11156
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1530115604400635,
+      "learning_rate": 2.2434408814664757e-05,
+      "loss": 5.5286,
+      "step": 11160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0784969329833984,
+      "learning_rate": 2.2424526903503136e-05,
+      "loss": 5.629,
+      "step": 11164
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7239807844161987,
+      "learning_rate": 2.2414644992341518e-05,
+      "loss": 5.4354,
+      "step": 11168
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1020421981811523,
+      "learning_rate": 2.2404763081179903e-05,
+      "loss": 5.4923,
+      "step": 11172
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8407636880874634,
+      "learning_rate": 2.2394881170018282e-05,
+      "loss": 5.5506,
+      "step": 11176
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8488260507583618,
+      "learning_rate": 2.2384999258856664e-05,
+      "loss": 5.6145,
+      "step": 11180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0274198055267334,
+      "learning_rate": 2.2375117347695046e-05,
+      "loss": 5.5115,
+      "step": 11184
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.122840642929077,
+      "learning_rate": 2.2365235436533428e-05,
+      "loss": 5.4538,
+      "step": 11188
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.121507406234741,
+      "learning_rate": 2.2355353525371807e-05,
+      "loss": 5.4959,
+      "step": 11192
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1240291595458984,
+      "learning_rate": 2.234547161421019e-05,
+      "loss": 5.3912,
+      "step": 11196
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7324938774108887,
+      "learning_rate": 2.233558970304857e-05,
+      "loss": 5.5702,
+      "step": 11200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7895362377166748,
+      "learning_rate": 2.2325707791886953e-05,
+      "loss": 5.5114,
+      "step": 11204
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8272205591201782,
+      "learning_rate": 2.2315825880725332e-05,
+      "loss": 5.5404,
+      "step": 11208
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2934823036193848,
+      "learning_rate": 2.2305943969563714e-05,
+      "loss": 5.6429,
+      "step": 11212
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1152985095977783,
+      "learning_rate": 2.2296062058402096e-05,
+      "loss": 5.5438,
+      "step": 11216
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9162452220916748,
+      "learning_rate": 2.2286180147240478e-05,
+      "loss": 5.574,
+      "step": 11220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9226047992706299,
+      "learning_rate": 2.2276298236078857e-05,
+      "loss": 5.5964,
+      "step": 11224
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2039763927459717,
+      "learning_rate": 2.2266416324917242e-05,
+      "loss": 5.5645,
+      "step": 11228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8182283639907837,
+      "learning_rate": 2.2256534413755624e-05,
+      "loss": 5.5464,
+      "step": 11232
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8267743587493896,
+      "learning_rate": 2.2246652502594003e-05,
+      "loss": 5.5826,
+      "step": 11236
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.933274269104004,
+      "learning_rate": 2.2236770591432385e-05,
+      "loss": 5.5138,
+      "step": 11240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.160489320755005,
+      "learning_rate": 2.2226888680270767e-05,
+      "loss": 5.5879,
+      "step": 11244
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8666636943817139,
+      "learning_rate": 2.2217006769109146e-05,
+      "loss": 5.5751,
+      "step": 11248
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0974009037017822,
+      "learning_rate": 2.2207124857947528e-05,
+      "loss": 5.4429,
+      "step": 11252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6826304197311401,
+      "learning_rate": 2.219724294678591e-05,
+      "loss": 5.5061,
+      "step": 11256
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9517406225204468,
+      "learning_rate": 2.2187361035624292e-05,
+      "loss": 5.5793,
+      "step": 11260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9135072231292725,
+      "learning_rate": 2.217747912446267e-05,
+      "loss": 5.4732,
+      "step": 11264
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7441056966781616,
+      "learning_rate": 2.2167597213301053e-05,
+      "loss": 5.4808,
+      "step": 11268
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.851120114326477,
+      "learning_rate": 2.2157715302139435e-05,
+      "loss": 5.6432,
+      "step": 11272
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8004862070083618,
+      "learning_rate": 2.2147833390977817e-05,
+      "loss": 5.5154,
+      "step": 11276
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6254485845565796,
+      "learning_rate": 2.2137951479816195e-05,
+      "loss": 5.6375,
+      "step": 11280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9577946662902832,
+      "learning_rate": 2.2128069568654577e-05,
+      "loss": 5.559,
+      "step": 11284
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.018347978591919,
+      "learning_rate": 2.2118187657492963e-05,
+      "loss": 5.484,
+      "step": 11288
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8874802589416504,
+      "learning_rate": 2.210830574633134e-05,
+      "loss": 5.577,
+      "step": 11292
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.764642357826233,
+      "learning_rate": 2.2098423835169724e-05,
+      "loss": 5.45,
+      "step": 11296
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.156693458557129,
+      "learning_rate": 2.2088541924008106e-05,
+      "loss": 5.5669,
+      "step": 11300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8049863576889038,
+      "learning_rate": 2.2078660012846484e-05,
+      "loss": 5.5686,
+      "step": 11304
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7862818241119385,
+      "learning_rate": 2.2068778101684866e-05,
+      "loss": 5.4679,
+      "step": 11308
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8136225938796997,
+      "learning_rate": 2.205889619052325e-05,
+      "loss": 5.5579,
+      "step": 11312
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.744735598564148,
+      "learning_rate": 2.204901427936163e-05,
+      "loss": 5.5756,
+      "step": 11316
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0922114849090576,
+      "learning_rate": 2.203913236820001e-05,
+      "loss": 5.3664,
+      "step": 11320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8184114694595337,
+      "learning_rate": 2.202925045703839e-05,
+      "loss": 5.3773,
+      "step": 11324
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.75265371799469,
+      "learning_rate": 2.2019368545876773e-05,
+      "loss": 5.5344,
+      "step": 11328
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.045360803604126,
+      "learning_rate": 2.2009486634715155e-05,
+      "loss": 5.5896,
+      "step": 11332
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7742873430252075,
+      "learning_rate": 2.1999604723553534e-05,
+      "loss": 5.5624,
+      "step": 11336
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0210537910461426,
+      "learning_rate": 2.1989722812391916e-05,
+      "loss": 5.6381,
+      "step": 11340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8388046026229858,
+      "learning_rate": 2.19798409012303e-05,
+      "loss": 5.5318,
+      "step": 11344
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9268219470977783,
+      "learning_rate": 2.196995899006868e-05,
+      "loss": 5.5328,
+      "step": 11348
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0054502487182617,
+      "learning_rate": 2.1960077078907062e-05,
+      "loss": 5.5936,
+      "step": 11352
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.741058349609375,
+      "learning_rate": 2.1950195167745444e-05,
+      "loss": 5.5389,
+      "step": 11356
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9942952394485474,
+      "learning_rate": 2.1940313256583826e-05,
+      "loss": 5.5587,
+      "step": 11360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9131739139556885,
+      "learning_rate": 2.1930431345422205e-05,
+      "loss": 5.417,
+      "step": 11364
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7850251197814941,
+      "learning_rate": 2.1920549434260587e-05,
+      "loss": 5.5102,
+      "step": 11368
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.221616744995117,
+      "learning_rate": 2.191066752309897e-05,
+      "loss": 5.5262,
+      "step": 11372
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.910980463027954,
+      "learning_rate": 2.1900785611937348e-05,
+      "loss": 5.7343,
+      "step": 11376
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8210395574569702,
+      "learning_rate": 2.189090370077573e-05,
+      "loss": 5.5456,
+      "step": 11380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9524825811386108,
+      "learning_rate": 2.1881021789614112e-05,
+      "loss": 5.5558,
+      "step": 11384
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5711989402770996,
+      "learning_rate": 2.1871139878452494e-05,
+      "loss": 5.4329,
+      "step": 11388
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0921053886413574,
+      "learning_rate": 2.1861257967290873e-05,
+      "loss": 5.6219,
+      "step": 11392
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.04020357131958,
+      "learning_rate": 2.1851376056129255e-05,
+      "loss": 5.4476,
+      "step": 11396
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7985339164733887,
+      "learning_rate": 2.184149414496764e-05,
+      "loss": 5.5278,
+      "step": 11400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8650219440460205,
+      "learning_rate": 2.183161223380602e-05,
+      "loss": 5.5657,
+      "step": 11404
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8160661458969116,
+      "learning_rate": 2.18217303226444e-05,
+      "loss": 5.525,
+      "step": 11408
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0409529209136963,
+      "learning_rate": 2.1811848411482783e-05,
+      "loss": 5.4098,
+      "step": 11412
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.731351375579834,
+      "learning_rate": 2.1801966500321165e-05,
+      "loss": 5.588,
+      "step": 11416
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9625564813613892,
+      "learning_rate": 2.1792084589159544e-05,
+      "loss": 5.4872,
+      "step": 11420
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0399959087371826,
+      "learning_rate": 2.1782202677997926e-05,
+      "loss": 5.5627,
+      "step": 11424
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.684001088142395,
+      "learning_rate": 2.1772320766836308e-05,
+      "loss": 5.6156,
+      "step": 11428
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7688792943954468,
+      "learning_rate": 2.176243885567469e-05,
+      "loss": 5.4949,
+      "step": 11432
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.069566011428833,
+      "learning_rate": 2.175255694451307e-05,
+      "loss": 5.6452,
+      "step": 11436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.030831813812256,
+      "learning_rate": 2.174267503335145e-05,
+      "loss": 5.3957,
+      "step": 11440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9303979873657227,
+      "learning_rate": 2.1732793122189833e-05,
+      "loss": 5.4431,
+      "step": 11444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6101690530776978,
+      "learning_rate": 2.172291121102821e-05,
+      "loss": 5.5681,
+      "step": 11448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9189858436584473,
+      "learning_rate": 2.1713029299866593e-05,
+      "loss": 5.4847,
+      "step": 11452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09655499458313,
+      "learning_rate": 2.1703147388704976e-05,
+      "loss": 5.5774,
+      "step": 11456
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.018876791000366,
+      "learning_rate": 2.1693265477543358e-05,
+      "loss": 5.614,
+      "step": 11460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.754034161567688,
+      "learning_rate": 2.168338356638174e-05,
+      "loss": 5.443,
+      "step": 11464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.950484275817871,
+      "learning_rate": 2.1673501655220122e-05,
+      "loss": 5.4982,
+      "step": 11468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8960100412368774,
+      "learning_rate": 2.1663619744058504e-05,
+      "loss": 5.478,
+      "step": 11472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7579559087753296,
+      "learning_rate": 2.1653737832896882e-05,
+      "loss": 5.5021,
+      "step": 11476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0521066188812256,
+      "learning_rate": 2.1643855921735264e-05,
+      "loss": 5.5329,
+      "step": 11480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.999937891960144,
+      "learning_rate": 2.1633974010573647e-05,
+      "loss": 5.4377,
+      "step": 11484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.099349021911621,
+      "learning_rate": 2.162409209941203e-05,
+      "loss": 5.562,
+      "step": 11488
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0938682556152344,
+      "learning_rate": 2.1614210188250407e-05,
+      "loss": 5.577,
+      "step": 11492
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6756892204284668,
+      "learning_rate": 2.160432827708879e-05,
+      "loss": 5.6301,
+      "step": 11496
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7941629886627197,
+      "learning_rate": 2.159444636592717e-05,
+      "loss": 5.5144,
+      "step": 11500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9828251600265503,
+      "learning_rate": 2.1584564454765553e-05,
+      "loss": 5.674,
+      "step": 11504
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8105112314224243,
+      "learning_rate": 2.1574682543603932e-05,
+      "loss": 5.5155,
+      "step": 11508
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7346547842025757,
+      "learning_rate": 2.1564800632442314e-05,
+      "loss": 5.6869,
+      "step": 11512
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7930268049240112,
+      "learning_rate": 2.15549187212807e-05,
+      "loss": 5.4631,
+      "step": 11516
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.799404263496399,
+      "learning_rate": 2.154503681011908e-05,
+      "loss": 5.5324,
+      "step": 11520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.948190450668335,
+      "learning_rate": 2.153515489895746e-05,
+      "loss": 5.6236,
+      "step": 11524
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0511763095855713,
+      "learning_rate": 2.1525272987795842e-05,
+      "loss": 5.6085,
+      "step": 11528
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1785106658935547,
+      "learning_rate": 2.151539107663422e-05,
+      "loss": 5.4754,
+      "step": 11532
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9326204061508179,
+      "learning_rate": 2.1505509165472603e-05,
+      "loss": 5.5052,
+      "step": 11536
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9747314453125,
+      "learning_rate": 2.1495627254310985e-05,
+      "loss": 5.4398,
+      "step": 11540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.743662714958191,
+      "learning_rate": 2.1485745343149367e-05,
+      "loss": 5.4521,
+      "step": 11544
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.823280930519104,
+      "learning_rate": 2.1475863431987746e-05,
+      "loss": 5.4957,
+      "step": 11548
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.047112464904785,
+      "learning_rate": 2.1465981520826128e-05,
+      "loss": 5.584,
+      "step": 11552
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7637749910354614,
+      "learning_rate": 2.145609960966451e-05,
+      "loss": 5.6439,
+      "step": 11556
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7925045490264893,
+      "learning_rate": 2.1446217698502892e-05,
+      "loss": 5.4745,
+      "step": 11560
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8895896673202515,
+      "learning_rate": 2.143633578734127e-05,
+      "loss": 5.4082,
+      "step": 11564
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.956284761428833,
+      "learning_rate": 2.1426453876179653e-05,
+      "loss": 5.3854,
+      "step": 11568
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9955167770385742,
+      "learning_rate": 2.141657196501804e-05,
+      "loss": 5.4486,
+      "step": 11572
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0636212825775146,
+      "learning_rate": 2.1406690053856417e-05,
+      "loss": 5.6201,
+      "step": 11576
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0052077770233154,
+      "learning_rate": 2.13968081426948e-05,
+      "loss": 5.5525,
+      "step": 11580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9207104444503784,
+      "learning_rate": 2.138692623153318e-05,
+      "loss": 5.6011,
+      "step": 11584
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9215471744537354,
+      "learning_rate": 2.137704432037156e-05,
+      "loss": 5.556,
+      "step": 11588
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6822532415390015,
+      "learning_rate": 2.1367162409209942e-05,
+      "loss": 5.521,
+      "step": 11592
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1839683055877686,
+      "learning_rate": 2.1357280498048324e-05,
+      "loss": 5.6367,
+      "step": 11596
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.855774998664856,
+      "learning_rate": 2.1347398586886706e-05,
+      "loss": 5.518,
+      "step": 11600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0112996101379395,
+      "learning_rate": 2.1337516675725085e-05,
+      "loss": 5.6125,
+      "step": 11604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0093090534210205,
+      "learning_rate": 2.1327634764563467e-05,
+      "loss": 5.6358,
+      "step": 11608
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6822510957717896,
+      "learning_rate": 2.131775285340185e-05,
+      "loss": 5.5177,
+      "step": 11612
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9259730577468872,
+      "learning_rate": 2.130787094224023e-05,
+      "loss": 5.5029,
+      "step": 11616
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9150243997573853,
+      "learning_rate": 2.129798903107861e-05,
+      "loss": 5.5636,
+      "step": 11620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8998494148254395,
+      "learning_rate": 2.128810711991699e-05,
+      "loss": 5.5628,
+      "step": 11624
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9951376914978027,
+      "learning_rate": 2.1278225208755374e-05,
+      "loss": 5.5671,
+      "step": 11628
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7849557399749756,
+      "learning_rate": 2.1268343297593756e-05,
+      "loss": 5.5127,
+      "step": 11632
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0039825439453125,
+      "learning_rate": 2.1258461386432138e-05,
+      "loss": 5.5114,
+      "step": 11636
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.007443904876709,
+      "learning_rate": 2.124857947527052e-05,
+      "loss": 5.5814,
+      "step": 11640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7764946222305298,
+      "learning_rate": 2.1238697564108902e-05,
+      "loss": 5.5585,
+      "step": 11644
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.218045473098755,
+      "learning_rate": 2.122881565294728e-05,
+      "loss": 5.4461,
+      "step": 11648
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.088454484939575,
+      "learning_rate": 2.1218933741785663e-05,
+      "loss": 5.492,
+      "step": 11652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8053348064422607,
+      "learning_rate": 2.1209051830624045e-05,
+      "loss": 5.5221,
+      "step": 11656
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.035414218902588,
+      "learning_rate": 2.1199169919462423e-05,
+      "loss": 5.6862,
+      "step": 11660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0980472564697266,
+      "learning_rate": 2.1189288008300805e-05,
+      "loss": 5.6149,
+      "step": 11664
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.033268690109253,
+      "learning_rate": 2.1179406097139187e-05,
+      "loss": 5.4371,
+      "step": 11668
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8143877983093262,
+      "learning_rate": 2.116952418597757e-05,
+      "loss": 5.5271,
+      "step": 11672
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7776703834533691,
+      "learning_rate": 2.1159642274815948e-05,
+      "loss": 5.6312,
+      "step": 11676
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8278911113739014,
+      "learning_rate": 2.114976036365433e-05,
+      "loss": 5.5128,
+      "step": 11680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8950855731964111,
+      "learning_rate": 2.1139878452492712e-05,
+      "loss": 5.4134,
+      "step": 11684
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.828503966331482,
+      "learning_rate": 2.1129996541331094e-05,
+      "loss": 5.5917,
+      "step": 11688
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9912033081054688,
+      "learning_rate": 2.1120114630169476e-05,
+      "loss": 5.4005,
+      "step": 11692
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.871800422668457,
+      "learning_rate": 2.111023271900786e-05,
+      "loss": 5.5202,
+      "step": 11696
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8771202564239502,
+      "learning_rate": 2.110035080784624e-05,
+      "loss": 5.5294,
+      "step": 11700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8930245637893677,
+      "learning_rate": 2.109046889668462e-05,
+      "loss": 5.482,
+      "step": 11704
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.910056471824646,
+      "learning_rate": 2.1080586985523e-05,
+      "loss": 5.625,
+      "step": 11708
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7149245738983154,
+      "learning_rate": 2.1070705074361383e-05,
+      "loss": 5.494,
+      "step": 11712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8306865692138672,
+      "learning_rate": 2.1060823163199765e-05,
+      "loss": 5.5143,
+      "step": 11716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9364176988601685,
+      "learning_rate": 2.1050941252038144e-05,
+      "loss": 5.5469,
+      "step": 11720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9254311323165894,
+      "learning_rate": 2.1041059340876526e-05,
+      "loss": 5.6064,
+      "step": 11724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8167616128921509,
+      "learning_rate": 2.1031177429714908e-05,
+      "loss": 5.6055,
+      "step": 11728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9327070713043213,
+      "learning_rate": 2.1021295518553287e-05,
+      "loss": 5.6255,
+      "step": 11732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0709402561187744,
+      "learning_rate": 2.101141360739167e-05,
+      "loss": 5.5823,
+      "step": 11736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6243914365768433,
+      "learning_rate": 2.100153169623005e-05,
+      "loss": 5.5357,
+      "step": 11740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0544323921203613,
+      "learning_rate": 2.0991649785068433e-05,
+      "loss": 5.4751,
+      "step": 11744
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8204375505447388,
+      "learning_rate": 2.0981767873906815e-05,
+      "loss": 5.4421,
+      "step": 11748
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8007164001464844,
+      "learning_rate": 2.0971885962745197e-05,
+      "loss": 5.5578,
+      "step": 11752
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0843026638031006,
+      "learning_rate": 2.096200405158358e-05,
+      "loss": 5.5082,
+      "step": 11756
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7756987810134888,
+      "learning_rate": 2.0952122140421958e-05,
+      "loss": 5.4814,
+      "step": 11760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9955602884292603,
+      "learning_rate": 2.094224022926034e-05,
+      "loss": 5.6497,
+      "step": 11764
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.000737190246582,
+      "learning_rate": 2.0932358318098722e-05,
+      "loss": 5.5241,
+      "step": 11768
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7721540927886963,
+      "learning_rate": 2.0922476406937104e-05,
+      "loss": 5.4961,
+      "step": 11772
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9845120906829834,
+      "learning_rate": 2.0912594495775483e-05,
+      "loss": 5.6158,
+      "step": 11776
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7041538953781128,
+      "learning_rate": 2.0902712584613865e-05,
+      "loss": 5.4892,
+      "step": 11780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8320084810256958,
+      "learning_rate": 2.0892830673452247e-05,
+      "loss": 5.6435,
+      "step": 11784
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7904802560806274,
+      "learning_rate": 2.088294876229063e-05,
+      "loss": 5.4774,
+      "step": 11788
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.894283413887024,
+      "learning_rate": 2.0873066851129008e-05,
+      "loss": 5.4308,
+      "step": 11792
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8176664113998413,
+      "learning_rate": 2.086318493996739e-05,
+      "loss": 5.4719,
+      "step": 11796
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9829813241958618,
+      "learning_rate": 2.0853303028805772e-05,
+      "loss": 5.4217,
+      "step": 11800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8619272708892822,
+      "learning_rate": 2.0843421117644154e-05,
+      "loss": 5.5205,
+      "step": 11804
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8735342025756836,
+      "learning_rate": 2.0833539206482536e-05,
+      "loss": 5.5359,
+      "step": 11808
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.860768437385559,
+      "learning_rate": 2.0823657295320918e-05,
+      "loss": 5.5111,
+      "step": 11812
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0333621501922607,
+      "learning_rate": 2.0813775384159297e-05,
+      "loss": 5.5294,
+      "step": 11816
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9842480421066284,
+      "learning_rate": 2.080389347299768e-05,
+      "loss": 5.521,
+      "step": 11820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7381541728973389,
+      "learning_rate": 2.079401156183606e-05,
+      "loss": 5.5587,
+      "step": 11824
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6994056701660156,
+      "learning_rate": 2.0784129650674443e-05,
+      "loss": 5.5389,
+      "step": 11828
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7459512948989868,
+      "learning_rate": 2.077424773951282e-05,
+      "loss": 5.6567,
+      "step": 11832
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9870339632034302,
+      "learning_rate": 2.0764365828351203e-05,
+      "loss": 5.5713,
+      "step": 11836
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1633095741271973,
+      "learning_rate": 2.0754483917189586e-05,
+      "loss": 5.4663,
+      "step": 11840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8806403875350952,
+      "learning_rate": 2.0744602006027968e-05,
+      "loss": 5.489,
+      "step": 11844
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.007995367050171,
+      "learning_rate": 2.0734720094866346e-05,
+      "loss": 5.669,
+      "step": 11848
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.024402141571045,
+      "learning_rate": 2.072483818370473e-05,
+      "loss": 5.5401,
+      "step": 11852
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9132201671600342,
+      "learning_rate": 2.071495627254311e-05,
+      "loss": 5.6524,
+      "step": 11856
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0311427116394043,
+      "learning_rate": 2.0705074361381492e-05,
+      "loss": 5.5418,
+      "step": 11860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0080606937408447,
+      "learning_rate": 2.0695192450219875e-05,
+      "loss": 5.5101,
+      "step": 11864
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.698283076286316,
+      "learning_rate": 2.0685310539058257e-05,
+      "loss": 5.6118,
+      "step": 11868
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9465456008911133,
+      "learning_rate": 2.0675428627896635e-05,
+      "loss": 5.545,
+      "step": 11872
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7435317039489746,
+      "learning_rate": 2.0665546716735017e-05,
+      "loss": 5.4387,
+      "step": 11876
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.106904983520508,
+      "learning_rate": 2.06556648055734e-05,
+      "loss": 5.6137,
+      "step": 11880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1040077209472656,
+      "learning_rate": 2.064578289441178e-05,
+      "loss": 5.5537,
+      "step": 11884
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9074550867080688,
+      "learning_rate": 2.063590098325016e-05,
+      "loss": 5.5837,
+      "step": 11888
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0350706577301025,
+      "learning_rate": 2.0626019072088542e-05,
+      "loss": 5.6268,
+      "step": 11892
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2680063247680664,
+      "learning_rate": 2.0616137160926924e-05,
+      "loss": 5.4717,
+      "step": 11896
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8204529285430908,
+      "learning_rate": 2.0606255249765306e-05,
+      "loss": 5.5205,
+      "step": 11900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.01955246925354,
+      "learning_rate": 2.0596373338603685e-05,
+      "loss": 5.4586,
+      "step": 11904
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.047470808029175,
+      "learning_rate": 2.0586491427442067e-05,
+      "loss": 5.5496,
+      "step": 11908
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9742119312286377,
+      "learning_rate": 2.057660951628045e-05,
+      "loss": 5.5521,
+      "step": 11912
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9781845808029175,
+      "learning_rate": 2.056672760511883e-05,
+      "loss": 5.4699,
+      "step": 11916
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7147916555404663,
+      "learning_rate": 2.0556845693957213e-05,
+      "loss": 5.5772,
+      "step": 11920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8772165775299072,
+      "learning_rate": 2.0546963782795595e-05,
+      "loss": 5.4976,
+      "step": 11924
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9535537958145142,
+      "learning_rate": 2.0537081871633977e-05,
+      "loss": 5.4462,
+      "step": 11928
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8837189674377441,
+      "learning_rate": 2.0527199960472356e-05,
+      "loss": 5.5724,
+      "step": 11932
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9887135028839111,
+      "learning_rate": 2.0517318049310738e-05,
+      "loss": 5.6435,
+      "step": 11936
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9106159210205078,
+      "learning_rate": 2.050743613814912e-05,
+      "loss": 5.6322,
+      "step": 11940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9000262022018433,
+      "learning_rate": 2.04975542269875e-05,
+      "loss": 5.5926,
+      "step": 11944
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8226341009140015,
+      "learning_rate": 2.048767231582588e-05,
+      "loss": 5.5135,
+      "step": 11948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8902435302734375,
+      "learning_rate": 2.0477790404664263e-05,
+      "loss": 5.4756,
+      "step": 11952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.058504819869995,
+      "learning_rate": 2.0467908493502645e-05,
+      "loss": 5.6203,
+      "step": 11956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.977247714996338,
+      "learning_rate": 2.0458026582341024e-05,
+      "loss": 5.5297,
+      "step": 11960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6700700521469116,
+      "learning_rate": 2.0448144671179406e-05,
+      "loss": 5.5585,
+      "step": 11964
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7290418148040771,
+      "learning_rate": 2.0438262760017788e-05,
+      "loss": 5.498,
+      "step": 11968
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7892743349075317,
+      "learning_rate": 2.042838084885617e-05,
+      "loss": 5.4366,
+      "step": 11972
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1159372329711914,
+      "learning_rate": 2.0418498937694552e-05,
+      "loss": 5.5055,
+      "step": 11976
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0965609550476074,
+      "learning_rate": 2.0408617026532934e-05,
+      "loss": 5.6062,
+      "step": 11980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8796151876449585,
+      "learning_rate": 2.0398735115371316e-05,
+      "loss": 5.5831,
+      "step": 11984
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8930494785308838,
+      "learning_rate": 2.0388853204209695e-05,
+      "loss": 5.5128,
+      "step": 11988
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1627285480499268,
+      "learning_rate": 2.0378971293048077e-05,
+      "loss": 5.5642,
+      "step": 11992
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7648162841796875,
+      "learning_rate": 2.036908938188646e-05,
+      "loss": 5.4877,
+      "step": 11996
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2223527431488037,
+      "learning_rate": 2.035920747072484e-05,
+      "loss": 5.5078,
+      "step": 12000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.301929473876953,
+      "learning_rate": 2.034932555956322e-05,
+      "loss": 5.5219,
+      "step": 12004
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9763400554656982,
+      "learning_rate": 2.03394436484016e-05,
+      "loss": 5.5175,
+      "step": 12008
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0059380531311035,
+      "learning_rate": 2.0329561737239984e-05,
+      "loss": 5.5533,
+      "step": 12012
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3033435344696045,
+      "learning_rate": 2.0319679826078362e-05,
+      "loss": 5.5641,
+      "step": 12016
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9601203203201294,
+      "learning_rate": 2.0309797914916744e-05,
+      "loss": 5.4841,
+      "step": 12020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.893579363822937,
+      "learning_rate": 2.0299916003755126e-05,
+      "loss": 5.418,
+      "step": 12024
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.857035517692566,
+      "learning_rate": 2.029003409259351e-05,
+      "loss": 5.5092,
+      "step": 12028
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.068701982498169,
+      "learning_rate": 2.028015218143189e-05,
+      "loss": 5.4992,
+      "step": 12032
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8823144435882568,
+      "learning_rate": 2.0270270270270273e-05,
+      "loss": 5.4709,
+      "step": 12036
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.873325228691101,
+      "learning_rate": 2.0260388359108655e-05,
+      "loss": 5.4815,
+      "step": 12040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.88863205909729,
+      "learning_rate": 2.0250506447947033e-05,
+      "loss": 5.6217,
+      "step": 12044
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.813723087310791,
+      "learning_rate": 2.0240624536785415e-05,
+      "loss": 5.5736,
+      "step": 12048
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8674362897872925,
+      "learning_rate": 2.0230742625623797e-05,
+      "loss": 5.4345,
+      "step": 12052
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.840463399887085,
+      "learning_rate": 2.022086071446218e-05,
+      "loss": 5.4454,
+      "step": 12056
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8874510526657104,
+      "learning_rate": 2.0210978803300558e-05,
+      "loss": 5.4834,
+      "step": 12060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8396440744400024,
+      "learning_rate": 2.020109689213894e-05,
+      "loss": 5.5029,
+      "step": 12064
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.950766921043396,
+      "learning_rate": 2.0191214980977322e-05,
+      "loss": 5.4637,
+      "step": 12068
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1061501502990723,
+      "learning_rate": 2.0181333069815704e-05,
+      "loss": 5.4081,
+      "step": 12072
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9160481691360474,
+      "learning_rate": 2.0171451158654083e-05,
+      "loss": 5.484,
+      "step": 12076
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.151904344558716,
+      "learning_rate": 2.0161569247492465e-05,
+      "loss": 5.5189,
+      "step": 12080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8847110271453857,
+      "learning_rate": 2.0151687336330847e-05,
+      "loss": 5.4264,
+      "step": 12084
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.914305329322815,
+      "learning_rate": 2.014180542516923e-05,
+      "loss": 5.551,
+      "step": 12088
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.034773111343384,
+      "learning_rate": 2.013192351400761e-05,
+      "loss": 5.5728,
+      "step": 12092
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7983940839767456,
+      "learning_rate": 2.0122041602845993e-05,
+      "loss": 5.4991,
+      "step": 12096
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.019416570663452,
+      "learning_rate": 2.0112159691684372e-05,
+      "loss": 5.6503,
+      "step": 12100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9403501749038696,
+      "learning_rate": 2.0102277780522754e-05,
+      "loss": 5.5295,
+      "step": 12104
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.006972074508667,
+      "learning_rate": 2.0092395869361136e-05,
+      "loss": 5.4612,
+      "step": 12108
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0223989486694336,
+      "learning_rate": 2.0082513958199518e-05,
+      "loss": 5.4774,
+      "step": 12112
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7617640495300293,
+      "learning_rate": 2.0072632047037897e-05,
+      "loss": 5.5604,
+      "step": 12116
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.961089849472046,
+      "learning_rate": 2.006275013587628e-05,
+      "loss": 5.4945,
+      "step": 12120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.054691791534424,
+      "learning_rate": 2.005286822471466e-05,
+      "loss": 5.4596,
+      "step": 12124
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8954715728759766,
+      "learning_rate": 2.0042986313553043e-05,
+      "loss": 5.5597,
+      "step": 12128
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9412333965301514,
+      "learning_rate": 2.0033104402391422e-05,
+      "loss": 5.5437,
+      "step": 12132
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7679616212844849,
+      "learning_rate": 2.0023222491229804e-05,
+      "loss": 5.3381,
+      "step": 12136
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.884602665901184,
+      "learning_rate": 2.0013340580068186e-05,
+      "loss": 5.5533,
+      "step": 12140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0963213443756104,
+      "learning_rate": 2.0003458668906565e-05,
+      "loss": 5.5015,
+      "step": 12144
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.796938180923462,
+      "learning_rate": 1.999357675774495e-05,
+      "loss": 5.4519,
+      "step": 12148
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2093305587768555,
+      "learning_rate": 1.9983694846583332e-05,
+      "loss": 5.6164,
+      "step": 12152
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8350774049758911,
+      "learning_rate": 1.997381293542171e-05,
+      "loss": 5.583,
+      "step": 12156
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8653216361999512,
+      "learning_rate": 1.9963931024260093e-05,
+      "loss": 5.656,
+      "step": 12160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7206474542617798,
+      "learning_rate": 1.9954049113098475e-05,
+      "loss": 5.5226,
+      "step": 12164
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8878382444381714,
+      "learning_rate": 1.9944167201936857e-05,
+      "loss": 5.6363,
+      "step": 12168
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0668506622314453,
+      "learning_rate": 1.9934285290775236e-05,
+      "loss": 5.5106,
+      "step": 12172
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0027477741241455,
+      "learning_rate": 1.9924403379613618e-05,
+      "loss": 5.5217,
+      "step": 12176
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1123950481414795,
+      "learning_rate": 1.9914521468452e-05,
+      "loss": 5.6275,
+      "step": 12180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1055514812469482,
+      "learning_rate": 1.9904639557290382e-05,
+      "loss": 5.6186,
+      "step": 12184
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0140175819396973,
+      "learning_rate": 1.989475764612876e-05,
+      "loss": 5.5965,
+      "step": 12188
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.902254343032837,
+      "learning_rate": 1.9884875734967142e-05,
+      "loss": 5.5557,
+      "step": 12192
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9510375261306763,
+      "learning_rate": 1.9874993823805525e-05,
+      "loss": 5.4804,
+      "step": 12196
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.950716495513916,
+      "learning_rate": 1.9865111912643907e-05,
+      "loss": 5.5887,
+      "step": 12200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0936696529388428,
+      "learning_rate": 1.985523000148229e-05,
+      "loss": 5.5463,
+      "step": 12204
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6361807584762573,
+      "learning_rate": 1.984534809032067e-05,
+      "loss": 5.5516,
+      "step": 12208
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1499600410461426,
+      "learning_rate": 1.9835466179159053e-05,
+      "loss": 5.5467,
+      "step": 12212
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9551507234573364,
+      "learning_rate": 1.982558426799743e-05,
+      "loss": 5.549,
+      "step": 12216
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9426565170288086,
+      "learning_rate": 1.9815702356835814e-05,
+      "loss": 5.5726,
+      "step": 12220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9303926229476929,
+      "learning_rate": 1.9805820445674196e-05,
+      "loss": 5.5571,
+      "step": 12224
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.859390139579773,
+      "learning_rate": 1.9795938534512574e-05,
+      "loss": 5.499,
+      "step": 12228
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8791084289550781,
+      "learning_rate": 1.9786056623350956e-05,
+      "loss": 5.5617,
+      "step": 12232
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1466195583343506,
+      "learning_rate": 1.977617471218934e-05,
+      "loss": 5.3845,
+      "step": 12236
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7159942388534546,
+      "learning_rate": 1.976629280102772e-05,
+      "loss": 5.551,
+      "step": 12240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8822176456451416,
+      "learning_rate": 1.97564108898661e-05,
+      "loss": 5.4277,
+      "step": 12244
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.798642635345459,
+      "learning_rate": 1.974652897870448e-05,
+      "loss": 5.6463,
+      "step": 12248
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9936386346817017,
+      "learning_rate": 1.9736647067542863e-05,
+      "loss": 5.6342,
+      "step": 12252
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1285390853881836,
+      "learning_rate": 1.9726765156381245e-05,
+      "loss": 5.5005,
+      "step": 12256
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.88554048538208,
+      "learning_rate": 1.9716883245219624e-05,
+      "loss": 5.5671,
+      "step": 12260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1367435455322266,
+      "learning_rate": 1.970700133405801e-05,
+      "loss": 5.4724,
+      "step": 12264
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9615224599838257,
+      "learning_rate": 1.969711942289639e-05,
+      "loss": 5.4797,
+      "step": 12268
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0672008991241455,
+      "learning_rate": 1.968723751173477e-05,
+      "loss": 5.4903,
+      "step": 12272
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8822599649429321,
+      "learning_rate": 1.9677355600573152e-05,
+      "loss": 5.4196,
+      "step": 12276
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9987417459487915,
+      "learning_rate": 1.9667473689411534e-05,
+      "loss": 5.5575,
+      "step": 12280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8995615243911743,
+      "learning_rate": 1.9657591778249916e-05,
+      "loss": 5.5343,
+      "step": 12284
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9003994464874268,
+      "learning_rate": 1.9647709867088295e-05,
+      "loss": 5.5341,
+      "step": 12288
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9870491027832031,
+      "learning_rate": 1.9637827955926677e-05,
+      "loss": 5.5928,
+      "step": 12292
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.072319269180298,
+      "learning_rate": 1.962794604476506e-05,
+      "loss": 5.4821,
+      "step": 12296
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9005805253982544,
+      "learning_rate": 1.9618064133603438e-05,
+      "loss": 5.5296,
+      "step": 12300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8590092658996582,
+      "learning_rate": 1.960818222244182e-05,
+      "loss": 5.568,
+      "step": 12304
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.030334949493408,
+      "learning_rate": 1.9598300311280202e-05,
+      "loss": 5.4239,
+      "step": 12308
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9241563081741333,
+      "learning_rate": 1.9588418400118584e-05,
+      "loss": 5.5618,
+      "step": 12312
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9295763969421387,
+      "learning_rate": 1.9578536488956963e-05,
+      "loss": 5.6011,
+      "step": 12316
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9351352453231812,
+      "learning_rate": 1.9568654577795348e-05,
+      "loss": 5.5549,
+      "step": 12320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0557003021240234,
+      "learning_rate": 1.955877266663373e-05,
+      "loss": 5.531,
+      "step": 12324
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0774848461151123,
+      "learning_rate": 1.954889075547211e-05,
+      "loss": 5.5031,
+      "step": 12328
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9636493921279907,
+      "learning_rate": 1.953900884431049e-05,
+      "loss": 5.5733,
+      "step": 12332
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.006387710571289,
+      "learning_rate": 1.9529126933148873e-05,
+      "loss": 5.5184,
+      "step": 12336
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8912975788116455,
+      "learning_rate": 1.9519245021987255e-05,
+      "loss": 5.4529,
+      "step": 12340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.028090715408325,
+      "learning_rate": 1.9509363110825634e-05,
+      "loss": 5.4242,
+      "step": 12344
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.042482376098633,
+      "learning_rate": 1.9499481199664016e-05,
+      "loss": 5.5455,
+      "step": 12348
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1823337078094482,
+      "learning_rate": 1.9489599288502398e-05,
+      "loss": 5.4165,
+      "step": 12352
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8578866720199585,
+      "learning_rate": 1.947971737734078e-05,
+      "loss": 5.4873,
+      "step": 12356
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.964311957359314,
+      "learning_rate": 1.946983546617916e-05,
+      "loss": 5.5457,
+      "step": 12360
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.030364513397217,
+      "learning_rate": 1.945995355501754e-05,
+      "loss": 5.474,
+      "step": 12364
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0638349056243896,
+      "learning_rate": 1.9450071643855923e-05,
+      "loss": 5.5748,
+      "step": 12368
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.002610445022583,
+      "learning_rate": 1.94401897326943e-05,
+      "loss": 5.546,
+      "step": 12372
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8753662109375,
+      "learning_rate": 1.9430307821532687e-05,
+      "loss": 5.5053,
+      "step": 12376
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1950864791870117,
+      "learning_rate": 1.942042591037107e-05,
+      "loss": 5.4716,
+      "step": 12380
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.067065715789795,
+      "learning_rate": 1.9410543999209447e-05,
+      "loss": 5.5028,
+      "step": 12384
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0129029750823975,
+      "learning_rate": 1.940066208804783e-05,
+      "loss": 5.522,
+      "step": 12388
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5361533164978027,
+      "learning_rate": 1.939078017688621e-05,
+      "loss": 5.4896,
+      "step": 12392
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.945295810699463,
+      "learning_rate": 1.9380898265724594e-05,
+      "loss": 5.5306,
+      "step": 12396
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7914199829101562,
+      "learning_rate": 1.9371016354562972e-05,
+      "loss": 5.5514,
+      "step": 12400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.061509609222412,
+      "learning_rate": 1.9361134443401354e-05,
+      "loss": 5.5803,
+      "step": 12404
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9697644710540771,
+      "learning_rate": 1.9351252532239736e-05,
+      "loss": 5.4989,
+      "step": 12408
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.801199197769165,
+      "learning_rate": 1.934137062107812e-05,
+      "loss": 5.6458,
+      "step": 12412
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.985129475593567,
+      "learning_rate": 1.9331488709916497e-05,
+      "loss": 5.4291,
+      "step": 12416
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8330814838409424,
+      "learning_rate": 1.932160679875488e-05,
+      "loss": 5.3954,
+      "step": 12420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2082693576812744,
+      "learning_rate": 1.931172488759326e-05,
+      "loss": 5.5882,
+      "step": 12424
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6590445041656494,
+      "learning_rate": 1.930184297643164e-05,
+      "loss": 5.4728,
+      "step": 12428
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0365848541259766,
+      "learning_rate": 1.9291961065270022e-05,
+      "loss": 5.5032,
+      "step": 12432
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9523799419403076,
+      "learning_rate": 1.9282079154108407e-05,
+      "loss": 5.4974,
+      "step": 12436
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9144923686981201,
+      "learning_rate": 1.9272197242946786e-05,
+      "loss": 5.5348,
+      "step": 12440
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7671104669570923,
+      "learning_rate": 1.9262315331785168e-05,
+      "loss": 5.5292,
+      "step": 12444
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0515549182891846,
+      "learning_rate": 1.925243342062355e-05,
+      "loss": 5.4845,
+      "step": 12448
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9381266832351685,
+      "learning_rate": 1.9242551509461932e-05,
+      "loss": 5.5455,
+      "step": 12452
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.017817497253418,
+      "learning_rate": 1.923266959830031e-05,
+      "loss": 5.3844,
+      "step": 12456
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8400537967681885,
+      "learning_rate": 1.9222787687138693e-05,
+      "loss": 5.4878,
+      "step": 12460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.263641119003296,
+      "learning_rate": 1.9212905775977075e-05,
+      "loss": 5.5841,
+      "step": 12464
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8722437620162964,
+      "learning_rate": 1.9203023864815457e-05,
+      "loss": 5.4782,
+      "step": 12468
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.020585060119629,
+      "learning_rate": 1.9193141953653836e-05,
+      "loss": 5.4668,
+      "step": 12472
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7479088306427002,
+      "learning_rate": 1.9183260042492218e-05,
+      "loss": 5.5514,
+      "step": 12476
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9156551361083984,
+      "learning_rate": 1.91733781313306e-05,
+      "loss": 5.4994,
+      "step": 12480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.882408618927002,
+      "learning_rate": 1.9163496220168982e-05,
+      "loss": 5.4834,
+      "step": 12484
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.083282232284546,
+      "learning_rate": 1.915361430900736e-05,
+      "loss": 5.5116,
+      "step": 12488
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9320555925369263,
+      "learning_rate": 1.9143732397845746e-05,
+      "loss": 5.5378,
+      "step": 12492
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.215940475463867,
+      "learning_rate": 1.9133850486684128e-05,
+      "loss": 5.4937,
+      "step": 12496
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.308119297027588,
+      "learning_rate": 1.9123968575522507e-05,
+      "loss": 5.555,
+      "step": 12500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.178675413131714,
+      "learning_rate": 1.911408666436089e-05,
+      "loss": 5.487,
+      "step": 12504
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.004458427429199,
+      "learning_rate": 1.910420475319927e-05,
+      "loss": 5.5416,
+      "step": 12508
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0435168743133545,
+      "learning_rate": 1.909432284203765e-05,
+      "loss": 5.4122,
+      "step": 12512
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0281968116760254,
+      "learning_rate": 1.9084440930876032e-05,
+      "loss": 5.5032,
+      "step": 12516
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9901241064071655,
+      "learning_rate": 1.9074559019714414e-05,
+      "loss": 5.5346,
+      "step": 12520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0608649253845215,
+      "learning_rate": 1.9064677108552796e-05,
+      "loss": 5.5318,
+      "step": 12524
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1132655143737793,
+      "learning_rate": 1.9054795197391175e-05,
+      "loss": 5.5227,
+      "step": 12528
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1006295680999756,
+      "learning_rate": 1.9044913286229557e-05,
+      "loss": 5.4704,
+      "step": 12532
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8386894464492798,
+      "learning_rate": 1.903503137506794e-05,
+      "loss": 5.4159,
+      "step": 12536
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9647696018218994,
+      "learning_rate": 1.902514946390632e-05,
+      "loss": 5.6805,
+      "step": 12540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1188244819641113,
+      "learning_rate": 1.90152675527447e-05,
+      "loss": 5.4004,
+      "step": 12544
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0802998542785645,
+      "learning_rate": 1.9005385641583085e-05,
+      "loss": 5.4416,
+      "step": 12548
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.834084153175354,
+      "learning_rate": 1.8995503730421467e-05,
+      "loss": 5.6306,
+      "step": 12552
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8112331628799438,
+      "learning_rate": 1.8985621819259846e-05,
+      "loss": 5.5333,
+      "step": 12556
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.899707317352295,
+      "learning_rate": 1.8975739908098228e-05,
+      "loss": 5.4976,
+      "step": 12560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.298161506652832,
+      "learning_rate": 1.896585799693661e-05,
+      "loss": 5.594,
+      "step": 12564
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.173597574234009,
+      "learning_rate": 1.8955976085774992e-05,
+      "loss": 5.572,
+      "step": 12568
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8348718881607056,
+      "learning_rate": 1.894609417461337e-05,
+      "loss": 5.4743,
+      "step": 12572
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7967536449432373,
+      "learning_rate": 1.8936212263451753e-05,
+      "loss": 5.5196,
+      "step": 12576
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9703553915023804,
+      "learning_rate": 1.8926330352290135e-05,
+      "loss": 5.532,
+      "step": 12580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7581743001937866,
+      "learning_rate": 1.8916448441128513e-05,
+      "loss": 5.5114,
+      "step": 12584
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0711758136749268,
+      "learning_rate": 1.8906566529966895e-05,
+      "loss": 5.4094,
+      "step": 12588
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7996923923492432,
+      "learning_rate": 1.8896684618805277e-05,
+      "loss": 5.5836,
+      "step": 12592
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.97800612449646,
+      "learning_rate": 1.888680270764366e-05,
+      "loss": 5.4898,
+      "step": 12596
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7940218448638916,
+      "learning_rate": 1.8876920796482038e-05,
+      "loss": 5.6005,
+      "step": 12600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8298521041870117,
+      "learning_rate": 1.886703888532042e-05,
+      "loss": 5.55,
+      "step": 12604
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.040109872817993,
+      "learning_rate": 1.8857156974158806e-05,
+      "loss": 5.5685,
+      "step": 12608
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8531662225723267,
+      "learning_rate": 1.8847275062997184e-05,
+      "loss": 5.5343,
+      "step": 12612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1842970848083496,
+      "learning_rate": 1.8837393151835566e-05,
+      "loss": 5.4704,
+      "step": 12616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.876779556274414,
+      "learning_rate": 1.882751124067395e-05,
+      "loss": 5.5272,
+      "step": 12620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9100033044815063,
+      "learning_rate": 1.881762932951233e-05,
+      "loss": 5.4669,
+      "step": 12624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.233772039413452,
+      "learning_rate": 1.880774741835071e-05,
+      "loss": 5.5786,
+      "step": 12628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.021141767501831,
+      "learning_rate": 1.879786550718909e-05,
+      "loss": 5.5469,
+      "step": 12632
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8748712539672852,
+      "learning_rate": 1.8787983596027473e-05,
+      "loss": 5.4186,
+      "step": 12636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0556745529174805,
+      "learning_rate": 1.8778101684865855e-05,
+      "loss": 5.6404,
+      "step": 12640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.089085102081299,
+      "learning_rate": 1.8768219773704234e-05,
+      "loss": 5.6056,
+      "step": 12644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8434518575668335,
+      "learning_rate": 1.8758337862542616e-05,
+      "loss": 5.6488,
+      "step": 12648
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.003434658050537,
+      "learning_rate": 1.8748455951380998e-05,
+      "loss": 5.5057,
+      "step": 12652
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.292663335800171,
+      "learning_rate": 1.8738574040219377e-05,
+      "loss": 5.5022,
+      "step": 12656
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9476063251495361,
+      "learning_rate": 1.872869212905776e-05,
+      "loss": 5.5844,
+      "step": 12660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.138032913208008,
+      "learning_rate": 1.8718810217896144e-05,
+      "loss": 5.4383,
+      "step": 12664
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8477308750152588,
+      "learning_rate": 1.8708928306734523e-05,
+      "loss": 5.5665,
+      "step": 12668
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9181241989135742,
+      "learning_rate": 1.8699046395572905e-05,
+      "loss": 5.5661,
+      "step": 12672
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.928312063217163,
+      "learning_rate": 1.8689164484411287e-05,
+      "loss": 5.5366,
+      "step": 12676
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.210855484008789,
+      "learning_rate": 1.867928257324967e-05,
+      "loss": 5.5136,
+      "step": 12680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.030755043029785,
+      "learning_rate": 1.8669400662088048e-05,
+      "loss": 5.4813,
+      "step": 12684
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1717166900634766,
+      "learning_rate": 1.865951875092643e-05,
+      "loss": 5.579,
+      "step": 12688
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.068718671798706,
+      "learning_rate": 1.8649636839764812e-05,
+      "loss": 5.4381,
+      "step": 12692
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0134997367858887,
+      "learning_rate": 1.8639754928603194e-05,
+      "loss": 5.6513,
+      "step": 12696
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.061288356781006,
+      "learning_rate": 1.8629873017441573e-05,
+      "loss": 5.4894,
+      "step": 12700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0297648906707764,
+      "learning_rate": 1.8619991106279955e-05,
+      "loss": 5.5485,
+      "step": 12704
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0792784690856934,
+      "learning_rate": 1.8610109195118337e-05,
+      "loss": 5.5622,
+      "step": 12708
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.003371238708496,
+      "learning_rate": 1.8600227283956715e-05,
+      "loss": 5.5616,
+      "step": 12712
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0913472175598145,
+      "learning_rate": 1.8590345372795098e-05,
+      "loss": 5.6751,
+      "step": 12716
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9094600677490234,
+      "learning_rate": 1.8580463461633483e-05,
+      "loss": 5.5047,
+      "step": 12720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1928985118865967,
+      "learning_rate": 1.857058155047186e-05,
+      "loss": 5.6044,
+      "step": 12724
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9102288484573364,
+      "learning_rate": 1.8560699639310244e-05,
+      "loss": 5.4843,
+      "step": 12728
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.908156156539917,
+      "learning_rate": 1.8550817728148626e-05,
+      "loss": 5.5217,
+      "step": 12732
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8940017223358154,
+      "learning_rate": 1.8540935816987008e-05,
+      "loss": 5.5211,
+      "step": 12736
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6733046770095825,
+      "learning_rate": 1.8531053905825386e-05,
+      "loss": 5.6507,
+      "step": 12740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9494881629943848,
+      "learning_rate": 1.852117199466377e-05,
+      "loss": 5.6028,
+      "step": 12744
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.361642360687256,
+      "learning_rate": 1.851129008350215e-05,
+      "loss": 5.5678,
+      "step": 12748
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9810757637023926,
+      "learning_rate": 1.8501408172340533e-05,
+      "loss": 5.5055,
+      "step": 12752
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.196544885635376,
+      "learning_rate": 1.849152626117891e-05,
+      "loss": 5.4622,
+      "step": 12756
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.841874599456787,
+      "learning_rate": 1.8481644350017293e-05,
+      "loss": 5.525,
+      "step": 12760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.982703685760498,
+      "learning_rate": 1.8471762438855675e-05,
+      "loss": 5.5797,
+      "step": 12764
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.193528413772583,
+      "learning_rate": 1.8461880527694058e-05,
+      "loss": 5.4744,
+      "step": 12768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6755714416503906,
+      "learning_rate": 1.8451998616532436e-05,
+      "loss": 5.4944,
+      "step": 12772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9214802980422974,
+      "learning_rate": 1.8442116705370818e-05,
+      "loss": 5.4525,
+      "step": 12776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1332216262817383,
+      "learning_rate": 1.8432234794209204e-05,
+      "loss": 5.4974,
+      "step": 12780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9983022212982178,
+      "learning_rate": 1.8422352883047582e-05,
+      "loss": 5.4737,
+      "step": 12784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.090367555618286,
+      "learning_rate": 1.8412470971885964e-05,
+      "loss": 5.4413,
+      "step": 12788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.903393030166626,
+      "learning_rate": 1.8402589060724346e-05,
+      "loss": 5.5338,
+      "step": 12792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8425928354263306,
+      "learning_rate": 1.8392707149562725e-05,
+      "loss": 5.5461,
+      "step": 12796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7911487817764282,
+      "learning_rate": 1.8382825238401107e-05,
+      "loss": 5.5144,
+      "step": 12800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9388378858566284,
+      "learning_rate": 1.837294332723949e-05,
+      "loss": 5.4777,
+      "step": 12804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9651098251342773,
+      "learning_rate": 1.836306141607787e-05,
+      "loss": 5.583,
+      "step": 12808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.097846269607544,
+      "learning_rate": 1.835317950491625e-05,
+      "loss": 5.4835,
+      "step": 12812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8683522939682007,
+      "learning_rate": 1.8343297593754632e-05,
+      "loss": 5.542,
+      "step": 12816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9476234912872314,
+      "learning_rate": 1.8333415682593014e-05,
+      "loss": 5.5376,
+      "step": 12820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.049328088760376,
+      "learning_rate": 1.8323533771431396e-05,
+      "loss": 5.4911,
+      "step": 12824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8876453638076782,
+      "learning_rate": 1.8313651860269775e-05,
+      "loss": 5.5226,
+      "step": 12828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9597487449645996,
+      "learning_rate": 1.8303769949108157e-05,
+      "loss": 5.4785,
+      "step": 12832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0029759407043457,
+      "learning_rate": 1.8293888037946542e-05,
+      "loss": 5.5109,
+      "step": 12836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9219965934753418,
+      "learning_rate": 1.828400612678492e-05,
+      "loss": 5.508,
+      "step": 12840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9963774681091309,
+      "learning_rate": 1.8274124215623303e-05,
+      "loss": 5.664,
+      "step": 12844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.181628942489624,
+      "learning_rate": 1.8264242304461685e-05,
+      "loss": 5.4549,
+      "step": 12848
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9105952978134155,
+      "learning_rate": 1.8254360393300067e-05,
+      "loss": 5.6288,
+      "step": 12852
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8461229801177979,
+      "learning_rate": 1.8244478482138446e-05,
+      "loss": 5.5195,
+      "step": 12856
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9446773529052734,
+      "learning_rate": 1.8234596570976828e-05,
+      "loss": 5.4313,
+      "step": 12860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.007297992706299,
+      "learning_rate": 1.822471465981521e-05,
+      "loss": 5.6664,
+      "step": 12864
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0537166595458984,
+      "learning_rate": 1.821483274865359e-05,
+      "loss": 5.5447,
+      "step": 12868
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0639407634735107,
+      "learning_rate": 1.820495083749197e-05,
+      "loss": 5.4532,
+      "step": 12872
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9469093084335327,
+      "learning_rate": 1.8195068926330353e-05,
+      "loss": 5.5461,
+      "step": 12876
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.766298770904541,
+      "learning_rate": 1.8185187015168735e-05,
+      "loss": 5.4775,
+      "step": 12880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7954472303390503,
+      "learning_rate": 1.8175305104007114e-05,
+      "loss": 5.5326,
+      "step": 12884
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8361804485321045,
+      "learning_rate": 1.8165423192845496e-05,
+      "loss": 5.5217,
+      "step": 12888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2192041873931885,
+      "learning_rate": 1.815554128168388e-05,
+      "loss": 5.4166,
+      "step": 12892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.092569351196289,
+      "learning_rate": 1.814565937052226e-05,
+      "loss": 5.5973,
+      "step": 12896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8529601097106934,
+      "learning_rate": 1.8135777459360642e-05,
+      "loss": 5.4088,
+      "step": 12900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7156609296798706,
+      "learning_rate": 1.8125895548199024e-05,
+      "loss": 5.5506,
+      "step": 12904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3277995586395264,
+      "learning_rate": 1.8116013637037406e-05,
+      "loss": 5.4852,
+      "step": 12908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.039177894592285,
+      "learning_rate": 1.8106131725875785e-05,
+      "loss": 5.509,
+      "step": 12912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2541489601135254,
+      "learning_rate": 1.8096249814714167e-05,
+      "loss": 5.5551,
+      "step": 12916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9068487882614136,
+      "learning_rate": 1.808636790355255e-05,
+      "loss": 5.5223,
+      "step": 12920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8464723825454712,
+      "learning_rate": 1.807648599239093e-05,
+      "loss": 5.4688,
+      "step": 12924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8630855083465576,
+      "learning_rate": 1.806660408122931e-05,
+      "loss": 5.521,
+      "step": 12928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9354201555252075,
+      "learning_rate": 1.805672217006769e-05,
+      "loss": 5.5234,
+      "step": 12932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2278544902801514,
+      "learning_rate": 1.8046840258906074e-05,
+      "loss": 5.563,
+      "step": 12936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0102896690368652,
+      "learning_rate": 1.8036958347744452e-05,
+      "loss": 5.6071,
+      "step": 12940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.12506103515625,
+      "learning_rate": 1.8027076436582834e-05,
+      "loss": 5.5518,
+      "step": 12944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.134568929672241,
+      "learning_rate": 1.8017194525421216e-05,
+      "loss": 5.4289,
+      "step": 12948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.866938829421997,
+      "learning_rate": 1.80073126142596e-05,
+      "loss": 5.4047,
+      "step": 12952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0489606857299805,
+      "learning_rate": 1.799743070309798e-05,
+      "loss": 5.4668,
+      "step": 12956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.130350351333618,
+      "learning_rate": 1.7987548791936363e-05,
+      "loss": 5.599,
+      "step": 12960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1276466846466064,
+      "learning_rate": 1.7977666880774745e-05,
+      "loss": 5.3925,
+      "step": 12964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8620883226394653,
+      "learning_rate": 1.7967784969613123e-05,
+      "loss": 5.4918,
+      "step": 12968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7407227754592896,
+      "learning_rate": 1.7957903058451505e-05,
+      "loss": 5.5406,
+      "step": 12972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9221688508987427,
+      "learning_rate": 1.7948021147289887e-05,
+      "loss": 5.5111,
+      "step": 12976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0765016078948975,
+      "learning_rate": 1.793813923612827e-05,
+      "loss": 5.6063,
+      "step": 12980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.040132999420166,
+      "learning_rate": 1.7928257324966648e-05,
+      "loss": 5.5904,
+      "step": 12984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8433319330215454,
+      "learning_rate": 1.791837541380503e-05,
+      "loss": 5.4806,
+      "step": 12988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.065800428390503,
+      "learning_rate": 1.7908493502643412e-05,
+      "loss": 5.563,
+      "step": 12992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.199831485748291,
+      "learning_rate": 1.789861159148179e-05,
+      "loss": 5.5233,
+      "step": 12996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8309836387634277,
+      "learning_rate": 1.7888729680320173e-05,
+      "loss": 5.4911,
+      "step": 13000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.900347113609314,
+      "learning_rate": 1.7878847769158555e-05,
+      "loss": 5.5083,
+      "step": 13004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8298702239990234,
+      "learning_rate": 1.786896585799694e-05,
+      "loss": 5.5197,
+      "step": 13008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8966771364212036,
+      "learning_rate": 1.785908394683532e-05,
+      "loss": 5.4034,
+      "step": 13012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9861708879470825,
+      "learning_rate": 1.78492020356737e-05,
+      "loss": 5.3773,
+      "step": 13016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1633896827697754,
+      "learning_rate": 1.7839320124512083e-05,
+      "loss": 5.5735,
+      "step": 13020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8944169282913208,
+      "learning_rate": 1.7829438213350462e-05,
+      "loss": 5.4982,
+      "step": 13024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.358996868133545,
+      "learning_rate": 1.7819556302188844e-05,
+      "loss": 5.5851,
+      "step": 13028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8002705574035645,
+      "learning_rate": 1.7809674391027226e-05,
+      "loss": 5.5903,
+      "step": 13032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.174081325531006,
+      "learning_rate": 1.7799792479865608e-05,
+      "loss": 5.5327,
+      "step": 13036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.115267515182495,
+      "learning_rate": 1.7789910568703987e-05,
+      "loss": 5.6085,
+      "step": 13040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.197908878326416,
+      "learning_rate": 1.778002865754237e-05,
+      "loss": 5.5499,
+      "step": 13044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2714781761169434,
+      "learning_rate": 1.777014674638075e-05,
+      "loss": 5.6221,
+      "step": 13048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1368441581726074,
+      "learning_rate": 1.7760264835219133e-05,
+      "loss": 5.5613,
+      "step": 13052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0283219814300537,
+      "learning_rate": 1.775038292405751e-05,
+      "loss": 5.4734,
+      "step": 13056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7801271677017212,
+      "learning_rate": 1.7740501012895894e-05,
+      "loss": 5.4928,
+      "step": 13060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8141615390777588,
+      "learning_rate": 1.7730619101734276e-05,
+      "loss": 5.4943,
+      "step": 13064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9424062967300415,
+      "learning_rate": 1.7720737190572658e-05,
+      "loss": 5.4888,
+      "step": 13068
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7956730127334595,
+      "learning_rate": 1.771085527941104e-05,
+      "loss": 5.3367,
+      "step": 13072
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8769028186798096,
+      "learning_rate": 1.7700973368249422e-05,
+      "loss": 5.5365,
+      "step": 13076
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8470765352249146,
+      "learning_rate": 1.76910914570878e-05,
+      "loss": 5.6281,
+      "step": 13080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9143494367599487,
+      "learning_rate": 1.7681209545926183e-05,
+      "loss": 5.403,
+      "step": 13084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8906290531158447,
+      "learning_rate": 1.7671327634764565e-05,
+      "loss": 5.4201,
+      "step": 13088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9759694337844849,
+      "learning_rate": 1.7661445723602947e-05,
+      "loss": 5.5,
+      "step": 13092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.16597580909729,
+      "learning_rate": 1.7651563812441325e-05,
+      "loss": 5.4831,
+      "step": 13096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.142273187637329,
+      "learning_rate": 1.7641681901279708e-05,
+      "loss": 5.5608,
+      "step": 13100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0143542289733887,
+      "learning_rate": 1.763179999011809e-05,
+      "loss": 5.4914,
+      "step": 13104
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0241010189056396,
+      "learning_rate": 1.762191807895647e-05,
+      "loss": 5.5402,
+      "step": 13108
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.111691951751709,
+      "learning_rate": 1.761203616779485e-05,
+      "loss": 5.5417,
+      "step": 13112
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0966546535491943,
+      "learning_rate": 1.7602154256633232e-05,
+      "loss": 5.4544,
+      "step": 13116
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2063889503479004,
+      "learning_rate": 1.7592272345471614e-05,
+      "loss": 5.5447,
+      "step": 13120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0416488647460938,
+      "learning_rate": 1.7582390434309997e-05,
+      "loss": 5.56,
+      "step": 13124
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1015825271606445,
+      "learning_rate": 1.757250852314838e-05,
+      "loss": 5.5632,
+      "step": 13128
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.154283046722412,
+      "learning_rate": 1.756262661198676e-05,
+      "loss": 5.4369,
+      "step": 13132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9613300561904907,
+      "learning_rate": 1.7552744700825143e-05,
+      "loss": 5.4493,
+      "step": 13136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.936897873878479,
+      "learning_rate": 1.754286278966352e-05,
+      "loss": 5.6429,
+      "step": 13140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1174933910369873,
+      "learning_rate": 1.7532980878501903e-05,
+      "loss": 5.5222,
+      "step": 13144
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8853952884674072,
+      "learning_rate": 1.7523098967340285e-05,
+      "loss": 5.5823,
+      "step": 13148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7702313661575317,
+      "learning_rate": 1.7513217056178664e-05,
+      "loss": 5.5205,
+      "step": 13152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8901150226593018,
+      "learning_rate": 1.7503335145017046e-05,
+      "loss": 5.5227,
+      "step": 13156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9898940324783325,
+      "learning_rate": 1.7493453233855428e-05,
+      "loss": 5.5064,
+      "step": 13160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8458938598632812,
+      "learning_rate": 1.748357132269381e-05,
+      "loss": 5.4032,
+      "step": 13164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1321139335632324,
+      "learning_rate": 1.747368941153219e-05,
+      "loss": 5.5491,
+      "step": 13168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.055555820465088,
+      "learning_rate": 1.746380750037057e-05,
+      "loss": 5.4821,
+      "step": 13172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0164570808410645,
+      "learning_rate": 1.7453925589208953e-05,
+      "loss": 5.4772,
+      "step": 13176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8798964023590088,
+      "learning_rate": 1.7444043678047335e-05,
+      "loss": 5.6292,
+      "step": 13180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8820830583572388,
+      "learning_rate": 1.7434161766885717e-05,
+      "loss": 5.5503,
+      "step": 13184
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.768707513809204,
+      "learning_rate": 1.74242798557241e-05,
+      "loss": 5.5907,
+      "step": 13188
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7819797992706299,
+      "learning_rate": 1.741439794456248e-05,
+      "loss": 5.3631,
+      "step": 13192
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9888780117034912,
+      "learning_rate": 1.740451603340086e-05,
+      "loss": 5.5318,
+      "step": 13196
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7733293771743774,
+      "learning_rate": 1.7394634122239242e-05,
+      "loss": 5.4794,
+      "step": 13200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.08803653717041,
+      "learning_rate": 1.7384752211077624e-05,
+      "loss": 5.5392,
+      "step": 13204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7837083339691162,
+      "learning_rate": 1.7374870299916006e-05,
+      "loss": 5.5612,
+      "step": 13208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.200601100921631,
+      "learning_rate": 1.7364988388754385e-05,
+      "loss": 5.4618,
+      "step": 13212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.062946319580078,
+      "learning_rate": 1.7355106477592767e-05,
+      "loss": 5.4031,
+      "step": 13216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8929904699325562,
+      "learning_rate": 1.734522456643115e-05,
+      "loss": 5.6011,
+      "step": 13220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.175036907196045,
+      "learning_rate": 1.7335342655269528e-05,
+      "loss": 5.5038,
+      "step": 13224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1107425689697266,
+      "learning_rate": 1.732546074410791e-05,
+      "loss": 5.4624,
+      "step": 13228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8881957530975342,
+      "learning_rate": 1.7315578832946292e-05,
+      "loss": 5.4579,
+      "step": 13232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9419997930526733,
+      "learning_rate": 1.7305696921784674e-05,
+      "loss": 5.5691,
+      "step": 13236
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.006504535675049,
+      "learning_rate": 1.7295815010623056e-05,
+      "loss": 5.604,
+      "step": 13240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9548697471618652,
+      "learning_rate": 1.7285933099461438e-05,
+      "loss": 5.5202,
+      "step": 13244
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.903361201286316,
+      "learning_rate": 1.727605118829982e-05,
+      "loss": 5.5459,
+      "step": 13248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9035142660140991,
+      "learning_rate": 1.72661692771382e-05,
+      "loss": 5.5067,
+      "step": 13252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9801918268203735,
+      "learning_rate": 1.725628736597658e-05,
+      "loss": 5.502,
+      "step": 13256
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8996813297271729,
+      "learning_rate": 1.7246405454814963e-05,
+      "loss": 5.4709,
+      "step": 13260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.903206467628479,
+      "learning_rate": 1.7236523543653345e-05,
+      "loss": 5.4252,
+      "step": 13264
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.280048370361328,
+      "learning_rate": 1.7226641632491724e-05,
+      "loss": 5.6236,
+      "step": 13268
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.916698932647705,
+      "learning_rate": 1.7216759721330106e-05,
+      "loss": 5.5417,
+      "step": 13272
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8944514989852905,
+      "learning_rate": 1.7206877810168488e-05,
+      "loss": 5.5273,
+      "step": 13276
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.160426139831543,
+      "learning_rate": 1.7196995899006866e-05,
+      "loss": 5.4925,
+      "step": 13280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9661693572998047,
+      "learning_rate": 1.718711398784525e-05,
+      "loss": 5.5566,
+      "step": 13284
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1579537391662598,
+      "learning_rate": 1.717723207668363e-05,
+      "loss": 5.5,
+      "step": 13288
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8311907052993774,
+      "learning_rate": 1.7167350165522013e-05,
+      "loss": 5.4402,
+      "step": 13292
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.067732810974121,
+      "learning_rate": 1.7157468254360395e-05,
+      "loss": 5.5876,
+      "step": 13296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9030920267105103,
+      "learning_rate": 1.7147586343198777e-05,
+      "loss": 5.5544,
+      "step": 13300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9689793586730957,
+      "learning_rate": 1.713770443203716e-05,
+      "loss": 5.5876,
+      "step": 13304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.319972038269043,
+      "learning_rate": 1.7127822520875537e-05,
+      "loss": 5.6179,
+      "step": 13308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.05124568939209,
+      "learning_rate": 1.711794060971392e-05,
+      "loss": 5.4808,
+      "step": 13312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.054259777069092,
+      "learning_rate": 1.71080586985523e-05,
+      "loss": 5.571,
+      "step": 13316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1697633266448975,
+      "learning_rate": 1.7098176787390684e-05,
+      "loss": 5.5687,
+      "step": 13320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.158599853515625,
+      "learning_rate": 1.7088294876229062e-05,
+      "loss": 5.4535,
+      "step": 13324
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.263106107711792,
+      "learning_rate": 1.7078412965067444e-05,
+      "loss": 5.5662,
+      "step": 13328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9761734008789062,
+      "learning_rate": 1.7068531053905826e-05,
+      "loss": 5.4655,
+      "step": 13332
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1491572856903076,
+      "learning_rate": 1.705864914274421e-05,
+      "loss": 5.5403,
+      "step": 13336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9614084959030151,
+      "learning_rate": 1.7048767231582587e-05,
+      "loss": 5.4883,
+      "step": 13340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1084208488464355,
+      "learning_rate": 1.703888532042097e-05,
+      "loss": 5.4915,
+      "step": 13344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9315608739852905,
+      "learning_rate": 1.702900340925935e-05,
+      "loss": 5.531,
+      "step": 13348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9300886392593384,
+      "learning_rate": 1.7019121498097733e-05,
+      "loss": 5.5294,
+      "step": 13352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9439579248428345,
+      "learning_rate": 1.7009239586936115e-05,
+      "loss": 5.5372,
+      "step": 13356
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8347561359405518,
+      "learning_rate": 1.6999357675774497e-05,
+      "loss": 5.5304,
+      "step": 13360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8642383813858032,
+      "learning_rate": 1.6989475764612876e-05,
+      "loss": 5.3163,
+      "step": 13364
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9651672840118408,
+      "learning_rate": 1.6979593853451258e-05,
+      "loss": 5.4271,
+      "step": 13368
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.146197557449341,
+      "learning_rate": 1.696971194228964e-05,
+      "loss": 5.6012,
+      "step": 13372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9717754125595093,
+      "learning_rate": 1.6959830031128022e-05,
+      "loss": 5.4703,
+      "step": 13376
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8049124479293823,
+      "learning_rate": 1.69499481199664e-05,
+      "loss": 5.4398,
+      "step": 13380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8642454147338867,
+      "learning_rate": 1.6940066208804783e-05,
+      "loss": 5.5652,
+      "step": 13384
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8307348489761353,
+      "learning_rate": 1.6930184297643165e-05,
+      "loss": 5.6079,
+      "step": 13388
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8231866359710693,
+      "learning_rate": 1.6920302386481547e-05,
+      "loss": 5.4477,
+      "step": 13392
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7777388095855713,
+      "learning_rate": 1.6910420475319926e-05,
+      "loss": 5.5014,
+      "step": 13396
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1190943717956543,
+      "learning_rate": 1.6900538564158308e-05,
+      "loss": 5.512,
+      "step": 13400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7859139442443848,
+      "learning_rate": 1.689065665299669e-05,
+      "loss": 5.4176,
+      "step": 13404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.937577486038208,
+      "learning_rate": 1.6880774741835072e-05,
+      "loss": 5.5242,
+      "step": 13408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9069881439208984,
+      "learning_rate": 1.6870892830673454e-05,
+      "loss": 5.5248,
+      "step": 13412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3434953689575195,
+      "learning_rate": 1.6861010919511836e-05,
+      "loss": 5.5887,
+      "step": 13416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0236103534698486,
+      "learning_rate": 1.6851129008350218e-05,
+      "loss": 5.5241,
+      "step": 13420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.807458519935608,
+      "learning_rate": 1.6841247097188597e-05,
+      "loss": 5.5018,
+      "step": 13424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7654306888580322,
+      "learning_rate": 1.683136518602698e-05,
+      "loss": 5.5346,
+      "step": 13428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.796311378479004,
+      "learning_rate": 1.682148327486536e-05,
+      "loss": 5.4943,
+      "step": 13432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1759448051452637,
+      "learning_rate": 1.681160136370374e-05,
+      "loss": 5.5745,
+      "step": 13436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9799151420593262,
+      "learning_rate": 1.680171945254212e-05,
+      "loss": 5.5721,
+      "step": 13440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8375133275985718,
+      "learning_rate": 1.6791837541380504e-05,
+      "loss": 5.5058,
+      "step": 13444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0697977542877197,
+      "learning_rate": 1.6781955630218886e-05,
+      "loss": 5.49,
+      "step": 13448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.968781590461731,
+      "learning_rate": 1.6772073719057264e-05,
+      "loss": 5.414,
+      "step": 13452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.182166814804077,
+      "learning_rate": 1.6762191807895647e-05,
+      "loss": 5.6226,
+      "step": 13456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.110377073287964,
+      "learning_rate": 1.675230989673403e-05,
+      "loss": 5.5626,
+      "step": 13460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.979601263999939,
+      "learning_rate": 1.674242798557241e-05,
+      "loss": 5.62,
+      "step": 13464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2129158973693848,
+      "learning_rate": 1.6732546074410793e-05,
+      "loss": 5.5565,
+      "step": 13468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0401570796966553,
+      "learning_rate": 1.6722664163249175e-05,
+      "loss": 5.4781,
+      "step": 13472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7625924348831177,
+      "learning_rate": 1.6712782252087557e-05,
+      "loss": 5.4104,
+      "step": 13476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8981072902679443,
+      "learning_rate": 1.6702900340925936e-05,
+      "loss": 5.5208,
+      "step": 13480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.141097068786621,
+      "learning_rate": 1.6693018429764318e-05,
+      "loss": 5.505,
+      "step": 13484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0228660106658936,
+      "learning_rate": 1.66831365186027e-05,
+      "loss": 5.5653,
+      "step": 13488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9627779722213745,
+      "learning_rate": 1.667325460744108e-05,
+      "loss": 5.5383,
+      "step": 13492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.80388605594635,
+      "learning_rate": 1.666337269627946e-05,
+      "loss": 5.4095,
+      "step": 13496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125562906265259,
+      "learning_rate": 1.6653490785117842e-05,
+      "loss": 5.5157,
+      "step": 13500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3643898963928223,
+      "learning_rate": 1.6643608873956224e-05,
+      "loss": 5.4778,
+      "step": 13504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9573434591293335,
+      "learning_rate": 1.6633726962794603e-05,
+      "loss": 5.3951,
+      "step": 13508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.767749309539795,
+      "learning_rate": 1.6623845051632985e-05,
+      "loss": 5.5775,
+      "step": 13512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7242648601531982,
+      "learning_rate": 1.6613963140471367e-05,
+      "loss": 5.5681,
+      "step": 13516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.098914384841919,
+      "learning_rate": 1.660408122930975e-05,
+      "loss": 5.4892,
+      "step": 13520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9471126794815063,
+      "learning_rate": 1.659419931814813e-05,
+      "loss": 5.5672,
+      "step": 13524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9518860578536987,
+      "learning_rate": 1.6584317406986513e-05,
+      "loss": 5.5318,
+      "step": 13528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9697927236557007,
+      "learning_rate": 1.6574435495824896e-05,
+      "loss": 5.5275,
+      "step": 13532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.005502700805664,
+      "learning_rate": 1.6564553584663274e-05,
+      "loss": 5.6122,
+      "step": 13536
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.06354022026062,
+      "learning_rate": 1.6554671673501656e-05,
+      "loss": 5.3431,
+      "step": 13540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9697012901306152,
+      "learning_rate": 1.6544789762340038e-05,
+      "loss": 5.5376,
+      "step": 13544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9158987998962402,
+      "learning_rate": 1.653490785117842e-05,
+      "loss": 5.5176,
+      "step": 13548
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0514771938323975,
+      "learning_rate": 1.65250259400168e-05,
+      "loss": 5.5724,
+      "step": 13552
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8229410648345947,
+      "learning_rate": 1.651514402885518e-05,
+      "loss": 5.4933,
+      "step": 13556
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6895649433135986,
+      "learning_rate": 1.6505262117693563e-05,
+      "loss": 5.5462,
+      "step": 13560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9636627435684204,
+      "learning_rate": 1.6495380206531942e-05,
+      "loss": 5.4918,
+      "step": 13564
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.728353500366211,
+      "learning_rate": 1.6485498295370324e-05,
+      "loss": 5.4126,
+      "step": 13568
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8992080688476562,
+      "learning_rate": 1.6475616384208706e-05,
+      "loss": 5.4954,
+      "step": 13572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8971209526062012,
+      "learning_rate": 1.6465734473047088e-05,
+      "loss": 5.5205,
+      "step": 13576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9099971055984497,
+      "learning_rate": 1.6455852561885467e-05,
+      "loss": 5.5401,
+      "step": 13580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7198883295059204,
+      "learning_rate": 1.6445970650723852e-05,
+      "loss": 5.5187,
+      "step": 13584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.854477882385254,
+      "learning_rate": 1.6436088739562234e-05,
+      "loss": 5.5558,
+      "step": 13588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9261908531188965,
+      "learning_rate": 1.6426206828400613e-05,
+      "loss": 5.5156,
+      "step": 13592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.166208505630493,
+      "learning_rate": 1.6416324917238995e-05,
+      "loss": 5.5002,
+      "step": 13596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8934109210968018,
+      "learning_rate": 1.6406443006077377e-05,
+      "loss": 5.5092,
+      "step": 13600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.830947756767273,
+      "learning_rate": 1.639656109491576e-05,
+      "loss": 5.5542,
+      "step": 13604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.86097252368927,
+      "learning_rate": 1.6386679183754138e-05,
+      "loss": 5.502,
+      "step": 13608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9578551054000854,
+      "learning_rate": 1.637679727259252e-05,
+      "loss": 5.5351,
+      "step": 13612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9985218048095703,
+      "learning_rate": 1.6366915361430902e-05,
+      "loss": 5.403,
+      "step": 13616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.346946954727173,
+      "learning_rate": 1.6357033450269284e-05,
+      "loss": 5.4314,
+      "step": 13620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9613852500915527,
+      "learning_rate": 1.6347151539107663e-05,
+      "loss": 5.4851,
+      "step": 13624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2844598293304443,
+      "learning_rate": 1.6337269627946045e-05,
+      "loss": 5.4646,
+      "step": 13628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0212109088897705,
+      "learning_rate": 1.6327387716784427e-05,
+      "loss": 5.5153,
+      "step": 13632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0245635509490967,
+      "learning_rate": 1.6317505805622805e-05,
+      "loss": 5.5516,
+      "step": 13636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.051738977432251,
+      "learning_rate": 1.630762389446119e-05,
+      "loss": 5.6327,
+      "step": 13640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0657126903533936,
+      "learning_rate": 1.6297741983299573e-05,
+      "loss": 5.5882,
+      "step": 13644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.979304313659668,
+      "learning_rate": 1.628786007213795e-05,
+      "loss": 5.3553,
+      "step": 13648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.10304856300354,
+      "learning_rate": 1.6277978160976334e-05,
+      "loss": 5.5474,
+      "step": 13652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0482354164123535,
+      "learning_rate": 1.6268096249814716e-05,
+      "loss": 5.5004,
+      "step": 13656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0741443634033203,
+      "learning_rate": 1.6258214338653098e-05,
+      "loss": 5.4887,
+      "step": 13660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9850010871887207,
+      "learning_rate": 1.6248332427491476e-05,
+      "loss": 5.4449,
+      "step": 13664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.456059217453003,
+      "learning_rate": 1.623845051632986e-05,
+      "loss": 5.4471,
+      "step": 13668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8275492191314697,
+      "learning_rate": 1.622856860516824e-05,
+      "loss": 5.5139,
+      "step": 13672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.166975736618042,
+      "learning_rate": 1.6218686694006623e-05,
+      "loss": 5.5475,
+      "step": 13676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0723702907562256,
+      "learning_rate": 1.6208804782845e-05,
+      "loss": 5.6062,
+      "step": 13680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.227518320083618,
+      "learning_rate": 1.6198922871683383e-05,
+      "loss": 5.4955,
+      "step": 13684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9052515029907227,
+      "learning_rate": 1.6189040960521765e-05,
+      "loss": 5.5553,
+      "step": 13688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.945573329925537,
+      "learning_rate": 1.6179159049360147e-05,
+      "loss": 5.3597,
+      "step": 13692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.874603271484375,
+      "learning_rate": 1.616927713819853e-05,
+      "loss": 5.391,
+      "step": 13696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.339505672454834,
+      "learning_rate": 1.615939522703691e-05,
+      "loss": 5.4194,
+      "step": 13700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7062286138534546,
+      "learning_rate": 1.6149513315875294e-05,
+      "loss": 5.4586,
+      "step": 13704
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8526628017425537,
+      "learning_rate": 1.6139631404713672e-05,
+      "loss": 5.3614,
+      "step": 13708
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8179652690887451,
+      "learning_rate": 1.6129749493552054e-05,
+      "loss": 5.4977,
+      "step": 13712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.889673113822937,
+      "learning_rate": 1.6119867582390436e-05,
+      "loss": 5.5591,
+      "step": 13716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.039731740951538,
+      "learning_rate": 1.6109985671228815e-05,
+      "loss": 5.4809,
+      "step": 13720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3024516105651855,
+      "learning_rate": 1.6100103760067197e-05,
+      "loss": 5.4518,
+      "step": 13724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9836472272872925,
+      "learning_rate": 1.609022184890558e-05,
+      "loss": 5.4812,
+      "step": 13728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8883979320526123,
+      "learning_rate": 1.608033993774396e-05,
+      "loss": 5.4914,
+      "step": 13732
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9850894212722778,
+      "learning_rate": 1.607045802658234e-05,
+      "loss": 5.6323,
+      "step": 13736
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.124187707901001,
+      "learning_rate": 1.6060576115420722e-05,
+      "loss": 5.5609,
+      "step": 13740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9509978294372559,
+      "learning_rate": 1.6050694204259104e-05,
+      "loss": 5.5234,
+      "step": 13744
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9340142011642456,
+      "learning_rate": 1.6040812293097486e-05,
+      "loss": 5.491,
+      "step": 13748
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0254647731781006,
+      "learning_rate": 1.6030930381935865e-05,
+      "loss": 5.4735,
+      "step": 13752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.143260955810547,
+      "learning_rate": 1.602104847077425e-05,
+      "loss": 5.5477,
+      "step": 13756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6931935548782349,
+      "learning_rate": 1.6011166559612632e-05,
+      "loss": 5.5821,
+      "step": 13760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9342801570892334,
+      "learning_rate": 1.600128464845101e-05,
+      "loss": 5.4517,
+      "step": 13764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8290477991104126,
+      "learning_rate": 1.5991402737289393e-05,
+      "loss": 5.4992,
+      "step": 13768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.020256757736206,
+      "learning_rate": 1.5981520826127775e-05,
+      "loss": 5.5175,
+      "step": 13772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.106531858444214,
+      "learning_rate": 1.5971638914966157e-05,
+      "loss": 5.4482,
+      "step": 13776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9939652681350708,
+      "learning_rate": 1.5961757003804536e-05,
+      "loss": 5.4952,
+      "step": 13780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0970144271850586,
+      "learning_rate": 1.5951875092642918e-05,
+      "loss": 5.5438,
+      "step": 13784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9521361589431763,
+      "learning_rate": 1.59419931814813e-05,
+      "loss": 5.5225,
+      "step": 13788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0148961544036865,
+      "learning_rate": 1.593211127031968e-05,
+      "loss": 5.4785,
+      "step": 13792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1378774642944336,
+      "learning_rate": 1.592222935915806e-05,
+      "loss": 5.5553,
+      "step": 13796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.312819719314575,
+      "learning_rate": 1.5912347447996443e-05,
+      "loss": 5.4033,
+      "step": 13800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2060418128967285,
+      "learning_rate": 1.5902465536834825e-05,
+      "loss": 5.4713,
+      "step": 13804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8953157663345337,
+      "learning_rate": 1.5892583625673203e-05,
+      "loss": 5.5035,
+      "step": 13808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7217421531677246,
+      "learning_rate": 1.588270171451159e-05,
+      "loss": 5.6139,
+      "step": 13812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9969755411148071,
+      "learning_rate": 1.587281980334997e-05,
+      "loss": 5.44,
+      "step": 13816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.071129560470581,
+      "learning_rate": 1.586293789218835e-05,
+      "loss": 5.4202,
+      "step": 13820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8004354238510132,
+      "learning_rate": 1.5853055981026732e-05,
+      "loss": 5.5286,
+      "step": 13824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0194168090820312,
+      "learning_rate": 1.5843174069865114e-05,
+      "loss": 5.426,
+      "step": 13828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7910535335540771,
+      "learning_rate": 1.5833292158703496e-05,
+      "loss": 5.529,
+      "step": 13832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0648398399353027,
+      "learning_rate": 1.5823410247541874e-05,
+      "loss": 5.5632,
+      "step": 13836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.881971001625061,
+      "learning_rate": 1.5813528336380257e-05,
+      "loss": 5.6164,
+      "step": 13840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.088383913040161,
+      "learning_rate": 1.580364642521864e-05,
+      "loss": 5.5221,
+      "step": 13844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9117428064346313,
+      "learning_rate": 1.5793764514057017e-05,
+      "loss": 5.6231,
+      "step": 13848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0459342002868652,
+      "learning_rate": 1.57838826028954e-05,
+      "loss": 5.3971,
+      "step": 13852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8177950382232666,
+      "learning_rate": 1.577400069173378e-05,
+      "loss": 5.4996,
+      "step": 13856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.090265989303589,
+      "learning_rate": 1.5764118780572163e-05,
+      "loss": 5.5479,
+      "step": 13860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1461987495422363,
+      "learning_rate": 1.5754236869410542e-05,
+      "loss": 5.4959,
+      "step": 13864
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.008942127227783,
+      "learning_rate": 1.5744354958248928e-05,
+      "loss": 5.3582,
+      "step": 13868
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7721225023269653,
+      "learning_rate": 1.573447304708731e-05,
+      "loss": 5.4618,
+      "step": 13872
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.161746025085449,
+      "learning_rate": 1.572459113592569e-05,
+      "loss": 5.5163,
+      "step": 13876
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0746963024139404,
+      "learning_rate": 1.571470922476407e-05,
+      "loss": 5.4898,
+      "step": 13880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": Infinity,
+      "learning_rate": 1.5707297791392857e-05,
+      "loss": 5.5354,
+      "step": 13884
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0238466262817383,
+      "learning_rate": 1.5697415880231236e-05,
+      "loss": 5.5099,
+      "step": 13888
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9300113916397095,
+      "learning_rate": 1.5687533969069618e-05,
+      "loss": 5.682,
+      "step": 13892
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9683018922805786,
+      "learning_rate": 1.5677652057908e-05,
+      "loss": 5.5058,
+      "step": 13896
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.975710391998291,
+      "learning_rate": 1.5667770146746382e-05,
+      "loss": 5.5139,
+      "step": 13900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9731028079986572,
+      "learning_rate": 1.5657888235584764e-05,
+      "loss": 5.565,
+      "step": 13904
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0347113609313965,
+      "learning_rate": 1.5648006324423146e-05,
+      "loss": 5.6587,
+      "step": 13908
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9666558504104614,
+      "learning_rate": 1.5638124413261525e-05,
+      "loss": 5.6072,
+      "step": 13912
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.808203935623169,
+      "learning_rate": 1.5628242502099907e-05,
+      "loss": 5.4888,
+      "step": 13916
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0783326625823975,
+      "learning_rate": 1.561836059093829e-05,
+      "loss": 5.6407,
+      "step": 13920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.006601572036743,
+      "learning_rate": 1.560847867977667e-05,
+      "loss": 5.5085,
+      "step": 13924
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8633679151535034,
+      "learning_rate": 1.559859676861505e-05,
+      "loss": 5.4967,
+      "step": 13928
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9505640268325806,
+      "learning_rate": 1.558871485745343e-05,
+      "loss": 5.4678,
+      "step": 13932
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0986247062683105,
+      "learning_rate": 1.5578832946291814e-05,
+      "loss": 5.4932,
+      "step": 13936
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.256755828857422,
+      "learning_rate": 1.5568951035130196e-05,
+      "loss": 5.5958,
+      "step": 13940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2147109508514404,
+      "learning_rate": 1.5559069123968574e-05,
+      "loss": 5.5194,
+      "step": 13944
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.88095223903656,
+      "learning_rate": 1.5549187212806956e-05,
+      "loss": 5.4609,
+      "step": 13948
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.867007851600647,
+      "learning_rate": 1.553930530164534e-05,
+      "loss": 5.4802,
+      "step": 13952
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.029303789138794,
+      "learning_rate": 1.552942339048372e-05,
+      "loss": 5.5337,
+      "step": 13956
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9142879247665405,
+      "learning_rate": 1.5519541479322103e-05,
+      "loss": 5.4932,
+      "step": 13960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8500617742538452,
+      "learning_rate": 1.5509659568160485e-05,
+      "loss": 5.4916,
+      "step": 13964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0694692134857178,
+      "learning_rate": 1.5499777656998867e-05,
+      "loss": 5.4766,
+      "step": 13968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.122901201248169,
+      "learning_rate": 1.5489895745837245e-05,
+      "loss": 5.6467,
+      "step": 13972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.108297109603882,
+      "learning_rate": 1.5480013834675627e-05,
+      "loss": 5.5902,
+      "step": 13976
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9898920059204102,
+      "learning_rate": 1.547013192351401e-05,
+      "loss": 5.5772,
+      "step": 13980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2552645206451416,
+      "learning_rate": 1.5460250012352388e-05,
+      "loss": 5.507,
+      "step": 13984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9198163747787476,
+      "learning_rate": 1.545036810119077e-05,
+      "loss": 5.4749,
+      "step": 13988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.101717233657837,
+      "learning_rate": 1.5440486190029152e-05,
+      "loss": 5.444,
+      "step": 13992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1184744834899902,
+      "learning_rate": 1.5430604278867534e-05,
+      "loss": 5.6017,
+      "step": 13996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7170560359954834,
+      "learning_rate": 1.5420722367705913e-05,
+      "loss": 5.4235,
+      "step": 14000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1068732738494873,
+      "learning_rate": 1.5410840456544295e-05,
+      "loss": 5.504,
+      "step": 14004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.275245189666748,
+      "learning_rate": 1.5400958545382677e-05,
+      "loss": 5.5617,
+      "step": 14008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2603347301483154,
+      "learning_rate": 1.539107663422106e-05,
+      "loss": 5.4309,
+      "step": 14012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.082984685897827,
+      "learning_rate": 1.538119472305944e-05,
+      "loss": 5.408,
+      "step": 14016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8457398414611816,
+      "learning_rate": 1.5371312811897823e-05,
+      "loss": 5.4611,
+      "step": 14020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0834155082702637,
+      "learning_rate": 1.5361430900736205e-05,
+      "loss": 5.4901,
+      "step": 14024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9145028591156006,
+      "learning_rate": 1.5351548989574584e-05,
+      "loss": 5.5115,
+      "step": 14028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7850406169891357,
+      "learning_rate": 1.5341667078412966e-05,
+      "loss": 5.4384,
+      "step": 14032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.964521884918213,
+      "learning_rate": 1.5331785167251348e-05,
+      "loss": 5.3728,
+      "step": 14036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125605821609497,
+      "learning_rate": 1.532190325608973e-05,
+      "loss": 5.4972,
+      "step": 14040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7855336666107178,
+      "learning_rate": 1.531202134492811e-05,
+      "loss": 5.5624,
+      "step": 14044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9751635789871216,
+      "learning_rate": 1.530213943376649e-05,
+      "loss": 5.4388,
+      "step": 14048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.943023681640625,
+      "learning_rate": 1.5292257522604873e-05,
+      "loss": 5.4861,
+      "step": 14052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7670375108718872,
+      "learning_rate": 1.528237561144325e-05,
+      "loss": 5.5438,
+      "step": 14056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.040696859359741,
+      "learning_rate": 1.5272493700281634e-05,
+      "loss": 5.4905,
+      "step": 14060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2104179859161377,
+      "learning_rate": 1.5262611789120016e-05,
+      "loss": 5.4197,
+      "step": 14064
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.070598840713501,
+      "learning_rate": 1.5252729877958396e-05,
+      "loss": 5.3342,
+      "step": 14068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1541106700897217,
+      "learning_rate": 1.524284796679678e-05,
+      "loss": 5.5594,
+      "step": 14072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8858451843261719,
+      "learning_rate": 1.5232966055635162e-05,
+      "loss": 5.5494,
+      "step": 14076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0127689838409424,
+      "learning_rate": 1.5223084144473542e-05,
+      "loss": 5.5005,
+      "step": 14080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.70602548122406,
+      "learning_rate": 1.5213202233311924e-05,
+      "loss": 5.4694,
+      "step": 14084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1480917930603027,
+      "learning_rate": 1.5203320322150305e-05,
+      "loss": 5.5891,
+      "step": 14088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8127115964889526,
+      "learning_rate": 1.5193438410988687e-05,
+      "loss": 5.497,
+      "step": 14092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.768389105796814,
+      "learning_rate": 1.5183556499827067e-05,
+      "loss": 5.5075,
+      "step": 14096
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2110512256622314,
+      "learning_rate": 1.517367458866545e-05,
+      "loss": 5.4944,
+      "step": 14100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2322652339935303,
+      "learning_rate": 1.516379267750383e-05,
+      "loss": 5.5072,
+      "step": 14104
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.103665351867676,
+      "learning_rate": 1.5153910766342212e-05,
+      "loss": 5.5093,
+      "step": 14108
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0604758262634277,
+      "learning_rate": 1.5144028855180592e-05,
+      "loss": 5.4539,
+      "step": 14112
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8728346824645996,
+      "learning_rate": 1.5134146944018972e-05,
+      "loss": 5.401,
+      "step": 14116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.031858444213867,
+      "learning_rate": 1.5124265032857354e-05,
+      "loss": 5.499,
+      "step": 14120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.828974962234497,
+      "learning_rate": 1.5114383121695735e-05,
+      "loss": 5.4467,
+      "step": 14124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8760355710983276,
+      "learning_rate": 1.5104501210534119e-05,
+      "loss": 5.3958,
+      "step": 14128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8751137256622314,
+      "learning_rate": 1.50946192993725e-05,
+      "loss": 5.4515,
+      "step": 14132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9187356233596802,
+      "learning_rate": 1.5084737388210881e-05,
+      "loss": 5.5581,
+      "step": 14136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1077044010162354,
+      "learning_rate": 1.5074855477049263e-05,
+      "loss": 5.5018,
+      "step": 14140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9304084777832031,
+      "learning_rate": 1.5064973565887643e-05,
+      "loss": 5.6017,
+      "step": 14144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7740778923034668,
+      "learning_rate": 1.5055091654726025e-05,
+      "loss": 5.5123,
+      "step": 14148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9732328653335571,
+      "learning_rate": 1.5045209743564406e-05,
+      "loss": 5.4148,
+      "step": 14152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.98763906955719,
+      "learning_rate": 1.5035327832402788e-05,
+      "loss": 5.5671,
+      "step": 14156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.025677442550659,
+      "learning_rate": 1.5025445921241168e-05,
+      "loss": 5.5388,
+      "step": 14160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1304235458374023,
+      "learning_rate": 1.501556401007955e-05,
+      "loss": 5.549,
+      "step": 14164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8666431903839111,
+      "learning_rate": 1.500568209891793e-05,
+      "loss": 5.5133,
+      "step": 14168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.001925468444824,
+      "learning_rate": 1.4995800187756313e-05,
+      "loss": 5.4878,
+      "step": 14172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0786337852478027,
+      "learning_rate": 1.4985918276594693e-05,
+      "loss": 5.5475,
+      "step": 14176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0157644748687744,
+      "learning_rate": 1.4976036365433073e-05,
+      "loss": 5.4686,
+      "step": 14180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1091387271881104,
+      "learning_rate": 1.4966154454271456e-05,
+      "loss": 5.5175,
+      "step": 14184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8892532587051392,
+      "learning_rate": 1.495627254310984e-05,
+      "loss": 5.57,
+      "step": 14188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0681941509246826,
+      "learning_rate": 1.4946390631948221e-05,
+      "loss": 5.4503,
+      "step": 14192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.028324842453003,
+      "learning_rate": 1.4936508720786602e-05,
+      "loss": 5.5231,
+      "step": 14196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8590697050094604,
+      "learning_rate": 1.4926626809624982e-05,
+      "loss": 5.4046,
+      "step": 14200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1684091091156006,
+      "learning_rate": 1.4916744898463364e-05,
+      "loss": 5.5428,
+      "step": 14204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4600932598114014,
+      "learning_rate": 1.4906862987301745e-05,
+      "loss": 5.4973,
+      "step": 14208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.182546615600586,
+      "learning_rate": 1.4896981076140127e-05,
+      "loss": 5.5168,
+      "step": 14212
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0207765102386475,
+      "learning_rate": 1.4887099164978507e-05,
+      "loss": 5.4892,
+      "step": 14216
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.07978892326355,
+      "learning_rate": 1.4877217253816889e-05,
+      "loss": 5.5074,
+      "step": 14220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0122103691101074,
+      "learning_rate": 1.486733534265527e-05,
+      "loss": 5.5007,
+      "step": 14224
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.951704502105713,
+      "learning_rate": 1.4857453431493651e-05,
+      "loss": 5.5281,
+      "step": 14228
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.903686761856079,
+      "learning_rate": 1.4847571520332032e-05,
+      "loss": 5.483,
+      "step": 14232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7512837648391724,
+      "learning_rate": 1.4837689609170414e-05,
+      "loss": 5.5026,
+      "step": 14236
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9716506004333496,
+      "learning_rate": 1.4827807698008794e-05,
+      "loss": 5.5547,
+      "step": 14240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1064813137054443,
+      "learning_rate": 1.4817925786847178e-05,
+      "loss": 5.6868,
+      "step": 14244
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.920245885848999,
+      "learning_rate": 1.480804387568556e-05,
+      "loss": 5.5195,
+      "step": 14248
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8807624578475952,
+      "learning_rate": 1.479816196452394e-05,
+      "loss": 5.5605,
+      "step": 14252
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1558022499084473,
+      "learning_rate": 1.4788280053362322e-05,
+      "loss": 5.54,
+      "step": 14256
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9825403690338135,
+      "learning_rate": 1.4778398142200703e-05,
+      "loss": 5.5888,
+      "step": 14260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.974697232246399,
+      "learning_rate": 1.4768516231039083e-05,
+      "loss": 5.5791,
+      "step": 14264
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0667243003845215,
+      "learning_rate": 1.4758634319877465e-05,
+      "loss": 5.4354,
+      "step": 14268
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8902255296707153,
+      "learning_rate": 1.4748752408715846e-05,
+      "loss": 5.5167,
+      "step": 14272
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.882251501083374,
+      "learning_rate": 1.4738870497554228e-05,
+      "loss": 5.4551,
+      "step": 14276
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0265276432037354,
+      "learning_rate": 1.4728988586392608e-05,
+      "loss": 5.3989,
+      "step": 14280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1499834060668945,
+      "learning_rate": 1.471910667523099e-05,
+      "loss": 5.3913,
+      "step": 14284
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9374068975448608,
+      "learning_rate": 1.470922476406937e-05,
+      "loss": 5.4951,
+      "step": 14288
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.04496693611145,
+      "learning_rate": 1.4699342852907753e-05,
+      "loss": 5.5568,
+      "step": 14292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1718828678131104,
+      "learning_rate": 1.4689460941746133e-05,
+      "loss": 5.4003,
+      "step": 14296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.170172929763794,
+      "learning_rate": 1.4679579030584517e-05,
+      "loss": 5.4535,
+      "step": 14300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9548550844192505,
+      "learning_rate": 1.4669697119422899e-05,
+      "loss": 5.5077,
+      "step": 14304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7335338592529297,
+      "learning_rate": 1.4659815208261279e-05,
+      "loss": 5.4265,
+      "step": 14308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0178778171539307,
+      "learning_rate": 1.4649933297099661e-05,
+      "loss": 5.3215,
+      "step": 14312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.198841094970703,
+      "learning_rate": 1.4640051385938042e-05,
+      "loss": 5.5392,
+      "step": 14316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.093581438064575,
+      "learning_rate": 1.4630169474776424e-05,
+      "loss": 5.4572,
+      "step": 14320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7876406908035278,
+      "learning_rate": 1.4620287563614804e-05,
+      "loss": 5.3976,
+      "step": 14324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3379602432250977,
+      "learning_rate": 1.4610405652453184e-05,
+      "loss": 5.5972,
+      "step": 14328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0745928287506104,
+      "learning_rate": 1.4600523741291566e-05,
+      "loss": 5.4489,
+      "step": 14332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9868686199188232,
+      "learning_rate": 1.4590641830129947e-05,
+      "loss": 5.5921,
+      "step": 14336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.052441358566284,
+      "learning_rate": 1.4580759918968329e-05,
+      "loss": 5.5621,
+      "step": 14340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.250218629837036,
+      "learning_rate": 1.457087800780671e-05,
+      "loss": 5.3768,
+      "step": 14344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0471935272216797,
+      "learning_rate": 1.4560996096645091e-05,
+      "loss": 5.5863,
+      "step": 14348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0607481002807617,
+      "learning_rate": 1.4551114185483472e-05,
+      "loss": 5.5404,
+      "step": 14352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7724336385726929,
+      "learning_rate": 1.4541232274321854e-05,
+      "loss": 5.4719,
+      "step": 14356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1439015865325928,
+      "learning_rate": 1.4531350363160237e-05,
+      "loss": 5.5045,
+      "step": 14360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.067152500152588,
+      "learning_rate": 1.4521468451998618e-05,
+      "loss": 5.5518,
+      "step": 14364
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8666131496429443,
+      "learning_rate": 1.4511586540837e-05,
+      "loss": 5.4791,
+      "step": 14368
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9369677305221558,
+      "learning_rate": 1.450170462967538e-05,
+      "loss": 5.4379,
+      "step": 14372
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0378055572509766,
+      "learning_rate": 1.4491822718513762e-05,
+      "loss": 5.4874,
+      "step": 14376
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8923200368881226,
+      "learning_rate": 1.4481940807352143e-05,
+      "loss": 5.5813,
+      "step": 14380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9572350978851318,
+      "learning_rate": 1.4472058896190525e-05,
+      "loss": 5.5161,
+      "step": 14384
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9131194353103638,
+      "learning_rate": 1.4462176985028905e-05,
+      "loss": 5.6163,
+      "step": 14388
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5276191234588623,
+      "learning_rate": 1.4452295073867287e-05,
+      "loss": 5.5442,
+      "step": 14392
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0798611640930176,
+      "learning_rate": 1.4442413162705667e-05,
+      "loss": 5.6086,
+      "step": 14396
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1294023990631104,
+      "learning_rate": 1.4432531251544048e-05,
+      "loss": 5.3953,
+      "step": 14400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.01753568649292,
+      "learning_rate": 1.442264934038243e-05,
+      "loss": 5.537,
+      "step": 14404
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0679104328155518,
+      "learning_rate": 1.441276742922081e-05,
+      "loss": 5.4632,
+      "step": 14408
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.921613097190857,
+      "learning_rate": 1.4402885518059192e-05,
+      "loss": 5.2852,
+      "step": 14412
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9771466255187988,
+      "learning_rate": 1.4393003606897576e-05,
+      "loss": 5.5808,
+      "step": 14416
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8879215717315674,
+      "learning_rate": 1.4383121695735956e-05,
+      "loss": 5.4957,
+      "step": 14420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0186474323272705,
+      "learning_rate": 1.4373239784574339e-05,
+      "loss": 5.4864,
+      "step": 14424
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.147420883178711,
+      "learning_rate": 1.4363357873412719e-05,
+      "loss": 5.5098,
+      "step": 14428
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9963911771774292,
+      "learning_rate": 1.4353475962251101e-05,
+      "loss": 5.3886,
+      "step": 14432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03368878364563,
+      "learning_rate": 1.4343594051089481e-05,
+      "loss": 5.4533,
+      "step": 14436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.274022340774536,
+      "learning_rate": 1.4333712139927863e-05,
+      "loss": 5.5222,
+      "step": 14440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0925748348236084,
+      "learning_rate": 1.4323830228766244e-05,
+      "loss": 5.3865,
+      "step": 14444
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25508451461792,
+      "learning_rate": 1.4313948317604626e-05,
+      "loss": 5.388,
+      "step": 14448
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9655667543411255,
+      "learning_rate": 1.4304066406443006e-05,
+      "loss": 5.5274,
+      "step": 14452
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7187186479568481,
+      "learning_rate": 1.4294184495281388e-05,
+      "loss": 5.5024,
+      "step": 14456
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9573662281036377,
+      "learning_rate": 1.4284302584119769e-05,
+      "loss": 5.5709,
+      "step": 14460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9070782661437988,
+      "learning_rate": 1.4274420672958149e-05,
+      "loss": 5.5638,
+      "step": 14464
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.99477219581604,
+      "learning_rate": 1.4264538761796531e-05,
+      "loss": 5.554,
+      "step": 14468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1351230144500732,
+      "learning_rate": 1.4254656850634915e-05,
+      "loss": 5.4401,
+      "step": 14472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.12768292427063,
+      "learning_rate": 1.4244774939473297e-05,
+      "loss": 5.5931,
+      "step": 14476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.834502100944519,
+      "learning_rate": 1.4234893028311677e-05,
+      "loss": 5.3383,
+      "step": 14480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9401111602783203,
+      "learning_rate": 1.4225011117150058e-05,
+      "loss": 5.5206,
+      "step": 14484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1304268836975098,
+      "learning_rate": 1.421512920598844e-05,
+      "loss": 5.4791,
+      "step": 14488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8855656385421753,
+      "learning_rate": 1.420524729482682e-05,
+      "loss": 5.4404,
+      "step": 14492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1601414680480957,
+      "learning_rate": 1.4195365383665202e-05,
+      "loss": 5.509,
+      "step": 14496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.054011821746826,
+      "learning_rate": 1.4185483472503582e-05,
+      "loss": 5.5097,
+      "step": 14500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2468581199645996,
+      "learning_rate": 1.4175601561341964e-05,
+      "loss": 5.5658,
+      "step": 14504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8028780221939087,
+      "learning_rate": 1.4165719650180345e-05,
+      "loss": 5.4528,
+      "step": 14508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7557624578475952,
+      "learning_rate": 1.4155837739018727e-05,
+      "loss": 5.4267,
+      "step": 14512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2908835411071777,
+      "learning_rate": 1.4145955827857107e-05,
+      "loss": 5.5676,
+      "step": 14516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8863370418548584,
+      "learning_rate": 1.413607391669549e-05,
+      "loss": 5.5873,
+      "step": 14520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9381186962127686,
+      "learning_rate": 1.412619200553387e-05,
+      "loss": 5.5573,
+      "step": 14524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9330207109451294,
+      "learning_rate": 1.4116310094372252e-05,
+      "loss": 5.3673,
+      "step": 14528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0471458435058594,
+      "learning_rate": 1.4106428183210636e-05,
+      "loss": 5.4435,
+      "step": 14532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.922804594039917,
+      "learning_rate": 1.4096546272049016e-05,
+      "loss": 5.4511,
+      "step": 14536
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.946153998374939,
+      "learning_rate": 1.4086664360887398e-05,
+      "loss": 5.4641,
+      "step": 14540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0340843200683594,
+      "learning_rate": 1.4076782449725778e-05,
+      "loss": 5.5205,
+      "step": 14544
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.218997001647949,
+      "learning_rate": 1.4066900538564159e-05,
+      "loss": 5.4572,
+      "step": 14548
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8929450511932373,
+      "learning_rate": 1.405701862740254e-05,
+      "loss": 5.4428,
+      "step": 14552
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.828822374343872,
+      "learning_rate": 1.4047136716240921e-05,
+      "loss": 5.4737,
+      "step": 14556
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1129636764526367,
+      "learning_rate": 1.4037254805079303e-05,
+      "loss": 5.4952,
+      "step": 14560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0485777854919434,
+      "learning_rate": 1.4027372893917684e-05,
+      "loss": 5.5007,
+      "step": 14564
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2085585594177246,
+      "learning_rate": 1.4017490982756066e-05,
+      "loss": 5.5036,
+      "step": 14568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0381059646606445,
+      "learning_rate": 1.4007609071594446e-05,
+      "loss": 5.5335,
+      "step": 14572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.994460940361023,
+      "learning_rate": 1.3997727160432828e-05,
+      "loss": 5.5262,
+      "step": 14576
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9527006149291992,
+      "learning_rate": 1.3987845249271208e-05,
+      "loss": 5.4994,
+      "step": 14580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.07161808013916,
+      "learning_rate": 1.397796333810959e-05,
+      "loss": 5.5022,
+      "step": 14584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.083160638809204,
+      "learning_rate": 1.3968081426947974e-05,
+      "loss": 5.46,
+      "step": 14588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9139795303344727,
+      "learning_rate": 1.3958199515786355e-05,
+      "loss": 5.3918,
+      "step": 14592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.016361951828003,
+      "learning_rate": 1.3948317604624737e-05,
+      "loss": 5.3577,
+      "step": 14596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9085208177566528,
+      "learning_rate": 1.3938435693463117e-05,
+      "loss": 5.5672,
+      "step": 14600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.156883716583252,
+      "learning_rate": 1.3928553782301499e-05,
+      "loss": 5.4402,
+      "step": 14604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1130199432373047,
+      "learning_rate": 1.391867187113988e-05,
+      "loss": 5.5375,
+      "step": 14608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0779714584350586,
+      "learning_rate": 1.3908789959978261e-05,
+      "loss": 5.4568,
+      "step": 14612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0961239337921143,
+      "learning_rate": 1.3898908048816642e-05,
+      "loss": 5.5331,
+      "step": 14616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9321538209915161,
+      "learning_rate": 1.3889026137655022e-05,
+      "loss": 5.4607,
+      "step": 14620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171811103820801,
+      "learning_rate": 1.3879144226493404e-05,
+      "loss": 5.5705,
+      "step": 14624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0111582279205322,
+      "learning_rate": 1.3869262315331785e-05,
+      "loss": 5.5365,
+      "step": 14628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2230725288391113,
+      "learning_rate": 1.3859380404170167e-05,
+      "loss": 5.4544,
+      "step": 14632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1781952381134033,
+      "learning_rate": 1.3849498493008547e-05,
+      "loss": 5.5207,
+      "step": 14636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1648108959198,
+      "learning_rate": 1.3839616581846929e-05,
+      "loss": 5.542,
+      "step": 14640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.008898973464966,
+      "learning_rate": 1.3829734670685313e-05,
+      "loss": 5.4819,
+      "step": 14644
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8579167127609253,
+      "learning_rate": 1.3819852759523693e-05,
+      "loss": 5.5307,
+      "step": 14648
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.951011300086975,
+      "learning_rate": 1.3809970848362075e-05,
+      "loss": 5.5083,
+      "step": 14652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.622732639312744,
+      "learning_rate": 1.3800088937200456e-05,
+      "loss": 5.4687,
+      "step": 14656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8241809606552124,
+      "learning_rate": 1.3790207026038838e-05,
+      "loss": 5.5702,
+      "step": 14660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2202184200286865,
+      "learning_rate": 1.3780325114877218e-05,
+      "loss": 5.5717,
+      "step": 14664
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125215530395508,
+      "learning_rate": 1.37704432037156e-05,
+      "loss": 5.4724,
+      "step": 14668
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0341951847076416,
+      "learning_rate": 1.376056129255398e-05,
+      "loss": 5.4859,
+      "step": 14672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9950518608093262,
+      "learning_rate": 1.3750679381392363e-05,
+      "loss": 5.4983,
+      "step": 14676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9778791666030884,
+      "learning_rate": 1.3740797470230743e-05,
+      "loss": 5.3464,
+      "step": 14680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.052849054336548,
+      "learning_rate": 1.3730915559069123e-05,
+      "loss": 5.4766,
+      "step": 14684
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0208585262298584,
+      "learning_rate": 1.3721033647907505e-05,
+      "loss": 5.4662,
+      "step": 14688
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0381686687469482,
+      "learning_rate": 1.3711151736745886e-05,
+      "loss": 5.4048,
+      "step": 14692
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0504868030548096,
+      "learning_rate": 1.3701269825584268e-05,
+      "loss": 5.5422,
+      "step": 14696
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9056216478347778,
+      "learning_rate": 1.3691387914422648e-05,
+      "loss": 5.5829,
+      "step": 14700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1584672927856445,
+      "learning_rate": 1.3681506003261032e-05,
+      "loss": 5.4341,
+      "step": 14704
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8544187545776367,
+      "learning_rate": 1.3671624092099414e-05,
+      "loss": 5.5552,
+      "step": 14708
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9360532760620117,
+      "learning_rate": 1.3661742180937794e-05,
+      "loss": 5.413,
+      "step": 14712
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7825835943222046,
+      "learning_rate": 1.3651860269776176e-05,
+      "loss": 5.4916,
+      "step": 14716
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0297420024871826,
+      "learning_rate": 1.3641978358614557e-05,
+      "loss": 5.4825,
+      "step": 14720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0158278942108154,
+      "learning_rate": 1.3632096447452939e-05,
+      "loss": 5.4502,
+      "step": 14724
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265287160873413,
+      "learning_rate": 1.362221453629132e-05,
+      "loss": 5.5644,
+      "step": 14728
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0918755531311035,
+      "learning_rate": 1.3612332625129701e-05,
+      "loss": 5.6084,
+      "step": 14732
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.141103744506836,
+      "learning_rate": 1.3602450713968082e-05,
+      "loss": 5.4327,
+      "step": 14736
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.070760488510132,
+      "learning_rate": 1.3592568802806464e-05,
+      "loss": 5.5944,
+      "step": 14740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9426864385604858,
+      "learning_rate": 1.3582686891644844e-05,
+      "loss": 5.4326,
+      "step": 14744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8541244268417358,
+      "learning_rate": 1.3572804980483224e-05,
+      "loss": 5.5438,
+      "step": 14748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.081083297729492,
+      "learning_rate": 1.3562923069321606e-05,
+      "loss": 5.5656,
+      "step": 14752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.923309326171875,
+      "learning_rate": 1.3553041158159987e-05,
+      "loss": 5.5546,
+      "step": 14756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.746881365776062,
+      "learning_rate": 1.3543159246998372e-05,
+      "loss": 5.4799,
+      "step": 14760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.91403067111969,
+      "learning_rate": 1.3533277335836753e-05,
+      "loss": 5.4892,
+      "step": 14764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0929675102233887,
+      "learning_rate": 1.3523395424675133e-05,
+      "loss": 5.5852,
+      "step": 14768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9422674179077148,
+      "learning_rate": 1.3513513513513515e-05,
+      "loss": 5.5581,
+      "step": 14772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.684131383895874,
+      "learning_rate": 1.3503631602351895e-05,
+      "loss": 5.553,
+      "step": 14776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8600718975067139,
+      "learning_rate": 1.3493749691190278e-05,
+      "loss": 5.5774,
+      "step": 14780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7786033153533936,
+      "learning_rate": 1.3483867780028658e-05,
+      "loss": 5.4997,
+      "step": 14784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.239494562149048,
+      "learning_rate": 1.347398586886704e-05,
+      "loss": 5.6186,
+      "step": 14788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078833818435669,
+      "learning_rate": 1.346410395770542e-05,
+      "loss": 5.5609,
+      "step": 14792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.867530107498169,
+      "learning_rate": 1.3454222046543802e-05,
+      "loss": 5.5624,
+      "step": 14796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8725926876068115,
+      "learning_rate": 1.3444340135382183e-05,
+      "loss": 5.5675,
+      "step": 14800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9592565298080444,
+      "learning_rate": 1.3434458224220565e-05,
+      "loss": 5.4278,
+      "step": 14804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0653085708618164,
+      "learning_rate": 1.3424576313058945e-05,
+      "loss": 5.4732,
+      "step": 14808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0164620876312256,
+      "learning_rate": 1.3414694401897327e-05,
+      "loss": 5.4636,
+      "step": 14812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9894605875015259,
+      "learning_rate": 1.3404812490735708e-05,
+      "loss": 5.5102,
+      "step": 14816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0743892192840576,
+      "learning_rate": 1.3394930579574091e-05,
+      "loss": 5.4852,
+      "step": 14820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.000945806503296,
+      "learning_rate": 1.3385048668412473e-05,
+      "loss": 5.4698,
+      "step": 14824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.219153881072998,
+      "learning_rate": 1.3375166757250854e-05,
+      "loss": 5.5966,
+      "step": 14828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0266921520233154,
+      "learning_rate": 1.3365284846089234e-05,
+      "loss": 5.5036,
+      "step": 14832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1432504653930664,
+      "learning_rate": 1.3355402934927616e-05,
+      "loss": 5.5122,
+      "step": 14836
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.975110411643982,
+      "learning_rate": 1.3345521023765997e-05,
+      "loss": 5.4677,
+      "step": 14840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1612823009490967,
+      "learning_rate": 1.3335639112604379e-05,
+      "loss": 5.544,
+      "step": 14844
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0813982486724854,
+      "learning_rate": 1.3325757201442759e-05,
+      "loss": 5.3651,
+      "step": 14848
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8772978782653809,
+      "learning_rate": 1.3315875290281141e-05,
+      "loss": 5.4319,
+      "step": 14852
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8763823509216309,
+      "learning_rate": 1.3305993379119521e-05,
+      "loss": 5.6508,
+      "step": 14856
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0054590702056885,
+      "learning_rate": 1.3296111467957903e-05,
+      "loss": 5.4899,
+      "step": 14860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0155956745147705,
+      "learning_rate": 1.3286229556796284e-05,
+      "loss": 5.4597,
+      "step": 14864
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9215456247329712,
+      "learning_rate": 1.3276347645634666e-05,
+      "loss": 5.5587,
+      "step": 14868
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2192320823669434,
+      "learning_rate": 1.3266465734473046e-05,
+      "loss": 5.5397,
+      "step": 14872
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8037538528442383,
+      "learning_rate": 1.325658382331143e-05,
+      "loss": 5.503,
+      "step": 14876
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.096147298812866,
+      "learning_rate": 1.3246701912149812e-05,
+      "loss": 5.4808,
+      "step": 14880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0961217880249023,
+      "learning_rate": 1.3236820000988192e-05,
+      "loss": 5.5546,
+      "step": 14884
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1718828678131104,
+      "learning_rate": 1.3226938089826574e-05,
+      "loss": 5.4201,
+      "step": 14888
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.062138795852661,
+      "learning_rate": 1.3217056178664955e-05,
+      "loss": 5.46,
+      "step": 14892
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0585410594940186,
+      "learning_rate": 1.3207174267503337e-05,
+      "loss": 5.5545,
+      "step": 14896
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8028877973556519,
+      "learning_rate": 1.3197292356341717e-05,
+      "loss": 5.4255,
+      "step": 14900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.002005100250244,
+      "learning_rate": 1.3187410445180098e-05,
+      "loss": 5.5847,
+      "step": 14904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1855287551879883,
+      "learning_rate": 1.317752853401848e-05,
+      "loss": 5.536,
+      "step": 14908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.193171977996826,
+      "learning_rate": 1.316764662285686e-05,
+      "loss": 5.4059,
+      "step": 14912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0406908988952637,
+      "learning_rate": 1.3157764711695242e-05,
+      "loss": 5.5851,
+      "step": 14916
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.342682123184204,
+      "learning_rate": 1.3147882800533623e-05,
+      "loss": 5.5068,
+      "step": 14920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0584092140197754,
+      "learning_rate": 1.3138000889372005e-05,
+      "loss": 5.3973,
+      "step": 14924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7771698236465454,
+      "learning_rate": 1.3128118978210385e-05,
+      "loss": 5.4812,
+      "step": 14928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.923229694366455,
+      "learning_rate": 1.3118237067048769e-05,
+      "loss": 5.4176,
+      "step": 14932
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9571110010147095,
+      "learning_rate": 1.310835515588715e-05,
+      "loss": 5.5085,
+      "step": 14936
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9827457666397095,
+      "learning_rate": 1.3098473244725531e-05,
+      "loss": 5.4454,
+      "step": 14940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2986981868743896,
+      "learning_rate": 1.3088591333563913e-05,
+      "loss": 5.562,
+      "step": 14944
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.087019920349121,
+      "learning_rate": 1.3078709422402294e-05,
+      "loss": 5.5164,
+      "step": 14948
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9258410930633545,
+      "learning_rate": 1.3068827511240676e-05,
+      "loss": 5.5252,
+      "step": 14952
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.833736538887024,
+      "learning_rate": 1.3058945600079056e-05,
+      "loss": 5.4423,
+      "step": 14956
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1392202377319336,
+      "learning_rate": 1.3049063688917438e-05,
+      "loss": 5.5637,
+      "step": 14960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9656065702438354,
+      "learning_rate": 1.3039181777755818e-05,
+      "loss": 5.474,
+      "step": 14964
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8016414642333984,
+      "learning_rate": 1.3029299866594199e-05,
+      "loss": 5.4101,
+      "step": 14968
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.141220808029175,
+      "learning_rate": 1.301941795543258e-05,
+      "loss": 5.4747,
+      "step": 14972
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1080868244171143,
+      "learning_rate": 1.3009536044270961e-05,
+      "loss": 5.3842,
+      "step": 14976
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9357255697250366,
+      "learning_rate": 1.2999654133109343e-05,
+      "loss": 5.5716,
+      "step": 14980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3347742557525635,
+      "learning_rate": 1.2989772221947724e-05,
+      "loss": 5.5935,
+      "step": 14984
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.999286413192749,
+      "learning_rate": 1.2979890310786106e-05,
+      "loss": 5.4976,
+      "step": 14988
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.040515184402466,
+      "learning_rate": 1.297000839962449e-05,
+      "loss": 5.6029,
+      "step": 14992
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9321497678756714,
+      "learning_rate": 1.296012648846287e-05,
+      "loss": 5.503,
+      "step": 14996
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.834435224533081,
+      "learning_rate": 1.2950244577301252e-05,
+      "loss": 5.5076,
+      "step": 15000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.410569906234741,
+      "learning_rate": 1.2940362666139632e-05,
+      "loss": 5.6112,
+      "step": 15004
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.110943555831909,
+      "learning_rate": 1.2930480754978014e-05,
+      "loss": 5.4589,
+      "step": 15008
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2104272842407227,
+      "learning_rate": 1.2920598843816395e-05,
+      "loss": 5.4494,
+      "step": 15012
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0038228034973145,
+      "learning_rate": 1.2910716932654777e-05,
+      "loss": 5.4229,
+      "step": 15016
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0080738067626953,
+      "learning_rate": 1.2900835021493157e-05,
+      "loss": 5.5645,
+      "step": 15020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.075934886932373,
+      "learning_rate": 1.2890953110331539e-05,
+      "loss": 5.6104,
+      "step": 15024
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9630203247070312,
+      "learning_rate": 1.288107119916992e-05,
+      "loss": 5.4888,
+      "step": 15028
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1975886821746826,
+      "learning_rate": 1.28711892880083e-05,
+      "loss": 5.5283,
+      "step": 15032
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9507403373718262,
+      "learning_rate": 1.2861307376846682e-05,
+      "loss": 5.6203,
+      "step": 15036
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0065085887908936,
+      "learning_rate": 1.2851425465685062e-05,
+      "loss": 5.4493,
+      "step": 15040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.170250177383423,
+      "learning_rate": 1.2841543554523444e-05,
+      "loss": 5.5062,
+      "step": 15044
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9755939245224,
+      "learning_rate": 1.2831661643361828e-05,
+      "loss": 5.4901,
+      "step": 15048
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0391461849212646,
+      "learning_rate": 1.2821779732200208e-05,
+      "loss": 5.4896,
+      "step": 15052
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8779929876327515,
+      "learning_rate": 1.281189782103859e-05,
+      "loss": 5.4311,
+      "step": 15056
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.123504400253296,
+      "learning_rate": 1.2802015909876971e-05,
+      "loss": 5.5117,
+      "step": 15060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.885321021080017,
+      "learning_rate": 1.2792133998715353e-05,
+      "loss": 5.515,
+      "step": 15064
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.281585454940796,
+      "learning_rate": 1.2782252087553733e-05,
+      "loss": 5.4979,
+      "step": 15068
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1277706623077393,
+      "learning_rate": 1.2772370176392115e-05,
+      "loss": 5.4285,
+      "step": 15072
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0085339546203613,
+      "learning_rate": 1.2762488265230496e-05,
+      "loss": 5.4741,
+      "step": 15076
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0528817176818848,
+      "learning_rate": 1.2752606354068878e-05,
+      "loss": 5.5861,
+      "step": 15080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0713131427764893,
+      "learning_rate": 1.2742724442907258e-05,
+      "loss": 5.5236,
+      "step": 15084
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9570958614349365,
+      "learning_rate": 1.273284253174564e-05,
+      "loss": 5.4983,
+      "step": 15088
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8373630046844482,
+      "learning_rate": 1.272296062058402e-05,
+      "loss": 5.4913,
+      "step": 15092
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0442957878112793,
+      "learning_rate": 1.2713078709422403e-05,
+      "loss": 5.4757,
+      "step": 15096
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.090407609939575,
+      "learning_rate": 1.2703196798260783e-05,
+      "loss": 5.4915,
+      "step": 15100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1005630493164062,
+      "learning_rate": 1.2693314887099167e-05,
+      "loss": 5.5734,
+      "step": 15104
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.062366008758545,
+      "learning_rate": 1.2683432975937549e-05,
+      "loss": 5.5959,
+      "step": 15108
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9203088283538818,
+      "learning_rate": 1.267355106477593e-05,
+      "loss": 5.4388,
+      "step": 15112
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0604312419891357,
+      "learning_rate": 1.266366915361431e-05,
+      "loss": 5.5115,
+      "step": 15116
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.230130672454834,
+      "learning_rate": 1.2653787242452692e-05,
+      "loss": 5.596,
+      "step": 15120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0392181873321533,
+      "learning_rate": 1.2643905331291072e-05,
+      "loss": 5.57,
+      "step": 15124
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.253793954849243,
+      "learning_rate": 1.2634023420129454e-05,
+      "loss": 5.5402,
+      "step": 15128
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9613460302352905,
+      "learning_rate": 1.2624141508967834e-05,
+      "loss": 5.3467,
+      "step": 15132
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8354169130325317,
+      "learning_rate": 1.2614259597806216e-05,
+      "loss": 5.5377,
+      "step": 15136
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8670376539230347,
+      "learning_rate": 1.2604377686644597e-05,
+      "loss": 5.4526,
+      "step": 15140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1439993381500244,
+      "learning_rate": 1.2594495775482979e-05,
+      "loss": 5.4915,
+      "step": 15144
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9621843099594116,
+      "learning_rate": 1.258461386432136e-05,
+      "loss": 5.415,
+      "step": 15148
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0130698680877686,
+      "learning_rate": 1.2574731953159741e-05,
+      "loss": 5.5646,
+      "step": 15152
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1963608264923096,
+      "learning_rate": 1.2564850041998122e-05,
+      "loss": 5.5678,
+      "step": 15156
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.007394790649414,
+      "learning_rate": 1.2554968130836504e-05,
+      "loss": 5.4874,
+      "step": 15160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03348708152771,
+      "learning_rate": 1.2545086219674888e-05,
+      "loss": 5.3595,
+      "step": 15164
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1427929401397705,
+      "learning_rate": 1.2535204308513268e-05,
+      "loss": 5.5416,
+      "step": 15168
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0188114643096924,
+      "learning_rate": 1.252532239735165e-05,
+      "loss": 5.4923,
+      "step": 15172
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9621504545211792,
+      "learning_rate": 1.251544048619003e-05,
+      "loss": 5.5167,
+      "step": 15176
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1401383876800537,
+      "learning_rate": 1.2505558575028412e-05,
+      "loss": 5.3883,
+      "step": 15180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0091779232025146,
+      "learning_rate": 1.2495676663866793e-05,
+      "loss": 5.5855,
+      "step": 15184
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0588550567626953,
+      "learning_rate": 1.2485794752705173e-05,
+      "loss": 5.4961,
+      "step": 15188
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0023372173309326,
+      "learning_rate": 1.2475912841543555e-05,
+      "loss": 5.3909,
+      "step": 15192
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.395747184753418,
+      "learning_rate": 1.2466030930381936e-05,
+      "loss": 5.4807,
+      "step": 15196
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8098511695861816,
+      "learning_rate": 1.2456149019220318e-05,
+      "loss": 5.4387,
+      "step": 15200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9918867349624634,
+      "learning_rate": 1.24462671080587e-05,
+      "loss": 5.396,
+      "step": 15204
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0140202045440674,
+      "learning_rate": 1.2436385196897082e-05,
+      "loss": 5.4098,
+      "step": 15208
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9749282598495483,
+      "learning_rate": 1.2426503285735462e-05,
+      "loss": 5.4362,
+      "step": 15212
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1992135047912598,
+      "learning_rate": 1.2416621374573842e-05,
+      "loss": 5.5986,
+      "step": 15216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3077468872070312,
+      "learning_rate": 1.2406739463412225e-05,
+      "loss": 5.5058,
+      "step": 15220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.964931845664978,
+      "learning_rate": 1.2396857552250605e-05,
+      "loss": 5.3997,
+      "step": 15224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9411903619766235,
+      "learning_rate": 1.2386975641088987e-05,
+      "loss": 5.4008,
+      "step": 15228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.01727557182312,
+      "learning_rate": 1.2377093729927367e-05,
+      "loss": 5.6187,
+      "step": 15232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9451463222503662,
+      "learning_rate": 1.2367211818765751e-05,
+      "loss": 5.5272,
+      "step": 15236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1012978553771973,
+      "learning_rate": 1.2357329907604131e-05,
+      "loss": 5.3657,
+      "step": 15240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.219510078430176,
+      "learning_rate": 1.2347447996442513e-05,
+      "loss": 5.526,
+      "step": 15244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.013444185256958,
+      "learning_rate": 1.2337566085280894e-05,
+      "loss": 5.4749,
+      "step": 15248
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7549456357955933,
+      "learning_rate": 1.2327684174119274e-05,
+      "loss": 5.3478,
+      "step": 15252
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8471717834472656,
+      "learning_rate": 1.2317802262957656e-05,
+      "loss": 5.5201,
+      "step": 15256
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.324028491973877,
+      "learning_rate": 1.2307920351796037e-05,
+      "loss": 5.5589,
+      "step": 15260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0255486965179443,
+      "learning_rate": 1.229803844063442e-05,
+      "loss": 5.5575,
+      "step": 15264
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1873011589050293,
+      "learning_rate": 1.22881565294728e-05,
+      "loss": 5.5165,
+      "step": 15268
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1346065998077393,
+      "learning_rate": 1.2278274618311183e-05,
+      "loss": 5.5053,
+      "step": 15272
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1076815128326416,
+      "learning_rate": 1.2268392707149563e-05,
+      "loss": 5.4105,
+      "step": 15276
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9122978448867798,
+      "learning_rate": 1.2258510795987945e-05,
+      "loss": 5.4235,
+      "step": 15280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.054979085922241,
+      "learning_rate": 1.2248628884826326e-05,
+      "loss": 5.5838,
+      "step": 15284
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9995005130767822,
+      "learning_rate": 1.2238746973664706e-05,
+      "loss": 5.4584,
+      "step": 15288
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0757248401641846,
+      "learning_rate": 1.222886506250309e-05,
+      "loss": 5.5064,
+      "step": 15292
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.831465721130371,
+      "learning_rate": 1.221898315134147e-05,
+      "loss": 5.5416,
+      "step": 15296
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3364105224609375,
+      "learning_rate": 1.2209101240179852e-05,
+      "loss": 5.5332,
+      "step": 15300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9546422958374023,
+      "learning_rate": 1.2199219329018233e-05,
+      "loss": 5.4726,
+      "step": 15304
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8395370244979858,
+      "learning_rate": 1.2189337417856615e-05,
+      "loss": 5.4443,
+      "step": 15308
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.080458402633667,
+      "learning_rate": 1.2179455506694995e-05,
+      "loss": 5.5154,
+      "step": 15312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.017116069793701,
+      "learning_rate": 1.2169573595533375e-05,
+      "loss": 5.5048,
+      "step": 15316
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0325398445129395,
+      "learning_rate": 1.2159691684371759e-05,
+      "loss": 5.3668,
+      "step": 15320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.246100664138794,
+      "learning_rate": 1.214980977321014e-05,
+      "loss": 5.5176,
+      "step": 15324
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0287086963653564,
+      "learning_rate": 1.2139927862048522e-05,
+      "loss": 5.5352,
+      "step": 15328
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.278211832046509,
+      "learning_rate": 1.2130045950886902e-05,
+      "loss": 5.441,
+      "step": 15332
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.143902063369751,
+      "learning_rate": 1.2120164039725284e-05,
+      "loss": 5.5263,
+      "step": 15336
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0122079849243164,
+      "learning_rate": 1.2110282128563664e-05,
+      "loss": 5.3925,
+      "step": 15340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.094726324081421,
+      "learning_rate": 1.2100400217402046e-05,
+      "loss": 5.4837,
+      "step": 15344
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0904650688171387,
+      "learning_rate": 1.2090518306240428e-05,
+      "loss": 5.4732,
+      "step": 15348
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.869597315788269,
+      "learning_rate": 1.2080636395078809e-05,
+      "loss": 5.4292,
+      "step": 15352
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0959110260009766,
+      "learning_rate": 1.2070754483917191e-05,
+      "loss": 5.5033,
+      "step": 15356
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.153909921646118,
+      "learning_rate": 1.2060872572755571e-05,
+      "loss": 5.4692,
+      "step": 15360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0445823669433594,
+      "learning_rate": 1.2050990661593953e-05,
+      "loss": 5.3999,
+      "step": 15364
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0181336402893066,
+      "learning_rate": 1.2041108750432334e-05,
+      "loss": 5.4045,
+      "step": 15368
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.008654832839966,
+      "learning_rate": 1.2031226839270716e-05,
+      "loss": 5.6017,
+      "step": 15372
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.126469850540161,
+      "learning_rate": 1.2021344928109096e-05,
+      "loss": 5.4578,
+      "step": 15376
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.031398057937622,
+      "learning_rate": 1.2011463016947478e-05,
+      "loss": 5.5091,
+      "step": 15380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.102151870727539,
+      "learning_rate": 1.200158110578586e-05,
+      "loss": 5.4637,
+      "step": 15384
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9239962100982666,
+      "learning_rate": 1.199169919462424e-05,
+      "loss": 5.5014,
+      "step": 15388
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9613525867462158,
+      "learning_rate": 1.1981817283462623e-05,
+      "loss": 5.5283,
+      "step": 15392
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.973003625869751,
+      "learning_rate": 1.1971935372301003e-05,
+      "loss": 5.4865,
+      "step": 15396
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3486452102661133,
+      "learning_rate": 1.1962053461139385e-05,
+      "loss": 5.5125,
+      "step": 15400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.088740825653076,
+      "learning_rate": 1.1952171549977765e-05,
+      "loss": 5.4416,
+      "step": 15404
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9701464176177979,
+      "learning_rate": 1.1942289638816147e-05,
+      "loss": 5.2871,
+      "step": 15408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2388153076171875,
+      "learning_rate": 1.193240772765453e-05,
+      "loss": 5.551,
+      "step": 15412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2453413009643555,
+      "learning_rate": 1.192252581649291e-05,
+      "loss": 5.4372,
+      "step": 15416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1105456352233887,
+      "learning_rate": 1.1912643905331292e-05,
+      "loss": 5.5284,
+      "step": 15420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.197547197341919,
+      "learning_rate": 1.1902761994169672e-05,
+      "loss": 5.5067,
+      "step": 15424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9585908651351929,
+      "learning_rate": 1.1892880083008054e-05,
+      "loss": 5.4292,
+      "step": 15428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.037917137145996,
+      "learning_rate": 1.1882998171846435e-05,
+      "loss": 5.455,
+      "step": 15432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.797452449798584,
+      "learning_rate": 1.1873116260684817e-05,
+      "loss": 5.4833,
+      "step": 15436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8614048957824707,
+      "learning_rate": 1.1863234349523199e-05,
+      "loss": 5.5127,
+      "step": 15440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9821441173553467,
+      "learning_rate": 1.185335243836158e-05,
+      "loss": 5.5038,
+      "step": 15444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3709757328033447,
+      "learning_rate": 1.1843470527199961e-05,
+      "loss": 5.4745,
+      "step": 15448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8125073909759521,
+      "learning_rate": 1.1833588616038342e-05,
+      "loss": 5.4642,
+      "step": 15452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0087850093841553,
+      "learning_rate": 1.1823706704876724e-05,
+      "loss": 5.4628,
+      "step": 15456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9169509410858154,
+      "learning_rate": 1.1813824793715104e-05,
+      "loss": 5.4821,
+      "step": 15460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.01556396484375,
+      "learning_rate": 1.1803942882553488e-05,
+      "loss": 5.4142,
+      "step": 15464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9918155670166016,
+      "learning_rate": 1.1794060971391868e-05,
+      "loss": 5.6412,
+      "step": 15468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.052454710006714,
+      "learning_rate": 1.1784179060230249e-05,
+      "loss": 5.4347,
+      "step": 15472
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8537468910217285,
+      "learning_rate": 1.177429714906863e-05,
+      "loss": 5.5036,
+      "step": 15476
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9864877462387085,
+      "learning_rate": 1.1764415237907011e-05,
+      "loss": 5.4979,
+      "step": 15480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.200070858001709,
+      "learning_rate": 1.1754533326745393e-05,
+      "loss": 5.5684,
+      "step": 15484
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7679542303085327,
+      "learning_rate": 1.1744651415583773e-05,
+      "loss": 5.4283,
+      "step": 15488
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8458738327026367,
+      "learning_rate": 1.1734769504422157e-05,
+      "loss": 5.5566,
+      "step": 15492
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3362741470336914,
+      "learning_rate": 1.1724887593260538e-05,
+      "loss": 5.4377,
+      "step": 15496
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7595479488372803,
+      "learning_rate": 1.1715005682098918e-05,
+      "loss": 5.4785,
+      "step": 15500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1757829189300537,
+      "learning_rate": 1.17051237709373e-05,
+      "loss": 5.5964,
+      "step": 15504
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0845088958740234,
+      "learning_rate": 1.169524185977568e-05,
+      "loss": 5.5294,
+      "step": 15508
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.937070608139038,
+      "learning_rate": 1.1685359948614062e-05,
+      "loss": 5.5338,
+      "step": 15512
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.137470006942749,
+      "learning_rate": 1.1675478037452443e-05,
+      "loss": 5.4788,
+      "step": 15516
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1320180892944336,
+      "learning_rate": 1.1665596126290827e-05,
+      "loss": 5.6173,
+      "step": 15520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0288352966308594,
+      "learning_rate": 1.1655714215129207e-05,
+      "loss": 5.3012,
+      "step": 15524
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.784881830215454,
+      "learning_rate": 1.1645832303967589e-05,
+      "loss": 5.5805,
+      "step": 15528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.08077073097229,
+      "learning_rate": 1.163595039280597e-05,
+      "loss": 5.4992,
+      "step": 15532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7977101802825928,
+      "learning_rate": 1.162606848164435e-05,
+      "loss": 5.4872,
+      "step": 15536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.186459541320801,
+      "learning_rate": 1.1616186570482732e-05,
+      "loss": 5.578,
+      "step": 15540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.183048725128174,
+      "learning_rate": 1.1606304659321112e-05,
+      "loss": 5.4738,
+      "step": 15544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8500030040740967,
+      "learning_rate": 1.1596422748159494e-05,
+      "loss": 5.5345,
+      "step": 15548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8691611289978027,
+      "learning_rate": 1.1586540836997876e-05,
+      "loss": 5.5283,
+      "step": 15552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1216142177581787,
+      "learning_rate": 1.1576658925836258e-05,
+      "loss": 5.581,
+      "step": 15556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0468342304229736,
+      "learning_rate": 1.1566777014674639e-05,
+      "loss": 5.5649,
+      "step": 15560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.060667037963867,
+      "learning_rate": 1.155689510351302e-05,
+      "loss": 5.4901,
+      "step": 15564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.803126573562622,
+      "learning_rate": 1.1547013192351401e-05,
+      "loss": 5.4285,
+      "step": 15568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.05906081199646,
+      "learning_rate": 1.1537131281189781e-05,
+      "loss": 5.4268,
+      "step": 15572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0089757442474365,
+      "learning_rate": 1.1527249370028164e-05,
+      "loss": 5.5307,
+      "step": 15576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9152356386184692,
+      "learning_rate": 1.1517367458866546e-05,
+      "loss": 5.4297,
+      "step": 15580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9012370109558105,
+      "learning_rate": 1.1507485547704928e-05,
+      "loss": 5.5271,
+      "step": 15584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0786564350128174,
+      "learning_rate": 1.1497603636543308e-05,
+      "loss": 5.5957,
+      "step": 15588
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.03715443611145,
+      "learning_rate": 1.148772172538169e-05,
+      "loss": 5.5501,
+      "step": 15592
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1424834728240967,
+      "learning_rate": 1.147783981422007e-05,
+      "loss": 5.5153,
+      "step": 15596
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9324986934661865,
+      "learning_rate": 1.146795790305845e-05,
+      "loss": 5.5928,
+      "step": 15600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.986244559288025,
+      "learning_rate": 1.1458075991896833e-05,
+      "loss": 5.4466,
+      "step": 15604
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3215315341949463,
+      "learning_rate": 1.1448194080735215e-05,
+      "loss": 5.4948,
+      "step": 15608
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2035694122314453,
+      "learning_rate": 1.1438312169573597e-05,
+      "loss": 5.6002,
+      "step": 15612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0092086791992188,
+      "learning_rate": 1.1428430258411977e-05,
+      "loss": 5.4083,
+      "step": 15616
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.008274555206299,
+      "learning_rate": 1.141854834725036e-05,
+      "loss": 5.3977,
+      "step": 15620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.084843873977661,
+      "learning_rate": 1.140866643608874e-05,
+      "loss": 5.4896,
+      "step": 15624
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.138129949569702,
+      "learning_rate": 1.1398784524927122e-05,
+      "loss": 5.556,
+      "step": 15628
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.161590814590454,
+      "learning_rate": 1.1388902613765502e-05,
+      "loss": 5.4909,
+      "step": 15632
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1914632320404053,
+      "learning_rate": 1.1379020702603884e-05,
+      "loss": 5.5578,
+      "step": 15636
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.254403829574585,
+      "learning_rate": 1.1369138791442266e-05,
+      "loss": 5.6521,
+      "step": 15640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9875483512878418,
+      "learning_rate": 1.1359256880280647e-05,
+      "loss": 5.4972,
+      "step": 15644
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2201669216156006,
+      "learning_rate": 1.1349374969119029e-05,
+      "loss": 5.6261,
+      "step": 15648
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.054435968399048,
+      "learning_rate": 1.1339493057957409e-05,
+      "loss": 5.5603,
+      "step": 15652
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.902565598487854,
+      "learning_rate": 1.1329611146795791e-05,
+      "loss": 5.6618,
+      "step": 15656
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2018725872039795,
+      "learning_rate": 1.1319729235634172e-05,
+      "loss": 5.429,
+      "step": 15660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9127063751220703,
+      "learning_rate": 1.1309847324472554e-05,
+      "loss": 5.5171,
+      "step": 15664
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8942164182662964,
+      "learning_rate": 1.1299965413310936e-05,
+      "loss": 5.5357,
+      "step": 15668
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2651684284210205,
+      "learning_rate": 1.1290083502149316e-05,
+      "loss": 5.4591,
+      "step": 15672
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.189774990081787,
+      "learning_rate": 1.1280201590987698e-05,
+      "loss": 5.4338,
+      "step": 15676
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1134681701660156,
+      "learning_rate": 1.1270319679826078e-05,
+      "loss": 5.5294,
+      "step": 15680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2125091552734375,
+      "learning_rate": 1.126043776866446e-05,
+      "loss": 5.5269,
+      "step": 15684
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9919397830963135,
+      "learning_rate": 1.1250555857502841e-05,
+      "loss": 5.4072,
+      "step": 15688
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.958975911140442,
+      "learning_rate": 1.1240673946341223e-05,
+      "loss": 5.4421,
+      "step": 15692
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.033118724822998,
+      "learning_rate": 1.1230792035179605e-05,
+      "loss": 5.5095,
+      "step": 15696
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1201207637786865,
+      "learning_rate": 1.1220910124017985e-05,
+      "loss": 5.4767,
+      "step": 15700
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9773985147476196,
+      "learning_rate": 1.1211028212856367e-05,
+      "loss": 5.424,
+      "step": 15704
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.841871976852417,
+      "learning_rate": 1.1201146301694748e-05,
+      "loss": 5.521,
+      "step": 15708
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9351176023483276,
+      "learning_rate": 1.119126439053313e-05,
+      "loss": 5.5646,
+      "step": 15712
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0061535835266113,
+      "learning_rate": 1.118138247937151e-05,
+      "loss": 5.36,
+      "step": 15716
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4020817279815674,
+      "learning_rate": 1.1171500568209892e-05,
+      "loss": 5.4701,
+      "step": 15720
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9213863611221313,
+      "learning_rate": 1.1161618657048274e-05,
+      "loss": 5.4668,
+      "step": 15724
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3538432121276855,
+      "learning_rate": 1.1151736745886655e-05,
+      "loss": 5.5128,
+      "step": 15728
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.147163152694702,
+      "learning_rate": 1.1141854834725037e-05,
+      "loss": 5.4579,
+      "step": 15732
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1335911750793457,
+      "learning_rate": 1.1131972923563417e-05,
+      "loss": 5.4277,
+      "step": 15736
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1882131099700928,
+      "learning_rate": 1.11220910124018e-05,
+      "loss": 5.6003,
+      "step": 15740
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9644808769226074,
+      "learning_rate": 1.111220910124018e-05,
+      "loss": 5.5355,
+      "step": 15744
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7521815299987793,
+      "learning_rate": 1.1102327190078562e-05,
+      "loss": 5.4743,
+      "step": 15748
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1638858318328857,
+      "learning_rate": 1.1092445278916944e-05,
+      "loss": 5.3701,
+      "step": 15752
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8966166973114014,
+      "learning_rate": 1.1082563367755324e-05,
+      "loss": 5.5355,
+      "step": 15756
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.981217861175537,
+      "learning_rate": 1.1072681456593706e-05,
+      "loss": 5.4366,
+      "step": 15760
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0059125423431396,
+      "learning_rate": 1.1062799545432086e-05,
+      "loss": 5.4749,
+      "step": 15764
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.042475461959839,
+      "learning_rate": 1.1052917634270469e-05,
+      "loss": 5.4066,
+      "step": 15768
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1431596279144287,
+      "learning_rate": 1.1043035723108849e-05,
+      "loss": 5.4831,
+      "step": 15772
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.903968095779419,
+      "learning_rate": 1.1033153811947231e-05,
+      "loss": 5.4749,
+      "step": 15776
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2087221145629883,
+      "learning_rate": 1.1023271900785613e-05,
+      "loss": 5.5542,
+      "step": 15780
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.990248680114746,
+      "learning_rate": 1.1013389989623993e-05,
+      "loss": 5.4592,
+      "step": 15784
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9822028875350952,
+      "learning_rate": 1.1003508078462375e-05,
+      "loss": 5.5245,
+      "step": 15788
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0607571601867676,
+      "learning_rate": 1.0993626167300756e-05,
+      "loss": 5.5105,
+      "step": 15792
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2510719299316406,
+      "learning_rate": 1.0983744256139138e-05,
+      "loss": 5.4485,
+      "step": 15796
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9407929182052612,
+      "learning_rate": 1.0973862344977518e-05,
+      "loss": 5.4835,
+      "step": 15800
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0228731632232666,
+      "learning_rate": 1.09639804338159e-05,
+      "loss": 5.4121,
+      "step": 15804
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0110514163970947,
+      "learning_rate": 1.0954098522654282e-05,
+      "loss": 5.4097,
+      "step": 15808
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.005176305770874,
+      "learning_rate": 1.0944216611492664e-05,
+      "loss": 5.5651,
+      "step": 15812
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.08233380317688,
+      "learning_rate": 1.0934334700331045e-05,
+      "loss": 5.5942,
+      "step": 15816
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.76272714138031,
+      "learning_rate": 1.0924452789169425e-05,
+      "loss": 5.3883,
+      "step": 15820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8395804166793823,
+      "learning_rate": 1.0914570878007807e-05,
+      "loss": 5.372,
+      "step": 15824
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.197016954421997,
+      "learning_rate": 1.0904688966846188e-05,
+      "loss": 5.4239,
+      "step": 15828
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1127779483795166,
+      "learning_rate": 1.089480705568457e-05,
+      "loss": 5.358,
+      "step": 15832
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1906561851501465,
+      "learning_rate": 1.0884925144522952e-05,
+      "loss": 5.528,
+      "step": 15836
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2963180541992188,
+      "learning_rate": 1.0875043233361334e-05,
+      "loss": 5.6399,
+      "step": 15840
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1306509971618652,
+      "learning_rate": 1.0865161322199714e-05,
+      "loss": 5.4986,
+      "step": 15844
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8235065937042236,
+      "learning_rate": 1.0855279411038096e-05,
+      "loss": 5.5502,
+      "step": 15848
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.058922529220581,
+      "learning_rate": 1.0845397499876477e-05,
+      "loss": 5.5531,
+      "step": 15852
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.818724274635315,
+      "learning_rate": 1.0835515588714857e-05,
+      "loss": 5.4607,
+      "step": 15856
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0503363609313965,
+      "learning_rate": 1.0825633677553239e-05,
+      "loss": 5.4722,
+      "step": 15860
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.076927423477173,
+      "learning_rate": 1.081575176639162e-05,
+      "loss": 5.4039,
+      "step": 15864
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.991584300994873,
+      "learning_rate": 1.0805869855230003e-05,
+      "loss": 5.4651,
+      "step": 15868
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.276181936264038,
+      "learning_rate": 1.0795987944068383e-05,
+      "loss": 5.5605,
+      "step": 15872
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0964319705963135,
+      "learning_rate": 1.0786106032906766e-05,
+      "loss": 5.4315,
+      "step": 15876
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0152671337127686,
+      "learning_rate": 1.0776224121745146e-05,
+      "loss": 5.5034,
+      "step": 15880
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.065906524658203,
+      "learning_rate": 1.0766342210583528e-05,
+      "loss": 5.5383,
+      "step": 15884
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9010733366012573,
+      "learning_rate": 1.0756460299421908e-05,
+      "loss": 5.5424,
+      "step": 15888
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9711081981658936,
+      "learning_rate": 1.0746578388260289e-05,
+      "loss": 5.4899,
+      "step": 15892
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9570908546447754,
+      "learning_rate": 1.0736696477098672e-05,
+      "loss": 5.5072,
+      "step": 15896
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1524243354797363,
+      "learning_rate": 1.0726814565937053e-05,
+      "loss": 5.4421,
+      "step": 15900
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8811511993408203,
+      "learning_rate": 1.0716932654775435e-05,
+      "loss": 5.3984,
+      "step": 15904
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.914753794670105,
+      "learning_rate": 1.0707050743613815e-05,
+      "loss": 5.5874,
+      "step": 15908
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9757344722747803,
+      "learning_rate": 1.0697168832452197e-05,
+      "loss": 5.5402,
+      "step": 15912
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1234004497528076,
+      "learning_rate": 1.0689757399080982e-05,
+      "loss": 5.4948,
+      "step": 15916
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9321492910385132,
+      "learning_rate": 1.0679875487919364e-05,
+      "loss": 5.3983,
+      "step": 15920
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0924885272979736,
+      "learning_rate": 1.0669993576757746e-05,
+      "loss": 5.3964,
+      "step": 15924
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1206881999969482,
+      "learning_rate": 1.0660111665596127e-05,
+      "loss": 5.4496,
+      "step": 15928
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.230121374130249,
+      "learning_rate": 1.0650229754434509e-05,
+      "loss": 5.3782,
+      "step": 15932
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9307682514190674,
+      "learning_rate": 1.0640347843272889e-05,
+      "loss": 5.4723,
+      "step": 15936
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.326744794845581,
+      "learning_rate": 1.0630465932111271e-05,
+      "loss": 5.4944,
+      "step": 15940
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0075511932373047,
+      "learning_rate": 1.0620584020949651e-05,
+      "loss": 5.4213,
+      "step": 15944
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0154924392700195,
+      "learning_rate": 1.0610702109788034e-05,
+      "loss": 5.5201,
+      "step": 15948
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3839402198791504,
+      "learning_rate": 1.0600820198626416e-05,
+      "loss": 5.5373,
+      "step": 15952
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8833853006362915,
+      "learning_rate": 1.0590938287464796e-05,
+      "loss": 5.5104,
+      "step": 15956
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0425021648406982,
+      "learning_rate": 1.0581056376303178e-05,
+      "loss": 5.5934,
+      "step": 15960
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8929284811019897,
+      "learning_rate": 1.0571174465141558e-05,
+      "loss": 5.5095,
+      "step": 15964
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8911198377609253,
+      "learning_rate": 1.056129255397994e-05,
+      "loss": 5.4591,
+      "step": 15968
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.856986403465271,
+      "learning_rate": 1.055141064281832e-05,
+      "loss": 5.35,
+      "step": 15972
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8045761585235596,
+      "learning_rate": 1.0541528731656703e-05,
+      "loss": 5.5255,
+      "step": 15976
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9339076280593872,
+      "learning_rate": 1.0531646820495085e-05,
+      "loss": 5.4632,
+      "step": 15980
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.076307535171509,
+      "learning_rate": 1.0521764909333465e-05,
+      "loss": 5.453,
+      "step": 15984
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0790350437164307,
+      "learning_rate": 1.0511882998171847e-05,
+      "loss": 5.4504,
+      "step": 15988
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.015014171600342,
+      "learning_rate": 1.0502001087010228e-05,
+      "loss": 5.4082,
+      "step": 15992
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9248104095458984,
+      "learning_rate": 1.049211917584861e-05,
+      "loss": 5.5261,
+      "step": 15996
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9845161437988281,
+      "learning_rate": 1.048223726468699e-05,
+      "loss": 5.495,
+      "step": 16000
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0011472702026367,
+      "learning_rate": 1.0472355353525372e-05,
+      "loss": 5.3597,
+      "step": 16004
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0203723907470703,
+      "learning_rate": 1.0462473442363753e-05,
+      "loss": 5.5997,
+      "step": 16008
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9304739236831665,
+      "learning_rate": 1.0452591531202136e-05,
+      "loss": 5.4136,
+      "step": 16012
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1026952266693115,
+      "learning_rate": 1.0442709620040517e-05,
+      "loss": 5.5538,
+      "step": 16016
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9498411417007446,
+      "learning_rate": 1.0432827708878897e-05,
+      "loss": 5.5066,
+      "step": 16020
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.043534517288208,
+      "learning_rate": 1.0422945797717279e-05,
+      "loss": 5.4616,
+      "step": 16024
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9264411926269531,
+      "learning_rate": 1.041306388655566e-05,
+      "loss": 5.5081,
+      "step": 16028
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9634205102920532,
+      "learning_rate": 1.0403181975394042e-05,
+      "loss": 5.436,
+      "step": 16032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9793167114257812,
+      "learning_rate": 1.0393300064232422e-05,
+      "loss": 5.4457,
+      "step": 16036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0029139518737793,
+      "learning_rate": 1.0383418153070806e-05,
+      "loss": 5.5097,
+      "step": 16040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.93049156665802,
+      "learning_rate": 1.0373536241909186e-05,
+      "loss": 5.5088,
+      "step": 16044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9566268920898438,
+      "learning_rate": 1.0363654330747566e-05,
+      "loss": 5.509,
+      "step": 16048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9150116443634033,
+      "learning_rate": 1.0353772419585948e-05,
+      "loss": 5.432,
+      "step": 16052
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9457064867019653,
+      "learning_rate": 1.0343890508424329e-05,
+      "loss": 5.5555,
+      "step": 16056
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.037177562713623,
+      "learning_rate": 1.0334008597262711e-05,
+      "loss": 5.4792,
+      "step": 16060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0516912937164307,
+      "learning_rate": 1.0324126686101091e-05,
+      "loss": 5.5053,
+      "step": 16064
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7669730186462402,
+      "learning_rate": 1.0314244774939475e-05,
+      "loss": 5.3994,
+      "step": 16068
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1283509731292725,
+      "learning_rate": 1.0304362863777855e-05,
+      "loss": 5.466,
+      "step": 16072
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7946540117263794,
+      "learning_rate": 1.0294480952616237e-05,
+      "loss": 5.4215,
+      "step": 16076
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.273894786834717,
+      "learning_rate": 1.0284599041454618e-05,
+      "loss": 5.408,
+      "step": 16080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.055126667022705,
+      "learning_rate": 1.0274717130292998e-05,
+      "loss": 5.5304,
+      "step": 16084
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0089943408966064,
+      "learning_rate": 1.026483521913138e-05,
+      "loss": 5.4147,
+      "step": 16088
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.053406000137329,
+      "learning_rate": 1.025495330796976e-05,
+      "loss": 5.5624,
+      "step": 16092
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9701050519943237,
+      "learning_rate": 1.0245071396808144e-05,
+      "loss": 5.3615,
+      "step": 16096
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9590353965759277,
+      "learning_rate": 1.0235189485646525e-05,
+      "loss": 5.4707,
+      "step": 16100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.097073793411255,
+      "learning_rate": 1.0225307574484907e-05,
+      "loss": 5.4783,
+      "step": 16104
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9956692457199097,
+      "learning_rate": 1.0215425663323287e-05,
+      "loss": 5.4288,
+      "step": 16108
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8568942546844482,
+      "learning_rate": 1.020554375216167e-05,
+      "loss": 5.5691,
+      "step": 16112
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9717293977737427,
+      "learning_rate": 1.019566184100005e-05,
+      "loss": 5.4763,
+      "step": 16116
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.085775375366211,
+      "learning_rate": 1.018577992983843e-05,
+      "loss": 5.5406,
+      "step": 16120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9938271045684814,
+      "learning_rate": 1.0175898018676814e-05,
+      "loss": 5.52,
+      "step": 16124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1408092975616455,
+      "learning_rate": 1.0166016107515194e-05,
+      "loss": 5.4718,
+      "step": 16128
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.084689140319824,
+      "learning_rate": 1.0156134196353576e-05,
+      "loss": 5.5263,
+      "step": 16132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8502254486083984,
+      "learning_rate": 1.0146252285191956e-05,
+      "loss": 5.4056,
+      "step": 16136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.832261323928833,
+      "learning_rate": 1.0136370374030339e-05,
+      "loss": 5.4597,
+      "step": 16140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9201068878173828,
+      "learning_rate": 1.0126488462868719e-05,
+      "loss": 5.5378,
+      "step": 16144
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8467752933502197,
+      "learning_rate": 1.0116606551707101e-05,
+      "loss": 5.4567,
+      "step": 16148
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1131222248077393,
+      "learning_rate": 1.0106724640545481e-05,
+      "loss": 5.35,
+      "step": 16152
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0209567546844482,
+      "learning_rate": 1.0096842729383863e-05,
+      "loss": 5.5339,
+      "step": 16156
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1684625148773193,
+      "learning_rate": 1.0086960818222245e-05,
+      "loss": 5.4267,
+      "step": 16160
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1010987758636475,
+      "learning_rate": 1.0077078907060626e-05,
+      "loss": 5.4042,
+      "step": 16164
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9701296091079712,
+      "learning_rate": 1.0067196995899008e-05,
+      "loss": 5.6227,
+      "step": 16168
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8271695375442505,
+      "learning_rate": 1.0057315084737388e-05,
+      "loss": 5.4151,
+      "step": 16172
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2199959754943848,
+      "learning_rate": 1.004743317357577e-05,
+      "loss": 5.5352,
+      "step": 16176
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2069809436798096,
+      "learning_rate": 1.003755126241415e-05,
+      "loss": 5.4987,
+      "step": 16180
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.027318239212036,
+      "learning_rate": 1.0027669351252533e-05,
+      "loss": 5.5284,
+      "step": 16184
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9697614908218384,
+      "learning_rate": 1.0017787440090915e-05,
+      "loss": 5.5906,
+      "step": 16188
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.252358913421631,
+      "learning_rate": 1.0007905528929295e-05,
+      "loss": 5.4911,
+      "step": 16192
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1323323249816895,
+      "learning_rate": 9.998023617767677e-06,
+      "loss": 5.5276,
+      "step": 16196
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.277160167694092,
+      "learning_rate": 9.988141706606058e-06,
+      "loss": 5.4718,
+      "step": 16200
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8769733905792236,
+      "learning_rate": 9.97825979544444e-06,
+      "loss": 5.3622,
+      "step": 16204
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8731465339660645,
+      "learning_rate": 9.96837788428282e-06,
+      "loss": 5.5215,
+      "step": 16208
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8808999061584473,
+      "learning_rate": 9.958495973121202e-06,
+      "loss": 5.3953,
+      "step": 16212
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8657152652740479,
+      "learning_rate": 9.948614061959584e-06,
+      "loss": 5.3316,
+      "step": 16216
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.849173665046692,
+      "learning_rate": 9.938732150797965e-06,
+      "loss": 5.4471,
+      "step": 16220
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.220717668533325,
+      "learning_rate": 9.928850239636347e-06,
+      "loss": 5.5204,
+      "step": 16224
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0210342407226562,
+      "learning_rate": 9.918968328474727e-06,
+      "loss": 5.3364,
+      "step": 16228
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9695372581481934,
+      "learning_rate": 9.909086417313109e-06,
+      "loss": 5.485,
+      "step": 16232
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.875001311302185,
+      "learning_rate": 9.89920450615149e-06,
+      "loss": 5.3977,
+      "step": 16236
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.136852741241455,
+      "learning_rate": 9.889322594989871e-06,
+      "loss": 5.5784,
+      "step": 16240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8972970247268677,
+      "learning_rate": 9.879440683828253e-06,
+      "loss": 5.4973,
+      "step": 16244
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.14034366607666,
+      "learning_rate": 9.869558772666634e-06,
+      "loss": 5.546,
+      "step": 16248
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0028293132781982,
+      "learning_rate": 9.859676861505016e-06,
+      "loss": 5.3869,
+      "step": 16252
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9485490322113037,
+      "learning_rate": 9.849794950343396e-06,
+      "loss": 5.4852,
+      "step": 16256
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9260238409042358,
+      "learning_rate": 9.839913039181778e-06,
+      "loss": 5.473,
+      "step": 16260
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1630001068115234,
+      "learning_rate": 9.830031128020159e-06,
+      "loss": 5.4207,
+      "step": 16264
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9328508377075195,
+      "learning_rate": 9.82014921685854e-06,
+      "loss": 5.4433,
+      "step": 16268
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9832642078399658,
+      "learning_rate": 9.810267305696923e-06,
+      "loss": 5.4387,
+      "step": 16272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0616111755371094,
+      "learning_rate": 9.800385394535303e-06,
+      "loss": 5.3895,
+      "step": 16276
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1709606647491455,
+      "learning_rate": 9.790503483373685e-06,
+      "loss": 5.5639,
+      "step": 16280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1553335189819336,
+      "learning_rate": 9.780621572212066e-06,
+      "loss": 5.5419,
+      "step": 16284
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7869793176651,
+      "learning_rate": 9.770739661050448e-06,
+      "loss": 5.4425,
+      "step": 16288
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.902388095855713,
+      "learning_rate": 9.760857749888828e-06,
+      "loss": 5.3651,
+      "step": 16292
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.135723829269409,
+      "learning_rate": 9.750975838727212e-06,
+      "loss": 5.5838,
+      "step": 16296
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3749701976776123,
+      "learning_rate": 9.741093927565592e-06,
+      "loss": 5.5562,
+      "step": 16300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9421731233596802,
+      "learning_rate": 9.731212016403973e-06,
+      "loss": 5.432,
+      "step": 16304
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.87296462059021,
+      "learning_rate": 9.721330105242355e-06,
+      "loss": 5.5429,
+      "step": 16308
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0159077644348145,
+      "learning_rate": 9.711448194080735e-06,
+      "loss": 5.5088,
+      "step": 16312
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9539657831192017,
+      "learning_rate": 9.701566282919117e-06,
+      "loss": 5.49,
+      "step": 16316
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0840256214141846,
+      "learning_rate": 9.691684371757497e-06,
+      "loss": 5.4972,
+      "step": 16320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0817222595214844,
+      "learning_rate": 9.68180246059588e-06,
+      "loss": 5.5185,
+      "step": 16324
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.172551393508911,
+      "learning_rate": 9.671920549434262e-06,
+      "loss": 5.5465,
+      "step": 16328
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8779146671295166,
+      "learning_rate": 9.662038638272644e-06,
+      "loss": 5.5047,
+      "step": 16332
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9017332792282104,
+      "learning_rate": 9.652156727111024e-06,
+      "loss": 5.4705,
+      "step": 16336
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0198209285736084,
+      "learning_rate": 9.642274815949404e-06,
+      "loss": 5.4886,
+      "step": 16340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0686516761779785,
+      "learning_rate": 9.632392904787786e-06,
+      "loss": 5.5188,
+      "step": 16344
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.962902545928955,
+      "learning_rate": 9.622510993626167e-06,
+      "loss": 5.5414,
+      "step": 16348
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0159122943878174,
+      "learning_rate": 9.612629082464549e-06,
+      "loss": 5.422,
+      "step": 16352
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2383811473846436,
+      "learning_rate": 9.60274717130293e-06,
+      "loss": 5.5384,
+      "step": 16356
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.993220567703247,
+      "learning_rate": 9.592865260141313e-06,
+      "loss": 5.3551,
+      "step": 16360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.130995035171509,
+      "learning_rate": 9.582983348979693e-06,
+      "loss": 5.4926,
+      "step": 16364
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9019581079483032,
+      "learning_rate": 9.573101437818074e-06,
+      "loss": 5.5623,
+      "step": 16368
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1175732612609863,
+      "learning_rate": 9.563219526656456e-06,
+      "loss": 5.4529,
+      "step": 16372
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8680050373077393,
+      "learning_rate": 9.553337615494836e-06,
+      "loss": 5.4273,
+      "step": 16376
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0988821983337402,
+      "learning_rate": 9.543455704333218e-06,
+      "loss": 5.477,
+      "step": 16380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.198651075363159,
+      "learning_rate": 9.5335737931716e-06,
+      "loss": 5.5427,
+      "step": 16384
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1054868698120117,
+      "learning_rate": 9.523691882009982e-06,
+      "loss": 5.4173,
+      "step": 16388
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0211637020111084,
+      "learning_rate": 9.513809970848363e-06,
+      "loss": 5.6035,
+      "step": 16392
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0547540187835693,
+      "learning_rate": 9.503928059686745e-06,
+      "loss": 5.4983,
+      "step": 16396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9957647323608398,
+      "learning_rate": 9.494046148525125e-06,
+      "loss": 5.5374,
+      "step": 16400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0535998344421387,
+      "learning_rate": 9.484164237363505e-06,
+      "loss": 5.5643,
+      "step": 16404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1046228408813477,
+      "learning_rate": 9.474282326201887e-06,
+      "loss": 5.4727,
+      "step": 16408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1698873043060303,
+      "learning_rate": 9.46440041504027e-06,
+      "loss": 5.5637,
+      "step": 16412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9448730945587158,
+      "learning_rate": 9.454518503878652e-06,
+      "loss": 5.4932,
+      "step": 16416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0663201808929443,
+      "learning_rate": 9.444636592717032e-06,
+      "loss": 5.5067,
+      "step": 16420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8125361204147339,
+      "learning_rate": 9.434754681555414e-06,
+      "loss": 5.5381,
+      "step": 16424
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.077420473098755,
+      "learning_rate": 9.424872770393794e-06,
+      "loss": 5.5642,
+      "step": 16428
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9312140941619873,
+      "learning_rate": 9.414990859232176e-06,
+      "loss": 5.3661,
+      "step": 16432
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8529763221740723,
+      "learning_rate": 9.405108948070557e-06,
+      "loss": 5.4776,
+      "step": 16436
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8764407634735107,
+      "learning_rate": 9.395227036908939e-06,
+      "loss": 5.4928,
+      "step": 16440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.139594316482544,
+      "learning_rate": 9.385345125747321e-06,
+      "loss": 5.4108,
+      "step": 16444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1105079650878906,
+      "learning_rate": 9.375463214585701e-06,
+      "loss": 5.4876,
+      "step": 16448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9484140872955322,
+      "learning_rate": 9.365581303424083e-06,
+      "loss": 5.4148,
+      "step": 16452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0571391582489014,
+      "learning_rate": 9.355699392262464e-06,
+      "loss": 5.6282,
+      "step": 16456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7625576257705688,
+      "learning_rate": 9.345817481100846e-06,
+      "loss": 5.5073,
+      "step": 16460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1183722019195557,
+      "learning_rate": 9.335935569939226e-06,
+      "loss": 5.4659,
+      "step": 16464
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.135255813598633,
+      "learning_rate": 9.326053658777608e-06,
+      "loss": 5.4262,
+      "step": 16468
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7497916221618652,
+      "learning_rate": 9.31617174761599e-06,
+      "loss": 5.537,
+      "step": 16472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9297901391983032,
+      "learning_rate": 9.30628983645437e-06,
+      "loss": 5.4848,
+      "step": 16476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.992133617401123,
+      "learning_rate": 9.296407925292753e-06,
+      "loss": 5.5307,
+      "step": 16480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8346421718597412,
+      "learning_rate": 9.286526014131133e-06,
+      "loss": 5.4071,
+      "step": 16484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9083247184753418,
+      "learning_rate": 9.276644102969515e-06,
+      "loss": 5.4248,
+      "step": 16488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0954606533050537,
+      "learning_rate": 9.266762191807895e-06,
+      "loss": 5.5175,
+      "step": 16492
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.101158857345581,
+      "learning_rate": 9.256880280646278e-06,
+      "loss": 5.5345,
+      "step": 16496
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.170283794403076,
+      "learning_rate": 9.24699836948466e-06,
+      "loss": 5.5719,
+      "step": 16500
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.078697681427002,
+      "learning_rate": 9.23711645832304e-06,
+      "loss": 5.4199,
+      "step": 16504
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1254682540893555,
+      "learning_rate": 9.227234547161422e-06,
+      "loss": 5.3647,
+      "step": 16508
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2468132972717285,
+      "learning_rate": 9.217352635999802e-06,
+      "loss": 5.4673,
+      "step": 16512
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.053579330444336,
+      "learning_rate": 9.207470724838184e-06,
+      "loss": 5.4083,
+      "step": 16516
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2088301181793213,
+      "learning_rate": 9.197588813676565e-06,
+      "loss": 5.5694,
+      "step": 16520
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9974719285964966,
+      "learning_rate": 9.187706902514947e-06,
+      "loss": 5.4396,
+      "step": 16524
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.066420555114746,
+      "learning_rate": 9.177824991353329e-06,
+      "loss": 5.5418,
+      "step": 16528
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2263917922973633,
+      "learning_rate": 9.16794308019171e-06,
+      "loss": 5.4904,
+      "step": 16532
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0956313610076904,
+      "learning_rate": 9.158061169030091e-06,
+      "loss": 5.47,
+      "step": 16536
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.120701551437378,
+      "learning_rate": 9.148179257868472e-06,
+      "loss": 5.4498,
+      "step": 16540
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.068040609359741,
+      "learning_rate": 9.138297346706854e-06,
+      "loss": 5.5373,
+      "step": 16544
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.125322103500366,
+      "learning_rate": 9.128415435545234e-06,
+      "loss": 5.4782,
+      "step": 16548
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8984034061431885,
+      "learning_rate": 9.118533524383616e-06,
+      "loss": 5.5614,
+      "step": 16552
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.038201332092285,
+      "learning_rate": 9.108651613221998e-06,
+      "loss": 5.5308,
+      "step": 16556
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.083704948425293,
+      "learning_rate": 9.098769702060379e-06,
+      "loss": 5.3986,
+      "step": 16560
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2222559452056885,
+      "learning_rate": 9.08888779089876e-06,
+      "loss": 5.4164,
+      "step": 16564
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.319937229156494,
+      "learning_rate": 9.079005879737141e-06,
+      "loss": 5.5564,
+      "step": 16568
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.930305004119873,
+      "learning_rate": 9.069123968575523e-06,
+      "loss": 5.5181,
+      "step": 16572
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.175090789794922,
+      "learning_rate": 9.059242057413904e-06,
+      "loss": 5.4816,
+      "step": 16576
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9041398763656616,
+      "learning_rate": 9.049360146252286e-06,
+      "loss": 5.4462,
+      "step": 16580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.92721426486969,
+      "learning_rate": 9.039478235090668e-06,
+      "loss": 5.4993,
+      "step": 16584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.098320484161377,
+      "learning_rate": 9.029596323929048e-06,
+      "loss": 5.4334,
+      "step": 16588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.988296389579773,
+      "learning_rate": 9.01971441276743e-06,
+      "loss": 5.5241,
+      "step": 16592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9492675065994263,
+      "learning_rate": 9.00983250160581e-06,
+      "loss": 5.4024,
+      "step": 16596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0381886959075928,
+      "learning_rate": 8.999950590444192e-06,
+      "loss": 5.4861,
+      "step": 16600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2015647888183594,
+      "learning_rate": 8.990068679282573e-06,
+      "loss": 5.4589,
+      "step": 16604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.970509648323059,
+      "learning_rate": 8.980186768120955e-06,
+      "loss": 5.5072,
+      "step": 16608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.048265218734741,
+      "learning_rate": 8.970304856959337e-06,
+      "loss": 5.4781,
+      "step": 16612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1781177520751953,
+      "learning_rate": 8.960422945797719e-06,
+      "loss": 5.4382,
+      "step": 16616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3206918239593506,
+      "learning_rate": 8.9505410346361e-06,
+      "loss": 5.5371,
+      "step": 16620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.129166603088379,
+      "learning_rate": 8.94065912347448e-06,
+      "loss": 5.4773,
+      "step": 16624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8786160945892334,
+      "learning_rate": 8.930777212312862e-06,
+      "loss": 5.3599,
+      "step": 16628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.114015579223633,
+      "learning_rate": 8.920895301151242e-06,
+      "loss": 5.5262,
+      "step": 16632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9730268716812134,
+      "learning_rate": 8.911013389989624e-06,
+      "loss": 5.3921,
+      "step": 16636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.007050037384033,
+      "learning_rate": 8.901131478828005e-06,
+      "loss": 5.4521,
+      "step": 16640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9782698154449463,
+      "learning_rate": 8.891249567666388e-06,
+      "loss": 5.4024,
+      "step": 16644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.118109941482544,
+      "learning_rate": 8.881367656504769e-06,
+      "loss": 5.4549,
+      "step": 16648
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9269788265228271,
+      "learning_rate": 8.871485745343149e-06,
+      "loss": 5.3879,
+      "step": 16652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0356998443603516,
+      "learning_rate": 8.861603834181531e-06,
+      "loss": 5.4729,
+      "step": 16656
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9945244789123535,
+      "learning_rate": 8.851721923019912e-06,
+      "loss": 5.489,
+      "step": 16660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1839029788970947,
+      "learning_rate": 8.841840011858294e-06,
+      "loss": 5.3626,
+      "step": 16664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9056282043457031,
+      "learning_rate": 8.831958100696674e-06,
+      "loss": 5.5155,
+      "step": 16668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.971134066581726,
+      "learning_rate": 8.822076189535058e-06,
+      "loss": 5.4582,
+      "step": 16672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2608683109283447,
+      "learning_rate": 8.812194278373438e-06,
+      "loss": 5.5671,
+      "step": 16676
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8873885869979858,
+      "learning_rate": 8.80231236721182e-06,
+      "loss": 5.5378,
+      "step": 16680
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1268815994262695,
+      "learning_rate": 8.7924304560502e-06,
+      "loss": 5.5464,
+      "step": 16684
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.120333671569824,
+      "learning_rate": 8.782548544888581e-06,
+      "loss": 5.4566,
+      "step": 16688
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.202099323272705,
+      "learning_rate": 8.772666633726963e-06,
+      "loss": 5.3333,
+      "step": 16692
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8824645280838013,
+      "learning_rate": 8.762784722565343e-06,
+      "loss": 5.5016,
+      "step": 16696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0147457122802734,
+      "learning_rate": 8.752902811403727e-06,
+      "loss": 5.4586,
+      "step": 16700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.041895627975464,
+      "learning_rate": 8.743020900242107e-06,
+      "loss": 5.5022,
+      "step": 16704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.077690362930298,
+      "learning_rate": 8.73313898908049e-06,
+      "loss": 5.5781,
+      "step": 16708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9477964639663696,
+      "learning_rate": 8.72325707791887e-06,
+      "loss": 5.5584,
+      "step": 16712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8398356437683105,
+      "learning_rate": 8.713375166757252e-06,
+      "loss": 5.5032,
+      "step": 16716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9973992109298706,
+      "learning_rate": 8.703493255595632e-06,
+      "loss": 5.37,
+      "step": 16720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0176520347595215,
+      "learning_rate": 8.693611344434013e-06,
+      "loss": 5.5138,
+      "step": 16724
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1837217807769775,
+      "learning_rate": 8.683729433272396e-06,
+      "loss": 5.5308,
+      "step": 16728
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0156595706939697,
+      "learning_rate": 8.673847522110777e-06,
+      "loss": 5.5722,
+      "step": 16732
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3676466941833496,
+      "learning_rate": 8.663965610949159e-06,
+      "loss": 5.5775,
+      "step": 16736
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.270716667175293,
+      "learning_rate": 8.65408369978754e-06,
+      "loss": 5.4107,
+      "step": 16740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9551721811294556,
+      "learning_rate": 8.644201788625921e-06,
+      "loss": 5.6022,
+      "step": 16744
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9827896356582642,
+      "learning_rate": 8.634319877464302e-06,
+      "loss": 5.4923,
+      "step": 16748
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1360533237457275,
+      "learning_rate": 8.624437966302682e-06,
+      "loss": 5.4328,
+      "step": 16752
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.094109296798706,
+      "learning_rate": 8.614556055141066e-06,
+      "loss": 5.416,
+      "step": 16756
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9513869285583496,
+      "learning_rate": 8.604674143979446e-06,
+      "loss": 5.403,
+      "step": 16760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0020523071289062,
+      "learning_rate": 8.594792232817828e-06,
+      "loss": 5.4628,
+      "step": 16764
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.405801296234131,
+      "learning_rate": 8.584910321656209e-06,
+      "loss": 5.5799,
+      "step": 16768
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.114650249481201,
+      "learning_rate": 8.57502841049459e-06,
+      "loss": 5.4151,
+      "step": 16772
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9269883632659912,
+      "learning_rate": 8.565146499332971e-06,
+      "loss": 5.4321,
+      "step": 16776
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.894822597503662,
+      "learning_rate": 8.555264588171353e-06,
+      "loss": 5.4419,
+      "step": 16780
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0186209678649902,
+      "learning_rate": 8.545382677009735e-06,
+      "loss": 5.5657,
+      "step": 16784
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.920372486114502,
+      "learning_rate": 8.535500765848115e-06,
+      "loss": 5.5334,
+      "step": 16788
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1055715084075928,
+      "learning_rate": 8.525618854686497e-06,
+      "loss": 5.4952,
+      "step": 16792
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0137712955474854,
+      "learning_rate": 8.515736943524878e-06,
+      "loss": 5.5128,
+      "step": 16796
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0473227500915527,
+      "learning_rate": 8.50585503236326e-06,
+      "loss": 5.6793,
+      "step": 16800
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0757429599761963,
+      "learning_rate": 8.49597312120164e-06,
+      "loss": 5.4888,
+      "step": 16804
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9529187679290771,
+      "learning_rate": 8.486091210040022e-06,
+      "loss": 5.2731,
+      "step": 16808
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.897220492362976,
+      "learning_rate": 8.476209298878403e-06,
+      "loss": 5.4103,
+      "step": 16812
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0591204166412354,
+      "learning_rate": 8.466327387716785e-06,
+      "loss": 5.5609,
+      "step": 16816
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8429813385009766,
+      "learning_rate": 8.456445476555167e-06,
+      "loss": 5.3453,
+      "step": 16820
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.854067087173462,
+      "learning_rate": 8.446563565393547e-06,
+      "loss": 5.4855,
+      "step": 16824
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.87723708152771,
+      "learning_rate": 8.43668165423193e-06,
+      "loss": 5.5087,
+      "step": 16828
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.25486159324646,
+      "learning_rate": 8.42679974307031e-06,
+      "loss": 5.6419,
+      "step": 16832
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1054129600524902,
+      "learning_rate": 8.416917831908692e-06,
+      "loss": 5.4713,
+      "step": 16836
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9546363353729248,
+      "learning_rate": 8.407035920747072e-06,
+      "loss": 5.51,
+      "step": 16840
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8574483394622803,
+      "learning_rate": 8.397154009585454e-06,
+      "loss": 5.4273,
+      "step": 16844
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.135690927505493,
+      "learning_rate": 8.387272098423836e-06,
+      "loss": 5.4408,
+      "step": 16848
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9898631572723389,
+      "learning_rate": 8.377390187262217e-06,
+      "loss": 5.2905,
+      "step": 16852
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.016470193862915,
+      "learning_rate": 8.367508276100599e-06,
+      "loss": 5.3266,
+      "step": 16856
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9973735809326172,
+      "learning_rate": 8.357626364938979e-06,
+      "loss": 5.4987,
+      "step": 16860
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.116567373275757,
+      "learning_rate": 8.347744453777361e-06,
+      "loss": 5.4306,
+      "step": 16864
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.044475793838501,
+      "learning_rate": 8.337862542615741e-06,
+      "loss": 5.4907,
+      "step": 16868
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.090527296066284,
+      "learning_rate": 8.327980631454123e-06,
+      "loss": 5.3698,
+      "step": 16872
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.98384428024292,
+      "learning_rate": 8.318098720292506e-06,
+      "loss": 5.469,
+      "step": 16876
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.775121808052063,
+      "learning_rate": 8.308216809130886e-06,
+      "loss": 5.461,
+      "step": 16880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9661427736282349,
+      "learning_rate": 8.298334897969268e-06,
+      "loss": 5.494,
+      "step": 16884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0031895637512207,
+      "learning_rate": 8.288452986807648e-06,
+      "loss": 5.5011,
+      "step": 16888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.221911907196045,
+      "learning_rate": 8.27857107564603e-06,
+      "loss": 5.4296,
+      "step": 16892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0504343509674072,
+      "learning_rate": 8.26868916448441e-06,
+      "loss": 5.5495,
+      "step": 16896
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1068339347839355,
+      "learning_rate": 8.258807253322794e-06,
+      "loss": 5.408,
+      "step": 16900
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0044867992401123,
+      "learning_rate": 8.248925342161175e-06,
+      "loss": 5.512,
+      "step": 16904
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3192813396453857,
+      "learning_rate": 8.239043430999555e-06,
+      "loss": 5.4185,
+      "step": 16908
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8410991430282593,
+      "learning_rate": 8.229161519837937e-06,
+      "loss": 5.4222,
+      "step": 16912
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0134191513061523,
+      "learning_rate": 8.219279608676318e-06,
+      "loss": 5.4357,
+      "step": 16916
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0390844345092773,
+      "learning_rate": 8.2093976975147e-06,
+      "loss": 5.5363,
+      "step": 16920
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.12786602973938,
+      "learning_rate": 8.19951578635308e-06,
+      "loss": 5.31,
+      "step": 16924
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9766027927398682,
+      "learning_rate": 8.189633875191464e-06,
+      "loss": 5.534,
+      "step": 16928
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7689497470855713,
+      "learning_rate": 8.179751964029844e-06,
+      "loss": 5.3465,
+      "step": 16932
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.117271900177002,
+      "learning_rate": 8.169870052868225e-06,
+      "loss": 5.4583,
+      "step": 16936
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0808498859405518,
+      "learning_rate": 8.159988141706607e-06,
+      "loss": 5.477,
+      "step": 16940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0178062915802,
+      "learning_rate": 8.150106230544987e-06,
+      "loss": 5.5178,
+      "step": 16944
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7878342866897583,
+      "learning_rate": 8.140224319383369e-06,
+      "loss": 5.533,
+      "step": 16948
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0112874507904053,
+      "learning_rate": 8.13034240822175e-06,
+      "loss": 5.342,
+      "step": 16952
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.224484443664551,
+      "learning_rate": 8.120460497060133e-06,
+      "loss": 5.4574,
+      "step": 16956
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.29886531829834,
+      "learning_rate": 8.110578585898514e-06,
+      "loss": 5.5541,
+      "step": 16960
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8924994468688965,
+      "learning_rate": 8.100696674736896e-06,
+      "loss": 5.4196,
+      "step": 16964
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2159488201141357,
+      "learning_rate": 8.090814763575276e-06,
+      "loss": 5.4989,
+      "step": 16968
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.170715570449829,
+      "learning_rate": 8.080932852413656e-06,
+      "loss": 5.5074,
+      "step": 16972
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0515708923339844,
+      "learning_rate": 8.071050941252038e-06,
+      "loss": 5.4033,
+      "step": 16976
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0467865467071533,
+      "learning_rate": 8.061169030090419e-06,
+      "loss": 5.4566,
+      "step": 16980
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9163670539855957,
+      "learning_rate": 8.0512871189288e-06,
+      "loss": 5.5335,
+      "step": 16984
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1390318870544434,
+      "learning_rate": 8.041405207767183e-06,
+      "loss": 5.5707,
+      "step": 16988
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9965319633483887,
+      "learning_rate": 8.031523296605565e-06,
+      "loss": 5.4165,
+      "step": 16992
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.137233257293701,
+      "learning_rate": 8.021641385443945e-06,
+      "loss": 5.3596,
+      "step": 16996
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.152256727218628,
+      "learning_rate": 8.011759474282327e-06,
+      "loss": 5.3396,
+      "step": 17000
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.28680682182312,
+      "learning_rate": 8.001877563120708e-06,
+      "loss": 5.5507,
+      "step": 17004
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.26821231842041,
+      "learning_rate": 7.991995651959088e-06,
+      "loss": 5.4969,
+      "step": 17008
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.275667428970337,
+      "learning_rate": 7.98211374079747e-06,
+      "loss": 5.441,
+      "step": 17012
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.080756902694702,
+      "learning_rate": 7.972231829635852e-06,
+      "loss": 5.4398,
+      "step": 17016
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.10422420501709,
+      "learning_rate": 7.962349918474234e-06,
+      "loss": 5.437,
+      "step": 17020
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.858323335647583,
+      "learning_rate": 7.952468007312615e-06,
+      "loss": 5.5432,
+      "step": 17024
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4101650714874268,
+      "learning_rate": 7.942586096150997e-06,
+      "loss": 5.412,
+      "step": 17028
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2219436168670654,
+      "learning_rate": 7.932704184989377e-06,
+      "loss": 5.5279,
+      "step": 17032
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.182474374771118,
+      "learning_rate": 7.922822273827757e-06,
+      "loss": 5.4662,
+      "step": 17036
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.060351610183716,
+      "learning_rate": 7.91294036266614e-06,
+      "loss": 5.504,
+      "step": 17040
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1096701622009277,
+      "learning_rate": 7.903058451504522e-06,
+      "loss": 5.4539,
+      "step": 17044
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0492708683013916,
+      "learning_rate": 7.893176540342904e-06,
+      "loss": 5.4398,
+      "step": 17048
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.032947301864624,
+      "learning_rate": 7.883294629181284e-06,
+      "loss": 5.5253,
+      "step": 17052
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0764636993408203,
+      "learning_rate": 7.873412718019666e-06,
+      "loss": 5.4655,
+      "step": 17056
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.105656862258911,
+      "learning_rate": 7.863530806858046e-06,
+      "loss": 5.4761,
+      "step": 17060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.975953459739685,
+      "learning_rate": 7.853648895696428e-06,
+      "loss": 5.5364,
+      "step": 17064
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0592944622039795,
+      "learning_rate": 7.843766984534809e-06,
+      "loss": 5.4987,
+      "step": 17068
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1122117042541504,
+      "learning_rate": 7.833885073373191e-06,
+      "loss": 5.4162,
+      "step": 17072
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.143172264099121,
+      "learning_rate": 7.824003162211573e-06,
+      "loss": 5.4959,
+      "step": 17076
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9919787645339966,
+      "learning_rate": 7.814121251049953e-06,
+      "loss": 5.469,
+      "step": 17080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9146004915237427,
+      "learning_rate": 7.804239339888335e-06,
+      "loss": 5.4748,
+      "step": 17084
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3150486946105957,
+      "learning_rate": 7.794357428726716e-06,
+      "loss": 5.4056,
+      "step": 17088
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1717705726623535,
+      "learning_rate": 7.784475517565098e-06,
+      "loss": 5.5389,
+      "step": 17092
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1674489974975586,
+      "learning_rate": 7.774593606403478e-06,
+      "loss": 5.3378,
+      "step": 17096
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.17425537109375,
+      "learning_rate": 7.76471169524186e-06,
+      "loss": 5.5094,
+      "step": 17100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2170867919921875,
+      "learning_rate": 7.754829784080242e-06,
+      "loss": 5.4591,
+      "step": 17104
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0710206031799316,
+      "learning_rate": 7.744947872918623e-06,
+      "loss": 5.461,
+      "step": 17108
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9662617444992065,
+      "learning_rate": 7.735065961757005e-06,
+      "loss": 5.5232,
+      "step": 17112
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1950018405914307,
+      "learning_rate": 7.725184050595385e-06,
+      "loss": 5.4845,
+      "step": 17116
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.166281223297119,
+      "learning_rate": 7.715302139433767e-06,
+      "loss": 5.4899,
+      "step": 17120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0825867652893066,
+      "learning_rate": 7.705420228272148e-06,
+      "loss": 5.5889,
+      "step": 17124
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0458121299743652,
+      "learning_rate": 7.69553831711053e-06,
+      "loss": 5.4465,
+      "step": 17128
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.972931146621704,
+      "learning_rate": 7.685656405948912e-06,
+      "loss": 5.447,
+      "step": 17132
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2071616649627686,
+      "learning_rate": 7.675774494787292e-06,
+      "loss": 5.5405,
+      "step": 17136
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.24798583984375,
+      "learning_rate": 7.665892583625674e-06,
+      "loss": 5.5034,
+      "step": 17140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3352463245391846,
+      "learning_rate": 7.656010672464054e-06,
+      "loss": 5.5322,
+      "step": 17144
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1701347827911377,
+      "learning_rate": 7.646128761302436e-06,
+      "loss": 5.6042,
+      "step": 17148
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0654942989349365,
+      "learning_rate": 7.636246850140817e-06,
+      "loss": 5.5102,
+      "step": 17152
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.410454273223877,
+      "learning_rate": 7.626364938979198e-06,
+      "loss": 5.4408,
+      "step": 17156
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0221352577209473,
+      "learning_rate": 7.616483027817581e-06,
+      "loss": 5.4086,
+      "step": 17160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.211092233657837,
+      "learning_rate": 7.606601116655962e-06,
+      "loss": 5.4589,
+      "step": 17164
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9467920064926147,
+      "learning_rate": 7.596719205494343e-06,
+      "loss": 5.4867,
+      "step": 17168
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1144025325775146,
+      "learning_rate": 7.586837294332725e-06,
+      "loss": 5.5143,
+      "step": 17172
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1652915477752686,
+      "learning_rate": 7.576955383171106e-06,
+      "loss": 5.533,
+      "step": 17176
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9289984703063965,
+      "learning_rate": 7.567073472009486e-06,
+      "loss": 5.5406,
+      "step": 17180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.028322458267212,
+      "learning_rate": 7.557191560847867e-06,
+      "loss": 5.4118,
+      "step": 17184
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2385053634643555,
+      "learning_rate": 7.54730964968625e-06,
+      "loss": 5.5023,
+      "step": 17188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8756178617477417,
+      "learning_rate": 7.5374277385246315e-06,
+      "loss": 5.4028,
+      "step": 17192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0008492469787598,
+      "learning_rate": 7.527545827363013e-06,
+      "loss": 5.4572,
+      "step": 17196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9606680870056152,
+      "learning_rate": 7.517663916201394e-06,
+      "loss": 5.3813,
+      "step": 17200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.765757441520691,
+      "learning_rate": 7.507782005039775e-06,
+      "loss": 5.433,
+      "step": 17204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2999327182769775,
+      "learning_rate": 7.497900093878156e-06,
+      "loss": 5.4443,
+      "step": 17208
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.831790804862976,
+      "learning_rate": 7.488018182716537e-06,
+      "loss": 5.4745,
+      "step": 17212
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0281448364257812,
+      "learning_rate": 7.47813627155492e-06,
+      "loss": 5.3546,
+      "step": 17216
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.163875102996826,
+      "learning_rate": 7.468254360393301e-06,
+      "loss": 5.4957,
+      "step": 17220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0201468467712402,
+      "learning_rate": 7.458372449231682e-06,
+      "loss": 5.3797,
+      "step": 17224
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9520927667617798,
+      "learning_rate": 7.448490538070063e-06,
+      "loss": 5.5663,
+      "step": 17228
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9706037044525146,
+      "learning_rate": 7.4386086269084445e-06,
+      "loss": 5.4462,
+      "step": 17232
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2334280014038086,
+      "learning_rate": 7.428726715746826e-06,
+      "loss": 5.5028,
+      "step": 17236
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.089432716369629,
+      "learning_rate": 7.418844804585207e-06,
+      "loss": 5.5537,
+      "step": 17240
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0354325771331787,
+      "learning_rate": 7.408962893423589e-06,
+      "loss": 5.4,
+      "step": 17244
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9282554388046265,
+      "learning_rate": 7.39908098226197e-06,
+      "loss": 5.3915,
+      "step": 17248
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.133868455886841,
+      "learning_rate": 7.389199071100351e-06,
+      "loss": 5.4868,
+      "step": 17252
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0873701572418213,
+      "learning_rate": 7.379317159938733e-06,
+      "loss": 5.6375,
+      "step": 17256
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9751038551330566,
+      "learning_rate": 7.369435248777114e-06,
+      "loss": 5.4291,
+      "step": 17260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8549004793167114,
+      "learning_rate": 7.359553337615495e-06,
+      "loss": 5.4953,
+      "step": 17264
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9882365465164185,
+      "learning_rate": 7.349671426453876e-06,
+      "loss": 5.5191,
+      "step": 17268
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0008509159088135,
+      "learning_rate": 7.339789515292258e-06,
+      "loss": 5.573,
+      "step": 17272
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9084336757659912,
+      "learning_rate": 7.3299076041306395e-06,
+      "loss": 5.5864,
+      "step": 17276
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3234941959381104,
+      "learning_rate": 7.320025692969021e-06,
+      "loss": 5.4648,
+      "step": 17280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.033445358276367,
+      "learning_rate": 7.310143781807402e-06,
+      "loss": 5.5663,
+      "step": 17284
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.036726474761963,
+      "learning_rate": 7.300261870645783e-06,
+      "loss": 5.3706,
+      "step": 17288
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.135927677154541,
+      "learning_rate": 7.290379959484164e-06,
+      "loss": 5.4365,
+      "step": 17292
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0434539318084717,
+      "learning_rate": 7.280498048322546e-06,
+      "loss": 5.4914,
+      "step": 17296
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9364794492721558,
+      "learning_rate": 7.270616137160927e-06,
+      "loss": 5.358,
+      "step": 17300
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1555495262145996,
+      "learning_rate": 7.260734225999309e-06,
+      "loss": 5.4667,
+      "step": 17304
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0687687397003174,
+      "learning_rate": 7.25085231483769e-06,
+      "loss": 5.5075,
+      "step": 17308
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2169644832611084,
+      "learning_rate": 7.240970403676071e-06,
+      "loss": 5.5186,
+      "step": 17312
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0690207481384277,
+      "learning_rate": 7.2310884925144525e-06,
+      "loss": 5.4944,
+      "step": 17316
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.172851324081421,
+      "learning_rate": 7.221206581352834e-06,
+      "loss": 5.5178,
+      "step": 17320
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.178602457046509,
+      "learning_rate": 7.211324670191215e-06,
+      "loss": 5.5693,
+      "step": 17324
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9525049924850464,
+      "learning_rate": 7.201442759029596e-06,
+      "loss": 5.3773,
+      "step": 17328
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0250043869018555,
+      "learning_rate": 7.191560847867978e-06,
+      "loss": 5.4672,
+      "step": 17332
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.229799747467041,
+      "learning_rate": 7.1816789367063594e-06,
+      "loss": 5.4451,
+      "step": 17336
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2048773765563965,
+      "learning_rate": 7.171797025544741e-06,
+      "loss": 5.5388,
+      "step": 17340
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1948986053466797,
+      "learning_rate": 7.161915114383122e-06,
+      "loss": 5.4121,
+      "step": 17344
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.410446882247925,
+      "learning_rate": 7.152033203221503e-06,
+      "loss": 5.5068,
+      "step": 17348
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0198326110839844,
+      "learning_rate": 7.142151292059884e-06,
+      "loss": 5.4786,
+      "step": 17352
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1943955421447754,
+      "learning_rate": 7.1322693808982655e-06,
+      "loss": 5.4957,
+      "step": 17356
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1132426261901855,
+      "learning_rate": 7.122387469736648e-06,
+      "loss": 5.3965,
+      "step": 17360
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0462357997894287,
+      "learning_rate": 7.112505558575029e-06,
+      "loss": 5.5153,
+      "step": 17364
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0501723289489746,
+      "learning_rate": 7.10262364741341e-06,
+      "loss": 5.5062,
+      "step": 17368
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.148674726486206,
+      "learning_rate": 7.092741736251791e-06,
+      "loss": 5.2946,
+      "step": 17372
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0384411811828613,
+      "learning_rate": 7.082859825090172e-06,
+      "loss": 5.6131,
+      "step": 17376
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.235848903656006,
+      "learning_rate": 7.072977913928554e-06,
+      "loss": 5.3982,
+      "step": 17380
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0050299167633057,
+      "learning_rate": 7.063096002766935e-06,
+      "loss": 5.5681,
+      "step": 17384
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9482308626174927,
+      "learning_rate": 7.053214091605318e-06,
+      "loss": 5.4242,
+      "step": 17388
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.077125072479248,
+      "learning_rate": 7.043332180443699e-06,
+      "loss": 5.4092,
+      "step": 17392
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2242355346679688,
+      "learning_rate": 7.033450269282079e-06,
+      "loss": 5.4268,
+      "step": 17396
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2366597652435303,
+      "learning_rate": 7.0235683581204605e-06,
+      "loss": 5.4304,
+      "step": 17400
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3268561363220215,
+      "learning_rate": 7.013686446958842e-06,
+      "loss": 5.4141,
+      "step": 17404
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1040186882019043,
+      "learning_rate": 7.003804535797223e-06,
+      "loss": 5.4129,
+      "step": 17408
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0050957202911377,
+      "learning_rate": 6.993922624635604e-06,
+      "loss": 5.5029,
+      "step": 17412
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.914214849472046,
+      "learning_rate": 6.984040713473987e-06,
+      "loss": 5.5394,
+      "step": 17416
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.112946033477783,
+      "learning_rate": 6.974158802312368e-06,
+      "loss": 5.4547,
+      "step": 17420
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.980510950088501,
+      "learning_rate": 6.9642768911507495e-06,
+      "loss": 5.4997,
+      "step": 17424
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.985985517501831,
+      "learning_rate": 6.954394979989131e-06,
+      "loss": 5.3671,
+      "step": 17428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.897262454032898,
+      "learning_rate": 6.944513068827511e-06,
+      "loss": 5.4578,
+      "step": 17432
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9851828813552856,
+      "learning_rate": 6.934631157665892e-06,
+      "loss": 5.4318,
+      "step": 17436
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8977246284484863,
+      "learning_rate": 6.9247492465042735e-06,
+      "loss": 5.4777,
+      "step": 17440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1280171871185303,
+      "learning_rate": 6.9148673353426564e-06,
+      "loss": 5.583,
+      "step": 17444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3764641284942627,
+      "learning_rate": 6.904985424181038e-06,
+      "loss": 5.5559,
+      "step": 17448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9994136095046997,
+      "learning_rate": 6.895103513019419e-06,
+      "loss": 5.5662,
+      "step": 17452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.108659267425537,
+      "learning_rate": 6.8852216018578e-06,
+      "loss": 5.4098,
+      "step": 17456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9477959871292114,
+      "learning_rate": 6.875339690696181e-06,
+      "loss": 5.4161,
+      "step": 17460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2120134830474854,
+      "learning_rate": 6.865457779534562e-06,
+      "loss": 5.5344,
+      "step": 17464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9351931810379028,
+      "learning_rate": 6.855575868372943e-06,
+      "loss": 5.5115,
+      "step": 17468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9376587867736816,
+      "learning_rate": 6.845693957211324e-06,
+      "loss": 5.4218,
+      "step": 17472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8527143001556396,
+      "learning_rate": 6.835812046049707e-06,
+      "loss": 5.4781,
+      "step": 17476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9370919466018677,
+      "learning_rate": 6.825930134888088e-06,
+      "loss": 5.6185,
+      "step": 17480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8956094980239868,
+      "learning_rate": 6.816048223726469e-06,
+      "loss": 5.5335,
+      "step": 17484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1358373165130615,
+      "learning_rate": 6.806166312564851e-06,
+      "loss": 5.5501,
+      "step": 17488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8500255346298218,
+      "learning_rate": 6.796284401403232e-06,
+      "loss": 5.4718,
+      "step": 17492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9620647430419922,
+      "learning_rate": 6.786402490241612e-06,
+      "loss": 5.4566,
+      "step": 17496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0902743339538574,
+      "learning_rate": 6.776520579079993e-06,
+      "loss": 5.484,
+      "step": 17500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.329399824142456,
+      "learning_rate": 6.766638667918376e-06,
+      "loss": 5.5337,
+      "step": 17504
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9751675128936768,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 5.4638,
+      "step": 17508
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.188885450363159,
+      "learning_rate": 6.746874845595139e-06,
+      "loss": 5.5174,
+      "step": 17512
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1230249404907227,
+      "learning_rate": 6.73699293443352e-06,
+      "loss": 5.4525,
+      "step": 17516
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0748202800750732,
+      "learning_rate": 6.727111023271901e-06,
+      "loss": 5.5633,
+      "step": 17520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8766546249389648,
+      "learning_rate": 6.717229112110282e-06,
+      "loss": 5.5295,
+      "step": 17524
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1104044914245605,
+      "learning_rate": 6.707347200948664e-06,
+      "loss": 5.5265,
+      "step": 17528
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9020673036575317,
+      "learning_rate": 6.697465289787046e-06,
+      "loss": 5.4684,
+      "step": 17532
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9456652402877808,
+      "learning_rate": 6.687583378625427e-06,
+      "loss": 5.3826,
+      "step": 17536
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.117117166519165,
+      "learning_rate": 6.677701467463808e-06,
+      "loss": 5.4525,
+      "step": 17540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0873782634735107,
+      "learning_rate": 6.667819556302189e-06,
+      "loss": 5.5402,
+      "step": 17544
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0289838314056396,
+      "learning_rate": 6.6579376451405705e-06,
+      "loss": 5.5603,
+      "step": 17548
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2275471687316895,
+      "learning_rate": 6.648055733978952e-06,
+      "loss": 5.3792,
+      "step": 17552
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9133155345916748,
+      "learning_rate": 6.638173822817333e-06,
+      "loss": 5.4289,
+      "step": 17556
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.193645477294922,
+      "learning_rate": 6.628291911655715e-06,
+      "loss": 5.5224,
+      "step": 17560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1608972549438477,
+      "learning_rate": 6.618410000494096e-06,
+      "loss": 5.5412,
+      "step": 17564
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.141594648361206,
+      "learning_rate": 6.6085280893324774e-06,
+      "loss": 5.4787,
+      "step": 17568
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9416935443878174,
+      "learning_rate": 6.598646178170859e-06,
+      "loss": 5.3674,
+      "step": 17572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.110677480697632,
+      "learning_rate": 6.58876426700924e-06,
+      "loss": 5.4024,
+      "step": 17576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2235372066497803,
+      "learning_rate": 6.578882355847621e-06,
+      "loss": 5.4458,
+      "step": 17580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.280282974243164,
+      "learning_rate": 6.569000444686002e-06,
+      "loss": 5.5981,
+      "step": 17584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1084625720977783,
+      "learning_rate": 6.559118533524384e-06,
+      "loss": 5.4475,
+      "step": 17588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.004232406616211,
+      "learning_rate": 6.5492366223627656e-06,
+      "loss": 5.4314,
+      "step": 17592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9286199808120728,
+      "learning_rate": 6.539354711201147e-06,
+      "loss": 5.3943,
+      "step": 17596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9742239713668823,
+      "learning_rate": 6.529472800039528e-06,
+      "loss": 5.4633,
+      "step": 17600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1503305435180664,
+      "learning_rate": 6.519590888877909e-06,
+      "loss": 5.4654,
+      "step": 17604
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.996319055557251,
+      "learning_rate": 6.50970897771629e-06,
+      "loss": 5.4479,
+      "step": 17608
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1689870357513428,
+      "learning_rate": 6.499827066554672e-06,
+      "loss": 5.4242,
+      "step": 17612
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0061464309692383,
+      "learning_rate": 6.489945155393053e-06,
+      "loss": 5.3833,
+      "step": 17616
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1201388835906982,
+      "learning_rate": 6.480063244231435e-06,
+      "loss": 5.4374,
+      "step": 17620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.196545124053955,
+      "learning_rate": 6.470181333069816e-06,
+      "loss": 5.3564,
+      "step": 17624
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.073232412338257,
+      "learning_rate": 6.460299421908197e-06,
+      "loss": 5.4057,
+      "step": 17628
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9354524612426758,
+      "learning_rate": 6.4504175107465785e-06,
+      "loss": 5.4718,
+      "step": 17632
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.032994508743286,
+      "learning_rate": 6.44053559958496e-06,
+      "loss": 5.358,
+      "step": 17636
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.130598545074463,
+      "learning_rate": 6.430653688423341e-06,
+      "loss": 5.4658,
+      "step": 17640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8692468404769897,
+      "learning_rate": 6.420771777261722e-06,
+      "loss": 5.4004,
+      "step": 17644
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9368531703948975,
+      "learning_rate": 6.410889866100104e-06,
+      "loss": 5.4809,
+      "step": 17648
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.235506534576416,
+      "learning_rate": 6.4010079549384855e-06,
+      "loss": 5.4532,
+      "step": 17652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9980324506759644,
+      "learning_rate": 6.391126043776867e-06,
+      "loss": 5.5156,
+      "step": 17656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.947649598121643,
+      "learning_rate": 6.381244132615248e-06,
+      "loss": 5.3676,
+      "step": 17660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.365041971206665,
+      "learning_rate": 6.371362221453629e-06,
+      "loss": 5.4387,
+      "step": 17664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2820627689361572,
+      "learning_rate": 6.36148031029201e-06,
+      "loss": 5.4541,
+      "step": 17668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.02691650390625,
+      "learning_rate": 6.3515983991303915e-06,
+      "loss": 5.5084,
+      "step": 17672
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0064783096313477,
+      "learning_rate": 6.3417164879687744e-06,
+      "loss": 5.5205,
+      "step": 17676
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9961150884628296,
+      "learning_rate": 6.331834576807155e-06,
+      "loss": 5.3904,
+      "step": 17680
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2273404598236084,
+      "learning_rate": 6.321952665645536e-06,
+      "loss": 5.4756,
+      "step": 17684
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.078472852706909,
+      "learning_rate": 6.312070754483917e-06,
+      "loss": 5.5535,
+      "step": 17688
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9248629808425903,
+      "learning_rate": 6.3021888433222984e-06,
+      "loss": 5.4378,
+      "step": 17692
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1985530853271484,
+      "learning_rate": 6.29230693216068e-06,
+      "loss": 5.4345,
+      "step": 17696
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.087536096572876,
+      "learning_rate": 6.282425020999061e-06,
+      "loss": 5.5135,
+      "step": 17700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0699515342712402,
+      "learning_rate": 6.272543109837444e-06,
+      "loss": 5.3821,
+      "step": 17704
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.309680223464966,
+      "learning_rate": 6.262661198675825e-06,
+      "loss": 5.3426,
+      "step": 17708
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9877557754516602,
+      "learning_rate": 6.252779287514206e-06,
+      "loss": 5.554,
+      "step": 17712
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1621484756469727,
+      "learning_rate": 6.2428973763525866e-06,
+      "loss": 5.4897,
+      "step": 17716
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0568161010742188,
+      "learning_rate": 6.233015465190968e-06,
+      "loss": 5.4726,
+      "step": 17720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8659361600875854,
+      "learning_rate": 6.22313355402935e-06,
+      "loss": 5.461,
+      "step": 17724
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8461517095565796,
+      "learning_rate": 6.213251642867731e-06,
+      "loss": 5.3407,
+      "step": 17728
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2194485664367676,
+      "learning_rate": 6.203369731706112e-06,
+      "loss": 5.3863,
+      "step": 17732
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2594525814056396,
+      "learning_rate": 6.1934878205444935e-06,
+      "loss": 5.4737,
+      "step": 17736
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.067777156829834,
+      "learning_rate": 6.1836059093828755e-06,
+      "loss": 5.4817,
+      "step": 17740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9696800708770752,
+      "learning_rate": 6.173723998221257e-06,
+      "loss": 5.5215,
+      "step": 17744
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9700802564620972,
+      "learning_rate": 6.163842087059637e-06,
+      "loss": 5.5821,
+      "step": 17748
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2519845962524414,
+      "learning_rate": 6.153960175898018e-06,
+      "loss": 5.5501,
+      "step": 17752
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1531550884246826,
+      "learning_rate": 6.1440782647364e-06,
+      "loss": 5.354,
+      "step": 17756
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.299639940261841,
+      "learning_rate": 6.134196353574782e-06,
+      "loss": 5.5487,
+      "step": 17760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9032407999038696,
+      "learning_rate": 6.124314442413163e-06,
+      "loss": 5.4551,
+      "step": 17764
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.121720552444458,
+      "learning_rate": 6.114432531251545e-06,
+      "loss": 5.4185,
+      "step": 17768
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.955588698387146,
+      "learning_rate": 6.104550620089926e-06,
+      "loss": 5.5947,
+      "step": 17772
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9518580436706543,
+      "learning_rate": 6.094668708928307e-06,
+      "loss": 5.4159,
+      "step": 17776
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2284739017486572,
+      "learning_rate": 6.084786797766688e-06,
+      "loss": 5.4806,
+      "step": 17780
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9473198652267456,
+      "learning_rate": 6.07490488660507e-06,
+      "loss": 5.3457,
+      "step": 17784
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.25762939453125,
+      "learning_rate": 6.065022975443451e-06,
+      "loss": 5.3905,
+      "step": 17788
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.13055682182312,
+      "learning_rate": 6.055141064281832e-06,
+      "loss": 5.4596,
+      "step": 17792
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.078608751296997,
+      "learning_rate": 6.045259153120214e-06,
+      "loss": 5.4836,
+      "step": 17796
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.032860040664673,
+      "learning_rate": 6.0353772419585954e-06,
+      "loss": 5.3315,
+      "step": 17800
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.18186616897583,
+      "learning_rate": 6.025495330796977e-06,
+      "loss": 5.5165,
+      "step": 17804
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8949894905090332,
+      "learning_rate": 6.015613419635358e-06,
+      "loss": 5.4296,
+      "step": 17808
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9019147157669067,
+      "learning_rate": 6.005731508473739e-06,
+      "loss": 5.3742,
+      "step": 17812
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9749938249588013,
+      "learning_rate": 5.99584959731212e-06,
+      "loss": 5.4702,
+      "step": 17816
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9528026580810547,
+      "learning_rate": 5.9859676861505015e-06,
+      "loss": 5.4179,
+      "step": 17820
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.04555082321167,
+      "learning_rate": 5.976085774988883e-06,
+      "loss": 5.4807,
+      "step": 17824
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2078750133514404,
+      "learning_rate": 5.966203863827265e-06,
+      "loss": 5.4206,
+      "step": 17828
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1232731342315674,
+      "learning_rate": 5.956321952665646e-06,
+      "loss": 5.5452,
+      "step": 17832
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.724265217781067,
+      "learning_rate": 5.946440041504027e-06,
+      "loss": 5.3831,
+      "step": 17836
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0802602767944336,
+      "learning_rate": 5.936558130342408e-06,
+      "loss": 5.3925,
+      "step": 17840
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.913464069366455,
+      "learning_rate": 5.92667621918079e-06,
+      "loss": 5.4786,
+      "step": 17844
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9357552528381348,
+      "learning_rate": 5.916794308019171e-06,
+      "loss": 5.4415,
+      "step": 17848
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9869678020477295,
+      "learning_rate": 5.906912396857552e-06,
+      "loss": 5.4731,
+      "step": 17852
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1964402198791504,
+      "learning_rate": 5.897030485695934e-06,
+      "loss": 5.4538,
+      "step": 17856
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2341887950897217,
+      "learning_rate": 5.887148574534315e-06,
+      "loss": 5.4484,
+      "step": 17860
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2685790061950684,
+      "learning_rate": 5.8772666633726965e-06,
+      "loss": 5.4935,
+      "step": 17864
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9977366924285889,
+      "learning_rate": 5.867384752211079e-06,
+      "loss": 5.4866,
+      "step": 17868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8025336265563965,
+      "learning_rate": 5.857502841049459e-06,
+      "loss": 5.5315,
+      "step": 17872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8028703927993774,
+      "learning_rate": 5.84762092988784e-06,
+      "loss": 5.4048,
+      "step": 17876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0206375122070312,
+      "learning_rate": 5.837739018726221e-06,
+      "loss": 5.5057,
+      "step": 17880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2146549224853516,
+      "learning_rate": 5.8278571075646034e-06,
+      "loss": 5.4375,
+      "step": 17884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1485488414764404,
+      "learning_rate": 5.817975196402985e-06,
+      "loss": 5.4375,
+      "step": 17888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9976389408111572,
+      "learning_rate": 5.808093285241366e-06,
+      "loss": 5.5645,
+      "step": 17892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1577677726745605,
+      "learning_rate": 5.798211374079747e-06,
+      "loss": 5.4917,
+      "step": 17896
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.085784673690796,
+      "learning_rate": 5.788329462918129e-06,
+      "loss": 5.5115,
+      "step": 17900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.250061511993408,
+      "learning_rate": 5.77844755175651e-06,
+      "loss": 5.4932,
+      "step": 17904
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.076542854309082,
+      "learning_rate": 5.768565640594891e-06,
+      "loss": 5.3858,
+      "step": 17908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0458431243896484,
+      "learning_rate": 5.761154207223677e-06,
+      "loss": 5.5596,
+      "step": 17912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.075693130493164,
+      "learning_rate": 5.7512722960620585e-06,
+      "loss": 5.4962,
+      "step": 17916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.09002685546875,
+      "learning_rate": 5.74139038490044e-06,
+      "loss": 5.3926,
+      "step": 17920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0116419792175293,
+      "learning_rate": 5.731508473738821e-06,
+      "loss": 5.4649,
+      "step": 17924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0170979499816895,
+      "learning_rate": 5.721626562577203e-06,
+      "loss": 5.4906,
+      "step": 17928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.033344268798828,
+      "learning_rate": 5.711744651415584e-06,
+      "loss": 5.5144,
+      "step": 17932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9842782020568848,
+      "learning_rate": 5.701862740253965e-06,
+      "loss": 5.4478,
+      "step": 17936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.059737205505371,
+      "learning_rate": 5.691980829092347e-06,
+      "loss": 5.4969,
+      "step": 17940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.116508960723877,
+      "learning_rate": 5.682098917930728e-06,
+      "loss": 5.5734,
+      "step": 17944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.062220573425293,
+      "learning_rate": 5.672217006769109e-06,
+      "loss": 5.3472,
+      "step": 17948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.176339626312256,
+      "learning_rate": 5.66233509560749e-06,
+      "loss": 5.3356,
+      "step": 17952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2417047023773193,
+      "learning_rate": 5.652453184445872e-06,
+      "loss": 5.429,
+      "step": 17956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9376815557479858,
+      "learning_rate": 5.6425712732842535e-06,
+      "loss": 5.4263,
+      "step": 17960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9888767004013062,
+      "learning_rate": 5.632689362122635e-06,
+      "loss": 5.5286,
+      "step": 17964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1054704189300537,
+      "learning_rate": 5.622807450961016e-06,
+      "loss": 5.4473,
+      "step": 17968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3070156574249268,
+      "learning_rate": 5.612925539799398e-06,
+      "loss": 5.6394,
+      "step": 17972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.956694483757019,
+      "learning_rate": 5.603043628637779e-06,
+      "loss": 5.3773,
+      "step": 17976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1512622833251953,
+      "learning_rate": 5.59316171747616e-06,
+      "loss": 5.5051,
+      "step": 17980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9715803861618042,
+      "learning_rate": 5.583279806314542e-06,
+      "loss": 5.4564,
+      "step": 17984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.07094144821167,
+      "learning_rate": 5.573397895152923e-06,
+      "loss": 5.486,
+      "step": 17988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0776047706604004,
+      "learning_rate": 5.563515983991304e-06,
+      "loss": 5.4101,
+      "step": 17992
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0641090869903564,
+      "learning_rate": 5.553634072829685e-06,
+      "loss": 5.4413,
+      "step": 17996
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.069200038909912,
+      "learning_rate": 5.543752161668067e-06,
+      "loss": 5.4239,
+      "step": 18000
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9378856420516968,
+      "learning_rate": 5.5338702505064486e-06,
+      "loss": 5.4994,
+      "step": 18004
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9170506000518799,
+      "learning_rate": 5.52398833934483e-06,
+      "loss": 5.4589,
+      "step": 18008
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1123738288879395,
+      "learning_rate": 5.514106428183211e-06,
+      "loss": 5.4226,
+      "step": 18012
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9895274639129639,
+      "learning_rate": 5.504224517021592e-06,
+      "loss": 5.522,
+      "step": 18016
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.216384172439575,
+      "learning_rate": 5.494342605859973e-06,
+      "loss": 5.4323,
+      "step": 18020
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1332645416259766,
+      "learning_rate": 5.484460694698355e-06,
+      "loss": 5.3666,
+      "step": 18024
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0537400245666504,
+      "learning_rate": 5.474578783536737e-06,
+      "loss": 5.4988,
+      "step": 18028
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9897429943084717,
+      "learning_rate": 5.464696872375118e-06,
+      "loss": 5.4833,
+      "step": 18032
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.066513776779175,
+      "learning_rate": 5.454814961213499e-06,
+      "loss": 5.4636,
+      "step": 18036
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.251376152038574,
+      "learning_rate": 5.44493305005188e-06,
+      "loss": 5.5205,
+      "step": 18040
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1560051441192627,
+      "learning_rate": 5.4350511388902615e-06,
+      "loss": 5.4888,
+      "step": 18044
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9921746253967285,
+      "learning_rate": 5.425169227728643e-06,
+      "loss": 5.503,
+      "step": 18048
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9806662797927856,
+      "learning_rate": 5.415287316567024e-06,
+      "loss": 5.5299,
+      "step": 18052
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.086308479309082,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 5.3775,
+      "step": 18056
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9652711153030396,
+      "learning_rate": 5.395523494243787e-06,
+      "loss": 5.5683,
+      "step": 18060
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8376736640930176,
+      "learning_rate": 5.3856415830821685e-06,
+      "loss": 5.3905,
+      "step": 18064
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1515750885009766,
+      "learning_rate": 5.37575967192055e-06,
+      "loss": 5.421,
+      "step": 18068
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.18635630607605,
+      "learning_rate": 5.365877760758931e-06,
+      "loss": 5.4513,
+      "step": 18072
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2322137355804443,
+      "learning_rate": 5.355995849597312e-06,
+      "loss": 5.5696,
+      "step": 18076
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1388771533966064,
+      "learning_rate": 5.346113938435693e-06,
+      "loss": 5.5154,
+      "step": 18080
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1032564640045166,
+      "learning_rate": 5.336232027274075e-06,
+      "loss": 5.4737,
+      "step": 18084
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2419564723968506,
+      "learning_rate": 5.326350116112457e-06,
+      "loss": 5.4804,
+      "step": 18088
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1092734336853027,
+      "learning_rate": 5.316468204950838e-06,
+      "loss": 5.523,
+      "step": 18092
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9642736911773682,
+      "learning_rate": 5.306586293789219e-06,
+      "loss": 5.4615,
+      "step": 18096
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.285712480545044,
+      "learning_rate": 5.296704382627601e-06,
+      "loss": 5.4944,
+      "step": 18100
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8048274517059326,
+      "learning_rate": 5.2868224714659814e-06,
+      "loss": 5.3293,
+      "step": 18104
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9001215696334839,
+      "learning_rate": 5.276940560304363e-06,
+      "loss": 5.3799,
+      "step": 18108
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.052248954772949,
+      "learning_rate": 5.267058649142744e-06,
+      "loss": 5.54,
+      "step": 18112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9618264436721802,
+      "learning_rate": 5.257176737981126e-06,
+      "loss": 5.5206,
+      "step": 18116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.086357355117798,
+      "learning_rate": 5.247294826819507e-06,
+      "loss": 5.4713,
+      "step": 18120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.989790678024292,
+      "learning_rate": 5.237412915657888e-06,
+      "loss": 5.4183,
+      "step": 18124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4142305850982666,
+      "learning_rate": 5.22753100449627e-06,
+      "loss": 5.4867,
+      "step": 18128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9405925273895264,
+      "learning_rate": 5.217649093334652e-06,
+      "loss": 5.4259,
+      "step": 18132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9569774866104126,
+      "learning_rate": 5.207767182173033e-06,
+      "loss": 5.4619,
+      "step": 18136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1723146438598633,
+      "learning_rate": 5.197885271011413e-06,
+      "loss": 5.4719,
+      "step": 18140
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3420112133026123,
+      "learning_rate": 5.188003359849795e-06,
+      "loss": 5.4274,
+      "step": 18144
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2983791828155518,
+      "learning_rate": 5.1781214486881765e-06,
+      "loss": 5.5184,
+      "step": 18148
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.061795711517334,
+      "learning_rate": 5.168239537526558e-06,
+      "loss": 5.4416,
+      "step": 18152
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.11879301071167,
+      "learning_rate": 5.15835762636494e-06,
+      "loss": 5.3997,
+      "step": 18156
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.072601556777954,
+      "learning_rate": 5.148475715203321e-06,
+      "loss": 5.4744,
+      "step": 18160
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.037374258041382,
+      "learning_rate": 5.138593804041702e-06,
+      "loss": 5.5549,
+      "step": 18164
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1012215614318848,
+      "learning_rate": 5.128711892880083e-06,
+      "loss": 5.5601,
+      "step": 18168
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9614689350128174,
+      "learning_rate": 5.118829981718465e-06,
+      "loss": 5.3775,
+      "step": 18172
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3375091552734375,
+      "learning_rate": 5.108948070556846e-06,
+      "loss": 5.5373,
+      "step": 18176
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9838519096374512,
+      "learning_rate": 5.099066159395227e-06,
+      "loss": 5.3491,
+      "step": 18180
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0909204483032227,
+      "learning_rate": 5.089184248233608e-06,
+      "loss": 5.3386,
+      "step": 18184
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.396127223968506,
+      "learning_rate": 5.07930233707199e-06,
+      "loss": 5.4791,
+      "step": 18188
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.117344379425049,
+      "learning_rate": 5.0694204259103715e-06,
+      "loss": 5.3322,
+      "step": 18192
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0109291076660156,
+      "learning_rate": 5.059538514748753e-06,
+      "loss": 5.4152,
+      "step": 18196
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.051154851913452,
+      "learning_rate": 5.049656603587134e-06,
+      "loss": 5.504,
+      "step": 18200
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9939186573028564,
+      "learning_rate": 5.039774692425515e-06,
+      "loss": 5.377,
+      "step": 18204
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0900635719299316,
+      "learning_rate": 5.029892781263896e-06,
+      "loss": 5.4448,
+      "step": 18208
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7526922225952148,
+      "learning_rate": 5.020010870102278e-06,
+      "loss": 5.2948,
+      "step": 18212
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.082535982131958,
+      "learning_rate": 5.01012895894066e-06,
+      "loss": 5.4976,
+      "step": 18216
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.088693380355835,
+      "learning_rate": 5.000247047779041e-06,
+      "loss": 5.4579,
+      "step": 18220
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9704039096832275,
+      "learning_rate": 4.990365136617422e-06,
+      "loss": 5.5093,
+      "step": 18224
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2970244884490967,
+      "learning_rate": 4.980483225455803e-06,
+      "loss": 5.5017,
+      "step": 18228
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0478765964508057,
+      "learning_rate": 4.9706013142941845e-06,
+      "loss": 5.4761,
+      "step": 18232
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1008124351501465,
+      "learning_rate": 4.960719403132566e-06,
+      "loss": 5.5902,
+      "step": 18236
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.41119384765625,
+      "learning_rate": 4.950837491970947e-06,
+      "loss": 5.613,
+      "step": 18240
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9696173667907715,
+      "learning_rate": 4.940955580809329e-06,
+      "loss": 5.4408,
+      "step": 18244
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0599782466888428,
+      "learning_rate": 4.93107366964771e-06,
+      "loss": 5.455,
+      "step": 18248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1011674404144287,
+      "learning_rate": 4.921191758486091e-06,
+      "loss": 5.4402,
+      "step": 18252
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.209735870361328,
+      "learning_rate": 4.9113098473244735e-06,
+      "loss": 5.378,
+      "step": 18256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.093017101287842,
+      "learning_rate": 4.901427936162855e-06,
+      "loss": 5.4387,
+      "step": 18260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.14595103263855,
+      "learning_rate": 4.891546025001235e-06,
+      "loss": 5.5165,
+      "step": 18264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3589043617248535,
+      "learning_rate": 4.881664113839616e-06,
+      "loss": 5.4973,
+      "step": 18268
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.041437864303589,
+      "learning_rate": 4.871782202677998e-06,
+      "loss": 5.5404,
+      "step": 18272
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.244915723800659,
+      "learning_rate": 4.8619002915163795e-06,
+      "loss": 5.3885,
+      "step": 18276
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1960060596466064,
+      "learning_rate": 4.852018380354761e-06,
+      "loss": 5.534,
+      "step": 18280
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.032874345779419,
+      "learning_rate": 4.842136469193142e-06,
+      "loss": 5.4292,
+      "step": 18284
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9136545658111572,
+      "learning_rate": 4.832254558031524e-06,
+      "loss": 5.511,
+      "step": 18288
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0950520038604736,
+      "learning_rate": 4.822372646869905e-06,
+      "loss": 5.414,
+      "step": 18292
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0211408138275146,
+      "learning_rate": 4.812490735708286e-06,
+      "loss": 5.4139,
+      "step": 18296
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.994523286819458,
+      "learning_rate": 4.802608824546668e-06,
+      "loss": 5.4814,
+      "step": 18300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0099375247955322,
+      "learning_rate": 4.792726913385049e-06,
+      "loss": 5.4478,
+      "step": 18304
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9458421468734741,
+      "learning_rate": 4.78284500222343e-06,
+      "loss": 5.4538,
+      "step": 18308
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.13274884223938,
+      "learning_rate": 4.772963091061811e-06,
+      "loss": 5.5146,
+      "step": 18312
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.223994731903076,
+      "learning_rate": 4.763081179900193e-06,
+      "loss": 5.3984,
+      "step": 18316
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0752246379852295,
+      "learning_rate": 4.753199268738575e-06,
+      "loss": 5.4198,
+      "step": 18320
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.913928508758545,
+      "learning_rate": 4.743317357576956e-06,
+      "loss": 5.4548,
+      "step": 18324
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9709479808807373,
+      "learning_rate": 4.733435446415337e-06,
+      "loss": 5.432,
+      "step": 18328
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.19464111328125,
+      "learning_rate": 4.723553535253718e-06,
+      "loss": 5.5339,
+      "step": 18332
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9611873626708984,
+      "learning_rate": 4.7136716240920994e-06,
+      "loss": 5.6126,
+      "step": 18336
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9830701351165771,
+      "learning_rate": 4.703789712930481e-06,
+      "loss": 5.413,
+      "step": 18340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.189093828201294,
+      "learning_rate": 4.693907801768863e-06,
+      "loss": 5.4765,
+      "step": 18344
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.831568956375122,
+      "learning_rate": 4.684025890607244e-06,
+      "loss": 5.4116,
+      "step": 18348
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8038347959518433,
+      "learning_rate": 4.674143979445625e-06,
+      "loss": 5.3937,
+      "step": 18352
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.043757200241089,
+      "learning_rate": 4.664262068284006e-06,
+      "loss": 5.3792,
+      "step": 18356
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.10916805267334,
+      "learning_rate": 4.6543801571223876e-06,
+      "loss": 5.4894,
+      "step": 18360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1079790592193604,
+      "learning_rate": 4.644498245960769e-06,
+      "loss": 5.418,
+      "step": 18364
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.068610191345215,
+      "learning_rate": 4.63461633479915e-06,
+      "loss": 5.4669,
+      "step": 18368
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.140131950378418,
+      "learning_rate": 4.624734423637532e-06,
+      "loss": 5.4256,
+      "step": 18372
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9517799615859985,
+      "learning_rate": 4.614852512475913e-06,
+      "loss": 5.4603,
+      "step": 18376
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0555922985076904,
+      "learning_rate": 4.6049706013142945e-06,
+      "loss": 5.594,
+      "step": 18380
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1756505966186523,
+      "learning_rate": 4.595088690152676e-06,
+      "loss": 5.4045,
+      "step": 18384
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0365309715270996,
+      "learning_rate": 4.585206778991057e-06,
+      "loss": 5.4668,
+      "step": 18388
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.979575753211975,
+      "learning_rate": 4.575324867829438e-06,
+      "loss": 5.5546,
+      "step": 18392
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1645045280456543,
+      "learning_rate": 4.565442956667819e-06,
+      "loss": 5.4158,
+      "step": 18396
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.400611162185669,
+      "learning_rate": 4.555561045506201e-06,
+      "loss": 5.4139,
+      "step": 18400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1293954849243164,
+      "learning_rate": 4.545679134344583e-06,
+      "loss": 5.3817,
+      "step": 18404
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0387656688690186,
+      "learning_rate": 4.535797223182964e-06,
+      "loss": 5.533,
+      "step": 18408
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.231621265411377,
+      "learning_rate": 4.525915312021345e-06,
+      "loss": 5.4445,
+      "step": 18412
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0062594413757324,
+      "learning_rate": 4.516033400859727e-06,
+      "loss": 5.3092,
+      "step": 18416
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9188722372055054,
+      "learning_rate": 4.506151489698108e-06,
+      "loss": 5.4884,
+      "step": 18420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9304651021957397,
+      "learning_rate": 4.496269578536489e-06,
+      "loss": 5.5277,
+      "step": 18424
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.186587333679199,
+      "learning_rate": 4.48638766737487e-06,
+      "loss": 5.494,
+      "step": 18428
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.170325994491577,
+      "learning_rate": 4.476505756213252e-06,
+      "loss": 5.5372,
+      "step": 18432
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1663074493408203,
+      "learning_rate": 4.466623845051633e-06,
+      "loss": 5.4398,
+      "step": 18436
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.090306043624878,
+      "learning_rate": 4.456741933890014e-06,
+      "loss": 5.5293,
+      "step": 18440
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9806874990463257,
+      "learning_rate": 4.4468600227283964e-06,
+      "loss": 5.4678,
+      "step": 18444
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.011406183242798,
+      "learning_rate": 4.436978111566778e-06,
+      "loss": 5.4633,
+      "step": 18448
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8348437547683716,
+      "learning_rate": 4.427096200405159e-06,
+      "loss": 5.4313,
+      "step": 18452
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4174129962921143,
+      "learning_rate": 4.417214289243539e-06,
+      "loss": 5.5076,
+      "step": 18456
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0693600177764893,
+      "learning_rate": 4.407332378081921e-06,
+      "loss": 5.505,
+      "step": 18460
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9076710939407349,
+      "learning_rate": 4.3974504669203025e-06,
+      "loss": 5.3791,
+      "step": 18464
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2067015171051025,
+      "learning_rate": 4.387568555758684e-06,
+      "loss": 5.4183,
+      "step": 18468
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.229694128036499,
+      "learning_rate": 4.377686644597066e-06,
+      "loss": 5.4481,
+      "step": 18472
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3484323024749756,
+      "learning_rate": 4.367804733435447e-06,
+      "loss": 5.55,
+      "step": 18476
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2795610427856445,
+      "learning_rate": 4.357922822273828e-06,
+      "loss": 5.4688,
+      "step": 18480
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2929630279541016,
+      "learning_rate": 4.348040911112209e-06,
+      "loss": 5.5401,
+      "step": 18484
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.089888334274292,
+      "learning_rate": 4.338158999950591e-06,
+      "loss": 5.3467,
+      "step": 18488
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0417628288269043,
+      "learning_rate": 4.328277088788972e-06,
+      "loss": 5.3768,
+      "step": 18492
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.065882682800293,
+      "learning_rate": 4.318395177627353e-06,
+      "loss": 5.4829,
+      "step": 18496
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1247379779815674,
+      "learning_rate": 4.308513266465734e-06,
+      "loss": 5.4237,
+      "step": 18500
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8752723932266235,
+      "learning_rate": 4.298631355304116e-06,
+      "loss": 5.426,
+      "step": 18504
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2132692337036133,
+      "learning_rate": 4.2887494441424975e-06,
+      "loss": 5.4543,
+      "step": 18508
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0311672687530518,
+      "learning_rate": 4.278867532980879e-06,
+      "loss": 5.6122,
+      "step": 18512
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.083220958709717,
+      "learning_rate": 4.26898562181926e-06,
+      "loss": 5.4746,
+      "step": 18516
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9595463275909424,
+      "learning_rate": 4.259103710657641e-06,
+      "loss": 5.3808,
+      "step": 18520
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2142648696899414,
+      "learning_rate": 4.249221799496022e-06,
+      "loss": 5.4515,
+      "step": 18524
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9906036853790283,
+      "learning_rate": 4.239339888334404e-06,
+      "loss": 5.4185,
+      "step": 18528
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.193829298019409,
+      "learning_rate": 4.229457977172786e-06,
+      "loss": 5.5095,
+      "step": 18532
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1080212593078613,
+      "learning_rate": 4.219576066011167e-06,
+      "loss": 5.4778,
+      "step": 18536
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9774274826049805,
+      "learning_rate": 4.209694154849548e-06,
+      "loss": 5.557,
+      "step": 18540
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9373048543930054,
+      "learning_rate": 4.19981224368793e-06,
+      "loss": 5.2974,
+      "step": 18544
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1534974575042725,
+      "learning_rate": 4.1899303325263105e-06,
+      "loss": 5.3349,
+      "step": 18548
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8971552848815918,
+      "learning_rate": 4.180048421364692e-06,
+      "loss": 5.4853,
+      "step": 18552
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.132871627807617,
+      "learning_rate": 4.170166510203073e-06,
+      "loss": 5.4839,
+      "step": 18556
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.068107843399048,
+      "learning_rate": 4.160284599041455e-06,
+      "loss": 5.5167,
+      "step": 18560
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9894949197769165,
+      "learning_rate": 4.150402687879836e-06,
+      "loss": 5.5169,
+      "step": 18564
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.914143443107605,
+      "learning_rate": 4.1405207767182174e-06,
+      "loss": 5.484,
+      "step": 18568
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.168916940689087,
+      "learning_rate": 4.1306388655565995e-06,
+      "loss": 5.4795,
+      "step": 18572
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0595452785491943,
+      "learning_rate": 4.120756954394981e-06,
+      "loss": 5.5072,
+      "step": 18576
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9776825904846191,
+      "learning_rate": 4.110875043233361e-06,
+      "loss": 5.5014,
+      "step": 18580
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.120623826980591,
+      "learning_rate": 4.100993132071742e-06,
+      "loss": 5.4891,
+      "step": 18584
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.789355754852295,
+      "learning_rate": 4.091111220910124e-06,
+      "loss": 5.5514,
+      "step": 18588
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0536649227142334,
+      "learning_rate": 4.0812293097485056e-06,
+      "loss": 5.3581,
+      "step": 18592
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.785683035850525,
+      "learning_rate": 4.071347398586887e-06,
+      "loss": 5.4293,
+      "step": 18596
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9793387651443481,
+      "learning_rate": 4.061465487425268e-06,
+      "loss": 5.4698,
+      "step": 18600
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0507895946502686,
+      "learning_rate": 4.05158357626365e-06,
+      "loss": 5.3736,
+      "step": 18604
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2948641777038574,
+      "learning_rate": 4.041701665102031e-06,
+      "loss": 5.4733,
+      "step": 18608
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0238101482391357,
+      "learning_rate": 4.0318197539404125e-06,
+      "loss": 5.4463,
+      "step": 18612
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.103618860244751,
+      "learning_rate": 4.021937842778794e-06,
+      "loss": 5.4943,
+      "step": 18616
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.038139581680298,
+      "learning_rate": 4.012055931617175e-06,
+      "loss": 5.5276,
+      "step": 18620
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.837792158126831,
+      "learning_rate": 4.002174020455556e-06,
+      "loss": 5.4706,
+      "step": 18624
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0344648361206055,
+      "learning_rate": 3.992292109293937e-06,
+      "loss": 5.3494,
+      "step": 18628
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0213499069213867,
+      "learning_rate": 3.982410198132319e-06,
+      "loss": 5.5867,
+      "step": 18632
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.209096670150757,
+      "learning_rate": 3.972528286970701e-06,
+      "loss": 5.5146,
+      "step": 18636
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.924126148223877,
+      "learning_rate": 3.962646375809082e-06,
+      "loss": 5.4478,
+      "step": 18640
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2356016635894775,
+      "learning_rate": 3.952764464647463e-06,
+      "loss": 5.5293,
+      "step": 18644
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2954514026641846,
+      "learning_rate": 3.942882553485844e-06,
+      "loss": 5.4561,
+      "step": 18648
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.275831460952759,
+      "learning_rate": 3.9330006423242254e-06,
+      "loss": 5.4746,
+      "step": 18652
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.116305112838745,
+      "learning_rate": 3.923118731162607e-06,
+      "loss": 5.6006,
+      "step": 18656
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1635780334472656,
+      "learning_rate": 3.913236820000989e-06,
+      "loss": 5.5606,
+      "step": 18660
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8676637411117554,
+      "learning_rate": 3.90335490883937e-06,
+      "loss": 5.4456,
+      "step": 18664
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.344409227371216,
+      "learning_rate": 3.893472997677751e-06,
+      "loss": 5.5425,
+      "step": 18668
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0696935653686523,
+      "learning_rate": 3.883591086516132e-06,
+      "loss": 5.4989,
+      "step": 18672
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1459434032440186,
+      "learning_rate": 3.8737091753545136e-06,
+      "loss": 5.4776,
+      "step": 18676
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0918266773223877,
+      "learning_rate": 3.863827264192895e-06,
+      "loss": 5.408,
+      "step": 18680
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6926395893096924,
+      "learning_rate": 3.853945353031276e-06,
+      "loss": 5.4212,
+      "step": 18684
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9440195560455322,
+      "learning_rate": 3.844063441869658e-06,
+      "loss": 5.4617,
+      "step": 18688
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8561819791793823,
+      "learning_rate": 3.834181530708039e-06,
+      "loss": 5.5185,
+      "step": 18692
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1534957885742188,
+      "learning_rate": 3.8242996195464205e-06,
+      "loss": 5.3607,
+      "step": 18696
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9762316942214966,
+      "learning_rate": 3.8144177083848017e-06,
+      "loss": 5.4843,
+      "step": 18700
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9604766368865967,
+      "learning_rate": 3.8045357972231833e-06,
+      "loss": 5.5596,
+      "step": 18704
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.186389684677124,
+      "learning_rate": 3.7946538860615645e-06,
+      "loss": 5.5269,
+      "step": 18708
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.211263656616211,
+      "learning_rate": 3.7847719748999458e-06,
+      "loss": 5.4103,
+      "step": 18712
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0040395259857178,
+      "learning_rate": 3.7748900637383274e-06,
+      "loss": 5.4427,
+      "step": 18716
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.08552885055542,
+      "learning_rate": 3.7650081525767086e-06,
+      "loss": 5.3101,
+      "step": 18720
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.090424060821533,
+      "learning_rate": 3.75512624141509e-06,
+      "loss": 5.4502,
+      "step": 18724
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8602705001831055,
+      "learning_rate": 3.745244330253471e-06,
+      "loss": 5.4708,
+      "step": 18728
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9746719598770142,
+      "learning_rate": 3.7353624190918527e-06,
+      "loss": 5.3835,
+      "step": 18732
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9959601163864136,
+      "learning_rate": 3.725480507930234e-06,
+      "loss": 5.4041,
+      "step": 18736
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.080343008041382,
+      "learning_rate": 3.715598596768615e-06,
+      "loss": 5.5822,
+      "step": 18740
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.260117292404175,
+      "learning_rate": 3.7057166856069963e-06,
+      "loss": 5.4475,
+      "step": 18744
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2036492824554443,
+      "learning_rate": 3.695834774445378e-06,
+      "loss": 5.5317,
+      "step": 18748
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4585494995117188,
+      "learning_rate": 3.685952863283759e-06,
+      "loss": 5.5552,
+      "step": 18752
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2832350730895996,
+      "learning_rate": 3.6760709521221404e-06,
+      "loss": 5.4459,
+      "step": 18756
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.144115924835205,
+      "learning_rate": 3.666189040960522e-06,
+      "loss": 5.3545,
+      "step": 18760
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.214642286300659,
+      "learning_rate": 3.6563071297989032e-06,
+      "loss": 5.3576,
+      "step": 18764
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9967424869537354,
+      "learning_rate": 3.6464252186372844e-06,
+      "loss": 5.4909,
+      "step": 18768
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0453543663024902,
+      "learning_rate": 3.6365433074756657e-06,
+      "loss": 5.5049,
+      "step": 18772
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.09987473487854,
+      "learning_rate": 3.6266613963140473e-06,
+      "loss": 5.5611,
+      "step": 18776
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9165294170379639,
+      "learning_rate": 3.6167794851524285e-06,
+      "loss": 5.3851,
+      "step": 18780
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.128594160079956,
+      "learning_rate": 3.6068975739908097e-06,
+      "loss": 5.444,
+      "step": 18784
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.029412031173706,
+      "learning_rate": 3.5970156628291918e-06,
+      "loss": 5.6027,
+      "step": 18788
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0921339988708496,
+      "learning_rate": 3.587133751667573e-06,
+      "loss": 5.4492,
+      "step": 18792
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.300293445587158,
+      "learning_rate": 3.5772518405059538e-06,
+      "loss": 5.5684,
+      "step": 18796
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.034621000289917,
+      "learning_rate": 3.567369929344335e-06,
+      "loss": 5.5198,
+      "step": 18800
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3296048641204834,
+      "learning_rate": 3.557488018182717e-06,
+      "loss": 5.5555,
+      "step": 18804
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2251405715942383,
+      "learning_rate": 3.5476061070210983e-06,
+      "loss": 5.4131,
+      "step": 18808
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1167004108428955,
+      "learning_rate": 3.537724195859479e-06,
+      "loss": 5.3361,
+      "step": 18812
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.959514856338501,
+      "learning_rate": 3.527842284697861e-06,
+      "loss": 5.3026,
+      "step": 18816
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3255879878997803,
+      "learning_rate": 3.5179603735362423e-06,
+      "loss": 5.3271,
+      "step": 18820
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0048890113830566,
+      "learning_rate": 3.5080784623746235e-06,
+      "loss": 5.4256,
+      "step": 18824
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1140854358673096,
+      "learning_rate": 3.4981965512130043e-06,
+      "loss": 5.3592,
+      "step": 18828
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.857484221458435,
+      "learning_rate": 3.4883146400513864e-06,
+      "loss": 5.4143,
+      "step": 18832
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.063621997833252,
+      "learning_rate": 3.4784327288897676e-06,
+      "loss": 5.5095,
+      "step": 18836
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9271105527877808,
+      "learning_rate": 3.468550817728149e-06,
+      "loss": 5.4678,
+      "step": 18840
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9842108488082886,
+      "learning_rate": 3.4586689065665296e-06,
+      "loss": 5.4825,
+      "step": 18844
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.128753662109375,
+      "learning_rate": 3.4487869954049117e-06,
+      "loss": 5.4128,
+      "step": 18848
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0321123600006104,
+      "learning_rate": 3.438905084243293e-06,
+      "loss": 5.4735,
+      "step": 18852
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1090190410614014,
+      "learning_rate": 3.429023173081674e-06,
+      "loss": 5.4828,
+      "step": 18856
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.016251802444458,
+      "learning_rate": 3.4191412619200557e-06,
+      "loss": 5.56,
+      "step": 18860
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.863065481185913,
+      "learning_rate": 3.409259350758437e-06,
+      "loss": 5.456,
+      "step": 18864
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.226012706756592,
+      "learning_rate": 3.399377439596818e-06,
+      "loss": 5.5674,
+      "step": 18868
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3455286026000977,
+      "learning_rate": 3.3894955284351994e-06,
+      "loss": 5.5108,
+      "step": 18872
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.991924524307251,
+      "learning_rate": 3.379613617273581e-06,
+      "loss": 5.5043,
+      "step": 18876
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.287811279296875,
+      "learning_rate": 3.3697317061119622e-06,
+      "loss": 5.4942,
+      "step": 18880
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.099695920944214,
+      "learning_rate": 3.3598497949503434e-06,
+      "loss": 5.4295,
+      "step": 18884
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0772879123687744,
+      "learning_rate": 3.349967883788725e-06,
+      "loss": 5.4999,
+      "step": 18888
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.076507329940796,
+      "learning_rate": 3.3400859726271063e-06,
+      "loss": 5.3828,
+      "step": 18892
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9823633432388306,
+      "learning_rate": 3.3302040614654875e-06,
+      "loss": 5.4312,
+      "step": 18896
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.120603322982788,
+      "learning_rate": 3.3203221503038687e-06,
+      "loss": 5.4585,
+      "step": 18900
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9160094261169434,
+      "learning_rate": 3.3104402391422504e-06,
+      "loss": 5.4471,
+      "step": 18904
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0175113677978516,
+      "learning_rate": 3.3005583279806316e-06,
+      "loss": 5.3894,
+      "step": 18908
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.052076578140259,
+      "learning_rate": 3.2906764168190128e-06,
+      "loss": 5.5049,
+      "step": 18912
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0342178344726562,
+      "learning_rate": 3.280794505657394e-06,
+      "loss": 5.4793,
+      "step": 18916
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.004908561706543,
+      "learning_rate": 3.2709125944957756e-06,
+      "loss": 5.5442,
+      "step": 18920
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1404547691345215,
+      "learning_rate": 3.261030683334157e-06,
+      "loss": 5.4585,
+      "step": 18924
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1052112579345703,
+      "learning_rate": 3.251148772172538e-06,
+      "loss": 5.5015,
+      "step": 18928
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0737648010253906,
+      "learning_rate": 3.24126686101092e-06,
+      "loss": 5.5249,
+      "step": 18932
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.041416883468628,
+      "learning_rate": 3.231384949849301e-06,
+      "loss": 5.5562,
+      "step": 18936
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.186037063598633,
+      "learning_rate": 3.221503038687682e-06,
+      "loss": 5.4618,
+      "step": 18940
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2970268726348877,
+      "learning_rate": 3.2116211275260633e-06,
+      "loss": 5.3398,
+      "step": 18944
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2040741443634033,
+      "learning_rate": 3.2017392163644454e-06,
+      "loss": 5.48,
+      "step": 18948
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0961127281188965,
+      "learning_rate": 3.191857305202826e-06,
+      "loss": 5.5835,
+      "step": 18952
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.112224578857422,
+      "learning_rate": 3.1819753940412074e-06,
+      "loss": 5.5615,
+      "step": 18956
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9072363376617432,
+      "learning_rate": 3.1720934828795895e-06,
+      "loss": 5.4394,
+      "step": 18960
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0917303562164307,
+      "learning_rate": 3.1622115717179707e-06,
+      "loss": 5.4445,
+      "step": 18964
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9776668548583984,
+      "learning_rate": 3.152329660556352e-06,
+      "loss": 5.4521,
+      "step": 18968
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.028456449508667,
+      "learning_rate": 3.1424477493947327e-06,
+      "loss": 5.4723,
+      "step": 18972
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.263448715209961,
+      "learning_rate": 3.1325658382331147e-06,
+      "loss": 5.4633,
+      "step": 18976
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.058852434158325,
+      "learning_rate": 3.122683927071496e-06,
+      "loss": 5.4923,
+      "step": 18980
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.173109769821167,
+      "learning_rate": 3.112802015909877e-06,
+      "loss": 5.3346,
+      "step": 18984
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0743250846862793,
+      "learning_rate": 3.1029201047482584e-06,
+      "loss": 5.5121,
+      "step": 18988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.125735282897949,
+      "learning_rate": 3.09303819358664e-06,
+      "loss": 5.3462,
+      "step": 18992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9164544343948364,
+      "learning_rate": 3.0831562824250212e-06,
+      "loss": 5.4138,
+      "step": 18996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9155079126358032,
+      "learning_rate": 3.0732743712634024e-06,
+      "loss": 5.3672,
+      "step": 19000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.128096342086792,
+      "learning_rate": 3.0633924601017836e-06,
+      "loss": 5.4021,
+      "step": 19004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0933046340942383,
+      "learning_rate": 3.0535105489401653e-06,
+      "loss": 5.5964,
+      "step": 19008
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9437744617462158,
+      "learning_rate": 3.0436286377785465e-06,
+      "loss": 5.5408,
+      "step": 19012
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2945642471313477,
+      "learning_rate": 3.0337467266169277e-06,
+      "loss": 5.4818,
+      "step": 19016
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.067274570465088,
+      "learning_rate": 3.023864815455309e-06,
+      "loss": 5.4306,
+      "step": 19020
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.847240686416626,
+      "learning_rate": 3.0139829042936906e-06,
+      "loss": 5.451,
+      "step": 19024
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.06215238571167,
+      "learning_rate": 3.004100993132072e-06,
+      "loss": 5.406,
+      "step": 19028
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9967362880706787,
+      "learning_rate": 2.994219081970453e-06,
+      "loss": 5.4419,
+      "step": 19032
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.127033233642578,
+      "learning_rate": 2.9843371708088346e-06,
+      "loss": 5.4244,
+      "step": 19036
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.093514919281006,
+      "learning_rate": 2.974455259647216e-06,
+      "loss": 5.4256,
+      "step": 19040
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.253448963165283,
+      "learning_rate": 2.9645733484855975e-06,
+      "loss": 5.5029,
+      "step": 19044
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.996626377105713,
+      "learning_rate": 2.9546914373239783e-06,
+      "loss": 5.5046,
+      "step": 19048
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.028315782546997,
+      "learning_rate": 2.94480952616236e-06,
+      "loss": 5.5133,
+      "step": 19052
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8789162635803223,
+      "learning_rate": 2.934927615000741e-06,
+      "loss": 5.5341,
+      "step": 19056
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1197164058685303,
+      "learning_rate": 2.9250457038391228e-06,
+      "loss": 5.426,
+      "step": 19060
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.14929461479187,
+      "learning_rate": 2.915163792677504e-06,
+      "loss": 5.5187,
+      "step": 19064
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1319830417633057,
+      "learning_rate": 2.905281881515885e-06,
+      "loss": 5.4122,
+      "step": 19068
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1806631088256836,
+      "learning_rate": 2.895399970354267e-06,
+      "loss": 5.483,
+      "step": 19072
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.130634307861328,
+      "learning_rate": 2.885518059192648e-06,
+      "loss": 5.352,
+      "step": 19076
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8736112117767334,
+      "learning_rate": 2.8756361480310292e-06,
+      "loss": 5.4776,
+      "step": 19080
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9261388778686523,
+      "learning_rate": 2.8657542368694105e-06,
+      "loss": 5.4317,
+      "step": 19084
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2391560077667236,
+      "learning_rate": 2.855872325707792e-06,
+      "loss": 5.4883,
+      "step": 19088
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0670218467712402,
+      "learning_rate": 2.8459904145461733e-06,
+      "loss": 5.4033,
+      "step": 19092
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.906471610069275,
+      "learning_rate": 2.8361085033845545e-06,
+      "loss": 5.4008,
+      "step": 19096
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9118762016296387,
+      "learning_rate": 2.826226592222936e-06,
+      "loss": 5.4519,
+      "step": 19100
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1099867820739746,
+      "learning_rate": 2.8163446810613174e-06,
+      "loss": 5.4299,
+      "step": 19104
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8047434091567993,
+      "learning_rate": 2.806462769899699e-06,
+      "loss": 5.4628,
+      "step": 19108
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.058469772338867,
+      "learning_rate": 2.79658085873808e-06,
+      "loss": 5.4885,
+      "step": 19112
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.310060501098633,
+      "learning_rate": 2.7866989475764614e-06,
+      "loss": 5.5399,
+      "step": 19116
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1850757598876953,
+      "learning_rate": 2.7768170364148426e-06,
+      "loss": 5.397,
+      "step": 19120
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.979308009147644,
+      "learning_rate": 2.7669351252532243e-06,
+      "loss": 5.4149,
+      "step": 19124
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1982696056365967,
+      "learning_rate": 2.7570532140916055e-06,
+      "loss": 5.4355,
+      "step": 19128
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1380321979522705,
+      "learning_rate": 2.7471713029299867e-06,
+      "loss": 5.4576,
+      "step": 19132
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1724483966827393,
+      "learning_rate": 2.7372893917683683e-06,
+      "loss": 5.4849,
+      "step": 19136
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1920723915100098,
+      "learning_rate": 2.7274074806067496e-06,
+      "loss": 5.5044,
+      "step": 19140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0876924991607666,
+      "learning_rate": 2.7175255694451308e-06,
+      "loss": 5.5164,
+      "step": 19144
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.028972864151001,
+      "learning_rate": 2.707643658283512e-06,
+      "loss": 5.4324,
+      "step": 19148
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1649794578552246,
+      "learning_rate": 2.6977617471218936e-06,
+      "loss": 5.4183,
+      "step": 19152
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0742039680480957,
+      "learning_rate": 2.687879835960275e-06,
+      "loss": 5.5804,
+      "step": 19156
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0070149898529053,
+      "learning_rate": 2.677997924798656e-06,
+      "loss": 5.4086,
+      "step": 19160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.270585298538208,
+      "learning_rate": 2.6681160136370377e-06,
+      "loss": 5.3827,
+      "step": 19164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9209606647491455,
+      "learning_rate": 2.658234102475419e-06,
+      "loss": 5.4162,
+      "step": 19168
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2000796794891357,
+      "learning_rate": 2.6483521913138005e-06,
+      "loss": 5.5244,
+      "step": 19172
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1105525493621826,
+      "learning_rate": 2.6384702801521813e-06,
+      "loss": 5.554,
+      "step": 19176
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1889917850494385,
+      "learning_rate": 2.628588368990563e-06,
+      "loss": 5.434,
+      "step": 19180
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.020902395248413,
+      "learning_rate": 2.618706457828944e-06,
+      "loss": 5.5205,
+      "step": 19184
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8815635442733765,
+      "learning_rate": 2.608824546667326e-06,
+      "loss": 5.3136,
+      "step": 19188
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1251041889190674,
+      "learning_rate": 2.5989426355057066e-06,
+      "loss": 5.4888,
+      "step": 19192
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0390384197235107,
+      "learning_rate": 2.5890607243440882e-06,
+      "loss": 5.487,
+      "step": 19196
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1622934341430664,
+      "learning_rate": 2.57917881318247e-06,
+      "loss": 5.4759,
+      "step": 19200
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.190906524658203,
+      "learning_rate": 2.569296902020851e-06,
+      "loss": 5.505,
+      "step": 19204
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2829856872558594,
+      "learning_rate": 2.5594149908592323e-06,
+      "loss": 5.5217,
+      "step": 19208
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9600523710250854,
+      "learning_rate": 2.5495330796976135e-06,
+      "loss": 5.5042,
+      "step": 19212
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.987808346748352,
+      "learning_rate": 2.539651168535995e-06,
+      "loss": 5.5081,
+      "step": 19216
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.160310745239258,
+      "learning_rate": 2.5297692573743764e-06,
+      "loss": 5.41,
+      "step": 19220
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.06400728225708,
+      "learning_rate": 2.5198873462127576e-06,
+      "loss": 5.4132,
+      "step": 19224
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9328960180282593,
+      "learning_rate": 2.510005435051139e-06,
+      "loss": 5.4076,
+      "step": 19228
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1214709281921387,
+      "learning_rate": 2.5001235238895204e-06,
+      "loss": 5.4865,
+      "step": 19232
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.868773102760315,
+      "learning_rate": 2.4902416127279016e-06,
+      "loss": 5.41,
+      "step": 19236
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9432061910629272,
+      "learning_rate": 2.480359701566283e-06,
+      "loss": 5.4604,
+      "step": 19240
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9144906997680664,
+      "learning_rate": 2.4704777904046645e-06,
+      "loss": 5.4058,
+      "step": 19244
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.061899185180664,
+      "learning_rate": 2.4605958792430457e-06,
+      "loss": 5.3946,
+      "step": 19248
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.211974859237671,
+      "learning_rate": 2.4507139680814273e-06,
+      "loss": 5.5185,
+      "step": 19252
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.155259847640991,
+      "learning_rate": 2.440832056919808e-06,
+      "loss": 5.5318,
+      "step": 19256
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0900609493255615,
+      "learning_rate": 2.4309501457581898e-06,
+      "loss": 5.3854,
+      "step": 19260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.067147970199585,
+      "learning_rate": 2.421068234596571e-06,
+      "loss": 5.5763,
+      "step": 19264
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9264382123947144,
+      "learning_rate": 2.4111863234349526e-06,
+      "loss": 5.4579,
+      "step": 19268
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9205288887023926,
+      "learning_rate": 2.401304412273334e-06,
+      "loss": 5.5091,
+      "step": 19272
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4955880641937256,
+      "learning_rate": 2.391422501111715e-06,
+      "loss": 5.4334,
+      "step": 19276
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0582950115203857,
+      "learning_rate": 2.3815405899500967e-06,
+      "loss": 5.5041,
+      "step": 19280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9333152770996094,
+      "learning_rate": 2.371658678788478e-06,
+      "loss": 5.5563,
+      "step": 19284
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0240728855133057,
+      "learning_rate": 2.361776767626859e-06,
+      "loss": 5.4964,
+      "step": 19288
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.198359251022339,
+      "learning_rate": 2.3518948564652403e-06,
+      "loss": 5.5561,
+      "step": 19292
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.372915506362915,
+      "learning_rate": 2.342012945303622e-06,
+      "loss": 5.5221,
+      "step": 19296
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.183767557144165,
+      "learning_rate": 2.332131034142003e-06,
+      "loss": 5.4475,
+      "step": 19300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.019258499145508,
+      "learning_rate": 2.3222491229803844e-06,
+      "loss": 5.4531,
+      "step": 19304
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.098344564437866,
+      "learning_rate": 2.312367211818766e-06,
+      "loss": 5.4763,
+      "step": 19308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9019840955734253,
+      "learning_rate": 2.3024853006571472e-06,
+      "loss": 5.5028,
+      "step": 19312
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9219045639038086,
+      "learning_rate": 2.2926033894955284e-06,
+      "loss": 5.4599,
+      "step": 19316
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0754764080047607,
+      "learning_rate": 2.2827214783339097e-06,
+      "loss": 5.5103,
+      "step": 19320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8941829204559326,
+      "learning_rate": 2.2728395671722913e-06,
+      "loss": 5.4677,
+      "step": 19324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.03706431388855,
+      "learning_rate": 2.2629576560106725e-06,
+      "loss": 5.3742,
+      "step": 19328
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2529075145721436,
+      "learning_rate": 2.253075744849054e-06,
+      "loss": 5.4839,
+      "step": 19332
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.053737163543701,
+      "learning_rate": 2.243193833687435e-06,
+      "loss": 5.466,
+      "step": 19336
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1574110984802246,
+      "learning_rate": 2.2333119225258166e-06,
+      "loss": 5.3834,
+      "step": 19340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3568503856658936,
+      "learning_rate": 2.2234300113641982e-06,
+      "loss": 5.4472,
+      "step": 19344
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0544795989990234,
+      "learning_rate": 2.2135481002025794e-06,
+      "loss": 5.3323,
+      "step": 19348
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.053724527359009,
+      "learning_rate": 2.2036661890409606e-06,
+      "loss": 5.4349,
+      "step": 19352
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.132596015930176,
+      "learning_rate": 2.193784277879342e-06,
+      "loss": 5.4897,
+      "step": 19356
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1658577919006348,
+      "learning_rate": 2.1839023667177235e-06,
+      "loss": 5.4621,
+      "step": 19360
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2812771797180176,
+      "learning_rate": 2.1740204555561047e-06,
+      "loss": 5.3762,
+      "step": 19364
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.056185483932495,
+      "learning_rate": 2.164138544394486e-06,
+      "loss": 5.4444,
+      "step": 19368
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1531660556793213,
+      "learning_rate": 2.154256633232867e-06,
+      "loss": 5.5484,
+      "step": 19372
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3134000301361084,
+      "learning_rate": 2.1443747220712488e-06,
+      "loss": 5.5518,
+      "step": 19376
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.110779047012329,
+      "learning_rate": 2.13449281090963e-06,
+      "loss": 5.5833,
+      "step": 19380
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.998799443244934,
+      "learning_rate": 2.124610899748011e-06,
+      "loss": 5.4789,
+      "step": 19384
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.902309536933899,
+      "learning_rate": 2.114728988586393e-06,
+      "loss": 5.4095,
+      "step": 19388
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1755199432373047,
+      "learning_rate": 2.104847077424774e-06,
+      "loss": 5.5097,
+      "step": 19392
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.089999198913574,
+      "learning_rate": 2.0949651662631553e-06,
+      "loss": 5.4806,
+      "step": 19396
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.029437780380249,
+      "learning_rate": 2.0850832551015365e-06,
+      "loss": 5.406,
+      "step": 19400
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0628230571746826,
+      "learning_rate": 2.075201343939918e-06,
+      "loss": 5.5123,
+      "step": 19404
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2448441982269287,
+      "learning_rate": 2.0653194327782997e-06,
+      "loss": 5.5619,
+      "step": 19408
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9651387929916382,
+      "learning_rate": 2.0554375216166805e-06,
+      "loss": 5.4864,
+      "step": 19412
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1724853515625,
+      "learning_rate": 2.045555610455062e-06,
+      "loss": 5.4135,
+      "step": 19416
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.981934666633606,
+      "learning_rate": 2.0356736992934434e-06,
+      "loss": 5.5062,
+      "step": 19420
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0844616889953613,
+      "learning_rate": 2.025791788131825e-06,
+      "loss": 5.4076,
+      "step": 19424
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2012956142425537,
+      "learning_rate": 2.0159098769702062e-06,
+      "loss": 5.4093,
+      "step": 19428
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.386598587036133,
+      "learning_rate": 2.0060279658085874e-06,
+      "loss": 5.4528,
+      "step": 19432
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9763364791870117,
+      "learning_rate": 1.9961460546469687e-06,
+      "loss": 5.5794,
+      "step": 19436
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9587762355804443,
+      "learning_rate": 1.9862641434853503e-06,
+      "loss": 5.4508,
+      "step": 19440
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9256550073623657,
+      "learning_rate": 1.9763822323237315e-06,
+      "loss": 5.4952,
+      "step": 19444
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1667470932006836,
+      "learning_rate": 1.9665003211621127e-06,
+      "loss": 5.4601,
+      "step": 19448
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.241722583770752,
+      "learning_rate": 1.9566184100004944e-06,
+      "loss": 5.4552,
+      "step": 19452
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.051391363143921,
+      "learning_rate": 1.9467364988388756e-06,
+      "loss": 5.4654,
+      "step": 19456
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4016683101654053,
+      "learning_rate": 1.9368545876772568e-06,
+      "loss": 5.4888,
+      "step": 19460
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.049546480178833,
+      "learning_rate": 1.926972676515638e-06,
+      "loss": 5.4593,
+      "step": 19464
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.011448621749878,
+      "learning_rate": 1.9170907653540196e-06,
+      "loss": 5.4591,
+      "step": 19468
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.133927345275879,
+      "learning_rate": 1.9072088541924008e-06,
+      "loss": 5.5276,
+      "step": 19472
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1823043823242188,
+      "learning_rate": 1.8973269430307823e-06,
+      "loss": 5.4671,
+      "step": 19476
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2862277030944824,
+      "learning_rate": 1.8874450318691637e-06,
+      "loss": 5.3897,
+      "step": 19480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9178924560546875,
+      "learning_rate": 1.877563120707545e-06,
+      "loss": 5.4061,
+      "step": 19484
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.043980836868286,
+      "learning_rate": 1.8676812095459263e-06,
+      "loss": 5.4469,
+      "step": 19488
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0812246799468994,
+      "learning_rate": 1.8577992983843076e-06,
+      "loss": 5.4019,
+      "step": 19492
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1293153762817383,
+      "learning_rate": 1.847917387222689e-06,
+      "loss": 5.5026,
+      "step": 19496
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.150707960128784,
+      "learning_rate": 1.8380354760610702e-06,
+      "loss": 5.5382,
+      "step": 19500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0206191539764404,
+      "learning_rate": 1.8281535648994516e-06,
+      "loss": 5.3712,
+      "step": 19504
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0948994159698486,
+      "learning_rate": 1.8182716537378328e-06,
+      "loss": 5.3896,
+      "step": 19508
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.020963191986084,
+      "learning_rate": 1.8083897425762143e-06,
+      "loss": 5.3902,
+      "step": 19512
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0603950023651123,
+      "learning_rate": 1.7985078314145959e-06,
+      "loss": 5.4755,
+      "step": 19516
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9533694982528687,
+      "learning_rate": 1.7886259202529769e-06,
+      "loss": 5.3764,
+      "step": 19520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0614960193634033,
+      "learning_rate": 1.7787440090913585e-06,
+      "loss": 5.5436,
+      "step": 19524
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9552912712097168,
+      "learning_rate": 1.7688620979297395e-06,
+      "loss": 5.4593,
+      "step": 19528
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1261250972747803,
+      "learning_rate": 1.7589801867681212e-06,
+      "loss": 5.3616,
+      "step": 19532
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9987773895263672,
+      "learning_rate": 1.7490982756065022e-06,
+      "loss": 5.4697,
+      "step": 19536
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9402105808258057,
+      "learning_rate": 1.7392163644448838e-06,
+      "loss": 5.4345,
+      "step": 19540
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9414345026016235,
+      "learning_rate": 1.7293344532832648e-06,
+      "loss": 5.3996,
+      "step": 19544
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1511659622192383,
+      "learning_rate": 1.7194525421216464e-06,
+      "loss": 5.3444,
+      "step": 19548
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.011024236679077,
+      "learning_rate": 1.7095706309600279e-06,
+      "loss": 5.5161,
+      "step": 19552
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.05631422996521,
+      "learning_rate": 1.699688719798409e-06,
+      "loss": 5.451,
+      "step": 19556
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.112424373626709,
+      "learning_rate": 1.6898068086367905e-06,
+      "loss": 5.5912,
+      "step": 19560
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.210801839828491,
+      "learning_rate": 1.6799248974751717e-06,
+      "loss": 5.5403,
+      "step": 19564
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0372581481933594,
+      "learning_rate": 1.6700429863135531e-06,
+      "loss": 5.3212,
+      "step": 19568
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9437897205352783,
+      "learning_rate": 1.6601610751519344e-06,
+      "loss": 5.49,
+      "step": 19572
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0939133167266846,
+      "learning_rate": 1.6502791639903158e-06,
+      "loss": 5.5276,
+      "step": 19576
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0417065620422363,
+      "learning_rate": 1.640397252828697e-06,
+      "loss": 5.4277,
+      "step": 19580
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9575787782669067,
+      "learning_rate": 1.6305153416670784e-06,
+      "loss": 5.4794,
+      "step": 19584
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.010903835296631,
+      "learning_rate": 1.62063343050546e-06,
+      "loss": 5.5096,
+      "step": 19588
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9882851839065552,
+      "learning_rate": 1.610751519343841e-06,
+      "loss": 5.3729,
+      "step": 19592
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.00439190864563,
+      "learning_rate": 1.6008696081822227e-06,
+      "loss": 5.4072,
+      "step": 19596
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.154120445251465,
+      "learning_rate": 1.5909876970206037e-06,
+      "loss": 5.546,
+      "step": 19600
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1139557361602783,
+      "learning_rate": 1.5811057858589853e-06,
+      "loss": 5.5374,
+      "step": 19604
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.38275408744812,
+      "learning_rate": 1.5712238746973663e-06,
+      "loss": 5.6017,
+      "step": 19608
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0247836112976074,
+      "learning_rate": 1.561341963535748e-06,
+      "loss": 5.4871,
+      "step": 19612
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.97475004196167,
+      "learning_rate": 1.5514600523741292e-06,
+      "loss": 5.5351,
+      "step": 19616
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0974037647247314,
+      "learning_rate": 1.5415781412125106e-06,
+      "loss": 5.5301,
+      "step": 19620
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1904966831207275,
+      "learning_rate": 1.5316962300508918e-06,
+      "loss": 5.4878,
+      "step": 19624
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0616374015808105,
+      "learning_rate": 1.5218143188892732e-06,
+      "loss": 5.4575,
+      "step": 19628
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0660183429718018,
+      "learning_rate": 1.5119324077276545e-06,
+      "loss": 5.3544,
+      "step": 19632
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9736340045928955,
+      "learning_rate": 1.502050496566036e-06,
+      "loss": 5.4619,
+      "step": 19636
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.014892339706421,
+      "learning_rate": 1.4921685854044173e-06,
+      "loss": 5.3977,
+      "step": 19640
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8507190942764282,
+      "learning_rate": 1.4822866742427987e-06,
+      "loss": 5.4187,
+      "step": 19644
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1220788955688477,
+      "learning_rate": 1.47240476308118e-06,
+      "loss": 5.4812,
+      "step": 19648
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1677944660186768,
+      "learning_rate": 1.4625228519195614e-06,
+      "loss": 5.3252,
+      "step": 19652
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.054081439971924,
+      "learning_rate": 1.4526409407579426e-06,
+      "loss": 5.5602,
+      "step": 19656
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1197407245635986,
+      "learning_rate": 1.442759029596324e-06,
+      "loss": 5.4585,
+      "step": 19660
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.005307674407959,
+      "learning_rate": 1.4328771184347052e-06,
+      "loss": 5.4986,
+      "step": 19664
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.428250312805176,
+      "learning_rate": 1.4229952072730867e-06,
+      "loss": 5.3953,
+      "step": 19668
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.051632881164551,
+      "learning_rate": 1.413113296111468e-06,
+      "loss": 5.4902,
+      "step": 19672
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1171364784240723,
+      "learning_rate": 1.4032313849498495e-06,
+      "loss": 5.411,
+      "step": 19676
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9378799200057983,
+      "learning_rate": 1.3933494737882307e-06,
+      "loss": 5.401,
+      "step": 19680
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.151566743850708,
+      "learning_rate": 1.3834675626266121e-06,
+      "loss": 5.4693,
+      "step": 19684
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.948789358139038,
+      "learning_rate": 1.3735856514649934e-06,
+      "loss": 5.4927,
+      "step": 19688
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4134631156921387,
+      "learning_rate": 1.3637037403033748e-06,
+      "loss": 5.5223,
+      "step": 19692
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0687310695648193,
+      "learning_rate": 1.353821829141756e-06,
+      "loss": 5.6205,
+      "step": 19696
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.03088641166687,
+      "learning_rate": 1.3439399179801374e-06,
+      "loss": 5.4587,
+      "step": 19700
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0663702487945557,
+      "learning_rate": 1.3340580068185188e-06,
+      "loss": 5.4831,
+      "step": 19704
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.904789686203003,
+      "learning_rate": 1.3241760956569003e-06,
+      "loss": 5.5511,
+      "step": 19708
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.119781970977783,
+      "learning_rate": 1.3142941844952815e-06,
+      "loss": 5.4318,
+      "step": 19712
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9469795227050781,
+      "learning_rate": 1.304412273333663e-06,
+      "loss": 5.4406,
+      "step": 19716
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.30818247795105,
+      "learning_rate": 1.2945303621720441e-06,
+      "loss": 5.4206,
+      "step": 19720
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9821704626083374,
+      "learning_rate": 1.2846484510104255e-06,
+      "loss": 5.5063,
+      "step": 19724
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.095536470413208,
+      "learning_rate": 1.2747665398488068e-06,
+      "loss": 5.545,
+      "step": 19728
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.048340320587158,
+      "learning_rate": 1.2648846286871882e-06,
+      "loss": 5.386,
+      "step": 19732
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1303012371063232,
+      "learning_rate": 1.2550027175255694e-06,
+      "loss": 5.433,
+      "step": 19736
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1081206798553467,
+      "learning_rate": 1.2451208063639508e-06,
+      "loss": 5.4296,
+      "step": 19740
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0982227325439453,
+      "learning_rate": 1.2352388952023322e-06,
+      "loss": 5.4267,
+      "step": 19744
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9972717761993408,
+      "learning_rate": 1.2253569840407137e-06,
+      "loss": 5.55,
+      "step": 19748
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1928958892822266,
+      "learning_rate": 1.2154750728790949e-06,
+      "loss": 5.5068,
+      "step": 19752
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.973193883895874,
+      "learning_rate": 1.2055931617174763e-06,
+      "loss": 5.4586,
+      "step": 19756
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.975972056388855,
+      "learning_rate": 1.1957112505558575e-06,
+      "loss": 5.2701,
+      "step": 19760
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.053972005844116,
+      "learning_rate": 1.185829339394239e-06,
+      "loss": 5.4591,
+      "step": 19764
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.108062267303467,
+      "learning_rate": 1.1759474282326202e-06,
+      "loss": 5.4726,
+      "step": 19768
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1291093826293945,
+      "learning_rate": 1.1660655170710016e-06,
+      "loss": 5.459,
+      "step": 19772
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2266008853912354,
+      "learning_rate": 1.156183605909383e-06,
+      "loss": 5.4267,
+      "step": 19776
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.031205892562866,
+      "learning_rate": 1.1463016947477642e-06,
+      "loss": 5.3383,
+      "step": 19780
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1661624908447266,
+      "learning_rate": 1.1364197835861456e-06,
+      "loss": 5.4843,
+      "step": 19784
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1752281188964844,
+      "learning_rate": 1.126537872424527e-06,
+      "loss": 5.4579,
+      "step": 19788
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.279290199279785,
+      "learning_rate": 1.1166559612629083e-06,
+      "loss": 5.4887,
+      "step": 19792
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.843815803527832,
+      "learning_rate": 1.1067740501012897e-06,
+      "loss": 5.4187,
+      "step": 19796
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9832565784454346,
+      "learning_rate": 1.096892138939671e-06,
+      "loss": 5.5083,
+      "step": 19800
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7962124347686768,
+      "learning_rate": 1.0870102277780524e-06,
+      "loss": 5.5071,
+      "step": 19804
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0890607833862305,
+      "learning_rate": 1.0771283166164336e-06,
+      "loss": 5.56,
+      "step": 19808
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.972830057144165,
+      "learning_rate": 1.067246405454815e-06,
+      "loss": 5.4381,
+      "step": 19812
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.968068242073059,
+      "learning_rate": 1.0573644942931964e-06,
+      "loss": 5.5253,
+      "step": 19816
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.201663017272949,
+      "learning_rate": 1.0474825831315776e-06,
+      "loss": 5.5018,
+      "step": 19820
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8504890203475952,
+      "learning_rate": 1.037600671969959e-06,
+      "loss": 5.4115,
+      "step": 19824
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.112267017364502,
+      "learning_rate": 1.0277187608083403e-06,
+      "loss": 5.3483,
+      "step": 19828
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.159766435623169,
+      "learning_rate": 1.0178368496467217e-06,
+      "loss": 5.381,
+      "step": 19832
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1608057022094727,
+      "learning_rate": 1.0079549384851031e-06,
+      "loss": 5.4216,
+      "step": 19836
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8269460201263428,
+      "learning_rate": 9.980730273234843e-07,
+      "loss": 5.4642,
+      "step": 19840
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0793392658233643,
+      "learning_rate": 9.881911161618658e-07,
+      "loss": 5.4693,
+      "step": 19844
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1695215702056885,
+      "learning_rate": 9.783092050002472e-07,
+      "loss": 5.5016,
+      "step": 19848
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9932477474212646,
+      "learning_rate": 9.684272938386284e-07,
+      "loss": 5.4885,
+      "step": 19852
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9905871152877808,
+      "learning_rate": 9.585453826770098e-07,
+      "loss": 5.4697,
+      "step": 19856
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9750021696090698,
+      "learning_rate": 9.486634715153911e-07,
+      "loss": 5.3438,
+      "step": 19860
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9781215190887451,
+      "learning_rate": 9.387815603537725e-07,
+      "loss": 5.4243,
+      "step": 19864
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.112760543823242,
+      "learning_rate": 9.288996491921538e-07,
+      "loss": 5.496,
+      "step": 19868
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.029419422149658,
+      "learning_rate": 9.190177380305351e-07,
+      "loss": 5.4413,
+      "step": 19872
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0092504024505615,
+      "learning_rate": 9.091358268689164e-07,
+      "loss": 5.3783,
+      "step": 19876
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1386382579803467,
+      "learning_rate": 8.992539157072979e-07,
+      "loss": 5.4708,
+      "step": 19880
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9855784177780151,
+      "learning_rate": 8.893720045456793e-07,
+      "loss": 5.5455,
+      "step": 19884
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.007969856262207,
+      "learning_rate": 8.794900933840606e-07,
+      "loss": 5.5531,
+      "step": 19888
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9627690315246582,
+      "learning_rate": 8.696081822224419e-07,
+      "loss": 5.4983,
+      "step": 19892
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0733351707458496,
+      "learning_rate": 8.597262710608232e-07,
+      "loss": 5.4475,
+      "step": 19896
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1221306324005127,
+      "learning_rate": 8.498443598992045e-07,
+      "loss": 5.5056,
+      "step": 19900
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9957540035247803,
+      "learning_rate": 8.399624487375859e-07,
+      "loss": 5.5603,
+      "step": 19904
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1626851558685303,
+      "learning_rate": 8.300805375759672e-07,
+      "loss": 5.485,
+      "step": 19908
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9834141731262207,
+      "learning_rate": 8.201986264143485e-07,
+      "loss": 5.3205,
+      "step": 19912
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.065784454345703,
+      "learning_rate": 8.1031671525273e-07,
+      "loss": 5.5108,
+      "step": 19916
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.299056053161621,
+      "learning_rate": 8.004348040911113e-07,
+      "loss": 5.5952,
+      "step": 19920
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0861730575561523,
+      "learning_rate": 7.905528929294927e-07,
+      "loss": 5.4168,
+      "step": 19924
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2378146648406982,
+      "learning_rate": 7.80670981767874e-07,
+      "loss": 5.4531,
+      "step": 19928
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9151175022125244,
+      "learning_rate": 7.707890706062553e-07,
+      "loss": 5.4655,
+      "step": 19932
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.122283697128296,
+      "learning_rate": 7.609071594446366e-07,
+      "loss": 5.4724,
+      "step": 19936
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1069324016571045,
+      "learning_rate": 7.51025248283018e-07,
+      "loss": 5.5641,
+      "step": 19940
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.006833791732788,
+      "learning_rate": 7.411433371213994e-07,
+      "loss": 5.4192,
+      "step": 19944
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0627007484436035,
+      "learning_rate": 7.312614259597807e-07,
+      "loss": 5.5582,
+      "step": 19948
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9530029296875,
+      "learning_rate": 7.21379514798162e-07,
+      "loss": 5.5445,
+      "step": 19952
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1795194149017334,
+      "learning_rate": 7.114976036365433e-07,
+      "loss": 5.5048,
+      "step": 19956
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1359763145446777,
+      "learning_rate": 7.016156924749248e-07,
+      "loss": 5.432,
+      "step": 19960
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1916072368621826,
+      "learning_rate": 6.917337813133061e-07,
+      "loss": 5.4901,
+      "step": 19964
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.025327205657959,
+      "learning_rate": 6.818518701516874e-07,
+      "loss": 5.3755,
+      "step": 19968
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0348243713378906,
+      "learning_rate": 6.719699589900687e-07,
+      "loss": 5.5042,
+      "step": 19972
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.128100633621216,
+      "learning_rate": 6.620880478284501e-07,
+      "loss": 5.4892,
+      "step": 19976
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0665409564971924,
+      "learning_rate": 6.522061366668315e-07,
+      "loss": 5.4291,
+      "step": 19980
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2713510990142822,
+      "learning_rate": 6.423242255052128e-07,
+      "loss": 5.5485,
+      "step": 19984
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9565412998199463,
+      "learning_rate": 6.324423143435941e-07,
+      "loss": 5.3597,
+      "step": 19988
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.122183084487915,
+      "learning_rate": 6.225604031819754e-07,
+      "loss": 5.5183,
+      "step": 19992
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2601876258850098,
+      "learning_rate": 6.126784920203568e-07,
+      "loss": 5.3787,
+      "step": 19996
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9778292179107666,
+      "learning_rate": 6.027965808587382e-07,
+      "loss": 5.3719,
+      "step": 20000
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0290167331695557,
+      "learning_rate": 5.929146696971195e-07,
+      "loss": 5.4688,
+      "step": 20004
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.108403444290161,
+      "learning_rate": 5.830327585355008e-07,
+      "loss": 5.3245,
+      "step": 20008
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3024027347564697,
+      "learning_rate": 5.731508473738821e-07,
+      "loss": 5.4963,
+      "step": 20012
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.154008150100708,
+      "learning_rate": 5.632689362122635e-07,
+      "loss": 5.4071,
+      "step": 20016
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2379016876220703,
+      "learning_rate": 5.533870250506449e-07,
+      "loss": 5.4005,
+      "step": 20020
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0857393741607666,
+      "learning_rate": 5.435051138890262e-07,
+      "loss": 5.5153,
+      "step": 20024
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9951947927474976,
+      "learning_rate": 5.336232027274075e-07,
+      "loss": 5.4283,
+      "step": 20028
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.357630729675293,
+      "learning_rate": 5.237412915657888e-07,
+      "loss": 5.4484,
+      "step": 20032
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1432507038116455,
+      "learning_rate": 5.138593804041701e-07,
+      "loss": 5.4331,
+      "step": 20036
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8710029125213623,
+      "learning_rate": 5.039774692425516e-07,
+      "loss": 5.331,
+      "step": 20040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1204705238342285,
+      "learning_rate": 4.940955580809329e-07,
+      "loss": 5.3974,
+      "step": 20044
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2792441844940186,
+      "learning_rate": 4.842136469193142e-07,
+      "loss": 5.3835,
+      "step": 20048
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.204084873199463,
+      "learning_rate": 4.7433173575769557e-07,
+      "loss": 5.457,
+      "step": 20052
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0620133876800537,
+      "learning_rate": 4.644498245960769e-07,
+      "loss": 5.4719,
+      "step": 20056
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0034713745117188,
+      "learning_rate": 4.545679134344582e-07,
+      "loss": 5.3951,
+      "step": 20060
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9210227727890015,
+      "learning_rate": 4.4468600227283963e-07,
+      "loss": 5.5225,
+      "step": 20064
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.102591037750244,
+      "learning_rate": 4.3480409111122095e-07,
+      "loss": 5.5268,
+      "step": 20068
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0443971157073975,
+      "learning_rate": 4.2492217994960227e-07,
+      "loss": 5.5364,
+      "step": 20072
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.135408878326416,
+      "learning_rate": 4.150402687879836e-07,
+      "loss": 5.4666,
+      "step": 20076
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1176159381866455,
+      "learning_rate": 4.05158357626365e-07,
+      "loss": 5.4501,
+      "step": 20080
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8199737071990967,
+      "learning_rate": 3.9527644646474633e-07,
+      "loss": 5.4657,
+      "step": 20084
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.200198173522949,
+      "learning_rate": 3.8539453530312765e-07,
+      "loss": 5.6004,
+      "step": 20088
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9499452114105225,
+      "learning_rate": 3.75512624141509e-07,
+      "loss": 5.5219,
+      "step": 20092
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.084623098373413,
+      "learning_rate": 3.6563071297989034e-07,
+      "loss": 5.5269,
+      "step": 20096
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1379587650299072,
+      "learning_rate": 3.5574880181827166e-07,
+      "loss": 5.2866,
+      "step": 20100
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1811280250549316,
+      "learning_rate": 3.4586689065665304e-07,
+      "loss": 5.5607,
+      "step": 20104
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0732221603393555,
+      "learning_rate": 3.3598497949503435e-07,
+      "loss": 5.5122,
+      "step": 20108
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.181933641433716,
+      "learning_rate": 3.2610306833341573e-07,
+      "loss": 5.5222,
+      "step": 20112
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8368821144104004,
+      "learning_rate": 3.1622115717179705e-07,
+      "loss": 5.3693,
+      "step": 20116
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0082316398620605,
+      "learning_rate": 3.063392460101784e-07,
+      "loss": 5.3343,
+      "step": 20120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.952849268913269,
+      "learning_rate": 2.9645733484855974e-07,
+      "loss": 5.377,
+      "step": 20124
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9544694423675537,
+      "learning_rate": 2.8657542368694106e-07,
+      "loss": 5.4439,
+      "step": 20128
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0548641681671143,
+      "learning_rate": 2.7669351252532243e-07,
+      "loss": 5.5092,
+      "step": 20132
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.101043939590454,
+      "learning_rate": 2.6681160136370375e-07,
+      "loss": 5.3574,
+      "step": 20136
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1047868728637695,
+      "learning_rate": 2.5692969020208507e-07,
+      "loss": 5.513,
+      "step": 20140
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9184880256652832,
+      "learning_rate": 2.4704777904046644e-07,
+      "loss": 5.3459,
+      "step": 20144
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2628798484802246,
+      "learning_rate": 2.3716586787884778e-07,
+      "loss": 5.4285,
+      "step": 20148
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3330535888671875,
+      "learning_rate": 2.272839567172291e-07,
+      "loss": 5.4673,
+      "step": 20152
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0739777088165283,
+      "learning_rate": 2.1740204555561048e-07,
+      "loss": 5.4847,
+      "step": 20156
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.010349750518799,
+      "learning_rate": 2.075201343939918e-07,
+      "loss": 5.4364,
+      "step": 20160
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8513480424880981,
+      "learning_rate": 1.9763822323237317e-07,
+      "loss": 5.528,
+      "step": 20164
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0048437118530273,
+      "learning_rate": 1.877563120707545e-07,
+      "loss": 5.3088,
+      "step": 20168
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.953896403312683,
+      "learning_rate": 1.7787440090913583e-07,
+      "loss": 5.4566,
+      "step": 20172
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9878010749816895,
+      "learning_rate": 1.6799248974751718e-07,
+      "loss": 5.3852,
+      "step": 20176
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.017378330230713,
+      "learning_rate": 1.5811057858589852e-07,
+      "loss": 5.334,
+      "step": 20180
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9125686883926392,
+      "learning_rate": 1.4822866742427987e-07,
+      "loss": 5.4209,
+      "step": 20184
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.205244302749634,
+      "learning_rate": 1.3834675626266121e-07,
+      "loss": 5.443,
+      "step": 20188
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9783563613891602,
+      "learning_rate": 1.2846484510104253e-07,
+      "loss": 5.5629,
+      "step": 20192
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.003634214401245,
+      "learning_rate": 1.1858293393942389e-07,
+      "loss": 5.4012,
+      "step": 20196
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.183382511138916,
+      "learning_rate": 1.0870102277780524e-07,
+      "loss": 5.4401,
+      "step": 20200
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.161017656326294,
+      "learning_rate": 9.881911161618658e-08,
+      "loss": 5.5125,
+      "step": 20204
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1500487327575684,
+      "learning_rate": 8.893720045456792e-08,
+      "loss": 5.5267,
+      "step": 20208
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1386756896972656,
+      "learning_rate": 7.905528929294926e-08,
+      "loss": 5.4015,
+      "step": 20212
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1625399589538574,
+      "learning_rate": 6.917337813133061e-08,
+      "loss": 5.467,
+      "step": 20216
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1189723014831543,
+      "learning_rate": 5.9291466969711946e-08,
+      "loss": 5.4626,
+      "step": 20220
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1535654067993164,
+      "learning_rate": 4.940955580809329e-08,
+      "loss": 5.507,
+      "step": 20224
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8377280235290527,
+      "learning_rate": 3.952764464647463e-08,
+      "loss": 5.4731,
+      "step": 20228
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.746285080909729,
+      "learning_rate": 2.9645733484855973e-08,
+      "loss": 5.3422,
+      "step": 20232
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.335073947906494,
+      "learning_rate": 1.9763822323237315e-08,
+      "loss": 5.4435,
+      "step": 20236
+    },
+    {
+      "epoch": 1.0,
+      "step": 20239,
+      "total_flos": 8.525167704460493e+16,
+      "train_loss": 5.640737008268616,
+      "train_runtime": 3394.3955,
+      "train_samples_per_second": 95.396,
+      "train_steps_per_second": 5.962
+    }
+  ],
+  "logging_steps": 4,
+  "max_steps": 20239,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2024,
+  "total_flos": 8.525167704460493e+16,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}