diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6018 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9947460595446582,
+  "eval_steps": 500,
+  "global_step": 855,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0035026269702276708,
+      "grad_norm": 52.48405456542969,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 4.5474,
+      "step": 1
+    },
+    {
+      "epoch": 0.0070052539404553416,
+      "grad_norm": 15.457415580749512,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 3.8101,
+      "step": 2
+    },
+    {
+      "epoch": 0.010507880910683012,
+      "grad_norm": 38.854068756103516,
+      "learning_rate": 1.3953488372093025e-06,
+      "loss": 4.5719,
+      "step": 3
+    },
+    {
+      "epoch": 0.014010507880910683,
+      "grad_norm": 32.52230453491211,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 4.2939,
+      "step": 4
+    },
+    {
+      "epoch": 0.017513134851138354,
+      "grad_norm": 62.512908935546875,
+      "learning_rate": 2.3255813953488376e-06,
+      "loss": 4.9511,
+      "step": 5
+    },
+    {
+      "epoch": 0.021015761821366025,
+      "grad_norm": 14.109648704528809,
+      "learning_rate": 2.790697674418605e-06,
+      "loss": 3.9829,
+      "step": 6
+    },
+    {
+      "epoch": 0.024518388791593695,
+      "grad_norm": 39.390140533447266,
+      "learning_rate": 3.2558139534883724e-06,
+      "loss": 3.8983,
+      "step": 7
+    },
+    {
+      "epoch": 0.028021015761821366,
+      "grad_norm": 47.82828903198242,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 4.5586,
+      "step": 8
+    },
+    {
+      "epoch": 0.03152364273204904,
+      "grad_norm": 27.250408172607422,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 3.2116,
+      "step": 9
+    },
+    {
+      "epoch": 0.03502626970227671,
+      "grad_norm": 15.391385078430176,
+      "learning_rate": 4.651162790697675e-06,
+      "loss": 3.864,
+      "step": 10
+    },
+    {
+      "epoch": 0.03852889667250438,
+      "grad_norm": 15.911999702453613,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 3.4548,
+      "step": 11
+    },
+    {
+      "epoch": 0.04203152364273205,
+      "grad_norm": 11.446131706237793,
+      "learning_rate": 5.58139534883721e-06,
+      "loss": 3.3845,
+      "step": 12
+    },
+    {
+      "epoch": 0.04553415061295972,
+      "grad_norm": 21.21734619140625,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 3.9722,
+      "step": 13
+    },
+    {
+      "epoch": 0.04903677758318739,
+      "grad_norm": 19.188844680786133,
+      "learning_rate": 6.511627906976745e-06,
+      "loss": 4.4815,
+      "step": 14
+    },
+    {
+      "epoch": 0.05253940455341506,
+      "grad_norm": 15.476765632629395,
+      "learning_rate": 6.976744186046513e-06,
+      "loss": 3.9449,
+      "step": 15
+    },
+    {
+      "epoch": 0.05604203152364273,
+      "grad_norm": 8.746248245239258,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 3.0432,
+      "step": 16
+    },
+    {
+      "epoch": 0.0595446584938704,
+      "grad_norm": 14.563529014587402,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 3.4222,
+      "step": 17
+    },
+    {
+      "epoch": 0.06304728546409807,
+      "grad_norm": 9.088067054748535,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 3.2242,
+      "step": 18
+    },
+    {
+      "epoch": 0.06654991243432574,
+      "grad_norm": 18.267414093017578,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 3.4081,
+      "step": 19
+    },
+    {
+      "epoch": 0.07005253940455342,
+      "grad_norm": 19.622844696044922,
+      "learning_rate": 9.30232558139535e-06,
+      "loss": 3.1812,
+      "step": 20
+    },
+    {
+      "epoch": 0.07355516637478109,
+      "grad_norm": 14.69324016571045,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 2.7359,
+      "step": 21
+    },
+    {
+      "epoch": 0.07705779334500876,
+      "grad_norm": 16.522323608398438,
+      "learning_rate": 1.0232558139534884e-05,
+      "loss": 3.3083,
+      "step": 22
+    },
+    {
+      "epoch": 0.08056042031523643,
+      "grad_norm": 27.648794174194336,
+      "learning_rate": 1.0697674418604651e-05,
+      "loss": 3.7333,
+      "step": 23
+    },
+    {
+      "epoch": 0.0840630472854641,
+      "grad_norm": 18.54423713684082,
+      "learning_rate": 1.116279069767442e-05,
+      "loss": 3.013,
+      "step": 24
+    },
+    {
+      "epoch": 0.08756567425569177,
+      "grad_norm": 26.133230209350586,
+      "learning_rate": 1.1627906976744187e-05,
+      "loss": 3.5883,
+      "step": 25
+    },
+    {
+      "epoch": 0.09106830122591944,
+      "grad_norm": 11.354973793029785,
+      "learning_rate": 1.2093023255813954e-05,
+      "loss": 3.2263,
+      "step": 26
+    },
+    {
+      "epoch": 0.09457092819614711,
+      "grad_norm": 28.07408332824707,
+      "learning_rate": 1.2558139534883723e-05,
+      "loss": 3.4547,
+      "step": 27
+    },
+    {
+      "epoch": 0.09807355516637478,
+      "grad_norm": 34.68488693237305,
+      "learning_rate": 1.302325581395349e-05,
+      "loss": 4.8204,
+      "step": 28
+    },
+    {
+      "epoch": 0.10157618213660245,
+      "grad_norm": 11.963549613952637,
+      "learning_rate": 1.3488372093023257e-05,
+      "loss": 3.1184,
+      "step": 29
+    },
+    {
+      "epoch": 0.10507880910683012,
+      "grad_norm": 16.531707763671875,
+      "learning_rate": 1.3953488372093025e-05,
+      "loss": 3.4912,
+      "step": 30
+    },
+    {
+      "epoch": 0.1085814360770578,
+      "grad_norm": 20.2814998626709,
+      "learning_rate": 1.441860465116279e-05,
+      "loss": 3.6366,
+      "step": 31
+    },
+    {
+      "epoch": 0.11208406304728546,
+      "grad_norm": 19.060791015625,
+      "learning_rate": 1.488372093023256e-05,
+      "loss": 3.1751,
+      "step": 32
+    },
+    {
+      "epoch": 0.11558669001751314,
+      "grad_norm": 11.113019943237305,
+      "learning_rate": 1.5348837209302328e-05,
+      "loss": 3.1345,
+      "step": 33
+    },
+    {
+      "epoch": 0.1190893169877408,
+      "grad_norm": 17.613746643066406,
+      "learning_rate": 1.5813953488372095e-05,
+      "loss": 3.4793,
+      "step": 34
+    },
+    {
+      "epoch": 0.12259194395796848,
+      "grad_norm": 10.551315307617188,
+      "learning_rate": 1.6279069767441862e-05,
+      "loss": 3.0938,
+      "step": 35
+    },
+    {
+      "epoch": 0.12609457092819615,
+      "grad_norm": 30.171554565429688,
+      "learning_rate": 1.674418604651163e-05,
+      "loss": 4.7277,
+      "step": 36
+    },
+    {
+      "epoch": 0.1295971978984238,
+      "grad_norm": 12.733338356018066,
+      "learning_rate": 1.7209302325581396e-05,
+      "loss": 3.4881,
+      "step": 37
+    },
+    {
+      "epoch": 0.1330998248686515,
+      "grad_norm": 18.498449325561523,
+      "learning_rate": 1.7674418604651163e-05,
+      "loss": 4.3251,
+      "step": 38
+    },
+    {
+      "epoch": 0.13660245183887915,
+      "grad_norm": 18.08729362487793,
+      "learning_rate": 1.813953488372093e-05,
+      "loss": 3.8999,
+      "step": 39
+    },
+    {
+      "epoch": 0.14010507880910683,
+      "grad_norm": 12.728877067565918,
+      "learning_rate": 1.86046511627907e-05,
+      "loss": 3.3771,
+      "step": 40
+    },
+    {
+      "epoch": 0.1436077057793345,
+      "grad_norm": 19.182889938354492,
+      "learning_rate": 1.9069767441860468e-05,
+      "loss": 4.3478,
+      "step": 41
+    },
+    {
+      "epoch": 0.14711033274956217,
+      "grad_norm": 14.268775939941406,
+      "learning_rate": 1.9534883720930235e-05,
+      "loss": 3.8552,
+      "step": 42
+    },
+    {
+      "epoch": 0.15061295971978983,
+      "grad_norm": 32.71632766723633,
+      "learning_rate": 2e-05,
+      "loss": 2.8895,
+      "step": 43
+    },
+    {
+      "epoch": 0.15411558669001751,
+      "grad_norm": 15.663018226623535,
+      "learning_rate": 2.046511627906977e-05,
+      "loss": 3.9284,
+      "step": 44
+    },
+    {
+      "epoch": 0.15761821366024517,
+      "grad_norm": 12.49007797241211,
+      "learning_rate": 2.093023255813954e-05,
+      "loss": 3.397,
+      "step": 45
+    },
+    {
+      "epoch": 0.16112084063047286,
+      "grad_norm": 18.972797393798828,
+      "learning_rate": 2.1395348837209303e-05,
+      "loss": 3.4676,
+      "step": 46
+    },
+    {
+      "epoch": 0.1646234676007005,
+      "grad_norm": 17.138303756713867,
+      "learning_rate": 2.186046511627907e-05,
+      "loss": 3.5788,
+      "step": 47
+    },
+    {
+      "epoch": 0.1681260945709282,
+      "grad_norm": 13.487051010131836,
+      "learning_rate": 2.232558139534884e-05,
+      "loss": 3.618,
+      "step": 48
+    },
+    {
+      "epoch": 0.17162872154115585,
+      "grad_norm": 17.096302032470703,
+      "learning_rate": 2.2790697674418607e-05,
+      "loss": 3.3029,
+      "step": 49
+    },
+    {
+      "epoch": 0.17513134851138354,
+      "grad_norm": 9.788946151733398,
+      "learning_rate": 2.3255813953488374e-05,
+      "loss": 2.7551,
+      "step": 50
+    },
+    {
+      "epoch": 0.1786339754816112,
+      "grad_norm": 20.02460479736328,
+      "learning_rate": 2.3720930232558144e-05,
+      "loss": 2.5158,
+      "step": 51
+    },
+    {
+      "epoch": 0.18213660245183888,
+      "grad_norm": 13.219809532165527,
+      "learning_rate": 2.4186046511627908e-05,
+      "loss": 2.8995,
+      "step": 52
+    },
+    {
+      "epoch": 0.18563922942206654,
+      "grad_norm": 14.657979011535645,
+      "learning_rate": 2.4651162790697675e-05,
+      "loss": 3.384,
+      "step": 53
+    },
+    {
+      "epoch": 0.18914185639229422,
+      "grad_norm": 20.963659286499023,
+      "learning_rate": 2.5116279069767445e-05,
+      "loss": 3.3458,
+      "step": 54
+    },
+    {
+      "epoch": 0.19264448336252188,
+      "grad_norm": 12.124914169311523,
+      "learning_rate": 2.5581395348837212e-05,
+      "loss": 3.0666,
+      "step": 55
+    },
+    {
+      "epoch": 0.19614711033274956,
+      "grad_norm": 11.544058799743652,
+      "learning_rate": 2.604651162790698e-05,
+      "loss": 2.8553,
+      "step": 56
+    },
+    {
+      "epoch": 0.19964973730297722,
+      "grad_norm": 16.375518798828125,
+      "learning_rate": 2.6511627906976743e-05,
+      "loss": 3.2603,
+      "step": 57
+    },
+    {
+      "epoch": 0.2031523642732049,
+      "grad_norm": 14.338220596313477,
+      "learning_rate": 2.6976744186046514e-05,
+      "loss": 3.6208,
+      "step": 58
+    },
+    {
+      "epoch": 0.20665499124343256,
+      "grad_norm": 10.745012283325195,
+      "learning_rate": 2.744186046511628e-05,
+      "loss": 3.0888,
+      "step": 59
+    },
+    {
+      "epoch": 0.21015761821366025,
+      "grad_norm": 11.644838333129883,
+      "learning_rate": 2.790697674418605e-05,
+      "loss": 2.7899,
+      "step": 60
+    },
+    {
+      "epoch": 0.2136602451838879,
+      "grad_norm": 10.003059387207031,
+      "learning_rate": 2.8372093023255818e-05,
+      "loss": 3.314,
+      "step": 61
+    },
+    {
+      "epoch": 0.2171628721541156,
+      "grad_norm": 24.950681686401367,
+      "learning_rate": 2.883720930232558e-05,
+      "loss": 3.5718,
+      "step": 62
+    },
+    {
+      "epoch": 0.22066549912434325,
+      "grad_norm": 13.138785362243652,
+      "learning_rate": 2.930232558139535e-05,
+      "loss": 2.705,
+      "step": 63
+    },
+    {
+      "epoch": 0.22416812609457093,
+      "grad_norm": 10.878084182739258,
+      "learning_rate": 2.976744186046512e-05,
+      "loss": 3.1842,
+      "step": 64
+    },
+    {
+      "epoch": 0.2276707530647986,
+      "grad_norm": 16.99833106994629,
+      "learning_rate": 3.0232558139534886e-05,
+      "loss": 3.5814,
+      "step": 65
+    },
+    {
+      "epoch": 0.23117338003502627,
+      "grad_norm": 15.462761878967285,
+      "learning_rate": 3.0697674418604656e-05,
+      "loss": 3.1773,
+      "step": 66
+    },
+    {
+      "epoch": 0.23467600700525393,
+      "grad_norm": 13.846227645874023,
+      "learning_rate": 3.116279069767442e-05,
+      "loss": 2.386,
+      "step": 67
+    },
+    {
+      "epoch": 0.2381786339754816,
+      "grad_norm": 19.097707748413086,
+      "learning_rate": 3.162790697674419e-05,
+      "loss": 2.8701,
+      "step": 68
+    },
+    {
+      "epoch": 0.24168126094570927,
+      "grad_norm": 13.607734680175781,
+      "learning_rate": 3.2093023255813954e-05,
+      "loss": 3.9503,
+      "step": 69
+    },
+    {
+      "epoch": 0.24518388791593695,
+      "grad_norm": 11.81222152709961,
+      "learning_rate": 3.2558139534883724e-05,
+      "loss": 2.8775,
+      "step": 70
+    },
+    {
+      "epoch": 0.2486865148861646,
+      "grad_norm": 18.59412956237793,
+      "learning_rate": 3.3023255813953495e-05,
+      "loss": 3.9074,
+      "step": 71
+    },
+    {
+      "epoch": 0.2521891418563923,
+      "grad_norm": 25.35107421875,
+      "learning_rate": 3.348837209302326e-05,
+      "loss": 3.8891,
+      "step": 72
+    },
+    {
+      "epoch": 0.25569176882662,
+      "grad_norm": 16.84481430053711,
+      "learning_rate": 3.395348837209302e-05,
+      "loss": 3.8725,
+      "step": 73
+    },
+    {
+      "epoch": 0.2591943957968476,
+      "grad_norm": 17.052255630493164,
+      "learning_rate": 3.441860465116279e-05,
+      "loss": 3.0218,
+      "step": 74
+    },
+    {
+      "epoch": 0.2626970227670753,
+      "grad_norm": 36.11460876464844,
+      "learning_rate": 3.488372093023256e-05,
+      "loss": 4.1979,
+      "step": 75
+    },
+    {
+      "epoch": 0.266199649737303,
+      "grad_norm": 10.964736938476562,
+      "learning_rate": 3.5348837209302326e-05,
+      "loss": 3.1011,
+      "step": 76
+    },
+    {
+      "epoch": 0.26970227670753066,
+      "grad_norm": 13.844853401184082,
+      "learning_rate": 3.58139534883721e-05,
+      "loss": 3.6975,
+      "step": 77
+    },
+    {
+      "epoch": 0.2732049036777583,
+      "grad_norm": 12.19601821899414,
+      "learning_rate": 3.627906976744186e-05,
+      "loss": 3.1434,
+      "step": 78
+    },
+    {
+      "epoch": 0.276707530647986,
+      "grad_norm": 12.09930419921875,
+      "learning_rate": 3.674418604651163e-05,
+      "loss": 3.3209,
+      "step": 79
+    },
+    {
+      "epoch": 0.28021015761821366,
+      "grad_norm": 16.452251434326172,
+      "learning_rate": 3.72093023255814e-05,
+      "loss": 2.9085,
+      "step": 80
+    },
+    {
+      "epoch": 0.28371278458844135,
+      "grad_norm": 10.51562213897705,
+      "learning_rate": 3.7674418604651165e-05,
+      "loss": 3.3766,
+      "step": 81
+    },
+    {
+      "epoch": 0.287215411558669,
+      "grad_norm": 10.312739372253418,
+      "learning_rate": 3.8139534883720935e-05,
+      "loss": 3.2133,
+      "step": 82
+    },
+    {
+      "epoch": 0.29071803852889666,
+      "grad_norm": 15.232990264892578,
+      "learning_rate": 3.8604651162790706e-05,
+      "loss": 4.067,
+      "step": 83
+    },
+    {
+      "epoch": 0.29422066549912435,
+      "grad_norm": 13.83324146270752,
+      "learning_rate": 3.906976744186047e-05,
+      "loss": 3.2642,
+      "step": 84
+    },
+    {
+      "epoch": 0.29772329246935203,
+      "grad_norm": 14.54284954071045,
+      "learning_rate": 3.953488372093023e-05,
+      "loss": 3.2644,
+      "step": 85
+    },
+    {
+      "epoch": 0.30122591943957966,
+      "grad_norm": 18.364408493041992,
+      "learning_rate": 4e-05,
+      "loss": 3.0344,
+      "step": 86
+    },
+    {
+      "epoch": 0.30472854640980734,
+      "grad_norm": 11.513346672058105,
+      "learning_rate": 3.99998331037949e-05,
+      "loss": 2.5797,
+      "step": 87
+    },
+    {
+      "epoch": 0.30823117338003503,
+      "grad_norm": 19.537710189819336,
+      "learning_rate": 3.9999332417965036e-05,
+      "loss": 3.5762,
+      "step": 88
+    },
+    {
+      "epoch": 0.3117338003502627,
+      "grad_norm": 16.66529655456543,
+      "learning_rate": 3.999849795086666e-05,
+      "loss": 3.181,
+      "step": 89
+    },
+    {
+      "epoch": 0.31523642732049034,
+      "grad_norm": 8.610281944274902,
+      "learning_rate": 3.999732971642672e-05,
+      "loss": 2.9271,
+      "step": 90
+    },
+    {
+      "epoch": 0.318739054290718,
+      "grad_norm": 26.244932174682617,
+      "learning_rate": 3.999582773414259e-05,
+      "loss": 2.9479,
+      "step": 91
+    },
+    {
+      "epoch": 0.3222416812609457,
+      "grad_norm": 10.8726167678833,
+      "learning_rate": 3.99939920290818e-05,
+      "loss": 2.5413,
+      "step": 92
+    },
+    {
+      "epoch": 0.3257443082311734,
+      "grad_norm": 12.799675941467285,
+      "learning_rate": 3.999182263188156e-05,
+      "loss": 3.3469,
+      "step": 93
+    },
+    {
+      "epoch": 0.329246935201401,
+      "grad_norm": 13.35167121887207,
+      "learning_rate": 3.998931957874829e-05,
+      "loss": 3.5572,
+      "step": 94
+    },
+    {
+      "epoch": 0.3327495621716287,
+      "grad_norm": 10.501457214355469,
+      "learning_rate": 3.998648291145701e-05,
+      "loss": 2.5409,
+      "step": 95
+    },
+    {
+      "epoch": 0.3362521891418564,
+      "grad_norm": 13.109798431396484,
+      "learning_rate": 3.99833126773506e-05,
+      "loss": 3.4712,
+      "step": 96
+    },
+    {
+      "epoch": 0.3397548161120841,
+      "grad_norm": 12.54065227508545,
+      "learning_rate": 3.9979808929339066e-05,
+      "loss": 3.4943,
+      "step": 97
+    },
+    {
+      "epoch": 0.3432574430823117,
+      "grad_norm": 18.664926528930664,
+      "learning_rate": 3.997597172589865e-05,
+      "loss": 4.0018,
+      "step": 98
+    },
+    {
+      "epoch": 0.3467600700525394,
+      "grad_norm": 9.06640911102295,
+      "learning_rate": 3.9971801131070805e-05,
+      "loss": 3.5354,
+      "step": 99
+    },
+    {
+      "epoch": 0.3502626970227671,
+      "grad_norm": 10.051581382751465,
+      "learning_rate": 3.996729721446118e-05,
+      "loss": 3.6732,
+      "step": 100
+    },
+    {
+      "epoch": 0.35376532399299476,
+      "grad_norm": 10.670345306396484,
+      "learning_rate": 3.996246005123843e-05,
+      "loss": 3.422,
+      "step": 101
+    },
+    {
+      "epoch": 0.3572679509632224,
+      "grad_norm": 12.128016471862793,
+      "learning_rate": 3.995728972213299e-05,
+      "loss": 3.37,
+      "step": 102
+    },
+    {
+      "epoch": 0.3607705779334501,
+      "grad_norm": 15.131268501281738,
+      "learning_rate": 3.995178631343567e-05,
+      "loss": 3.8201,
+      "step": 103
+    },
+    {
+      "epoch": 0.36427320490367776,
+      "grad_norm": 11.882003784179688,
+      "learning_rate": 3.994594991699629e-05,
+      "loss": 2.4535,
+      "step": 104
+    },
+    {
+      "epoch": 0.36777583187390545,
+      "grad_norm": 13.227598190307617,
+      "learning_rate": 3.993978063022208e-05,
+      "loss": 3.5045,
+      "step": 105
+    },
+    {
+      "epoch": 0.3712784588441331,
+      "grad_norm": 9.773946762084961,
+      "learning_rate": 3.99332785560761e-05,
+      "loss": 2.7874,
+      "step": 106
+    },
+    {
+      "epoch": 0.37478108581436076,
+      "grad_norm": 27.06616973876953,
+      "learning_rate": 3.99264438030755e-05,
+      "loss": 3.155,
+      "step": 107
+    },
+    {
+      "epoch": 0.37828371278458844,
+      "grad_norm": 9.595654487609863,
+      "learning_rate": 3.991927648528971e-05,
+      "loss": 2.9589,
+      "step": 108
+    },
+    {
+      "epoch": 0.38178633975481613,
+      "grad_norm": 14.137282371520996,
+      "learning_rate": 3.9911776722338544e-05,
+      "loss": 3.6373,
+      "step": 109
+    },
+    {
+      "epoch": 0.38528896672504376,
+      "grad_norm": 11.451871871948242,
+      "learning_rate": 3.990394463939021e-05,
+      "loss": 3.41,
+      "step": 110
+    },
+    {
+      "epoch": 0.38879159369527144,
+      "grad_norm": 11.46674919128418,
+      "learning_rate": 3.989578036715919e-05,
+      "loss": 3.5972,
+      "step": 111
+    },
+    {
+      "epoch": 0.3922942206654991,
+      "grad_norm": 16.71035385131836,
+      "learning_rate": 3.9887284041904085e-05,
+      "loss": 3.734,
+      "step": 112
+    },
+    {
+      "epoch": 0.3957968476357268,
+      "grad_norm": 12.229024887084961,
+      "learning_rate": 3.987845580542535e-05,
+      "loss": 3.4147,
+      "step": 113
+    },
+    {
+      "epoch": 0.39929947460595444,
+      "grad_norm": 19.44122886657715,
+      "learning_rate": 3.9869295805062906e-05,
+      "loss": 3.7344,
+      "step": 114
+    },
+    {
+      "epoch": 0.4028021015761821,
+      "grad_norm": 19.81208038330078,
+      "learning_rate": 3.9859804193693665e-05,
+      "loss": 4.4243,
+      "step": 115
+    },
+    {
+      "epoch": 0.4063047285464098,
+      "grad_norm": 12.858543395996094,
+      "learning_rate": 3.9849981129729026e-05,
+      "loss": 4.4602,
+      "step": 116
+    },
+    {
+      "epoch": 0.4098073555166375,
+      "grad_norm": 14.306174278259277,
+      "learning_rate": 3.983982677711221e-05,
+      "loss": 3.147,
+      "step": 117
+    },
+    {
+      "epoch": 0.4133099824868651,
+      "grad_norm": 9.356593132019043,
+      "learning_rate": 3.98293413053155e-05,
+      "loss": 3.8658,
+      "step": 118
+    },
+    {
+      "epoch": 0.4168126094570928,
+      "grad_norm": 9.299378395080566,
+      "learning_rate": 3.981852488933743e-05,
+      "loss": 4.0554,
+      "step": 119
+    },
+    {
+      "epoch": 0.4203152364273205,
+      "grad_norm": 9.975719451904297,
+      "learning_rate": 3.9807377709699895e-05,
+      "loss": 4.179,
+      "step": 120
+    },
+    {
+      "epoch": 0.4238178633975482,
+      "grad_norm": 13.278733253479004,
+      "learning_rate": 3.9795899952445086e-05,
+      "loss": 2.7921,
+      "step": 121
+    },
+    {
+      "epoch": 0.4273204903677758,
+      "grad_norm": 10.308444023132324,
+      "learning_rate": 3.9784091809132424e-05,
+      "loss": 3.8932,
+      "step": 122
+    },
+    {
+      "epoch": 0.4308231173380035,
+      "grad_norm": 17.627866744995117,
+      "learning_rate": 3.9771953476835323e-05,
+      "loss": 3.3324,
+      "step": 123
+    },
+    {
+      "epoch": 0.4343257443082312,
+      "grad_norm": 14.405228614807129,
+      "learning_rate": 3.975948515813795e-05,
+      "loss": 3.3283,
+      "step": 124
+    },
+    {
+      "epoch": 0.43782837127845886,
+      "grad_norm": 6.409677028656006,
+      "learning_rate": 3.9746687061131826e-05,
+      "loss": 2.5538,
+      "step": 125
+    },
+    {
+      "epoch": 0.4413309982486865,
+      "grad_norm": 21.610538482666016,
+      "learning_rate": 3.973355939941232e-05,
+      "loss": 3.2504,
+      "step": 126
+    },
+    {
+      "epoch": 0.4448336252189142,
+      "grad_norm": 11.399911880493164,
+      "learning_rate": 3.972010239207512e-05,
+      "loss": 3.8677,
+      "step": 127
+    },
+    {
+      "epoch": 0.44833625218914186,
+      "grad_norm": 11.416419982910156,
+      "learning_rate": 3.970631626371258e-05,
+      "loss": 2.8155,
+      "step": 128
+    },
+    {
+      "epoch": 0.45183887915936954,
+      "grad_norm": 11.055480003356934,
+      "learning_rate": 3.969220124440995e-05,
+      "loss": 3.0696,
+      "step": 129
+    },
+    {
+      "epoch": 0.4553415061295972,
+      "grad_norm": 24.55332374572754,
+      "learning_rate": 3.9677757569741544e-05,
+      "loss": 4.3794,
+      "step": 130
+    },
+    {
+      "epoch": 0.45884413309982486,
+      "grad_norm": 14.385886192321777,
+      "learning_rate": 3.966298548076681e-05,
+      "loss": 2.9191,
+      "step": 131
+    },
+    {
+      "epoch": 0.46234676007005254,
+      "grad_norm": 16.254535675048828,
+      "learning_rate": 3.9647885224026306e-05,
+      "loss": 3.7145,
+      "step": 132
+    },
+    {
+      "epoch": 0.4658493870402802,
+      "grad_norm": 17.73900604248047,
+      "learning_rate": 3.96324570515376e-05,
+      "loss": 3.6577,
+      "step": 133
+    },
+    {
+      "epoch": 0.46935201401050786,
+      "grad_norm": 10.10382080078125,
+      "learning_rate": 3.961670122079101e-05,
+      "loss": 3.0099,
+      "step": 134
+    },
+    {
+      "epoch": 0.47285464098073554,
+      "grad_norm": 19.65964698791504,
+      "learning_rate": 3.960061799474539e-05,
+      "loss": 3.1526,
+      "step": 135
+    },
+    {
+      "epoch": 0.4763572679509632,
+      "grad_norm": 8.950761795043945,
+      "learning_rate": 3.958420764182368e-05,
+      "loss": 3.0122,
+      "step": 136
+    },
+    {
+      "epoch": 0.4798598949211909,
+      "grad_norm": 20.611316680908203,
+      "learning_rate": 3.956747043590844e-05,
+      "loss": 4.1029,
+      "step": 137
+    },
+    {
+      "epoch": 0.48336252189141854,
+      "grad_norm": 11.411933898925781,
+      "learning_rate": 3.9550406656337286e-05,
+      "loss": 3.5931,
+      "step": 138
+    },
+    {
+      "epoch": 0.4868651488616462,
+      "grad_norm": 14.644675254821777,
+      "learning_rate": 3.953301658789822e-05,
+      "loss": 3.5152,
+      "step": 139
+    },
+    {
+      "epoch": 0.4903677758318739,
+      "grad_norm": 16.666471481323242,
+      "learning_rate": 3.9515300520824877e-05,
+      "loss": 3.8511,
+      "step": 140
+    },
+    {
+      "epoch": 0.4938704028021016,
+      "grad_norm": 8.767988204956055,
+      "learning_rate": 3.949725875079171e-05,
+      "loss": 2.9792,
+      "step": 141
+    },
+    {
+      "epoch": 0.4973730297723292,
+      "grad_norm": 7.37055778503418,
+      "learning_rate": 3.947889157890901e-05,
+      "loss": 2.7419,
+      "step": 142
+    },
+    {
+      "epoch": 0.500875656742557,
+      "grad_norm": 9.221246719360352,
+      "learning_rate": 3.9460199311717905e-05,
+      "loss": 3.4633,
+      "step": 143
+    },
+    {
+      "epoch": 0.5043782837127846,
+      "grad_norm": 8.684052467346191,
+      "learning_rate": 3.944118226118523e-05,
+      "loss": 2.51,
+      "step": 144
+    },
+    {
+      "epoch": 0.5078809106830122,
+      "grad_norm": 13.942377090454102,
+      "learning_rate": 3.942184074469835e-05,
+      "loss": 2.778,
+      "step": 145
+    },
+    {
+      "epoch": 0.51138353765324,
+      "grad_norm": 8.383426666259766,
+      "learning_rate": 3.940217508505984e-05,
+      "loss": 3.2499,
+      "step": 146
+    },
+    {
+      "epoch": 0.5148861646234676,
+      "grad_norm": 11.05321979522705,
+      "learning_rate": 3.938218561048209e-05,
+      "loss": 2.8737,
+      "step": 147
+    },
+    {
+      "epoch": 0.5183887915936952,
+      "grad_norm": 15.219361305236816,
+      "learning_rate": 3.936187265458185e-05,
+      "loss": 3.2706,
+      "step": 148
+    },
+    {
+      "epoch": 0.521891418563923,
+      "grad_norm": 12.02111530303955,
+      "learning_rate": 3.9341236556374635e-05,
+      "loss": 2.4901,
+      "step": 149
+    },
+    {
+      "epoch": 0.5253940455341506,
+      "grad_norm": 12.307247161865234,
+      "learning_rate": 3.93202776602691e-05,
+      "loss": 3.0426,
+      "step": 150
+    },
+    {
+      "epoch": 0.5288966725043783,
+      "grad_norm": 17.756227493286133,
+      "learning_rate": 3.929899631606128e-05,
+      "loss": 3.8817,
+      "step": 151
+    },
+    {
+      "epoch": 0.532399299474606,
+      "grad_norm": 15.380708694458008,
+      "learning_rate": 3.9277392878928714e-05,
+      "loss": 2.2106,
+      "step": 152
+    },
+    {
+      "epoch": 0.5359019264448336,
+      "grad_norm": 11.024133682250977,
+      "learning_rate": 3.925546770942457e-05,
+      "loss": 3.1595,
+      "step": 153
+    },
+    {
+      "epoch": 0.5394045534150613,
+      "grad_norm": 12.819958686828613,
+      "learning_rate": 3.9233221173471616e-05,
+      "loss": 2.6395,
+      "step": 154
+    },
+    {
+      "epoch": 0.542907180385289,
+      "grad_norm": 9.612881660461426,
+      "learning_rate": 3.9210653642356094e-05,
+      "loss": 3.2601,
+      "step": 155
+    },
+    {
+      "epoch": 0.5464098073555166,
+      "grad_norm": 16.474605560302734,
+      "learning_rate": 3.9187765492721534e-05,
+      "loss": 3.6263,
+      "step": 156
+    },
+    {
+      "epoch": 0.5499124343257443,
+      "grad_norm": 7.583954334259033,
+      "learning_rate": 3.916455710656246e-05,
+      "loss": 3.0502,
+      "step": 157
+    },
+    {
+      "epoch": 0.553415061295972,
+      "grad_norm": 10.567022323608398,
+      "learning_rate": 3.914102887121804e-05,
+      "loss": 2.5924,
+      "step": 158
+    },
+    {
+      "epoch": 0.5569176882661997,
+      "grad_norm": 9.386385917663574,
+      "learning_rate": 3.911718117936559e-05,
+      "loss": 2.6265,
+      "step": 159
+    },
+    {
+      "epoch": 0.5604203152364273,
+      "grad_norm": 11.7581787109375,
+      "learning_rate": 3.9093014429014034e-05,
+      "loss": 3.7777,
+      "step": 160
+    },
+    {
+      "epoch": 0.563922942206655,
+      "grad_norm": 9.84007453918457,
+      "learning_rate": 3.906852902349726e-05,
+      "loss": 3.1266,
+      "step": 161
+    },
+    {
+      "epoch": 0.5674255691768827,
+      "grad_norm": 18.16061782836914,
+      "learning_rate": 3.9043725371467406e-05,
+      "loss": 3.04,
+      "step": 162
+    },
+    {
+      "epoch": 0.5709281961471103,
+      "grad_norm": 8.399232864379883,
+      "learning_rate": 3.9018603886888e-05,
+      "loss": 3.1878,
+      "step": 163
+    },
+    {
+      "epoch": 0.574430823117338,
+      "grad_norm": 28.759723663330078,
+      "learning_rate": 3.89931649890271e-05,
+      "loss": 2.6301,
+      "step": 164
+    },
+    {
+      "epoch": 0.5779334500875657,
+      "grad_norm": 13.141151428222656,
+      "learning_rate": 3.896740910245024e-05,
+      "loss": 2.3805,
+      "step": 165
+    },
+    {
+      "epoch": 0.5814360770577933,
+      "grad_norm": 9.281489372253418,
+      "learning_rate": 3.894133665701341e-05,
+      "loss": 2.4934,
+      "step": 166
+    },
+    {
+      "epoch": 0.5849387040280211,
+      "grad_norm": 12.659279823303223,
+      "learning_rate": 3.8914948087855816e-05,
+      "loss": 2.809,
+      "step": 167
+    },
+    {
+      "epoch": 0.5884413309982487,
+      "grad_norm": 9.575944900512695,
+      "learning_rate": 3.888824383539267e-05,
+      "loss": 3.4147,
+      "step": 168
+    },
+    {
+      "epoch": 0.5919439579684763,
+      "grad_norm": 12.01468276977539,
+      "learning_rate": 3.886122434530782e-05,
+      "loss": 4.2984,
+      "step": 169
+    },
+    {
+      "epoch": 0.5954465849387041,
+      "grad_norm": 17.843990325927734,
+      "learning_rate": 3.8833890068546284e-05,
+      "loss": 3.9301,
+      "step": 170
+    },
+    {
+      "epoch": 0.5989492119089317,
+      "grad_norm": 10.337265014648438,
+      "learning_rate": 3.8806241461306774e-05,
+      "loss": 2.9173,
+      "step": 171
+    },
+    {
+      "epoch": 0.6024518388791593,
+      "grad_norm": 10.002184867858887,
+      "learning_rate": 3.877827898503406e-05,
+      "loss": 3.1344,
+      "step": 172
+    },
+    {
+      "epoch": 0.6059544658493871,
+      "grad_norm": 9.526917457580566,
+      "learning_rate": 3.875000310641125e-05,
+      "loss": 2.7306,
+      "step": 173
+    },
+    {
+      "epoch": 0.6094570928196147,
+      "grad_norm": 15.251102447509766,
+      "learning_rate": 3.8721414297352044e-05,
+      "loss": 2.6925,
+      "step": 174
+    },
+    {
+      "epoch": 0.6129597197898424,
+      "grad_norm": 15.054774284362793,
+      "learning_rate": 3.86925130349928e-05,
+      "loss": 2.9571,
+      "step": 175
+    },
+    {
+      "epoch": 0.6164623467600701,
+      "grad_norm": 6.836494445800781,
+      "learning_rate": 3.866329980168462e-05,
+      "loss": 2.8877,
+      "step": 176
+    },
+    {
+      "epoch": 0.6199649737302977,
+      "grad_norm": 66.23248291015625,
+      "learning_rate": 3.863377508498529e-05,
+      "loss": 3.5541,
+      "step": 177
+    },
+    {
+      "epoch": 0.6234676007005254,
+      "grad_norm": 8.304655075073242,
+      "learning_rate": 3.860393937765112e-05,
+      "loss": 2.7798,
+      "step": 178
+    },
+    {
+      "epoch": 0.626970227670753,
+      "grad_norm": 12.527859687805176,
+      "learning_rate": 3.8573793177628746e-05,
+      "loss": 2.7926,
+      "step": 179
+    },
+    {
+      "epoch": 0.6304728546409807,
+      "grad_norm": 15.866973876953125,
+      "learning_rate": 3.854333698804681e-05,
+      "loss": 2.7737,
+      "step": 180
+    },
+    {
+      "epoch": 0.6339754816112084,
+      "grad_norm": 9.761703491210938,
+      "learning_rate": 3.8512571317207545e-05,
+      "loss": 2.9336,
+      "step": 181
+    },
+    {
+      "epoch": 0.637478108581436,
+      "grad_norm": 19.601985931396484,
+      "learning_rate": 3.848149667857834e-05,
+      "loss": 3.7819,
+      "step": 182
+    },
+    {
+      "epoch": 0.6409807355516638,
+      "grad_norm": 8.4579439163208,
+      "learning_rate": 3.845011359078311e-05,
+      "loss": 3.8168,
+      "step": 183
+    },
+    {
+      "epoch": 0.6444833625218914,
+      "grad_norm": 9.35549545288086,
+      "learning_rate": 3.841842257759368e-05,
+      "loss": 3.1408,
+      "step": 184
+    },
+    {
+      "epoch": 0.647985989492119,
+      "grad_norm": 11.803943634033203,
+      "learning_rate": 3.8386424167921044e-05,
+      "loss": 3.3967,
+      "step": 185
+    },
+    {
+      "epoch": 0.6514886164623468,
+      "grad_norm": 10.127033233642578,
+      "learning_rate": 3.83541188958065e-05,
+      "loss": 3.4275,
+      "step": 186
+    },
+    {
+      "epoch": 0.6549912434325744,
+      "grad_norm": 15.094486236572266,
+      "learning_rate": 3.83215073004128e-05,
+      "loss": 3.5596,
+      "step": 187
+    },
+    {
+      "epoch": 0.658493870402802,
+      "grad_norm": 5.772601127624512,
+      "learning_rate": 3.8288589926015085e-05,
+      "loss": 1.9319,
+      "step": 188
+    },
+    {
+      "epoch": 0.6619964973730298,
+      "grad_norm": 6.778163433074951,
+      "learning_rate": 3.825536732199183e-05,
+      "loss": 2.1441,
+      "step": 189
+    },
+    {
+      "epoch": 0.6654991243432574,
+      "grad_norm": 28.084945678710938,
+      "learning_rate": 3.822184004281571e-05,
+      "loss": 3.33,
+      "step": 190
+    },
+    {
+      "epoch": 0.6690017513134852,
+      "grad_norm": 24.842683792114258,
+      "learning_rate": 3.818800864804428e-05,
+      "loss": 3.7081,
+      "step": 191
+    },
+    {
+      "epoch": 0.6725043782837128,
+      "grad_norm": 24.76456069946289,
+      "learning_rate": 3.815387370231068e-05,
+      "loss": 3.9048,
+      "step": 192
+    },
+    {
+      "epoch": 0.6760070052539404,
+      "grad_norm": 16.95564842224121,
+      "learning_rate": 3.8119435775314195e-05,
+      "loss": 3.8501,
+      "step": 193
+    },
+    {
+      "epoch": 0.6795096322241682,
+      "grad_norm": 14.68836784362793,
+      "learning_rate": 3.808469544181078e-05,
+      "loss": 2.9176,
+      "step": 194
+    },
+    {
+      "epoch": 0.6830122591943958,
+      "grad_norm": 12.532986640930176,
+      "learning_rate": 3.804965328160339e-05,
+      "loss": 2.9567,
+      "step": 195
+    },
+    {
+      "epoch": 0.6865148861646234,
+      "grad_norm": 12.127058982849121,
+      "learning_rate": 3.801430987953239e-05,
+      "loss": 2.9253,
+      "step": 196
+    },
+    {
+      "epoch": 0.6900175131348512,
+      "grad_norm": 16.993106842041016,
+      "learning_rate": 3.797866582546577e-05,
+      "loss": 3.7068,
+      "step": 197
+    },
+    {
+      "epoch": 0.6935201401050788,
+      "grad_norm": 8.354787826538086,
+      "learning_rate": 3.794272171428923e-05,
+      "loss": 2.4388,
+      "step": 198
+    },
+    {
+      "epoch": 0.6970227670753065,
+      "grad_norm": 20.402502059936523,
+      "learning_rate": 3.790647814589637e-05,
+      "loss": 3.7772,
+      "step": 199
+    },
+    {
+      "epoch": 0.7005253940455342,
+      "grad_norm": 12.042376518249512,
+      "learning_rate": 3.7869935725178574e-05,
+      "loss": 2.7738,
+      "step": 200
+    },
+    {
+      "epoch": 0.7040280210157618,
+      "grad_norm": 9.22277545928955,
+      "learning_rate": 3.7833095062015e-05,
+      "loss": 2.7279,
+      "step": 201
+    },
+    {
+      "epoch": 0.7075306479859895,
+      "grad_norm": 16.998899459838867,
+      "learning_rate": 3.7795956771262315e-05,
+      "loss": 3.5442,
+      "step": 202
+    },
+    {
+      "epoch": 0.7110332749562172,
+      "grad_norm": 17.048124313354492,
+      "learning_rate": 3.77585214727445e-05,
+      "loss": 3.5084,
+      "step": 203
+    },
+    {
+      "epoch": 0.7145359019264448,
+      "grad_norm": 18.509761810302734,
+      "learning_rate": 3.77207897912425e-05,
+      "loss": 3.6591,
+      "step": 204
+    },
+    {
+      "epoch": 0.7180385288966725,
+      "grad_norm": 13.900635719299316,
+      "learning_rate": 3.768276235648373e-05,
+      "loss": 3.0523,
+      "step": 205
+    },
+    {
+      "epoch": 0.7215411558669002,
+      "grad_norm": 8.05377197265625,
+      "learning_rate": 3.764443980313168e-05,
+      "loss": 2.7942,
+      "step": 206
+    },
+    {
+      "epoch": 0.7250437828371279,
+      "grad_norm": 8.520611763000488,
+      "learning_rate": 3.760582277077519e-05,
+      "loss": 3.3686,
+      "step": 207
+    },
+    {
+      "epoch": 0.7285464098073555,
+      "grad_norm": 10.90099811553955,
+      "learning_rate": 3.7566911903917896e-05,
+      "loss": 3.7414,
+      "step": 208
+    },
+    {
+      "epoch": 0.7320490367775832,
+      "grad_norm": 11.195393562316895,
+      "learning_rate": 3.752770785196739e-05,
+      "loss": 3.1357,
+      "step": 209
+    },
+    {
+      "epoch": 0.7355516637478109,
+      "grad_norm": 8.983715057373047,
+      "learning_rate": 3.748821126922442e-05,
+      "loss": 3.5049,
+      "step": 210
+    },
+    {
+      "epoch": 0.7390542907180385,
+      "grad_norm": 20.425249099731445,
+      "learning_rate": 3.744842281487198e-05,
+      "loss": 4.4525,
+      "step": 211
+    },
+    {
+      "epoch": 0.7425569176882661,
+      "grad_norm": 11.836134910583496,
+      "learning_rate": 3.740834315296426e-05,
+      "loss": 2.8092,
+      "step": 212
+    },
+    {
+      "epoch": 0.7460595446584939,
+      "grad_norm": 20.9670352935791,
+      "learning_rate": 3.73679729524156e-05,
+      "loss": 3.6011,
+      "step": 213
+    },
+    {
+      "epoch": 0.7495621716287215,
+      "grad_norm": 6.383330821990967,
+      "learning_rate": 3.7327312886989345e-05,
+      "loss": 1.8692,
+      "step": 214
+    },
+    {
+      "epoch": 0.7530647985989493,
+      "grad_norm": 10.559625625610352,
+      "learning_rate": 3.728636363528654e-05,
+      "loss": 3.7165,
+      "step": 215
+    },
+    {
+      "epoch": 0.7565674255691769,
+      "grad_norm": 20.34081268310547,
+      "learning_rate": 3.724512588073467e-05,
+      "loss": 3.1359,
+      "step": 216
+    },
+    {
+      "epoch": 0.7600700525394045,
+      "grad_norm": 12.730659484863281,
+      "learning_rate": 3.72036003115762e-05,
+      "loss": 3.0624,
+      "step": 217
+    },
+    {
+      "epoch": 0.7635726795096323,
+      "grad_norm": 15.829120635986328,
+      "learning_rate": 3.7161787620857124e-05,
+      "loss": 3.6834,
+      "step": 218
+    },
+    {
+      "epoch": 0.7670753064798599,
+      "grad_norm": 14.584449768066406,
+      "learning_rate": 3.711968850641539e-05,
+      "loss": 3.8689,
+      "step": 219
+    },
+    {
+      "epoch": 0.7705779334500875,
+      "grad_norm": 11.392684936523438,
+      "learning_rate": 3.707730367086923e-05,
+      "loss": 3.0575,
+      "step": 220
+    },
+    {
+      "epoch": 0.7740805604203153,
+      "grad_norm": 7.998726844787598,
+      "learning_rate": 3.703463382160546e-05,
+      "loss": 3.0557,
+      "step": 221
+    },
+    {
+      "epoch": 0.7775831873905429,
+      "grad_norm": 8.74062728881836,
+      "learning_rate": 3.699167967076769e-05,
+      "loss": 2.7828,
+      "step": 222
+    },
+    {
+      "epoch": 0.7810858143607706,
+      "grad_norm": 9.26952075958252,
+      "learning_rate": 3.694844193524438e-05,
+      "loss": 2.6036,
+      "step": 223
+    },
+    {
+      "epoch": 0.7845884413309983,
+      "grad_norm": 20.752344131469727,
+      "learning_rate": 3.690492133665695e-05,
+      "loss": 4.1958,
+      "step": 224
+    },
+    {
+      "epoch": 0.7880910683012259,
+      "grad_norm": 8.477578163146973,
+      "learning_rate": 3.686111860134765e-05,
+      "loss": 3.269,
+      "step": 225
+    },
+    {
+      "epoch": 0.7915936952714536,
+      "grad_norm": 12.902408599853516,
+      "learning_rate": 3.681703446036751e-05,
+      "loss": 2.917,
+      "step": 226
+    },
+    {
+      "epoch": 0.7950963222416813,
+      "grad_norm": 5.842979431152344,
+      "learning_rate": 3.677266964946414e-05,
+      "loss": 2.2078,
+      "step": 227
+    },
+    {
+      "epoch": 0.7985989492119089,
+      "grad_norm": 6.6894755363464355,
+      "learning_rate": 3.672802490906937e-05,
+      "loss": 2.2943,
+      "step": 228
+    },
+    {
+      "epoch": 0.8021015761821366,
+      "grad_norm": 12.93032455444336,
+      "learning_rate": 3.6683100984286994e-05,
+      "loss": 3.2858,
+      "step": 229
+    },
+    {
+      "epoch": 0.8056042031523643,
+      "grad_norm": 16.70753288269043,
+      "learning_rate": 3.663789862488025e-05,
+      "loss": 3.2081,
+      "step": 230
+    },
+    {
+      "epoch": 0.809106830122592,
+      "grad_norm": 9.086612701416016,
+      "learning_rate": 3.659241858525938e-05,
+      "loss": 3.0141,
+      "step": 231
+    },
+    {
+      "epoch": 0.8126094570928196,
+      "grad_norm": 13.422370910644531,
+      "learning_rate": 3.6546661624468976e-05,
+      "loss": 3.8186,
+      "step": 232
+    },
+    {
+      "epoch": 0.8161120840630472,
+      "grad_norm": 20.846773147583008,
+      "learning_rate": 3.6500628506175353e-05,
+      "loss": 4.2557,
+      "step": 233
+    },
+    {
+      "epoch": 0.819614711033275,
+      "grad_norm": 11.69239616394043,
+      "learning_rate": 3.645431999865379e-05,
+      "loss": 2.3714,
+      "step": 234
+    },
+    {
+      "epoch": 0.8231173380035026,
+      "grad_norm": 16.386404037475586,
+      "learning_rate": 3.6407736874775694e-05,
+      "loss": 4.0155,
+      "step": 235
+    },
+    {
+      "epoch": 0.8266199649737302,
+      "grad_norm": 16.324825286865234,
+      "learning_rate": 3.6360879911995735e-05,
+      "loss": 4.509,
+      "step": 236
+    },
+    {
+      "epoch": 0.830122591943958,
+      "grad_norm": 7.930265426635742,
+      "learning_rate": 3.631374989233883e-05,
+      "loss": 2.2893,
+      "step": 237
+    },
+    {
+      "epoch": 0.8336252189141856,
+      "grad_norm": 9.163013458251953,
+      "learning_rate": 3.626634760238712e-05,
+      "loss": 2.7087,
+      "step": 238
+    },
+    {
+      "epoch": 0.8371278458844134,
+      "grad_norm": 10.82229995727539,
+      "learning_rate": 3.621867383326685e-05,
+      "loss": 3.1594,
+      "step": 239
+    },
+    {
+      "epoch": 0.840630472854641,
+      "grad_norm": 13.09811019897461,
+      "learning_rate": 3.617072938063513e-05,
+      "loss": 3.5272,
+      "step": 240
+    },
+    {
+      "epoch": 0.8441330998248686,
+      "grad_norm": 11.873635292053223,
+      "learning_rate": 3.6122515044666676e-05,
+      "loss": 3.0622,
+      "step": 241
+    },
+    {
+      "epoch": 0.8476357267950964,
+      "grad_norm": 11.226723670959473,
+      "learning_rate": 3.607403163004046e-05,
+      "loss": 2.3039,
+      "step": 242
+    },
+    {
+      "epoch": 0.851138353765324,
+      "grad_norm": 8.14344596862793,
+      "learning_rate": 3.602527994592627e-05,
+      "loss": 1.8582,
+      "step": 243
+    },
+    {
+      "epoch": 0.8546409807355516,
+      "grad_norm": 15.325976371765137,
+      "learning_rate": 3.5976260805971216e-05,
+      "loss": 2.7251,
+      "step": 244
+    },
+    {
+      "epoch": 0.8581436077057794,
+      "grad_norm": 9.784257888793945,
+      "learning_rate": 3.5926975028286145e-05,
+      "loss": 2.6414,
+      "step": 245
+    },
+    {
+      "epoch": 0.861646234676007,
+      "grad_norm": 12.891518592834473,
+      "learning_rate": 3.587742343543198e-05,
+      "loss": 3.6431,
+      "step": 246
+    },
+    {
+      "epoch": 0.8651488616462347,
+      "grad_norm": 12.887101173400879,
+      "learning_rate": 3.5827606854405995e-05,
+      "loss": 2.1479,
+      "step": 247
+    },
+    {
+      "epoch": 0.8686514886164624,
+      "grad_norm": 8.942181587219238,
+      "learning_rate": 3.577752611662803e-05,
+      "loss": 2.7682,
+      "step": 248
+    },
+    {
+      "epoch": 0.87215411558669,
+      "grad_norm": 10.70305061340332,
+      "learning_rate": 3.5727182057926594e-05,
+      "loss": 1.6457,
+      "step": 249
+    },
+    {
+      "epoch": 0.8756567425569177,
+      "grad_norm": 20.12733268737793,
+      "learning_rate": 3.567657551852492e-05,
+      "loss": 3.6745,
+      "step": 250
+    },
+    {
+      "epoch": 0.8791593695271454,
+      "grad_norm": 11.07908821105957,
+      "learning_rate": 3.5625707343026943e-05,
+      "loss": 3.3239,
+      "step": 251
+    },
+    {
+      "epoch": 0.882661996497373,
+      "grad_norm": 17.630199432373047,
+      "learning_rate": 3.557457838040321e-05,
+      "loss": 3.6983,
+      "step": 252
+    },
+    {
+      "epoch": 0.8861646234676007,
+      "grad_norm": 13.158476829528809,
+      "learning_rate": 3.552318948397671e-05,
+      "loss": 2.9929,
+      "step": 253
+    },
+    {
+      "epoch": 0.8896672504378283,
+      "grad_norm": 18.34935188293457,
+      "learning_rate": 3.547154151140862e-05,
+      "loss": 2.9516,
+      "step": 254
+    },
+    {
+      "epoch": 0.8931698774080561,
+      "grad_norm": 9.750235557556152,
+      "learning_rate": 3.5419635324683996e-05,
+      "loss": 3.2622,
+      "step": 255
+    },
+    {
+      "epoch": 0.8966725043782837,
+      "grad_norm": 13.19760513305664,
+      "learning_rate": 3.5367471790097395e-05,
+      "loss": 3.4669,
+      "step": 256
+    },
+    {
+      "epoch": 0.9001751313485113,
+      "grad_norm": 13.441046714782715,
+      "learning_rate": 3.5315051778238425e-05,
+      "loss": 3.3668,
+      "step": 257
+    },
+    {
+      "epoch": 0.9036777583187391,
+      "grad_norm": 14.647550582885742,
+      "learning_rate": 3.526237616397718e-05,
+      "loss": 3.6369,
+      "step": 258
+    },
+    {
+      "epoch": 0.9071803852889667,
+      "grad_norm": 18.50444984436035,
+      "learning_rate": 3.520944582644968e-05,
+      "loss": 3.3555,
+      "step": 259
+    },
+    {
+      "epoch": 0.9106830122591943,
+      "grad_norm": 13.95717716217041,
+      "learning_rate": 3.515626164904317e-05,
+      "loss": 3.3408,
+      "step": 260
+    },
+    {
+      "epoch": 0.9141856392294221,
+      "grad_norm": 10.345568656921387,
+      "learning_rate": 3.510282451938139e-05,
+      "loss": 3.7049,
+      "step": 261
+    },
+    {
+      "epoch": 0.9176882661996497,
+      "grad_norm": 16.413713455200195,
+      "learning_rate": 3.5049135329309746e-05,
+      "loss": 3.2264,
+      "step": 262
+    },
+    {
+      "epoch": 0.9211908931698775,
+      "grad_norm": 8.679278373718262,
+      "learning_rate": 3.499519497488046e-05,
+      "loss": 4.0202,
+      "step": 263
+    },
+    {
+      "epoch": 0.9246935201401051,
+      "grad_norm": 13.41882038116455,
+      "learning_rate": 3.4941004356337566e-05,
+      "loss": 3.2065,
+      "step": 264
+    },
+    {
+      "epoch": 0.9281961471103327,
+      "grad_norm": 25.00071144104004,
+      "learning_rate": 3.488656437810193e-05,
+      "loss": 3.2281,
+      "step": 265
+    },
+    {
+      "epoch": 0.9316987740805605,
+      "grad_norm": 8.694498062133789,
+      "learning_rate": 3.4831875948756115e-05,
+      "loss": 2.7791,
+      "step": 266
+    },
+    {
+      "epoch": 0.9352014010507881,
+      "grad_norm": 8.920536041259766,
+      "learning_rate": 3.477693998102927e-05,
+      "loss": 2.8562,
+      "step": 267
+    },
+    {
+      "epoch": 0.9387040280210157,
+      "grad_norm": 13.899048805236816,
+      "learning_rate": 3.472175739178184e-05,
+      "loss": 2.759,
+      "step": 268
+    },
+    {
+      "epoch": 0.9422066549912435,
+      "grad_norm": 17.14980125427246,
+      "learning_rate": 3.4666329101990305e-05,
+      "loss": 3.4728,
+      "step": 269
+    },
+    {
+      "epoch": 0.9457092819614711,
+      "grad_norm": 12.87341022491455,
+      "learning_rate": 3.461065603673178e-05,
+      "loss": 3.6059,
+      "step": 270
+    },
+    {
+      "epoch": 0.9492119089316988,
+      "grad_norm": 17.3095703125,
+      "learning_rate": 3.45547391251686e-05,
+      "loss": 4.1194,
+      "step": 271
+    },
+    {
+      "epoch": 0.9527145359019265,
+      "grad_norm": 12.699674606323242,
+      "learning_rate": 3.4498579300532803e-05,
+      "loss": 2.7921,
+      "step": 272
+    },
+    {
+      "epoch": 0.9562171628721541,
+      "grad_norm": 11.289113998413086,
+      "learning_rate": 3.444217750011054e-05,
+      "loss": 2.6439,
+      "step": 273
+    },
+    {
+      "epoch": 0.9597197898423818,
+      "grad_norm": 10.24327564239502,
+      "learning_rate": 3.438553466522647e-05,
+      "loss": 3.1692,
+      "step": 274
+    },
+    {
+      "epoch": 0.9632224168126094,
+      "grad_norm": 7.742884159088135,
+      "learning_rate": 3.4328651741227997e-05,
+      "loss": 3.4076,
+      "step": 275
+    },
+    {
+      "epoch": 0.9667250437828371,
+      "grad_norm": 18.92707061767578,
+      "learning_rate": 3.4271529677469546e-05,
+      "loss": 3.0746,
+      "step": 276
+    },
+    {
+      "epoch": 0.9702276707530648,
+      "grad_norm": 9.217591285705566,
+      "learning_rate": 3.421416942729668e-05,
+      "loss": 2.6312,
+      "step": 277
+    },
+    {
+      "epoch": 0.9737302977232924,
+      "grad_norm": 18.405925750732422,
+      "learning_rate": 3.4156571948030206e-05,
+      "loss": 2.9252,
+      "step": 278
+    },
+    {
+      "epoch": 0.9772329246935202,
+      "grad_norm": 17.508012771606445,
+      "learning_rate": 3.40987382009502e-05,
+      "loss": 3.0746,
+      "step": 279
+    },
+    {
+      "epoch": 0.9807355516637478,
+      "grad_norm": 11.160324096679688,
+      "learning_rate": 3.4040669151279945e-05,
+      "loss": 3.445,
+      "step": 280
+    },
+    {
+      "epoch": 0.9842381786339754,
+      "grad_norm": 19.364736557006836,
+      "learning_rate": 3.3982365768169856e-05,
+      "loss": 3.5094,
+      "step": 281
+    },
+    {
+      "epoch": 0.9877408056042032,
+      "grad_norm": 13.00851821899414,
+      "learning_rate": 3.392382902468126e-05,
+      "loss": 3.5123,
+      "step": 282
+    },
+    {
+      "epoch": 0.9912434325744308,
+      "grad_norm": 7.9619550704956055,
+      "learning_rate": 3.38650598977702e-05,
+      "loss": 2.4051,
+      "step": 283
+    },
+    {
+      "epoch": 0.9947460595446584,
+      "grad_norm": 14.466473579406738,
+      "learning_rate": 3.380605936827109e-05,
+      "loss": 3.7423,
+      "step": 284
+    },
+    {
+      "epoch": 0.9982486865148862,
+      "grad_norm": 11.79201889038086,
+      "learning_rate": 3.374682842088039e-05,
+      "loss": 3.7114,
+      "step": 285
+    },
+    {
+      "epoch": 1.001751313485114,
+      "grad_norm": 8.342573165893555,
+      "learning_rate": 3.3687368044140125e-05,
+      "loss": 2.7779,
+      "step": 286
+    },
+    {
+      "epoch": 1.0052539404553416,
+      "grad_norm": 7.175319194793701,
+      "learning_rate": 3.362767923042142e-05,
+      "loss": 2.4556,
+      "step": 287
+    },
+    {
+      "epoch": 1.0087565674255692,
+      "grad_norm": 9.89644718170166,
+      "learning_rate": 3.356776297590794e-05,
+      "loss": 1.8897,
+      "step": 288
+    },
+    {
+      "epoch": 1.0122591943957968,
+      "grad_norm": 8.86550521850586,
+      "learning_rate": 3.3507620280579215e-05,
+      "loss": 1.3645,
+      "step": 289
+    },
+    {
+      "epoch": 1.0157618213660244,
+      "grad_norm": 7.4619855880737305,
+      "learning_rate": 3.3447252148194014e-05,
+      "loss": 1.0798,
+      "step": 290
+    },
+    {
+      "epoch": 1.0192644483362523,
+      "grad_norm": 14.268264770507812,
+      "learning_rate": 3.338665958627356e-05,
+      "loss": 3.0727,
+      "step": 291
+    },
+    {
+      "epoch": 1.02276707530648,
+      "grad_norm": 9.76750373840332,
+      "learning_rate": 3.332584360608471e-05,
+      "loss": 1.7199,
+      "step": 292
+    },
+    {
+      "epoch": 1.0262697022767076,
+      "grad_norm": 18.44536018371582,
+      "learning_rate": 3.32648052226231e-05,
+      "loss": 2.139,
+      "step": 293
+    },
+    {
+      "epoch": 1.0297723292469352,
+      "grad_norm": 11.522726058959961,
+      "learning_rate": 3.320354545459619e-05,
+      "loss": 0.8731,
+      "step": 294
+    },
+    {
+      "epoch": 1.0332749562171628,
+      "grad_norm": 10.644336700439453,
+      "learning_rate": 3.314206532440625e-05,
+      "loss": 0.7796,
+      "step": 295
+    },
+    {
+      "epoch": 1.0367775831873904,
+      "grad_norm": 9.179198265075684,
+      "learning_rate": 3.3080365858133335e-05,
+      "loss": 0.4986,
+      "step": 296
+    },
+    {
+      "epoch": 1.0402802101576183,
+      "grad_norm": 14.007453918457031,
+      "learning_rate": 3.301844808551811e-05,
+      "loss": 0.7696,
+      "step": 297
+    },
+    {
+      "epoch": 1.043782837127846,
+      "grad_norm": 15.713719367980957,
+      "learning_rate": 3.295631303994471e-05,
+      "loss": 0.8191,
+      "step": 298
+    },
+    {
+      "epoch": 1.0472854640980735,
+      "grad_norm": 24.446319580078125,
+      "learning_rate": 3.289396175842346e-05,
+      "loss": 1.3376,
+      "step": 299
+    },
+    {
+      "epoch": 1.0507880910683012,
+      "grad_norm": 18.64075469970703,
+      "learning_rate": 3.2831395281573605e-05,
+      "loss": 0.9942,
+      "step": 300
+    },
+    {
+      "epoch": 1.0542907180385288,
+      "grad_norm": 27.9279842376709,
+      "learning_rate": 3.276861465360587e-05,
+      "loss": 2.8875,
+      "step": 301
+    },
+    {
+      "epoch": 1.0577933450087567,
+      "grad_norm": 13.904041290283203,
+      "learning_rate": 3.270562092230514e-05,
+      "loss": 1.2454,
+      "step": 302
+    },
+    {
+      "epoch": 1.0612959719789843,
+      "grad_norm": 15.944894790649414,
+      "learning_rate": 3.264241513901287e-05,
+      "loss": 0.6807,
+      "step": 303
+    },
+    {
+      "epoch": 1.064798598949212,
+      "grad_norm": 31.337432861328125,
+      "learning_rate": 3.2578998358609595e-05,
+      "loss": 1.8232,
+      "step": 304
+    },
+    {
+      "epoch": 1.0683012259194395,
+      "grad_norm": 14.853253364562988,
+      "learning_rate": 3.251537163949732e-05,
+      "loss": 2.0792,
+      "step": 305
+    },
+    {
+      "epoch": 1.0718038528896672,
+      "grad_norm": 11.485796928405762,
+      "learning_rate": 3.245153604358184e-05,
+      "loss": 1.0212,
+      "step": 306
+    },
+    {
+      "epoch": 1.0753064798598948,
+      "grad_norm": 10.490492820739746,
+      "learning_rate": 3.238749263625503e-05,
+      "loss": 0.937,
+      "step": 307
+    },
+    {
+      "epoch": 1.0788091068301227,
+      "grad_norm": 10.13619613647461,
+      "learning_rate": 3.2323242486377056e-05,
+      "loss": 1.597,
+      "step": 308
+    },
+    {
+      "epoch": 1.0823117338003503,
+      "grad_norm": 17.22042465209961,
+      "learning_rate": 3.2258786666258524e-05,
+      "loss": 1.7487,
+      "step": 309
+    },
+    {
+      "epoch": 1.085814360770578,
+      "grad_norm": 6.475918769836426,
+      "learning_rate": 3.219412625164262e-05,
+      "loss": 0.645,
+      "step": 310
+    },
+    {
+      "epoch": 1.0893169877408055,
+      "grad_norm": 12.947412490844727,
+      "learning_rate": 3.212926232168712e-05,
+      "loss": 2.1994,
+      "step": 311
+    },
+    {
+      "epoch": 1.0928196147110332,
+      "grad_norm": 9.059617042541504,
+      "learning_rate": 3.2064195958946406e-05,
+      "loss": 1.1179,
+      "step": 312
+    },
+    {
+      "epoch": 1.096322241681261,
+      "grad_norm": 8.391510009765625,
+      "learning_rate": 3.1998928249353394e-05,
+      "loss": 1.2839,
+      "step": 313
+    },
+    {
+      "epoch": 1.0998248686514887,
+      "grad_norm": 11.310205459594727,
+      "learning_rate": 3.193346028220136e-05,
+      "loss": 1.6124,
+      "step": 314
+    },
+    {
+      "epoch": 1.1033274956217163,
+      "grad_norm": 10.316460609436035,
+      "learning_rate": 3.1867793150125844e-05,
+      "loss": 0.9333,
+      "step": 315
+    },
+    {
+      "epoch": 1.106830122591944,
+      "grad_norm": 13.748690605163574,
+      "learning_rate": 3.180192794908636e-05,
+      "loss": 1.2929,
+      "step": 316
+    },
+    {
+      "epoch": 1.1103327495621715,
+      "grad_norm": 6.411694526672363,
+      "learning_rate": 3.173586577834812e-05,
+      "loss": 0.5326,
+      "step": 317
+    },
+    {
+      "epoch": 1.1138353765323994,
+      "grad_norm": 16.91478729248047,
+      "learning_rate": 3.166960774046369e-05,
+      "loss": 1.9846,
+      "step": 318
+    },
+    {
+      "epoch": 1.117338003502627,
+      "grad_norm": 10.801108360290527,
+      "learning_rate": 3.160315494125457e-05,
+      "loss": 1.1411,
+      "step": 319
+    },
+    {
+      "epoch": 1.1208406304728546,
+      "grad_norm": 28.392133712768555,
+      "learning_rate": 3.153650848979276e-05,
+      "loss": 2.0427,
+      "step": 320
+    },
+    {
+      "epoch": 1.1243432574430823,
+      "grad_norm": 13.292547225952148,
+      "learning_rate": 3.146966949838224e-05,
+      "loss": 1.7108,
+      "step": 321
+    },
+    {
+      "epoch": 1.12784588441331,
+      "grad_norm": 18.15999412536621,
+      "learning_rate": 3.140263908254042e-05,
+      "loss": 1.8602,
+      "step": 322
+    },
+    {
+      "epoch": 1.1313485113835378,
+      "grad_norm": 13.406962394714355,
+      "learning_rate": 3.13354183609795e-05,
+      "loss": 0.7865,
+      "step": 323
+    },
+    {
+      "epoch": 1.1348511383537654,
+      "grad_norm": 10.282164573669434,
+      "learning_rate": 3.126800845558782e-05,
+      "loss": 1.0902,
+      "step": 324
+    },
+    {
+      "epoch": 1.138353765323993,
+      "grad_norm": 12.66031551361084,
+      "learning_rate": 3.1200410491411105e-05,
+      "loss": 1.9786,
+      "step": 325
+    },
+    {
+      "epoch": 1.1418563922942206,
+      "grad_norm": 13.621040344238281,
+      "learning_rate": 3.1132625596633734e-05,
+      "loss": 1.8057,
+      "step": 326
+    },
+    {
+      "epoch": 1.1453590192644483,
+      "grad_norm": 13.055573463439941,
+      "learning_rate": 3.1064654902559875e-05,
+      "loss": 1.4152,
+      "step": 327
+    },
+    {
+      "epoch": 1.1488616462346761,
+      "grad_norm": 9.842079162597656,
+      "learning_rate": 3.099649954359462e-05,
+      "loss": 1.4,
+      "step": 328
+    },
+    {
+      "epoch": 1.1523642732049038,
+      "grad_norm": 5.827723026275635,
+      "learning_rate": 3.0928160657225044e-05,
+      "loss": 0.4828,
+      "step": 329
+    },
+    {
+      "epoch": 1.1558669001751314,
+      "grad_norm": 14.849542617797852,
+      "learning_rate": 3.085963938400122e-05,
+      "loss": 2.3794,
+      "step": 330
+    },
+    {
+      "epoch": 1.159369527145359,
+      "grad_norm": 11.149800300598145,
+      "learning_rate": 3.079093686751721e-05,
+      "loss": 1.6826,
+      "step": 331
+    },
+    {
+      "epoch": 1.1628721541155866,
+      "grad_norm": 12.088582992553711,
+      "learning_rate": 3.072205425439193e-05,
+      "loss": 1.2977,
+      "step": 332
+    },
+    {
+      "epoch": 1.1663747810858143,
+      "grad_norm": 9.657583236694336,
+      "learning_rate": 3.0652992694250055e-05,
+      "loss": 1.1603,
+      "step": 333
+    },
+    {
+      "epoch": 1.1698774080560421,
+      "grad_norm": 14.046825408935547,
+      "learning_rate": 3.0583753339702816e-05,
+      "loss": 0.9929,
+      "step": 334
+    },
+    {
+      "epoch": 1.1733800350262698,
+      "grad_norm": 8.845903396606445,
+      "learning_rate": 3.0514337346328768e-05,
+      "loss": 0.764,
+      "step": 335
+    },
+    {
+      "epoch": 1.1768826619964974,
+      "grad_norm": 8.96368408203125,
+      "learning_rate": 3.04447458726545e-05,
+      "loss": 0.8944,
+      "step": 336
+    },
+    {
+      "epoch": 1.180385288966725,
+      "grad_norm": 12.853364944458008,
+      "learning_rate": 3.0374980080135292e-05,
+      "loss": 1.6053,
+      "step": 337
+    },
+    {
+      "epoch": 1.1838879159369526,
+      "grad_norm": 11.627004623413086,
+      "learning_rate": 3.0305041133135745e-05,
+      "loss": 1.8025,
+      "step": 338
+    },
+    {
+      "epoch": 1.1873905429071803,
+      "grad_norm": 13.911933898925781,
+      "learning_rate": 3.0234930198910346e-05,
+      "loss": 1.7916,
+      "step": 339
+    },
+    {
+      "epoch": 1.1908931698774081,
+      "grad_norm": 19.774444580078125,
+      "learning_rate": 3.0164648447583977e-05,
+      "loss": 0.8744,
+      "step": 340
+    },
+    {
+      "epoch": 1.1943957968476357,
+      "grad_norm": 14.323461532592773,
+      "learning_rate": 3.0094197052132414e-05,
+      "loss": 1.6673,
+      "step": 341
+    },
+    {
+      "epoch": 1.1978984238178634,
+      "grad_norm": 13.647873878479004,
+      "learning_rate": 3.002357718836268e-05,
+      "loss": 1.2698,
+      "step": 342
+    },
+    {
+      "epoch": 1.201401050788091,
+      "grad_norm": 17.837255477905273,
+      "learning_rate": 2.9952790034893534e-05,
+      "loss": 2.0078,
+      "step": 343
+    },
+    {
+      "epoch": 1.2049036777583186,
+      "grad_norm": 13.053879737854004,
+      "learning_rate": 2.988183677313568e-05,
+      "loss": 1.9082,
+      "step": 344
+    },
+    {
+      "epoch": 1.2084063047285465,
+      "grad_norm": 10.994709014892578,
+      "learning_rate": 2.981071858727215e-05,
+      "loss": 1.9751,
+      "step": 345
+    },
+    {
+      "epoch": 1.2119089316987741,
+      "grad_norm": 10.098788261413574,
+      "learning_rate": 2.9739436664238464e-05,
+      "loss": 0.869,
+      "step": 346
+    },
+    {
+      "epoch": 1.2154115586690017,
+      "grad_norm": 14.354055404663086,
+      "learning_rate": 2.9667992193702865e-05,
+      "loss": 3.0441,
+      "step": 347
+    },
+    {
+      "epoch": 1.2189141856392294,
+      "grad_norm": 9.282268524169922,
+      "learning_rate": 2.9596386368046466e-05,
+      "loss": 0.9235,
+      "step": 348
+    },
+    {
+      "epoch": 1.222416812609457,
+      "grad_norm": 15.490865707397461,
+      "learning_rate": 2.9524620382343313e-05,
+      "loss": 1.7872,
+      "step": 349
+    },
+    {
+      "epoch": 1.2259194395796849,
+      "grad_norm": 11.079964637756348,
+      "learning_rate": 2.945269543434048e-05,
+      "loss": 1.258,
+      "step": 350
+    },
+    {
+      "epoch": 1.2294220665499125,
+      "grad_norm": 11.71127986907959,
+      "learning_rate": 2.9380612724438048e-05,
+      "loss": 1.9081,
+      "step": 351
+    },
+    {
+      "epoch": 1.2329246935201401,
+      "grad_norm": 11.974512100219727,
+      "learning_rate": 2.93083734556691e-05,
+      "loss": 1.5215,
+      "step": 352
+    },
+    {
+      "epoch": 1.2364273204903677,
+      "grad_norm": 8.614048957824707,
+      "learning_rate": 2.9235978833679604e-05,
+      "loss": 0.9446,
+      "step": 353
+    },
+    {
+      "epoch": 1.2399299474605954,
+      "grad_norm": 9.980677604675293,
+      "learning_rate": 2.916343006670834e-05,
+      "loss": 1.3451,
+      "step": 354
+    },
+    {
+      "epoch": 1.2434325744308232,
+      "grad_norm": 6.885375499725342,
+      "learning_rate": 2.9090728365566687e-05,
+      "loss": 0.3762,
+      "step": 355
+    },
+    {
+      "epoch": 1.2469352014010509,
+      "grad_norm": 9.016031265258789,
+      "learning_rate": 2.9017874943618465e-05,
+      "loss": 1.0411,
+      "step": 356
+    },
+    {
+      "epoch": 1.2504378283712785,
+      "grad_norm": 17.361536026000977,
+      "learning_rate": 2.8944871016759616e-05,
+      "loss": 2.4325,
+      "step": 357
+    },
+    {
+      "epoch": 1.253940455341506,
+      "grad_norm": 11.617949485778809,
+      "learning_rate": 2.887171780339799e-05,
+      "loss": 0.9287,
+      "step": 358
+    },
+    {
+      "epoch": 1.2574430823117337,
+      "grad_norm": 14.221748352050781,
+      "learning_rate": 2.879841652443295e-05,
+      "loss": 0.9818,
+      "step": 359
+    },
+    {
+      "epoch": 1.2609457092819616,
+      "grad_norm": 11.184192657470703,
+      "learning_rate": 2.8724968403235038e-05,
+      "loss": 1.5024,
+      "step": 360
+    },
+    {
+      "epoch": 1.2644483362521892,
+      "grad_norm": 17.768596649169922,
+      "learning_rate": 2.8651374665625507e-05,
+      "loss": 1.4243,
+      "step": 361
+    },
+    {
+      "epoch": 1.2679509632224168,
+      "grad_norm": 17.50186538696289,
+      "learning_rate": 2.857763653985592e-05,
+      "loss": 1.8902,
+      "step": 362
+    },
+    {
+      "epoch": 1.2714535901926445,
+      "grad_norm": 18.82529640197754,
+      "learning_rate": 2.8503755256587608e-05,
+      "loss": 1.1333,
+      "step": 363
+    },
+    {
+      "epoch": 1.274956217162872,
+      "grad_norm": 8.98492431640625,
+      "learning_rate": 2.8429732048871152e-05,
+      "loss": 0.6526,
+      "step": 364
+    },
+    {
+      "epoch": 1.2784588441331,
+      "grad_norm": 11.492777824401855,
+      "learning_rate": 2.83555681521258e-05,
+      "loss": 0.9215,
+      "step": 365
+    },
+    {
+      "epoch": 1.2819614711033274,
+      "grad_norm": 10.343647003173828,
+      "learning_rate": 2.8281264804118848e-05,
+      "loss": 0.7322,
+      "step": 366
+    },
+    {
+      "epoch": 1.2854640980735552,
+      "grad_norm": 16.036224365234375,
+      "learning_rate": 2.8206823244944966e-05,
+      "loss": 1.6768,
+      "step": 367
+    },
+    {
+      "epoch": 1.2889667250437828,
+      "grad_norm": 28.502471923828125,
+      "learning_rate": 2.8132244717005545e-05,
+      "loss": 1.3644,
+      "step": 368
+    },
+    {
+      "epoch": 1.2924693520140105,
+      "grad_norm": 19.65816879272461,
+      "learning_rate": 2.8057530464987883e-05,
+      "loss": 1.1389,
+      "step": 369
+    },
+    {
+      "epoch": 1.295971978984238,
+      "grad_norm": 15.79315185546875,
+      "learning_rate": 2.7982681735844532e-05,
+      "loss": 1.4806,
+      "step": 370
+    },
+    {
+      "epoch": 1.2994746059544657,
+      "grad_norm": 11.223061561584473,
+      "learning_rate": 2.7907699778772345e-05,
+      "loss": 0.761,
+      "step": 371
+    },
+    {
+      "epoch": 1.3029772329246936,
+      "grad_norm": 14.808821678161621,
+      "learning_rate": 2.783258584519175e-05,
+      "loss": 1.2235,
+      "step": 372
+    },
+    {
+      "epoch": 1.3064798598949212,
+      "grad_norm": 14.618925094604492,
+      "learning_rate": 2.7757341188725786e-05,
+      "loss": 2.7571,
+      "step": 373
+    },
+    {
+      "epoch": 1.3099824868651488,
+      "grad_norm": 14.738235473632812,
+      "learning_rate": 2.7681967065179212e-05,
+      "loss": 1.443,
+      "step": 374
+    },
+    {
+      "epoch": 1.3134851138353765,
+      "grad_norm": 11.83553695678711,
+      "learning_rate": 2.7606464732517548e-05,
+      "loss": 1.0723,
+      "step": 375
+    },
+    {
+      "epoch": 1.316987740805604,
+      "grad_norm": 11.12104606628418,
+      "learning_rate": 2.753083545084608e-05,
+      "loss": 2.5116,
+      "step": 376
+    },
+    {
+      "epoch": 1.320490367775832,
+      "grad_norm": 9.57516860961914,
+      "learning_rate": 2.7455080482388817e-05,
+      "loss": 0.9263,
+      "step": 377
+    },
+    {
+      "epoch": 1.3239929947460596,
+      "grad_norm": 9.881256103515625,
+      "learning_rate": 2.7379201091467425e-05,
+      "loss": 1.1542,
+      "step": 378
+    },
+    {
+      "epoch": 1.3274956217162872,
+      "grad_norm": 12.401742935180664,
+      "learning_rate": 2.7303198544480155e-05,
+      "loss": 2.2727,
+      "step": 379
+    },
+    {
+      "epoch": 1.3309982486865148,
+      "grad_norm": 19.053503036499023,
+      "learning_rate": 2.7227074109880662e-05,
+      "loss": 1.367,
+      "step": 380
+    },
+    {
+      "epoch": 1.3345008756567425,
+      "grad_norm": 10.002474784851074,
+      "learning_rate": 2.7150829058156886e-05,
+      "loss": 0.9753,
+      "step": 381
+    },
+    {
+      "epoch": 1.3380035026269703,
+      "grad_norm": 9.08540153503418,
+      "learning_rate": 2.707446466180979e-05,
+      "loss": 1.0053,
+      "step": 382
+    },
+    {
+      "epoch": 1.341506129597198,
+      "grad_norm": 9.479249000549316,
+      "learning_rate": 2.699798219533218e-05,
+      "loss": 1.0939,
+      "step": 383
+    },
+    {
+      "epoch": 1.3450087565674256,
+      "grad_norm": 7.634374618530273,
+      "learning_rate": 2.6921382935187393e-05,
+      "loss": 0.7234,
+      "step": 384
+    },
+    {
+      "epoch": 1.3485113835376532,
+      "grad_norm": 10.296411514282227,
+      "learning_rate": 2.6844668159788015e-05,
+      "loss": 1.1631,
+      "step": 385
+    },
+    {
+      "epoch": 1.3520140105078808,
+      "grad_norm": 11.623985290527344,
+      "learning_rate": 2.6767839149474533e-05,
+      "loss": 1.2477,
+      "step": 386
+    },
+    {
+      "epoch": 1.3555166374781087,
+      "grad_norm": 11.007031440734863,
+      "learning_rate": 2.6690897186493972e-05,
+      "loss": 1.1851,
+      "step": 387
+    },
+    {
+      "epoch": 1.3590192644483363,
+      "grad_norm": 12.727971076965332,
+      "learning_rate": 2.66138435549785e-05,
+      "loss": 1.2956,
+      "step": 388
+    },
+    {
+      "epoch": 1.362521891418564,
+      "grad_norm": 11.943631172180176,
+      "learning_rate": 2.6536679540923977e-05,
+      "loss": 1.3679,
+      "step": 389
+    },
+    {
+      "epoch": 1.3660245183887916,
+      "grad_norm": 14.202005386352539,
+      "learning_rate": 2.6459406432168525e-05,
+      "loss": 1.1852,
+      "step": 390
+    },
+    {
+      "epoch": 1.3695271453590192,
+      "grad_norm": 8.993060111999512,
+      "learning_rate": 2.6382025518371e-05,
+      "loss": 0.7622,
+      "step": 391
+    },
+    {
+      "epoch": 1.373029772329247,
+      "grad_norm": 21.68409538269043,
+      "learning_rate": 2.6304538090989488e-05,
+      "loss": 1.4973,
+      "step": 392
+    },
+    {
+      "epoch": 1.3765323992994747,
+      "grad_norm": 17.034027099609375,
+      "learning_rate": 2.6226945443259742e-05,
+      "loss": 1.6121,
+      "step": 393
+    },
+    {
+      "epoch": 1.3800350262697023,
+      "grad_norm": 16.731870651245117,
+      "learning_rate": 2.6149248870173618e-05,
+      "loss": 1.2101,
+      "step": 394
+    },
+    {
+      "epoch": 1.38353765323993,
+      "grad_norm": 17.604406356811523,
+      "learning_rate": 2.6071449668457426e-05,
+      "loss": 1.6075,
+      "step": 395
+    },
+    {
+      "epoch": 1.3870402802101576,
+      "grad_norm": 8.639128684997559,
+      "learning_rate": 2.5993549136550326e-05,
+      "loss": 0.7712,
+      "step": 396
+    },
+    {
+      "epoch": 1.3905429071803854,
+      "grad_norm": 17.338247299194336,
+      "learning_rate": 2.5915548574582622e-05,
+      "loss": 1.3931,
+      "step": 397
+    },
+    {
+      "epoch": 1.3940455341506128,
+      "grad_norm": 10.645469665527344,
+      "learning_rate": 2.583744928435411e-05,
+      "loss": 0.7641,
+      "step": 398
+    },
+    {
+      "epoch": 1.3975481611208407,
+      "grad_norm": 13.74074935913086,
+      "learning_rate": 2.5759252569312282e-05,
+      "loss": 1.5749,
+      "step": 399
+    },
+    {
+      "epoch": 1.4010507880910683,
+      "grad_norm": 11.76068115234375,
+      "learning_rate": 2.5680959734530663e-05,
+      "loss": 0.7086,
+      "step": 400
+    },
+    {
+      "epoch": 1.404553415061296,
+      "grad_norm": 8.68877124786377,
+      "learning_rate": 2.5602572086686935e-05,
+      "loss": 0.9056,
+      "step": 401
+    },
+    {
+      "epoch": 1.4080560420315236,
+      "grad_norm": 10.641279220581055,
+      "learning_rate": 2.5524090934041203e-05,
+      "loss": 1.1332,
+      "step": 402
+    },
+    {
+      "epoch": 1.4115586690017512,
+      "grad_norm": 12.472097396850586,
+      "learning_rate": 2.544551758641412e-05,
+      "loss": 1.6939,
+      "step": 403
+    },
+    {
+      "epoch": 1.415061295971979,
+      "grad_norm": 6.378911972045898,
+      "learning_rate": 2.5366853355165036e-05,
+      "loss": 0.4686,
+      "step": 404
+    },
+    {
+      "epoch": 1.4185639229422067,
+      "grad_norm": 14.799468994140625,
+      "learning_rate": 2.5288099553170123e-05,
+      "loss": 1.4848,
+      "step": 405
+    },
+    {
+      "epoch": 1.4220665499124343,
+      "grad_norm": 16.260482788085938,
+      "learning_rate": 2.5209257494800454e-05,
+      "loss": 2.419,
+      "step": 406
+    },
+    {
+      "epoch": 1.425569176882662,
+      "grad_norm": 16.565237045288086,
+      "learning_rate": 2.5130328495900046e-05,
+      "loss": 0.7131,
+      "step": 407
+    },
+    {
+      "epoch": 1.4290718038528896,
+      "grad_norm": 14.373163223266602,
+      "learning_rate": 2.505131387376396e-05,
+      "loss": 1.1584,
+      "step": 408
+    },
+    {
+      "epoch": 1.4325744308231174,
+      "grad_norm": 7.218545436859131,
+      "learning_rate": 2.4972214947116235e-05,
+      "loss": 0.6338,
+      "step": 409
+    },
+    {
+      "epoch": 1.436077057793345,
+      "grad_norm": 16.41472816467285,
+      "learning_rate": 2.489303303608796e-05,
+      "loss": 0.5099,
+      "step": 410
+    },
+    {
+      "epoch": 1.4395796847635727,
+      "grad_norm": 11.42400074005127,
+      "learning_rate": 2.4813769462195164e-05,
+      "loss": 1.1297,
+      "step": 411
+    },
+    {
+      "epoch": 1.4430823117338003,
+      "grad_norm": 15.395111083984375,
+      "learning_rate": 2.473442554831682e-05,
+      "loss": 0.5138,
+      "step": 412
+    },
+    {
+      "epoch": 1.446584938704028,
+      "grad_norm": 22.0821533203125,
+      "learning_rate": 2.4655002618672742e-05,
+      "loss": 2.1099,
+      "step": 413
+    },
+    {
+      "epoch": 1.4500875656742558,
+      "grad_norm": 13.769172668457031,
+      "learning_rate": 2.4575501998801487e-05,
+      "loss": 1.1711,
+      "step": 414
+    },
+    {
+      "epoch": 1.4535901926444834,
+      "grad_norm": 11.085646629333496,
+      "learning_rate": 2.4495925015538225e-05,
+      "loss": 1.5601,
+      "step": 415
+    },
+    {
+      "epoch": 1.457092819614711,
+      "grad_norm": 15.132203102111816,
+      "learning_rate": 2.4416272996992614e-05,
+      "loss": 2.413,
+      "step": 416
+    },
+    {
+      "epoch": 1.4605954465849387,
+      "grad_norm": 9.576119422912598,
+      "learning_rate": 2.4336547272526615e-05,
+      "loss": 1.0864,
+      "step": 417
+    },
+    {
+      "epoch": 1.4640980735551663,
+      "grad_norm": 10.316338539123535,
+      "learning_rate": 2.4256749172732317e-05,
+      "loss": 1.3956,
+      "step": 418
+    },
+    {
+      "epoch": 1.4676007005253942,
+      "grad_norm": 14.369219779968262,
+      "learning_rate": 2.4176880029409717e-05,
+      "loss": 1.4616,
+      "step": 419
+    },
+    {
+      "epoch": 1.4711033274956218,
+      "grad_norm": 11.594289779663086,
+      "learning_rate": 2.4096941175544514e-05,
+      "loss": 1.387,
+      "step": 420
+    },
+    {
+      "epoch": 1.4746059544658494,
+      "grad_norm": 13.779914855957031,
+      "learning_rate": 2.4016933945285842e-05,
+      "loss": 0.8608,
+      "step": 421
+    },
+    {
+      "epoch": 1.478108581436077,
+      "grad_norm": 11.94723129272461,
+      "learning_rate": 2.3936859673924e-05,
+      "loss": 0.7525,
+      "step": 422
+    },
+    {
+      "epoch": 1.4816112084063047,
+      "grad_norm": 22.267183303833008,
+      "learning_rate": 2.3856719697868202e-05,
+      "loss": 0.5829,
+      "step": 423
+    },
+    {
+      "epoch": 1.4851138353765325,
+      "grad_norm": 11.79100227355957,
+      "learning_rate": 2.3776515354624238e-05,
+      "loss": 1.3584,
+      "step": 424
+    },
+    {
+      "epoch": 1.4886164623467601,
+      "grad_norm": 14.353073120117188,
+      "learning_rate": 2.3696247982772146e-05,
+      "loss": 1.8197,
+      "step": 425
+    },
+    {
+      "epoch": 1.4921190893169878,
+      "grad_norm": 8.39911937713623,
+      "learning_rate": 2.3615918921943916e-05,
+      "loss": 1.4571,
+      "step": 426
+    },
+    {
+      "epoch": 1.4956217162872154,
+      "grad_norm": 16.525100708007812,
+      "learning_rate": 2.3535529512801083e-05,
+      "loss": 1.1179,
+      "step": 427
+    },
+    {
+      "epoch": 1.499124343257443,
+      "grad_norm": 11.500781059265137,
+      "learning_rate": 2.345508109701238e-05,
+      "loss": 1.1656,
+      "step": 428
+    },
+    {
+      "epoch": 1.5026269702276709,
+      "grad_norm": 22.588510513305664,
+      "learning_rate": 2.337457501723134e-05,
+      "loss": 1.6281,
+      "step": 429
+    },
+    {
+      "epoch": 1.5061295971978983,
+      "grad_norm": 10.75312328338623,
+      "learning_rate": 2.3294012617073874e-05,
+      "loss": 1.3504,
+      "step": 430
+    },
+    {
+      "epoch": 1.5096322241681261,
+      "grad_norm": 12.925887107849121,
+      "learning_rate": 2.321339524109588e-05,
+      "loss": 0.9368,
+      "step": 431
+    },
+    {
+      "epoch": 1.5131348511383538,
+      "grad_norm": 10.410888671875,
+      "learning_rate": 2.313272423477076e-05,
+      "loss": 0.9173,
+      "step": 432
+    },
+    {
+      "epoch": 1.5166374781085814,
+      "grad_norm": 10.296022415161133,
+      "learning_rate": 2.3052000944467e-05,
+      "loss": 0.5623,
+      "step": 433
+    },
+    {
+      "epoch": 1.5201401050788093,
+      "grad_norm": 9.609498977661133,
+      "learning_rate": 2.2971226717425677e-05,
+      "loss": 1.3033,
+      "step": 434
+    },
+    {
+      "epoch": 1.5236427320490367,
+      "grad_norm": 10.810608863830566,
+      "learning_rate": 2.2890402901737997e-05,
+      "loss": 1.5715,
+      "step": 435
+    },
+    {
+      "epoch": 1.5271453590192645,
+      "grad_norm": 11.768206596374512,
+      "learning_rate": 2.2809530846322762e-05,
+      "loss": 0.7532,
+      "step": 436
+    },
+    {
+      "epoch": 1.5306479859894921,
+      "grad_norm": 14.947583198547363,
+      "learning_rate": 2.272861190090389e-05,
+      "loss": 1.1824,
+      "step": 437
+    },
+    {
+      "epoch": 1.5341506129597198,
+      "grad_norm": 7.3182454109191895,
+      "learning_rate": 2.2647647415987874e-05,
+      "loss": 0.8167,
+      "step": 438
+    },
+    {
+      "epoch": 1.5376532399299476,
+      "grad_norm": 30.437318801879883,
+      "learning_rate": 2.256663874284124e-05,
+      "loss": 2.3371,
+      "step": 439
+    },
+    {
+      "epoch": 1.541155866900175,
+      "grad_norm": 9.188960075378418,
+      "learning_rate": 2.2485587233468004e-05,
+      "loss": 0.7048,
+      "step": 440
+    },
+    {
+      "epoch": 1.5446584938704029,
+      "grad_norm": 13.94011402130127,
+      "learning_rate": 2.2404494240587097e-05,
+      "loss": 1.1928,
+      "step": 441
+    },
+    {
+      "epoch": 1.5481611208406305,
+      "grad_norm": 12.979096412658691,
+      "learning_rate": 2.2323361117609793e-05,
+      "loss": 0.995,
+      "step": 442
+    },
+    {
+      "epoch": 1.5516637478108581,
+      "grad_norm": 10.994502067565918,
+      "learning_rate": 2.224218921861713e-05,
+      "loss": 1.1992,
+      "step": 443
+    },
+    {
+      "epoch": 1.5551663747810858,
+      "grad_norm": 10.981045722961426,
+      "learning_rate": 2.2160979898337302e-05,
+      "loss": 0.824,
+      "step": 444
+    },
+    {
+      "epoch": 1.5586690017513134,
+      "grad_norm": 16.379398345947266,
+      "learning_rate": 2.2079734512123033e-05,
+      "loss": 1.4839,
+      "step": 445
+    },
+    {
+      "epoch": 1.5621716287215412,
+      "grad_norm": 12.315567970275879,
+      "learning_rate": 2.1998454415929e-05,
+      "loss": 1.313,
+      "step": 446
+    },
+    {
+      "epoch": 1.5656742556917689,
+      "grad_norm": 28.550540924072266,
+      "learning_rate": 2.1917140966289155e-05,
+      "loss": 1.549,
+      "step": 447
+    },
+    {
+      "epoch": 1.5691768826619965,
+      "grad_norm": 22.290969848632812,
+      "learning_rate": 2.183579552029412e-05,
+      "loss": 1.1429,
+      "step": 448
+    },
+    {
+      "epoch": 1.5726795096322241,
+      "grad_norm": 9.34792709350586,
+      "learning_rate": 2.175441943556852e-05,
+      "loss": 0.661,
+      "step": 449
+    },
+    {
+      "epoch": 1.5761821366024518,
+      "grad_norm": 7.548702239990234,
+      "learning_rate": 2.167301407024832e-05,
+      "loss": 0.4753,
+      "step": 450
+    },
+    {
+      "epoch": 1.5796847635726796,
+      "grad_norm": 18.04340362548828,
+      "learning_rate": 2.159158078295818e-05,
+      "loss": 1.06,
+      "step": 451
+    },
+    {
+      "epoch": 1.583187390542907,
+      "grad_norm": 11.318816184997559,
+      "learning_rate": 2.1510120932788766e-05,
+      "loss": 0.5676,
+      "step": 452
+    },
+    {
+      "epoch": 1.5866900175131349,
+      "grad_norm": 10.302929878234863,
+      "learning_rate": 2.1428635879274056e-05,
+      "loss": 0.8679,
+      "step": 453
+    },
+    {
+      "epoch": 1.5901926444833625,
+      "grad_norm": 15.172941207885742,
+      "learning_rate": 2.134712698236868e-05,
+      "loss": 2.0982,
+      "step": 454
+    },
+    {
+      "epoch": 1.5936952714535901,
+      "grad_norm": 9.808448791503906,
+      "learning_rate": 2.1265595602425182e-05,
+      "loss": 0.4706,
+      "step": 455
+    },
+    {
+      "epoch": 1.597197898423818,
+      "grad_norm": 25.450702667236328,
+      "learning_rate": 2.1184043100171367e-05,
+      "loss": 1.0416,
+      "step": 456
+    },
+    {
+      "epoch": 1.6007005253940454,
+      "grad_norm": 16.862489700317383,
+      "learning_rate": 2.1102470836687532e-05,
+      "loss": 1.4711,
+      "step": 457
+    },
+    {
+      "epoch": 1.6042031523642732,
+      "grad_norm": 14.062199592590332,
+      "learning_rate": 2.1020880173383822e-05,
+      "loss": 1.8195,
+      "step": 458
+    },
+    {
+      "epoch": 1.6077057793345009,
+      "grad_norm": 17.54671859741211,
+      "learning_rate": 2.0939272471977422e-05,
+      "loss": 2.3865,
+      "step": 459
+    },
+    {
+      "epoch": 1.6112084063047285,
+      "grad_norm": 14.068134307861328,
+      "learning_rate": 2.0857649094469912e-05,
+      "loss": 1.135,
+      "step": 460
+    },
+    {
+      "epoch": 1.6147110332749564,
+      "grad_norm": 18.243608474731445,
+      "learning_rate": 2.077601140312449e-05,
+      "loss": 1.3811,
+      "step": 461
+    },
+    {
+      "epoch": 1.6182136602451838,
+      "grad_norm": 9.20577335357666,
+      "learning_rate": 2.0694360760443236e-05,
+      "loss": 0.946,
+      "step": 462
+    },
+    {
+      "epoch": 1.6217162872154116,
+      "grad_norm": 10.629399299621582,
+      "learning_rate": 2.061269852914439e-05,
+      "loss": 0.9641,
+      "step": 463
+    },
+    {
+      "epoch": 1.6252189141856392,
+      "grad_norm": 12.289473533630371,
+      "learning_rate": 2.0531026072139606e-05,
+      "loss": 1.6794,
+      "step": 464
+    },
+    {
+      "epoch": 1.6287215411558669,
+      "grad_norm": 6.732086658477783,
+      "learning_rate": 2.0449344752511197e-05,
+      "loss": 0.467,
+      "step": 465
+    },
+    {
+      "epoch": 1.6322241681260947,
+      "grad_norm": 7.626420974731445,
+      "learning_rate": 2.036765593348939e-05,
+      "loss": 0.821,
+      "step": 466
+    },
+    {
+      "epoch": 1.6357267950963221,
+      "grad_norm": 7.220170497894287,
+      "learning_rate": 2.0285960978429576e-05,
+      "loss": 0.7901,
+      "step": 467
+    },
+    {
+      "epoch": 1.63922942206655,
+      "grad_norm": 12.389723777770996,
+      "learning_rate": 2.020426125078955e-05,
+      "loss": 1.4001,
+      "step": 468
+    },
+    {
+      "epoch": 1.6427320490367776,
+      "grad_norm": 7.4357171058654785,
+      "learning_rate": 2.012255811410677e-05,
+      "loss": 0.7152,
+      "step": 469
+    },
+    {
+      "epoch": 1.6462346760070052,
+      "grad_norm": 6.436247825622559,
+      "learning_rate": 2.0040852931975565e-05,
+      "loss": 0.6231,
+      "step": 470
+    },
+    {
+      "epoch": 1.649737302977233,
+      "grad_norm": 13.278726577758789,
+      "learning_rate": 1.9959147068024435e-05,
+      "loss": 1.2873,
+      "step": 471
+    },
+    {
+      "epoch": 1.6532399299474605,
+      "grad_norm": 12.931772232055664,
+      "learning_rate": 1.9877441885893233e-05,
+      "loss": 1.2275,
+      "step": 472
+    },
+    {
+      "epoch": 1.6567425569176883,
+      "grad_norm": 8.590225219726562,
+      "learning_rate": 1.9795738749210452e-05,
+      "loss": 0.5315,
+      "step": 473
+    },
+    {
+      "epoch": 1.660245183887916,
+      "grad_norm": 16.391572952270508,
+      "learning_rate": 1.971403902157043e-05,
+      "loss": 1.7926,
+      "step": 474
+    },
+    {
+      "epoch": 1.6637478108581436,
+      "grad_norm": 9.678115844726562,
+      "learning_rate": 1.9632344066510615e-05,
+      "loss": 1.1219,
+      "step": 475
+    },
+    {
+      "epoch": 1.6672504378283712,
+      "grad_norm": 12.876395225524902,
+      "learning_rate": 1.9550655247488806e-05,
+      "loss": 1.5153,
+      "step": 476
+    },
+    {
+      "epoch": 1.6707530647985989,
+      "grad_norm": 12.242756843566895,
+      "learning_rate": 1.94689739278604e-05,
+      "loss": 1.1275,
+      "step": 477
+    },
+    {
+      "epoch": 1.6742556917688267,
+      "grad_norm": 9.798916816711426,
+      "learning_rate": 1.9387301470855616e-05,
+      "loss": 1.3813,
+      "step": 478
+    },
+    {
+      "epoch": 1.6777583187390543,
+      "grad_norm": 7.478942394256592,
+      "learning_rate": 1.9305639239556774e-05,
+      "loss": 0.5468,
+      "step": 479
+    },
+    {
+      "epoch": 1.681260945709282,
+      "grad_norm": 13.860644340515137,
+      "learning_rate": 1.922398859687552e-05,
+      "loss": 1.3181,
+      "step": 480
+    },
+    {
+      "epoch": 1.6847635726795096,
+      "grad_norm": 10.224373817443848,
+      "learning_rate": 1.914235090553009e-05,
+      "loss": 1.3759,
+      "step": 481
+    },
+    {
+      "epoch": 1.6882661996497372,
+      "grad_norm": 13.883950233459473,
+      "learning_rate": 1.906072752802258e-05,
+      "loss": 1.993,
+      "step": 482
+    },
+    {
+      "epoch": 1.691768826619965,
+      "grad_norm": 13.97096061706543,
+      "learning_rate": 1.8979119826616185e-05,
+      "loss": 1.236,
+      "step": 483
+    },
+    {
+      "epoch": 1.6952714535901925,
+      "grad_norm": 10.855804443359375,
+      "learning_rate": 1.8897529163312475e-05,
+      "loss": 0.9965,
+      "step": 484
+    },
+    {
+      "epoch": 1.6987740805604203,
+      "grad_norm": 14.875264167785645,
+      "learning_rate": 1.8815956899828643e-05,
+      "loss": 1.0194,
+      "step": 485
+    },
+    {
+      "epoch": 1.702276707530648,
+      "grad_norm": 16.03706932067871,
+      "learning_rate": 1.873440439757482e-05,
+      "loss": 1.2932,
+      "step": 486
+    },
+    {
+      "epoch": 1.7057793345008756,
+      "grad_norm": 20.21481704711914,
+      "learning_rate": 1.8652873017631325e-05,
+      "loss": 1.2415,
+      "step": 487
+    },
+    {
+      "epoch": 1.7092819614711035,
+      "grad_norm": 20.23651123046875,
+      "learning_rate": 1.8571364120725947e-05,
+      "loss": 1.7298,
+      "step": 488
+    },
+    {
+      "epoch": 1.7127845884413309,
+      "grad_norm": 13.188529968261719,
+      "learning_rate": 1.8489879067211237e-05,
+      "loss": 1.1726,
+      "step": 489
+    },
+    {
+      "epoch": 1.7162872154115587,
+      "grad_norm": 9.632075309753418,
+      "learning_rate": 1.8408419217041825e-05,
+      "loss": 0.5913,
+      "step": 490
+    },
+    {
+      "epoch": 1.7197898423817863,
+      "grad_norm": 14.285216331481934,
+      "learning_rate": 1.832698592975168e-05,
+      "loss": 1.5428,
+      "step": 491
+    },
+    {
+      "epoch": 1.723292469352014,
+      "grad_norm": 19.85175132751465,
+      "learning_rate": 1.8245580564431486e-05,
+      "loss": 2.9951,
+      "step": 492
+    },
+    {
+      "epoch": 1.7267950963222418,
+      "grad_norm": 9.290555000305176,
+      "learning_rate": 1.8164204479705884e-05,
+      "loss": 0.8093,
+      "step": 493
+    },
+    {
+      "epoch": 1.7302977232924692,
+      "grad_norm": 10.422537803649902,
+      "learning_rate": 1.8082859033710855e-05,
+      "loss": 1.3501,
+      "step": 494
+    },
+    {
+      "epoch": 1.733800350262697,
+      "grad_norm": 5.478065013885498,
+      "learning_rate": 1.800154558407101e-05,
+      "loss": 0.4231,
+      "step": 495
+    },
+    {
+      "epoch": 1.7373029772329247,
+      "grad_norm": 9.295846939086914,
+      "learning_rate": 1.792026548787697e-05,
+      "loss": 1.4291,
+      "step": 496
+    },
+    {
+      "epoch": 1.7408056042031523,
+      "grad_norm": 9.88309097290039,
+      "learning_rate": 1.78390201016627e-05,
+      "loss": 0.8173,
+      "step": 497
+    },
+    {
+      "epoch": 1.7443082311733802,
+      "grad_norm": 15.843478202819824,
+      "learning_rate": 1.7757810781382875e-05,
+      "loss": 1.7793,
+      "step": 498
+    },
+    {
+      "epoch": 1.7478108581436076,
+      "grad_norm": 12.055654525756836,
+      "learning_rate": 1.767663888239021e-05,
+      "loss": 0.4423,
+      "step": 499
+    },
+    {
+      "epoch": 1.7513134851138354,
+      "grad_norm": 6.999086380004883,
+      "learning_rate": 1.759550575941291e-05,
+      "loss": 0.561,
+      "step": 500
+    },
+    {
+      "epoch": 1.754816112084063,
+      "grad_norm": 10.667252540588379,
+      "learning_rate": 1.7514412766531995e-05,
+      "loss": 1.246,
+      "step": 501
+    },
+    {
+      "epoch": 1.7583187390542907,
+      "grad_norm": 10.708955764770508,
+      "learning_rate": 1.7433361257158764e-05,
+      "loss": 0.9596,
+      "step": 502
+    },
+    {
+      "epoch": 1.7618213660245186,
+      "grad_norm": 12.158077239990234,
+      "learning_rate": 1.735235258401213e-05,
+      "loss": 0.5837,
+      "step": 503
+    },
+    {
+      "epoch": 1.765323992994746,
+      "grad_norm": 9.02990436553955,
+      "learning_rate": 1.7271388099096115e-05,
+      "loss": 1.3399,
+      "step": 504
+    },
+    {
+      "epoch": 1.7688266199649738,
+      "grad_norm": 13.258699417114258,
+      "learning_rate": 1.7190469153677248e-05,
+      "loss": 1.6198,
+      "step": 505
+    },
+    {
+      "epoch": 1.7723292469352014,
+      "grad_norm": 12.630635261535645,
+      "learning_rate": 1.710959709826201e-05,
+      "loss": 1.5737,
+      "step": 506
+    },
+    {
+      "epoch": 1.775831873905429,
+      "grad_norm": 6.134144306182861,
+      "learning_rate": 1.7028773282574326e-05,
+      "loss": 0.5122,
+      "step": 507
+    },
+    {
+      "epoch": 1.7793345008756567,
+      "grad_norm": 11.753387451171875,
+      "learning_rate": 1.6947999055533006e-05,
+      "loss": 1.0792,
+      "step": 508
+    },
+    {
+      "epoch": 1.7828371278458843,
+      "grad_norm": 4.01309061050415,
+      "learning_rate": 1.6867275765229248e-05,
+      "loss": 0.1886,
+      "step": 509
+    },
+    {
+      "epoch": 1.7863397548161122,
+      "grad_norm": 9.992500305175781,
+      "learning_rate": 1.678660475890413e-05,
+      "loss": 1.2433,
+      "step": 510
+    },
+    {
+      "epoch": 1.7898423817863398,
+      "grad_norm": 15.76321792602539,
+      "learning_rate": 1.6705987382926126e-05,
+      "loss": 1.9447,
+      "step": 511
+    },
+    {
+      "epoch": 1.7933450087565674,
+      "grad_norm": 16.34722900390625,
+      "learning_rate": 1.6625424982768663e-05,
+      "loss": 1.7307,
+      "step": 512
+    },
+    {
+      "epoch": 1.796847635726795,
+      "grad_norm": 8.70386028289795,
+      "learning_rate": 1.6544918902987623e-05,
+      "loss": 0.7112,
+      "step": 513
+    },
+    {
+      "epoch": 1.8003502626970227,
+      "grad_norm": 23.811084747314453,
+      "learning_rate": 1.646447048719892e-05,
+      "loss": 1.7522,
+      "step": 514
+    },
+    {
+      "epoch": 1.8038528896672505,
+      "grad_norm": 9.175934791564941,
+      "learning_rate": 1.6384081078056094e-05,
+      "loss": 1.4078,
+      "step": 515
+    },
+    {
+      "epoch": 1.807355516637478,
+      "grad_norm": 11.81799602508545,
+      "learning_rate": 1.6303752017227854e-05,
+      "loss": 0.7936,
+      "step": 516
+    },
+    {
+      "epoch": 1.8108581436077058,
+      "grad_norm": 10.776399612426758,
+      "learning_rate": 1.6223484645375772e-05,
+      "loss": 1.6203,
+      "step": 517
+    },
+    {
+      "epoch": 1.8143607705779334,
+      "grad_norm": 14.465993881225586,
+      "learning_rate": 1.6143280302131804e-05,
+      "loss": 2.6888,
+      "step": 518
+    },
+    {
+      "epoch": 1.817863397548161,
+      "grad_norm": 10.451493263244629,
+      "learning_rate": 1.606314032607601e-05,
+      "loss": 0.9084,
+      "step": 519
+    },
+    {
+      "epoch": 1.821366024518389,
+      "grad_norm": 11.41504192352295,
+      "learning_rate": 1.598306605471417e-05,
+      "loss": 0.6167,
+      "step": 520
+    },
+    {
+      "epoch": 1.8248686514886163,
+      "grad_norm": 7.008203983306885,
+      "learning_rate": 1.590305882445549e-05,
+      "loss": 0.4968,
+      "step": 521
+    },
+    {
+      "epoch": 1.8283712784588442,
+      "grad_norm": 15.883779525756836,
+      "learning_rate": 1.5823119970590283e-05,
+      "loss": 1.2069,
+      "step": 522
+    },
+    {
+      "epoch": 1.8318739054290718,
+      "grad_norm": 13.613566398620605,
+      "learning_rate": 1.5743250827267686e-05,
+      "loss": 1.4072,
+      "step": 523
+    },
+    {
+      "epoch": 1.8353765323992994,
+      "grad_norm": 19.81671142578125,
+      "learning_rate": 1.5663452727473388e-05,
+      "loss": 1.2059,
+      "step": 524
+    },
+    {
+      "epoch": 1.8388791593695273,
+      "grad_norm": 9.457639694213867,
+      "learning_rate": 1.558372700300739e-05,
+      "loss": 0.8344,
+      "step": 525
+    },
+    {
+      "epoch": 1.8423817863397547,
+      "grad_norm": 18.07425880432129,
+      "learning_rate": 1.5504074984461775e-05,
+      "loss": 2.2088,
+      "step": 526
+    },
+    {
+      "epoch": 1.8458844133099825,
+      "grad_norm": 22.424962997436523,
+      "learning_rate": 1.542449800119852e-05,
+      "loss": 1.6593,
+      "step": 527
+    },
+    {
+      "epoch": 1.8493870402802102,
+      "grad_norm": 16.835895538330078,
+      "learning_rate": 1.5344997381327265e-05,
+      "loss": 0.8791,
+      "step": 528
+    },
+    {
+      "epoch": 1.8528896672504378,
+      "grad_norm": 11.287313461303711,
+      "learning_rate": 1.5265574451683186e-05,
+      "loss": 0.8198,
+      "step": 529
+    },
+    {
+      "epoch": 1.8563922942206657,
+      "grad_norm": 25.34825897216797,
+      "learning_rate": 1.5186230537804846e-05,
+      "loss": 1.4374,
+      "step": 530
+    },
+    {
+      "epoch": 1.859894921190893,
+      "grad_norm": 19.87013816833496,
+      "learning_rate": 1.5106966963912047e-05,
+      "loss": 1.2982,
+      "step": 531
+    },
+    {
+      "epoch": 1.863397548161121,
+      "grad_norm": 15.213961601257324,
+      "learning_rate": 1.5027785052883766e-05,
+      "loss": 0.515,
+      "step": 532
+    },
+    {
+      "epoch": 1.8669001751313485,
+      "grad_norm": 17.84857749938965,
+      "learning_rate": 1.4948686126236046e-05,
+      "loss": 0.9899,
+      "step": 533
+    },
+    {
+      "epoch": 1.8704028021015762,
+      "grad_norm": 10.338643074035645,
+      "learning_rate": 1.4869671504099961e-05,
+      "loss": 0.9608,
+      "step": 534
+    },
+    {
+      "epoch": 1.873905429071804,
+      "grad_norm": 26.768953323364258,
+      "learning_rate": 1.4790742505199558e-05,
+      "loss": 3.1679,
+      "step": 535
+    },
+    {
+      "epoch": 1.8774080560420314,
+      "grad_norm": 8.469191551208496,
+      "learning_rate": 1.4711900446829879e-05,
+      "loss": 0.4198,
+      "step": 536
+    },
+    {
+      "epoch": 1.8809106830122593,
+      "grad_norm": 12.908656120300293,
+      "learning_rate": 1.4633146644834966e-05,
+      "loss": 1.1931,
+      "step": 537
+    },
+    {
+      "epoch": 1.884413309982487,
+      "grad_norm": 13.781142234802246,
+      "learning_rate": 1.4554482413585888e-05,
+      "loss": 1.6451,
+      "step": 538
+    },
+    {
+      "epoch": 1.8879159369527145,
+      "grad_norm": 14.904067993164062,
+      "learning_rate": 1.4475909065958802e-05,
+      "loss": 1.1193,
+      "step": 539
+    },
+    {
+      "epoch": 1.8914185639229422,
+      "grad_norm": 13.490738868713379,
+      "learning_rate": 1.4397427913313076e-05,
+      "loss": 1.4777,
+      "step": 540
+    },
+    {
+      "epoch": 1.8949211908931698,
+      "grad_norm": 18.033458709716797,
+      "learning_rate": 1.4319040265469339e-05,
+      "loss": 0.9049,
+      "step": 541
+    },
+    {
+      "epoch": 1.8984238178633976,
+      "grad_norm": 13.456507682800293,
+      "learning_rate": 1.424074743068772e-05,
+      "loss": 3.3861,
+      "step": 542
+    },
+    {
+      "epoch": 1.9019264448336253,
+      "grad_norm": 12.611906051635742,
+      "learning_rate": 1.41625507156459e-05,
+      "loss": 1.6322,
+      "step": 543
+    },
+    {
+      "epoch": 1.905429071803853,
+      "grad_norm": 8.617958068847656,
+      "learning_rate": 1.4084451425417385e-05,
+      "loss": 0.4577,
+      "step": 544
+    },
+    {
+      "epoch": 1.9089316987740805,
+      "grad_norm": 14.570892333984375,
+      "learning_rate": 1.4006450863449683e-05,
+      "loss": 2.1014,
+      "step": 545
+    },
+    {
+      "epoch": 1.9124343257443082,
+      "grad_norm": 8.288101196289062,
+      "learning_rate": 1.3928550331542576e-05,
+      "loss": 0.6519,
+      "step": 546
+    },
+    {
+      "epoch": 1.915936952714536,
+      "grad_norm": 7.784424304962158,
+      "learning_rate": 1.3850751129826385e-05,
+      "loss": 0.6649,
+      "step": 547
+    },
+    {
+      "epoch": 1.9194395796847634,
+      "grad_norm": 10.094883918762207,
+      "learning_rate": 1.3773054556740261e-05,
+      "loss": 1.3525,
+      "step": 548
+    },
+    {
+      "epoch": 1.9229422066549913,
+      "grad_norm": 7.993988990783691,
+      "learning_rate": 1.369546190901052e-05,
+      "loss": 0.5407,
+      "step": 549
+    },
+    {
+      "epoch": 1.926444833625219,
+      "grad_norm": 15.709915161132812,
+      "learning_rate": 1.3617974481629008e-05,
+      "loss": 1.8723,
+      "step": 550
+    },
+    {
+      "epoch": 1.9299474605954465,
+      "grad_norm": 9.49156379699707,
+      "learning_rate": 1.3540593567831477e-05,
+      "loss": 0.5052,
+      "step": 551
+    },
+    {
+      "epoch": 1.9334500875656744,
+      "grad_norm": 13.911781311035156,
+      "learning_rate": 1.3463320459076028e-05,
+      "loss": 1.2281,
+      "step": 552
+    },
+    {
+      "epoch": 1.9369527145359018,
+      "grad_norm": 7.616586685180664,
+      "learning_rate": 1.3386156445021507e-05,
+      "loss": 0.763,
+      "step": 553
+    },
+    {
+      "epoch": 1.9404553415061296,
+      "grad_norm": 12.074398040771484,
+      "learning_rate": 1.3309102813506035e-05,
+      "loss": 1.1219,
+      "step": 554
+    },
+    {
+      "epoch": 1.9439579684763573,
+      "grad_norm": 18.879173278808594,
+      "learning_rate": 1.3232160850525477e-05,
+      "loss": 1.2152,
+      "step": 555
+    },
+    {
+      "epoch": 1.947460595446585,
+      "grad_norm": 8.424382209777832,
+      "learning_rate": 1.3155331840211992e-05,
+      "loss": 1.1847,
+      "step": 556
+    },
+    {
+      "epoch": 1.9509632224168127,
+      "grad_norm": 9.01657485961914,
+      "learning_rate": 1.3078617064812612e-05,
+      "loss": 0.722,
+      "step": 557
+    },
+    {
+      "epoch": 1.9544658493870402,
+      "grad_norm": 7.651902198791504,
+      "learning_rate": 1.3002017804667824e-05,
+      "loss": 0.3635,
+      "step": 558
+    },
+    {
+      "epoch": 1.957968476357268,
+      "grad_norm": 10.501350402832031,
+      "learning_rate": 1.2925535338190218e-05,
+      "loss": 1.3115,
+      "step": 559
+    },
+    {
+      "epoch": 1.9614711033274956,
+      "grad_norm": 10.333868026733398,
+      "learning_rate": 1.2849170941843126e-05,
+      "loss": 1.0494,
+      "step": 560
+    },
+    {
+      "epoch": 1.9649737302977233,
+      "grad_norm": 14.399115562438965,
+      "learning_rate": 1.2772925890119339e-05,
+      "loss": 1.755,
+      "step": 561
+    },
+    {
+      "epoch": 1.9684763572679511,
+      "grad_norm": 9.45056438446045,
+      "learning_rate": 1.269680145551985e-05,
+      "loss": 1.8133,
+      "step": 562
+    },
+    {
+      "epoch": 1.9719789842381785,
+      "grad_norm": 12.52590560913086,
+      "learning_rate": 1.2620798908532578e-05,
+      "loss": 0.8922,
+      "step": 563
+    },
+    {
+      "epoch": 1.9754816112084064,
+      "grad_norm": 20.283485412597656,
+      "learning_rate": 1.2544919517611191e-05,
+      "loss": 1.8731,
+      "step": 564
+    },
+    {
+      "epoch": 1.978984238178634,
+      "grad_norm": 16.538997650146484,
+      "learning_rate": 1.2469164549153922e-05,
+      "loss": 0.9377,
+      "step": 565
+    },
+    {
+      "epoch": 1.9824868651488616,
+      "grad_norm": 11.674798965454102,
+      "learning_rate": 1.239353526748245e-05,
+      "loss": 0.907,
+      "step": 566
+    },
+    {
+      "epoch": 1.9859894921190895,
+      "grad_norm": 11.961647033691406,
+      "learning_rate": 1.2318032934820794e-05,
+      "loss": 1.5268,
+      "step": 567
+    },
+    {
+      "epoch": 1.989492119089317,
+      "grad_norm": 10.949502944946289,
+      "learning_rate": 1.2242658811274222e-05,
+      "loss": 1.4954,
+      "step": 568
+    },
+    {
+      "epoch": 1.9929947460595447,
+      "grad_norm": 6.9351935386657715,
+      "learning_rate": 1.2167414154808258e-05,
+      "loss": 0.4464,
+      "step": 569
+    },
+    {
+      "epoch": 1.9964973730297724,
+      "grad_norm": 8.37053394317627,
+      "learning_rate": 1.2092300221227661e-05,
+      "loss": 0.642,
+      "step": 570
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.300679206848145,
+      "learning_rate": 1.201731826415548e-05,
+      "loss": 1.9645,
+      "step": 571
+    },
+    {
+      "epoch": 2.003502626970228,
+      "grad_norm": 4.111511707305908,
+      "learning_rate": 1.1942469535012115e-05,
+      "loss": 0.2615,
+      "step": 572
+    },
+    {
+      "epoch": 2.0070052539404553,
+      "grad_norm": 3.611501455307007,
+      "learning_rate": 1.1867755282994464e-05,
+      "loss": 0.2108,
+      "step": 573
+    },
+    {
+      "epoch": 2.010507880910683,
+      "grad_norm": 6.634884834289551,
+      "learning_rate": 1.1793176755055037e-05,
+      "loss": 0.8019,
+      "step": 574
+    },
+    {
+      "epoch": 2.0140105078809105,
+      "grad_norm": 11.04240608215332,
+      "learning_rate": 1.1718735195881159e-05,
+      "loss": 1.0215,
+      "step": 575
+    },
+    {
+      "epoch": 2.0175131348511384,
+      "grad_norm": 2.4960787296295166,
+      "learning_rate": 1.164443184787421e-05,
+      "loss": 0.2112,
+      "step": 576
+    },
+    {
+      "epoch": 2.021015761821366,
+      "grad_norm": 6.72168493270874,
+      "learning_rate": 1.1570267951128853e-05,
+      "loss": 0.2771,
+      "step": 577
+    },
+    {
+      "epoch": 2.0245183887915936,
+      "grad_norm": 7.282315254211426,
+      "learning_rate": 1.1496244743412395e-05,
+      "loss": 0.3761,
+      "step": 578
+    },
+    {
+      "epoch": 2.0280210157618215,
+      "grad_norm": 7.026294708251953,
+      "learning_rate": 1.1422363460144087e-05,
+      "loss": 0.207,
+      "step": 579
+    },
+    {
+      "epoch": 2.031523642732049,
+      "grad_norm": 9.867234230041504,
+      "learning_rate": 1.1348625334374501e-05,
+      "loss": 0.4578,
+      "step": 580
+    },
+    {
+      "epoch": 2.0350262697022767,
+      "grad_norm": 6.709160327911377,
+      "learning_rate": 1.1275031596764975e-05,
+      "loss": 0.2914,
+      "step": 581
+    },
+    {
+      "epoch": 2.0385288966725046,
+      "grad_norm": 6.458284854888916,
+      "learning_rate": 1.1201583475567046e-05,
+      "loss": 0.6003,
+      "step": 582
+    },
+    {
+      "epoch": 2.042031523642732,
+      "grad_norm": 6.811922550201416,
+      "learning_rate": 1.1128282196602011e-05,
+      "loss": 0.4343,
+      "step": 583
+    },
+    {
+      "epoch": 2.04553415061296,
+      "grad_norm": 13.446271896362305,
+      "learning_rate": 1.105512898324039e-05,
+      "loss": 0.1694,
+      "step": 584
+    },
+    {
+      "epoch": 2.0490367775831873,
+      "grad_norm": 3.9686508178710938,
+      "learning_rate": 1.098212505638155e-05,
+      "loss": 0.2998,
+      "step": 585
+    },
+    {
+      "epoch": 2.052539404553415,
+      "grad_norm": 11.05303955078125,
+      "learning_rate": 1.0909271634433317e-05,
+      "loss": 0.268,
+      "step": 586
+    },
+    {
+      "epoch": 2.056042031523643,
+      "grad_norm": 14.222155570983887,
+      "learning_rate": 1.0836569933291663e-05,
+      "loss": 0.4609,
+      "step": 587
+    },
+    {
+      "epoch": 2.0595446584938704,
+      "grad_norm": 5.8074259757995605,
+      "learning_rate": 1.0764021166320403e-05,
+      "loss": 0.3518,
+      "step": 588
+    },
+    {
+      "epoch": 2.063047285464098,
+      "grad_norm": 7.573509216308594,
+      "learning_rate": 1.0691626544330913e-05,
+      "loss": 0.1935,
+      "step": 589
+    },
+    {
+      "epoch": 2.0665499124343256,
+      "grad_norm": 7.35963773727417,
+      "learning_rate": 1.0619387275561957e-05,
+      "loss": 0.142,
+      "step": 590
+    },
+    {
+      "epoch": 2.0700525394045535,
+      "grad_norm": 26.056726455688477,
+      "learning_rate": 1.0547304565659528e-05,
+      "loss": 0.5094,
+      "step": 591
+    },
+    {
+      "epoch": 2.073555166374781,
+      "grad_norm": 12.973404884338379,
+      "learning_rate": 1.0475379617656692e-05,
+      "loss": 0.4237,
+      "step": 592
+    },
+    {
+      "epoch": 2.0770577933450087,
+      "grad_norm": 15.458786964416504,
+      "learning_rate": 1.0403613631953544e-05,
+      "loss": 0.3591,
+      "step": 593
+    },
+    {
+      "epoch": 2.0805604203152366,
+      "grad_norm": 9.443426132202148,
+      "learning_rate": 1.0332007806297137e-05,
+      "loss": 0.3133,
+      "step": 594
+    },
+    {
+      "epoch": 2.084063047285464,
+      "grad_norm": 11.876457214355469,
+      "learning_rate": 1.0260563335761548e-05,
+      "loss": 0.191,
+      "step": 595
+    },
+    {
+      "epoch": 2.087565674255692,
+      "grad_norm": 16.497386932373047,
+      "learning_rate": 1.0189281412727863e-05,
+      "loss": 0.3268,
+      "step": 596
+    },
+    {
+      "epoch": 2.0910683012259192,
+      "grad_norm": 6.713415622711182,
+      "learning_rate": 1.0118163226864324e-05,
+      "loss": 0.1679,
+      "step": 597
+    },
+    {
+      "epoch": 2.094570928196147,
+      "grad_norm": 8.332857131958008,
+      "learning_rate": 1.0047209965106471e-05,
+      "loss": 0.3882,
+      "step": 598
+    },
+    {
+      "epoch": 2.098073555166375,
+      "grad_norm": 15.426718711853027,
+      "learning_rate": 9.976422811637323e-06,
+      "loss": 0.6588,
+      "step": 599
+    },
+    {
+      "epoch": 2.1015761821366024,
+      "grad_norm": 19.400615692138672,
+      "learning_rate": 9.9058029478676e-06,
+      "loss": 0.5528,
+      "step": 600
+    },
+    {
+      "epoch": 2.10507880910683,
+      "grad_norm": 10.129969596862793,
+      "learning_rate": 9.835351552416026e-06,
+      "loss": 0.4625,
+      "step": 601
+    },
+    {
+      "epoch": 2.1085814360770576,
+      "grad_norm": 12.683780670166016,
+      "learning_rate": 9.765069801089658e-06,
+      "loss": 0.7214,
+      "step": 602
+    },
+    {
+      "epoch": 2.1120840630472855,
+      "grad_norm": 9.236724853515625,
+      "learning_rate": 9.694958866864262e-06,
+      "loss": 0.4221,
+      "step": 603
+    },
+    {
+      "epoch": 2.1155866900175133,
+      "grad_norm": 5.072461128234863,
+      "learning_rate": 9.62501991986472e-06,
+      "loss": 0.1352,
+      "step": 604
+    },
+    {
+      "epoch": 2.1190893169877407,
+      "grad_norm": 5.266534805297852,
+      "learning_rate": 9.555254127345506e-06,
+      "loss": 0.3825,
+      "step": 605
+    },
+    {
+      "epoch": 2.1225919439579686,
+      "grad_norm": 2.0258290767669678,
+      "learning_rate": 9.485662653671237e-06,
+      "loss": 0.1042,
+      "step": 606
+    },
+    {
+      "epoch": 2.126094570928196,
+      "grad_norm": 7.214817047119141,
+      "learning_rate": 9.416246660297186e-06,
+      "loss": 0.1874,
+      "step": 607
+    },
+    {
+      "epoch": 2.129597197898424,
+      "grad_norm": 14.512842178344727,
+      "learning_rate": 9.347007305749954e-06,
+      "loss": 0.2769,
+      "step": 608
+    },
+    {
+      "epoch": 2.1330998248686517,
+      "grad_norm": 12.86308765411377,
+      "learning_rate": 9.277945745608076e-06,
+      "loss": 0.4056,
+      "step": 609
+    },
+    {
+      "epoch": 2.136602451838879,
+      "grad_norm": 8.935503959655762,
+      "learning_rate": 9.2090631324828e-06,
+      "loss": 0.267,
+      "step": 610
+    },
+    {
+      "epoch": 2.140105078809107,
+      "grad_norm": 10.452750205993652,
+      "learning_rate": 9.140360615998789e-06,
+      "loss": 0.9766,
+      "step": 611
+    },
+    {
+      "epoch": 2.1436077057793343,
+      "grad_norm": 12.40889835357666,
+      "learning_rate": 9.071839342774966e-06,
+      "loss": 0.6476,
+      "step": 612
+    },
+    {
+      "epoch": 2.147110332749562,
+      "grad_norm": 7.821516513824463,
+      "learning_rate": 9.003500456405385e-06,
+      "loss": 0.2597,
+      "step": 613
+    },
+    {
+      "epoch": 2.1506129597197896,
+      "grad_norm": 32.6134147644043,
+      "learning_rate": 8.935345097440131e-06,
+      "loss": 0.3772,
+      "step": 614
+    },
+    {
+      "epoch": 2.1541155866900175,
+      "grad_norm": 8.472585678100586,
+      "learning_rate": 8.867374403366276e-06,
+      "loss": 0.2884,
+      "step": 615
+    },
+    {
+      "epoch": 2.1576182136602453,
+      "grad_norm": 8.967446327209473,
+      "learning_rate": 8.7995895085889e-06,
+      "loss": 0.2359,
+      "step": 616
+    },
+    {
+      "epoch": 2.1611208406304727,
+      "grad_norm": 4.8656415939331055,
+      "learning_rate": 8.731991544412184e-06,
+      "loss": 0.201,
+      "step": 617
+    },
+    {
+      "epoch": 2.1646234676007006,
+      "grad_norm": 17.89019012451172,
+      "learning_rate": 8.6645816390205e-06,
+      "loss": 0.5396,
+      "step": 618
+    },
+    {
+      "epoch": 2.168126094570928,
+      "grad_norm": 4.795694351196289,
+      "learning_rate": 8.597360917459585e-06,
+      "loss": 0.3591,
+      "step": 619
+    },
+    {
+      "epoch": 2.171628721541156,
+      "grad_norm": 9.445202827453613,
+      "learning_rate": 8.53033050161776e-06,
+      "loss": 0.275,
+      "step": 620
+    },
+    {
+      "epoch": 2.1751313485113837,
+      "grad_norm": 12.556199073791504,
+      "learning_rate": 8.46349151020725e-06,
+      "loss": 0.3625,
+      "step": 621
+    },
+    {
+      "epoch": 2.178633975481611,
+      "grad_norm": 3.0647783279418945,
+      "learning_rate": 8.396845058745433e-06,
+      "loss": 0.1984,
+      "step": 622
+    },
+    {
+      "epoch": 2.182136602451839,
+      "grad_norm": 13.919934272766113,
+      "learning_rate": 8.330392259536315e-06,
+      "loss": 0.3918,
+      "step": 623
+    },
+    {
+      "epoch": 2.1856392294220663,
+      "grad_norm": 5.176346302032471,
+      "learning_rate": 8.264134221651879e-06,
+      "loss": 0.3719,
+      "step": 624
+    },
+    {
+      "epoch": 2.189141856392294,
+      "grad_norm": 6.9213762283325195,
+      "learning_rate": 8.198072050913643e-06,
+      "loss": 0.3471,
+      "step": 625
+    },
+    {
+      "epoch": 2.192644483362522,
+      "grad_norm": 2.5226728916168213,
+      "learning_rate": 8.132206849874166e-06,
+      "loss": 0.2453,
+      "step": 626
+    },
+    {
+      "epoch": 2.1961471103327495,
+      "grad_norm": 4.744520664215088,
+      "learning_rate": 8.066539717798648e-06,
+      "loss": 0.2357,
+      "step": 627
+    },
+    {
+      "epoch": 2.1996497373029773,
+      "grad_norm": 7.503551006317139,
+      "learning_rate": 8.001071750646615e-06,
+      "loss": 0.5094,
+      "step": 628
+    },
+    {
+      "epoch": 2.2031523642732047,
+      "grad_norm": 5.823261737823486,
+      "learning_rate": 7.935804041053594e-06,
+      "loss": 0.2253,
+      "step": 629
+    },
+    {
+      "epoch": 2.2066549912434326,
+      "grad_norm": 10.464018821716309,
+      "learning_rate": 7.870737678312887e-06,
+      "loss": 0.3054,
+      "step": 630
+    },
+    {
+      "epoch": 2.2101576182136604,
+      "grad_norm": 4.4330735206604,
+      "learning_rate": 7.805873748357387e-06,
+      "loss": 0.1626,
+      "step": 631
+    },
+    {
+      "epoch": 2.213660245183888,
+      "grad_norm": 4.565890312194824,
+      "learning_rate": 7.74121333374148e-06,
+      "loss": 0.2295,
+      "step": 632
+    },
+    {
+      "epoch": 2.2171628721541157,
+      "grad_norm": 8.96017074584961,
+      "learning_rate": 7.67675751362295e-06,
+      "loss": 0.6048,
+      "step": 633
+    },
+    {
+      "epoch": 2.220665499124343,
+      "grad_norm": 10.378385543823242,
+      "learning_rate": 7.612507363744972e-06,
+      "loss": 0.8332,
+      "step": 634
+    },
+    {
+      "epoch": 2.224168126094571,
+      "grad_norm": 7.120364665985107,
+      "learning_rate": 7.548463956418157e-06,
+      "loss": 0.2532,
+      "step": 635
+    },
+    {
+      "epoch": 2.227670753064799,
+      "grad_norm": 3.8531832695007324,
+      "learning_rate": 7.484628360502683e-06,
+      "loss": 0.5231,
+      "step": 636
+    },
+    {
+      "epoch": 2.231173380035026,
+      "grad_norm": 8.29166316986084,
+      "learning_rate": 7.421001641390406e-06,
+      "loss": 0.74,
+      "step": 637
+    },
+    {
+      "epoch": 2.234676007005254,
+      "grad_norm": 14.644600868225098,
+      "learning_rate": 7.357584860987139e-06,
+      "loss": 0.712,
+      "step": 638
+    },
+    {
+      "epoch": 2.2381786339754814,
+      "grad_norm": 8.683403968811035,
+      "learning_rate": 7.2943790776948645e-06,
+      "loss": 0.1712,
+      "step": 639
+    },
+    {
+      "epoch": 2.2416812609457093,
+      "grad_norm": 11.98154067993164,
+      "learning_rate": 7.231385346394133e-06,
+      "loss": 1.1342,
+      "step": 640
+    },
+    {
+      "epoch": 2.245183887915937,
+      "grad_norm": 5.7232136726379395,
+      "learning_rate": 7.168604718426408e-06,
+      "loss": 0.2113,
+      "step": 641
+    },
+    {
+      "epoch": 2.2486865148861646,
+      "grad_norm": 7.635982990264893,
+      "learning_rate": 7.106038241576541e-06,
+      "loss": 0.2895,
+      "step": 642
+    },
+    {
+      "epoch": 2.2521891418563924,
+      "grad_norm": 16.10260772705078,
+      "learning_rate": 7.043686960055292e-06,
+      "loss": 0.2543,
+      "step": 643
+    },
+    {
+      "epoch": 2.25569176882662,
+      "grad_norm": 11.05922794342041,
+      "learning_rate": 6.981551914481894e-06,
+      "loss": 0.431,
+      "step": 644
+    },
+    {
+      "epoch": 2.2591943957968477,
+      "grad_norm": 5.137984275817871,
+      "learning_rate": 6.919634141866674e-06,
+      "loss": 0.2008,
+      "step": 645
+    },
+    {
+      "epoch": 2.2626970227670755,
+      "grad_norm": 10.650192260742188,
+      "learning_rate": 6.857934675593758e-06,
+      "loss": 0.4342,
+      "step": 646
+    },
+    {
+      "epoch": 2.266199649737303,
+      "grad_norm": 19.132814407348633,
+      "learning_rate": 6.79645454540381e-06,
+      "loss": 0.4469,
+      "step": 647
+    },
+    {
+      "epoch": 2.2697022767075308,
+      "grad_norm": 7.868581771850586,
+      "learning_rate": 6.735194777376901e-06,
+      "loss": 0.4569,
+      "step": 648
+    },
+    {
+      "epoch": 2.273204903677758,
+      "grad_norm": 5.338744163513184,
+      "learning_rate": 6.6741563939152946e-06,
+      "loss": 0.5621,
+      "step": 649
+    },
+    {
+      "epoch": 2.276707530647986,
+      "grad_norm": 8.34805679321289,
+      "learning_rate": 6.61334041372645e-06,
+      "loss": 0.5147,
+      "step": 650
+    },
+    {
+      "epoch": 2.280210157618214,
+      "grad_norm": 7.190115451812744,
+      "learning_rate": 6.552747851805989e-06,
+      "loss": 0.3024,
+      "step": 651
+    },
+    {
+      "epoch": 2.2837127845884413,
+      "grad_norm": 6.776481628417969,
+      "learning_rate": 6.492379719420787e-06,
+      "loss": 0.1938,
+      "step": 652
+    },
+    {
+      "epoch": 2.287215411558669,
+      "grad_norm": 8.968748092651367,
+      "learning_rate": 6.432237024092065e-06,
+      "loss": 0.2207,
+      "step": 653
+    },
+    {
+      "epoch": 2.2907180385288965,
+      "grad_norm": 4.696019172668457,
+      "learning_rate": 6.372320769578581e-06,
+      "loss": 0.3357,
+      "step": 654
+    },
+    {
+      "epoch": 2.2942206654991244,
+      "grad_norm": 1.6001213788986206,
+      "learning_rate": 6.312631955859878e-06,
+      "loss": 0.1304,
+      "step": 655
+    },
+    {
+      "epoch": 2.2977232924693523,
+      "grad_norm": 5.230320930480957,
+      "learning_rate": 6.253171579119619e-06,
+      "loss": 0.4285,
+      "step": 656
+    },
+    {
+      "epoch": 2.3012259194395797,
+      "grad_norm": 12.429251670837402,
+      "learning_rate": 6.1939406317289165e-06,
+      "loss": 0.4969,
+      "step": 657
+    },
+    {
+      "epoch": 2.3047285464098075,
+      "grad_norm": 6.5600266456604,
+      "learning_rate": 6.134940102229811e-06,
+      "loss": 0.256,
+      "step": 658
+    },
+    {
+      "epoch": 2.308231173380035,
+      "grad_norm": 8.233277320861816,
+      "learning_rate": 6.076170975318745e-06,
+      "loss": 0.2827,
+      "step": 659
+    },
+    {
+      "epoch": 2.3117338003502628,
+      "grad_norm": 6.555598735809326,
+      "learning_rate": 6.0176342318301515e-06,
+      "loss": 0.32,
+      "step": 660
+    },
+    {
+      "epoch": 2.31523642732049,
+      "grad_norm": 8.620915412902832,
+      "learning_rate": 5.959330848720062e-06,
+      "loss": 0.1916,
+      "step": 661
+    },
+    {
+      "epoch": 2.318739054290718,
+      "grad_norm": 9.808154106140137,
+      "learning_rate": 5.901261799049807e-06,
+      "loss": 0.2856,
+      "step": 662
+    },
+    {
+      "epoch": 2.322241681260946,
+      "grad_norm": 8.030816078186035,
+      "learning_rate": 5.843428051969797e-06,
+      "loss": 0.3204,
+      "step": 663
+    },
+    {
+      "epoch": 2.3257443082311733,
+      "grad_norm": 5.74130916595459,
+      "learning_rate": 5.785830572703326e-06,
+      "loss": 0.2252,
+      "step": 664
+    },
+    {
+      "epoch": 2.329246935201401,
+      "grad_norm": 7.626296043395996,
+      "learning_rate": 5.728470322530464e-06,
+      "loss": 0.2844,
+      "step": 665
+    },
+    {
+      "epoch": 2.3327495621716285,
+      "grad_norm": 13.404093742370605,
+      "learning_rate": 5.671348258772007e-06,
+      "loss": 0.5269,
+      "step": 666
+    },
+    {
+      "epoch": 2.3362521891418564,
+      "grad_norm": 14.98967170715332,
+      "learning_rate": 5.614465334773533e-06,
+      "loss": 0.6109,
+      "step": 667
+    },
+    {
+      "epoch": 2.3397548161120842,
+      "grad_norm": 11.958736419677734,
+      "learning_rate": 5.5578224998894625e-06,
+      "loss": 0.214,
+      "step": 668
+    },
+    {
+      "epoch": 2.3432574430823117,
+      "grad_norm": 11.929584503173828,
+      "learning_rate": 5.501420699467204e-06,
+      "loss": 0.7827,
+      "step": 669
+    },
+    {
+      "epoch": 2.3467600700525395,
+      "grad_norm": 19.051668167114258,
+      "learning_rate": 5.445260874831402e-06,
+      "loss": 0.4145,
+      "step": 670
+    },
+    {
+      "epoch": 2.350262697022767,
+      "grad_norm": 8.200482368469238,
+      "learning_rate": 5.389343963268225e-06,
+      "loss": 0.1799,
+      "step": 671
+    },
+    {
+      "epoch": 2.3537653239929948,
+      "grad_norm": 9.073884963989258,
+      "learning_rate": 5.3336708980097e-06,
+      "loss": 0.1788,
+      "step": 672
+    },
+    {
+      "epoch": 2.357267950963222,
+      "grad_norm": 15.959501266479492,
+      "learning_rate": 5.2782426082181645e-06,
+      "loss": 0.659,
+      "step": 673
+    },
+    {
+      "epoch": 2.36077057793345,
+      "grad_norm": 7.414675712585449,
+      "learning_rate": 5.223060018970733e-06,
+      "loss": 0.335,
+      "step": 674
+    },
+    {
+      "epoch": 2.364273204903678,
+      "grad_norm": 9.217352867126465,
+      "learning_rate": 5.168124051243893e-06,
+      "loss": 0.4305,
+      "step": 675
+    },
+    {
+      "epoch": 2.3677758318739053,
+      "grad_norm": 12.484313011169434,
+      "learning_rate": 5.1134356218980866e-06,
+      "loss": 0.5912,
+      "step": 676
+    },
+    {
+      "epoch": 2.371278458844133,
+      "grad_norm": 9.667016983032227,
+      "learning_rate": 5.058995643662443e-06,
+      "loss": 0.4746,
+      "step": 677
+    },
+    {
+      "epoch": 2.3747810858143605,
+      "grad_norm": 15.732196807861328,
+      "learning_rate": 5.004805025119546e-06,
+      "loss": 0.4182,
+      "step": 678
+    },
+    {
+      "epoch": 2.3782837127845884,
+      "grad_norm": 5.530452728271484,
+      "learning_rate": 4.950864670690257e-06,
+      "loss": 0.3474,
+      "step": 679
+    },
+    {
+      "epoch": 2.3817863397548162,
+      "grad_norm": 7.507744312286377,
+      "learning_rate": 4.897175480618619e-06,
+      "loss": 0.2377,
+      "step": 680
+    },
+    {
+      "epoch": 2.3852889667250436,
+      "grad_norm": 0.8480662703514099,
+      "learning_rate": 4.8437383509568324e-06,
+      "loss": 0.1067,
+      "step": 681
+    },
+    {
+      "epoch": 2.3887915936952715,
+      "grad_norm": 8.764185905456543,
+      "learning_rate": 4.790554173550319e-06,
+      "loss": 0.4602,
+      "step": 682
+    },
+    {
+      "epoch": 2.392294220665499,
+      "grad_norm": 8.413671493530273,
+      "learning_rate": 4.737623836022822e-06,
+      "loss": 0.4463,
+      "step": 683
+    },
+    {
+      "epoch": 2.3957968476357268,
+      "grad_norm": 6.1764068603515625,
+      "learning_rate": 4.684948221761583e-06,
+      "loss": 0.354,
+      "step": 684
+    },
+    {
+      "epoch": 2.3992994746059546,
+      "grad_norm": 5.312464714050293,
+      "learning_rate": 4.632528209902607e-06,
+      "loss": 0.1352,
+      "step": 685
+    },
+    {
+      "epoch": 2.402802101576182,
+      "grad_norm": 10.079058647155762,
+      "learning_rate": 4.580364675316014e-06,
+      "loss": 0.4412,
+      "step": 686
+    },
+    {
+      "epoch": 2.40630472854641,
+      "grad_norm": 6.57410192489624,
+      "learning_rate": 4.528458488591385e-06,
+      "loss": 0.6779,
+      "step": 687
+    },
+    {
+      "epoch": 2.4098073555166373,
+      "grad_norm": 7.1317009925842285,
+      "learning_rate": 4.476810516023293e-06,
+      "loss": 0.1367,
+      "step": 688
+    },
+    {
+      "epoch": 2.413309982486865,
+      "grad_norm": 9.098307609558105,
+      "learning_rate": 4.425421619596786e-06,
+      "loss": 0.4837,
+      "step": 689
+    },
+    {
+      "epoch": 2.416812609457093,
+      "grad_norm": 12.327546119689941,
+      "learning_rate": 4.374292656973058e-06,
+      "loss": 0.2527,
+      "step": 690
+    },
+    {
+      "epoch": 2.4203152364273204,
+      "grad_norm": 5.387122631072998,
+      "learning_rate": 4.323424481475085e-06,
+      "loss": 0.2458,
+      "step": 691
+    },
+    {
+      "epoch": 2.4238178633975482,
+      "grad_norm": 9.385303497314453,
+      "learning_rate": 4.272817942073408e-06,
+      "loss": 0.2781,
+      "step": 692
+    },
+    {
+      "epoch": 2.4273204903677756,
+      "grad_norm": 3.463916778564453,
+      "learning_rate": 4.2224738833719695e-06,
+      "loss": 0.115,
+      "step": 693
+    },
+    {
+      "epoch": 2.4308231173380035,
+      "grad_norm": 10.816231727600098,
+      "learning_rate": 4.172393145594009e-06,
+      "loss": 0.2425,
+      "step": 694
+    },
+    {
+      "epoch": 2.4343257443082313,
+      "grad_norm": 6.746379375457764,
+      "learning_rate": 4.122576564568031e-06,
+      "loss": 0.2906,
+      "step": 695
+    },
+    {
+      "epoch": 2.4378283712784588,
+      "grad_norm": 1.8436380624771118,
+      "learning_rate": 4.073024971713859e-06,
+      "loss": 0.0874,
+      "step": 696
+    },
+    {
+      "epoch": 2.4413309982486866,
+      "grad_norm": 4.27414608001709,
+      "learning_rate": 4.023739194028784e-06,
+      "loss": 0.254,
+      "step": 697
+    },
+    {
+      "epoch": 2.444833625218914,
+      "grad_norm": 18.09396743774414,
+      "learning_rate": 3.974720054073733e-06,
+      "loss": 0.7811,
+      "step": 698
+    },
+    {
+      "epoch": 2.448336252189142,
+      "grad_norm": 13.307997703552246,
+      "learning_rate": 3.925968369959548e-06,
+      "loss": 0.5651,
+      "step": 699
+    },
+    {
+      "epoch": 2.4518388791593697,
+      "grad_norm": 11.251547813415527,
+      "learning_rate": 3.877484955333326e-06,
+      "loss": 0.7353,
+      "step": 700
+    },
+    {
+      "epoch": 2.455341506129597,
+      "grad_norm": 5.3191046714782715,
+      "learning_rate": 3.829270619364873e-06,
+      "loss": 0.2166,
+      "step": 701
+    },
+    {
+      "epoch": 2.458844133099825,
+      "grad_norm": 17.071352005004883,
+      "learning_rate": 3.781326166733148e-06,
+      "loss": 0.3852,
+      "step": 702
+    },
+    {
+      "epoch": 2.4623467600700524,
+      "grad_norm": 42.43695831298828,
+      "learning_rate": 3.7336523976128812e-06,
+      "loss": 0.8098,
+      "step": 703
+    },
+    {
+      "epoch": 2.4658493870402802,
+      "grad_norm": 15.574095726013184,
+      "learning_rate": 3.686250107661176e-06,
+      "loss": 0.6515,
+      "step": 704
+    },
+    {
+      "epoch": 2.469352014010508,
+      "grad_norm": 5.729953765869141,
+      "learning_rate": 3.6391200880042754e-06,
+      "loss": 0.4977,
+      "step": 705
+    },
+    {
+      "epoch": 2.4728546409807355,
+      "grad_norm": 7.877871513366699,
+      "learning_rate": 3.5922631252243132e-06,
+      "loss": 0.2416,
+      "step": 706
+    },
+    {
+      "epoch": 2.4763572679509633,
+      "grad_norm": 6.631842613220215,
+      "learning_rate": 3.5456800013462166e-06,
+      "loss": 0.216,
+      "step": 707
+    },
+    {
+      "epoch": 2.4798598949211907,
+      "grad_norm": 11.070491790771484,
+      "learning_rate": 3.499371493824646e-06,
+      "loss": 0.5384,
+      "step": 708
+    },
+    {
+      "epoch": 2.4833625218914186,
+      "grad_norm": 8.516698837280273,
+      "learning_rate": 3.4533383755310257e-06,
+      "loss": 0.3293,
+      "step": 709
+    },
+    {
+      "epoch": 2.4868651488616464,
+      "grad_norm": 4.230687618255615,
+      "learning_rate": 3.407581414740626e-06,
+      "loss": 0.1136,
+      "step": 710
+    },
+    {
+      "epoch": 2.490367775831874,
+      "grad_norm": 6.080500602722168,
+      "learning_rate": 3.362101375119755e-06,
+      "loss": 0.434,
+      "step": 711
+    },
+    {
+      "epoch": 2.4938704028021017,
+      "grad_norm": 6.872626781463623,
+      "learning_rate": 3.316899015713009e-06,
+      "loss": 0.1984,
+      "step": 712
+    },
+    {
+      "epoch": 2.497373029772329,
+      "grad_norm": 8.121652603149414,
+      "learning_rate": 3.2719750909306303e-06,
+      "loss": 0.2077,
+      "step": 713
+    },
+    {
+      "epoch": 2.500875656742557,
+      "grad_norm": 7.671825408935547,
+      "learning_rate": 3.2273303505358667e-06,
+      "loss": 0.1965,
+      "step": 714
+    },
+    {
+      "epoch": 2.504378283712785,
+      "grad_norm": 5.99457311630249,
+      "learning_rate": 3.182965539632492e-06,
+      "loss": 0.2283,
+      "step": 715
+    },
+    {
+      "epoch": 2.507880910683012,
+      "grad_norm": 12.298001289367676,
+      "learning_rate": 3.138881398652358e-06,
+      "loss": 0.2892,
+      "step": 716
+    },
+    {
+      "epoch": 2.51138353765324,
+      "grad_norm": 5.526932716369629,
+      "learning_rate": 3.0950786633430564e-06,
+      "loss": 0.208,
+      "step": 717
+    },
+    {
+      "epoch": 2.5148861646234675,
+      "grad_norm": 7.946258544921875,
+      "learning_rate": 3.05155806475562e-06,
+      "loss": 1.0295,
+      "step": 718
+    },
+    {
+      "epoch": 2.5183887915936953,
+      "grad_norm": 7.10674524307251,
+      "learning_rate": 3.008320329232317e-06,
+      "loss": 0.3865,
+      "step": 719
+    },
+    {
+      "epoch": 2.521891418563923,
+      "grad_norm": 5.112523555755615,
+      "learning_rate": 2.965366178394542e-06,
+      "loss": 0.1231,
+      "step": 720
+    },
+    {
+      "epoch": 2.5253940455341506,
+      "grad_norm": 7.176146030426025,
+      "learning_rate": 2.922696329130783e-06,
+      "loss": 0.4793,
+      "step": 721
+    },
+    {
+      "epoch": 2.5288966725043784,
+      "grad_norm": 3.4290506839752197,
+      "learning_rate": 2.8803114935846178e-06,
+      "loss": 0.301,
+      "step": 722
+    },
+    {
+      "epoch": 2.532399299474606,
+      "grad_norm": 6.2297210693359375,
+      "learning_rate": 2.8382123791428795e-06,
+      "loss": 0.1753,
+      "step": 723
+    },
+    {
+      "epoch": 2.5359019264448337,
+      "grad_norm": 5.336044788360596,
+      "learning_rate": 2.7963996884238034e-06,
+      "loss": 0.2035,
+      "step": 724
+    },
+    {
+      "epoch": 2.5394045534150615,
+      "grad_norm": 7.6555891036987305,
+      "learning_rate": 2.754874119265336e-06,
+      "loss": 0.359,
+      "step": 725
+    },
+    {
+      "epoch": 2.542907180385289,
+      "grad_norm": 8.17387866973877,
+      "learning_rate": 2.7136363647134655e-06,
+      "loss": 0.7163,
+      "step": 726
+    },
+    {
+      "epoch": 2.5464098073555164,
+      "grad_norm": 7.864010334014893,
+      "learning_rate": 2.6726871130106637e-06,
+      "loss": 0.3539,
+      "step": 727
+    },
+    {
+      "epoch": 2.549912434325744,
+      "grad_norm": 6.670307159423828,
+      "learning_rate": 2.6320270475844023e-06,
+      "loss": 0.19,
+      "step": 728
+    },
+    {
+      "epoch": 2.553415061295972,
+      "grad_norm": 7.956202507019043,
+      "learning_rate": 2.5916568470357486e-06,
+      "loss": 0.4138,
+      "step": 729
+    },
+    {
+      "epoch": 2.5569176882662,
+      "grad_norm": 2.7980082035064697,
+      "learning_rate": 2.5515771851280248e-06,
+      "loss": 0.1076,
+      "step": 730
+    },
+    {
+      "epoch": 2.5604203152364273,
+      "grad_norm": 9.661548614501953,
+      "learning_rate": 2.5117887307755775e-06,
+      "loss": 0.2974,
+      "step": 731
+    },
+    {
+      "epoch": 2.5639229422066547,
+      "grad_norm": 8.196139335632324,
+      "learning_rate": 2.472292148032609e-06,
+      "loss": 0.3056,
+      "step": 732
+    },
+    {
+      "epoch": 2.5674255691768826,
+      "grad_norm": 7.685372829437256,
+      "learning_rate": 2.433088096082108e-06,
+      "loss": 0.3848,
+      "step": 733
+    },
+    {
+      "epoch": 2.5709281961471104,
+      "grad_norm": 6.782233715057373,
+      "learning_rate": 2.3941772292248146e-06,
+      "loss": 0.1736,
+      "step": 734
+    },
+    {
+      "epoch": 2.574430823117338,
+      "grad_norm": 3.7146668434143066,
+      "learning_rate": 2.3555601968683273e-06,
+      "loss": 0.265,
+      "step": 735
+    },
+    {
+      "epoch": 2.5779334500875657,
+      "grad_norm": 6.972784996032715,
+      "learning_rate": 2.317237643516268e-06,
+      "loss": 0.189,
+      "step": 736
+    },
+    {
+      "epoch": 2.581436077057793,
+      "grad_norm": 3.0132017135620117,
+      "learning_rate": 2.2792102087575053e-06,
+      "loss": 0.1583,
+      "step": 737
+    },
+    {
+      "epoch": 2.584938704028021,
+      "grad_norm": 5.253518104553223,
+      "learning_rate": 2.2414785272555005e-06,
+      "loss": 0.3482,
+      "step": 738
+    },
+    {
+      "epoch": 2.588441330998249,
+      "grad_norm": 4.389585018157959,
+      "learning_rate": 2.204043228737689e-06,
+      "loss": 0.2313,
+      "step": 739
+    },
+    {
+      "epoch": 2.591943957968476,
+      "grad_norm": 11.996556282043457,
+      "learning_rate": 2.166904937985006e-06,
+      "loss": 0.4344,
+      "step": 740
+    },
+    {
+      "epoch": 2.595446584938704,
+      "grad_norm": 9.555136680603027,
+      "learning_rate": 2.1300642748214284e-06,
+      "loss": 0.2702,
+      "step": 741
+    },
+    {
+      "epoch": 2.5989492119089315,
+      "grad_norm": 10.910072326660156,
+      "learning_rate": 2.0935218541036373e-06,
+      "loss": 0.3017,
+      "step": 742
+    },
+    {
+      "epoch": 2.6024518388791593,
+      "grad_norm": 4.974386692047119,
+      "learning_rate": 2.057278285710771e-06,
+      "loss": 0.3118,
+      "step": 743
+    },
+    {
+      "epoch": 2.605954465849387,
+      "grad_norm": 7.93409538269043,
+      "learning_rate": 2.0213341745342373e-06,
+      "loss": 0.4271,
+      "step": 744
+    },
+    {
+      "epoch": 2.6094570928196146,
+      "grad_norm": 3.470421314239502,
+      "learning_rate": 1.9856901204676093e-06,
+      "loss": 0.167,
+      "step": 745
+    },
+    {
+      "epoch": 2.6129597197898424,
+      "grad_norm": 11.024426460266113,
+      "learning_rate": 1.950346718396614e-06,
+      "loss": 0.5929,
+      "step": 746
+    },
+    {
+      "epoch": 2.61646234676007,
+      "grad_norm": 4.000313758850098,
+      "learning_rate": 1.915304558189226e-06,
+      "loss": 0.4232,
+      "step": 747
+    },
+    {
+      "epoch": 2.6199649737302977,
+      "grad_norm": 4.251917839050293,
+      "learning_rate": 1.8805642246858035e-06,
+      "loss": 0.332,
+      "step": 748
+    },
+    {
+      "epoch": 2.6234676007005255,
+      "grad_norm": 14.466004371643066,
+      "learning_rate": 1.846126297689328e-06,
+      "loss": 0.5515,
+      "step": 749
+    },
+    {
+      "epoch": 2.626970227670753,
+      "grad_norm": 7.978627681732178,
+      "learning_rate": 1.8119913519557264e-06,
+      "loss": 0.2431,
+      "step": 750
+    },
+    {
+      "epoch": 2.630472854640981,
+      "grad_norm": 8.531673431396484,
+      "learning_rate": 1.7781599571842956e-06,
+      "loss": 0.4857,
+      "step": 751
+    },
+    {
+      "epoch": 2.633975481611208,
+      "grad_norm": 6.550644874572754,
+      "learning_rate": 1.7446326780081713e-06,
+      "loss": 0.5275,
+      "step": 752
+    },
+    {
+      "epoch": 2.637478108581436,
+      "grad_norm": 8.497308731079102,
+      "learning_rate": 1.7114100739849249e-06,
+      "loss": 0.3287,
+      "step": 753
+    },
+    {
+      "epoch": 2.640980735551664,
+      "grad_norm": 3.7064123153686523,
+      "learning_rate": 1.678492699587204e-06,
+      "loss": 0.1416,
+      "step": 754
+    },
+    {
+      "epoch": 2.6444833625218913,
+      "grad_norm": 1.5083993673324585,
+      "learning_rate": 1.6458811041935007e-06,
+      "loss": 0.1093,
+      "step": 755
+    },
+    {
+      "epoch": 2.647985989492119,
+      "grad_norm": 6.478839874267578,
+      "learning_rate": 1.6135758320789641e-06,
+      "loss": 0.3954,
+      "step": 756
+    },
+    {
+      "epoch": 2.6514886164623466,
+      "grad_norm": 9.450393676757812,
+      "learning_rate": 1.5815774224063218e-06,
+      "loss": 0.4096,
+      "step": 757
+    },
+    {
+      "epoch": 2.6549912434325744,
+      "grad_norm": 5.808910369873047,
+      "learning_rate": 1.5498864092168941e-06,
+      "loss": 0.2206,
+      "step": 758
+    },
+    {
+      "epoch": 2.6584938704028023,
+      "grad_norm": 3.4636878967285156,
+      "learning_rate": 1.5185033214216649e-06,
+      "loss": 0.3483,
+      "step": 759
+    },
+    {
+      "epoch": 2.6619964973730297,
+      "grad_norm": 11.797184944152832,
+      "learning_rate": 1.4874286827924599e-06,
+      "loss": 0.625,
+      "step": 760
+    },
+    {
+      "epoch": 2.6654991243432575,
+      "grad_norm": 11.192656517028809,
+      "learning_rate": 1.456663011953201e-06,
+      "loss": 0.2856,
+      "step": 761
+    },
+    {
+      "epoch": 2.669001751313485,
+      "grad_norm": 17.491886138916016,
+      "learning_rate": 1.4262068223712587e-06,
+      "loss": 0.6158,
+      "step": 762
+    },
+    {
+      "epoch": 2.672504378283713,
+      "grad_norm": 8.262967109680176,
+      "learning_rate": 1.3960606223488848e-06,
+      "loss": 0.4461,
+      "step": 763
+    },
+    {
+      "epoch": 2.6760070052539406,
+      "grad_norm": 25.136058807373047,
+      "learning_rate": 1.3662249150147155e-06,
+      "loss": 0.8565,
+      "step": 764
+    },
+    {
+      "epoch": 2.679509632224168,
+      "grad_norm": 8.498688697814941,
+      "learning_rate": 1.3367001983153837e-06,
+      "loss": 0.405,
+      "step": 765
+    },
+    {
+      "epoch": 2.683012259194396,
+      "grad_norm": 3.612231492996216,
+      "learning_rate": 1.3074869650072053e-06,
+      "loss": 0.2578,
+      "step": 766
+    },
+    {
+      "epoch": 2.6865148861646233,
+      "grad_norm": 6.869486331939697,
+      "learning_rate": 1.2785857026479609e-06,
+      "loss": 0.3389,
+      "step": 767
+    },
+    {
+      "epoch": 2.690017513134851,
+      "grad_norm": 9.957880973815918,
+      "learning_rate": 1.2499968935887474e-06,
+      "loss": 0.2081,
+      "step": 768
+    },
+    {
+      "epoch": 2.693520140105079,
+      "grad_norm": 4.789759635925293,
+      "learning_rate": 1.221721014965942e-06,
+      "loss": 0.3021,
+      "step": 769
+    },
+    {
+      "epoch": 2.6970227670753064,
+      "grad_norm": 4.189669609069824,
+      "learning_rate": 1.1937585386932281e-06,
+      "loss": 0.1325,
+      "step": 770
+    },
+    {
+      "epoch": 2.7005253940455343,
+      "grad_norm": 12.296935081481934,
+      "learning_rate": 1.1661099314537228e-06,
+      "loss": 0.1675,
+      "step": 771
+    },
+    {
+      "epoch": 2.7040280210157617,
+      "grad_norm": 1.7808837890625,
+      "learning_rate": 1.1387756546921902e-06,
+      "loss": 0.0979,
+      "step": 772
+    },
+    {
+      "epoch": 2.7075306479859895,
+      "grad_norm": 9.73708438873291,
+      "learning_rate": 1.1117561646073316e-06,
+      "loss": 0.3332,
+      "step": 773
+    },
+    {
+      "epoch": 2.7110332749562174,
+      "grad_norm": 10.950447082519531,
+      "learning_rate": 1.085051912144186e-06,
+      "loss": 0.4934,
+      "step": 774
+    },
+    {
+      "epoch": 2.714535901926445,
+      "grad_norm": 7.5403618812561035,
+      "learning_rate": 1.0586633429865966e-06,
+      "loss": 0.2746,
+      "step": 775
+    },
+    {
+      "epoch": 2.7180385288966726,
+      "grad_norm": 7.8029279708862305,
+      "learning_rate": 1.0325908975497657e-06,
+      "loss": 0.4395,
+      "step": 776
+    },
+    {
+      "epoch": 2.7215411558669,
+      "grad_norm": 4.058933258056641,
+      "learning_rate": 1.0068350109729063e-06,
+      "loss": 0.1307,
+      "step": 777
+    },
+    {
+      "epoch": 2.725043782837128,
+      "grad_norm": 7.272954940795898,
+      "learning_rate": 9.813961131120009e-07,
+      "loss": 0.204,
+      "step": 778
+    },
+    {
+      "epoch": 2.7285464098073557,
+      "grad_norm": 5.053043842315674,
+      "learning_rate": 9.562746285326007e-07,
+      "loss": 0.1737,
+      "step": 779
+    },
+    {
+      "epoch": 2.732049036777583,
+      "grad_norm": 4.146439075469971,
+      "learning_rate": 9.314709765027441e-07,
+      "loss": 0.3323,
+      "step": 780
+    },
+    {
+      "epoch": 2.735551663747811,
+      "grad_norm": 9.863334655761719,
+      "learning_rate": 9.069855709859743e-07,
+      "loss": 0.3481,
+      "step": 781
+    },
+    {
+      "epoch": 2.7390542907180384,
+      "grad_norm": 9.322667121887207,
+      "learning_rate": 8.828188206344168e-07,
+      "loss": 0.2881,
+      "step": 782
+    },
+    {
+      "epoch": 2.7425569176882663,
+      "grad_norm": 17.841449737548828,
+      "learning_rate": 8.589711287819624e-07,
+      "loss": 0.5504,
+      "step": 783
+    },
+    {
+      "epoch": 2.746059544658494,
+      "grad_norm": 5.811645984649658,
+      "learning_rate": 8.354428934375414e-07,
+      "loss": 0.2561,
+      "step": 784
+    },
+    {
+      "epoch": 2.7495621716287215,
+      "grad_norm": 8.1492338180542,
+      "learning_rate": 8.122345072784687e-07,
+      "loss": 0.3995,
+      "step": 785
+    },
+    {
+      "epoch": 2.7530647985989494,
+      "grad_norm": 16.187278747558594,
+      "learning_rate": 7.893463576439076e-07,
+      "loss": 0.7552,
+      "step": 786
+    },
+    {
+      "epoch": 2.7565674255691768,
+      "grad_norm": 3.650089979171753,
+      "learning_rate": 7.667788265283871e-07,
+      "loss": 0.2169,
+      "step": 787
+    },
+    {
+      "epoch": 2.7600700525394046,
+      "grad_norm": 6.4505743980407715,
+      "learning_rate": 7.445322905754326e-07,
+      "loss": 0.3677,
+      "step": 788
+    },
+    {
+      "epoch": 2.7635726795096325,
+      "grad_norm": 5.248019695281982,
+      "learning_rate": 7.226071210712926e-07,
+      "loss": 0.1273,
+      "step": 789
+    },
+    {
+      "epoch": 2.76707530647986,
+      "grad_norm": 5.4500837326049805,
+      "learning_rate": 7.010036839387257e-07,
+      "loss": 0.2377,
+      "step": 790
+    },
+    {
+      "epoch": 2.7705779334500873,
+      "grad_norm": 11.434654235839844,
+      "learning_rate": 6.797223397308994e-07,
+      "loss": 0.436,
+      "step": 791
+    },
+    {
+      "epoch": 2.774080560420315,
+      "grad_norm": 3.954472064971924,
+      "learning_rate": 6.587634436253654e-07,
+      "loss": 0.3242,
+      "step": 792
+    },
+    {
+      "epoch": 2.777583187390543,
+      "grad_norm": 7.352035045623779,
+      "learning_rate": 6.381273454181536e-07,
+      "loss": 0.4541,
+      "step": 793
+    },
+    {
+      "epoch": 2.781085814360771,
+      "grad_norm": 8.057424545288086,
+      "learning_rate": 6.178143895179123e-07,
+      "loss": 0.3616,
+      "step": 794
+    },
+    {
+      "epoch": 2.7845884413309983,
+      "grad_norm": 21.68024253845215,
+      "learning_rate": 5.978249149401638e-07,
+      "loss": 0.6163,
+      "step": 795
+    },
+    {
+      "epoch": 2.7880910683012257,
+      "grad_norm": 1.6756069660186768,
+      "learning_rate": 5.781592553016513e-07,
+      "loss": 0.0949,
+      "step": 796
+    },
+    {
+      "epoch": 2.7915936952714535,
+      "grad_norm": 15.292009353637695,
+      "learning_rate": 5.588177388147741e-07,
+      "loss": 0.195,
+      "step": 797
+    },
+    {
+      "epoch": 2.7950963222416814,
+      "grad_norm": 5.428353786468506,
+      "learning_rate": 5.398006882821016e-07,
+      "loss": 0.4268,
+      "step": 798
+    },
+    {
+      "epoch": 2.7985989492119088,
+      "grad_norm": 4.934905529022217,
+      "learning_rate": 5.211084210909922e-07,
+      "loss": 0.4207,
+      "step": 799
+    },
+    {
+      "epoch": 2.8021015761821366,
+      "grad_norm": 32.14579391479492,
+      "learning_rate": 5.027412492082895e-07,
+      "loss": 1.2459,
+      "step": 800
+    },
+    {
+      "epoch": 2.805604203152364,
+      "grad_norm": 5.984612464904785,
+      "learning_rate": 4.846994791751236e-07,
+      "loss": 0.2008,
+      "step": 801
+    },
+    {
+      "epoch": 2.809106830122592,
+      "grad_norm": 8.9047269821167,
+      "learning_rate": 4.669834121017891e-07,
+      "loss": 0.2059,
+      "step": 802
+    },
+    {
+      "epoch": 2.8126094570928197,
+      "grad_norm": 12.699865341186523,
+      "learning_rate": 4.495933436627198e-07,
+      "loss": 0.2789,
+      "step": 803
+    },
+    {
+      "epoch": 2.816112084063047,
+      "grad_norm": 6.121337413787842,
+      "learning_rate": 4.3252956409156164e-07,
+      "loss": 0.334,
+      "step": 804
+    },
+    {
+      "epoch": 2.819614711033275,
+      "grad_norm": 22.385419845581055,
+      "learning_rate": 4.1579235817632126e-07,
+      "loss": 0.7606,
+      "step": 805
+    },
+    {
+      "epoch": 2.8231173380035024,
+      "grad_norm": 2.840463161468506,
+      "learning_rate": 3.993820052546116e-07,
+      "loss": 0.2278,
+      "step": 806
+    },
+    {
+      "epoch": 2.8266199649737302,
+      "grad_norm": 9.669700622558594,
+      "learning_rate": 3.832987792089937e-07,
+      "loss": 0.2599,
+      "step": 807
+    },
+    {
+      "epoch": 2.830122591943958,
+      "grad_norm": 6.415764331817627,
+      "learning_rate": 3.6754294846240934e-07,
+      "loss": 0.4672,
+      "step": 808
+    },
+    {
+      "epoch": 2.8336252189141855,
+      "grad_norm": 12.454601287841797,
+      "learning_rate": 3.5211477597369535e-07,
+      "loss": 0.3072,
+      "step": 809
+    },
+    {
+      "epoch": 2.8371278458844134,
+      "grad_norm": 3.790520191192627,
+      "learning_rate": 3.370145192331964e-07,
+      "loss": 0.2364,
+      "step": 810
+    },
+    {
+      "epoch": 2.8406304728546408,
+      "grad_norm": 6.003750801086426,
+      "learning_rate": 3.222424302584615e-07,
+      "loss": 0.1876,
+      "step": 811
+    },
+    {
+      "epoch": 2.8441330998248686,
+      "grad_norm": 8.819941520690918,
+      "learning_rate": 3.0779875559005636e-07,
+      "loss": 0.4489,
+      "step": 812
+    },
+    {
+      "epoch": 2.8476357267950965,
+      "grad_norm": 9.377801895141602,
+      "learning_rate": 2.936837362874245e-07,
+      "loss": 0.3392,
+      "step": 813
+    },
+    {
+      "epoch": 2.851138353765324,
+      "grad_norm": 5.5884246826171875,
+      "learning_rate": 2.798976079248861e-07,
+      "loss": 0.3151,
+      "step": 814
+    },
+    {
+      "epoch": 2.8546409807355517,
+      "grad_norm": 29.302658081054688,
+      "learning_rate": 2.664406005876874e-07,
+      "loss": 0.6223,
+      "step": 815
+    },
+    {
+      "epoch": 2.858143607705779,
+      "grad_norm": 9.293455123901367,
+      "learning_rate": 2.5331293886817764e-07,
+      "loss": 0.379,
+      "step": 816
+    },
+    {
+      "epoch": 2.861646234676007,
+      "grad_norm": 3.9380757808685303,
+      "learning_rate": 2.4051484186204953e-07,
+      "loss": 0.3059,
+      "step": 817
+    },
+    {
+      "epoch": 2.865148861646235,
+      "grad_norm": 8.622139930725098,
+      "learning_rate": 2.2804652316468e-07,
+      "loss": 0.3891,
+      "step": 818
+    },
+    {
+      "epoch": 2.8686514886164622,
+      "grad_norm": 5.322511196136475,
+      "learning_rate": 2.1590819086758197e-07,
+      "loss": 0.3623,
+      "step": 819
+    },
+    {
+      "epoch": 2.87215411558669,
+      "grad_norm": 3.123744249343872,
+      "learning_rate": 2.0410004755491597e-07,
+      "loss": 0.0961,
+      "step": 820
+    },
+    {
+      "epoch": 2.8756567425569175,
+      "grad_norm": 8.360587120056152,
+      "learning_rate": 1.926222903001085e-07,
+      "loss": 0.4422,
+      "step": 821
+    },
+    {
+      "epoch": 2.8791593695271454,
+      "grad_norm": 8.696887969970703,
+      "learning_rate": 1.814751106625745e-07,
+      "loss": 0.3083,
+      "step": 822
+    },
+    {
+      "epoch": 2.882661996497373,
+      "grad_norm": 9.53421401977539,
+      "learning_rate": 1.7065869468450902e-07,
+      "loss": 0.4419,
+      "step": 823
+    },
+    {
+      "epoch": 2.8861646234676006,
+      "grad_norm": 8.00709056854248,
+      "learning_rate": 1.6017322288779614e-07,
+      "loss": 0.3319,
+      "step": 824
+    },
+    {
+      "epoch": 2.8896672504378285,
+      "grad_norm": 2.6990621089935303,
+      "learning_rate": 1.5001887027097594e-07,
+      "loss": 0.1647,
+      "step": 825
+    },
+    {
+      "epoch": 2.893169877408056,
+      "grad_norm": 6.979440689086914,
+      "learning_rate": 1.4019580630634244e-07,
+      "loss": 0.3581,
+      "step": 826
+    },
+    {
+      "epoch": 2.8966725043782837,
+      "grad_norm": 7.536319732666016,
+      "learning_rate": 1.307041949371013e-07,
+      "loss": 0.3931,
+      "step": 827
+    },
+    {
+      "epoch": 2.9001751313485116,
+      "grad_norm": 11.709566116333008,
+      "learning_rate": 1.2154419457464984e-07,
+      "loss": 0.6539,
+      "step": 828
+    },
+    {
+      "epoch": 2.903677758318739,
+      "grad_norm": 3.197880744934082,
+      "learning_rate": 1.1271595809591696e-07,
+      "loss": 0.2004,
+      "step": 829
+    },
+    {
+      "epoch": 2.907180385288967,
+      "grad_norm": 3.852081775665283,
+      "learning_rate": 1.0421963284081848e-07,
+      "loss": 0.2795,
+      "step": 830
+    },
+    {
+      "epoch": 2.9106830122591942,
+      "grad_norm": 2.3727164268493652,
+      "learning_rate": 9.60553606097947e-08,
+      "loss": 0.2725,
+      "step": 831
+    },
+    {
+      "epoch": 2.914185639229422,
+      "grad_norm": 23.161338806152344,
+      "learning_rate": 8.822327766145889e-08,
+      "loss": 0.7655,
+      "step": 832
+    },
+    {
+      "epoch": 2.91768826619965,
+      "grad_norm": 2.332108974456787,
+      "learning_rate": 8.072351471029471e-08,
+      "loss": 0.1882,
+      "step": 833
+    },
+    {
+      "epoch": 2.9211908931698773,
+      "grad_norm": 10.980060577392578,
+      "learning_rate": 7.355619692450688e-08,
+      "loss": 0.2514,
+      "step": 834
+    },
+    {
+      "epoch": 2.924693520140105,
+      "grad_norm": 6.918917179107666,
+      "learning_rate": 6.672144392390501e-08,
+      "loss": 0.1826,
+      "step": 835
+    },
+    {
+      "epoch": 2.9281961471103326,
+      "grad_norm": 11.677573204040527,
+      "learning_rate": 6.021936977792298e-08,
+      "loss": 0.4053,
+      "step": 836
+    },
+    {
+      "epoch": 2.9316987740805605,
+      "grad_norm": 9.314156532287598,
+      "learning_rate": 5.405008300371384e-08,
+      "loss": 0.3143,
+      "step": 837
+    },
+    {
+      "epoch": 2.9352014010507883,
+      "grad_norm": 5.6505045890808105,
+      "learning_rate": 4.821368656433123e-08,
+      "loss": 0.2023,
+      "step": 838
+    },
+    {
+      "epoch": 2.9387040280210157,
+      "grad_norm": 11.518875122070312,
+      "learning_rate": 4.2710277867015204e-08,
+      "loss": 0.3932,
+      "step": 839
+    },
+    {
+      "epoch": 2.9422066549912436,
+      "grad_norm": 4.484135150909424,
+      "learning_rate": 3.753994876156908e-08,
+      "loss": 0.2197,
+      "step": 840
+    },
+    {
+      "epoch": 2.945709281961471,
+      "grad_norm": 17.19270133972168,
+      "learning_rate": 3.270278553882511e-08,
+      "loss": 1.0666,
+      "step": 841
+    },
+    {
+      "epoch": 2.949211908931699,
+      "grad_norm": 2.880202293395996,
+      "learning_rate": 2.8198868929201206e-08,
+      "loss": 0.1144,
+      "step": 842
+    },
+    {
+      "epoch": 2.9527145359019267,
+      "grad_norm": 15.712217330932617,
+      "learning_rate": 2.402827410135533e-08,
+      "loss": 0.3268,
+      "step": 843
+    },
+    {
+      "epoch": 2.956217162872154,
+      "grad_norm": 4.731568813323975,
+      "learning_rate": 2.0191070660935396e-08,
+      "loss": 0.2733,
+      "step": 844
+    },
+    {
+      "epoch": 2.959719789842382,
+      "grad_norm": 9.0734224319458,
+      "learning_rate": 1.668732264940687e-08,
+      "loss": 0.2679,
+      "step": 845
+    },
+    {
+      "epoch": 2.9632224168126093,
+      "grad_norm": 13.673662185668945,
+      "learning_rate": 1.3517088542995827e-08,
+      "loss": 0.6694,
+      "step": 846
+    },
+    {
+      "epoch": 2.966725043782837,
+      "grad_norm": 9.462847709655762,
+      "learning_rate": 1.068042125170754e-08,
+      "loss": 0.2576,
+      "step": 847
+    },
+    {
+      "epoch": 2.970227670753065,
+      "grad_norm": 5.057038307189941,
+      "learning_rate": 8.177368118440499e-09,
+      "loss": 0.1687,
+      "step": 848
+    },
+    {
+      "epoch": 2.9737302977232924,
+      "grad_norm": 4.47308874130249,
+      "learning_rate": 6.007970918202599e-09,
+      "loss": 0.2286,
+      "step": 849
+    },
+    {
+      "epoch": 2.9772329246935203,
+      "grad_norm": 2.5368804931640625,
+      "learning_rate": 4.172265857411706e-09,
+      "loss": 0.2175,
+      "step": 850
+    },
+    {
+      "epoch": 2.9807355516637477,
+      "grad_norm": 6.23403263092041,
+      "learning_rate": 2.670283573285026e-09,
+      "loss": 0.2142,
+      "step": 851
+    },
+    {
+      "epoch": 2.9842381786339756,
+      "grad_norm": 9.441424369812012,
+      "learning_rate": 1.502049133341732e-09,
+      "loss": 0.3126,
+      "step": 852
+    },
+    {
+      "epoch": 2.9877408056042034,
+      "grad_norm": 5.0686163902282715,
+      "learning_rate": 6.675820349655304e-10,
+      "loss": 0.2322,
+      "step": 853
+    },
+    {
+      "epoch": 2.991243432574431,
+      "grad_norm": 17.517122268676758,
+      "learning_rate": 1.6689620510046412e-10,
+      "loss": 0.6579,
+      "step": 854
+    },
+    {
+      "epoch": 2.9947460595446582,
+      "grad_norm": 7.781464576721191,
+      "learning_rate": 0.0,
+      "loss": 0.3209,
+      "step": 855
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 855,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9951049450045440.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}