diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15407 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.994311717861206,
+  "eval_steps": 500,
+  "global_step": 2195,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0022753128555176336,
+      "grad_norm": 5.685660859809189,
+      "learning_rate": 1.8181818181818183e-07,
+      "loss": 0.8913,
+      "step": 1
+    },
+    {
+      "epoch": 0.004550625711035267,
+      "grad_norm": 5.711076889258604,
+      "learning_rate": 3.6363636363636366e-07,
+      "loss": 0.8872,
+      "step": 2
+    },
+    {
+      "epoch": 0.006825938566552901,
+      "grad_norm": 5.704778542595518,
+      "learning_rate": 5.454545454545455e-07,
+      "loss": 0.8818,
+      "step": 3
+    },
+    {
+      "epoch": 0.009101251422070534,
+      "grad_norm": 5.822601641411562,
+      "learning_rate": 7.272727272727273e-07,
+      "loss": 0.9099,
+      "step": 4
+    },
+    {
+      "epoch": 0.011376564277588168,
+      "grad_norm": 5.779461616420256,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.9119,
+      "step": 5
+    },
+    {
+      "epoch": 0.013651877133105802,
+      "grad_norm": 5.729877153470166,
+      "learning_rate": 1.090909090909091e-06,
+      "loss": 0.8553,
+      "step": 6
+    },
+    {
+      "epoch": 0.015927189988623434,
+      "grad_norm": 5.281427413434935,
+      "learning_rate": 1.2727272727272728e-06,
+      "loss": 0.8673,
+      "step": 7
+    },
+    {
+      "epoch": 0.01820250284414107,
+      "grad_norm": 5.052471409802485,
+      "learning_rate": 1.4545454545454546e-06,
+      "loss": 0.8509,
+      "step": 8
+    },
+    {
+      "epoch": 0.020477815699658702,
+      "grad_norm": 4.389244479229915,
+      "learning_rate": 1.6363636363636365e-06,
+      "loss": 0.8566,
+      "step": 9
+    },
+    {
+      "epoch": 0.022753128555176336,
+      "grad_norm": 4.094630788604825,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.8302,
+      "step": 10
+    },
+    {
+      "epoch": 0.02502844141069397,
+      "grad_norm": 2.4624238627507267,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.7387,
+      "step": 11
+    },
+    {
+      "epoch": 0.027303754266211604,
+      "grad_norm": 2.456278255907781,
+      "learning_rate": 2.181818181818182e-06,
+      "loss": 0.7747,
+      "step": 12
+    },
+    {
+      "epoch": 0.02957906712172924,
+      "grad_norm": 2.072663924996278,
+      "learning_rate": 2.363636363636364e-06,
+      "loss": 0.7314,
+      "step": 13
+    },
+    {
+      "epoch": 0.03185437997724687,
+      "grad_norm": 2.051452583818761,
+      "learning_rate": 2.5454545454545456e-06,
+      "loss": 0.7661,
+      "step": 14
+    },
+    {
+      "epoch": 0.034129692832764506,
+      "grad_norm": 2.837150084118399,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.8176,
+      "step": 15
+    },
+    {
+      "epoch": 0.03640500568828214,
+      "grad_norm": 3.2676216662546578,
+      "learning_rate": 2.9090909090909093e-06,
+      "loss": 0.7332,
+      "step": 16
+    },
+    {
+      "epoch": 0.038680318543799774,
+      "grad_norm": 3.478330510465565,
+      "learning_rate": 3.090909090909091e-06,
+      "loss": 0.8074,
+      "step": 17
+    },
+    {
+      "epoch": 0.040955631399317405,
+      "grad_norm": 3.3365962406701617,
+      "learning_rate": 3.272727272727273e-06,
+      "loss": 0.7319,
+      "step": 18
+    },
+    {
+      "epoch": 0.04323094425483504,
+      "grad_norm": 3.172713062908951,
+      "learning_rate": 3.454545454545455e-06,
+      "loss": 0.7348,
+      "step": 19
+    },
+    {
+      "epoch": 0.04550625711035267,
+      "grad_norm": 2.647469572423114,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.6978,
+      "step": 20
+    },
+    {
+      "epoch": 0.04778156996587031,
+      "grad_norm": 1.9123731673559052,
+      "learning_rate": 3.818181818181819e-06,
+      "loss": 0.6676,
+      "step": 21
+    },
+    {
+      "epoch": 0.05005688282138794,
+      "grad_norm": 1.5607974881146613,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6623,
+      "step": 22
+    },
+    {
+      "epoch": 0.05233219567690557,
+      "grad_norm": 1.3192020195203946,
+      "learning_rate": 4.181818181818182e-06,
+      "loss": 0.6766,
+      "step": 23
+    },
+    {
+      "epoch": 0.05460750853242321,
+      "grad_norm": 1.1681629362926533,
+      "learning_rate": 4.363636363636364e-06,
+      "loss": 0.6282,
+      "step": 24
+    },
+    {
+      "epoch": 0.05688282138794084,
+      "grad_norm": 1.4253417702456757,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.6829,
+      "step": 25
+    },
+    {
+      "epoch": 0.05915813424345848,
+      "grad_norm": 1.304707249259586,
+      "learning_rate": 4.727272727272728e-06,
+      "loss": 0.6547,
+      "step": 26
+    },
+    {
+      "epoch": 0.06143344709897611,
+      "grad_norm": 1.229005210666558,
+      "learning_rate": 4.90909090909091e-06,
+      "loss": 0.6336,
+      "step": 27
+    },
+    {
+      "epoch": 0.06370875995449374,
+      "grad_norm": 1.3050185528089813,
+      "learning_rate": 5.090909090909091e-06,
+      "loss": 0.6571,
+      "step": 28
+    },
+    {
+      "epoch": 0.06598407281001138,
+      "grad_norm": 1.1603661646747794,
+      "learning_rate": 5.272727272727273e-06,
+      "loss": 0.6185,
+      "step": 29
+    },
+    {
+      "epoch": 0.06825938566552901,
+      "grad_norm": 0.975017107538546,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.6272,
+      "step": 30
+    },
+    {
+      "epoch": 0.07053469852104664,
+      "grad_norm": 0.9775508605964778,
+      "learning_rate": 5.636363636363636e-06,
+      "loss": 0.5826,
+      "step": 31
+    },
+    {
+      "epoch": 0.07281001137656427,
+      "grad_norm": 1.1490334435580931,
+      "learning_rate": 5.8181818181818185e-06,
+      "loss": 0.6514,
+      "step": 32
+    },
+    {
+      "epoch": 0.07508532423208192,
+      "grad_norm": 1.0553018624073731,
+      "learning_rate": 6e-06,
+      "loss": 0.6011,
+      "step": 33
+    },
+    {
+      "epoch": 0.07736063708759955,
+      "grad_norm": 0.6887889050426664,
+      "learning_rate": 6.181818181818182e-06,
+      "loss": 0.589,
+      "step": 34
+    },
+    {
+      "epoch": 0.07963594994311718,
+      "grad_norm": 0.6451648576496538,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.5828,
+      "step": 35
+    },
+    {
+      "epoch": 0.08191126279863481,
+      "grad_norm": 0.769304174505827,
+      "learning_rate": 6.545454545454546e-06,
+      "loss": 0.5921,
+      "step": 36
+    },
+    {
+      "epoch": 0.08418657565415244,
+      "grad_norm": 0.8058219749803411,
+      "learning_rate": 6.7272727272727275e-06,
+      "loss": 0.5817,
+      "step": 37
+    },
+    {
+      "epoch": 0.08646188850967008,
+      "grad_norm": 0.6959095038632431,
+      "learning_rate": 6.90909090909091e-06,
+      "loss": 0.5872,
+      "step": 38
+    },
+    {
+      "epoch": 0.08873720136518772,
+      "grad_norm": 0.6998025818497815,
+      "learning_rate": 7.0909090909090916e-06,
+      "loss": 0.6163,
+      "step": 39
+    },
+    {
+      "epoch": 0.09101251422070535,
+      "grad_norm": 0.6481637742829409,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.596,
+      "step": 40
+    },
+    {
+      "epoch": 0.09328782707622298,
+      "grad_norm": 0.6701911156282011,
+      "learning_rate": 7.454545454545456e-06,
+      "loss": 0.5605,
+      "step": 41
+    },
+    {
+      "epoch": 0.09556313993174062,
+      "grad_norm": 0.621382473643511,
+      "learning_rate": 7.636363636363638e-06,
+      "loss": 0.5656,
+      "step": 42
+    },
+    {
+      "epoch": 0.09783845278725825,
+      "grad_norm": 0.6366513729903114,
+      "learning_rate": 7.81818181818182e-06,
+      "loss": 0.5677,
+      "step": 43
+    },
+    {
+      "epoch": 0.10011376564277588,
+      "grad_norm": 0.6530354813151484,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.5768,
+      "step": 44
+    },
+    {
+      "epoch": 0.10238907849829351,
+      "grad_norm": 0.5250961965862125,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.571,
+      "step": 45
+    },
+    {
+      "epoch": 0.10466439135381114,
+      "grad_norm": 0.5862346211984556,
+      "learning_rate": 8.363636363636365e-06,
+      "loss": 0.5502,
+      "step": 46
+    },
+    {
+      "epoch": 0.10693970420932879,
+      "grad_norm": 0.5645831108750284,
+      "learning_rate": 8.545454545454546e-06,
+      "loss": 0.5612,
+      "step": 47
+    },
+    {
+      "epoch": 0.10921501706484642,
+      "grad_norm": 0.5428516306287078,
+      "learning_rate": 8.727272727272728e-06,
+      "loss": 0.5941,
+      "step": 48
+    },
+    {
+      "epoch": 0.11149032992036405,
+      "grad_norm": 0.5470221016318203,
+      "learning_rate": 8.90909090909091e-06,
+      "loss": 0.5478,
+      "step": 49
+    },
+    {
+      "epoch": 0.11376564277588168,
+      "grad_norm": 0.5670301758284394,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.568,
+      "step": 50
+    },
+    {
+      "epoch": 0.11604095563139932,
+      "grad_norm": 0.5238687690636948,
+      "learning_rate": 9.272727272727273e-06,
+      "loss": 0.5352,
+      "step": 51
+    },
+    {
+      "epoch": 0.11831626848691695,
+      "grad_norm": 0.494334648795746,
+      "learning_rate": 9.454545454545456e-06,
+      "loss": 0.573,
+      "step": 52
+    },
+    {
+      "epoch": 0.12059158134243458,
+      "grad_norm": 0.6418508884639001,
+      "learning_rate": 9.636363636363638e-06,
+      "loss": 0.5716,
+      "step": 53
+    },
+    {
+      "epoch": 0.12286689419795221,
+      "grad_norm": 0.5544219094970814,
+      "learning_rate": 9.81818181818182e-06,
+      "loss": 0.5671,
+      "step": 54
+    },
+    {
+      "epoch": 0.12514220705346984,
+      "grad_norm": 0.5044191975148502,
+      "learning_rate": 1e-05,
+      "loss": 0.5649,
+      "step": 55
+    },
+    {
+      "epoch": 0.12741751990898748,
+      "grad_norm": 0.5352497333561101,
+      "learning_rate": 1.0181818181818182e-05,
+      "loss": 0.5418,
+      "step": 56
+    },
+    {
+      "epoch": 0.1296928327645051,
+      "grad_norm": 0.5683305195263919,
+      "learning_rate": 1.0363636363636364e-05,
+      "loss": 0.5346,
+      "step": 57
+    },
+    {
+      "epoch": 0.13196814562002276,
+      "grad_norm": 0.8310946445690662,
+      "learning_rate": 1.0545454545454546e-05,
+      "loss": 0.5503,
+      "step": 58
+    },
+    {
+      "epoch": 0.1342434584755404,
+      "grad_norm": 0.5692386390346372,
+      "learning_rate": 1.0727272727272729e-05,
+      "loss": 0.5641,
+      "step": 59
+    },
+    {
+      "epoch": 0.13651877133105803,
+      "grad_norm": 0.6552853434740772,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 0.5555,
+      "step": 60
+    },
+    {
+      "epoch": 0.13879408418657566,
+      "grad_norm": 0.5558986591762781,
+      "learning_rate": 1.1090909090909092e-05,
+      "loss": 0.544,
+      "step": 61
+    },
+    {
+      "epoch": 0.1410693970420933,
+      "grad_norm": 0.5898414892780288,
+      "learning_rate": 1.1272727272727272e-05,
+      "loss": 0.5564,
+      "step": 62
+    },
+    {
+      "epoch": 0.14334470989761092,
+      "grad_norm": 0.506476126864649,
+      "learning_rate": 1.1454545454545455e-05,
+      "loss": 0.5765,
+      "step": 63
+    },
+    {
+      "epoch": 0.14562002275312855,
+      "grad_norm": 0.5291323959711722,
+      "learning_rate": 1.1636363636363637e-05,
+      "loss": 0.5618,
+      "step": 64
+    },
+    {
+      "epoch": 0.14789533560864618,
+      "grad_norm": 0.6080011056361084,
+      "learning_rate": 1.181818181818182e-05,
+      "loss": 0.53,
+      "step": 65
+    },
+    {
+      "epoch": 0.15017064846416384,
+      "grad_norm": 0.5685623206194781,
+      "learning_rate": 1.2e-05,
+      "loss": 0.5719,
+      "step": 66
+    },
+    {
+      "epoch": 0.15244596131968147,
+      "grad_norm": 0.6292150667659716,
+      "learning_rate": 1.2181818181818184e-05,
+      "loss": 0.5512,
+      "step": 67
+    },
+    {
+      "epoch": 0.1547212741751991,
+      "grad_norm": 0.5914023995005103,
+      "learning_rate": 1.2363636363636364e-05,
+      "loss": 0.5624,
+      "step": 68
+    },
+    {
+      "epoch": 0.15699658703071673,
+      "grad_norm": 0.561334886574656,
+      "learning_rate": 1.2545454545454547e-05,
+      "loss": 0.5481,
+      "step": 69
+    },
+    {
+      "epoch": 0.15927189988623436,
+      "grad_norm": 0.6735343236499104,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 0.5932,
+      "step": 70
+    },
+    {
+      "epoch": 0.161547212741752,
+      "grad_norm": 0.5906206771821041,
+      "learning_rate": 1.2909090909090912e-05,
+      "loss": 0.5337,
+      "step": 71
+    },
+    {
+      "epoch": 0.16382252559726962,
+      "grad_norm": 0.6355464197194023,
+      "learning_rate": 1.3090909090909092e-05,
+      "loss": 0.5842,
+      "step": 72
+    },
+    {
+      "epoch": 0.16609783845278725,
+      "grad_norm": 0.6048043966163964,
+      "learning_rate": 1.3272727272727275e-05,
+      "loss": 0.5668,
+      "step": 73
+    },
+    {
+      "epoch": 0.16837315130830488,
+      "grad_norm": 0.7298080325337586,
+      "learning_rate": 1.3454545454545455e-05,
+      "loss": 0.5796,
+      "step": 74
+    },
+    {
+      "epoch": 0.17064846416382254,
+      "grad_norm": 0.6230989396592326,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.5466,
+      "step": 75
+    },
+    {
+      "epoch": 0.17292377701934017,
+      "grad_norm": 0.6218921985816483,
+      "learning_rate": 1.381818181818182e-05,
+      "loss": 0.5137,
+      "step": 76
+    },
+    {
+      "epoch": 0.1751990898748578,
+      "grad_norm": 0.5506392277633055,
+      "learning_rate": 1.4e-05,
+      "loss": 0.5271,
+      "step": 77
+    },
+    {
+      "epoch": 0.17747440273037543,
+      "grad_norm": 0.709091172729563,
+      "learning_rate": 1.4181818181818183e-05,
+      "loss": 0.5275,
+      "step": 78
+    },
+    {
+      "epoch": 0.17974971558589306,
+      "grad_norm": 0.7052601004280128,
+      "learning_rate": 1.4363636363636365e-05,
+      "loss": 0.5706,
+      "step": 79
+    },
+    {
+      "epoch": 0.1820250284414107,
+      "grad_norm": 0.5238245701618371,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 0.5656,
+      "step": 80
+    },
+    {
+      "epoch": 0.18430034129692832,
+      "grad_norm": 0.6687841687625173,
+      "learning_rate": 1.4727272727272728e-05,
+      "loss": 0.5197,
+      "step": 81
+    },
+    {
+      "epoch": 0.18657565415244595,
+      "grad_norm": 0.550803915665101,
+      "learning_rate": 1.4909090909090911e-05,
+      "loss": 0.4971,
+      "step": 82
+    },
+    {
+      "epoch": 0.18885096700796358,
+      "grad_norm": 0.5247163009385345,
+      "learning_rate": 1.5090909090909091e-05,
+      "loss": 0.5307,
+      "step": 83
+    },
+    {
+      "epoch": 0.19112627986348124,
+      "grad_norm": 0.6578165456867408,
+      "learning_rate": 1.5272727272727276e-05,
+      "loss": 0.5316,
+      "step": 84
+    },
+    {
+      "epoch": 0.19340159271899887,
+      "grad_norm": 0.6381427629949843,
+      "learning_rate": 1.5454545454545454e-05,
+      "loss": 0.5406,
+      "step": 85
+    },
+    {
+      "epoch": 0.1956769055745165,
+      "grad_norm": 0.6201059926158639,
+      "learning_rate": 1.563636363636364e-05,
+      "loss": 0.5359,
+      "step": 86
+    },
+    {
+      "epoch": 0.19795221843003413,
+      "grad_norm": 0.7319265448424545,
+      "learning_rate": 1.5818181818181818e-05,
+      "loss": 0.5494,
+      "step": 87
+    },
+    {
+      "epoch": 0.20022753128555176,
+      "grad_norm": 0.8254922009620973,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.5519,
+      "step": 88
+    },
+    {
+      "epoch": 0.2025028441410694,
+      "grad_norm": 0.6596770624786563,
+      "learning_rate": 1.6181818181818184e-05,
+      "loss": 0.5491,
+      "step": 89
+    },
+    {
+      "epoch": 0.20477815699658702,
+      "grad_norm": 0.7612010560559346,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 0.57,
+      "step": 90
+    },
+    {
+      "epoch": 0.20705346985210465,
+      "grad_norm": 0.6052327879452324,
+      "learning_rate": 1.6545454545454548e-05,
+      "loss": 0.5323,
+      "step": 91
+    },
+    {
+      "epoch": 0.20932878270762229,
+      "grad_norm": 0.6173795688483125,
+      "learning_rate": 1.672727272727273e-05,
+      "loss": 0.5033,
+      "step": 92
+    },
+    {
+      "epoch": 0.21160409556313994,
+      "grad_norm": 0.592019998832286,
+      "learning_rate": 1.690909090909091e-05,
+      "loss": 0.5387,
+      "step": 93
+    },
+    {
+      "epoch": 0.21387940841865757,
+      "grad_norm": 0.5933082150936967,
+      "learning_rate": 1.7090909090909092e-05,
+      "loss": 0.4999,
+      "step": 94
+    },
+    {
+      "epoch": 0.2161547212741752,
+      "grad_norm": 0.5173687814283479,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 0.5254,
+      "step": 95
+    },
+    {
+      "epoch": 0.21843003412969283,
+      "grad_norm": 0.6135332216240957,
+      "learning_rate": 1.7454545454545456e-05,
+      "loss": 0.5215,
+      "step": 96
+    },
+    {
+      "epoch": 0.22070534698521047,
+      "grad_norm": 0.6176285527248535,
+      "learning_rate": 1.7636363636363637e-05,
+      "loss": 0.5283,
+      "step": 97
+    },
+    {
+      "epoch": 0.2229806598407281,
+      "grad_norm": 0.51907123478383,
+      "learning_rate": 1.781818181818182e-05,
+      "loss": 0.5077,
+      "step": 98
+    },
+    {
+      "epoch": 0.22525597269624573,
+      "grad_norm": 0.5985820930171443,
+      "learning_rate": 1.8e-05,
+      "loss": 0.5314,
+      "step": 99
+    },
+    {
+      "epoch": 0.22753128555176336,
+      "grad_norm": 0.5990332411759242,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.5608,
+      "step": 100
+    },
+    {
+      "epoch": 0.229806598407281,
+      "grad_norm": 0.539437803681028,
+      "learning_rate": 1.8363636363636367e-05,
+      "loss": 0.5294,
+      "step": 101
+    },
+    {
+      "epoch": 0.23208191126279865,
+      "grad_norm": 0.6583588524475489,
+      "learning_rate": 1.8545454545454545e-05,
+      "loss": 0.5217,
+      "step": 102
+    },
+    {
+      "epoch": 0.23435722411831628,
+      "grad_norm": 0.6082065004111803,
+      "learning_rate": 1.872727272727273e-05,
+      "loss": 0.5735,
+      "step": 103
+    },
+    {
+      "epoch": 0.2366325369738339,
+      "grad_norm": 0.6830805339768707,
+      "learning_rate": 1.8909090909090912e-05,
+      "loss": 0.4806,
+      "step": 104
+    },
+    {
+      "epoch": 0.23890784982935154,
+      "grad_norm": 0.7127039583647218,
+      "learning_rate": 1.9090909090909094e-05,
+      "loss": 0.5464,
+      "step": 105
+    },
+    {
+      "epoch": 0.24118316268486917,
+      "grad_norm": 0.5799258813203299,
+      "learning_rate": 1.9272727272727275e-05,
+      "loss": 0.5456,
+      "step": 106
+    },
+    {
+      "epoch": 0.2434584755403868,
+      "grad_norm": 0.626832121785901,
+      "learning_rate": 1.9454545454545457e-05,
+      "loss": 0.5155,
+      "step": 107
+    },
+    {
+      "epoch": 0.24573378839590443,
+      "grad_norm": 0.626545364067402,
+      "learning_rate": 1.963636363636364e-05,
+      "loss": 0.5728,
+      "step": 108
+    },
+    {
+      "epoch": 0.24800910125142206,
+      "grad_norm": 0.6756228608832343,
+      "learning_rate": 1.981818181818182e-05,
+      "loss": 0.499,
+      "step": 109
+    },
+    {
+      "epoch": 0.2502844141069397,
+      "grad_norm": 0.6597283708892084,
+      "learning_rate": 2e-05,
+      "loss": 0.5033,
+      "step": 110
+    },
+    {
+      "epoch": 0.2525597269624573,
+      "grad_norm": 0.5982551151532118,
+      "learning_rate": 2.0181818181818183e-05,
+      "loss": 0.5017,
+      "step": 111
+    },
+    {
+      "epoch": 0.25483503981797495,
+      "grad_norm": 0.744281625979201,
+      "learning_rate": 2.0363636363636365e-05,
+      "loss": 0.5251,
+      "step": 112
+    },
+    {
+      "epoch": 0.2571103526734926,
+      "grad_norm": 0.6016672894534352,
+      "learning_rate": 2.054545454545455e-05,
+      "loss": 0.5551,
+      "step": 113
+    },
+    {
+      "epoch": 0.2593856655290102,
+      "grad_norm": 0.587932618261952,
+      "learning_rate": 2.0727272727272728e-05,
+      "loss": 0.5354,
+      "step": 114
+    },
+    {
+      "epoch": 0.2616609783845279,
+      "grad_norm": 0.708698769469988,
+      "learning_rate": 2.090909090909091e-05,
+      "loss": 0.5264,
+      "step": 115
+    },
+    {
+      "epoch": 0.26393629124004553,
+      "grad_norm": 0.5828643101576894,
+      "learning_rate": 2.109090909090909e-05,
+      "loss": 0.5196,
+      "step": 116
+    },
+    {
+      "epoch": 0.26621160409556316,
+      "grad_norm": 0.6139046114610022,
+      "learning_rate": 2.1272727272727276e-05,
+      "loss": 0.5304,
+      "step": 117
+    },
+    {
+      "epoch": 0.2684869169510808,
+      "grad_norm": 0.5890111951541503,
+      "learning_rate": 2.1454545454545458e-05,
+      "loss": 0.5259,
+      "step": 118
+    },
+    {
+      "epoch": 0.2707622298065984,
+      "grad_norm": 0.5627119961934509,
+      "learning_rate": 2.1636363636363636e-05,
+      "loss": 0.544,
+      "step": 119
+    },
+    {
+      "epoch": 0.27303754266211605,
+      "grad_norm": 0.562616066387851,
+      "learning_rate": 2.1818181818181818e-05,
+      "loss": 0.5183,
+      "step": 120
+    },
+    {
+      "epoch": 0.2753128555176337,
+      "grad_norm": 0.640542051027252,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.5378,
+      "step": 121
+    },
+    {
+      "epoch": 0.2775881683731513,
+      "grad_norm": 0.8129933150812961,
+      "learning_rate": 2.2181818181818184e-05,
+      "loss": 0.5439,
+      "step": 122
+    },
+    {
+      "epoch": 0.27986348122866894,
+      "grad_norm": 0.6024730739925737,
+      "learning_rate": 2.2363636363636366e-05,
+      "loss": 0.5456,
+      "step": 123
+    },
+    {
+      "epoch": 0.2821387940841866,
+      "grad_norm": 0.5556244301627785,
+      "learning_rate": 2.2545454545454544e-05,
+      "loss": 0.498,
+      "step": 124
+    },
+    {
+      "epoch": 0.2844141069397042,
+      "grad_norm": 0.5805044282934706,
+      "learning_rate": 2.2727272727272733e-05,
+      "loss": 0.4867,
+      "step": 125
+    },
+    {
+      "epoch": 0.28668941979522183,
+      "grad_norm": 0.7621407399983073,
+      "learning_rate": 2.290909090909091e-05,
+      "loss": 0.5054,
+      "step": 126
+    },
+    {
+      "epoch": 0.28896473265073946,
+      "grad_norm": 0.6072052484471204,
+      "learning_rate": 2.3090909090909093e-05,
+      "loss": 0.5057,
+      "step": 127
+    },
+    {
+      "epoch": 0.2912400455062571,
+      "grad_norm": 0.750258183088792,
+      "learning_rate": 2.3272727272727274e-05,
+      "loss": 0.4932,
+      "step": 128
+    },
+    {
+      "epoch": 0.2935153583617747,
+      "grad_norm": 0.7769802914476007,
+      "learning_rate": 2.3454545454545456e-05,
+      "loss": 0.5352,
+      "step": 129
+    },
+    {
+      "epoch": 0.29579067121729236,
+      "grad_norm": 0.6654153752837011,
+      "learning_rate": 2.363636363636364e-05,
+      "loss": 0.5173,
+      "step": 130
+    },
+    {
+      "epoch": 0.29806598407281,
+      "grad_norm": 0.7441767179318566,
+      "learning_rate": 2.381818181818182e-05,
+      "loss": 0.5169,
+      "step": 131
+    },
+    {
+      "epoch": 0.3003412969283277,
+      "grad_norm": 0.7967429150652531,
+      "learning_rate": 2.4e-05,
+      "loss": 0.5314,
+      "step": 132
+    },
+    {
+      "epoch": 0.3026166097838453,
+      "grad_norm": 0.8000301237971132,
+      "learning_rate": 2.4181818181818182e-05,
+      "loss": 0.5533,
+      "step": 133
+    },
+    {
+      "epoch": 0.30489192263936293,
+      "grad_norm": 0.939455311035287,
+      "learning_rate": 2.4363636363636367e-05,
+      "loss": 0.5511,
+      "step": 134
+    },
+    {
+      "epoch": 0.30716723549488056,
+      "grad_norm": 0.9480823859443015,
+      "learning_rate": 2.454545454545455e-05,
+      "loss": 0.5384,
+      "step": 135
+    },
+    {
+      "epoch": 0.3094425483503982,
+      "grad_norm": 0.7533352675603919,
+      "learning_rate": 2.4727272727272727e-05,
+      "loss": 0.5401,
+      "step": 136
+    },
+    {
+      "epoch": 0.3117178612059158,
+      "grad_norm": 0.9474216004746516,
+      "learning_rate": 2.490909090909091e-05,
+      "loss": 0.5068,
+      "step": 137
+    },
+    {
+      "epoch": 0.31399317406143346,
+      "grad_norm": 0.8947273707720739,
+      "learning_rate": 2.5090909090909094e-05,
+      "loss": 0.5472,
+      "step": 138
+    },
+    {
+      "epoch": 0.3162684869169511,
+      "grad_norm": 0.9068890803831923,
+      "learning_rate": 2.5272727272727275e-05,
+      "loss": 0.5413,
+      "step": 139
+    },
+    {
+      "epoch": 0.3185437997724687,
+      "grad_norm": 0.7543198462881926,
+      "learning_rate": 2.5454545454545457e-05,
+      "loss": 0.5188,
+      "step": 140
+    },
+    {
+      "epoch": 0.32081911262798635,
+      "grad_norm": 0.7278501214873285,
+      "learning_rate": 2.563636363636364e-05,
+      "loss": 0.4928,
+      "step": 141
+    },
+    {
+      "epoch": 0.323094425483504,
+      "grad_norm": 0.8008535696995278,
+      "learning_rate": 2.5818181818181824e-05,
+      "loss": 0.5598,
+      "step": 142
+    },
+    {
+      "epoch": 0.3253697383390216,
+      "grad_norm": 0.9543254203313492,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.5543,
+      "step": 143
+    },
+    {
+      "epoch": 0.32764505119453924,
+      "grad_norm": 0.58106713770697,
+      "learning_rate": 2.6181818181818183e-05,
+      "loss": 0.4953,
+      "step": 144
+    },
+    {
+      "epoch": 0.32992036405005687,
+      "grad_norm": 0.8674304291572694,
+      "learning_rate": 2.6363636363636365e-05,
+      "loss": 0.5151,
+      "step": 145
+    },
+    {
+      "epoch": 0.3321956769055745,
+      "grad_norm": 0.8493822180548894,
+      "learning_rate": 2.654545454545455e-05,
+      "loss": 0.5163,
+      "step": 146
+    },
+    {
+      "epoch": 0.33447098976109213,
+      "grad_norm": 0.8187615665950908,
+      "learning_rate": 2.672727272727273e-05,
+      "loss": 0.5566,
+      "step": 147
+    },
+    {
+      "epoch": 0.33674630261660976,
+      "grad_norm": 0.9950204016308096,
+      "learning_rate": 2.690909090909091e-05,
+      "loss": 0.5371,
+      "step": 148
+    },
+    {
+      "epoch": 0.3390216154721274,
+      "grad_norm": 0.8808677119498094,
+      "learning_rate": 2.709090909090909e-05,
+      "loss": 0.5233,
+      "step": 149
+    },
+    {
+      "epoch": 0.3412969283276451,
+      "grad_norm": 0.8457240122266904,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.516,
+      "step": 150
+    },
+    {
+      "epoch": 0.3435722411831627,
+      "grad_norm": 0.8283847683700987,
+      "learning_rate": 2.7454545454545458e-05,
+      "loss": 0.503,
+      "step": 151
+    },
+    {
+      "epoch": 0.34584755403868034,
+      "grad_norm": 0.8742562793745727,
+      "learning_rate": 2.763636363636364e-05,
+      "loss": 0.5842,
+      "step": 152
+    },
+    {
+      "epoch": 0.34812286689419797,
+      "grad_norm": 0.73368286028384,
+      "learning_rate": 2.781818181818182e-05,
+      "loss": 0.4992,
+      "step": 153
+    },
+    {
+      "epoch": 0.3503981797497156,
+      "grad_norm": 0.9115501625619696,
+      "learning_rate": 2.8e-05,
+      "loss": 0.544,
+      "step": 154
+    },
+    {
+      "epoch": 0.35267349260523323,
+      "grad_norm": 0.7823471567359226,
+      "learning_rate": 2.8181818181818185e-05,
+      "loss": 0.5255,
+      "step": 155
+    },
+    {
+      "epoch": 0.35494880546075086,
+      "grad_norm": 0.8817254934737524,
+      "learning_rate": 2.8363636363636366e-05,
+      "loss": 0.5192,
+      "step": 156
+    },
+    {
+      "epoch": 0.3572241183162685,
+      "grad_norm": 0.6669781410669484,
+      "learning_rate": 2.8545454545454548e-05,
+      "loss": 0.4928,
+      "step": 157
+    },
+    {
+      "epoch": 0.3594994311717861,
+      "grad_norm": 0.7564492150228576,
+      "learning_rate": 2.872727272727273e-05,
+      "loss": 0.4826,
+      "step": 158
+    },
+    {
+      "epoch": 0.36177474402730375,
+      "grad_norm": 0.8455103889650585,
+      "learning_rate": 2.8909090909090914e-05,
+      "loss": 0.4954,
+      "step": 159
+    },
+    {
+      "epoch": 0.3640500568828214,
+      "grad_norm": 0.8961703984285883,
+      "learning_rate": 2.9090909090909093e-05,
+      "loss": 0.4845,
+      "step": 160
+    },
+    {
+      "epoch": 0.366325369738339,
+      "grad_norm": 0.6730757645383194,
+      "learning_rate": 2.9272727272727274e-05,
+      "loss": 0.5241,
+      "step": 161
+    },
+    {
+      "epoch": 0.36860068259385664,
+      "grad_norm": 1.0257851109411285,
+      "learning_rate": 2.9454545454545456e-05,
+      "loss": 0.4895,
+      "step": 162
+    },
+    {
+      "epoch": 0.3708759954493743,
+      "grad_norm": 0.9565249938146152,
+      "learning_rate": 2.963636363636364e-05,
+      "loss": 0.5158,
+      "step": 163
+    },
+    {
+      "epoch": 0.3731513083048919,
+      "grad_norm": 0.8580475728945989,
+      "learning_rate": 2.9818181818181823e-05,
+      "loss": 0.5131,
+      "step": 164
+    },
+    {
+      "epoch": 0.37542662116040953,
+      "grad_norm": 1.0407795891737115,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5551,
+      "step": 165
+    },
+    {
+      "epoch": 0.37770193401592717,
+      "grad_norm": 0.9461477598875092,
+      "learning_rate": 3.0181818181818182e-05,
+      "loss": 0.5128,
+      "step": 166
+    },
+    {
+      "epoch": 0.3799772468714448,
+      "grad_norm": 1.1750520423935205,
+      "learning_rate": 3.0363636363636364e-05,
+      "loss": 0.4988,
+      "step": 167
+    },
+    {
+      "epoch": 0.3822525597269625,
+      "grad_norm": 0.9619657535301448,
+      "learning_rate": 3.054545454545455e-05,
+      "loss": 0.5443,
+      "step": 168
+    },
+    {
+      "epoch": 0.3845278725824801,
+      "grad_norm": 0.8084615561755145,
+      "learning_rate": 3.072727272727273e-05,
+      "loss": 0.5538,
+      "step": 169
+    },
+    {
+      "epoch": 0.38680318543799774,
+      "grad_norm": 0.9166403915982619,
+      "learning_rate": 3.090909090909091e-05,
+      "loss": 0.5025,
+      "step": 170
+    },
+    {
+      "epoch": 0.3890784982935154,
+      "grad_norm": 0.706974515649925,
+      "learning_rate": 3.1090909090909094e-05,
+      "loss": 0.5192,
+      "step": 171
+    },
+    {
+      "epoch": 0.391353811149033,
+      "grad_norm": 0.744738485367252,
+      "learning_rate": 3.127272727272728e-05,
+      "loss": 0.5452,
+      "step": 172
+    },
+    {
+      "epoch": 0.39362912400455063,
+      "grad_norm": 1.185609845688004,
+      "learning_rate": 3.145454545454546e-05,
+      "loss": 0.5392,
+      "step": 173
+    },
+    {
+      "epoch": 0.39590443686006827,
+      "grad_norm": 0.6812824017739959,
+      "learning_rate": 3.1636363636363635e-05,
+      "loss": 0.538,
+      "step": 174
+    },
+    {
+      "epoch": 0.3981797497155859,
+      "grad_norm": 0.8921095831871404,
+      "learning_rate": 3.181818181818182e-05,
+      "loss": 0.5047,
+      "step": 175
+    },
+    {
+      "epoch": 0.4004550625711035,
+      "grad_norm": 1.0272787280734288,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.5,
+      "step": 176
+    },
+    {
+      "epoch": 0.40273037542662116,
+      "grad_norm": 0.7627496866984136,
+      "learning_rate": 3.2181818181818184e-05,
+      "loss": 0.5271,
+      "step": 177
+    },
+    {
+      "epoch": 0.4050056882821388,
+      "grad_norm": 0.783437681835504,
+      "learning_rate": 3.236363636363637e-05,
+      "loss": 0.5053,
+      "step": 178
+    },
+    {
+      "epoch": 0.4072810011376564,
+      "grad_norm": 0.9198593954163056,
+      "learning_rate": 3.254545454545455e-05,
+      "loss": 0.526,
+      "step": 179
+    },
+    {
+      "epoch": 0.40955631399317405,
+      "grad_norm": 0.990543263500737,
+      "learning_rate": 3.272727272727273e-05,
+      "loss": 0.5486,
+      "step": 180
+    },
+    {
+      "epoch": 0.4118316268486917,
+      "grad_norm": 0.7674137416061684,
+      "learning_rate": 3.290909090909091e-05,
+      "loss": 0.542,
+      "step": 181
+    },
+    {
+      "epoch": 0.4141069397042093,
+      "grad_norm": 0.7097774464178122,
+      "learning_rate": 3.3090909090909095e-05,
+      "loss": 0.5414,
+      "step": 182
+    },
+    {
+      "epoch": 0.41638225255972694,
+      "grad_norm": 0.7301988738783638,
+      "learning_rate": 3.327272727272727e-05,
+      "loss": 0.4837,
+      "step": 183
+    },
+    {
+      "epoch": 0.41865756541524457,
+      "grad_norm": 0.6586636192320678,
+      "learning_rate": 3.345454545454546e-05,
+      "loss": 0.5704,
+      "step": 184
+    },
+    {
+      "epoch": 0.42093287827076226,
+      "grad_norm": 0.7051322380639373,
+      "learning_rate": 3.363636363636364e-05,
+      "loss": 0.5087,
+      "step": 185
+    },
+    {
+      "epoch": 0.4232081911262799,
+      "grad_norm": 0.6614159680980038,
+      "learning_rate": 3.381818181818182e-05,
+      "loss": 0.5356,
+      "step": 186
+    },
+    {
+      "epoch": 0.4254835039817975,
+      "grad_norm": 0.6616643424557804,
+      "learning_rate": 3.4e-05,
+      "loss": 0.5236,
+      "step": 187
+    },
+    {
+      "epoch": 0.42775881683731515,
+      "grad_norm": 0.7001439685159849,
+      "learning_rate": 3.4181818181818185e-05,
+      "loss": 0.4978,
+      "step": 188
+    },
+    {
+      "epoch": 0.4300341296928328,
+      "grad_norm": 1.0004953566379289,
+      "learning_rate": 3.436363636363637e-05,
+      "loss": 0.5361,
+      "step": 189
+    },
+    {
+      "epoch": 0.4323094425483504,
+      "grad_norm": 0.9763247007839354,
+      "learning_rate": 3.454545454545455e-05,
+      "loss": 0.5032,
+      "step": 190
+    },
+    {
+      "epoch": 0.43458475540386804,
+      "grad_norm": 0.7796740868074972,
+      "learning_rate": 3.4727272727272726e-05,
+      "loss": 0.5303,
+      "step": 191
+    },
+    {
+      "epoch": 0.43686006825938567,
+      "grad_norm": 0.933786254284513,
+      "learning_rate": 3.490909090909091e-05,
+      "loss": 0.481,
+      "step": 192
+    },
+    {
+      "epoch": 0.4391353811149033,
+      "grad_norm": 1.1059297147921487,
+      "learning_rate": 3.5090909090909096e-05,
+      "loss": 0.5371,
+      "step": 193
+    },
+    {
+      "epoch": 0.44141069397042093,
+      "grad_norm": 0.9239737001173246,
+      "learning_rate": 3.5272727272727274e-05,
+      "loss": 0.5298,
+      "step": 194
+    },
+    {
+      "epoch": 0.44368600682593856,
+      "grad_norm": 0.7224979164330604,
+      "learning_rate": 3.545454545454546e-05,
+      "loss": 0.4723,
+      "step": 195
+    },
+    {
+      "epoch": 0.4459613196814562,
+      "grad_norm": 0.8163841859106732,
+      "learning_rate": 3.563636363636364e-05,
+      "loss": 0.5182,
+      "step": 196
+    },
+    {
+      "epoch": 0.4482366325369738,
+      "grad_norm": 0.7793873365990543,
+      "learning_rate": 3.581818181818182e-05,
+      "loss": 0.5175,
+      "step": 197
+    },
+    {
+      "epoch": 0.45051194539249145,
+      "grad_norm": 0.9106557759854739,
+      "learning_rate": 3.6e-05,
+      "loss": 0.5099,
+      "step": 198
+    },
+    {
+      "epoch": 0.4527872582480091,
+      "grad_norm": 0.8616846220679973,
+      "learning_rate": 3.6181818181818186e-05,
+      "loss": 0.5269,
+      "step": 199
+    },
+    {
+      "epoch": 0.4550625711035267,
+      "grad_norm": 0.9492125869073801,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 0.4923,
+      "step": 200
+    },
+    {
+      "epoch": 0.45733788395904434,
+      "grad_norm": 0.9382423826449089,
+      "learning_rate": 3.654545454545455e-05,
+      "loss": 0.5067,
+      "step": 201
+    },
+    {
+      "epoch": 0.459613196814562,
+      "grad_norm": 0.853602390271063,
+      "learning_rate": 3.6727272727272734e-05,
+      "loss": 0.5372,
+      "step": 202
+    },
+    {
+      "epoch": 0.46188850967007966,
+      "grad_norm": 0.8967497126984583,
+      "learning_rate": 3.690909090909091e-05,
+      "loss": 0.4968,
+      "step": 203
+    },
+    {
+      "epoch": 0.4641638225255973,
+      "grad_norm": 0.738267328983595,
+      "learning_rate": 3.709090909090909e-05,
+      "loss": 0.4992,
+      "step": 204
+    },
+    {
+      "epoch": 0.4664391353811149,
+      "grad_norm": 0.7188663644275407,
+      "learning_rate": 3.7272727272727276e-05,
+      "loss": 0.4772,
+      "step": 205
+    },
+    {
+      "epoch": 0.46871444823663255,
+      "grad_norm": 0.6515238182183732,
+      "learning_rate": 3.745454545454546e-05,
+      "loss": 0.5144,
+      "step": 206
+    },
+    {
+      "epoch": 0.4709897610921502,
+      "grad_norm": 0.8468875442175026,
+      "learning_rate": 3.763636363636364e-05,
+      "loss": 0.5449,
+      "step": 207
+    },
+    {
+      "epoch": 0.4732650739476678,
+      "grad_norm": 0.9072495234464792,
+      "learning_rate": 3.7818181818181824e-05,
+      "loss": 0.4696,
+      "step": 208
+    },
+    {
+      "epoch": 0.47554038680318544,
+      "grad_norm": 0.7954574695888201,
+      "learning_rate": 3.8e-05,
+      "loss": 0.4927,
+      "step": 209
+    },
+    {
+      "epoch": 0.4778156996587031,
+      "grad_norm": 0.8327389753927068,
+      "learning_rate": 3.818181818181819e-05,
+      "loss": 0.521,
+      "step": 210
+    },
+    {
+      "epoch": 0.4800910125142207,
+      "grad_norm": 0.722228238156621,
+      "learning_rate": 3.8363636363636365e-05,
+      "loss": 0.4953,
+      "step": 211
+    },
+    {
+      "epoch": 0.48236632536973834,
+      "grad_norm": 0.8705537435893771,
+      "learning_rate": 3.854545454545455e-05,
+      "loss": 0.5362,
+      "step": 212
+    },
+    {
+      "epoch": 0.48464163822525597,
+      "grad_norm": 0.8338745144603515,
+      "learning_rate": 3.872727272727273e-05,
+      "loss": 0.4901,
+      "step": 213
+    },
+    {
+      "epoch": 0.4869169510807736,
+      "grad_norm": 1.0078139079315223,
+      "learning_rate": 3.8909090909090914e-05,
+      "loss": 0.5112,
+      "step": 214
+    },
+    {
+      "epoch": 0.4891922639362912,
+      "grad_norm": 1.0811992758776074,
+      "learning_rate": 3.909090909090909e-05,
+      "loss": 0.4931,
+      "step": 215
+    },
+    {
+      "epoch": 0.49146757679180886,
+      "grad_norm": 0.8803716614397517,
+      "learning_rate": 3.927272727272728e-05,
+      "loss": 0.4868,
+      "step": 216
+    },
+    {
+      "epoch": 0.4937428896473265,
+      "grad_norm": 0.803266647344771,
+      "learning_rate": 3.9454545454545455e-05,
+      "loss": 0.4902,
+      "step": 217
+    },
+    {
+      "epoch": 0.4960182025028441,
+      "grad_norm": 1.00833874965682,
+      "learning_rate": 3.963636363636364e-05,
+      "loss": 0.464,
+      "step": 218
+    },
+    {
+      "epoch": 0.49829351535836175,
+      "grad_norm": 0.96083659528256,
+      "learning_rate": 3.9818181818181825e-05,
+      "loss": 0.4919,
+      "step": 219
+    },
+    {
+      "epoch": 0.5005688282138794,
+      "grad_norm": 0.7790449810495192,
+      "learning_rate": 4e-05,
+      "loss": 0.4965,
+      "step": 220
+    },
+    {
+      "epoch": 0.502844141069397,
+      "grad_norm": 0.8905458268931085,
+      "learning_rate": 3.9999974697382296e-05,
+      "loss": 0.53,
+      "step": 221
+    },
+    {
+      "epoch": 0.5051194539249146,
+      "grad_norm": 0.7625724515485385,
+      "learning_rate": 3.9999898789593194e-05,
+      "loss": 0.5027,
+      "step": 222
+    },
+    {
+      "epoch": 0.5073947667804323,
+      "grad_norm": 1.0155453849101541,
+      "learning_rate": 3.999977227682476e-05,
+      "loss": 0.508,
+      "step": 223
+    },
+    {
+      "epoch": 0.5096700796359499,
+      "grad_norm": 0.6681764192465899,
+      "learning_rate": 3.99995951593971e-05,
+      "loss": 0.5049,
+      "step": 224
+    },
+    {
+      "epoch": 0.5119453924914675,
+      "grad_norm": 0.9099406965338225,
+      "learning_rate": 3.999936743775839e-05,
+      "loss": 0.5252,
+      "step": 225
+    },
+    {
+      "epoch": 0.5142207053469852,
+      "grad_norm": 0.6923524004307658,
+      "learning_rate": 3.999908911248481e-05,
+      "loss": 0.4751,
+      "step": 226
+    },
+    {
+      "epoch": 0.5164960182025028,
+      "grad_norm": 0.881151320708067,
+      "learning_rate": 3.999876018428059e-05,
+      "loss": 0.5008,
+      "step": 227
+    },
+    {
+      "epoch": 0.5187713310580204,
+      "grad_norm": 0.8293172103871226,
+      "learning_rate": 3.999838065397801e-05,
+      "loss": 0.5198,
+      "step": 228
+    },
+    {
+      "epoch": 0.5210466439135382,
+      "grad_norm": 0.7096703324834885,
+      "learning_rate": 3.9997950522537385e-05,
+      "loss": 0.5272,
+      "step": 229
+    },
+    {
+      "epoch": 0.5233219567690558,
+      "grad_norm": 0.7127876942337593,
+      "learning_rate": 3.999746979104705e-05,
+      "loss": 0.5225,
+      "step": 230
+    },
+    {
+      "epoch": 0.5255972696245734,
+      "grad_norm": 0.7792466358382388,
+      "learning_rate": 3.999693846072339e-05,
+      "loss": 0.5182,
+      "step": 231
+    },
+    {
+      "epoch": 0.5278725824800911,
+      "grad_norm": 0.779669905296092,
+      "learning_rate": 3.9996356532910814e-05,
+      "loss": 0.4981,
+      "step": 232
+    },
+    {
+      "epoch": 0.5301478953356087,
+      "grad_norm": 0.800598136561164,
+      "learning_rate": 3.9995724009081745e-05,
+      "loss": 0.5603,
+      "step": 233
+    },
+    {
+      "epoch": 0.5324232081911263,
+      "grad_norm": 0.9619367505895898,
+      "learning_rate": 3.9995040890836635e-05,
+      "loss": 0.5583,
+      "step": 234
+    },
+    {
+      "epoch": 0.534698521046644,
+      "grad_norm": 0.9453874433476237,
+      "learning_rate": 3.999430717990395e-05,
+      "loss": 0.5082,
+      "step": 235
+    },
+    {
+      "epoch": 0.5369738339021616,
+      "grad_norm": 0.9382827233401629,
+      "learning_rate": 3.9993522878140174e-05,
+      "loss": 0.4989,
+      "step": 236
+    },
+    {
+      "epoch": 0.5392491467576792,
+      "grad_norm": 0.78252208920523,
+      "learning_rate": 3.999268798752979e-05,
+      "loss": 0.5234,
+      "step": 237
+    },
+    {
+      "epoch": 0.5415244596131968,
+      "grad_norm": 0.8195622765128195,
+      "learning_rate": 3.9991802510185296e-05,
+      "loss": 0.556,
+      "step": 238
+    },
+    {
+      "epoch": 0.5437997724687145,
+      "grad_norm": 0.8208206901010974,
+      "learning_rate": 3.999086644834719e-05,
+      "loss": 0.4887,
+      "step": 239
+    },
+    {
+      "epoch": 0.5460750853242321,
+      "grad_norm": 0.7551681915194937,
+      "learning_rate": 3.998987980438393e-05,
+      "loss": 0.4936,
+      "step": 240
+    },
+    {
+      "epoch": 0.5483503981797497,
+      "grad_norm": 0.7495264868649975,
+      "learning_rate": 3.998884258079199e-05,
+      "loss": 0.4627,
+      "step": 241
+    },
+    {
+      "epoch": 0.5506257110352674,
+      "grad_norm": 0.8060313156395092,
+      "learning_rate": 3.998775478019584e-05,
+      "loss": 0.5042,
+      "step": 242
+    },
+    {
+      "epoch": 0.552901023890785,
+      "grad_norm": 0.7867672669067826,
+      "learning_rate": 3.998661640534787e-05,
+      "loss": 0.5115,
+      "step": 243
+    },
+    {
+      "epoch": 0.5551763367463026,
+      "grad_norm": 0.7557722856697018,
+      "learning_rate": 3.9985427459128496e-05,
+      "loss": 0.5077,
+      "step": 244
+    },
+    {
+      "epoch": 0.5574516496018203,
+      "grad_norm": 0.7390570724379956,
+      "learning_rate": 3.998418794454604e-05,
+      "loss": 0.5583,
+      "step": 245
+    },
+    {
+      "epoch": 0.5597269624573379,
+      "grad_norm": 0.8305288417970496,
+      "learning_rate": 3.998289786473681e-05,
+      "loss": 0.5342,
+      "step": 246
+    },
+    {
+      "epoch": 0.5620022753128555,
+      "grad_norm": 0.7858042344970474,
+      "learning_rate": 3.998155722296504e-05,
+      "loss": 0.5082,
+      "step": 247
+    },
+    {
+      "epoch": 0.5642775881683731,
+      "grad_norm": 0.7423633752656493,
+      "learning_rate": 3.99801660226229e-05,
+      "loss": 0.511,
+      "step": 248
+    },
+    {
+      "epoch": 0.5665529010238908,
+      "grad_norm": 0.9185613255353122,
+      "learning_rate": 3.9978724267230495e-05,
+      "loss": 0.5457,
+      "step": 249
+    },
+    {
+      "epoch": 0.5688282138794084,
+      "grad_norm": 1.0074100879377919,
+      "learning_rate": 3.997723196043585e-05,
+      "loss": 0.5201,
+      "step": 250
+    },
+    {
+      "epoch": 0.571103526734926,
+      "grad_norm": 0.8713246831114032,
+      "learning_rate": 3.997568910601489e-05,
+      "loss": 0.541,
+      "step": 251
+    },
+    {
+      "epoch": 0.5733788395904437,
+      "grad_norm": 1.0445942432402573,
+      "learning_rate": 3.997409570787144e-05,
+      "loss": 0.5242,
+      "step": 252
+    },
+    {
+      "epoch": 0.5756541524459613,
+      "grad_norm": 0.7312036738233352,
+      "learning_rate": 3.997245177003721e-05,
+      "loss": 0.4896,
+      "step": 253
+    },
+    {
+      "epoch": 0.5779294653014789,
+      "grad_norm": 0.8072632910129269,
+      "learning_rate": 3.9970757296671795e-05,
+      "loss": 0.4972,
+      "step": 254
+    },
+    {
+      "epoch": 0.5802047781569966,
+      "grad_norm": 1.0752550763412985,
+      "learning_rate": 3.9969012292062655e-05,
+      "loss": 0.4786,
+      "step": 255
+    },
+    {
+      "epoch": 0.5824800910125142,
+      "grad_norm": 0.7203656254008417,
+      "learning_rate": 3.9967216760625113e-05,
+      "loss": 0.5189,
+      "step": 256
+    },
+    {
+      "epoch": 0.5847554038680318,
+      "grad_norm": 0.6613667360116985,
+      "learning_rate": 3.9965370706902324e-05,
+      "loss": 0.4769,
+      "step": 257
+    },
+    {
+      "epoch": 0.5870307167235495,
+      "grad_norm": 0.7822122386302286,
+      "learning_rate": 3.99634741355653e-05,
+      "loss": 0.5005,
+      "step": 258
+    },
+    {
+      "epoch": 0.5893060295790671,
+      "grad_norm": 0.9385427267856482,
+      "learning_rate": 3.9961527051412854e-05,
+      "loss": 0.5257,
+      "step": 259
+    },
+    {
+      "epoch": 0.5915813424345847,
+      "grad_norm": 0.6735881526097904,
+      "learning_rate": 3.9959529459371624e-05,
+      "loss": 0.5049,
+      "step": 260
+    },
+    {
+      "epoch": 0.5938566552901023,
+      "grad_norm": 0.8457373957055475,
+      "learning_rate": 3.9957481364496044e-05,
+      "loss": 0.5102,
+      "step": 261
+    },
+    {
+      "epoch": 0.59613196814562,
+      "grad_norm": 1.0973039674832767,
+      "learning_rate": 3.9955382771968316e-05,
+      "loss": 0.5099,
+      "step": 262
+    },
+    {
+      "epoch": 0.5984072810011376,
+      "grad_norm": 0.8203260675634886,
+      "learning_rate": 3.9953233687098435e-05,
+      "loss": 0.5309,
+      "step": 263
+    },
+    {
+      "epoch": 0.6006825938566553,
+      "grad_norm": 0.7874965178599942,
+      "learning_rate": 3.9951034115324156e-05,
+      "loss": 0.568,
+      "step": 264
+    },
+    {
+      "epoch": 0.602957906712173,
+      "grad_norm": 1.2512774514947018,
+      "learning_rate": 3.994878406221097e-05,
+      "loss": 0.5319,
+      "step": 265
+    },
+    {
+      "epoch": 0.6052332195676906,
+      "grad_norm": 0.8249352132277762,
+      "learning_rate": 3.994648353345208e-05,
+      "loss": 0.5227,
+      "step": 266
+    },
+    {
+      "epoch": 0.6075085324232082,
+      "grad_norm": 1.034821442073306,
+      "learning_rate": 3.994413253486846e-05,
+      "loss": 0.4714,
+      "step": 267
+    },
+    {
+      "epoch": 0.6097838452787259,
+      "grad_norm": 1.3292869026587193,
+      "learning_rate": 3.994173107240872e-05,
+      "loss": 0.4962,
+      "step": 268
+    },
+    {
+      "epoch": 0.6120591581342435,
+      "grad_norm": 0.8212065544130984,
+      "learning_rate": 3.9939279152149216e-05,
+      "loss": 0.5242,
+      "step": 269
+    },
+    {
+      "epoch": 0.6143344709897611,
+      "grad_norm": 1.5092198444566256,
+      "learning_rate": 3.993677678029392e-05,
+      "loss": 0.5343,
+      "step": 270
+    },
+    {
+      "epoch": 0.6166097838452788,
+      "grad_norm": 0.9359978887113274,
+      "learning_rate": 3.993422396317451e-05,
+      "loss": 0.5203,
+      "step": 271
+    },
+    {
+      "epoch": 0.6188850967007964,
+      "grad_norm": 1.4237602720094844,
+      "learning_rate": 3.993162070725027e-05,
+      "loss": 0.5501,
+      "step": 272
+    },
+    {
+      "epoch": 0.621160409556314,
+      "grad_norm": 0.9726057260183041,
+      "learning_rate": 3.992896701910813e-05,
+      "loss": 0.5163,
+      "step": 273
+    },
+    {
+      "epoch": 0.6234357224118316,
+      "grad_norm": 1.4550093737222574,
+      "learning_rate": 3.99262629054626e-05,
+      "loss": 0.5206,
+      "step": 274
+    },
+    {
+      "epoch": 0.6257110352673493,
+      "grad_norm": 0.9555548052797809,
+      "learning_rate": 3.992350837315581e-05,
+      "loss": 0.5023,
+      "step": 275
+    },
+    {
+      "epoch": 0.6279863481228669,
+      "grad_norm": 1.369245874866738,
+      "learning_rate": 3.9920703429157436e-05,
+      "loss": 0.4886,
+      "step": 276
+    },
+    {
+      "epoch": 0.6302616609783845,
+      "grad_norm": 1.1851345853900797,
+      "learning_rate": 3.991784808056473e-05,
+      "loss": 0.5148,
+      "step": 277
+    },
+    {
+      "epoch": 0.6325369738339022,
+      "grad_norm": 1.179793591277137,
+      "learning_rate": 3.9914942334602464e-05,
+      "loss": 0.5318,
+      "step": 278
+    },
+    {
+      "epoch": 0.6348122866894198,
+      "grad_norm": 1.199565301581228,
+      "learning_rate": 3.991198619862294e-05,
+      "loss": 0.5194,
+      "step": 279
+    },
+    {
+      "epoch": 0.6370875995449374,
+      "grad_norm": 0.9094867274398518,
+      "learning_rate": 3.990897968010596e-05,
+      "loss": 0.5411,
+      "step": 280
+    },
+    {
+      "epoch": 0.6393629124004551,
+      "grad_norm": 1.1094913036582985,
+      "learning_rate": 3.99059227866588e-05,
+      "loss": 0.5076,
+      "step": 281
+    },
+    {
+      "epoch": 0.6416382252559727,
+      "grad_norm": 0.826685561156455,
+      "learning_rate": 3.9902815526016196e-05,
+      "loss": 0.5031,
+      "step": 282
+    },
+    {
+      "epoch": 0.6439135381114903,
+      "grad_norm": 0.9716995105435796,
+      "learning_rate": 3.989965790604033e-05,
+      "loss": 0.4603,
+      "step": 283
+    },
+    {
+      "epoch": 0.646188850967008,
+      "grad_norm": 0.900885856862581,
+      "learning_rate": 3.9896449934720814e-05,
+      "loss": 0.5107,
+      "step": 284
+    },
+    {
+      "epoch": 0.6484641638225256,
+      "grad_norm": 0.8778940794295121,
+      "learning_rate": 3.989319162017465e-05,
+      "loss": 0.5502,
+      "step": 285
+    },
+    {
+      "epoch": 0.6507394766780432,
+      "grad_norm": 0.9622322529521133,
+      "learning_rate": 3.988988297064623e-05,
+      "loss": 0.5083,
+      "step": 286
+    },
+    {
+      "epoch": 0.6530147895335608,
+      "grad_norm": 0.7756328265585737,
+      "learning_rate": 3.98865239945073e-05,
+      "loss": 0.5028,
+      "step": 287
+    },
+    {
+      "epoch": 0.6552901023890785,
+      "grad_norm": 0.6822282919532825,
+      "learning_rate": 3.988311470025695e-05,
+      "loss": 0.5036,
+      "step": 288
+    },
+    {
+      "epoch": 0.6575654152445961,
+      "grad_norm": 0.8959561359324182,
+      "learning_rate": 3.987965509652159e-05,
+      "loss": 0.4908,
+      "step": 289
+    },
+    {
+      "epoch": 0.6598407281001137,
+      "grad_norm": 0.6601859038879857,
+      "learning_rate": 3.987614519205493e-05,
+      "loss": 0.478,
+      "step": 290
+    },
+    {
+      "epoch": 0.6621160409556314,
+      "grad_norm": 0.9825085087311318,
+      "learning_rate": 3.987258499573792e-05,
+      "loss": 0.5338,
+      "step": 291
+    },
+    {
+      "epoch": 0.664391353811149,
+      "grad_norm": 1.022289539818651,
+      "learning_rate": 3.986897451657882e-05,
+      "loss": 0.5124,
+      "step": 292
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.659127791072242,
+      "learning_rate": 3.986531376371307e-05,
+      "loss": 0.4858,
+      "step": 293
+    },
+    {
+      "epoch": 0.6689419795221843,
+      "grad_norm": 0.9679175230064438,
+      "learning_rate": 3.9861602746403336e-05,
+      "loss": 0.5302,
+      "step": 294
+    },
+    {
+      "epoch": 0.6712172923777019,
+      "grad_norm": 0.7180146822347161,
+      "learning_rate": 3.985784147403947e-05,
+      "loss": 0.4594,
+      "step": 295
+    },
+    {
+      "epoch": 0.6734926052332195,
+      "grad_norm": 0.8559576954351592,
+      "learning_rate": 3.985402995613847e-05,
+      "loss": 0.4986,
+      "step": 296
+    },
+    {
+      "epoch": 0.6757679180887372,
+      "grad_norm": 0.9453576621806624,
+      "learning_rate": 3.985016820234447e-05,
+      "loss": 0.4866,
+      "step": 297
+    },
+    {
+      "epoch": 0.6780432309442548,
+      "grad_norm": 0.857442638332296,
+      "learning_rate": 3.984625622242872e-05,
+      "loss": 0.4996,
+      "step": 298
+    },
+    {
+      "epoch": 0.6803185437997725,
+      "grad_norm": 0.7396924998396898,
+      "learning_rate": 3.9842294026289565e-05,
+      "loss": 0.4805,
+      "step": 299
+    },
+    {
+      "epoch": 0.6825938566552902,
+      "grad_norm": 1.0151633646700187,
+      "learning_rate": 3.983828162395238e-05,
+      "loss": 0.4821,
+      "step": 300
+    },
+    {
+      "epoch": 0.6848691695108078,
+      "grad_norm": 0.9979683962499621,
+      "learning_rate": 3.98342190255696e-05,
+      "loss": 0.527,
+      "step": 301
+    },
+    {
+      "epoch": 0.6871444823663254,
+      "grad_norm": 0.899885752733397,
+      "learning_rate": 3.9830106241420666e-05,
+      "loss": 0.5062,
+      "step": 302
+    },
+    {
+      "epoch": 0.689419795221843,
+      "grad_norm": 0.8778242686510422,
+      "learning_rate": 3.9825943281912005e-05,
+      "loss": 0.5062,
+      "step": 303
+    },
+    {
+      "epoch": 0.6916951080773607,
+      "grad_norm": 0.637806414801961,
+      "learning_rate": 3.9821730157576975e-05,
+      "loss": 0.4887,
+      "step": 304
+    },
+    {
+      "epoch": 0.6939704209328783,
+      "grad_norm": 0.7447436139595571,
+      "learning_rate": 3.98174668790759e-05,
+      "loss": 0.5404,
+      "step": 305
+    },
+    {
+      "epoch": 0.6962457337883959,
+      "grad_norm": 0.5406588274310511,
+      "learning_rate": 3.9813153457195986e-05,
+      "loss": 0.4759,
+      "step": 306
+    },
+    {
+      "epoch": 0.6985210466439136,
+      "grad_norm": 0.776438137829388,
+      "learning_rate": 3.980878990285132e-05,
+      "loss": 0.5379,
+      "step": 307
+    },
+    {
+      "epoch": 0.7007963594994312,
+      "grad_norm": 0.6179783876468353,
+      "learning_rate": 3.9804376227082834e-05,
+      "loss": 0.5152,
+      "step": 308
+    },
+    {
+      "epoch": 0.7030716723549488,
+      "grad_norm": 0.6195544387076716,
+      "learning_rate": 3.9799912441058286e-05,
+      "loss": 0.5232,
+      "step": 309
+    },
+    {
+      "epoch": 0.7053469852104665,
+      "grad_norm": 0.6482103136431522,
+      "learning_rate": 3.979539855607222e-05,
+      "loss": 0.5045,
+      "step": 310
+    },
+    {
+      "epoch": 0.7076222980659841,
+      "grad_norm": 0.7357097196921872,
+      "learning_rate": 3.9790834583545946e-05,
+      "loss": 0.4881,
+      "step": 311
+    },
+    {
+      "epoch": 0.7098976109215017,
+      "grad_norm": 0.6461268041598386,
+      "learning_rate": 3.978622053502751e-05,
+      "loss": 0.4773,
+      "step": 312
+    },
+    {
+      "epoch": 0.7121729237770194,
+      "grad_norm": 0.6986308188983689,
+      "learning_rate": 3.978155642219167e-05,
+      "loss": 0.5217,
+      "step": 313
+    },
+    {
+      "epoch": 0.714448236632537,
+      "grad_norm": 0.568430245151543,
+      "learning_rate": 3.977684225683984e-05,
+      "loss": 0.5206,
+      "step": 314
+    },
+    {
+      "epoch": 0.7167235494880546,
+      "grad_norm": 0.736787180700768,
+      "learning_rate": 3.9772078050900105e-05,
+      "loss": 0.5388,
+      "step": 315
+    },
+    {
+      "epoch": 0.7189988623435722,
+      "grad_norm": 0.6623240192179631,
+      "learning_rate": 3.9767263816427146e-05,
+      "loss": 0.4748,
+      "step": 316
+    },
+    {
+      "epoch": 0.7212741751990899,
+      "grad_norm": 0.7564926165936612,
+      "learning_rate": 3.9762399565602233e-05,
+      "loss": 0.4889,
+      "step": 317
+    },
+    {
+      "epoch": 0.7235494880546075,
+      "grad_norm": 0.6534761363373418,
+      "learning_rate": 3.97574853107332e-05,
+      "loss": 0.5083,
+      "step": 318
+    },
+    {
+      "epoch": 0.7258248009101251,
+      "grad_norm": 0.7537961040113483,
+      "learning_rate": 3.97525210642544e-05,
+      "loss": 0.4951,
+      "step": 319
+    },
+    {
+      "epoch": 0.7281001137656428,
+      "grad_norm": 0.7578706031427168,
+      "learning_rate": 3.974750683872667e-05,
+      "loss": 0.4905,
+      "step": 320
+    },
+    {
+      "epoch": 0.7303754266211604,
+      "grad_norm": 0.7260807303288004,
+      "learning_rate": 3.9742442646837316e-05,
+      "loss": 0.5322,
+      "step": 321
+    },
+    {
+      "epoch": 0.732650739476678,
+      "grad_norm": 0.9173353438904912,
+      "learning_rate": 3.973732850140007e-05,
+      "loss": 0.52,
+      "step": 322
+    },
+    {
+      "epoch": 0.7349260523321957,
+      "grad_norm": 0.6310646643036116,
+      "learning_rate": 3.973216441535506e-05,
+      "loss": 0.532,
+      "step": 323
+    },
+    {
+      "epoch": 0.7372013651877133,
+      "grad_norm": 0.7197572927145501,
+      "learning_rate": 3.972695040176877e-05,
+      "loss": 0.5008,
+      "step": 324
+    },
+    {
+      "epoch": 0.7394766780432309,
+      "grad_norm": 0.7374860781338815,
+      "learning_rate": 3.972168647383402e-05,
+      "loss": 0.5578,
+      "step": 325
+    },
+    {
+      "epoch": 0.7417519908987485,
+      "grad_norm": 0.6128071895701607,
+      "learning_rate": 3.971637264486993e-05,
+      "loss": 0.5155,
+      "step": 326
+    },
+    {
+      "epoch": 0.7440273037542662,
+      "grad_norm": 0.5637948735002289,
+      "learning_rate": 3.971100892832188e-05,
+      "loss": 0.4719,
+      "step": 327
+    },
+    {
+      "epoch": 0.7463026166097838,
+      "grad_norm": 0.5424079474042907,
+      "learning_rate": 3.970559533776147e-05,
+      "loss": 0.4922,
+      "step": 328
+    },
+    {
+      "epoch": 0.7485779294653014,
+      "grad_norm": 0.6317180275054568,
+      "learning_rate": 3.9700131886886506e-05,
+      "loss": 0.5046,
+      "step": 329
+    },
+    {
+      "epoch": 0.7508532423208191,
+      "grad_norm": 0.7049941743535045,
+      "learning_rate": 3.9694618589520945e-05,
+      "loss": 0.5153,
+      "step": 330
+    },
+    {
+      "epoch": 0.7531285551763367,
+      "grad_norm": 0.7526126287558034,
+      "learning_rate": 3.968905545961487e-05,
+      "loss": 0.4978,
+      "step": 331
+    },
+    {
+      "epoch": 0.7554038680318543,
+      "grad_norm": 0.6914871893079282,
+      "learning_rate": 3.968344251124447e-05,
+      "loss": 0.5221,
+      "step": 332
+    },
+    {
+      "epoch": 0.757679180887372,
+      "grad_norm": 0.8497636918088499,
+      "learning_rate": 3.967777975861196e-05,
+      "loss": 0.495,
+      "step": 333
+    },
+    {
+      "epoch": 0.7599544937428896,
+      "grad_norm": 0.6810417416080219,
+      "learning_rate": 3.96720672160456e-05,
+      "loss": 0.538,
+      "step": 334
+    },
+    {
+      "epoch": 0.7622298065984073,
+      "grad_norm": 0.8161126575125076,
+      "learning_rate": 3.966630489799959e-05,
+      "loss": 0.5085,
+      "step": 335
+    },
+    {
+      "epoch": 0.764505119453925,
+      "grad_norm": 0.7738738143828285,
+      "learning_rate": 3.966049281905414e-05,
+      "loss": 0.4873,
+      "step": 336
+    },
+    {
+      "epoch": 0.7667804323094426,
+      "grad_norm": 0.8764493288845413,
+      "learning_rate": 3.9654630993915305e-05,
+      "loss": 0.4971,
+      "step": 337
+    },
+    {
+      "epoch": 0.7690557451649602,
+      "grad_norm": 0.7805156720414773,
+      "learning_rate": 3.964871943741504e-05,
+      "loss": 0.5122,
+      "step": 338
+    },
+    {
+      "epoch": 0.7713310580204779,
+      "grad_norm": 0.7201535220586814,
+      "learning_rate": 3.964275816451115e-05,
+      "loss": 0.5161,
+      "step": 339
+    },
+    {
+      "epoch": 0.7736063708759955,
+      "grad_norm": 0.7396481006919741,
+      "learning_rate": 3.96367471902872e-05,
+      "loss": 0.5104,
+      "step": 340
+    },
+    {
+      "epoch": 0.7758816837315131,
+      "grad_norm": 0.7123311898098197,
+      "learning_rate": 3.963068652995252e-05,
+      "loss": 0.5148,
+      "step": 341
+    },
+    {
+      "epoch": 0.7781569965870307,
+      "grad_norm": 0.6652575492753041,
+      "learning_rate": 3.962457619884218e-05,
+      "loss": 0.471,
+      "step": 342
+    },
+    {
+      "epoch": 0.7804323094425484,
+      "grad_norm": 0.8105783570316155,
+      "learning_rate": 3.961841621241692e-05,
+      "loss": 0.5463,
+      "step": 343
+    },
+    {
+      "epoch": 0.782707622298066,
+      "grad_norm": 0.606651260022466,
+      "learning_rate": 3.9612206586263104e-05,
+      "loss": 0.5066,
+      "step": 344
+    },
+    {
+      "epoch": 0.7849829351535836,
+      "grad_norm": 1.0361515753308552,
+      "learning_rate": 3.960594733609273e-05,
+      "loss": 0.4976,
+      "step": 345
+    },
+    {
+      "epoch": 0.7872582480091013,
+      "grad_norm": 0.6631800027893998,
+      "learning_rate": 3.959963847774332e-05,
+      "loss": 0.4864,
+      "step": 346
+    },
+    {
+      "epoch": 0.7895335608646189,
+      "grad_norm": 0.748533983699474,
+      "learning_rate": 3.959328002717795e-05,
+      "loss": 0.4753,
+      "step": 347
+    },
+    {
+      "epoch": 0.7918088737201365,
+      "grad_norm": 0.5971622138969933,
+      "learning_rate": 3.958687200048516e-05,
+      "loss": 0.4916,
+      "step": 348
+    },
+    {
+      "epoch": 0.7940841865756542,
+      "grad_norm": 0.660067189965713,
+      "learning_rate": 3.958041441387894e-05,
+      "loss": 0.5161,
+      "step": 349
+    },
+    {
+      "epoch": 0.7963594994311718,
+      "grad_norm": 0.5986830293577159,
+      "learning_rate": 3.957390728369867e-05,
+      "loss": 0.4895,
+      "step": 350
+    },
+    {
+      "epoch": 0.7986348122866894,
+      "grad_norm": 0.6004693311643707,
+      "learning_rate": 3.9567350626409094e-05,
+      "loss": 0.4915,
+      "step": 351
+    },
+    {
+      "epoch": 0.800910125142207,
+      "grad_norm": 0.6276703295004309,
+      "learning_rate": 3.956074445860027e-05,
+      "loss": 0.487,
+      "step": 352
+    },
+    {
+      "epoch": 0.8031854379977247,
+      "grad_norm": 0.6103277468343838,
+      "learning_rate": 3.955408879698753e-05,
+      "loss": 0.507,
+      "step": 353
+    },
+    {
+      "epoch": 0.8054607508532423,
+      "grad_norm": 0.6799400586797437,
+      "learning_rate": 3.954738365841144e-05,
+      "loss": 0.5003,
+      "step": 354
+    },
+    {
+      "epoch": 0.8077360637087599,
+      "grad_norm": 0.6505285949994575,
+      "learning_rate": 3.9540629059837767e-05,
+      "loss": 0.4729,
+      "step": 355
+    },
+    {
+      "epoch": 0.8100113765642776,
+      "grad_norm": 0.6586392259013478,
+      "learning_rate": 3.95338250183574e-05,
+      "loss": 0.5229,
+      "step": 356
+    },
+    {
+      "epoch": 0.8122866894197952,
+      "grad_norm": 0.7567869042648286,
+      "learning_rate": 3.952697155118635e-05,
+      "loss": 0.5088,
+      "step": 357
+    },
+    {
+      "epoch": 0.8145620022753128,
+      "grad_norm": 0.6347889342427536,
+      "learning_rate": 3.952006867566569e-05,
+      "loss": 0.4869,
+      "step": 358
+    },
+    {
+      "epoch": 0.8168373151308305,
+      "grad_norm": 0.6893249132631182,
+      "learning_rate": 3.951311640926148e-05,
+      "loss": 0.5289,
+      "step": 359
+    },
+    {
+      "epoch": 0.8191126279863481,
+      "grad_norm": 0.7138269635664551,
+      "learning_rate": 3.95061147695648e-05,
+      "loss": 0.4777,
+      "step": 360
+    },
+    {
+      "epoch": 0.8213879408418657,
+      "grad_norm": 0.7649893274903273,
+      "learning_rate": 3.949906377429162e-05,
+      "loss": 0.5008,
+      "step": 361
+    },
+    {
+      "epoch": 0.8236632536973834,
+      "grad_norm": 0.7777457016748831,
+      "learning_rate": 3.94919634412828e-05,
+      "loss": 0.5241,
+      "step": 362
+    },
+    {
+      "epoch": 0.825938566552901,
+      "grad_norm": 0.7206642217902817,
+      "learning_rate": 3.948481378850405e-05,
+      "loss": 0.4912,
+      "step": 363
+    },
+    {
+      "epoch": 0.8282138794084186,
+      "grad_norm": 1.1425011384581647,
+      "learning_rate": 3.947761483404585e-05,
+      "loss": 0.486,
+      "step": 364
+    },
+    {
+      "epoch": 0.8304891922639362,
+      "grad_norm": 0.7431045570776554,
+      "learning_rate": 3.947036659612345e-05,
+      "loss": 0.4874,
+      "step": 365
+    },
+    {
+      "epoch": 0.8327645051194539,
+      "grad_norm": 0.9260901341723518,
+      "learning_rate": 3.94630690930768e-05,
+      "loss": 0.5323,
+      "step": 366
+    },
+    {
+      "epoch": 0.8350398179749715,
+      "grad_norm": 0.6461587891988468,
+      "learning_rate": 3.945572234337046e-05,
+      "loss": 0.4979,
+      "step": 367
+    },
+    {
+      "epoch": 0.8373151308304891,
+      "grad_norm": 1.2599484128221388,
+      "learning_rate": 3.944832636559366e-05,
+      "loss": 0.551,
+      "step": 368
+    },
+    {
+      "epoch": 0.8395904436860068,
+      "grad_norm": 0.9305399142519398,
+      "learning_rate": 3.944088117846015e-05,
+      "loss": 0.5008,
+      "step": 369
+    },
+    {
+      "epoch": 0.8418657565415245,
+      "grad_norm": 1.121526764606612,
+      "learning_rate": 3.94333868008082e-05,
+      "loss": 0.5148,
+      "step": 370
+    },
+    {
+      "epoch": 0.8441410693970421,
+      "grad_norm": 0.9052152771663639,
+      "learning_rate": 3.9425843251600555e-05,
+      "loss": 0.491,
+      "step": 371
+    },
+    {
+      "epoch": 0.8464163822525598,
+      "grad_norm": 1.0494031846799934,
+      "learning_rate": 3.9418250549924355e-05,
+      "loss": 0.5177,
+      "step": 372
+    },
+    {
+      "epoch": 0.8486916951080774,
+      "grad_norm": 0.9888631313574925,
+      "learning_rate": 3.9410608714991136e-05,
+      "loss": 0.5107,
+      "step": 373
+    },
+    {
+      "epoch": 0.850967007963595,
+      "grad_norm": 1.1055928703416604,
+      "learning_rate": 3.9402917766136735e-05,
+      "loss": 0.5397,
+      "step": 374
+    },
+    {
+      "epoch": 0.8532423208191127,
+      "grad_norm": 1.115912993227309,
+      "learning_rate": 3.939517772282127e-05,
+      "loss": 0.4829,
+      "step": 375
+    },
+    {
+      "epoch": 0.8555176336746303,
+      "grad_norm": 0.9099392894966472,
+      "learning_rate": 3.938738860462907e-05,
+      "loss": 0.5003,
+      "step": 376
+    },
+    {
+      "epoch": 0.8577929465301479,
+      "grad_norm": 1.0706866454722113,
+      "learning_rate": 3.937955043126864e-05,
+      "loss": 0.5284,
+      "step": 377
+    },
+    {
+      "epoch": 0.8600682593856656,
+      "grad_norm": 0.909125878536022,
+      "learning_rate": 3.9371663222572625e-05,
+      "loss": 0.5121,
+      "step": 378
+    },
+    {
+      "epoch": 0.8623435722411832,
+      "grad_norm": 0.8423121951286772,
+      "learning_rate": 3.936372699849772e-05,
+      "loss": 0.4996,
+      "step": 379
+    },
+    {
+      "epoch": 0.8646188850967008,
+      "grad_norm": 0.8332263490241194,
+      "learning_rate": 3.935574177912465e-05,
+      "loss": 0.5245,
+      "step": 380
+    },
+    {
+      "epoch": 0.8668941979522184,
+      "grad_norm": 0.7530064300634705,
+      "learning_rate": 3.93477075846581e-05,
+      "loss": 0.5342,
+      "step": 381
+    },
+    {
+      "epoch": 0.8691695108077361,
+      "grad_norm": 0.819669590890119,
+      "learning_rate": 3.93396244354267e-05,
+      "loss": 0.4866,
+      "step": 382
+    },
+    {
+      "epoch": 0.8714448236632537,
+      "grad_norm": 0.6333438826999286,
+      "learning_rate": 3.933149235188294e-05,
+      "loss": 0.5043,
+      "step": 383
+    },
+    {
+      "epoch": 0.8737201365187713,
+      "grad_norm": 0.7459251590464333,
+      "learning_rate": 3.9323311354603086e-05,
+      "loss": 0.4805,
+      "step": 384
+    },
+    {
+      "epoch": 0.875995449374289,
+      "grad_norm": 0.6564470081159227,
+      "learning_rate": 3.931508146428724e-05,
+      "loss": 0.5067,
+      "step": 385
+    },
+    {
+      "epoch": 0.8782707622298066,
+      "grad_norm": 0.6998497277576152,
+      "learning_rate": 3.930680270175915e-05,
+      "loss": 0.5271,
+      "step": 386
+    },
+    {
+      "epoch": 0.8805460750853242,
+      "grad_norm": 0.5812403223556097,
+      "learning_rate": 3.929847508796628e-05,
+      "loss": 0.5124,
+      "step": 387
+    },
+    {
+      "epoch": 0.8828213879408419,
+      "grad_norm": 0.7056380734319831,
+      "learning_rate": 3.9290098643979654e-05,
+      "loss": 0.5205,
+      "step": 388
+    },
+    {
+      "epoch": 0.8850967007963595,
+      "grad_norm": 0.6313474071640174,
+      "learning_rate": 3.928167339099387e-05,
+      "loss": 0.503,
+      "step": 389
+    },
+    {
+      "epoch": 0.8873720136518771,
+      "grad_norm": 0.6315874701337566,
+      "learning_rate": 3.927319935032703e-05,
+      "loss": 0.4765,
+      "step": 390
+    },
+    {
+      "epoch": 0.8896473265073948,
+      "grad_norm": 0.7098100329539186,
+      "learning_rate": 3.926467654342067e-05,
+      "loss": 0.485,
+      "step": 391
+    },
+    {
+      "epoch": 0.8919226393629124,
+      "grad_norm": 0.5044092995092366,
+      "learning_rate": 3.9256104991839724e-05,
+      "loss": 0.4695,
+      "step": 392
+    },
+    {
+      "epoch": 0.89419795221843,
+      "grad_norm": 0.7272226279219421,
+      "learning_rate": 3.924748471727246e-05,
+      "loss": 0.4862,
+      "step": 393
+    },
+    {
+      "epoch": 0.8964732650739476,
+      "grad_norm": 0.5283682029284336,
+      "learning_rate": 3.923881574153043e-05,
+      "loss": 0.4937,
+      "step": 394
+    },
+    {
+      "epoch": 0.8987485779294653,
+      "grad_norm": 0.6049934119270775,
+      "learning_rate": 3.9230098086548414e-05,
+      "loss": 0.4581,
+      "step": 395
+    },
+    {
+      "epoch": 0.9010238907849829,
+      "grad_norm": 0.6190064097442485,
+      "learning_rate": 3.9221331774384356e-05,
+      "loss": 0.4966,
+      "step": 396
+    },
+    {
+      "epoch": 0.9032992036405005,
+      "grad_norm": 0.5778786608492407,
+      "learning_rate": 3.9212516827219323e-05,
+      "loss": 0.4709,
+      "step": 397
+    },
+    {
+      "epoch": 0.9055745164960182,
+      "grad_norm": 0.721457490979815,
+      "learning_rate": 3.920365326735745e-05,
+      "loss": 0.4943,
+      "step": 398
+    },
+    {
+      "epoch": 0.9078498293515358,
+      "grad_norm": 0.598488555520672,
+      "learning_rate": 3.919474111722585e-05,
+      "loss": 0.4643,
+      "step": 399
+    },
+    {
+      "epoch": 0.9101251422070534,
+      "grad_norm": 0.58655862603512,
+      "learning_rate": 3.918578039937459e-05,
+      "loss": 0.4839,
+      "step": 400
+    },
+    {
+      "epoch": 0.9124004550625711,
+      "grad_norm": 0.6909733580750689,
+      "learning_rate": 3.917677113647665e-05,
+      "loss": 0.4816,
+      "step": 401
+    },
+    {
+      "epoch": 0.9146757679180887,
+      "grad_norm": 0.7561177941628888,
+      "learning_rate": 3.916771335132781e-05,
+      "loss": 0.5565,
+      "step": 402
+    },
+    {
+      "epoch": 0.9169510807736063,
+      "grad_norm": 0.6312069877060036,
+      "learning_rate": 3.915860706684664e-05,
+      "loss": 0.5043,
+      "step": 403
+    },
+    {
+      "epoch": 0.919226393629124,
+      "grad_norm": 0.6867308599614804,
+      "learning_rate": 3.914945230607443e-05,
+      "loss": 0.4969,
+      "step": 404
+    },
+    {
+      "epoch": 0.9215017064846417,
+      "grad_norm": 0.6167566851414519,
+      "learning_rate": 3.914024909217511e-05,
+      "loss": 0.5155,
+      "step": 405
+    },
+    {
+      "epoch": 0.9237770193401593,
+      "grad_norm": 0.6703548395838873,
+      "learning_rate": 3.9130997448435235e-05,
+      "loss": 0.4922,
+      "step": 406
+    },
+    {
+      "epoch": 0.926052332195677,
+      "grad_norm": 0.6127702650380229,
+      "learning_rate": 3.9121697398263874e-05,
+      "loss": 0.4862,
+      "step": 407
+    },
+    {
+      "epoch": 0.9283276450511946,
+      "grad_norm": 0.6245821024752293,
+      "learning_rate": 3.911234896519259e-05,
+      "loss": 0.5229,
+      "step": 408
+    },
+    {
+      "epoch": 0.9306029579067122,
+      "grad_norm": 0.6846116775019015,
+      "learning_rate": 3.910295217287537e-05,
+      "loss": 0.5235,
+      "step": 409
+    },
+    {
+      "epoch": 0.9328782707622298,
+      "grad_norm": 0.6293963435933653,
+      "learning_rate": 3.909350704508856e-05,
+      "loss": 0.4991,
+      "step": 410
+    },
+    {
+      "epoch": 0.9351535836177475,
+      "grad_norm": 0.5096912924686995,
+      "learning_rate": 3.90840136057308e-05,
+      "loss": 0.5074,
+      "step": 411
+    },
+    {
+      "epoch": 0.9374288964732651,
+      "grad_norm": 0.6203975419670066,
+      "learning_rate": 3.9074471878822975e-05,
+      "loss": 0.4932,
+      "step": 412
+    },
+    {
+      "epoch": 0.9397042093287827,
+      "grad_norm": 0.5517618128678503,
+      "learning_rate": 3.906488188850816e-05,
+      "loss": 0.5003,
+      "step": 413
+    },
+    {
+      "epoch": 0.9419795221843004,
+      "grad_norm": 0.6999593057949012,
+      "learning_rate": 3.905524365905153e-05,
+      "loss": 0.4752,
+      "step": 414
+    },
+    {
+      "epoch": 0.944254835039818,
+      "grad_norm": 0.6627867781403664,
+      "learning_rate": 3.904555721484034e-05,
+      "loss": 0.4711,
+      "step": 415
+    },
+    {
+      "epoch": 0.9465301478953356,
+      "grad_norm": 0.5561928699593454,
+      "learning_rate": 3.903582258038382e-05,
+      "loss": 0.4742,
+      "step": 416
+    },
+    {
+      "epoch": 0.9488054607508533,
+      "grad_norm": 0.752856160747916,
+      "learning_rate": 3.902603978031315e-05,
+      "loss": 0.506,
+      "step": 417
+    },
+    {
+      "epoch": 0.9510807736063709,
+      "grad_norm": 0.7040158975065405,
+      "learning_rate": 3.901620883938137e-05,
+      "loss": 0.4762,
+      "step": 418
+    },
+    {
+      "epoch": 0.9533560864618885,
+      "grad_norm": 0.9078750042793132,
+      "learning_rate": 3.900632978246334e-05,
+      "loss": 0.4787,
+      "step": 419
+    },
+    {
+      "epoch": 0.9556313993174061,
+      "grad_norm": 0.6880784393167054,
+      "learning_rate": 3.899640263455566e-05,
+      "loss": 0.4921,
+      "step": 420
+    },
+    {
+      "epoch": 0.9579067121729238,
+      "grad_norm": 0.9616382722279428,
+      "learning_rate": 3.8986427420776604e-05,
+      "loss": 0.5105,
+      "step": 421
+    },
+    {
+      "epoch": 0.9601820250284414,
+      "grad_norm": 0.7137899331602579,
+      "learning_rate": 3.897640416636608e-05,
+      "loss": 0.4889,
+      "step": 422
+    },
+    {
+      "epoch": 0.962457337883959,
+      "grad_norm": 0.8857115772260385,
+      "learning_rate": 3.8966332896685545e-05,
+      "loss": 0.4662,
+      "step": 423
+    },
+    {
+      "epoch": 0.9647326507394767,
+      "grad_norm": 0.632315334400379,
+      "learning_rate": 3.895621363721795e-05,
+      "loss": 0.5228,
+      "step": 424
+    },
+    {
+      "epoch": 0.9670079635949943,
+      "grad_norm": 0.8832642529331565,
+      "learning_rate": 3.894604641356767e-05,
+      "loss": 0.5191,
+      "step": 425
+    },
+    {
+      "epoch": 0.9692832764505119,
+      "grad_norm": 0.6474869552718459,
+      "learning_rate": 3.893583125146043e-05,
+      "loss": 0.4918,
+      "step": 426
+    },
+    {
+      "epoch": 0.9715585893060296,
+      "grad_norm": 0.79708282939505,
+      "learning_rate": 3.892556817674328e-05,
+      "loss": 0.5068,
+      "step": 427
+    },
+    {
+      "epoch": 0.9738339021615472,
+      "grad_norm": 0.5979739475341149,
+      "learning_rate": 3.8915257215384485e-05,
+      "loss": 0.4915,
+      "step": 428
+    },
+    {
+      "epoch": 0.9761092150170648,
+      "grad_norm": 0.7710593720070672,
+      "learning_rate": 3.890489839347347e-05,
+      "loss": 0.4941,
+      "step": 429
+    },
+    {
+      "epoch": 0.9783845278725825,
+      "grad_norm": 0.5526776697686373,
+      "learning_rate": 3.889449173722077e-05,
+      "loss": 0.4752,
+      "step": 430
+    },
+    {
+      "epoch": 0.9806598407281001,
+      "grad_norm": 0.7049025640472182,
+      "learning_rate": 3.8884037272957936e-05,
+      "loss": 0.5074,
+      "step": 431
+    },
+    {
+      "epoch": 0.9829351535836177,
+      "grad_norm": 0.6095293538766196,
+      "learning_rate": 3.887353502713752e-05,
+      "loss": 0.5254,
+      "step": 432
+    },
+    {
+      "epoch": 0.9852104664391353,
+      "grad_norm": 0.5384079248688551,
+      "learning_rate": 3.886298502633294e-05,
+      "loss": 0.4802,
+      "step": 433
+    },
+    {
+      "epoch": 0.987485779294653,
+      "grad_norm": 0.679283569969809,
+      "learning_rate": 3.885238729723847e-05,
+      "loss": 0.4922,
+      "step": 434
+    },
+    {
+      "epoch": 0.9897610921501706,
+      "grad_norm": 0.6518320059827871,
+      "learning_rate": 3.8841741866669126e-05,
+      "loss": 0.5277,
+      "step": 435
+    },
+    {
+      "epoch": 0.9920364050056882,
+      "grad_norm": 0.790824660703856,
+      "learning_rate": 3.883104876156064e-05,
+      "loss": 0.5036,
+      "step": 436
+    },
+    {
+      "epoch": 0.9943117178612059,
+      "grad_norm": 0.512397127737013,
+      "learning_rate": 3.882030800896937e-05,
+      "loss": 0.4993,
+      "step": 437
+    },
+    {
+      "epoch": 0.9965870307167235,
+      "grad_norm": 0.7387832742784267,
+      "learning_rate": 3.880951963607222e-05,
+      "loss": 0.5484,
+      "step": 438
+    },
+    {
+      "epoch": 0.9988623435722411,
+      "grad_norm": 0.5472823524136017,
+      "learning_rate": 3.879868367016662e-05,
+      "loss": 0.4769,
+      "step": 439
+    },
+    {
+      "epoch": 1.0011376564277588,
+      "grad_norm": 1.1121215401964364,
+      "learning_rate": 3.878780013867038e-05,
+      "loss": 0.722,
+      "step": 440
+    },
+    {
+      "epoch": 1.0034129692832765,
+      "grad_norm": 0.5303547489884288,
+      "learning_rate": 3.877686906912168e-05,
+      "loss": 0.3795,
+      "step": 441
+    },
+    {
+      "epoch": 1.005688282138794,
+      "grad_norm": 0.6536985280427839,
+      "learning_rate": 3.876589048917901e-05,
+      "loss": 0.3871,
+      "step": 442
+    },
+    {
+      "epoch": 1.0079635949943118,
+      "grad_norm": 0.6972091381993293,
+      "learning_rate": 3.8754864426621045e-05,
+      "loss": 0.5004,
+      "step": 443
+    },
+    {
+      "epoch": 1.0102389078498293,
+      "grad_norm": 0.9103464551846269,
+      "learning_rate": 3.874379090934659e-05,
+      "loss": 0.448,
+      "step": 444
+    },
+    {
+      "epoch": 1.012514220705347,
+      "grad_norm": 0.7346432061843909,
+      "learning_rate": 3.873266996537456e-05,
+      "loss": 0.4681,
+      "step": 445
+    },
+    {
+      "epoch": 1.0147895335608645,
+      "grad_norm": 0.8594704129747518,
+      "learning_rate": 3.8721501622843846e-05,
+      "loss": 0.4278,
+      "step": 446
+    },
+    {
+      "epoch": 1.0170648464163823,
+      "grad_norm": 0.6520097950263782,
+      "learning_rate": 3.871028591001329e-05,
+      "loss": 0.4239,
+      "step": 447
+    },
+    {
+      "epoch": 1.0193401592718998,
+      "grad_norm": 0.9043382585956159,
+      "learning_rate": 3.869902285526157e-05,
+      "loss": 0.4701,
+      "step": 448
+    },
+    {
+      "epoch": 1.0216154721274175,
+      "grad_norm": 0.6783134828580611,
+      "learning_rate": 3.868771248708717e-05,
+      "loss": 0.4754,
+      "step": 449
+    },
+    {
+      "epoch": 1.023890784982935,
+      "grad_norm": 0.805096661530093,
+      "learning_rate": 3.867635483410827e-05,
+      "loss": 0.421,
+      "step": 450
+    },
+    {
+      "epoch": 1.0261660978384528,
+      "grad_norm": 0.6119555640585732,
+      "learning_rate": 3.8664949925062724e-05,
+      "loss": 0.4133,
+      "step": 451
+    },
+    {
+      "epoch": 1.0284414106939703,
+      "grad_norm": 0.940012119317623,
+      "learning_rate": 3.8653497788807926e-05,
+      "loss": 0.4146,
+      "step": 452
+    },
+    {
+      "epoch": 1.030716723549488,
+      "grad_norm": 0.5910709582131054,
+      "learning_rate": 3.864199845432077e-05,
+      "loss": 0.4616,
+      "step": 453
+    },
+    {
+      "epoch": 1.0329920364050056,
+      "grad_norm": 0.7239603105969694,
+      "learning_rate": 3.8630451950697605e-05,
+      "loss": 0.4484,
+      "step": 454
+    },
+    {
+      "epoch": 1.0352673492605233,
+      "grad_norm": 0.771047188430309,
+      "learning_rate": 3.8618858307154085e-05,
+      "loss": 0.4237,
+      "step": 455
+    },
+    {
+      "epoch": 1.0375426621160408,
+      "grad_norm": 0.5924033036723514,
+      "learning_rate": 3.8607217553025174e-05,
+      "loss": 0.4687,
+      "step": 456
+    },
+    {
+      "epoch": 1.0398179749715586,
+      "grad_norm": 0.7867388739225945,
+      "learning_rate": 3.859552971776503e-05,
+      "loss": 0.4366,
+      "step": 457
+    },
+    {
+      "epoch": 1.0420932878270763,
+      "grad_norm": 0.5435581352639793,
+      "learning_rate": 3.858379483094693e-05,
+      "loss": 0.3981,
+      "step": 458
+    },
+    {
+      "epoch": 1.0443686006825939,
+      "grad_norm": 0.7516548808366503,
+      "learning_rate": 3.857201292226322e-05,
+      "loss": 0.4714,
+      "step": 459
+    },
+    {
+      "epoch": 1.0466439135381116,
+      "grad_norm": 0.672487728953046,
+      "learning_rate": 3.8560184021525194e-05,
+      "loss": 0.4314,
+      "step": 460
+    },
+    {
+      "epoch": 1.0489192263936291,
+      "grad_norm": 0.721824130030649,
+      "learning_rate": 3.854830815866308e-05,
+      "loss": 0.4752,
+      "step": 461
+    },
+    {
+      "epoch": 1.0511945392491469,
+      "grad_norm": 0.554588452712049,
+      "learning_rate": 3.8536385363725914e-05,
+      "loss": 0.4099,
+      "step": 462
+    },
+    {
+      "epoch": 1.0534698521046644,
+      "grad_norm": 0.582636238778067,
+      "learning_rate": 3.8524415666881495e-05,
+      "loss": 0.4301,
+      "step": 463
+    },
+    {
+      "epoch": 1.0557451649601821,
+      "grad_norm": 0.6063203313474805,
+      "learning_rate": 3.851239909841629e-05,
+      "loss": 0.4351,
+      "step": 464
+    },
+    {
+      "epoch": 1.0580204778156996,
+      "grad_norm": 0.6530051032333519,
+      "learning_rate": 3.850033568873536e-05,
+      "loss": 0.4823,
+      "step": 465
+    },
+    {
+      "epoch": 1.0602957906712174,
+      "grad_norm": 0.5246299248110554,
+      "learning_rate": 3.8488225468362284e-05,
+      "loss": 0.3294,
+      "step": 466
+    },
+    {
+      "epoch": 1.062571103526735,
+      "grad_norm": 0.5622129573629325,
+      "learning_rate": 3.847606846793909e-05,
+      "loss": 0.4526,
+      "step": 467
+    },
+    {
+      "epoch": 1.0648464163822526,
+      "grad_norm": 0.5890629032138155,
+      "learning_rate": 3.846386471822618e-05,
+      "loss": 0.3856,
+      "step": 468
+    },
+    {
+      "epoch": 1.0671217292377702,
+      "grad_norm": 0.5275742634395962,
+      "learning_rate": 3.8451614250102234e-05,
+      "loss": 0.444,
+      "step": 469
+    },
+    {
+      "epoch": 1.069397042093288,
+      "grad_norm": 0.6842481934322107,
+      "learning_rate": 3.843931709456414e-05,
+      "loss": 0.4241,
+      "step": 470
+    },
+    {
+      "epoch": 1.0716723549488054,
+      "grad_norm": 0.48419973116649284,
+      "learning_rate": 3.8426973282726915e-05,
+      "loss": 0.3856,
+      "step": 471
+    },
+    {
+      "epoch": 1.0739476678043232,
+      "grad_norm": 0.6812528323951617,
+      "learning_rate": 3.841458284582364e-05,
+      "loss": 0.464,
+      "step": 472
+    },
+    {
+      "epoch": 1.0762229806598407,
+      "grad_norm": 0.4606611667546373,
+      "learning_rate": 3.8402145815205365e-05,
+      "loss": 0.4649,
+      "step": 473
+    },
+    {
+      "epoch": 1.0784982935153584,
+      "grad_norm": 0.6711071802996837,
+      "learning_rate": 3.838966222234104e-05,
+      "loss": 0.4304,
+      "step": 474
+    },
+    {
+      "epoch": 1.080773606370876,
+      "grad_norm": 0.5366181070379958,
+      "learning_rate": 3.83771320988174e-05,
+      "loss": 0.3844,
+      "step": 475
+    },
+    {
+      "epoch": 1.0830489192263937,
+      "grad_norm": 0.7505606804041826,
+      "learning_rate": 3.836455547633896e-05,
+      "loss": 0.4522,
+      "step": 476
+    },
+    {
+      "epoch": 1.0853242320819112,
+      "grad_norm": 0.5556699911110706,
+      "learning_rate": 3.835193238672786e-05,
+      "loss": 0.4654,
+      "step": 477
+    },
+    {
+      "epoch": 1.087599544937429,
+      "grad_norm": 0.7648233560530554,
+      "learning_rate": 3.833926286192382e-05,
+      "loss": 0.4287,
+      "step": 478
+    },
+    {
+      "epoch": 1.0898748577929465,
+      "grad_norm": 0.5346556144583974,
+      "learning_rate": 3.832654693398404e-05,
+      "loss": 0.4287,
+      "step": 479
+    },
+    {
+      "epoch": 1.0921501706484642,
+      "grad_norm": 0.6147545474740819,
+      "learning_rate": 3.831378463508318e-05,
+      "loss": 0.4292,
+      "step": 480
+    },
+    {
+      "epoch": 1.0944254835039817,
+      "grad_norm": 0.49199739831216216,
+      "learning_rate": 3.830097599751317e-05,
+      "loss": 0.3876,
+      "step": 481
+    },
+    {
+      "epoch": 1.0967007963594995,
+      "grad_norm": 0.6500922100346228,
+      "learning_rate": 3.828812105368323e-05,
+      "loss": 0.4149,
+      "step": 482
+    },
+    {
+      "epoch": 1.098976109215017,
+      "grad_norm": 0.46632381947018536,
+      "learning_rate": 3.8275219836119715e-05,
+      "loss": 0.3727,
+      "step": 483
+    },
+    {
+      "epoch": 1.1012514220705347,
+      "grad_norm": 0.8163109016449187,
+      "learning_rate": 3.82622723774661e-05,
+      "loss": 0.4258,
+      "step": 484
+    },
+    {
+      "epoch": 1.1035267349260522,
+      "grad_norm": 0.5606775924641534,
+      "learning_rate": 3.824927871048284e-05,
+      "loss": 0.4673,
+      "step": 485
+    },
+    {
+      "epoch": 1.10580204778157,
+      "grad_norm": 0.7046753680729659,
+      "learning_rate": 3.8236238868047315e-05,
+      "loss": 0.4333,
+      "step": 486
+    },
+    {
+      "epoch": 1.1080773606370875,
+      "grad_norm": 0.5036730787338999,
+      "learning_rate": 3.822315288315373e-05,
+      "loss": 0.4384,
+      "step": 487
+    },
+    {
+      "epoch": 1.1103526734926052,
+      "grad_norm": 0.6306392041070552,
+      "learning_rate": 3.821002078891307e-05,
+      "loss": 0.4187,
+      "step": 488
+    },
+    {
+      "epoch": 1.1126279863481228,
+      "grad_norm": 0.5202241380540945,
+      "learning_rate": 3.8196842618552953e-05,
+      "loss": 0.4482,
+      "step": 489
+    },
+    {
+      "epoch": 1.1149032992036405,
+      "grad_norm": 0.8084338237424731,
+      "learning_rate": 3.818361840541761e-05,
+      "loss": 0.4956,
+      "step": 490
+    },
+    {
+      "epoch": 1.117178612059158,
+      "grad_norm": 0.5612996846599624,
+      "learning_rate": 3.8170348182967764e-05,
+      "loss": 0.4056,
+      "step": 491
+    },
+    {
+      "epoch": 1.1194539249146758,
+      "grad_norm": 0.5924393050486125,
+      "learning_rate": 3.815703198478054e-05,
+      "loss": 0.4334,
+      "step": 492
+    },
+    {
+      "epoch": 1.1217292377701935,
+      "grad_norm": 0.5345743904225937,
+      "learning_rate": 3.814366984454941e-05,
+      "loss": 0.3805,
+      "step": 493
+    },
+    {
+      "epoch": 1.124004550625711,
+      "grad_norm": 0.5734287248659643,
+      "learning_rate": 3.81302617960841e-05,
+      "loss": 0.4137,
+      "step": 494
+    },
+    {
+      "epoch": 1.1262798634812285,
+      "grad_norm": 0.6624763213978735,
+      "learning_rate": 3.811680787331047e-05,
+      "loss": 0.4593,
+      "step": 495
+    },
+    {
+      "epoch": 1.1285551763367463,
+      "grad_norm": 0.8598311483755152,
+      "learning_rate": 3.810330811027046e-05,
+      "loss": 0.4163,
+      "step": 496
+    },
+    {
+      "epoch": 1.130830489192264,
+      "grad_norm": 0.5811168048723918,
+      "learning_rate": 3.8089762541122016e-05,
+      "loss": 0.3961,
+      "step": 497
+    },
+    {
+      "epoch": 1.1331058020477816,
+      "grad_norm": 0.6067424161701718,
+      "learning_rate": 3.807617120013897e-05,
+      "loss": 0.3828,
+      "step": 498
+    },
+    {
+      "epoch": 1.1353811149032993,
+      "grad_norm": 0.6027276686195756,
+      "learning_rate": 3.8062534121710974e-05,
+      "loss": 0.4223,
+      "step": 499
+    },
+    {
+      "epoch": 1.1376564277588168,
+      "grad_norm": 0.5265528740905795,
+      "learning_rate": 3.80488513403434e-05,
+      "loss": 0.3269,
+      "step": 500
+    },
+    {
+      "epoch": 1.1399317406143346,
+      "grad_norm": 0.619718442290863,
+      "learning_rate": 3.8035122890657276e-05,
+      "loss": 0.4754,
+      "step": 501
+    },
+    {
+      "epoch": 1.142207053469852,
+      "grad_norm": 0.6342725333529051,
+      "learning_rate": 3.802134880738916e-05,
+      "loss": 0.4512,
+      "step": 502
+    },
+    {
+      "epoch": 1.1444823663253698,
+      "grad_norm": 0.530485201522393,
+      "learning_rate": 3.80075291253911e-05,
+      "loss": 0.4104,
+      "step": 503
+    },
+    {
+      "epoch": 1.1467576791808873,
+      "grad_norm": 0.6627840837725466,
+      "learning_rate": 3.7993663879630516e-05,
+      "loss": 0.4043,
+      "step": 504
+    },
+    {
+      "epoch": 1.149032992036405,
+      "grad_norm": 0.5379023996005294,
+      "learning_rate": 3.797975310519009e-05,
+      "loss": 0.4411,
+      "step": 505
+    },
+    {
+      "epoch": 1.1513083048919226,
+      "grad_norm": 0.588214045835851,
+      "learning_rate": 3.796579683726774e-05,
+      "loss": 0.4187,
+      "step": 506
+    },
+    {
+      "epoch": 1.1535836177474403,
+      "grad_norm": 0.5678190724068664,
+      "learning_rate": 3.795179511117647e-05,
+      "loss": 0.4289,
+      "step": 507
+    },
+    {
+      "epoch": 1.1558589306029579,
+      "grad_norm": 0.688457625822847,
+      "learning_rate": 3.7937747962344295e-05,
+      "loss": 0.4943,
+      "step": 508
+    },
+    {
+      "epoch": 1.1581342434584756,
+      "grad_norm": 0.6814989607077934,
+      "learning_rate": 3.792365542631421e-05,
+      "loss": 0.4214,
+      "step": 509
+    },
+    {
+      "epoch": 1.1604095563139931,
+      "grad_norm": 0.5942736027664269,
+      "learning_rate": 3.7909517538744e-05,
+      "loss": 0.456,
+      "step": 510
+    },
+    {
+      "epoch": 1.1626848691695109,
+      "grad_norm": 0.7903713526208822,
+      "learning_rate": 3.789533433540623e-05,
+      "loss": 0.4608,
+      "step": 511
+    },
+    {
+      "epoch": 1.1649601820250284,
+      "grad_norm": 0.5337770537219456,
+      "learning_rate": 3.788110585218811e-05,
+      "loss": 0.4365,
+      "step": 512
+    },
+    {
+      "epoch": 1.1672354948805461,
+      "grad_norm": 0.5709386118048865,
+      "learning_rate": 3.7866832125091434e-05,
+      "loss": 0.3608,
+      "step": 513
+    },
+    {
+      "epoch": 1.1695108077360636,
+      "grad_norm": 0.6258401342708682,
+      "learning_rate": 3.7852513190232466e-05,
+      "loss": 0.4832,
+      "step": 514
+    },
+    {
+      "epoch": 1.1717861205915814,
+      "grad_norm": 0.5105926958234609,
+      "learning_rate": 3.7838149083841856e-05,
+      "loss": 0.4429,
+      "step": 515
+    },
+    {
+      "epoch": 1.174061433447099,
+      "grad_norm": 0.5353872958700819,
+      "learning_rate": 3.782373984226456e-05,
+      "loss": 0.4411,
+      "step": 516
+    },
+    {
+      "epoch": 1.1763367463026166,
+      "grad_norm": 0.49676244023632943,
+      "learning_rate": 3.7809285501959724e-05,
+      "loss": 0.3666,
+      "step": 517
+    },
+    {
+      "epoch": 1.1786120591581342,
+      "grad_norm": 0.5862119954620956,
+      "learning_rate": 3.7794786099500616e-05,
+      "loss": 0.4018,
+      "step": 518
+    },
+    {
+      "epoch": 1.180887372013652,
+      "grad_norm": 0.5036301639810965,
+      "learning_rate": 3.778024167157452e-05,
+      "loss": 0.4086,
+      "step": 519
+    },
+    {
+      "epoch": 1.1831626848691694,
+      "grad_norm": 0.6783189411364502,
+      "learning_rate": 3.776565225498264e-05,
+      "loss": 0.4399,
+      "step": 520
+    },
+    {
+      "epoch": 1.1854379977246872,
+      "grad_norm": 0.6912119898984184,
+      "learning_rate": 3.775101788664003e-05,
+      "loss": 0.4002,
+      "step": 521
+    },
+    {
+      "epoch": 1.1877133105802047,
+      "grad_norm": 0.7307828897177655,
+      "learning_rate": 3.773633860357547e-05,
+      "loss": 0.4303,
+      "step": 522
+    },
+    {
+      "epoch": 1.1899886234357224,
+      "grad_norm": 0.6451606846442599,
+      "learning_rate": 3.772161444293138e-05,
+      "loss": 0.4027,
+      "step": 523
+    },
+    {
+      "epoch": 1.1922639362912402,
+      "grad_norm": 0.7272217599681045,
+      "learning_rate": 3.7706845441963746e-05,
+      "loss": 0.4485,
+      "step": 524
+    },
+    {
+      "epoch": 1.1945392491467577,
+      "grad_norm": 0.6728295797405291,
+      "learning_rate": 3.769203163804202e-05,
+      "loss": 0.488,
+      "step": 525
+    },
+    {
+      "epoch": 1.1968145620022752,
+      "grad_norm": 0.7295058753840463,
+      "learning_rate": 3.7677173068648976e-05,
+      "loss": 0.4111,
+      "step": 526
+    },
+    {
+      "epoch": 1.199089874857793,
+      "grad_norm": 0.5720940993484606,
+      "learning_rate": 3.76622697713807e-05,
+      "loss": 0.457,
+      "step": 527
+    },
+    {
+      "epoch": 1.2013651877133107,
+      "grad_norm": 0.8869022231692593,
+      "learning_rate": 3.764732178394645e-05,
+      "loss": 0.4342,
+      "step": 528
+    },
+    {
+      "epoch": 1.2036405005688282,
+      "grad_norm": 0.43945315979194277,
+      "learning_rate": 3.763232914416851e-05,
+      "loss": 0.3231,
+      "step": 529
+    },
+    {
+      "epoch": 1.2059158134243457,
+      "grad_norm": 0.7635031106439869,
+      "learning_rate": 3.761729188998222e-05,
+      "loss": 0.4597,
+      "step": 530
+    },
+    {
+      "epoch": 1.2081911262798635,
+      "grad_norm": 0.6213178936947285,
+      "learning_rate": 3.760221005943575e-05,
+      "loss": 0.4321,
+      "step": 531
+    },
+    {
+      "epoch": 1.2104664391353812,
+      "grad_norm": 0.6346015751325879,
+      "learning_rate": 3.758708369069009e-05,
+      "loss": 0.4653,
+      "step": 532
+    },
+    {
+      "epoch": 1.2127417519908987,
+      "grad_norm": 0.8056204649209358,
+      "learning_rate": 3.75719128220189e-05,
+      "loss": 0.4568,
+      "step": 533
+    },
+    {
+      "epoch": 1.2150170648464165,
+      "grad_norm": 0.6513216153599897,
+      "learning_rate": 3.7556697491808455e-05,
+      "loss": 0.4,
+      "step": 534
+    },
+    {
+      "epoch": 1.217292377701934,
+      "grad_norm": 0.5983261337707148,
+      "learning_rate": 3.7541437738557524e-05,
+      "loss": 0.3973,
+      "step": 535
+    },
+    {
+      "epoch": 1.2195676905574517,
+      "grad_norm": 0.7107199080738904,
+      "learning_rate": 3.7526133600877275e-05,
+      "loss": 0.4443,
+      "step": 536
+    },
+    {
+      "epoch": 1.2218430034129693,
+      "grad_norm": 0.5139651203308719,
+      "learning_rate": 3.751078511749119e-05,
+      "loss": 0.4064,
+      "step": 537
+    },
+    {
+      "epoch": 1.224118316268487,
+      "grad_norm": 0.636147405866365,
+      "learning_rate": 3.7495392327234935e-05,
+      "loss": 0.4653,
+      "step": 538
+    },
+    {
+      "epoch": 1.2263936291240045,
+      "grad_norm": 0.7480412548964527,
+      "learning_rate": 3.747995526905632e-05,
+      "loss": 0.4806,
+      "step": 539
+    },
+    {
+      "epoch": 1.2286689419795223,
+      "grad_norm": 0.5622250258722541,
+      "learning_rate": 3.746447398201512e-05,
+      "loss": 0.4358,
+      "step": 540
+    },
+    {
+      "epoch": 1.2309442548350398,
+      "grad_norm": 0.5704468541719319,
+      "learning_rate": 3.744894850528306e-05,
+      "loss": 0.4279,
+      "step": 541
+    },
+    {
+      "epoch": 1.2332195676905575,
+      "grad_norm": 0.5392907143564795,
+      "learning_rate": 3.7433378878143654e-05,
+      "loss": 0.4302,
+      "step": 542
+    },
+    {
+      "epoch": 1.235494880546075,
+      "grad_norm": 0.6537941034798762,
+      "learning_rate": 3.741776513999214e-05,
+      "loss": 0.5026,
+      "step": 543
+    },
+    {
+      "epoch": 1.2377701934015928,
+      "grad_norm": 0.5784636449062069,
+      "learning_rate": 3.7402107330335346e-05,
+      "loss": 0.4013,
+      "step": 544
+    },
+    {
+      "epoch": 1.2400455062571103,
+      "grad_norm": 0.612882563229047,
+      "learning_rate": 3.738640548879166e-05,
+      "loss": 0.4175,
+      "step": 545
+    },
+    {
+      "epoch": 1.242320819112628,
+      "grad_norm": 0.7047522050978076,
+      "learning_rate": 3.7370659655090815e-05,
+      "loss": 0.4813,
+      "step": 546
+    },
+    {
+      "epoch": 1.2445961319681456,
+      "grad_norm": 0.5689652487980504,
+      "learning_rate": 3.7354869869073916e-05,
+      "loss": 0.4219,
+      "step": 547
+    },
+    {
+      "epoch": 1.2468714448236633,
+      "grad_norm": 0.48323792892632456,
+      "learning_rate": 3.733903617069325e-05,
+      "loss": 0.4291,
+      "step": 548
+    },
+    {
+      "epoch": 1.2491467576791808,
+      "grad_norm": 0.6225866345259096,
+      "learning_rate": 3.732315860001222e-05,
+      "loss": 0.4113,
+      "step": 549
+    },
+    {
+      "epoch": 1.2514220705346986,
+      "grad_norm": 0.4521098482775458,
+      "learning_rate": 3.730723719720523e-05,
+      "loss": 0.3784,
+      "step": 550
+    },
+    {
+      "epoch": 1.253697383390216,
+      "grad_norm": 0.5624605850411734,
+      "learning_rate": 3.72912720025576e-05,
+      "loss": 0.4234,
+      "step": 551
+    },
+    {
+      "epoch": 1.2559726962457338,
+      "grad_norm": 0.6099617567525514,
+      "learning_rate": 3.727526305646546e-05,
+      "loss": 0.4719,
+      "step": 552
+    },
+    {
+      "epoch": 1.2582480091012513,
+      "grad_norm": 0.6196907758321264,
+      "learning_rate": 3.725921039943561e-05,
+      "loss": 0.4149,
+      "step": 553
+    },
+    {
+      "epoch": 1.260523321956769,
+      "grad_norm": 0.5265431890255476,
+      "learning_rate": 3.72431140720855e-05,
+      "loss": 0.4067,
+      "step": 554
+    },
+    {
+      "epoch": 1.2627986348122868,
+      "grad_norm": 0.7141180841081363,
+      "learning_rate": 3.722697411514305e-05,
+      "loss": 0.4111,
+      "step": 555
+    },
+    {
+      "epoch": 1.2650739476678043,
+      "grad_norm": 0.7054376658478285,
+      "learning_rate": 3.7210790569446554e-05,
+      "loss": 0.4834,
+      "step": 556
+    },
+    {
+      "epoch": 1.2673492605233219,
+      "grad_norm": 0.544200246308691,
+      "learning_rate": 3.7194563475944645e-05,
+      "loss": 0.3724,
+      "step": 557
+    },
+    {
+      "epoch": 1.2696245733788396,
+      "grad_norm": 0.6859827142678692,
+      "learning_rate": 3.71782928756961e-05,
+      "loss": 0.4445,
+      "step": 558
+    },
+    {
+      "epoch": 1.2718998862343573,
+      "grad_norm": 0.49873706058808576,
+      "learning_rate": 3.7161978809869804e-05,
+      "loss": 0.377,
+      "step": 559
+    },
+    {
+      "epoch": 1.2741751990898749,
+      "grad_norm": 0.6905192832820776,
+      "learning_rate": 3.7145621319744614e-05,
+      "loss": 0.4957,
+      "step": 560
+    },
+    {
+      "epoch": 1.2764505119453924,
+      "grad_norm": 0.6674819555847167,
+      "learning_rate": 3.712922044670926e-05,
+      "loss": 0.4248,
+      "step": 561
+    },
+    {
+      "epoch": 1.2787258248009101,
+      "grad_norm": 0.5531069807879404,
+      "learning_rate": 3.7112776232262244e-05,
+      "loss": 0.4162,
+      "step": 562
+    },
+    {
+      "epoch": 1.2810011376564279,
+      "grad_norm": 0.793114505572383,
+      "learning_rate": 3.709628871801173e-05,
+      "loss": 0.4119,
+      "step": 563
+    },
+    {
+      "epoch": 1.2832764505119454,
+      "grad_norm": 0.6061615937073604,
+      "learning_rate": 3.7079757945675456e-05,
+      "loss": 0.4509,
+      "step": 564
+    },
+    {
+      "epoch": 1.285551763367463,
+      "grad_norm": 0.5801581834214365,
+      "learning_rate": 3.7063183957080594e-05,
+      "loss": 0.4292,
+      "step": 565
+    },
+    {
+      "epoch": 1.2878270762229806,
+      "grad_norm": 0.6531156166646916,
+      "learning_rate": 3.704656679416368e-05,
+      "loss": 0.4211,
+      "step": 566
+    },
+    {
+      "epoch": 1.2901023890784984,
+      "grad_norm": 0.4738840824742358,
+      "learning_rate": 3.702990649897047e-05,
+      "loss": 0.4046,
+      "step": 567
+    },
+    {
+      "epoch": 1.292377701934016,
+      "grad_norm": 0.6127788436356492,
+      "learning_rate": 3.70132031136559e-05,
+      "loss": 0.4367,
+      "step": 568
+    },
+    {
+      "epoch": 1.2946530147895334,
+      "grad_norm": 0.4464306345350417,
+      "learning_rate": 3.699645668048388e-05,
+      "loss": 0.407,
+      "step": 569
+    },
+    {
+      "epoch": 1.2969283276450512,
+      "grad_norm": 0.5847521547698988,
+      "learning_rate": 3.697966724182729e-05,
+      "loss": 0.443,
+      "step": 570
+    },
+    {
+      "epoch": 1.299203640500569,
+      "grad_norm": 0.5520684612319413,
+      "learning_rate": 3.6962834840167783e-05,
+      "loss": 0.4154,
+      "step": 571
+    },
+    {
+      "epoch": 1.3014789533560864,
+      "grad_norm": 0.577929816119576,
+      "learning_rate": 3.694595951809576e-05,
+      "loss": 0.4258,
+      "step": 572
+    },
+    {
+      "epoch": 1.3037542662116042,
+      "grad_norm": 0.6588178938708844,
+      "learning_rate": 3.6929041318310195e-05,
+      "loss": 0.4715,
+      "step": 573
+    },
+    {
+      "epoch": 1.3060295790671217,
+      "grad_norm": 0.5910947193408698,
+      "learning_rate": 3.691208028361857e-05,
+      "loss": 0.4136,
+      "step": 574
+    },
+    {
+      "epoch": 1.3083048919226394,
+      "grad_norm": 0.6437280595180613,
+      "learning_rate": 3.689507645693674e-05,
+      "loss": 0.4218,
+      "step": 575
+    },
+    {
+      "epoch": 1.310580204778157,
+      "grad_norm": 0.5321019507324688,
+      "learning_rate": 3.6878029881288824e-05,
+      "loss": 0.4413,
+      "step": 576
+    },
+    {
+      "epoch": 1.3128555176336747,
+      "grad_norm": 0.5655094990541256,
+      "learning_rate": 3.686094059980714e-05,
+      "loss": 0.4285,
+      "step": 577
+    },
+    {
+      "epoch": 1.3151308304891922,
+      "grad_norm": 0.5994309074467306,
+      "learning_rate": 3.684380865573203e-05,
+      "loss": 0.4059,
+      "step": 578
+    },
+    {
+      "epoch": 1.31740614334471,
+      "grad_norm": 0.5449339695520942,
+      "learning_rate": 3.6826634092411807e-05,
+      "loss": 0.4255,
+      "step": 579
+    },
+    {
+      "epoch": 1.3196814562002275,
+      "grad_norm": 0.6193425203295791,
+      "learning_rate": 3.6809416953302606e-05,
+      "loss": 0.4753,
+      "step": 580
+    },
+    {
+      "epoch": 1.3219567690557452,
+      "grad_norm": 0.5409147548110276,
+      "learning_rate": 3.6792157281968295e-05,
+      "loss": 0.3844,
+      "step": 581
+    },
+    {
+      "epoch": 1.3242320819112627,
+      "grad_norm": 0.5590317046964576,
+      "learning_rate": 3.677485512208037e-05,
+      "loss": 0.4503,
+      "step": 582
+    },
+    {
+      "epoch": 1.3265073947667805,
+      "grad_norm": 0.5496354734029336,
+      "learning_rate": 3.675751051741781e-05,
+      "loss": 0.4277,
+      "step": 583
+    },
+    {
+      "epoch": 1.328782707622298,
+      "grad_norm": 0.5723750685999135,
+      "learning_rate": 3.674012351186702e-05,
+      "loss": 0.4235,
+      "step": 584
+    },
+    {
+      "epoch": 1.3310580204778157,
+      "grad_norm": 0.5517032292934076,
+      "learning_rate": 3.672269414942166e-05,
+      "loss": 0.4246,
+      "step": 585
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.6287202109839667,
+      "learning_rate": 3.670522247418259e-05,
+      "loss": 0.4237,
+      "step": 586
+    },
+    {
+      "epoch": 1.335608646188851,
+      "grad_norm": 0.5113344624756003,
+      "learning_rate": 3.6687708530357726e-05,
+      "loss": 0.4012,
+      "step": 587
+    },
+    {
+      "epoch": 1.3378839590443685,
+      "grad_norm": 0.5574487232141875,
+      "learning_rate": 3.667015236226191e-05,
+      "loss": 0.4086,
+      "step": 588
+    },
+    {
+      "epoch": 1.3401592718998863,
+      "grad_norm": 0.5245768687215431,
+      "learning_rate": 3.665255401431687e-05,
+      "loss": 0.3862,
+      "step": 589
+    },
+    {
+      "epoch": 1.342434584755404,
+      "grad_norm": 0.49618557229950566,
+      "learning_rate": 3.663491353105101e-05,
+      "loss": 0.4012,
+      "step": 590
+    },
+    {
+      "epoch": 1.3447098976109215,
+      "grad_norm": 0.4983237488761633,
+      "learning_rate": 3.661723095709939e-05,
+      "loss": 0.3976,
+      "step": 591
+    },
+    {
+      "epoch": 1.346985210466439,
+      "grad_norm": 0.5376215526462558,
+      "learning_rate": 3.659950633720354e-05,
+      "loss": 0.4458,
+      "step": 592
+    },
+    {
+      "epoch": 1.3492605233219568,
+      "grad_norm": 0.5023259197769308,
+      "learning_rate": 3.658173971621139e-05,
+      "loss": 0.458,
+      "step": 593
+    },
+    {
+      "epoch": 1.3515358361774745,
+      "grad_norm": 0.4989521415561501,
+      "learning_rate": 3.6563931139077134e-05,
+      "loss": 0.4332,
+      "step": 594
+    },
+    {
+      "epoch": 1.353811149032992,
+      "grad_norm": 0.510072444206674,
+      "learning_rate": 3.654608065086115e-05,
+      "loss": 0.4081,
+      "step": 595
+    },
+    {
+      "epoch": 1.3560864618885096,
+      "grad_norm": 0.5959408340108707,
+      "learning_rate": 3.652818829672983e-05,
+      "loss": 0.4354,
+      "step": 596
+    },
+    {
+      "epoch": 1.3583617747440273,
+      "grad_norm": 0.5844735352057744,
+      "learning_rate": 3.651025412195552e-05,
+      "loss": 0.4262,
+      "step": 597
+    },
+    {
+      "epoch": 1.360637087599545,
+      "grad_norm": 0.5677541553759599,
+      "learning_rate": 3.649227817191639e-05,
+      "loss": 0.3893,
+      "step": 598
+    },
+    {
+      "epoch": 1.3629124004550626,
+      "grad_norm": 0.5729951500575374,
+      "learning_rate": 3.6474260492096274e-05,
+      "loss": 0.457,
+      "step": 599
+    },
+    {
+      "epoch": 1.36518771331058,
+      "grad_norm": 0.7503620931885405,
+      "learning_rate": 3.645620112808464e-05,
+      "loss": 0.4421,
+      "step": 600
+    },
+    {
+      "epoch": 1.3674630261660978,
+      "grad_norm": 0.7296856049994408,
+      "learning_rate": 3.643810012557639e-05,
+      "loss": 0.4508,
+      "step": 601
+    },
+    {
+      "epoch": 1.3697383390216156,
+      "grad_norm": 0.6002401423186757,
+      "learning_rate": 3.641995753037182e-05,
+      "loss": 0.3956,
+      "step": 602
+    },
+    {
+      "epoch": 1.372013651877133,
+      "grad_norm": 0.6180465269627985,
+      "learning_rate": 3.640177338837641e-05,
+      "loss": 0.4014,
+      "step": 603
+    },
+    {
+      "epoch": 1.3742889647326506,
+      "grad_norm": 0.6333745280579856,
+      "learning_rate": 3.638354774560084e-05,
+      "loss": 0.497,
+      "step": 604
+    },
+    {
+      "epoch": 1.3765642775881684,
+      "grad_norm": 0.6115951533043017,
+      "learning_rate": 3.636528064816073e-05,
+      "loss": 0.3805,
+      "step": 605
+    },
+    {
+      "epoch": 1.378839590443686,
+      "grad_norm": 0.5930980745694563,
+      "learning_rate": 3.6346972142276625e-05,
+      "loss": 0.4619,
+      "step": 606
+    },
+    {
+      "epoch": 1.3811149032992036,
+      "grad_norm": 0.634649972147651,
+      "learning_rate": 3.632862227427384e-05,
+      "loss": 0.4665,
+      "step": 607
+    },
+    {
+      "epoch": 1.3833902161547214,
+      "grad_norm": 0.6272867545095703,
+      "learning_rate": 3.631023109058235e-05,
+      "loss": 0.4541,
+      "step": 608
+    },
+    {
+      "epoch": 1.3856655290102389,
+      "grad_norm": 0.6317307195256453,
+      "learning_rate": 3.629179863773665e-05,
+      "loss": 0.4183,
+      "step": 609
+    },
+    {
+      "epoch": 1.3879408418657566,
+      "grad_norm": 0.5434046371600441,
+      "learning_rate": 3.6273324962375676e-05,
+      "loss": 0.3855,
+      "step": 610
+    },
+    {
+      "epoch": 1.3902161547212741,
+      "grad_norm": 0.7065598662541726,
+      "learning_rate": 3.625481011124267e-05,
+      "loss": 0.4573,
+      "step": 611
+    },
+    {
+      "epoch": 1.3924914675767919,
+      "grad_norm": 0.5852587577597338,
+      "learning_rate": 3.6236254131185046e-05,
+      "loss": 0.4478,
+      "step": 612
+    },
+    {
+      "epoch": 1.3947667804323094,
+      "grad_norm": 0.5653752038889909,
+      "learning_rate": 3.621765706915428e-05,
+      "loss": 0.4694,
+      "step": 613
+    },
+    {
+      "epoch": 1.3970420932878271,
+      "grad_norm": 0.5162422369201265,
+      "learning_rate": 3.6199018972205836e-05,
+      "loss": 0.3594,
+      "step": 614
+    },
+    {
+      "epoch": 1.3993174061433447,
+      "grad_norm": 0.6598966136350053,
+      "learning_rate": 3.6180339887498953e-05,
+      "loss": 0.5066,
+      "step": 615
+    },
+    {
+      "epoch": 1.4015927189988624,
+      "grad_norm": 0.5817663831618288,
+      "learning_rate": 3.616161986229661e-05,
+      "loss": 0.375,
+      "step": 616
+    },
+    {
+      "epoch": 1.40386803185438,
+      "grad_norm": 0.5876775258558006,
+      "learning_rate": 3.614285894396538e-05,
+      "loss": 0.4934,
+      "step": 617
+    },
+    {
+      "epoch": 1.4061433447098977,
+      "grad_norm": 0.5385625038563776,
+      "learning_rate": 3.612405717997529e-05,
+      "loss": 0.4172,
+      "step": 618
+    },
+    {
+      "epoch": 1.4084186575654152,
+      "grad_norm": 0.6459700228914548,
+      "learning_rate": 3.610521461789972e-05,
+      "loss": 0.4522,
+      "step": 619
+    },
+    {
+      "epoch": 1.410693970420933,
+      "grad_norm": 0.4936465756853313,
+      "learning_rate": 3.60863313054153e-05,
+      "loss": 0.3558,
+      "step": 620
+    },
+    {
+      "epoch": 1.4129692832764504,
+      "grad_norm": 0.5480391312766864,
+      "learning_rate": 3.606740729030174e-05,
+      "loss": 0.4218,
+      "step": 621
+    },
+    {
+      "epoch": 1.4152445961319682,
+      "grad_norm": 0.5907103782399543,
+      "learning_rate": 3.6048442620441754e-05,
+      "loss": 0.4813,
+      "step": 622
+    },
+    {
+      "epoch": 1.4175199089874857,
+      "grad_norm": 0.45735368041109864,
+      "learning_rate": 3.6029437343820925e-05,
+      "loss": 0.3461,
+      "step": 623
+    },
+    {
+      "epoch": 1.4197952218430034,
+      "grad_norm": 0.5431566682948396,
+      "learning_rate": 3.601039150852758e-05,
+      "loss": 0.4243,
+      "step": 624
+    },
+    {
+      "epoch": 1.4220705346985212,
+      "grad_norm": 0.5183924879891563,
+      "learning_rate": 3.599130516275266e-05,
+      "loss": 0.4595,
+      "step": 625
+    },
+    {
+      "epoch": 1.4243458475540387,
+      "grad_norm": 0.5720004646057972,
+      "learning_rate": 3.597217835478962e-05,
+      "loss": 0.4515,
+      "step": 626
+    },
+    {
+      "epoch": 1.4266211604095562,
+      "grad_norm": 0.47687964955658974,
+      "learning_rate": 3.595301113303429e-05,
+      "loss": 0.4063,
+      "step": 627
+    },
+    {
+      "epoch": 1.428896473265074,
+      "grad_norm": 0.5493519780479028,
+      "learning_rate": 3.593380354598476e-05,
+      "loss": 0.3963,
+      "step": 628
+    },
+    {
+      "epoch": 1.4311717861205917,
+      "grad_norm": 0.5656763414743938,
+      "learning_rate": 3.591455564224126e-05,
+      "loss": 0.417,
+      "step": 629
+    },
+    {
+      "epoch": 1.4334470989761092,
+      "grad_norm": 0.6174302910133038,
+      "learning_rate": 3.589526747050601e-05,
+      "loss": 0.4714,
+      "step": 630
+    },
+    {
+      "epoch": 1.4357224118316267,
+      "grad_norm": 0.5314357744461629,
+      "learning_rate": 3.587593907958314e-05,
+      "loss": 0.409,
+      "step": 631
+    },
+    {
+      "epoch": 1.4379977246871445,
+      "grad_norm": 0.5959045877414539,
+      "learning_rate": 3.585657051837855e-05,
+      "loss": 0.3917,
+      "step": 632
+    },
+    {
+      "epoch": 1.4402730375426622,
+      "grad_norm": 0.5925951139677259,
+      "learning_rate": 3.583716183589975e-05,
+      "loss": 0.4563,
+      "step": 633
+    },
+    {
+      "epoch": 1.4425483503981797,
+      "grad_norm": 0.5461579158393477,
+      "learning_rate": 3.58177130812558e-05,
+      "loss": 0.4687,
+      "step": 634
+    },
+    {
+      "epoch": 1.4448236632536973,
+      "grad_norm": 0.5705089648405642,
+      "learning_rate": 3.579822430365714e-05,
+      "loss": 0.4327,
+      "step": 635
+    },
+    {
+      "epoch": 1.447098976109215,
+      "grad_norm": 0.645475831271653,
+      "learning_rate": 3.577869555241548e-05,
+      "loss": 0.4536,
+      "step": 636
+    },
+    {
+      "epoch": 1.4493742889647327,
+      "grad_norm": 0.537595180808403,
+      "learning_rate": 3.5759126876943665e-05,
+      "loss": 0.4213,
+      "step": 637
+    },
+    {
+      "epoch": 1.4516496018202503,
+      "grad_norm": 0.5476762610652368,
+      "learning_rate": 3.573951832675557e-05,
+      "loss": 0.3823,
+      "step": 638
+    },
+    {
+      "epoch": 1.4539249146757678,
+      "grad_norm": 0.5303957445839604,
+      "learning_rate": 3.571986995146596e-05,
+      "loss": 0.451,
+      "step": 639
+    },
+    {
+      "epoch": 1.4562002275312855,
+      "grad_norm": 0.5134907930348338,
+      "learning_rate": 3.570018180079037e-05,
+      "loss": 0.449,
+      "step": 640
+    },
+    {
+      "epoch": 1.4584755403868033,
+      "grad_norm": 0.55728918847765,
+      "learning_rate": 3.568045392454498e-05,
+      "loss": 0.4201,
+      "step": 641
+    },
+    {
+      "epoch": 1.4607508532423208,
+      "grad_norm": 0.48252879342351945,
+      "learning_rate": 3.566068637264647e-05,
+      "loss": 0.3453,
+      "step": 642
+    },
+    {
+      "epoch": 1.4630261660978385,
+      "grad_norm": 0.6602277308990888,
+      "learning_rate": 3.564087919511193e-05,
+      "loss": 0.4534,
+      "step": 643
+    },
+    {
+      "epoch": 1.465301478953356,
+      "grad_norm": 0.6004785281379483,
+      "learning_rate": 3.562103244205869e-05,
+      "loss": 0.4352,
+      "step": 644
+    },
+    {
+      "epoch": 1.4675767918088738,
+      "grad_norm": 0.6718736260115368,
+      "learning_rate": 3.560114616370425e-05,
+      "loss": 0.4343,
+      "step": 645
+    },
+    {
+      "epoch": 1.4698521046643913,
+      "grad_norm": 0.7520904707301957,
+      "learning_rate": 3.558122041036608e-05,
+      "loss": 0.5003,
+      "step": 646
+    },
+    {
+      "epoch": 1.472127417519909,
+      "grad_norm": 0.589426691913399,
+      "learning_rate": 3.556125523246157e-05,
+      "loss": 0.3845,
+      "step": 647
+    },
+    {
+      "epoch": 1.4744027303754266,
+      "grad_norm": 0.8524632940765311,
+      "learning_rate": 3.554125068050783e-05,
+      "loss": 0.4957,
+      "step": 648
+    },
+    {
+      "epoch": 1.4766780432309443,
+      "grad_norm": 0.5164519133696035,
+      "learning_rate": 3.5521206805121626e-05,
+      "loss": 0.3878,
+      "step": 649
+    },
+    {
+      "epoch": 1.4789533560864618,
+      "grad_norm": 0.6042089039732726,
+      "learning_rate": 3.550112365701921e-05,
+      "loss": 0.4325,
+      "step": 650
+    },
+    {
+      "epoch": 1.4812286689419796,
+      "grad_norm": 0.5470513351715313,
+      "learning_rate": 3.548100128701619e-05,
+      "loss": 0.3819,
+      "step": 651
+    },
+    {
+      "epoch": 1.483503981797497,
+      "grad_norm": 0.5931353822631468,
+      "learning_rate": 3.546083974602745e-05,
+      "loss": 0.4815,
+      "step": 652
+    },
+    {
+      "epoch": 1.4857792946530148,
+      "grad_norm": 0.5390692476906889,
+      "learning_rate": 3.5440639085066944e-05,
+      "loss": 0.4074,
+      "step": 653
+    },
+    {
+      "epoch": 1.4880546075085324,
+      "grad_norm": 0.5454314551786908,
+      "learning_rate": 3.542039935524765e-05,
+      "loss": 0.434,
+      "step": 654
+    },
+    {
+      "epoch": 1.49032992036405,
+      "grad_norm": 0.6505407285781365,
+      "learning_rate": 3.540012060778137e-05,
+      "loss": 0.4386,
+      "step": 655
+    },
+    {
+      "epoch": 1.4926052332195676,
+      "grad_norm": 0.5343082656896795,
+      "learning_rate": 3.537980289397866e-05,
+      "loss": 0.3717,
+      "step": 656
+    },
+    {
+      "epoch": 1.4948805460750854,
+      "grad_norm": 0.641941733145819,
+      "learning_rate": 3.535944626524863e-05,
+      "loss": 0.4496,
+      "step": 657
+    },
+    {
+      "epoch": 1.4971558589306029,
+      "grad_norm": 0.5002958448477494,
+      "learning_rate": 3.53390507730989e-05,
+      "loss": 0.439,
+      "step": 658
+    },
+    {
+      "epoch": 1.4994311717861206,
+      "grad_norm": 0.7431898707356385,
+      "learning_rate": 3.53186164691354e-05,
+      "loss": 0.4425,
+      "step": 659
+    },
+    {
+      "epoch": 1.5017064846416384,
+      "grad_norm": 0.5625161811682998,
+      "learning_rate": 3.529814340506226e-05,
+      "loss": 0.3554,
+      "step": 660
+    },
+    {
+      "epoch": 1.5039817974971559,
+      "grad_norm": 0.7165022209644997,
+      "learning_rate": 3.52776316326817e-05,
+      "loss": 0.4264,
+      "step": 661
+    },
+    {
+      "epoch": 1.5062571103526734,
+      "grad_norm": 0.5406684891638924,
+      "learning_rate": 3.525708120389387e-05,
+      "loss": 0.471,
+      "step": 662
+    },
+    {
+      "epoch": 1.5085324232081911,
+      "grad_norm": 0.6878653332939912,
+      "learning_rate": 3.523649217069673e-05,
+      "loss": 0.4019,
+      "step": 663
+    },
+    {
+      "epoch": 1.5108077360637089,
+      "grad_norm": 0.7435693445914054,
+      "learning_rate": 3.521586458518593e-05,
+      "loss": 0.5242,
+      "step": 664
+    },
+    {
+      "epoch": 1.5130830489192264,
+      "grad_norm": 0.7244141769516639,
+      "learning_rate": 3.519519849955466e-05,
+      "loss": 0.3751,
+      "step": 665
+    },
+    {
+      "epoch": 1.515358361774744,
+      "grad_norm": 0.9823209050438462,
+      "learning_rate": 3.517449396609353e-05,
+      "loss": 0.5272,
+      "step": 666
+    },
+    {
+      "epoch": 1.5176336746302617,
+      "grad_norm": 0.9331506658748052,
+      "learning_rate": 3.515375103719042e-05,
+      "loss": 0.3914,
+      "step": 667
+    },
+    {
+      "epoch": 1.5199089874857794,
+      "grad_norm": 0.9043246185269522,
+      "learning_rate": 3.5132969765330384e-05,
+      "loss": 0.5045,
+      "step": 668
+    },
+    {
+      "epoch": 1.522184300341297,
+      "grad_norm": 0.8318586391712912,
+      "learning_rate": 3.5112150203095464e-05,
+      "loss": 0.4076,
+      "step": 669
+    },
+    {
+      "epoch": 1.5244596131968144,
+      "grad_norm": 0.6027291468147227,
+      "learning_rate": 3.509129240316461e-05,
+      "loss": 0.4436,
+      "step": 670
+    },
+    {
+      "epoch": 1.5267349260523322,
+      "grad_norm": 0.6823966178652825,
+      "learning_rate": 3.507039641831351e-05,
+      "loss": 0.43,
+      "step": 671
+    },
+    {
+      "epoch": 1.52901023890785,
+      "grad_norm": 0.49401985184456687,
+      "learning_rate": 3.5049462301414485e-05,
+      "loss": 0.3709,
+      "step": 672
+    },
+    {
+      "epoch": 1.5312855517633674,
+      "grad_norm": 0.7521950945695709,
+      "learning_rate": 3.502849010543633e-05,
+      "loss": 0.4482,
+      "step": 673
+    },
+    {
+      "epoch": 1.533560864618885,
+      "grad_norm": 0.4865749920113331,
+      "learning_rate": 3.500747988344418e-05,
+      "loss": 0.3553,
+      "step": 674
+    },
+    {
+      "epoch": 1.5358361774744027,
+      "grad_norm": 0.7833212324594446,
+      "learning_rate": 3.498643168859941e-05,
+      "loss": 0.4586,
+      "step": 675
+    },
+    {
+      "epoch": 1.5381114903299204,
+      "grad_norm": 0.6615305501265996,
+      "learning_rate": 3.496534557415945e-05,
+      "loss": 0.4285,
+      "step": 676
+    },
+    {
+      "epoch": 1.540386803185438,
+      "grad_norm": 0.9007038838265798,
+      "learning_rate": 3.49442215934777e-05,
+      "loss": 0.4671,
+      "step": 677
+    },
+    {
+      "epoch": 1.5426621160409555,
+      "grad_norm": 0.6800624160939918,
+      "learning_rate": 3.492305980000336e-05,
+      "loss": 0.3994,
+      "step": 678
+    },
+    {
+      "epoch": 1.5449374288964732,
+      "grad_norm": 0.7969418939760616,
+      "learning_rate": 3.49018602472813e-05,
+      "loss": 0.5462,
+      "step": 679
+    },
+    {
+      "epoch": 1.547212741751991,
+      "grad_norm": 0.8878652763061456,
+      "learning_rate": 3.488062298895194e-05,
+      "loss": 0.4025,
+      "step": 680
+    },
+    {
+      "epoch": 1.5494880546075085,
+      "grad_norm": 0.659068502289326,
+      "learning_rate": 3.4859348078751104e-05,
+      "loss": 0.4465,
+      "step": 681
+    },
+    {
+      "epoch": 1.551763367463026,
+      "grad_norm": 1.0400850492606901,
+      "learning_rate": 3.483803557050989e-05,
+      "loss": 0.4568,
+      "step": 682
+    },
+    {
+      "epoch": 1.5540386803185438,
+      "grad_norm": 0.5459464999028117,
+      "learning_rate": 3.481668551815451e-05,
+      "loss": 0.3927,
+      "step": 683
+    },
+    {
+      "epoch": 1.5563139931740615,
+      "grad_norm": 0.9155796738843736,
+      "learning_rate": 3.47952979757062e-05,
+      "loss": 0.4053,
+      "step": 684
+    },
+    {
+      "epoch": 1.558589306029579,
+      "grad_norm": 0.6377418061495952,
+      "learning_rate": 3.4773872997281026e-05,
+      "loss": 0.4688,
+      "step": 685
+    },
+    {
+      "epoch": 1.5608646188850968,
+      "grad_norm": 0.7829625967353625,
+      "learning_rate": 3.47524106370898e-05,
+      "loss": 0.4194,
+      "step": 686
+    },
+    {
+      "epoch": 1.5631399317406145,
+      "grad_norm": 0.5929924700099798,
+      "learning_rate": 3.473091094943791e-05,
+      "loss": 0.4175,
+      "step": 687
+    },
+    {
+      "epoch": 1.565415244596132,
+      "grad_norm": 0.6712086411389321,
+      "learning_rate": 3.47093739887252e-05,
+      "loss": 0.4458,
+      "step": 688
+    },
+    {
+      "epoch": 1.5676905574516495,
+      "grad_norm": 0.583321245931635,
+      "learning_rate": 3.468779980944581e-05,
+      "loss": 0.4219,
+      "step": 689
+    },
+    {
+      "epoch": 1.5699658703071673,
+      "grad_norm": 0.6812907714793384,
+      "learning_rate": 3.466618846618806e-05,
+      "loss": 0.406,
+      "step": 690
+    },
+    {
+      "epoch": 1.572241183162685,
+      "grad_norm": 0.5237917183630804,
+      "learning_rate": 3.4644540013634316e-05,
+      "loss": 0.3952,
+      "step": 691
+    },
+    {
+      "epoch": 1.5745164960182025,
+      "grad_norm": 0.6517368564518353,
+      "learning_rate": 3.4622854506560815e-05,
+      "loss": 0.4399,
+      "step": 692
+    },
+    {
+      "epoch": 1.57679180887372,
+      "grad_norm": 0.7673454679891567,
+      "learning_rate": 3.460113199983758e-05,
+      "loss": 0.4697,
+      "step": 693
+    },
+    {
+      "epoch": 1.5790671217292378,
+      "grad_norm": 0.5095156725474326,
+      "learning_rate": 3.457937254842823e-05,
+      "loss": 0.4047,
+      "step": 694
+    },
+    {
+      "epoch": 1.5813424345847555,
+      "grad_norm": 0.6908052191108716,
+      "learning_rate": 3.455757620738989e-05,
+      "loss": 0.4363,
+      "step": 695
+    },
+    {
+      "epoch": 1.583617747440273,
+      "grad_norm": 0.4786305700260121,
+      "learning_rate": 3.453574303187298e-05,
+      "loss": 0.3975,
+      "step": 696
+    },
+    {
+      "epoch": 1.5858930602957906,
+      "grad_norm": 0.7527935841962348,
+      "learning_rate": 3.451387307712117e-05,
+      "loss": 0.4569,
+      "step": 697
+    },
+    {
+      "epoch": 1.5881683731513083,
+      "grad_norm": 0.550349132523598,
+      "learning_rate": 3.4491966398471175e-05,
+      "loss": 0.4999,
+      "step": 698
+    },
+    {
+      "epoch": 1.590443686006826,
+      "grad_norm": 0.626223454214479,
+      "learning_rate": 3.447002305135261e-05,
+      "loss": 0.4665,
+      "step": 699
+    },
+    {
+      "epoch": 1.5927189988623436,
+      "grad_norm": 0.44085264091799026,
+      "learning_rate": 3.444804309128789e-05,
+      "loss": 0.3904,
+      "step": 700
+    },
+    {
+      "epoch": 1.594994311717861,
+      "grad_norm": 0.5320242737175317,
+      "learning_rate": 3.442602657389208e-05,
+      "loss": 0.4558,
+      "step": 701
+    },
+    {
+      "epoch": 1.5972696245733788,
+      "grad_norm": 0.512740152485577,
+      "learning_rate": 3.440397355487272e-05,
+      "loss": 0.3967,
+      "step": 702
+    },
+    {
+      "epoch": 1.5995449374288966,
+      "grad_norm": 0.586264919926901,
+      "learning_rate": 3.438188409002972e-05,
+      "loss": 0.4306,
+      "step": 703
+    },
+    {
+      "epoch": 1.601820250284414,
+      "grad_norm": 0.5613994999872779,
+      "learning_rate": 3.435975823525523e-05,
+      "loss": 0.4682,
+      "step": 704
+    },
+    {
+      "epoch": 1.6040955631399316,
+      "grad_norm": 0.48315507057673795,
+      "learning_rate": 3.4337596046533426e-05,
+      "loss": 0.3671,
+      "step": 705
+    },
+    {
+      "epoch": 1.6063708759954494,
+      "grad_norm": 0.6819301484592192,
+      "learning_rate": 3.4315397579940466e-05,
+      "loss": 0.4571,
+      "step": 706
+    },
+    {
+      "epoch": 1.608646188850967,
+      "grad_norm": 0.43527686529515514,
+      "learning_rate": 3.429316289164426e-05,
+      "loss": 0.4577,
+      "step": 707
+    },
+    {
+      "epoch": 1.6109215017064846,
+      "grad_norm": 0.6082554343422415,
+      "learning_rate": 3.427089203790442e-05,
+      "loss": 0.4239,
+      "step": 708
+    },
+    {
+      "epoch": 1.6131968145620021,
+      "grad_norm": 0.529694601924603,
+      "learning_rate": 3.424858507507202e-05,
+      "loss": 0.436,
+      "step": 709
+    },
+    {
+      "epoch": 1.6154721274175199,
+      "grad_norm": 0.4767337527915039,
+      "learning_rate": 3.42262420595895e-05,
+      "loss": 0.3503,
+      "step": 710
+    },
+    {
+      "epoch": 1.6177474402730376,
+      "grad_norm": 0.7631748467117545,
+      "learning_rate": 3.420386304799057e-05,
+      "loss": 0.4506,
+      "step": 711
+    },
+    {
+      "epoch": 1.6200227531285551,
+      "grad_norm": 0.5475209507603683,
+      "learning_rate": 3.418144809689996e-05,
+      "loss": 0.3797,
+      "step": 712
+    },
+    {
+      "epoch": 1.6222980659840727,
+      "grad_norm": 0.7219843622127368,
+      "learning_rate": 3.4158997263033384e-05,
+      "loss": 0.397,
+      "step": 713
+    },
+    {
+      "epoch": 1.6245733788395904,
+      "grad_norm": 0.6229913232346951,
+      "learning_rate": 3.413651060319732e-05,
+      "loss": 0.4915,
+      "step": 714
+    },
+    {
+      "epoch": 1.6268486916951082,
+      "grad_norm": 0.6116243374881276,
+      "learning_rate": 3.411398817428889e-05,
+      "loss": 0.4495,
+      "step": 715
+    },
+    {
+      "epoch": 1.6291240045506257,
+      "grad_norm": 0.6874023779192933,
+      "learning_rate": 3.409143003329575e-05,
+      "loss": 0.4577,
+      "step": 716
+    },
+    {
+      "epoch": 1.6313993174061432,
+      "grad_norm": 0.5331865010181484,
+      "learning_rate": 3.406883623729591e-05,
+      "loss": 0.389,
+      "step": 717
+    },
+    {
+      "epoch": 1.633674630261661,
+      "grad_norm": 0.541208507385322,
+      "learning_rate": 3.4046206843457576e-05,
+      "loss": 0.425,
+      "step": 718
+    },
+    {
+      "epoch": 1.6359499431171787,
+      "grad_norm": 0.5574749097907555,
+      "learning_rate": 3.4023541909039035e-05,
+      "loss": 0.4125,
+      "step": 719
+    },
+    {
+      "epoch": 1.6382252559726962,
+      "grad_norm": 0.525379595046595,
+      "learning_rate": 3.400084149138851e-05,
+      "loss": 0.4011,
+      "step": 720
+    },
+    {
+      "epoch": 1.640500568828214,
+      "grad_norm": 0.6349069920640864,
+      "learning_rate": 3.3978105647944e-05,
+      "loss": 0.5174,
+      "step": 721
+    },
+    {
+      "epoch": 1.6427758816837317,
+      "grad_norm": 0.6791512167624842,
+      "learning_rate": 3.3955334436233145e-05,
+      "loss": 0.4595,
+      "step": 722
+    },
+    {
+      "epoch": 1.6450511945392492,
+      "grad_norm": 0.5564269917656663,
+      "learning_rate": 3.393252791387306e-05,
+      "loss": 0.397,
+      "step": 723
+    },
+    {
+      "epoch": 1.6473265073947667,
+      "grad_norm": 0.5391279772289733,
+      "learning_rate": 3.3909686138570226e-05,
+      "loss": 0.4144,
+      "step": 724
+    },
+    {
+      "epoch": 1.6496018202502845,
+      "grad_norm": 0.6738091867356918,
+      "learning_rate": 3.388680916812031e-05,
+      "loss": 0.4434,
+      "step": 725
+    },
+    {
+      "epoch": 1.6518771331058022,
+      "grad_norm": 0.6126156918760496,
+      "learning_rate": 3.3863897060408036e-05,
+      "loss": 0.4501,
+      "step": 726
+    },
+    {
+      "epoch": 1.6541524459613197,
+      "grad_norm": 0.5590786991354324,
+      "learning_rate": 3.384094987340703e-05,
+      "loss": 0.4424,
+      "step": 727
+    },
+    {
+      "epoch": 1.6564277588168372,
+      "grad_norm": 0.5726606756254216,
+      "learning_rate": 3.3817967665179687e-05,
+      "loss": 0.4023,
+      "step": 728
+    },
+    {
+      "epoch": 1.658703071672355,
+      "grad_norm": 0.5308251504814927,
+      "learning_rate": 3.3794950493877014e-05,
+      "loss": 0.4373,
+      "step": 729
+    },
+    {
+      "epoch": 1.6609783845278727,
+      "grad_norm": 0.48318908441400044,
+      "learning_rate": 3.377189841773848e-05,
+      "loss": 0.4133,
+      "step": 730
+    },
+    {
+      "epoch": 1.6632536973833902,
+      "grad_norm": 0.6708760404159205,
+      "learning_rate": 3.374881149509186e-05,
+      "loss": 0.4377,
+      "step": 731
+    },
+    {
+      "epoch": 1.6655290102389078,
+      "grad_norm": 0.5173893189818557,
+      "learning_rate": 3.372568978435312e-05,
+      "loss": 0.4582,
+      "step": 732
+    },
+    {
+      "epoch": 1.6678043230944255,
+      "grad_norm": 0.5770282850563427,
+      "learning_rate": 3.370253334402624e-05,
+      "loss": 0.3747,
+      "step": 733
+    },
+    {
+      "epoch": 1.6700796359499432,
+      "grad_norm": 0.6804621006166472,
+      "learning_rate": 3.367934223270308e-05,
+      "loss": 0.4032,
+      "step": 734
+    },
+    {
+      "epoch": 1.6723549488054608,
+      "grad_norm": 0.5324344157986772,
+      "learning_rate": 3.365611650906321e-05,
+      "loss": 0.4575,
+      "step": 735
+    },
+    {
+      "epoch": 1.6746302616609783,
+      "grad_norm": 0.7892336868643832,
+      "learning_rate": 3.3632856231873806e-05,
+      "loss": 0.4448,
+      "step": 736
+    },
+    {
+      "epoch": 1.676905574516496,
+      "grad_norm": 0.4660395278953551,
+      "learning_rate": 3.3609561459989446e-05,
+      "loss": 0.3867,
+      "step": 737
+    },
+    {
+      "epoch": 1.6791808873720138,
+      "grad_norm": 0.6220290351552069,
+      "learning_rate": 3.358623225235201e-05,
+      "loss": 0.3934,
+      "step": 738
+    },
+    {
+      "epoch": 1.6814562002275313,
+      "grad_norm": 0.5003423448965053,
+      "learning_rate": 3.3562868667990487e-05,
+      "loss": 0.3917,
+      "step": 739
+    },
+    {
+      "epoch": 1.6837315130830488,
+      "grad_norm": 0.5023405186553996,
+      "learning_rate": 3.353947076602088e-05,
+      "loss": 0.4664,
+      "step": 740
+    },
+    {
+      "epoch": 1.6860068259385665,
+      "grad_norm": 0.4546077197981959,
+      "learning_rate": 3.3516038605645985e-05,
+      "loss": 0.432,
+      "step": 741
+    },
+    {
+      "epoch": 1.6882821387940843,
+      "grad_norm": 0.5142825872996363,
+      "learning_rate": 3.349257224615532e-05,
+      "loss": 0.4528,
+      "step": 742
+    },
+    {
+      "epoch": 1.6905574516496018,
+      "grad_norm": 0.5261812581068472,
+      "learning_rate": 3.346907174692491e-05,
+      "loss": 0.382,
+      "step": 743
+    },
+    {
+      "epoch": 1.6928327645051193,
+      "grad_norm": 0.48306162195031704,
+      "learning_rate": 3.3445537167417165e-05,
+      "loss": 0.4824,
+      "step": 744
+    },
+    {
+      "epoch": 1.695108077360637,
+      "grad_norm": 0.5253247800677288,
+      "learning_rate": 3.342196856718074e-05,
+      "loss": 0.4223,
+      "step": 745
+    },
+    {
+      "epoch": 1.6973833902161548,
+      "grad_norm": 0.501305461897868,
+      "learning_rate": 3.339836600585036e-05,
+      "loss": 0.4359,
+      "step": 746
+    },
+    {
+      "epoch": 1.6996587030716723,
+      "grad_norm": 0.4732093099233727,
+      "learning_rate": 3.3374729543146674e-05,
+      "loss": 0.3959,
+      "step": 747
+    },
+    {
+      "epoch": 1.7019340159271898,
+      "grad_norm": 0.4478272616866317,
+      "learning_rate": 3.335105923887614e-05,
+      "loss": 0.4071,
+      "step": 748
+    },
+    {
+      "epoch": 1.7042093287827076,
+      "grad_norm": 0.48865136328489367,
+      "learning_rate": 3.332735515293081e-05,
+      "loss": 0.3944,
+      "step": 749
+    },
+    {
+      "epoch": 1.7064846416382253,
+      "grad_norm": 0.5259957045989833,
+      "learning_rate": 3.330361734528823e-05,
+      "loss": 0.4227,
+      "step": 750
+    },
+    {
+      "epoch": 1.7087599544937428,
+      "grad_norm": 0.5497563708361985,
+      "learning_rate": 3.327984587601127e-05,
+      "loss": 0.4332,
+      "step": 751
+    },
+    {
+      "epoch": 1.7110352673492604,
+      "grad_norm": 0.5843800718238974,
+      "learning_rate": 3.325604080524796e-05,
+      "loss": 0.4286,
+      "step": 752
+    },
+    {
+      "epoch": 1.713310580204778,
+      "grad_norm": 0.45291954410142116,
+      "learning_rate": 3.323220219323137e-05,
+      "loss": 0.4476,
+      "step": 753
+    },
+    {
+      "epoch": 1.7155858930602959,
+      "grad_norm": 0.5160193476732419,
+      "learning_rate": 3.3208330100279425e-05,
+      "loss": 0.3927,
+      "step": 754
+    },
+    {
+      "epoch": 1.7178612059158134,
+      "grad_norm": 0.4931895088903224,
+      "learning_rate": 3.318442458679477e-05,
+      "loss": 0.4866,
+      "step": 755
+    },
+    {
+      "epoch": 1.7201365187713311,
+      "grad_norm": 0.5236604180952761,
+      "learning_rate": 3.316048571326461e-05,
+      "loss": 0.3987,
+      "step": 756
+    },
+    {
+      "epoch": 1.7224118316268489,
+      "grad_norm": 0.514426955191737,
+      "learning_rate": 3.313651354026057e-05,
+      "loss": 0.4476,
+      "step": 757
+    },
+    {
+      "epoch": 1.7246871444823664,
+      "grad_norm": 0.5196465533309423,
+      "learning_rate": 3.311250812843851e-05,
+      "loss": 0.5314,
+      "step": 758
+    },
+    {
+      "epoch": 1.726962457337884,
+      "grad_norm": 0.5443372341119862,
+      "learning_rate": 3.308846953853842e-05,
+      "loss": 0.4224,
+      "step": 759
+    },
+    {
+      "epoch": 1.7292377701934016,
+      "grad_norm": 0.5250443676651882,
+      "learning_rate": 3.306439783138421e-05,
+      "loss": 0.4563,
+      "step": 760
+    },
+    {
+      "epoch": 1.7315130830489194,
+      "grad_norm": 0.4465626870840488,
+      "learning_rate": 3.304029306788361e-05,
+      "loss": 0.4088,
+      "step": 761
+    },
+    {
+      "epoch": 1.733788395904437,
+      "grad_norm": 0.6510435095336619,
+      "learning_rate": 3.3016155309027985e-05,
+      "loss": 0.4655,
+      "step": 762
+    },
+    {
+      "epoch": 1.7360637087599544,
+      "grad_norm": 0.38834294904495403,
+      "learning_rate": 3.299198461589217e-05,
+      "loss": 0.4174,
+      "step": 763
+    },
+    {
+      "epoch": 1.7383390216154722,
+      "grad_norm": 0.5264237824600192,
+      "learning_rate": 3.2967781049634356e-05,
+      "loss": 0.3874,
+      "step": 764
+    },
+    {
+      "epoch": 1.74061433447099,
+      "grad_norm": 0.5879282657629377,
+      "learning_rate": 3.29435446714959e-05,
+      "loss": 0.4231,
+      "step": 765
+    },
+    {
+      "epoch": 1.7428896473265074,
+      "grad_norm": 0.5393876135695119,
+      "learning_rate": 3.291927554280118e-05,
+      "loss": 0.4339,
+      "step": 766
+    },
+    {
+      "epoch": 1.745164960182025,
+      "grad_norm": 0.49399211357173956,
+      "learning_rate": 3.289497372495744e-05,
+      "loss": 0.4163,
+      "step": 767
+    },
+    {
+      "epoch": 1.7474402730375427,
+      "grad_norm": 0.47019460412562336,
+      "learning_rate": 3.2870639279454665e-05,
+      "loss": 0.4354,
+      "step": 768
+    },
+    {
+      "epoch": 1.7497155858930604,
+      "grad_norm": 0.5685785048850405,
+      "learning_rate": 3.284627226786534e-05,
+      "loss": 0.4498,
+      "step": 769
+    },
+    {
+      "epoch": 1.751990898748578,
+      "grad_norm": 0.4753852695722544,
+      "learning_rate": 3.28218727518444e-05,
+      "loss": 0.3912,
+      "step": 770
+    },
+    {
+      "epoch": 1.7542662116040955,
+      "grad_norm": 0.562602566827557,
+      "learning_rate": 3.279744079312901e-05,
+      "loss": 0.4488,
+      "step": 771
+    },
+    {
+      "epoch": 1.7565415244596132,
+      "grad_norm": 0.5094239312668742,
+      "learning_rate": 3.277297645353842e-05,
+      "loss": 0.3791,
+      "step": 772
+    },
+    {
+      "epoch": 1.758816837315131,
+      "grad_norm": 0.620494439748899,
+      "learning_rate": 3.27484797949738e-05,
+      "loss": 0.3924,
+      "step": 773
+    },
+    {
+      "epoch": 1.7610921501706485,
+      "grad_norm": 0.6628863050037046,
+      "learning_rate": 3.272395087941812e-05,
+      "loss": 0.4488,
+      "step": 774
+    },
+    {
+      "epoch": 1.763367463026166,
+      "grad_norm": 0.637786898326395,
+      "learning_rate": 3.269938976893595e-05,
+      "loss": 0.4429,
+      "step": 775
+    },
+    {
+      "epoch": 1.7656427758816837,
+      "grad_norm": 0.510130999535126,
+      "learning_rate": 3.267479652567334e-05,
+      "loss": 0.4216,
+      "step": 776
+    },
+    {
+      "epoch": 1.7679180887372015,
+      "grad_norm": 0.5586834514509081,
+      "learning_rate": 3.2650171211857626e-05,
+      "loss": 0.4123,
+      "step": 777
+    },
+    {
+      "epoch": 1.770193401592719,
+      "grad_norm": 0.5411046048656755,
+      "learning_rate": 3.2625513889797307e-05,
+      "loss": 0.4633,
+      "step": 778
+    },
+    {
+      "epoch": 1.7724687144482365,
+      "grad_norm": 0.5448383040301249,
+      "learning_rate": 3.2600824621881846e-05,
+      "loss": 0.3844,
+      "step": 779
+    },
+    {
+      "epoch": 1.7747440273037542,
+      "grad_norm": 0.589811003209115,
+      "learning_rate": 3.2576103470581564e-05,
+      "loss": 0.4563,
+      "step": 780
+    },
+    {
+      "epoch": 1.777019340159272,
+      "grad_norm": 0.5584343787626713,
+      "learning_rate": 3.2551350498447446e-05,
+      "loss": 0.4303,
+      "step": 781
+    },
+    {
+      "epoch": 1.7792946530147895,
+      "grad_norm": 0.4810202970686725,
+      "learning_rate": 3.252656576811099e-05,
+      "loss": 0.3706,
+      "step": 782
+    },
+    {
+      "epoch": 1.781569965870307,
+      "grad_norm": 0.6156367306416939,
+      "learning_rate": 3.2501749342284044e-05,
+      "loss": 0.4231,
+      "step": 783
+    },
+    {
+      "epoch": 1.7838452787258248,
+      "grad_norm": 0.5204945060263102,
+      "learning_rate": 3.247690128375867e-05,
+      "loss": 0.4114,
+      "step": 784
+    },
+    {
+      "epoch": 1.7861205915813425,
+      "grad_norm": 0.7271145649976276,
+      "learning_rate": 3.245202165540697e-05,
+      "loss": 0.4466,
+      "step": 785
+    },
+    {
+      "epoch": 1.78839590443686,
+      "grad_norm": 0.4763873903676693,
+      "learning_rate": 3.24271105201809e-05,
+      "loss": 0.4805,
+      "step": 786
+    },
+    {
+      "epoch": 1.7906712172923775,
+      "grad_norm": 0.5144449537309139,
+      "learning_rate": 3.240216794111215e-05,
+      "loss": 0.374,
+      "step": 787
+    },
+    {
+      "epoch": 1.7929465301478953,
+      "grad_norm": 0.4718447371813775,
+      "learning_rate": 3.237719398131198e-05,
+      "loss": 0.4204,
+      "step": 788
+    },
+    {
+      "epoch": 1.795221843003413,
+      "grad_norm": 0.47782509060055806,
+      "learning_rate": 3.235218870397105e-05,
+      "loss": 0.4451,
+      "step": 789
+    },
+    {
+      "epoch": 1.7974971558589306,
+      "grad_norm": 0.5025689132907557,
+      "learning_rate": 3.232715217235927e-05,
+      "loss": 0.4412,
+      "step": 790
+    },
+    {
+      "epoch": 1.799772468714448,
+      "grad_norm": 0.4952509085855213,
+      "learning_rate": 3.2302084449825595e-05,
+      "loss": 0.4193,
+      "step": 791
+    },
+    {
+      "epoch": 1.802047781569966,
+      "grad_norm": 0.4720298981099759,
+      "learning_rate": 3.227698559979794e-05,
+      "loss": 0.3944,
+      "step": 792
+    },
+    {
+      "epoch": 1.8043230944254836,
+      "grad_norm": 0.6208004955993673,
+      "learning_rate": 3.225185568578295e-05,
+      "loss": 0.4374,
+      "step": 793
+    },
+    {
+      "epoch": 1.806598407281001,
+      "grad_norm": 0.43207309687869283,
+      "learning_rate": 3.2226694771365906e-05,
+      "loss": 0.3812,
+      "step": 794
+    },
+    {
+      "epoch": 1.8088737201365188,
+      "grad_norm": 0.6654873071833073,
+      "learning_rate": 3.220150292021049e-05,
+      "loss": 0.4411,
+      "step": 795
+    },
+    {
+      "epoch": 1.8111490329920366,
+      "grad_norm": 0.5760081189256475,
+      "learning_rate": 3.217628019605869e-05,
+      "loss": 0.4424,
+      "step": 796
+    },
+    {
+      "epoch": 1.813424345847554,
+      "grad_norm": 0.5018027239654591,
+      "learning_rate": 3.21510266627306e-05,
+      "loss": 0.4175,
+      "step": 797
+    },
+    {
+      "epoch": 1.8156996587030716,
+      "grad_norm": 0.539232908897014,
+      "learning_rate": 3.212574238412427e-05,
+      "loss": 0.443,
+      "step": 798
+    },
+    {
+      "epoch": 1.8179749715585893,
+      "grad_norm": 0.8883330790380443,
+      "learning_rate": 3.2100427424215536e-05,
+      "loss": 0.451,
+      "step": 799
+    },
+    {
+      "epoch": 1.820250284414107,
+      "grad_norm": 0.4607329856252554,
+      "learning_rate": 3.2075081847057886e-05,
+      "loss": 0.4077,
+      "step": 800
+    },
+    {
+      "epoch": 1.8225255972696246,
+      "grad_norm": 0.45881098461656095,
+      "learning_rate": 3.204970571678225e-05,
+      "loss": 0.3898,
+      "step": 801
+    },
+    {
+      "epoch": 1.8248009101251421,
+      "grad_norm": 0.4950046026315809,
+      "learning_rate": 3.20242990975969e-05,
+      "loss": 0.4913,
+      "step": 802
+    },
+    {
+      "epoch": 1.8270762229806599,
+      "grad_norm": 0.4284019412488292,
+      "learning_rate": 3.1998862053787214e-05,
+      "loss": 0.3605,
+      "step": 803
+    },
+    {
+      "epoch": 1.8293515358361776,
+      "grad_norm": 0.523837409188779,
+      "learning_rate": 3.197339464971558e-05,
+      "loss": 0.4482,
+      "step": 804
+    },
+    {
+      "epoch": 1.8316268486916951,
+      "grad_norm": 0.49142234581824934,
+      "learning_rate": 3.194789694982119e-05,
+      "loss": 0.4118,
+      "step": 805
+    },
+    {
+      "epoch": 1.8339021615472126,
+      "grad_norm": 0.49685520725633503,
+      "learning_rate": 3.192236901861992e-05,
+      "loss": 0.4309,
+      "step": 806
+    },
+    {
+      "epoch": 1.8361774744027304,
+      "grad_norm": 0.5783313475386348,
+      "learning_rate": 3.1896810920704095e-05,
+      "loss": 0.4543,
+      "step": 807
+    },
+    {
+      "epoch": 1.8384527872582481,
+      "grad_norm": 0.4663307446637632,
+      "learning_rate": 3.18712227207424e-05,
+      "loss": 0.3722,
+      "step": 808
+    },
+    {
+      "epoch": 1.8407281001137656,
+      "grad_norm": 0.5581189544180488,
+      "learning_rate": 3.184560448347969e-05,
+      "loss": 0.4143,
+      "step": 809
+    },
+    {
+      "epoch": 1.8430034129692832,
+      "grad_norm": 0.5691345254440217,
+      "learning_rate": 3.181995627373679e-05,
+      "loss": 0.4441,
+      "step": 810
+    },
+    {
+      "epoch": 1.845278725824801,
+      "grad_norm": 0.5066854792282622,
+      "learning_rate": 3.179427815641041e-05,
+      "loss": 0.3638,
+      "step": 811
+    },
+    {
+      "epoch": 1.8475540386803186,
+      "grad_norm": 0.7011362779760515,
+      "learning_rate": 3.176857019647289e-05,
+      "loss": 0.4866,
+      "step": 812
+    },
+    {
+      "epoch": 1.8498293515358362,
+      "grad_norm": 0.6094416012430611,
+      "learning_rate": 3.17428324589721e-05,
+      "loss": 0.4246,
+      "step": 813
+    },
+    {
+      "epoch": 1.8521046643913537,
+      "grad_norm": 0.48464772590754135,
+      "learning_rate": 3.171706500903126e-05,
+      "loss": 0.4035,
+      "step": 814
+    },
+    {
+      "epoch": 1.8543799772468714,
+      "grad_norm": 0.5786053028777935,
+      "learning_rate": 3.1691267911848765e-05,
+      "loss": 0.3657,
+      "step": 815
+    },
+    {
+      "epoch": 1.8566552901023892,
+      "grad_norm": 0.5861507724160404,
+      "learning_rate": 3.166544123269801e-05,
+      "loss": 0.4289,
+      "step": 816
+    },
+    {
+      "epoch": 1.8589306029579067,
+      "grad_norm": 0.6029563372781788,
+      "learning_rate": 3.163958503692727e-05,
+      "loss": 0.4721,
+      "step": 817
+    },
+    {
+      "epoch": 1.8612059158134242,
+      "grad_norm": 0.6129610821268773,
+      "learning_rate": 3.1613699389959484e-05,
+      "loss": 0.3952,
+      "step": 818
+    },
+    {
+      "epoch": 1.863481228668942,
+      "grad_norm": 0.7662158217676397,
+      "learning_rate": 3.158778435729211e-05,
+      "loss": 0.4303,
+      "step": 819
+    },
+    {
+      "epoch": 1.8657565415244597,
+      "grad_norm": 0.5796796368147477,
+      "learning_rate": 3.156184000449697e-05,
+      "loss": 0.4499,
+      "step": 820
+    },
+    {
+      "epoch": 1.8680318543799772,
+      "grad_norm": 0.7996782495584651,
+      "learning_rate": 3.153586639722006e-05,
+      "loss": 0.4245,
+      "step": 821
+    },
+    {
+      "epoch": 1.8703071672354947,
+      "grad_norm": 0.43525007544954886,
+      "learning_rate": 3.1509863601181424e-05,
+      "loss": 0.3829,
+      "step": 822
+    },
+    {
+      "epoch": 1.8725824800910125,
+      "grad_norm": 0.7527664756780577,
+      "learning_rate": 3.1483831682174914e-05,
+      "loss": 0.4542,
+      "step": 823
+    },
+    {
+      "epoch": 1.8748577929465302,
+      "grad_norm": 0.46604172701732616,
+      "learning_rate": 3.1457770706068125e-05,
+      "loss": 0.4255,
+      "step": 824
+    },
+    {
+      "epoch": 1.8771331058020477,
+      "grad_norm": 0.621671324916588,
+      "learning_rate": 3.143168073880214e-05,
+      "loss": 0.4562,
+      "step": 825
+    },
+    {
+      "epoch": 1.8794084186575652,
+      "grad_norm": 0.5812533736282571,
+      "learning_rate": 3.14055618463914e-05,
+      "loss": 0.4408,
+      "step": 826
+    },
+    {
+      "epoch": 1.8816837315130832,
+      "grad_norm": 0.48593039499357055,
+      "learning_rate": 3.1379414094923545e-05,
+      "loss": 0.3911,
+      "step": 827
+    },
+    {
+      "epoch": 1.8839590443686007,
+      "grad_norm": 0.5222388909677681,
+      "learning_rate": 3.135323755055923e-05,
+      "loss": 0.4699,
+      "step": 828
+    },
+    {
+      "epoch": 1.8862343572241183,
+      "grad_norm": 0.6048601049513829,
+      "learning_rate": 3.1327032279531966e-05,
+      "loss": 0.4389,
+      "step": 829
+    },
+    {
+      "epoch": 1.888509670079636,
+      "grad_norm": 0.49919715491157246,
+      "learning_rate": 3.1300798348147954e-05,
+      "loss": 0.4262,
+      "step": 830
+    },
+    {
+      "epoch": 1.8907849829351537,
+      "grad_norm": 0.6678286159773057,
+      "learning_rate": 3.127453582278589e-05,
+      "loss": 0.4593,
+      "step": 831
+    },
+    {
+      "epoch": 1.8930602957906713,
+      "grad_norm": 0.5906564126737536,
+      "learning_rate": 3.124824476989686e-05,
+      "loss": 0.3977,
+      "step": 832
+    },
+    {
+      "epoch": 1.8953356086461888,
+      "grad_norm": 0.5832034135440843,
+      "learning_rate": 3.122192525600409e-05,
+      "loss": 0.463,
+      "step": 833
+    },
+    {
+      "epoch": 1.8976109215017065,
+      "grad_norm": 0.7135832527092052,
+      "learning_rate": 3.119557734770285e-05,
+      "loss": 0.3795,
+      "step": 834
+    },
+    {
+      "epoch": 1.8998862343572243,
+      "grad_norm": 0.5621917499754232,
+      "learning_rate": 3.116920111166025e-05,
+      "loss": 0.4112,
+      "step": 835
+    },
+    {
+      "epoch": 1.9021615472127418,
+      "grad_norm": 0.7181751851282073,
+      "learning_rate": 3.114279661461506e-05,
+      "loss": 0.3811,
+      "step": 836
+    },
+    {
+      "epoch": 1.9044368600682593,
+      "grad_norm": 0.6743148704065208,
+      "learning_rate": 3.1116363923377576e-05,
+      "loss": 0.447,
+      "step": 837
+    },
+    {
+      "epoch": 1.906712172923777,
+      "grad_norm": 0.7349281938472108,
+      "learning_rate": 3.108990310482943e-05,
+      "loss": 0.4133,
+      "step": 838
+    },
+    {
+      "epoch": 1.9089874857792948,
+      "grad_norm": 0.6436988476892611,
+      "learning_rate": 3.1063414225923416e-05,
+      "loss": 0.4606,
+      "step": 839
+    },
+    {
+      "epoch": 1.9112627986348123,
+      "grad_norm": 1.0022956970748371,
+      "learning_rate": 3.103689735368333e-05,
+      "loss": 0.4597,
+      "step": 840
+    },
+    {
+      "epoch": 1.9135381114903298,
+      "grad_norm": 0.5116706876398635,
+      "learning_rate": 3.10103525552038e-05,
+      "loss": 0.4274,
+      "step": 841
+    },
+    {
+      "epoch": 1.9158134243458476,
+      "grad_norm": 0.8062209245544874,
+      "learning_rate": 3.098377989765011e-05,
+      "loss": 0.3849,
+      "step": 842
+    },
+    {
+      "epoch": 1.9180887372013653,
+      "grad_norm": 0.5590307923149567,
+      "learning_rate": 3.0957179448258053e-05,
+      "loss": 0.4467,
+      "step": 843
+    },
+    {
+      "epoch": 1.9203640500568828,
+      "grad_norm": 0.7522537029457251,
+      "learning_rate": 3.0930551274333715e-05,
+      "loss": 0.4432,
+      "step": 844
+    },
+    {
+      "epoch": 1.9226393629124003,
+      "grad_norm": 0.5051335750383762,
+      "learning_rate": 3.090389544325335e-05,
+      "loss": 0.3911,
+      "step": 845
+    },
+    {
+      "epoch": 1.924914675767918,
+      "grad_norm": 0.7913606783246964,
+      "learning_rate": 3.087721202246321e-05,
+      "loss": 0.4444,
+      "step": 846
+    },
+    {
+      "epoch": 1.9271899886234358,
+      "grad_norm": 0.4753111648665716,
+      "learning_rate": 3.08505010794793e-05,
+      "loss": 0.4163,
+      "step": 847
+    },
+    {
+      "epoch": 1.9294653014789533,
+      "grad_norm": 0.7035924858849998,
+      "learning_rate": 3.082376268188731e-05,
+      "loss": 0.4377,
+      "step": 848
+    },
+    {
+      "epoch": 1.9317406143344709,
+      "grad_norm": 0.5117975652669959,
+      "learning_rate": 3.0796996897342394e-05,
+      "loss": 0.4447,
+      "step": 849
+    },
+    {
+      "epoch": 1.9340159271899886,
+      "grad_norm": 0.5513898035344333,
+      "learning_rate": 3.0770203793568994e-05,
+      "loss": 0.3837,
+      "step": 850
+    },
+    {
+      "epoch": 1.9362912400455063,
+      "grad_norm": 0.4764564187002147,
+      "learning_rate": 3.0743383438360666e-05,
+      "loss": 0.4389,
+      "step": 851
+    },
+    {
+      "epoch": 1.9385665529010239,
+      "grad_norm": 0.5121515231941229,
+      "learning_rate": 3.0716535899579936e-05,
+      "loss": 0.417,
+      "step": 852
+    },
+    {
+      "epoch": 1.9408418657565414,
+      "grad_norm": 0.7254375435748747,
+      "learning_rate": 3.06896612451581e-05,
+      "loss": 0.4904,
+      "step": 853
+    },
+    {
+      "epoch": 1.9431171786120591,
+      "grad_norm": 0.5312185931920812,
+      "learning_rate": 3.066275954309507e-05,
+      "loss": 0.4833,
+      "step": 854
+    },
+    {
+      "epoch": 1.9453924914675769,
+      "grad_norm": 0.3906691207947801,
+      "learning_rate": 3.0635830861459204e-05,
+      "loss": 0.4172,
+      "step": 855
+    },
+    {
+      "epoch": 1.9476678043230944,
+      "grad_norm": 0.5490567237256371,
+      "learning_rate": 3.06088752683871e-05,
+      "loss": 0.4367,
+      "step": 856
+    },
+    {
+      "epoch": 1.949943117178612,
+      "grad_norm": 0.48830649429397904,
+      "learning_rate": 3.0581892832083484e-05,
+      "loss": 0.4098,
+      "step": 857
+    },
+    {
+      "epoch": 1.9522184300341296,
+      "grad_norm": 0.5668417266386482,
+      "learning_rate": 3.0554883620820955e-05,
+      "loss": 0.4232,
+      "step": 858
+    },
+    {
+      "epoch": 1.9544937428896474,
+      "grad_norm": 0.5490790985251669,
+      "learning_rate": 3.0527847702939915e-05,
+      "loss": 0.4491,
+      "step": 859
+    },
+    {
+      "epoch": 1.956769055745165,
+      "grad_norm": 0.6622999931743726,
+      "learning_rate": 3.0500785146848303e-05,
+      "loss": 0.486,
+      "step": 860
+    },
+    {
+      "epoch": 1.9590443686006824,
+      "grad_norm": 0.5954178521700658,
+      "learning_rate": 3.0473696021021466e-05,
+      "loss": 0.4428,
+      "step": 861
+    },
+    {
+      "epoch": 1.9613196814562004,
+      "grad_norm": 0.5210378657968507,
+      "learning_rate": 3.0446580394001986e-05,
+      "loss": 0.3706,
+      "step": 862
+    },
+    {
+      "epoch": 1.963594994311718,
+      "grad_norm": 0.47230222545638934,
+      "learning_rate": 3.04194383343995e-05,
+      "loss": 0.4372,
+      "step": 863
+    },
+    {
+      "epoch": 1.9658703071672354,
+      "grad_norm": 0.550474067886411,
+      "learning_rate": 3.0392269910890523e-05,
+      "loss": 0.4556,
+      "step": 864
+    },
+    {
+      "epoch": 1.9681456200227532,
+      "grad_norm": 0.5540535643108356,
+      "learning_rate": 3.0365075192218278e-05,
+      "loss": 0.4529,
+      "step": 865
+    },
+    {
+      "epoch": 1.970420932878271,
+      "grad_norm": 0.4896945623597814,
+      "learning_rate": 3.033785424719252e-05,
+      "loss": 0.3787,
+      "step": 866
+    },
+    {
+      "epoch": 1.9726962457337884,
+      "grad_norm": 0.5500626404394998,
+      "learning_rate": 3.0310607144689372e-05,
+      "loss": 0.4299,
+      "step": 867
+    },
+    {
+      "epoch": 1.974971558589306,
+      "grad_norm": 0.5546330084531744,
+      "learning_rate": 3.0283333953651124e-05,
+      "loss": 0.4132,
+      "step": 868
+    },
+    {
+      "epoch": 1.9772468714448237,
+      "grad_norm": 0.6542031906231642,
+      "learning_rate": 3.0256034743086103e-05,
+      "loss": 0.461,
+      "step": 869
+    },
+    {
+      "epoch": 1.9795221843003414,
+      "grad_norm": 0.46270201346529116,
+      "learning_rate": 3.022870958206845e-05,
+      "loss": 0.366,
+      "step": 870
+    },
+    {
+      "epoch": 1.981797497155859,
+      "grad_norm": 0.6868482518027066,
+      "learning_rate": 3.0201358539737976e-05,
+      "loss": 0.4794,
+      "step": 871
+    },
+    {
+      "epoch": 1.9840728100113765,
+      "grad_norm": 0.48697359161478826,
+      "learning_rate": 3.017398168529997e-05,
+      "loss": 0.3659,
+      "step": 872
+    },
+    {
+      "epoch": 1.9863481228668942,
+      "grad_norm": 0.7213911255238944,
+      "learning_rate": 3.0146579088025053e-05,
+      "loss": 0.4517,
+      "step": 873
+    },
+    {
+      "epoch": 1.988623435722412,
+      "grad_norm": 0.5315352661478399,
+      "learning_rate": 3.0119150817248966e-05,
+      "loss": 0.459,
+      "step": 874
+    },
+    {
+      "epoch": 1.9908987485779295,
+      "grad_norm": 0.5241257258878973,
+      "learning_rate": 3.0091696942372412e-05,
+      "loss": 0.4103,
+      "step": 875
+    },
+    {
+      "epoch": 1.993174061433447,
+      "grad_norm": 0.5019935569702207,
+      "learning_rate": 3.006421753286088e-05,
+      "loss": 0.4114,
+      "step": 876
+    },
+    {
+      "epoch": 1.9954493742889647,
+      "grad_norm": 0.5572188267356125,
+      "learning_rate": 3.0036712658244475e-05,
+      "loss": 0.4606,
+      "step": 877
+    },
+    {
+      "epoch": 1.9977246871444825,
+      "grad_norm": 0.4729138466778679,
+      "learning_rate": 3.0009182388117718e-05,
+      "loss": 0.4495,
+      "step": 878
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8733140039301539,
+      "learning_rate": 2.9981626792139414e-05,
+      "loss": 0.6445,
+      "step": 879
+    },
+    {
+      "epoch": 2.0022753128555175,
+      "grad_norm": 0.7126446895469605,
+      "learning_rate": 2.9954045940032423e-05,
+      "loss": 0.3322,
+      "step": 880
+    },
+    {
+      "epoch": 2.0045506257110355,
+      "grad_norm": 0.5799465206761989,
+      "learning_rate": 2.9926439901583524e-05,
+      "loss": 0.3348,
+      "step": 881
+    },
+    {
+      "epoch": 2.006825938566553,
+      "grad_norm": 0.7521511832974017,
+      "learning_rate": 2.9898808746643223e-05,
+      "loss": 0.3437,
+      "step": 882
+    },
+    {
+      "epoch": 2.0091012514220705,
+      "grad_norm": 0.634315449511539,
+      "learning_rate": 2.987115254512557e-05,
+      "loss": 0.3449,
+      "step": 883
+    },
+    {
+      "epoch": 2.011376564277588,
+      "grad_norm": 0.6844751831529167,
+      "learning_rate": 2.9843471367008003e-05,
+      "loss": 0.3331,
+      "step": 884
+    },
+    {
+      "epoch": 2.013651877133106,
+      "grad_norm": 0.4814286536478195,
+      "learning_rate": 2.981576528233114e-05,
+      "loss": 0.3264,
+      "step": 885
+    },
+    {
+      "epoch": 2.0159271899886235,
+      "grad_norm": 0.5467934829900717,
+      "learning_rate": 2.9788034361198633e-05,
+      "loss": 0.2875,
+      "step": 886
+    },
+    {
+      "epoch": 2.018202502844141,
+      "grad_norm": 0.5431794266712123,
+      "learning_rate": 2.976027867377697e-05,
+      "loss": 0.3433,
+      "step": 887
+    },
+    {
+      "epoch": 2.0204778156996586,
+      "grad_norm": 0.5683506352212406,
+      "learning_rate": 2.9732498290295296e-05,
+      "loss": 0.3436,
+      "step": 888
+    },
+    {
+      "epoch": 2.0227531285551765,
+      "grad_norm": 0.5409437069222102,
+      "learning_rate": 2.9704693281045272e-05,
+      "loss": 0.348,
+      "step": 889
+    },
+    {
+      "epoch": 2.025028441410694,
+      "grad_norm": 0.5595974396490498,
+      "learning_rate": 2.9676863716380845e-05,
+      "loss": 0.3002,
+      "step": 890
+    },
+    {
+      "epoch": 2.0273037542662116,
+      "grad_norm": 0.5072666611966352,
+      "learning_rate": 2.9649009666718086e-05,
+      "loss": 0.3334,
+      "step": 891
+    },
+    {
+      "epoch": 2.029579067121729,
+      "grad_norm": 0.44330222985587886,
+      "learning_rate": 2.9621131202535042e-05,
+      "loss": 0.308,
+      "step": 892
+    },
+    {
+      "epoch": 2.031854379977247,
+      "grad_norm": 0.5306386567420092,
+      "learning_rate": 2.9593228394371527e-05,
+      "loss": 0.3104,
+      "step": 893
+    },
+    {
+      "epoch": 2.0341296928327646,
+      "grad_norm": 0.4019692515169545,
+      "learning_rate": 2.9565301312828945e-05,
+      "loss": 0.3237,
+      "step": 894
+    },
+    {
+      "epoch": 2.036405005688282,
+      "grad_norm": 0.5227451720787013,
+      "learning_rate": 2.9537350028570126e-05,
+      "loss": 0.3499,
+      "step": 895
+    },
+    {
+      "epoch": 2.0386803185437996,
+      "grad_norm": 0.42757714521426105,
+      "learning_rate": 2.950937461231913e-05,
+      "loss": 0.3273,
+      "step": 896
+    },
+    {
+      "epoch": 2.0409556313993176,
+      "grad_norm": 0.47745497100808654,
+      "learning_rate": 2.948137513486109e-05,
+      "loss": 0.3053,
+      "step": 897
+    },
+    {
+      "epoch": 2.043230944254835,
+      "grad_norm": 0.5325164876648694,
+      "learning_rate": 2.9453351667042008e-05,
+      "loss": 0.3448,
+      "step": 898
+    },
+    {
+      "epoch": 2.0455062571103526,
+      "grad_norm": 0.47210366776632884,
+      "learning_rate": 2.94253042797686e-05,
+      "loss": 0.293,
+      "step": 899
+    },
+    {
+      "epoch": 2.04778156996587,
+      "grad_norm": 0.466854126612842,
+      "learning_rate": 2.9397233044008092e-05,
+      "loss": 0.2955,
+      "step": 900
+    },
+    {
+      "epoch": 2.050056882821388,
+      "grad_norm": 0.5005964391349274,
+      "learning_rate": 2.936913803078806e-05,
+      "loss": 0.3343,
+      "step": 901
+    },
+    {
+      "epoch": 2.0523321956769056,
+      "grad_norm": 0.4651126962723272,
+      "learning_rate": 2.9341019311196247e-05,
+      "loss": 0.3216,
+      "step": 902
+    },
+    {
+      "epoch": 2.054607508532423,
+      "grad_norm": 0.5278052326421759,
+      "learning_rate": 2.931287695638037e-05,
+      "loss": 0.3174,
+      "step": 903
+    },
+    {
+      "epoch": 2.0568828213879407,
+      "grad_norm": 0.4055235656855877,
+      "learning_rate": 2.9284711037547945e-05,
+      "loss": 0.3232,
+      "step": 904
+    },
+    {
+      "epoch": 2.0591581342434586,
+      "grad_norm": 0.557342098786895,
+      "learning_rate": 2.925652162596613e-05,
+      "loss": 0.3338,
+      "step": 905
+    },
+    {
+      "epoch": 2.061433447098976,
+      "grad_norm": 0.4535620497514134,
+      "learning_rate": 2.9228308792961523e-05,
+      "loss": 0.3455,
+      "step": 906
+    },
+    {
+      "epoch": 2.0637087599544937,
+      "grad_norm": 0.509067194030193,
+      "learning_rate": 2.9200072609919957e-05,
+      "loss": 0.3469,
+      "step": 907
+    },
+    {
+      "epoch": 2.065984072810011,
+      "grad_norm": 0.446324127596061,
+      "learning_rate": 2.917181314828638e-05,
+      "loss": 0.3484,
+      "step": 908
+    },
+    {
+      "epoch": 2.068259385665529,
+      "grad_norm": 0.5670420846161346,
+      "learning_rate": 2.914353047956463e-05,
+      "loss": 0.3565,
+      "step": 909
+    },
+    {
+      "epoch": 2.0705346985210467,
+      "grad_norm": 0.48304097423967896,
+      "learning_rate": 2.9115224675317252e-05,
+      "loss": 0.3396,
+      "step": 910
+    },
+    {
+      "epoch": 2.072810011376564,
+      "grad_norm": 0.4596146351472507,
+      "learning_rate": 2.908689580716534e-05,
+      "loss": 0.3507,
+      "step": 911
+    },
+    {
+      "epoch": 2.0750853242320817,
+      "grad_norm": 0.5183922247868473,
+      "learning_rate": 2.9058543946788357e-05,
+      "loss": 0.3166,
+      "step": 912
+    },
+    {
+      "epoch": 2.0773606370875997,
+      "grad_norm": 0.6011178626020073,
+      "learning_rate": 2.9030169165923926e-05,
+      "loss": 0.3329,
+      "step": 913
+    },
+    {
+      "epoch": 2.079635949943117,
+      "grad_norm": 0.4313193879440631,
+      "learning_rate": 2.9001771536367667e-05,
+      "loss": 0.2962,
+      "step": 914
+    },
+    {
+      "epoch": 2.0819112627986347,
+      "grad_norm": 0.5766905270202347,
+      "learning_rate": 2.897335112997302e-05,
+      "loss": 0.3343,
+      "step": 915
+    },
+    {
+      "epoch": 2.0841865756541527,
+      "grad_norm": 0.4304243154040951,
+      "learning_rate": 2.8944908018651055e-05,
+      "loss": 0.3284,
+      "step": 916
+    },
+    {
+      "epoch": 2.08646188850967,
+      "grad_norm": 0.534020213474336,
+      "learning_rate": 2.8916442274370283e-05,
+      "loss": 0.332,
+      "step": 917
+    },
+    {
+      "epoch": 2.0887372013651877,
+      "grad_norm": 0.4993577826939032,
+      "learning_rate": 2.88879539691565e-05,
+      "loss": 0.3085,
+      "step": 918
+    },
+    {
+      "epoch": 2.091012514220705,
+      "grad_norm": 0.4339374501373621,
+      "learning_rate": 2.8859443175092553e-05,
+      "loss": 0.3409,
+      "step": 919
+    },
+    {
+      "epoch": 2.093287827076223,
+      "grad_norm": 0.4476486984150351,
+      "learning_rate": 2.8830909964318242e-05,
+      "loss": 0.3524,
+      "step": 920
+    },
+    {
+      "epoch": 2.0955631399317407,
+      "grad_norm": 0.4915992580427168,
+      "learning_rate": 2.880235440903004e-05,
+      "loss": 0.3565,
+      "step": 921
+    },
+    {
+      "epoch": 2.0978384527872582,
+      "grad_norm": 0.4347367058608959,
+      "learning_rate": 2.8773776581480986e-05,
+      "loss": 0.346,
+      "step": 922
+    },
+    {
+      "epoch": 2.1001137656427757,
+      "grad_norm": 0.4023685534536312,
+      "learning_rate": 2.8745176553980454e-05,
+      "loss": 0.3116,
+      "step": 923
+    },
+    {
+      "epoch": 2.1023890784982937,
+      "grad_norm": 0.48203785089108603,
+      "learning_rate": 2.8716554398894015e-05,
+      "loss": 0.3392,
+      "step": 924
+    },
+    {
+      "epoch": 2.1046643913538112,
+      "grad_norm": 0.40290034314805556,
+      "learning_rate": 2.868791018864321e-05,
+      "loss": 0.3332,
+      "step": 925
+    },
+    {
+      "epoch": 2.1069397042093287,
+      "grad_norm": 0.48397626782125486,
+      "learning_rate": 2.865924399570538e-05,
+      "loss": 0.3565,
+      "step": 926
+    },
+    {
+      "epoch": 2.1092150170648463,
+      "grad_norm": 0.4737904462011868,
+      "learning_rate": 2.86305558926135e-05,
+      "loss": 0.3649,
+      "step": 927
+    },
+    {
+      "epoch": 2.1114903299203642,
+      "grad_norm": 0.4571656530020779,
+      "learning_rate": 2.860184595195599e-05,
+      "loss": 0.3334,
+      "step": 928
+    },
+    {
+      "epoch": 2.1137656427758817,
+      "grad_norm": 0.5089308015525444,
+      "learning_rate": 2.8573114246376508e-05,
+      "loss": 0.3304,
+      "step": 929
+    },
+    {
+      "epoch": 2.1160409556313993,
+      "grad_norm": 0.516422270099582,
+      "learning_rate": 2.854436084857379e-05,
+      "loss": 0.3197,
+      "step": 930
+    },
+    {
+      "epoch": 2.118316268486917,
+      "grad_norm": 0.42936517494553617,
+      "learning_rate": 2.8515585831301456e-05,
+      "loss": 0.3103,
+      "step": 931
+    },
+    {
+      "epoch": 2.1205915813424348,
+      "grad_norm": 0.4208030605335549,
+      "learning_rate": 2.848678926736784e-05,
+      "loss": 0.2928,
+      "step": 932
+    },
+    {
+      "epoch": 2.1228668941979523,
+      "grad_norm": 0.40310708599010403,
+      "learning_rate": 2.845797122963578e-05,
+      "loss": 0.296,
+      "step": 933
+    },
+    {
+      "epoch": 2.12514220705347,
+      "grad_norm": 0.3926937116571276,
+      "learning_rate": 2.8429131791022467e-05,
+      "loss": 0.2983,
+      "step": 934
+    },
+    {
+      "epoch": 2.1274175199089873,
+      "grad_norm": 0.4675033572182952,
+      "learning_rate": 2.8400271024499212e-05,
+      "loss": 0.3568,
+      "step": 935
+    },
+    {
+      "epoch": 2.1296928327645053,
+      "grad_norm": 0.3828039852443616,
+      "learning_rate": 2.837138900309133e-05,
+      "loss": 0.3506,
+      "step": 936
+    },
+    {
+      "epoch": 2.131968145620023,
+      "grad_norm": 0.47258551556235256,
+      "learning_rate": 2.834248579987787e-05,
+      "loss": 0.3265,
+      "step": 937
+    },
+    {
+      "epoch": 2.1342434584755403,
+      "grad_norm": 0.4112834564475498,
+      "learning_rate": 2.8313561487991527e-05,
+      "loss": 0.3123,
+      "step": 938
+    },
+    {
+      "epoch": 2.136518771331058,
+      "grad_norm": 0.4382940323504779,
+      "learning_rate": 2.8284616140618366e-05,
+      "loss": 0.3205,
+      "step": 939
+    },
+    {
+      "epoch": 2.138794084186576,
+      "grad_norm": 0.4005806317738167,
+      "learning_rate": 2.8255649830997704e-05,
+      "loss": 0.3504,
+      "step": 940
+    },
+    {
+      "epoch": 2.1410693970420933,
+      "grad_norm": 0.44219833189801894,
+      "learning_rate": 2.8226662632421874e-05,
+      "loss": 0.3387,
+      "step": 941
+    },
+    {
+      "epoch": 2.143344709897611,
+      "grad_norm": 0.39306548433110394,
+      "learning_rate": 2.8197654618236084e-05,
+      "loss": 0.3191,
+      "step": 942
+    },
+    {
+      "epoch": 2.1456200227531284,
+      "grad_norm": 0.43320014160920517,
+      "learning_rate": 2.8168625861838204e-05,
+      "loss": 0.3314,
+      "step": 943
+    },
+    {
+      "epoch": 2.1478953356086463,
+      "grad_norm": 0.4462810447120457,
+      "learning_rate": 2.813957643667859e-05,
+      "loss": 0.3615,
+      "step": 944
+    },
+    {
+      "epoch": 2.150170648464164,
+      "grad_norm": 0.5306187676101278,
+      "learning_rate": 2.811050641625988e-05,
+      "loss": 0.3316,
+      "step": 945
+    },
+    {
+      "epoch": 2.1524459613196814,
+      "grad_norm": 0.42718488947869915,
+      "learning_rate": 2.8081415874136846e-05,
+      "loss": 0.3358,
+      "step": 946
+    },
+    {
+      "epoch": 2.1547212741751993,
+      "grad_norm": 0.4008754463916256,
+      "learning_rate": 2.805230488391617e-05,
+      "loss": 0.3516,
+      "step": 947
+    },
+    {
+      "epoch": 2.156996587030717,
+      "grad_norm": 0.46697523347894887,
+      "learning_rate": 2.8023173519256278e-05,
+      "loss": 0.3283,
+      "step": 948
+    },
+    {
+      "epoch": 2.1592718998862344,
+      "grad_norm": 0.4814370655105727,
+      "learning_rate": 2.799402185386715e-05,
+      "loss": 0.348,
+      "step": 949
+    },
+    {
+      "epoch": 2.161547212741752,
+      "grad_norm": 0.40967124598455246,
+      "learning_rate": 2.796484996151013e-05,
+      "loss": 0.3429,
+      "step": 950
+    },
+    {
+      "epoch": 2.1638225255972694,
+      "grad_norm": 0.5135630313746052,
+      "learning_rate": 2.7935657915997746e-05,
+      "loss": 0.3305,
+      "step": 951
+    },
+    {
+      "epoch": 2.1660978384527874,
+      "grad_norm": 0.3787931785217832,
+      "learning_rate": 2.7906445791193504e-05,
+      "loss": 0.3069,
+      "step": 952
+    },
+    {
+      "epoch": 2.168373151308305,
+      "grad_norm": 0.4569615318817335,
+      "learning_rate": 2.787721366101173e-05,
+      "loss": 0.3118,
+      "step": 953
+    },
+    {
+      "epoch": 2.1706484641638224,
+      "grad_norm": 0.4481188682798795,
+      "learning_rate": 2.7847961599417375e-05,
+      "loss": 0.3252,
+      "step": 954
+    },
+    {
+      "epoch": 2.1729237770193404,
+      "grad_norm": 0.48221819740301913,
+      "learning_rate": 2.7818689680425807e-05,
+      "loss": 0.3387,
+      "step": 955
+    },
+    {
+      "epoch": 2.175199089874858,
+      "grad_norm": 0.48264529866841904,
+      "learning_rate": 2.7789397978102643e-05,
+      "loss": 0.3377,
+      "step": 956
+    },
+    {
+      "epoch": 2.1774744027303754,
+      "grad_norm": 0.41290712374130284,
+      "learning_rate": 2.7760086566563547e-05,
+      "loss": 0.3895,
+      "step": 957
+    },
+    {
+      "epoch": 2.179749715585893,
+      "grad_norm": 0.45132639708493355,
+      "learning_rate": 2.773075551997408e-05,
+      "loss": 0.3398,
+      "step": 958
+    },
+    {
+      "epoch": 2.182025028441411,
+      "grad_norm": 0.4444127278026595,
+      "learning_rate": 2.7701404912549465e-05,
+      "loss": 0.3029,
+      "step": 959
+    },
+    {
+      "epoch": 2.1843003412969284,
+      "grad_norm": 0.4282293855511231,
+      "learning_rate": 2.767203481855441e-05,
+      "loss": 0.3293,
+      "step": 960
+    },
+    {
+      "epoch": 2.186575654152446,
+      "grad_norm": 0.4222918245129199,
+      "learning_rate": 2.764264531230295e-05,
+      "loss": 0.3321,
+      "step": 961
+    },
+    {
+      "epoch": 2.1888509670079634,
+      "grad_norm": 0.4651754035829156,
+      "learning_rate": 2.7613236468158227e-05,
+      "loss": 0.3412,
+      "step": 962
+    },
+    {
+      "epoch": 2.1911262798634814,
+      "grad_norm": 0.4429753682559861,
+      "learning_rate": 2.758380836053232e-05,
+      "loss": 0.334,
+      "step": 963
+    },
+    {
+      "epoch": 2.193401592718999,
+      "grad_norm": 0.4506792671473799,
+      "learning_rate": 2.7554361063886042e-05,
+      "loss": 0.3802,
+      "step": 964
+    },
+    {
+      "epoch": 2.1956769055745164,
+      "grad_norm": 0.402698452950612,
+      "learning_rate": 2.7524894652728754e-05,
+      "loss": 0.3245,
+      "step": 965
+    },
+    {
+      "epoch": 2.197952218430034,
+      "grad_norm": 0.4514698831658316,
+      "learning_rate": 2.7495409201618204e-05,
+      "loss": 0.3699,
+      "step": 966
+    },
+    {
+      "epoch": 2.200227531285552,
+      "grad_norm": 0.44071254548311,
+      "learning_rate": 2.7465904785160286e-05,
+      "loss": 0.3095,
+      "step": 967
+    },
+    {
+      "epoch": 2.2025028441410694,
+      "grad_norm": 0.3974526433290124,
+      "learning_rate": 2.743638147800891e-05,
+      "loss": 0.3238,
+      "step": 968
+    },
+    {
+      "epoch": 2.204778156996587,
+      "grad_norm": 0.47712259703662935,
+      "learning_rate": 2.7406839354865768e-05,
+      "loss": 0.3298,
+      "step": 969
+    },
+    {
+      "epoch": 2.2070534698521045,
+      "grad_norm": 0.46205936519251023,
+      "learning_rate": 2.7377278490480157e-05,
+      "loss": 0.3389,
+      "step": 970
+    },
+    {
+      "epoch": 2.2093287827076225,
+      "grad_norm": 0.47965062020649035,
+      "learning_rate": 2.7347698959648817e-05,
+      "loss": 0.354,
+      "step": 971
+    },
+    {
+      "epoch": 2.21160409556314,
+      "grad_norm": 0.44557008156382216,
+      "learning_rate": 2.7318100837215688e-05,
+      "loss": 0.3425,
+      "step": 972
+    },
+    {
+      "epoch": 2.2138794084186575,
+      "grad_norm": 0.4269333534718295,
+      "learning_rate": 2.7288484198071778e-05,
+      "loss": 0.349,
+      "step": 973
+    },
+    {
+      "epoch": 2.216154721274175,
+      "grad_norm": 0.38466010268113765,
+      "learning_rate": 2.7258849117154934e-05,
+      "loss": 0.3247,
+      "step": 974
+    },
+    {
+      "epoch": 2.218430034129693,
+      "grad_norm": 0.5246724481013021,
+      "learning_rate": 2.7229195669449667e-05,
+      "loss": 0.3296,
+      "step": 975
+    },
+    {
+      "epoch": 2.2207053469852105,
+      "grad_norm": 0.4336227929999983,
+      "learning_rate": 2.7199523929986972e-05,
+      "loss": 0.3245,
+      "step": 976
+    },
+    {
+      "epoch": 2.222980659840728,
+      "grad_norm": 0.4608049394537325,
+      "learning_rate": 2.7169833973844095e-05,
+      "loss": 0.3348,
+      "step": 977
+    },
+    {
+      "epoch": 2.2252559726962455,
+      "grad_norm": 0.4335780434069244,
+      "learning_rate": 2.7140125876144416e-05,
+      "loss": 0.3621,
+      "step": 978
+    },
+    {
+      "epoch": 2.2275312855517635,
+      "grad_norm": 0.42545851610183605,
+      "learning_rate": 2.7110399712057195e-05,
+      "loss": 0.3233,
+      "step": 979
+    },
+    {
+      "epoch": 2.229806598407281,
+      "grad_norm": 0.41656534471160633,
+      "learning_rate": 2.7080655556797406e-05,
+      "loss": 0.3259,
+      "step": 980
+    },
+    {
+      "epoch": 2.2320819112627985,
+      "grad_norm": 0.5495549115860415,
+      "learning_rate": 2.7050893485625547e-05,
+      "loss": 0.3298,
+      "step": 981
+    },
+    {
+      "epoch": 2.234357224118316,
+      "grad_norm": 0.338719254100158,
+      "learning_rate": 2.7021113573847454e-05,
+      "loss": 0.3127,
+      "step": 982
+    },
+    {
+      "epoch": 2.236632536973834,
+      "grad_norm": 0.575628509160119,
+      "learning_rate": 2.6991315896814098e-05,
+      "loss": 0.3127,
+      "step": 983
+    },
+    {
+      "epoch": 2.2389078498293515,
+      "grad_norm": 0.4018107807328143,
+      "learning_rate": 2.6961500529921395e-05,
+      "loss": 0.3189,
+      "step": 984
+    },
+    {
+      "epoch": 2.241183162684869,
+      "grad_norm": 0.5309139349913046,
+      "learning_rate": 2.693166754861003e-05,
+      "loss": 0.4735,
+      "step": 985
+    },
+    {
+      "epoch": 2.243458475540387,
+      "grad_norm": 0.40738353010271605,
+      "learning_rate": 2.6901817028365263e-05,
+      "loss": 0.3144,
+      "step": 986
+    },
+    {
+      "epoch": 2.2457337883959045,
+      "grad_norm": 0.49702330123818667,
+      "learning_rate": 2.6871949044716723e-05,
+      "loss": 0.3382,
+      "step": 987
+    },
+    {
+      "epoch": 2.248009101251422,
+      "grad_norm": 0.47987163769127383,
+      "learning_rate": 2.6842063673238223e-05,
+      "loss": 0.3334,
+      "step": 988
+    },
+    {
+      "epoch": 2.2502844141069396,
+      "grad_norm": 0.45620904242825305,
+      "learning_rate": 2.681216098954757e-05,
+      "loss": 0.3373,
+      "step": 989
+    },
+    {
+      "epoch": 2.252559726962457,
+      "grad_norm": 0.44644654191757066,
+      "learning_rate": 2.6782241069306395e-05,
+      "loss": 0.3163,
+      "step": 990
+    },
+    {
+      "epoch": 2.254835039817975,
+      "grad_norm": 0.4068960377449261,
+      "learning_rate": 2.6752303988219915e-05,
+      "loss": 0.3197,
+      "step": 991
+    },
+    {
+      "epoch": 2.2571103526734926,
+      "grad_norm": 0.45285913624947255,
+      "learning_rate": 2.6722349822036796e-05,
+      "loss": 0.3497,
+      "step": 992
+    },
+    {
+      "epoch": 2.25938566552901,
+      "grad_norm": 0.40760134014949057,
+      "learning_rate": 2.6692378646548904e-05,
+      "loss": 0.3426,
+      "step": 993
+    },
+    {
+      "epoch": 2.261660978384528,
+      "grad_norm": 0.38386299423561715,
+      "learning_rate": 2.6662390537591175e-05,
+      "loss": 0.3494,
+      "step": 994
+    },
+    {
+      "epoch": 2.2639362912400456,
+      "grad_norm": 0.42369854867815504,
+      "learning_rate": 2.663238557104136e-05,
+      "loss": 0.3418,
+      "step": 995
+    },
+    {
+      "epoch": 2.266211604095563,
+      "grad_norm": 0.41909488435243325,
+      "learning_rate": 2.6602363822819887e-05,
+      "loss": 0.3225,
+      "step": 996
+    },
+    {
+      "epoch": 2.2684869169510806,
+      "grad_norm": 0.3994565368867542,
+      "learning_rate": 2.6572325368889633e-05,
+      "loss": 0.3045,
+      "step": 997
+    },
+    {
+      "epoch": 2.2707622298065986,
+      "grad_norm": 0.388146229645343,
+      "learning_rate": 2.6542270285255756e-05,
+      "loss": 0.2995,
+      "step": 998
+    },
+    {
+      "epoch": 2.273037542662116,
+      "grad_norm": 0.44156111760441435,
+      "learning_rate": 2.6512198647965478e-05,
+      "loss": 0.3505,
+      "step": 999
+    },
+    {
+      "epoch": 2.2753128555176336,
+      "grad_norm": 0.42089597928299294,
+      "learning_rate": 2.648211053310792e-05,
+      "loss": 0.3493,
+      "step": 1000
+    },
+    {
+      "epoch": 2.277588168373151,
+      "grad_norm": 0.45712285437388794,
+      "learning_rate": 2.6452006016813882e-05,
+      "loss": 0.3366,
+      "step": 1001
+    },
+    {
+      "epoch": 2.279863481228669,
+      "grad_norm": 0.4501732377023807,
+      "learning_rate": 2.6421885175255675e-05,
+      "loss": 0.344,
+      "step": 1002
+    },
+    {
+      "epoch": 2.2821387940841866,
+      "grad_norm": 0.4229283632947017,
+      "learning_rate": 2.6391748084646914e-05,
+      "loss": 0.3562,
+      "step": 1003
+    },
+    {
+      "epoch": 2.284414106939704,
+      "grad_norm": 0.4347442412553129,
+      "learning_rate": 2.6361594821242324e-05,
+      "loss": 0.3509,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2866894197952217,
+      "grad_norm": 0.3931610774927613,
+      "learning_rate": 2.6331425461337557e-05,
+      "loss": 0.3199,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2889647326507396,
+      "grad_norm": 0.408117532581314,
+      "learning_rate": 2.6301240081268987e-05,
+      "loss": 0.3504,
+      "step": 1006
+    },
+    {
+      "epoch": 2.291240045506257,
+      "grad_norm": 0.39819784875619857,
+      "learning_rate": 2.6271038757413535e-05,
+      "loss": 0.344,
+      "step": 1007
+    },
+    {
+      "epoch": 2.2935153583617747,
+      "grad_norm": 0.39159211366569074,
+      "learning_rate": 2.624082156618845e-05,
+      "loss": 0.331,
+      "step": 1008
+    },
+    {
+      "epoch": 2.295790671217292,
+      "grad_norm": 0.4014550533237517,
+      "learning_rate": 2.621058858405113e-05,
+      "loss": 0.3268,
+      "step": 1009
+    },
+    {
+      "epoch": 2.29806598407281,
+      "grad_norm": 0.4243208085102565,
+      "learning_rate": 2.618033988749895e-05,
+      "loss": 0.3068,
+      "step": 1010
+    },
+    {
+      "epoch": 2.3003412969283277,
+      "grad_norm": 0.4076296649805579,
+      "learning_rate": 2.615007555306902e-05,
+      "loss": 0.3571,
+      "step": 1011
+    },
+    {
+      "epoch": 2.302616609783845,
+      "grad_norm": 0.4686341511962635,
+      "learning_rate": 2.6119795657338028e-05,
+      "loss": 0.3403,
+      "step": 1012
+    },
+    {
+      "epoch": 2.3048919226393627,
+      "grad_norm": 0.4336300175984805,
+      "learning_rate": 2.6089500276922038e-05,
+      "loss": 0.3534,
+      "step": 1013
+    },
+    {
+      "epoch": 2.3071672354948807,
+      "grad_norm": 0.4230808109642132,
+      "learning_rate": 2.60591894884763e-05,
+      "loss": 0.3202,
+      "step": 1014
+    },
+    {
+      "epoch": 2.309442548350398,
+      "grad_norm": 0.40936239516570677,
+      "learning_rate": 2.602886336869503e-05,
+      "loss": 0.3003,
+      "step": 1015
+    },
+    {
+      "epoch": 2.3117178612059157,
+      "grad_norm": 0.408650560420394,
+      "learning_rate": 2.5998521994311264e-05,
+      "loss": 0.3318,
+      "step": 1016
+    },
+    {
+      "epoch": 2.3139931740614337,
+      "grad_norm": 0.3858722141552388,
+      "learning_rate": 2.5968165442096603e-05,
+      "loss": 0.3194,
+      "step": 1017
+    },
+    {
+      "epoch": 2.316268486916951,
+      "grad_norm": 0.3982980138390431,
+      "learning_rate": 2.593779378886109e-05,
+      "loss": 0.3201,
+      "step": 1018
+    },
+    {
+      "epoch": 2.3185437997724687,
+      "grad_norm": 0.4551971424360926,
+      "learning_rate": 2.5907407111452942e-05,
+      "loss": 0.3546,
+      "step": 1019
+    },
+    {
+      "epoch": 2.3208191126279862,
+      "grad_norm": 0.401481106912197,
+      "learning_rate": 2.5877005486758424e-05,
+      "loss": 0.3405,
+      "step": 1020
+    },
+    {
+      "epoch": 2.3230944254835038,
+      "grad_norm": 0.4136507043623277,
+      "learning_rate": 2.5846588991701585e-05,
+      "loss": 0.3313,
+      "step": 1021
+    },
+    {
+      "epoch": 2.3253697383390217,
+      "grad_norm": 0.39821934668137265,
+      "learning_rate": 2.5816157703244142e-05,
+      "loss": 0.3224,
+      "step": 1022
+    },
+    {
+      "epoch": 2.3276450511945392,
+      "grad_norm": 0.4242575172743546,
+      "learning_rate": 2.5785711698385204e-05,
+      "loss": 0.303,
+      "step": 1023
+    },
+    {
+      "epoch": 2.3299203640500568,
+      "grad_norm": 0.4126931267650357,
+      "learning_rate": 2.5755251054161136e-05,
+      "loss": 0.3496,
+      "step": 1024
+    },
+    {
+      "epoch": 2.3321956769055747,
+      "grad_norm": 0.4289455241021369,
+      "learning_rate": 2.5724775847645345e-05,
+      "loss": 0.3159,
+      "step": 1025
+    },
+    {
+      "epoch": 2.3344709897610922,
+      "grad_norm": 0.4102737686333417,
+      "learning_rate": 2.569428615594808e-05,
+      "loss": 0.3412,
+      "step": 1026
+    },
+    {
+      "epoch": 2.3367463026166098,
+      "grad_norm": 0.4498395660714342,
+      "learning_rate": 2.5663782056216242e-05,
+      "loss": 0.3309,
+      "step": 1027
+    },
+    {
+      "epoch": 2.3390216154721273,
+      "grad_norm": 0.4098791457012732,
+      "learning_rate": 2.5633263625633188e-05,
+      "loss": 0.3073,
+      "step": 1028
+    },
+    {
+      "epoch": 2.3412969283276452,
+      "grad_norm": 0.41319480076430887,
+      "learning_rate": 2.560273094141854e-05,
+      "loss": 0.3351,
+      "step": 1029
+    },
+    {
+      "epoch": 2.3435722411831628,
+      "grad_norm": 0.3883968682027695,
+      "learning_rate": 2.557218408082798e-05,
+      "loss": 0.3176,
+      "step": 1030
+    },
+    {
+      "epoch": 2.3458475540386803,
+      "grad_norm": 0.4411653876987364,
+      "learning_rate": 2.5541623121153053e-05,
+      "loss": 0.3256,
+      "step": 1031
+    },
+    {
+      "epoch": 2.348122866894198,
+      "grad_norm": 0.45631384309745954,
+      "learning_rate": 2.5511048139720993e-05,
+      "loss": 0.3578,
+      "step": 1032
+    },
+    {
+      "epoch": 2.3503981797497158,
+      "grad_norm": 0.39850839240719127,
+      "learning_rate": 2.5480459213894514e-05,
+      "loss": 0.3396,
+      "step": 1033
+    },
+    {
+      "epoch": 2.3526734926052333,
+      "grad_norm": 0.4207268711095779,
+      "learning_rate": 2.5449856421071603e-05,
+      "loss": 0.3104,
+      "step": 1034
+    },
+    {
+      "epoch": 2.354948805460751,
+      "grad_norm": 0.35535500820002625,
+      "learning_rate": 2.5419239838685325e-05,
+      "loss": 0.3425,
+      "step": 1035
+    },
+    {
+      "epoch": 2.3572241183162683,
+      "grad_norm": 0.4101822274977704,
+      "learning_rate": 2.5388609544203653e-05,
+      "loss": 0.3667,
+      "step": 1036
+    },
+    {
+      "epoch": 2.3594994311717863,
+      "grad_norm": 0.4104696919475045,
+      "learning_rate": 2.535796561512926e-05,
+      "loss": 0.3478,
+      "step": 1037
+    },
+    {
+      "epoch": 2.361774744027304,
+      "grad_norm": 0.4190560418201998,
+      "learning_rate": 2.5327308128999302e-05,
+      "loss": 0.3573,
+      "step": 1038
+    },
+    {
+      "epoch": 2.3640500568828213,
+      "grad_norm": 0.3606297237914685,
+      "learning_rate": 2.529663716338524e-05,
+      "loss": 0.3338,
+      "step": 1039
+    },
+    {
+      "epoch": 2.366325369738339,
+      "grad_norm": 0.474284737988354,
+      "learning_rate": 2.526595279589265e-05,
+      "loss": 0.3285,
+      "step": 1040
+    },
+    {
+      "epoch": 2.368600682593857,
+      "grad_norm": 0.3933117024944243,
+      "learning_rate": 2.5235255104161013e-05,
+      "loss": 0.3131,
+      "step": 1041
+    },
+    {
+      "epoch": 2.3708759954493743,
+      "grad_norm": 0.45174667825198905,
+      "learning_rate": 2.5204544165863528e-05,
+      "loss": 0.2943,
+      "step": 1042
+    },
+    {
+      "epoch": 2.373151308304892,
+      "grad_norm": 0.3950012310464477,
+      "learning_rate": 2.5173820058706902e-05,
+      "loss": 0.3455,
+      "step": 1043
+    },
+    {
+      "epoch": 2.3754266211604094,
+      "grad_norm": 0.4622525221261252,
+      "learning_rate": 2.5143082860431173e-05,
+      "loss": 0.3455,
+      "step": 1044
+    },
+    {
+      "epoch": 2.3777019340159273,
+      "grad_norm": 0.46850316974346173,
+      "learning_rate": 2.5112332648809495e-05,
+      "loss": 0.3464,
+      "step": 1045
+    },
+    {
+      "epoch": 2.379977246871445,
+      "grad_norm": 0.4148884520885508,
+      "learning_rate": 2.508156950164796e-05,
+      "loss": 0.3103,
+      "step": 1046
+    },
+    {
+      "epoch": 2.3822525597269624,
+      "grad_norm": 0.470572029624604,
+      "learning_rate": 2.505079349678538e-05,
+      "loss": 0.3459,
+      "step": 1047
+    },
+    {
+      "epoch": 2.3845278725824803,
+      "grad_norm": 0.4695289856062143,
+      "learning_rate": 2.50200047120931e-05,
+      "loss": 0.3265,
+      "step": 1048
+    },
+    {
+      "epoch": 2.386803185437998,
+      "grad_norm": 0.4100169364168893,
+      "learning_rate": 2.4989203225474807e-05,
+      "loss": 0.3053,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3890784982935154,
+      "grad_norm": 0.5321799194490549,
+      "learning_rate": 2.4958389114866326e-05,
+      "loss": 0.3325,
+      "step": 1050
+    },
+    {
+      "epoch": 2.391353811149033,
+      "grad_norm": 0.4159936423428224,
+      "learning_rate": 2.492756245823542e-05,
+      "loss": 0.3399,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3936291240045504,
+      "grad_norm": 0.4266756865011224,
+      "learning_rate": 2.4896723333581607e-05,
+      "loss": 0.365,
+      "step": 1052
+    },
+    {
+      "epoch": 2.3959044368600684,
+      "grad_norm": 0.4407767461713438,
+      "learning_rate": 2.4865871818935935e-05,
+      "loss": 0.3496,
+      "step": 1053
+    },
+    {
+      "epoch": 2.398179749715586,
+      "grad_norm": 0.39165662572831933,
+      "learning_rate": 2.4835007992360817e-05,
+      "loss": 0.3469,
+      "step": 1054
+    },
+    {
+      "epoch": 2.4004550625711034,
+      "grad_norm": 0.446346752400498,
+      "learning_rate": 2.4804131931949823e-05,
+      "loss": 0.3126,
+      "step": 1055
+    },
+    {
+      "epoch": 2.4027303754266214,
+      "grad_norm": 0.43032428277877155,
+      "learning_rate": 2.4773243715827452e-05,
+      "loss": 0.3185,
+      "step": 1056
+    },
+    {
+      "epoch": 2.405005688282139,
+      "grad_norm": 0.4507009034591805,
+      "learning_rate": 2.4742343422148986e-05,
+      "loss": 0.3161,
+      "step": 1057
+    },
+    {
+      "epoch": 2.4072810011376564,
+      "grad_norm": 0.4236196606036691,
+      "learning_rate": 2.4711431129100252e-05,
+      "loss": 0.322,
+      "step": 1058
+    },
+    {
+      "epoch": 2.409556313993174,
+      "grad_norm": 0.5011939156733887,
+      "learning_rate": 2.4680506914897456e-05,
+      "loss": 0.3487,
+      "step": 1059
+    },
+    {
+      "epoch": 2.4118316268486915,
+      "grad_norm": 0.431184104034181,
+      "learning_rate": 2.4649570857786928e-05,
+      "loss": 0.3408,
+      "step": 1060
+    },
+    {
+      "epoch": 2.4141069397042094,
+      "grad_norm": 0.46561161383770455,
+      "learning_rate": 2.461862303604502e-05,
+      "loss": 0.3393,
+      "step": 1061
+    },
+    {
+      "epoch": 2.416382252559727,
+      "grad_norm": 0.46307110671585516,
+      "learning_rate": 2.4587663527977802e-05,
+      "loss": 0.3212,
+      "step": 1062
+    },
+    {
+      "epoch": 2.4186575654152445,
+      "grad_norm": 0.38832680834240807,
+      "learning_rate": 2.455669241192095e-05,
+      "loss": 0.3281,
+      "step": 1063
+    },
+    {
+      "epoch": 2.4209328782707624,
+      "grad_norm": 0.5036416899794153,
+      "learning_rate": 2.452570976623948e-05,
+      "loss": 0.3463,
+      "step": 1064
+    },
+    {
+      "epoch": 2.42320819112628,
+      "grad_norm": 0.36768104349960495,
+      "learning_rate": 2.449471566932761e-05,
+      "loss": 0.3265,
+      "step": 1065
+    },
+    {
+      "epoch": 2.4254835039817975,
+      "grad_norm": 0.4706364182344478,
+      "learning_rate": 2.4463710199608503e-05,
+      "loss": 0.3456,
+      "step": 1066
+    },
+    {
+      "epoch": 2.427758816837315,
+      "grad_norm": 0.5005365274468633,
+      "learning_rate": 2.4432693435534128e-05,
+      "loss": 0.3667,
+      "step": 1067
+    },
+    {
+      "epoch": 2.430034129692833,
+      "grad_norm": 0.4230800225468822,
+      "learning_rate": 2.4401665455585004e-05,
+      "loss": 0.3327,
+      "step": 1068
+    },
+    {
+      "epoch": 2.4323094425483505,
+      "grad_norm": 0.4669416782022287,
+      "learning_rate": 2.4370626338270052e-05,
+      "loss": 0.3316,
+      "step": 1069
+    },
+    {
+      "epoch": 2.434584755403868,
+      "grad_norm": 0.4327164651401064,
+      "learning_rate": 2.4339576162126362e-05,
+      "loss": 0.3356,
+      "step": 1070
+    },
+    {
+      "epoch": 2.4368600682593855,
+      "grad_norm": 0.39994880861025606,
+      "learning_rate": 2.430851500571901e-05,
+      "loss": 0.3566,
+      "step": 1071
+    },
+    {
+      "epoch": 2.4391353811149035,
+      "grad_norm": 0.45847337574158525,
+      "learning_rate": 2.4277442947640845e-05,
+      "loss": 0.3484,
+      "step": 1072
+    },
+    {
+      "epoch": 2.441410693970421,
+      "grad_norm": 0.4744507089590578,
+      "learning_rate": 2.4246360066512316e-05,
+      "loss": 0.3315,
+      "step": 1073
+    },
+    {
+      "epoch": 2.4436860068259385,
+      "grad_norm": 0.3833324983793694,
+      "learning_rate": 2.4215266440981245e-05,
+      "loss": 0.3307,
+      "step": 1074
+    },
+    {
+      "epoch": 2.445961319681456,
+      "grad_norm": 0.50740565370416,
+      "learning_rate": 2.418416214972265e-05,
+      "loss": 0.3287,
+      "step": 1075
+    },
+    {
+      "epoch": 2.448236632536974,
+      "grad_norm": 0.46729101607621365,
+      "learning_rate": 2.415304727143851e-05,
+      "loss": 0.3284,
+      "step": 1076
+    },
+    {
+      "epoch": 2.4505119453924915,
+      "grad_norm": 0.42570906698675304,
+      "learning_rate": 2.4121921884857638e-05,
+      "loss": 0.3249,
+      "step": 1077
+    },
+    {
+      "epoch": 2.452787258248009,
+      "grad_norm": 0.4597186479797924,
+      "learning_rate": 2.4090786068735387e-05,
+      "loss": 0.3527,
+      "step": 1078
+    },
+    {
+      "epoch": 2.4550625711035265,
+      "grad_norm": 0.4425804779532458,
+      "learning_rate": 2.4059639901853552e-05,
+      "loss": 0.332,
+      "step": 1079
+    },
+    {
+      "epoch": 2.4573378839590445,
+      "grad_norm": 0.40481512380156626,
+      "learning_rate": 2.4028483463020053e-05,
+      "loss": 0.3341,
+      "step": 1080
+    },
+    {
+      "epoch": 2.459613196814562,
+      "grad_norm": 0.4030421736394406,
+      "learning_rate": 2.3997316831068858e-05,
+      "loss": 0.3066,
+      "step": 1081
+    },
+    {
+      "epoch": 2.4618885096700796,
+      "grad_norm": 0.43379563589411485,
+      "learning_rate": 2.3966140084859698e-05,
+      "loss": 0.371,
+      "step": 1082
+    },
+    {
+      "epoch": 2.464163822525597,
+      "grad_norm": 0.3959467073732541,
+      "learning_rate": 2.3934953303277912e-05,
+      "loss": 0.3271,
+      "step": 1083
+    },
+    {
+      "epoch": 2.466439135381115,
+      "grad_norm": 0.3831467603215834,
+      "learning_rate": 2.39037565652342e-05,
+      "loss": 0.3133,
+      "step": 1084
+    },
+    {
+      "epoch": 2.4687144482366326,
+      "grad_norm": 0.38031635551424137,
+      "learning_rate": 2.3872549949664493e-05,
+      "loss": 0.3427,
+      "step": 1085
+    },
+    {
+      "epoch": 2.47098976109215,
+      "grad_norm": 0.38672451087371595,
+      "learning_rate": 2.3841333535529695e-05,
+      "loss": 0.3223,
+      "step": 1086
+    },
+    {
+      "epoch": 2.473265073947668,
+      "grad_norm": 0.3887032117399429,
+      "learning_rate": 2.3810107401815507e-05,
+      "loss": 0.2995,
+      "step": 1087
+    },
+    {
+      "epoch": 2.4755403868031856,
+      "grad_norm": 0.3840230443404789,
+      "learning_rate": 2.3778871627532207e-05,
+      "loss": 0.3063,
+      "step": 1088
+    },
+    {
+      "epoch": 2.477815699658703,
+      "grad_norm": 0.41811193328879326,
+      "learning_rate": 2.37476262917145e-05,
+      "loss": 0.3278,
+      "step": 1089
+    },
+    {
+      "epoch": 2.4800910125142206,
+      "grad_norm": 0.3903229827000629,
+      "learning_rate": 2.3716371473421242e-05,
+      "loss": 0.3341,
+      "step": 1090
+    },
+    {
+      "epoch": 2.482366325369738,
+      "grad_norm": 0.43928485970506703,
+      "learning_rate": 2.3685107251735336e-05,
+      "loss": 0.3299,
+      "step": 1091
+    },
+    {
+      "epoch": 2.484641638225256,
+      "grad_norm": 0.4118161474697317,
+      "learning_rate": 2.3653833705763417e-05,
+      "loss": 0.3105,
+      "step": 1092
+    },
+    {
+      "epoch": 2.4869169510807736,
+      "grad_norm": 0.4089962324637304,
+      "learning_rate": 2.3622550914635758e-05,
+      "loss": 0.3171,
+      "step": 1093
+    },
+    {
+      "epoch": 2.489192263936291,
+      "grad_norm": 0.4075036878096672,
+      "learning_rate": 2.3591258957506007e-05,
+      "loss": 0.3277,
+      "step": 1094
+    },
+    {
+      "epoch": 2.491467576791809,
+      "grad_norm": 0.4130842827382354,
+      "learning_rate": 2.3559957913551014e-05,
+      "loss": 0.3551,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4937428896473266,
+      "grad_norm": 0.38827396600299435,
+      "learning_rate": 2.3528647861970596e-05,
+      "loss": 0.308,
+      "step": 1096
+    },
+    {
+      "epoch": 2.496018202502844,
+      "grad_norm": 0.3765239077400001,
+      "learning_rate": 2.349732888198739e-05,
+      "loss": 0.3047,
+      "step": 1097
+    },
+    {
+      "epoch": 2.4982935153583616,
+      "grad_norm": 0.42562109897231937,
+      "learning_rate": 2.3466001052846617e-05,
+      "loss": 0.3813,
+      "step": 1098
+    },
+    {
+      "epoch": 2.500568828213879,
+      "grad_norm": 0.43075331823376906,
+      "learning_rate": 2.3434664453815885e-05,
+      "loss": 0.3577,
+      "step": 1099
+    },
+    {
+      "epoch": 2.502844141069397,
+      "grad_norm": 0.4280771817374536,
+      "learning_rate": 2.3403319164184986e-05,
+      "loss": 0.3217,
+      "step": 1100
+    },
+    {
+      "epoch": 2.5051194539249146,
+      "grad_norm": 0.37875391419931537,
+      "learning_rate": 2.3371965263265714e-05,
+      "loss": 0.3036,
+      "step": 1101
+    },
+    {
+      "epoch": 2.507394766780432,
+      "grad_norm": 0.4286854105083951,
+      "learning_rate": 2.334060283039164e-05,
+      "loss": 0.3041,
+      "step": 1102
+    },
+    {
+      "epoch": 2.50967007963595,
+      "grad_norm": 0.42423899167875767,
+      "learning_rate": 2.3309231944917937e-05,
+      "loss": 0.3486,
+      "step": 1103
+    },
+    {
+      "epoch": 2.5119453924914676,
+      "grad_norm": 0.39380880826140496,
+      "learning_rate": 2.327785268622115e-05,
+      "loss": 0.3279,
+      "step": 1104
+    },
+    {
+      "epoch": 2.514220705346985,
+      "grad_norm": 0.44149400886970624,
+      "learning_rate": 2.3246465133699024e-05,
+      "loss": 0.2983,
+      "step": 1105
+    },
+    {
+      "epoch": 2.5164960182025027,
+      "grad_norm": 0.4220494995967289,
+      "learning_rate": 2.321506936677027e-05,
+      "loss": 0.3545,
+      "step": 1106
+    },
+    {
+      "epoch": 2.51877133105802,
+      "grad_norm": 0.39004032618224976,
+      "learning_rate": 2.318366546487442e-05,
+      "loss": 0.3285,
+      "step": 1107
+    },
+    {
+      "epoch": 2.521046643913538,
+      "grad_norm": 0.4293755406970271,
+      "learning_rate": 2.3152253507471543e-05,
+      "loss": 0.3224,
+      "step": 1108
+    },
+    {
+      "epoch": 2.5233219567690557,
+      "grad_norm": 0.4534148070646945,
+      "learning_rate": 2.312083357404213e-05,
+      "loss": 0.3233,
+      "step": 1109
+    },
+    {
+      "epoch": 2.5255972696245736,
+      "grad_norm": 0.4218339483415786,
+      "learning_rate": 2.3089405744086823e-05,
+      "loss": 0.3648,
+      "step": 1110
+    },
+    {
+      "epoch": 2.527872582480091,
+      "grad_norm": 0.4868020320366218,
+      "learning_rate": 2.305797009712628e-05,
+      "loss": 0.32,
+      "step": 1111
+    },
+    {
+      "epoch": 2.5301478953356087,
+      "grad_norm": 0.4063169680463677,
+      "learning_rate": 2.3026526712700893e-05,
+      "loss": 0.3233,
+      "step": 1112
+    },
+    {
+      "epoch": 2.532423208191126,
+      "grad_norm": 0.42980601020754616,
+      "learning_rate": 2.299507567037067e-05,
+      "loss": 0.3311,
+      "step": 1113
+    },
+    {
+      "epoch": 2.5346985210466437,
+      "grad_norm": 0.4881371234939724,
+      "learning_rate": 2.2963617049714974e-05,
+      "loss": 0.3578,
+      "step": 1114
+    },
+    {
+      "epoch": 2.5369738339021617,
+      "grad_norm": 0.4494329234196991,
+      "learning_rate": 2.2932150930332363e-05,
+      "loss": 0.3349,
+      "step": 1115
+    },
+    {
+      "epoch": 2.539249146757679,
+      "grad_norm": 0.5396261834748108,
+      "learning_rate": 2.2900677391840338e-05,
+      "loss": 0.3604,
+      "step": 1116
+    },
+    {
+      "epoch": 2.5415244596131967,
+      "grad_norm": 0.48364432647591565,
+      "learning_rate": 2.28691965138752e-05,
+      "loss": 0.3472,
+      "step": 1117
+    },
+    {
+      "epoch": 2.5437997724687147,
+      "grad_norm": 0.46639259119824894,
+      "learning_rate": 2.283770837609181e-05,
+      "loss": 0.3263,
+      "step": 1118
+    },
+    {
+      "epoch": 2.546075085324232,
+      "grad_norm": 0.4154978640587029,
+      "learning_rate": 2.28062130581634e-05,
+      "loss": 0.3114,
+      "step": 1119
+    },
+    {
+      "epoch": 2.5483503981797497,
+      "grad_norm": 0.5293824526466654,
+      "learning_rate": 2.277471063978137e-05,
+      "loss": 0.3668,
+      "step": 1120
+    },
+    {
+      "epoch": 2.5506257110352673,
+      "grad_norm": 0.4241598383229602,
+      "learning_rate": 2.2743201200655077e-05,
+      "loss": 0.3326,
+      "step": 1121
+    },
+    {
+      "epoch": 2.5529010238907848,
+      "grad_norm": 0.40894333863788457,
+      "learning_rate": 2.2711684820511662e-05,
+      "loss": 0.3624,
+      "step": 1122
+    },
+    {
+      "epoch": 2.5551763367463027,
+      "grad_norm": 0.43884989380427636,
+      "learning_rate": 2.268016157909581e-05,
+      "loss": 0.3501,
+      "step": 1123
+    },
+    {
+      "epoch": 2.5574516496018203,
+      "grad_norm": 0.3790395735860823,
+      "learning_rate": 2.2648631556169575e-05,
+      "loss": 0.3183,
+      "step": 1124
+    },
+    {
+      "epoch": 2.5597269624573378,
+      "grad_norm": 0.442061991986524,
+      "learning_rate": 2.2617094831512167e-05,
+      "loss": 0.3253,
+      "step": 1125
+    },
+    {
+      "epoch": 2.5620022753128557,
+      "grad_norm": 0.41881109276377154,
+      "learning_rate": 2.258555148491975e-05,
+      "loss": 0.3381,
+      "step": 1126
+    },
+    {
+      "epoch": 2.5642775881683733,
+      "grad_norm": 0.4437153525163446,
+      "learning_rate": 2.2554001596205262e-05,
+      "loss": 0.3525,
+      "step": 1127
+    },
+    {
+      "epoch": 2.5665529010238908,
+      "grad_norm": 0.47642938536248186,
+      "learning_rate": 2.2522445245198172e-05,
+      "loss": 0.331,
+      "step": 1128
+    },
+    {
+      "epoch": 2.5688282138794083,
+      "grad_norm": 0.4286456745031337,
+      "learning_rate": 2.249088251174431e-05,
+      "loss": 0.3145,
+      "step": 1129
+    },
+    {
+      "epoch": 2.571103526734926,
+      "grad_norm": 0.49626610344590827,
+      "learning_rate": 2.2459313475705645e-05,
+      "loss": 0.3579,
+      "step": 1130
+    },
+    {
+      "epoch": 2.573378839590444,
+      "grad_norm": 0.4400110077306319,
+      "learning_rate": 2.2427738216960116e-05,
+      "loss": 0.3362,
+      "step": 1131
+    },
+    {
+      "epoch": 2.5756541524459613,
+      "grad_norm": 0.40840187148347123,
+      "learning_rate": 2.239615681540139e-05,
+      "loss": 0.3255,
+      "step": 1132
+    },
+    {
+      "epoch": 2.577929465301479,
+      "grad_norm": 0.403734345115153,
+      "learning_rate": 2.236456935093867e-05,
+      "loss": 0.296,
+      "step": 1133
+    },
+    {
+      "epoch": 2.580204778156997,
+      "grad_norm": 0.4007096524305456,
+      "learning_rate": 2.233297590349652e-05,
+      "loss": 0.3683,
+      "step": 1134
+    },
+    {
+      "epoch": 2.5824800910125143,
+      "grad_norm": 0.43563774745733447,
+      "learning_rate": 2.2301376553014625e-05,
+      "loss": 0.3511,
+      "step": 1135
+    },
+    {
+      "epoch": 2.584755403868032,
+      "grad_norm": 0.42235690826665484,
+      "learning_rate": 2.226977137944762e-05,
+      "loss": 0.3442,
+      "step": 1136
+    },
+    {
+      "epoch": 2.5870307167235493,
+      "grad_norm": 0.41315770540383245,
+      "learning_rate": 2.223816046276487e-05,
+      "loss": 0.3364,
+      "step": 1137
+    },
+    {
+      "epoch": 2.589306029579067,
+      "grad_norm": 0.3999109408063849,
+      "learning_rate": 2.2206543882950257e-05,
+      "loss": 0.3463,
+      "step": 1138
+    },
+    {
+      "epoch": 2.591581342434585,
+      "grad_norm": 0.46969421630938196,
+      "learning_rate": 2.2174921720002018e-05,
+      "loss": 0.3375,
+      "step": 1139
+    },
+    {
+      "epoch": 2.5938566552901023,
+      "grad_norm": 0.41838610421310074,
+      "learning_rate": 2.214329405393249e-05,
+      "loss": 0.3319,
+      "step": 1140
+    },
+    {
+      "epoch": 2.59613196814562,
+      "grad_norm": 0.40684873288050794,
+      "learning_rate": 2.211166096476796e-05,
+      "loss": 0.3253,
+      "step": 1141
+    },
+    {
+      "epoch": 2.598407281001138,
+      "grad_norm": 0.4373340577549481,
+      "learning_rate": 2.208002253254841e-05,
+      "loss": 0.3209,
+      "step": 1142
+    },
+    {
+      "epoch": 2.6006825938566553,
+      "grad_norm": 0.4034129811630071,
+      "learning_rate": 2.204837883732737e-05,
+      "loss": 0.3126,
+      "step": 1143
+    },
+    {
+      "epoch": 2.602957906712173,
+      "grad_norm": 0.3918569162074108,
+      "learning_rate": 2.201672995917166e-05,
+      "loss": 0.3397,
+      "step": 1144
+    },
+    {
+      "epoch": 2.6052332195676904,
+      "grad_norm": 0.4467241134900881,
+      "learning_rate": 2.1985075978161236e-05,
+      "loss": 0.3191,
+      "step": 1145
+    },
+    {
+      "epoch": 2.6075085324232083,
+      "grad_norm": 0.4505057754744431,
+      "learning_rate": 2.1953416974388948e-05,
+      "loss": 0.2992,
+      "step": 1146
+    },
+    {
+      "epoch": 2.609783845278726,
+      "grad_norm": 0.4146029371414084,
+      "learning_rate": 2.192175302796037e-05,
+      "loss": 0.3285,
+      "step": 1147
+    },
+    {
+      "epoch": 2.6120591581342434,
+      "grad_norm": 0.4292927317289128,
+      "learning_rate": 2.1890084218993577e-05,
+      "loss": 0.3262,
+      "step": 1148
+    },
+    {
+      "epoch": 2.6143344709897613,
+      "grad_norm": 0.44875236167205973,
+      "learning_rate": 2.185841062761893e-05,
+      "loss": 0.3299,
+      "step": 1149
+    },
+    {
+      "epoch": 2.616609783845279,
+      "grad_norm": 0.38374058776773967,
+      "learning_rate": 2.1826732333978924e-05,
+      "loss": 0.3214,
+      "step": 1150
+    },
+    {
+      "epoch": 2.6188850967007964,
+      "grad_norm": 0.40845814985241696,
+      "learning_rate": 2.179504941822793e-05,
+      "loss": 0.2941,
+      "step": 1151
+    },
+    {
+      "epoch": 2.621160409556314,
+      "grad_norm": 0.4245944997869959,
+      "learning_rate": 2.1763361960532014e-05,
+      "loss": 0.3632,
+      "step": 1152
+    },
+    {
+      "epoch": 2.6234357224118314,
+      "grad_norm": 0.4124890178533331,
+      "learning_rate": 2.173167004106874e-05,
+      "loss": 0.3517,
+      "step": 1153
+    },
+    {
+      "epoch": 2.6257110352673494,
+      "grad_norm": 0.4703164987096634,
+      "learning_rate": 2.1699973740026962e-05,
+      "loss": 0.3438,
+      "step": 1154
+    },
+    {
+      "epoch": 2.627986348122867,
+      "grad_norm": 0.46930203738129295,
+      "learning_rate": 2.166827313760662e-05,
+      "loss": 0.3491,
+      "step": 1155
+    },
+    {
+      "epoch": 2.6302616609783844,
+      "grad_norm": 0.4391271679349783,
+      "learning_rate": 2.1636568314018534e-05,
+      "loss": 0.299,
+      "step": 1156
+    },
+    {
+      "epoch": 2.6325369738339024,
+      "grad_norm": 0.4299397208820665,
+      "learning_rate": 2.1604859349484203e-05,
+      "loss": 0.366,
+      "step": 1157
+    },
+    {
+      "epoch": 2.63481228668942,
+      "grad_norm": 0.4180615865892131,
+      "learning_rate": 2.1573146324235618e-05,
+      "loss": 0.3353,
+      "step": 1158
+    },
+    {
+      "epoch": 2.6370875995449374,
+      "grad_norm": 0.48576423446336753,
+      "learning_rate": 2.1541429318515026e-05,
+      "loss": 0.3575,
+      "step": 1159
+    },
+    {
+      "epoch": 2.639362912400455,
+      "grad_norm": 0.3848453312354331,
+      "learning_rate": 2.150970841257476e-05,
+      "loss": 0.3458,
+      "step": 1160
+    },
+    {
+      "epoch": 2.6416382252559725,
+      "grad_norm": 0.3907750301355957,
+      "learning_rate": 2.1477983686677003e-05,
+      "loss": 0.3027,
+      "step": 1161
+    },
+    {
+      "epoch": 2.6439135381114904,
+      "grad_norm": 0.3974183662564141,
+      "learning_rate": 2.1446255221093636e-05,
+      "loss": 0.338,
+      "step": 1162
+    },
+    {
+      "epoch": 2.646188850967008,
+      "grad_norm": 0.3845848149906779,
+      "learning_rate": 2.141452309610597e-05,
+      "loss": 0.3481,
+      "step": 1163
+    },
+    {
+      "epoch": 2.6484641638225255,
+      "grad_norm": 0.4373210160414838,
+      "learning_rate": 2.138278739200459e-05,
+      "loss": 0.345,
+      "step": 1164
+    },
+    {
+      "epoch": 2.6507394766780434,
+      "grad_norm": 0.3824463733252028,
+      "learning_rate": 2.135104818908913e-05,
+      "loss": 0.3208,
+      "step": 1165
+    },
+    {
+      "epoch": 2.653014789533561,
+      "grad_norm": 0.4528584606435942,
+      "learning_rate": 2.131930556766809e-05,
+      "loss": 0.3226,
+      "step": 1166
+    },
+    {
+      "epoch": 2.6552901023890785,
+      "grad_norm": 0.4058424243479965,
+      "learning_rate": 2.1287559608058607e-05,
+      "loss": 0.3113,
+      "step": 1167
+    },
+    {
+      "epoch": 2.657565415244596,
+      "grad_norm": 0.4202388416175268,
+      "learning_rate": 2.125581039058627e-05,
+      "loss": 0.3404,
+      "step": 1168
+    },
+    {
+      "epoch": 2.6598407281001135,
+      "grad_norm": 0.42098415966269187,
+      "learning_rate": 2.122405799558491e-05,
+      "loss": 0.3406,
+      "step": 1169
+    },
+    {
+      "epoch": 2.6621160409556315,
+      "grad_norm": 0.4094431284142787,
+      "learning_rate": 2.11923025033964e-05,
+      "loss": 0.3483,
+      "step": 1170
+    },
+    {
+      "epoch": 2.664391353811149,
+      "grad_norm": 0.3761811585723381,
+      "learning_rate": 2.116054399437044e-05,
+      "loss": 0.2999,
+      "step": 1171
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.37488136915012654,
+      "learning_rate": 2.112878254886439e-05,
+      "loss": 0.3414,
+      "step": 1172
+    },
+    {
+      "epoch": 2.6689419795221845,
+      "grad_norm": 0.40620205902134443,
+      "learning_rate": 2.1097018247243e-05,
+      "loss": 0.3159,
+      "step": 1173
+    },
+    {
+      "epoch": 2.671217292377702,
+      "grad_norm": 0.4038563460802831,
+      "learning_rate": 2.106525116987828e-05,
+      "loss": 0.3459,
+      "step": 1174
+    },
+    {
+      "epoch": 2.6734926052332195,
+      "grad_norm": 0.41926081828150746,
+      "learning_rate": 2.103348139714925e-05,
+      "loss": 0.3623,
+      "step": 1175
+    },
+    {
+      "epoch": 2.675767918088737,
+      "grad_norm": 0.4616141172796574,
+      "learning_rate": 2.1001709009441748e-05,
+      "loss": 0.328,
+      "step": 1176
+    },
+    {
+      "epoch": 2.6780432309442546,
+      "grad_norm": 0.4362223873012969,
+      "learning_rate": 2.096993408714823e-05,
+      "loss": 0.3581,
+      "step": 1177
+    },
+    {
+      "epoch": 2.6803185437997725,
+      "grad_norm": 0.44938021129645267,
+      "learning_rate": 2.0938156710667577e-05,
+      "loss": 0.3368,
+      "step": 1178
+    },
+    {
+      "epoch": 2.68259385665529,
+      "grad_norm": 0.3971849821913564,
+      "learning_rate": 2.090637696040486e-05,
+      "loss": 0.3143,
+      "step": 1179
+    },
+    {
+      "epoch": 2.684869169510808,
+      "grad_norm": 0.48071925305593577,
+      "learning_rate": 2.0874594916771174e-05,
+      "loss": 0.3384,
+      "step": 1180
+    },
+    {
+      "epoch": 2.6871444823663255,
+      "grad_norm": 0.4381908076397643,
+      "learning_rate": 2.0842810660183402e-05,
+      "loss": 0.3455,
+      "step": 1181
+    },
+    {
+      "epoch": 2.689419795221843,
+      "grad_norm": 0.41950514466306293,
+      "learning_rate": 2.0811024271064038e-05,
+      "loss": 0.3389,
+      "step": 1182
+    },
+    {
+      "epoch": 2.6916951080773606,
+      "grad_norm": 0.4563487817057512,
+      "learning_rate": 2.0779235829840962e-05,
+      "loss": 0.3111,
+      "step": 1183
+    },
+    {
+      "epoch": 2.693970420932878,
+      "grad_norm": 0.4075398413704911,
+      "learning_rate": 2.074744541694726e-05,
+      "loss": 0.3253,
+      "step": 1184
+    },
+    {
+      "epoch": 2.696245733788396,
+      "grad_norm": 0.4024468884794903,
+      "learning_rate": 2.071565311282099e-05,
+      "loss": 0.3174,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6985210466439136,
+      "grad_norm": 0.41381064381022864,
+      "learning_rate": 2.068385899790501e-05,
+      "loss": 0.3371,
+      "step": 1186
+    },
+    {
+      "epoch": 2.700796359499431,
+      "grad_norm": 0.45174817528660066,
+      "learning_rate": 2.065206315264675e-05,
+      "loss": 0.3092,
+      "step": 1187
+    },
+    {
+      "epoch": 2.703071672354949,
+      "grad_norm": 0.4081105192132397,
+      "learning_rate": 2.0620265657498027e-05,
+      "loss": 0.3376,
+      "step": 1188
+    },
+    {
+      "epoch": 2.7053469852104666,
+      "grad_norm": 0.46840498959393,
+      "learning_rate": 2.0588466592914814e-05,
+      "loss": 0.3236,
+      "step": 1189
+    },
+    {
+      "epoch": 2.707622298065984,
+      "grad_norm": 0.4219057713870426,
+      "learning_rate": 2.0556666039357084e-05,
+      "loss": 0.3657,
+      "step": 1190
+    },
+    {
+      "epoch": 2.7098976109215016,
+      "grad_norm": 0.4788694048573619,
+      "learning_rate": 2.052486407728855e-05,
+      "loss": 0.3299,
+      "step": 1191
+    },
+    {
+      "epoch": 2.712172923777019,
+      "grad_norm": 0.433904805616125,
+      "learning_rate": 2.0493060787176513e-05,
+      "loss": 0.3581,
+      "step": 1192
+    },
+    {
+      "epoch": 2.714448236632537,
+      "grad_norm": 0.3737957765771606,
+      "learning_rate": 2.046125624949161e-05,
+      "loss": 0.3351,
+      "step": 1193
+    },
+    {
+      "epoch": 2.7167235494880546,
+      "grad_norm": 0.4636212723722085,
+      "learning_rate": 2.042945054470766e-05,
+      "loss": 0.3623,
+      "step": 1194
+    },
+    {
+      "epoch": 2.718998862343572,
+      "grad_norm": 0.3880127067244525,
+      "learning_rate": 2.0397643753301403e-05,
+      "loss": 0.3585,
+      "step": 1195
+    },
+    {
+      "epoch": 2.72127417519909,
+      "grad_norm": 0.36654974143294056,
+      "learning_rate": 2.036583595575237e-05,
+      "loss": 0.3284,
+      "step": 1196
+    },
+    {
+      "epoch": 2.7235494880546076,
+      "grad_norm": 0.48034755356053316,
+      "learning_rate": 2.0334027232542595e-05,
+      "loss": 0.3599,
+      "step": 1197
+    },
+    {
+      "epoch": 2.725824800910125,
+      "grad_norm": 0.4418794637066644,
+      "learning_rate": 2.030221766415649e-05,
+      "loss": 0.3327,
+      "step": 1198
+    },
+    {
+      "epoch": 2.7281001137656427,
+      "grad_norm": 0.444200319161712,
+      "learning_rate": 2.0270407331080575e-05,
+      "loss": 0.3303,
+      "step": 1199
+    },
+    {
+      "epoch": 2.73037542662116,
+      "grad_norm": 0.4035643207657577,
+      "learning_rate": 2.0238596313803337e-05,
+      "loss": 0.3108,
+      "step": 1200
+    },
+    {
+      "epoch": 2.732650739476678,
+      "grad_norm": 0.3571362184850684,
+      "learning_rate": 2.0206784692814957e-05,
+      "loss": 0.3387,
+      "step": 1201
+    },
+    {
+      "epoch": 2.7349260523321957,
+      "grad_norm": 0.35134568741895283,
+      "learning_rate": 2.0174972548607183e-05,
+      "loss": 0.354,
+      "step": 1202
+    },
+    {
+      "epoch": 2.737201365187713,
+      "grad_norm": 0.41167143435227466,
+      "learning_rate": 2.014315996167305e-05,
+      "loss": 0.31,
+      "step": 1203
+    },
+    {
+      "epoch": 2.739476678043231,
+      "grad_norm": 0.3419684364181005,
+      "learning_rate": 2.0111347012506753e-05,
+      "loss": 0.3078,
+      "step": 1204
+    },
+    {
+      "epoch": 2.7417519908987487,
+      "grad_norm": 0.37900537937003825,
+      "learning_rate": 2.0079533781603352e-05,
+      "loss": 0.3555,
+      "step": 1205
+    },
+    {
+      "epoch": 2.744027303754266,
+      "grad_norm": 0.3933678389303421,
+      "learning_rate": 2.004772034945867e-05,
+      "loss": 0.3435,
+      "step": 1206
+    },
+    {
+      "epoch": 2.7463026166097837,
+      "grad_norm": 0.3721686834336099,
+      "learning_rate": 2.001590679656901e-05,
+      "loss": 0.333,
+      "step": 1207
+    },
+    {
+      "epoch": 2.748577929465301,
+      "grad_norm": 0.3885080139471526,
+      "learning_rate": 1.9984093203430996e-05,
+      "loss": 0.3395,
+      "step": 1208
+    },
+    {
+      "epoch": 2.750853242320819,
+      "grad_norm": 0.406519404842218,
+      "learning_rate": 1.9952279650541332e-05,
+      "loss": 0.3624,
+      "step": 1209
+    },
+    {
+      "epoch": 2.7531285551763367,
+      "grad_norm": 0.37505712731585505,
+      "learning_rate": 1.9920466218396655e-05,
+      "loss": 0.3239,
+      "step": 1210
+    },
+    {
+      "epoch": 2.755403868031854,
+      "grad_norm": 0.36610554514055926,
+      "learning_rate": 1.988865298749326e-05,
+      "loss": 0.3095,
+      "step": 1211
+    },
+    {
+      "epoch": 2.757679180887372,
+      "grad_norm": 0.39302726430386103,
+      "learning_rate": 1.985684003832695e-05,
+      "loss": 0.3196,
+      "step": 1212
+    },
+    {
+      "epoch": 2.7599544937428897,
+      "grad_norm": 0.39747897923075504,
+      "learning_rate": 1.982502745139282e-05,
+      "loss": 0.3199,
+      "step": 1213
+    },
+    {
+      "epoch": 2.7622298065984072,
+      "grad_norm": 0.3595503828398732,
+      "learning_rate": 1.9793215307185053e-05,
+      "loss": 0.305,
+      "step": 1214
+    },
+    {
+      "epoch": 2.7645051194539247,
+      "grad_norm": 0.396177560368738,
+      "learning_rate": 1.976140368619667e-05,
+      "loss": 0.3618,
+      "step": 1215
+    },
+    {
+      "epoch": 2.7667804323094427,
+      "grad_norm": 0.4676467159822378,
+      "learning_rate": 1.972959266891943e-05,
+      "loss": 0.3432,
+      "step": 1216
+    },
+    {
+      "epoch": 2.7690557451649602,
+      "grad_norm": 0.41532924076950134,
+      "learning_rate": 1.9697782335843512e-05,
+      "loss": 0.3342,
+      "step": 1217
+    },
+    {
+      "epoch": 2.7713310580204777,
+      "grad_norm": 0.386754144273086,
+      "learning_rate": 1.966597276745741e-05,
+      "loss": 0.3074,
+      "step": 1218
+    },
+    {
+      "epoch": 2.7736063708759957,
+      "grad_norm": 0.41603270499755085,
+      "learning_rate": 1.963416404424764e-05,
+      "loss": 0.3534,
+      "step": 1219
+    },
+    {
+      "epoch": 2.7758816837315132,
+      "grad_norm": 0.3754548451201784,
+      "learning_rate": 1.9602356246698597e-05,
+      "loss": 0.3066,
+      "step": 1220
+    },
+    {
+      "epoch": 2.7781569965870307,
+      "grad_norm": 0.40805461049030567,
+      "learning_rate": 1.9570549455292348e-05,
+      "loss": 0.3183,
+      "step": 1221
+    },
+    {
+      "epoch": 2.7804323094425483,
+      "grad_norm": 0.4006941290488259,
+      "learning_rate": 1.9538743750508394e-05,
+      "loss": 0.3018,
+      "step": 1222
+    },
+    {
+      "epoch": 2.782707622298066,
+      "grad_norm": 0.4156799110873144,
+      "learning_rate": 1.950693921282349e-05,
+      "loss": 0.3476,
+      "step": 1223
+    },
+    {
+      "epoch": 2.7849829351535837,
+      "grad_norm": 0.4697145256845964,
+      "learning_rate": 1.9475135922711457e-05,
+      "loss": 0.3574,
+      "step": 1224
+    },
+    {
+      "epoch": 2.7872582480091013,
+      "grad_norm": 0.3893570385324192,
+      "learning_rate": 1.944333396064292e-05,
+      "loss": 0.3426,
+      "step": 1225
+    },
+    {
+      "epoch": 2.789533560864619,
+      "grad_norm": 0.43090362034832597,
+      "learning_rate": 1.9411533407085192e-05,
+      "loss": 0.3393,
+      "step": 1226
+    },
+    {
+      "epoch": 2.7918088737201368,
+      "grad_norm": 0.42054757043063234,
+      "learning_rate": 1.9379734342501987e-05,
+      "loss": 0.3237,
+      "step": 1227
+    },
+    {
+      "epoch": 2.7940841865756543,
+      "grad_norm": 0.38221543475601394,
+      "learning_rate": 1.9347936847353254e-05,
+      "loss": 0.3028,
+      "step": 1228
+    },
+    {
+      "epoch": 2.796359499431172,
+      "grad_norm": 0.37763704583230345,
+      "learning_rate": 1.9316141002094995e-05,
+      "loss": 0.3236,
+      "step": 1229
+    },
+    {
+      "epoch": 2.7986348122866893,
+      "grad_norm": 0.4206697615376536,
+      "learning_rate": 1.9284346887179016e-05,
+      "loss": 0.3384,
+      "step": 1230
+    },
+    {
+      "epoch": 2.800910125142207,
+      "grad_norm": 0.3953685229204783,
+      "learning_rate": 1.9252554583052744e-05,
+      "loss": 0.3592,
+      "step": 1231
+    },
+    {
+      "epoch": 2.803185437997725,
+      "grad_norm": 0.3915164783946858,
+      "learning_rate": 1.9220764170159045e-05,
+      "loss": 0.3337,
+      "step": 1232
+    },
+    {
+      "epoch": 2.8054607508532423,
+      "grad_norm": 0.5187383959053034,
+      "learning_rate": 1.9188975728935965e-05,
+      "loss": 0.3502,
+      "step": 1233
+    },
+    {
+      "epoch": 2.80773606370876,
+      "grad_norm": 0.4636595479793326,
+      "learning_rate": 1.9157189339816608e-05,
+      "loss": 0.3574,
+      "step": 1234
+    },
+    {
+      "epoch": 2.810011376564278,
+      "grad_norm": 0.3752458006501017,
+      "learning_rate": 1.9125405083228833e-05,
+      "loss": 0.3211,
+      "step": 1235
+    },
+    {
+      "epoch": 2.8122866894197953,
+      "grad_norm": 0.422860682847553,
+      "learning_rate": 1.9093623039595144e-05,
+      "loss": 0.3473,
+      "step": 1236
+    },
+    {
+      "epoch": 2.814562002275313,
+      "grad_norm": 0.43453944438849534,
+      "learning_rate": 1.906184328933243e-05,
+      "loss": 0.3265,
+      "step": 1237
+    },
+    {
+      "epoch": 2.8168373151308304,
+      "grad_norm": 0.449663210415,
+      "learning_rate": 1.9030065912851775e-05,
+      "loss": 0.3485,
+      "step": 1238
+    },
+    {
+      "epoch": 2.819112627986348,
+      "grad_norm": 0.45165695123021987,
+      "learning_rate": 1.8998290990558255e-05,
+      "loss": 0.3074,
+      "step": 1239
+    },
+    {
+      "epoch": 2.821387940841866,
+      "grad_norm": 0.4446006954201955,
+      "learning_rate": 1.896651860285076e-05,
+      "loss": 0.3645,
+      "step": 1240
+    },
+    {
+      "epoch": 2.8236632536973834,
+      "grad_norm": 0.39348141345452986,
+      "learning_rate": 1.8934748830121722e-05,
+      "loss": 0.3201,
+      "step": 1241
+    },
+    {
+      "epoch": 2.825938566552901,
+      "grad_norm": 0.5341345156314935,
+      "learning_rate": 1.8902981752757004e-05,
+      "loss": 0.3291,
+      "step": 1242
+    },
+    {
+      "epoch": 2.828213879408419,
+      "grad_norm": 0.4285735656951308,
+      "learning_rate": 1.887121745113562e-05,
+      "loss": 0.332,
+      "step": 1243
+    },
+    {
+      "epoch": 2.8304891922639364,
+      "grad_norm": 0.4229697688971152,
+      "learning_rate": 1.8839456005629558e-05,
+      "loss": 0.3115,
+      "step": 1244
+    },
+    {
+      "epoch": 2.832764505119454,
+      "grad_norm": 0.4318299780335706,
+      "learning_rate": 1.8807697496603604e-05,
+      "loss": 0.4012,
+      "step": 1245
+    },
+    {
+      "epoch": 2.8350398179749714,
+      "grad_norm": 0.42178408290930636,
+      "learning_rate": 1.8775942004415094e-05,
+      "loss": 0.3323,
+      "step": 1246
+    },
+    {
+      "epoch": 2.837315130830489,
+      "grad_norm": 0.3723509665815239,
+      "learning_rate": 1.8744189609413733e-05,
+      "loss": 0.3237,
+      "step": 1247
+    },
+    {
+      "epoch": 2.839590443686007,
+      "grad_norm": 0.4305482774531959,
+      "learning_rate": 1.8712440391941396e-05,
+      "loss": 0.3472,
+      "step": 1248
+    },
+    {
+      "epoch": 2.8418657565415244,
+      "grad_norm": 0.46366347580937944,
+      "learning_rate": 1.8680694432331914e-05,
+      "loss": 0.3668,
+      "step": 1249
+    },
+    {
+      "epoch": 2.8441410693970424,
+      "grad_norm": 0.3705622266403713,
+      "learning_rate": 1.8648951810910878e-05,
+      "loss": 0.3143,
+      "step": 1250
+    },
+    {
+      "epoch": 2.84641638225256,
+      "grad_norm": 0.49100142288919724,
+      "learning_rate": 1.8617212607995423e-05,
+      "loss": 0.3751,
+      "step": 1251
+    },
+    {
+      "epoch": 2.8486916951080774,
+      "grad_norm": 0.3514112339230659,
+      "learning_rate": 1.8585476903894036e-05,
+      "loss": 0.2869,
+      "step": 1252
+    },
+    {
+      "epoch": 2.850967007963595,
+      "grad_norm": 0.4722074620404091,
+      "learning_rate": 1.855374477890637e-05,
+      "loss": 0.3468,
+      "step": 1253
+    },
+    {
+      "epoch": 2.8532423208191124,
+      "grad_norm": 0.45737612006959516,
+      "learning_rate": 1.8522016313323e-05,
+      "loss": 0.3847,
+      "step": 1254
+    },
+    {
+      "epoch": 2.8555176336746304,
+      "grad_norm": 0.4049743949427917,
+      "learning_rate": 1.849029158742525e-05,
+      "loss": 0.3611,
+      "step": 1255
+    },
+    {
+      "epoch": 2.857792946530148,
+      "grad_norm": 0.43874858432410946,
+      "learning_rate": 1.845857068148498e-05,
+      "loss": 0.3354,
+      "step": 1256
+    },
+    {
+      "epoch": 2.8600682593856654,
+      "grad_norm": 0.40492277366109236,
+      "learning_rate": 1.8426853675764386e-05,
+      "loss": 0.3296,
+      "step": 1257
+    },
+    {
+      "epoch": 2.8623435722411834,
+      "grad_norm": 0.3472960988318782,
+      "learning_rate": 1.8395140650515803e-05,
+      "loss": 0.3263,
+      "step": 1258
+    },
+    {
+      "epoch": 2.864618885096701,
+      "grad_norm": 0.44332665688391126,
+      "learning_rate": 1.8363431685981476e-05,
+      "loss": 0.331,
+      "step": 1259
+    },
+    {
+      "epoch": 2.8668941979522184,
+      "grad_norm": 0.42360608795436905,
+      "learning_rate": 1.8331726862393385e-05,
+      "loss": 0.3614,
+      "step": 1260
+    },
+    {
+      "epoch": 2.869169510807736,
+      "grad_norm": 0.38736693003836803,
+      "learning_rate": 1.830002625997304e-05,
+      "loss": 0.3335,
+      "step": 1261
+    },
+    {
+      "epoch": 2.8714448236632535,
+      "grad_norm": 0.4562383953624614,
+      "learning_rate": 1.8268329958931267e-05,
+      "loss": 0.3251,
+      "step": 1262
+    },
+    {
+      "epoch": 2.8737201365187715,
+      "grad_norm": 0.40212087556281695,
+      "learning_rate": 1.823663803946799e-05,
+      "loss": 0.3506,
+      "step": 1263
+    },
+    {
+      "epoch": 2.875995449374289,
+      "grad_norm": 0.4729884947349188,
+      "learning_rate": 1.8204950581772073e-05,
+      "loss": 0.3489,
+      "step": 1264
+    },
+    {
+      "epoch": 2.8782707622298065,
+      "grad_norm": 0.41992748172255673,
+      "learning_rate": 1.8173267666021075e-05,
+      "loss": 0.3485,
+      "step": 1265
+    },
+    {
+      "epoch": 2.8805460750853245,
+      "grad_norm": 0.4043140099911277,
+      "learning_rate": 1.8141589372381073e-05,
+      "loss": 0.3169,
+      "step": 1266
+    },
+    {
+      "epoch": 2.882821387940842,
+      "grad_norm": 0.40756018663020205,
+      "learning_rate": 1.8109915781006436e-05,
+      "loss": 0.3203,
+      "step": 1267
+    },
+    {
+      "epoch": 2.8850967007963595,
+      "grad_norm": 0.4684773385227538,
+      "learning_rate": 1.8078246972039633e-05,
+      "loss": 0.3073,
+      "step": 1268
+    },
+    {
+      "epoch": 2.887372013651877,
+      "grad_norm": 0.3723068919285976,
+      "learning_rate": 1.804658302561106e-05,
+      "loss": 0.3302,
+      "step": 1269
+    },
+    {
+      "epoch": 2.8896473265073945,
+      "grad_norm": 0.3673828907740494,
+      "learning_rate": 1.8014924021838774e-05,
+      "loss": 0.3176,
+      "step": 1270
+    },
+    {
+      "epoch": 2.8919226393629125,
+      "grad_norm": 0.4056625374216179,
+      "learning_rate": 1.7983270040828346e-05,
+      "loss": 0.3295,
+      "step": 1271
+    },
+    {
+      "epoch": 2.89419795221843,
+      "grad_norm": 0.32625489285710857,
+      "learning_rate": 1.795162116267264e-05,
+      "loss": 0.3181,
+      "step": 1272
+    },
+    {
+      "epoch": 2.8964732650739475,
+      "grad_norm": 0.3576177266546512,
+      "learning_rate": 1.7919977467451593e-05,
+      "loss": 0.3178,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8987485779294655,
+      "grad_norm": 0.3681071605577923,
+      "learning_rate": 1.7888339035232047e-05,
+      "loss": 0.3428,
+      "step": 1274
+    },
+    {
+      "epoch": 2.901023890784983,
+      "grad_norm": 0.34539372920147526,
+      "learning_rate": 1.785670594606752e-05,
+      "loss": 0.333,
+      "step": 1275
+    },
+    {
+      "epoch": 2.9032992036405005,
+      "grad_norm": 0.37170066511097377,
+      "learning_rate": 1.782507827999799e-05,
+      "loss": 0.3254,
+      "step": 1276
+    },
+    {
+      "epoch": 2.905574516496018,
+      "grad_norm": 0.39952313728459865,
+      "learning_rate": 1.7793456117049746e-05,
+      "loss": 0.3253,
+      "step": 1277
+    },
+    {
+      "epoch": 2.9078498293515356,
+      "grad_norm": 0.354210897652707,
+      "learning_rate": 1.776183953723514e-05,
+      "loss": 0.3317,
+      "step": 1278
+    },
+    {
+      "epoch": 2.9101251422070535,
+      "grad_norm": 0.3652645740006678,
+      "learning_rate": 1.773022862055238e-05,
+      "loss": 0.3138,
+      "step": 1279
+    },
+    {
+      "epoch": 2.912400455062571,
+      "grad_norm": 0.3888678131084162,
+      "learning_rate": 1.7698623446985378e-05,
+      "loss": 0.3177,
+      "step": 1280
+    },
+    {
+      "epoch": 2.9146757679180886,
+      "grad_norm": 0.3861332065384473,
+      "learning_rate": 1.7667024096503485e-05,
+      "loss": 0.3341,
+      "step": 1281
+    },
+    {
+      "epoch": 2.9169510807736065,
+      "grad_norm": 0.3968878850605964,
+      "learning_rate": 1.7635430649061333e-05,
+      "loss": 0.3119,
+      "step": 1282
+    },
+    {
+      "epoch": 2.919226393629124,
+      "grad_norm": 0.3692324698525607,
+      "learning_rate": 1.7603843184598623e-05,
+      "loss": 0.303,
+      "step": 1283
+    },
+    {
+      "epoch": 2.9215017064846416,
+      "grad_norm": 0.3809675871957472,
+      "learning_rate": 1.757226178303989e-05,
+      "loss": 0.319,
+      "step": 1284
+    },
+    {
+      "epoch": 2.923777019340159,
+      "grad_norm": 0.38478746620596943,
+      "learning_rate": 1.754068652429436e-05,
+      "loss": 0.3243,
+      "step": 1285
+    },
+    {
+      "epoch": 2.926052332195677,
+      "grad_norm": 0.46139888708407795,
+      "learning_rate": 1.75091174882557e-05,
+      "loss": 0.3146,
+      "step": 1286
+    },
+    {
+      "epoch": 2.9283276450511946,
+      "grad_norm": 0.39783512322413067,
+      "learning_rate": 1.7477554754801835e-05,
+      "loss": 0.3275,
+      "step": 1287
+    },
+    {
+      "epoch": 2.930602957906712,
+      "grad_norm": 0.48084823548951655,
+      "learning_rate": 1.7445998403794744e-05,
+      "loss": 0.3486,
+      "step": 1288
+    },
+    {
+      "epoch": 2.93287827076223,
+      "grad_norm": 0.40567899003023666,
+      "learning_rate": 1.7414448515080253e-05,
+      "loss": 0.3692,
+      "step": 1289
+    },
+    {
+      "epoch": 2.9351535836177476,
+      "grad_norm": 0.383209141100548,
+      "learning_rate": 1.738290516848784e-05,
+      "loss": 0.3181,
+      "step": 1290
+    },
+    {
+      "epoch": 2.937428896473265,
+      "grad_norm": 0.43877459749694897,
+      "learning_rate": 1.7351368443830438e-05,
+      "loss": 0.3048,
+      "step": 1291
+    },
+    {
+      "epoch": 2.9397042093287826,
+      "grad_norm": 0.40759673361135945,
+      "learning_rate": 1.7319838420904194e-05,
+      "loss": 0.3534,
+      "step": 1292
+    },
+    {
+      "epoch": 2.9419795221843,
+      "grad_norm": 0.39045373617754325,
+      "learning_rate": 1.7288315179488348e-05,
+      "loss": 0.3732,
+      "step": 1293
+    },
+    {
+      "epoch": 2.944254835039818,
+      "grad_norm": 0.408840585877605,
+      "learning_rate": 1.725679879934493e-05,
+      "loss": 0.3264,
+      "step": 1294
+    },
+    {
+      "epoch": 2.9465301478953356,
+      "grad_norm": 0.3816064165282624,
+      "learning_rate": 1.722528936021864e-05,
+      "loss": 0.3316,
+      "step": 1295
+    },
+    {
+      "epoch": 2.948805460750853,
+      "grad_norm": 0.40041531131659197,
+      "learning_rate": 1.7193786941836608e-05,
+      "loss": 0.344,
+      "step": 1296
+    },
+    {
+      "epoch": 2.951080773606371,
+      "grad_norm": 0.41058583751790734,
+      "learning_rate": 1.7162291623908194e-05,
+      "loss": 0.2957,
+      "step": 1297
+    },
+    {
+      "epoch": 2.9533560864618886,
+      "grad_norm": 0.3798093592582125,
+      "learning_rate": 1.7130803486124806e-05,
+      "loss": 0.3269,
+      "step": 1298
+    },
+    {
+      "epoch": 2.955631399317406,
+      "grad_norm": 0.4104075641164096,
+      "learning_rate": 1.7099322608159676e-05,
+      "loss": 0.3102,
+      "step": 1299
+    },
+    {
+      "epoch": 2.9579067121729237,
+      "grad_norm": 0.40594869779106296,
+      "learning_rate": 1.7067849069667644e-05,
+      "loss": 0.3479,
+      "step": 1300
+    },
+    {
+      "epoch": 2.960182025028441,
+      "grad_norm": 0.3826029772595492,
+      "learning_rate": 1.703638295028503e-05,
+      "loss": 0.3491,
+      "step": 1301
+    },
+    {
+      "epoch": 2.962457337883959,
+      "grad_norm": 0.3721508444462159,
+      "learning_rate": 1.7004924329629335e-05,
+      "loss": 0.3328,
+      "step": 1302
+    },
+    {
+      "epoch": 2.9647326507394767,
+      "grad_norm": 0.4020437609324411,
+      "learning_rate": 1.6973473287299114e-05,
+      "loss": 0.3334,
+      "step": 1303
+    },
+    {
+      "epoch": 2.967007963594994,
+      "grad_norm": 0.393328296256328,
+      "learning_rate": 1.694202990287373e-05,
+      "loss": 0.307,
+      "step": 1304
+    },
+    {
+      "epoch": 2.969283276450512,
+      "grad_norm": 0.3841804775756949,
+      "learning_rate": 1.6910594255913177e-05,
+      "loss": 0.3198,
+      "step": 1305
+    },
+    {
+      "epoch": 2.9715585893060297,
+      "grad_norm": 0.41379994210505555,
+      "learning_rate": 1.687916642595788e-05,
+      "loss": 0.3159,
+      "step": 1306
+    },
+    {
+      "epoch": 2.973833902161547,
+      "grad_norm": 0.3975357129785672,
+      "learning_rate": 1.6847746492528467e-05,
+      "loss": 0.3228,
+      "step": 1307
+    },
+    {
+      "epoch": 2.9761092150170647,
+      "grad_norm": 0.35636419669271596,
+      "learning_rate": 1.6816334535125583e-05,
+      "loss": 0.3512,
+      "step": 1308
+    },
+    {
+      "epoch": 2.9783845278725822,
+      "grad_norm": 0.38436643906411494,
+      "learning_rate": 1.6784930633229734e-05,
+      "loss": 0.3247,
+      "step": 1309
+    },
+    {
+      "epoch": 2.9806598407281,
+      "grad_norm": 0.339825615244889,
+      "learning_rate": 1.6753534866300983e-05,
+      "loss": 0.3105,
+      "step": 1310
+    },
+    {
+      "epoch": 2.9829351535836177,
+      "grad_norm": 0.36635896773075227,
+      "learning_rate": 1.6722147313778855e-05,
+      "loss": 0.2991,
+      "step": 1311
+    },
+    {
+      "epoch": 2.9852104664391352,
+      "grad_norm": 0.3696431495703968,
+      "learning_rate": 1.669076805508207e-05,
+      "loss": 0.324,
+      "step": 1312
+    },
+    {
+      "epoch": 2.987485779294653,
+      "grad_norm": 0.3931605856930635,
+      "learning_rate": 1.6659397169608363e-05,
+      "loss": 0.333,
+      "step": 1313
+    },
+    {
+      "epoch": 2.9897610921501707,
+      "grad_norm": 0.4032923555871362,
+      "learning_rate": 1.6628034736734293e-05,
+      "loss": 0.3486,
+      "step": 1314
+    },
+    {
+      "epoch": 2.9920364050056882,
+      "grad_norm": 0.37102792124122086,
+      "learning_rate": 1.6596680835815024e-05,
+      "loss": 0.325,
+      "step": 1315
+    },
+    {
+      "epoch": 2.9943117178612058,
+      "grad_norm": 0.3843008555876691,
+      "learning_rate": 1.6565335546184118e-05,
+      "loss": 0.3367,
+      "step": 1316
+    },
+    {
+      "epoch": 2.9965870307167233,
+      "grad_norm": 0.3416313482353009,
+      "learning_rate": 1.6533998947153393e-05,
+      "loss": 0.3025,
+      "step": 1317
+    },
+    {
+      "epoch": 2.9988623435722412,
+      "grad_norm": 0.4127951133201824,
+      "learning_rate": 1.650267111801261e-05,
+      "loss": 0.3555,
+      "step": 1318
+    },
+    {
+      "epoch": 3.0011376564277588,
+      "grad_norm": 0.6665390187415182,
+      "learning_rate": 1.6471352138029414e-05,
+      "loss": 0.388,
+      "step": 1319
+    },
+    {
+      "epoch": 3.0034129692832763,
+      "grad_norm": 0.5122750406971657,
+      "learning_rate": 1.6440042086449e-05,
+      "loss": 0.2487,
+      "step": 1320
+    },
+    {
+      "epoch": 3.0056882821387942,
+      "grad_norm": 0.42385291143597537,
+      "learning_rate": 1.6408741042493996e-05,
+      "loss": 0.2772,
+      "step": 1321
+    },
+    {
+      "epoch": 3.0079635949943118,
+      "grad_norm": 0.650654703167525,
+      "learning_rate": 1.6377449085364246e-05,
+      "loss": 0.2536,
+      "step": 1322
+    },
+    {
+      "epoch": 3.0102389078498293,
+      "grad_norm": 0.4489795360940919,
+      "learning_rate": 1.6346166294236593e-05,
+      "loss": 0.2202,
+      "step": 1323
+    },
+    {
+      "epoch": 3.012514220705347,
+      "grad_norm": 0.49251096342350414,
+      "learning_rate": 1.631489274826467e-05,
+      "loss": 0.2449,
+      "step": 1324
+    },
+    {
+      "epoch": 3.0147895335608648,
+      "grad_norm": 0.5871177283470034,
+      "learning_rate": 1.628362852657876e-05,
+      "loss": 0.2506,
+      "step": 1325
+    },
+    {
+      "epoch": 3.0170648464163823,
+      "grad_norm": 0.45158096813826065,
+      "learning_rate": 1.6252373708285505e-05,
+      "loss": 0.2425,
+      "step": 1326
+    },
+    {
+      "epoch": 3.0193401592719,
+      "grad_norm": 0.395446791937916,
+      "learning_rate": 1.62211283724678e-05,
+      "loss": 0.2401,
+      "step": 1327
+    },
+    {
+      "epoch": 3.0216154721274173,
+      "grad_norm": 0.5740116419937004,
+      "learning_rate": 1.6189892598184504e-05,
+      "loss": 0.2741,
+      "step": 1328
+    },
+    {
+      "epoch": 3.0238907849829353,
+      "grad_norm": 0.47098133857587854,
+      "learning_rate": 1.615866646447031e-05,
+      "loss": 0.2535,
+      "step": 1329
+    },
+    {
+      "epoch": 3.026166097838453,
+      "grad_norm": 0.4120517637411914,
+      "learning_rate": 1.612745005033551e-05,
+      "loss": 0.2304,
+      "step": 1330
+    },
+    {
+      "epoch": 3.0284414106939703,
+      "grad_norm": 0.4875077502609603,
+      "learning_rate": 1.6096243434765808e-05,
+      "loss": 0.2388,
+      "step": 1331
+    },
+    {
+      "epoch": 3.030716723549488,
+      "grad_norm": 0.3888715470441953,
+      "learning_rate": 1.6065046696722094e-05,
+      "loss": 0.1889,
+      "step": 1332
+    },
+    {
+      "epoch": 3.032992036405006,
+      "grad_norm": 0.4404723669577395,
+      "learning_rate": 1.603385991514031e-05,
+      "loss": 0.3357,
+      "step": 1333
+    },
+    {
+      "epoch": 3.0352673492605233,
+      "grad_norm": 0.3986029461724571,
+      "learning_rate": 1.600268316893114e-05,
+      "loss": 0.2574,
+      "step": 1334
+    },
+    {
+      "epoch": 3.037542662116041,
+      "grad_norm": 0.37205035189042457,
+      "learning_rate": 1.597151653697995e-05,
+      "loss": 0.2527,
+      "step": 1335
+    },
+    {
+      "epoch": 3.039817974971559,
+      "grad_norm": 0.3889281620378717,
+      "learning_rate": 1.594036009814646e-05,
+      "loss": 0.2165,
+      "step": 1336
+    },
+    {
+      "epoch": 3.0420932878270763,
+      "grad_norm": 0.4038841594833209,
+      "learning_rate": 1.590921393126461e-05,
+      "loss": 0.2385,
+      "step": 1337
+    },
+    {
+      "epoch": 3.044368600682594,
+      "grad_norm": 0.3538892185732872,
+      "learning_rate": 1.5878078115142372e-05,
+      "loss": 0.2484,
+      "step": 1338
+    },
+    {
+      "epoch": 3.0466439135381114,
+      "grad_norm": 0.4581473851645579,
+      "learning_rate": 1.5846952728561496e-05,
+      "loss": 0.2906,
+      "step": 1339
+    },
+    {
+      "epoch": 3.0489192263936293,
+      "grad_norm": 0.3344402981437661,
+      "learning_rate": 1.581583785027736e-05,
+      "loss": 0.2073,
+      "step": 1340
+    },
+    {
+      "epoch": 3.051194539249147,
+      "grad_norm": 0.4125810966065969,
+      "learning_rate": 1.5784733559018762e-05,
+      "loss": 0.2686,
+      "step": 1341
+    },
+    {
+      "epoch": 3.0534698521046644,
+      "grad_norm": 0.37740963579799397,
+      "learning_rate": 1.5753639933487684e-05,
+      "loss": 0.2303,
+      "step": 1342
+    },
+    {
+      "epoch": 3.055745164960182,
+      "grad_norm": 0.3212141400616197,
+      "learning_rate": 1.572255705235916e-05,
+      "loss": 0.203,
+      "step": 1343
+    },
+    {
+      "epoch": 3.0580204778157,
+      "grad_norm": 0.3867866079306578,
+      "learning_rate": 1.5691484994280996e-05,
+      "loss": 0.2369,
+      "step": 1344
+    },
+    {
+      "epoch": 3.0602957906712174,
+      "grad_norm": 0.34234519810414454,
+      "learning_rate": 1.566042383787364e-05,
+      "loss": 0.2245,
+      "step": 1345
+    },
+    {
+      "epoch": 3.062571103526735,
+      "grad_norm": 0.37375415604623985,
+      "learning_rate": 1.562937366172995e-05,
+      "loss": 0.2603,
+      "step": 1346
+    },
+    {
+      "epoch": 3.0648464163822524,
+      "grad_norm": 0.3658935669421099,
+      "learning_rate": 1.5598334544415e-05,
+      "loss": 0.2506,
+      "step": 1347
+    },
+    {
+      "epoch": 3.0671217292377704,
+      "grad_norm": 0.3435077709797847,
+      "learning_rate": 1.556730656446588e-05,
+      "loss": 0.2396,
+      "step": 1348
+    },
+    {
+      "epoch": 3.069397042093288,
+      "grad_norm": 0.3594998561611038,
+      "learning_rate": 1.5536289800391504e-05,
+      "loss": 0.2223,
+      "step": 1349
+    },
+    {
+      "epoch": 3.0716723549488054,
+      "grad_norm": 0.34511347334216075,
+      "learning_rate": 1.550528433067239e-05,
+      "loss": 0.2382,
+      "step": 1350
+    },
+    {
+      "epoch": 3.073947667804323,
+      "grad_norm": 0.3397290693234507,
+      "learning_rate": 1.5474290233760522e-05,
+      "loss": 0.2775,
+      "step": 1351
+    },
+    {
+      "epoch": 3.076222980659841,
+      "grad_norm": 0.36915212461759794,
+      "learning_rate": 1.544330758807906e-05,
+      "loss": 0.243,
+      "step": 1352
+    },
+    {
+      "epoch": 3.0784982935153584,
+      "grad_norm": 0.3343502303670519,
+      "learning_rate": 1.5412336472022198e-05,
+      "loss": 0.2081,
+      "step": 1353
+    },
+    {
+      "epoch": 3.080773606370876,
+      "grad_norm": 0.3612524770320426,
+      "learning_rate": 1.5381376963954986e-05,
+      "loss": 0.2513,
+      "step": 1354
+    },
+    {
+      "epoch": 3.0830489192263935,
+      "grad_norm": 0.3609904278246174,
+      "learning_rate": 1.5350429142213075e-05,
+      "loss": 0.2527,
+      "step": 1355
+    },
+    {
+      "epoch": 3.0853242320819114,
+      "grad_norm": 0.34163028966822784,
+      "learning_rate": 1.531949308510255e-05,
+      "loss": 0.2375,
+      "step": 1356
+    },
+    {
+      "epoch": 3.087599544937429,
+      "grad_norm": 0.43019417255945536,
+      "learning_rate": 1.5288568870899754e-05,
+      "loss": 0.2815,
+      "step": 1357
+    },
+    {
+      "epoch": 3.0898748577929465,
+      "grad_norm": 0.37950466145542605,
+      "learning_rate": 1.5257656577851014e-05,
+      "loss": 0.2347,
+      "step": 1358
+    },
+    {
+      "epoch": 3.092150170648464,
+      "grad_norm": 0.6184686430893538,
+      "learning_rate": 1.5226756284172555e-05,
+      "loss": 0.2191,
+      "step": 1359
+    },
+    {
+      "epoch": 3.094425483503982,
+      "grad_norm": 0.34092582499384705,
+      "learning_rate": 1.5195868068050185e-05,
+      "loss": 0.238,
+      "step": 1360
+    },
+    {
+      "epoch": 3.0967007963594995,
+      "grad_norm": 0.37356227814658555,
+      "learning_rate": 1.5164992007639183e-05,
+      "loss": 0.2783,
+      "step": 1361
+    },
+    {
+      "epoch": 3.098976109215017,
+      "grad_norm": 0.33800457672474493,
+      "learning_rate": 1.513412818106407e-05,
+      "loss": 0.2092,
+      "step": 1362
+    },
+    {
+      "epoch": 3.1012514220705345,
+      "grad_norm": 0.3345134707481074,
+      "learning_rate": 1.5103276666418404e-05,
+      "loss": 0.2522,
+      "step": 1363
+    },
+    {
+      "epoch": 3.1035267349260525,
+      "grad_norm": 0.35279615944744536,
+      "learning_rate": 1.5072437541764585e-05,
+      "loss": 0.201,
+      "step": 1364
+    },
+    {
+      "epoch": 3.10580204778157,
+      "grad_norm": 0.390946887337508,
+      "learning_rate": 1.5041610885133681e-05,
+      "loss": 0.293,
+      "step": 1365
+    },
+    {
+      "epoch": 3.1080773606370875,
+      "grad_norm": 0.33124360951531784,
+      "learning_rate": 1.5010796774525197e-05,
+      "loss": 0.2424,
+      "step": 1366
+    },
+    {
+      "epoch": 3.110352673492605,
+      "grad_norm": 0.3687294463727865,
+      "learning_rate": 1.497999528790691e-05,
+      "loss": 0.25,
+      "step": 1367
+    },
+    {
+      "epoch": 3.112627986348123,
+      "grad_norm": 0.3592837699433852,
+      "learning_rate": 1.4949206503214631e-05,
+      "loss": 0.2348,
+      "step": 1368
+    },
+    {
+      "epoch": 3.1149032992036405,
+      "grad_norm": 0.32555904443408795,
+      "learning_rate": 1.4918430498352043e-05,
+      "loss": 0.2153,
+      "step": 1369
+    },
+    {
+      "epoch": 3.117178612059158,
+      "grad_norm": 0.3604675192479075,
+      "learning_rate": 1.4887667351190508e-05,
+      "loss": 0.2582,
+      "step": 1370
+    },
+    {
+      "epoch": 3.1194539249146755,
+      "grad_norm": 0.34282164163989276,
+      "learning_rate": 1.4856917139568836e-05,
+      "loss": 0.2364,
+      "step": 1371
+    },
+    {
+      "epoch": 3.1217292377701935,
+      "grad_norm": 0.3370745107292784,
+      "learning_rate": 1.4826179941293104e-05,
+      "loss": 0.2358,
+      "step": 1372
+    },
+    {
+      "epoch": 3.124004550625711,
+      "grad_norm": 0.3347171832918944,
+      "learning_rate": 1.479545583413648e-05,
+      "loss": 0.2577,
+      "step": 1373
+    },
+    {
+      "epoch": 3.1262798634812285,
+      "grad_norm": 0.3828614334838476,
+      "learning_rate": 1.4764744895838989e-05,
+      "loss": 0.227,
+      "step": 1374
+    },
+    {
+      "epoch": 3.1285551763367465,
+      "grad_norm": 0.3428856010232504,
+      "learning_rate": 1.4734047204107358e-05,
+      "loss": 0.2004,
+      "step": 1375
+    },
+    {
+      "epoch": 3.130830489192264,
+      "grad_norm": 0.369809547171071,
+      "learning_rate": 1.4703362836614767e-05,
+      "loss": 0.2382,
+      "step": 1376
+    },
+    {
+      "epoch": 3.1331058020477816,
+      "grad_norm": 0.35765496336791786,
+      "learning_rate": 1.4672691871000706e-05,
+      "loss": 0.2573,
+      "step": 1377
+    },
+    {
+      "epoch": 3.135381114903299,
+      "grad_norm": 0.35443189050242446,
+      "learning_rate": 1.4642034384870747e-05,
+      "loss": 0.2508,
+      "step": 1378
+    },
+    {
+      "epoch": 3.137656427758817,
+      "grad_norm": 0.405676939837812,
+      "learning_rate": 1.461139045579635e-05,
+      "loss": 0.2552,
+      "step": 1379
+    },
+    {
+      "epoch": 3.1399317406143346,
+      "grad_norm": 0.31010535796515454,
+      "learning_rate": 1.458076016131468e-05,
+      "loss": 0.2213,
+      "step": 1380
+    },
+    {
+      "epoch": 3.142207053469852,
+      "grad_norm": 0.36707703136955583,
+      "learning_rate": 1.4550143578928406e-05,
+      "loss": 0.2364,
+      "step": 1381
+    },
+    {
+      "epoch": 3.1444823663253696,
+      "grad_norm": 0.4133050899168252,
+      "learning_rate": 1.4519540786105486e-05,
+      "loss": 0.2687,
+      "step": 1382
+    },
+    {
+      "epoch": 3.1467576791808876,
+      "grad_norm": 0.3381717821117065,
+      "learning_rate": 1.448895186027901e-05,
+      "loss": 0.2262,
+      "step": 1383
+    },
+    {
+      "epoch": 3.149032992036405,
+      "grad_norm": 0.38820845950823396,
+      "learning_rate": 1.4458376878846956e-05,
+      "loss": 0.2675,
+      "step": 1384
+    },
+    {
+      "epoch": 3.1513083048919226,
+      "grad_norm": 0.34061659959440704,
+      "learning_rate": 1.442781591917203e-05,
+      "loss": 0.2155,
+      "step": 1385
+    },
+    {
+      "epoch": 3.15358361774744,
+      "grad_norm": 0.3603746448529638,
+      "learning_rate": 1.4397269058581467e-05,
+      "loss": 0.2396,
+      "step": 1386
+    },
+    {
+      "epoch": 3.155858930602958,
+      "grad_norm": 0.3534849537926216,
+      "learning_rate": 1.4366736374366817e-05,
+      "loss": 0.2371,
+      "step": 1387
+    },
+    {
+      "epoch": 3.1581342434584756,
+      "grad_norm": 0.40457147006129407,
+      "learning_rate": 1.4336217943783761e-05,
+      "loss": 0.2599,
+      "step": 1388
+    },
+    {
+      "epoch": 3.160409556313993,
+      "grad_norm": 0.3666157730553102,
+      "learning_rate": 1.4305713844051925e-05,
+      "loss": 0.2114,
+      "step": 1389
+    },
+    {
+      "epoch": 3.1626848691695106,
+      "grad_norm": 0.3294560042638884,
+      "learning_rate": 1.4275224152354658e-05,
+      "loss": 0.2038,
+      "step": 1390
+    },
+    {
+      "epoch": 3.1649601820250286,
+      "grad_norm": 0.3631745191632101,
+      "learning_rate": 1.4244748945838869e-05,
+      "loss": 0.2696,
+      "step": 1391
+    },
+    {
+      "epoch": 3.167235494880546,
+      "grad_norm": 0.37962273786727296,
+      "learning_rate": 1.4214288301614806e-05,
+      "loss": 0.2634,
+      "step": 1392
+    },
+    {
+      "epoch": 3.1695108077360636,
+      "grad_norm": 0.35320895287602805,
+      "learning_rate": 1.4183842296755865e-05,
+      "loss": 0.2376,
+      "step": 1393
+    },
+    {
+      "epoch": 3.171786120591581,
+      "grad_norm": 0.3259971048184194,
+      "learning_rate": 1.4153411008298416e-05,
+      "loss": 0.2567,
+      "step": 1394
+    },
+    {
+      "epoch": 3.174061433447099,
+      "grad_norm": 0.35611087287793414,
+      "learning_rate": 1.4122994513241587e-05,
+      "loss": 0.2534,
+      "step": 1395
+    },
+    {
+      "epoch": 3.1763367463026166,
+      "grad_norm": 0.3148922392334282,
+      "learning_rate": 1.4092592888547061e-05,
+      "loss": 0.2313,
+      "step": 1396
+    },
+    {
+      "epoch": 3.178612059158134,
+      "grad_norm": 0.326721621928511,
+      "learning_rate": 1.4062206211138917e-05,
+      "loss": 0.2369,
+      "step": 1397
+    },
+    {
+      "epoch": 3.1808873720136517,
+      "grad_norm": 0.3471825310924632,
+      "learning_rate": 1.4031834557903399e-05,
+      "loss": 0.2141,
+      "step": 1398
+    },
+    {
+      "epoch": 3.1831626848691696,
+      "grad_norm": 0.3290346572256407,
+      "learning_rate": 1.4001478005688744e-05,
+      "loss": 0.2045,
+      "step": 1399
+    },
+    {
+      "epoch": 3.185437997724687,
+      "grad_norm": 0.36785750709054627,
+      "learning_rate": 1.3971136631304978e-05,
+      "loss": 0.24,
+      "step": 1400
+    },
+    {
+      "epoch": 3.1877133105802047,
+      "grad_norm": 0.3360590731380939,
+      "learning_rate": 1.3940810511523708e-05,
+      "loss": 0.2346,
+      "step": 1401
+    },
+    {
+      "epoch": 3.189988623435722,
+      "grad_norm": 0.33235200603477055,
+      "learning_rate": 1.3910499723077965e-05,
+      "loss": 0.213,
+      "step": 1402
+    },
+    {
+      "epoch": 3.19226393629124,
+      "grad_norm": 0.3554694875952007,
+      "learning_rate": 1.3880204342661979e-05,
+      "loss": 0.2308,
+      "step": 1403
+    },
+    {
+      "epoch": 3.1945392491467577,
+      "grad_norm": 0.3368960682666124,
+      "learning_rate": 1.3849924446930985e-05,
+      "loss": 0.2642,
+      "step": 1404
+    },
+    {
+      "epoch": 3.196814562002275,
+      "grad_norm": 0.34050458371850023,
+      "learning_rate": 1.3819660112501054e-05,
+      "loss": 0.271,
+      "step": 1405
+    },
+    {
+      "epoch": 3.199089874857793,
+      "grad_norm": 0.3240902133031251,
+      "learning_rate": 1.3789411415948868e-05,
+      "loss": 0.2089,
+      "step": 1406
+    },
+    {
+      "epoch": 3.2013651877133107,
+      "grad_norm": 0.34732714520787034,
+      "learning_rate": 1.3759178433811554e-05,
+      "loss": 0.2398,
+      "step": 1407
+    },
+    {
+      "epoch": 3.203640500568828,
+      "grad_norm": 0.3619506043549804,
+      "learning_rate": 1.3728961242586474e-05,
+      "loss": 0.2597,
+      "step": 1408
+    },
+    {
+      "epoch": 3.2059158134243457,
+      "grad_norm": 0.32982359532241645,
+      "learning_rate": 1.3698759918731018e-05,
+      "loss": 0.2488,
+      "step": 1409
+    },
+    {
+      "epoch": 3.2081911262798632,
+      "grad_norm": 0.36132873965131257,
+      "learning_rate": 1.3668574538662451e-05,
+      "loss": 0.2369,
+      "step": 1410
+    },
+    {
+      "epoch": 3.210466439135381,
+      "grad_norm": 0.34382437372925284,
+      "learning_rate": 1.3638405178757682e-05,
+      "loss": 0.2507,
+      "step": 1411
+    },
+    {
+      "epoch": 3.2127417519908987,
+      "grad_norm": 0.33668161325420715,
+      "learning_rate": 1.3608251915353092e-05,
+      "loss": 0.2482,
+      "step": 1412
+    },
+    {
+      "epoch": 3.2150170648464163,
+      "grad_norm": 0.3340776386733362,
+      "learning_rate": 1.357811482474433e-05,
+      "loss": 0.2056,
+      "step": 1413
+    },
+    {
+      "epoch": 3.217292377701934,
+      "grad_norm": 0.35745487006805743,
+      "learning_rate": 1.3547993983186122e-05,
+      "loss": 0.2439,
+      "step": 1414
+    },
+    {
+      "epoch": 3.2195676905574517,
+      "grad_norm": 0.32125462699455243,
+      "learning_rate": 1.3517889466892085e-05,
+      "loss": 0.2188,
+      "step": 1415
+    },
+    {
+      "epoch": 3.2218430034129693,
+      "grad_norm": 0.31510471304694576,
+      "learning_rate": 1.348780135203453e-05,
+      "loss": 0.2128,
+      "step": 1416
+    },
+    {
+      "epoch": 3.2241183162684868,
+      "grad_norm": 0.3405797577146921,
+      "learning_rate": 1.3457729714744247e-05,
+      "loss": 0.2204,
+      "step": 1417
+    },
+    {
+      "epoch": 3.2263936291240047,
+      "grad_norm": 0.3435449909933618,
+      "learning_rate": 1.3427674631110374e-05,
+      "loss": 0.2549,
+      "step": 1418
+    },
+    {
+      "epoch": 3.2286689419795223,
+      "grad_norm": 0.368268378202336,
+      "learning_rate": 1.3397636177180125e-05,
+      "loss": 0.2903,
+      "step": 1419
+    },
+    {
+      "epoch": 3.2309442548350398,
+      "grad_norm": 0.292427137855411,
+      "learning_rate": 1.3367614428958646e-05,
+      "loss": 0.1921,
+      "step": 1420
+    },
+    {
+      "epoch": 3.2332195676905573,
+      "grad_norm": 0.38271334568666937,
+      "learning_rate": 1.3337609462408834e-05,
+      "loss": 0.2656,
+      "step": 1421
+    },
+    {
+      "epoch": 3.2354948805460753,
+      "grad_norm": 0.31727091050560885,
+      "learning_rate": 1.3307621353451096e-05,
+      "loss": 0.231,
+      "step": 1422
+    },
+    {
+      "epoch": 3.2377701934015928,
+      "grad_norm": 0.36004182056870804,
+      "learning_rate": 1.3277650177963208e-05,
+      "loss": 0.2786,
+      "step": 1423
+    },
+    {
+      "epoch": 3.2400455062571103,
+      "grad_norm": 0.31907622977065264,
+      "learning_rate": 1.3247696011780091e-05,
+      "loss": 0.2327,
+      "step": 1424
+    },
+    {
+      "epoch": 3.242320819112628,
+      "grad_norm": 0.33736643952931383,
+      "learning_rate": 1.3217758930693608e-05,
+      "loss": 0.2309,
+      "step": 1425
+    },
+    {
+      "epoch": 3.244596131968146,
+      "grad_norm": 0.31832255649693225,
+      "learning_rate": 1.3187839010452437e-05,
+      "loss": 0.2307,
+      "step": 1426
+    },
+    {
+      "epoch": 3.2468714448236633,
+      "grad_norm": 0.3593354547706998,
+      "learning_rate": 1.3157936326761789e-05,
+      "loss": 0.2506,
+      "step": 1427
+    },
+    {
+      "epoch": 3.249146757679181,
+      "grad_norm": 0.33616548685144265,
+      "learning_rate": 1.312805095528328e-05,
+      "loss": 0.2205,
+      "step": 1428
+    },
+    {
+      "epoch": 3.2514220705346983,
+      "grad_norm": 0.3654579069097713,
+      "learning_rate": 1.3098182971634738e-05,
+      "loss": 0.2531,
+      "step": 1429
+    },
+    {
+      "epoch": 3.2536973833902163,
+      "grad_norm": 0.345959063129829,
+      "learning_rate": 1.3068332451389969e-05,
+      "loss": 0.2346,
+      "step": 1430
+    },
+    {
+      "epoch": 3.255972696245734,
+      "grad_norm": 0.3661676468618493,
+      "learning_rate": 1.3038499470078611e-05,
+      "loss": 0.2536,
+      "step": 1431
+    },
+    {
+      "epoch": 3.2582480091012513,
+      "grad_norm": 0.3420232850373675,
+      "learning_rate": 1.3008684103185915e-05,
+      "loss": 0.25,
+      "step": 1432
+    },
+    {
+      "epoch": 3.260523321956769,
+      "grad_norm": 0.3506690952818245,
+      "learning_rate": 1.2978886426152549e-05,
+      "loss": 0.2198,
+      "step": 1433
+    },
+    {
+      "epoch": 3.262798634812287,
+      "grad_norm": 0.35765375060393567,
+      "learning_rate": 1.2949106514374458e-05,
+      "loss": 0.2154,
+      "step": 1434
+    },
+    {
+      "epoch": 3.2650739476678043,
+      "grad_norm": 0.3514554551432673,
+      "learning_rate": 1.2919344443202602e-05,
+      "loss": 0.2188,
+      "step": 1435
+    },
+    {
+      "epoch": 3.267349260523322,
+      "grad_norm": 0.33219201290281225,
+      "learning_rate": 1.2889600287942809e-05,
+      "loss": 0.2124,
+      "step": 1436
+    },
+    {
+      "epoch": 3.26962457337884,
+      "grad_norm": 0.3547778577433831,
+      "learning_rate": 1.285987412385559e-05,
+      "loss": 0.2386,
+      "step": 1437
+    },
+    {
+      "epoch": 3.2718998862343573,
+      "grad_norm": 0.374012599713748,
+      "learning_rate": 1.2830166026155908e-05,
+      "loss": 0.2844,
+      "step": 1438
+    },
+    {
+      "epoch": 3.274175199089875,
+      "grad_norm": 0.32390850806717336,
+      "learning_rate": 1.2800476070013038e-05,
+      "loss": 0.2303,
+      "step": 1439
+    },
+    {
+      "epoch": 3.2764505119453924,
+      "grad_norm": 0.31711858818464106,
+      "learning_rate": 1.277080433055034e-05,
+      "loss": 0.2287,
+      "step": 1440
+    },
+    {
+      "epoch": 3.27872582480091,
+      "grad_norm": 0.30594860363235177,
+      "learning_rate": 1.2741150882845066e-05,
+      "loss": 0.2016,
+      "step": 1441
+    },
+    {
+      "epoch": 3.281001137656428,
+      "grad_norm": 0.35211838532005957,
+      "learning_rate": 1.2711515801928229e-05,
+      "loss": 0.2915,
+      "step": 1442
+    },
+    {
+      "epoch": 3.2832764505119454,
+      "grad_norm": 0.3070214436924776,
+      "learning_rate": 1.2681899162784318e-05,
+      "loss": 0.2281,
+      "step": 1443
+    },
+    {
+      "epoch": 3.285551763367463,
+      "grad_norm": 0.34635081909766463,
+      "learning_rate": 1.2652301040351191e-05,
+      "loss": 0.2296,
+      "step": 1444
+    },
+    {
+      "epoch": 3.287827076222981,
+      "grad_norm": 0.3643612010569894,
+      "learning_rate": 1.2622721509519846e-05,
+      "loss": 0.2802,
+      "step": 1445
+    },
+    {
+      "epoch": 3.2901023890784984,
+      "grad_norm": 0.3387586387244522,
+      "learning_rate": 1.2593160645134237e-05,
+      "loss": 0.2057,
+      "step": 1446
+    },
+    {
+      "epoch": 3.292377701934016,
+      "grad_norm": 0.35441126738561124,
+      "learning_rate": 1.2563618521991093e-05,
+      "loss": 0.2627,
+      "step": 1447
+    },
+    {
+      "epoch": 3.2946530147895334,
+      "grad_norm": 0.3270523514329843,
+      "learning_rate": 1.2534095214839721e-05,
+      "loss": 0.2328,
+      "step": 1448
+    },
+    {
+      "epoch": 3.296928327645051,
+      "grad_norm": 0.3485054867825353,
+      "learning_rate": 1.2504590798381799e-05,
+      "loss": 0.2654,
+      "step": 1449
+    },
+    {
+      "epoch": 3.299203640500569,
+      "grad_norm": 0.33361561605989926,
+      "learning_rate": 1.247510534727125e-05,
+      "loss": 0.2233,
+      "step": 1450
+    },
+    {
+      "epoch": 3.3014789533560864,
+      "grad_norm": 0.3332124162613073,
+      "learning_rate": 1.2445638936113968e-05,
+      "loss": 0.254,
+      "step": 1451
+    },
+    {
+      "epoch": 3.303754266211604,
+      "grad_norm": 0.33396306209540555,
+      "learning_rate": 1.2416191639467683e-05,
+      "loss": 0.2813,
+      "step": 1452
+    },
+    {
+      "epoch": 3.306029579067122,
+      "grad_norm": 0.3459951135041794,
+      "learning_rate": 1.2386763531841776e-05,
+      "loss": 0.2358,
+      "step": 1453
+    },
+    {
+      "epoch": 3.3083048919226394,
+      "grad_norm": 0.33833889844367776,
+      "learning_rate": 1.2357354687697055e-05,
+      "loss": 0.2552,
+      "step": 1454
+    },
+    {
+      "epoch": 3.310580204778157,
+      "grad_norm": 0.3826153444150679,
+      "learning_rate": 1.2327965181445593e-05,
+      "loss": 0.2411,
+      "step": 1455
+    },
+    {
+      "epoch": 3.3128555176336745,
+      "grad_norm": 0.3267318396869323,
+      "learning_rate": 1.2298595087450549e-05,
+      "loss": 0.2396,
+      "step": 1456
+    },
+    {
+      "epoch": 3.3151308304891924,
+      "grad_norm": 0.3728421172188106,
+      "learning_rate": 1.2269244480025919e-05,
+      "loss": 0.2389,
+      "step": 1457
+    },
+    {
+      "epoch": 3.31740614334471,
+      "grad_norm": 0.34763878883093824,
+      "learning_rate": 1.223991343343646e-05,
+      "loss": 0.2276,
+      "step": 1458
+    },
+    {
+      "epoch": 3.3196814562002275,
+      "grad_norm": 0.338468563084109,
+      "learning_rate": 1.2210602021897369e-05,
+      "loss": 0.233,
+      "step": 1459
+    },
+    {
+      "epoch": 3.321956769055745,
+      "grad_norm": 0.3323678746338327,
+      "learning_rate": 1.21813103195742e-05,
+      "loss": 0.2561,
+      "step": 1460
+    },
+    {
+      "epoch": 3.324232081911263,
+      "grad_norm": 0.3318887505578868,
+      "learning_rate": 1.2152038400582628e-05,
+      "loss": 0.254,
+      "step": 1461
+    },
+    {
+      "epoch": 3.3265073947667805,
+      "grad_norm": 0.3284576201359955,
+      "learning_rate": 1.2122786338988268e-05,
+      "loss": 0.2174,
+      "step": 1462
+    },
+    {
+      "epoch": 3.328782707622298,
+      "grad_norm": 0.3213797717013886,
+      "learning_rate": 1.20935542088065e-05,
+      "loss": 0.253,
+      "step": 1463
+    },
+    {
+      "epoch": 3.3310580204778155,
+      "grad_norm": 0.3461505829660388,
+      "learning_rate": 1.2064342084002264e-05,
+      "loss": 0.2435,
+      "step": 1464
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.352184817284991,
+      "learning_rate": 1.203515003848987e-05,
+      "loss": 0.2267,
+      "step": 1465
+    },
+    {
+      "epoch": 3.335608646188851,
+      "grad_norm": 0.32028901572619467,
+      "learning_rate": 1.2005978146132855e-05,
+      "loss": 0.2434,
+      "step": 1466
+    },
+    {
+      "epoch": 3.3378839590443685,
+      "grad_norm": 0.2982423511962225,
+      "learning_rate": 1.1976826480743729e-05,
+      "loss": 0.2111,
+      "step": 1467
+    },
+    {
+      "epoch": 3.3401592718998865,
+      "grad_norm": 0.3770540532242069,
+      "learning_rate": 1.1947695116083835e-05,
+      "loss": 0.278,
+      "step": 1468
+    },
+    {
+      "epoch": 3.342434584755404,
+      "grad_norm": 0.3281814863566095,
+      "learning_rate": 1.1918584125863163e-05,
+      "loss": 0.2202,
+      "step": 1469
+    },
+    {
+      "epoch": 3.3447098976109215,
+      "grad_norm": 0.3388970029576235,
+      "learning_rate": 1.1889493583740124e-05,
+      "loss": 0.2699,
+      "step": 1470
+    },
+    {
+      "epoch": 3.346985210466439,
+      "grad_norm": 0.33285749805546927,
+      "learning_rate": 1.1860423563321417e-05,
+      "loss": 0.2385,
+      "step": 1471
+    },
+    {
+      "epoch": 3.3492605233219566,
+      "grad_norm": 0.37062291383489476,
+      "learning_rate": 1.18313741381618e-05,
+      "loss": 0.2523,
+      "step": 1472
+    },
+    {
+      "epoch": 3.3515358361774745,
+      "grad_norm": 0.304526701299871,
+      "learning_rate": 1.1802345381763918e-05,
+      "loss": 0.2197,
+      "step": 1473
+    },
+    {
+      "epoch": 3.353811149032992,
+      "grad_norm": 0.35064431333493157,
+      "learning_rate": 1.1773337367578134e-05,
+      "loss": 0.2346,
+      "step": 1474
+    },
+    {
+      "epoch": 3.3560864618885096,
+      "grad_norm": 0.38588974905676426,
+      "learning_rate": 1.1744350169002308e-05,
+      "loss": 0.2433,
+      "step": 1475
+    },
+    {
+      "epoch": 3.3583617747440275,
+      "grad_norm": 0.34525997525679775,
+      "learning_rate": 1.1715383859381639e-05,
+      "loss": 0.224,
+      "step": 1476
+    },
+    {
+      "epoch": 3.360637087599545,
+      "grad_norm": 0.5393672418195411,
+      "learning_rate": 1.168643851200848e-05,
+      "loss": 0.278,
+      "step": 1477
+    },
+    {
+      "epoch": 3.3629124004550626,
+      "grad_norm": 0.3344984797575095,
+      "learning_rate": 1.1657514200122133e-05,
+      "loss": 0.2277,
+      "step": 1478
+    },
+    {
+      "epoch": 3.36518771331058,
+      "grad_norm": 0.3807232368526946,
+      "learning_rate": 1.1628610996908685e-05,
+      "loss": 0.247,
+      "step": 1479
+    },
+    {
+      "epoch": 3.3674630261660976,
+      "grad_norm": 0.34520588066288926,
+      "learning_rate": 1.159972897550079e-05,
+      "loss": 0.2553,
+      "step": 1480
+    },
+    {
+      "epoch": 3.3697383390216156,
+      "grad_norm": 0.31387145207674094,
+      "learning_rate": 1.157086820897754e-05,
+      "loss": 0.2114,
+      "step": 1481
+    },
+    {
+      "epoch": 3.372013651877133,
+      "grad_norm": 0.3459709809459857,
+      "learning_rate": 1.1542028770364225e-05,
+      "loss": 0.2086,
+      "step": 1482
+    },
+    {
+      "epoch": 3.3742889647326506,
+      "grad_norm": 0.3248421852769576,
+      "learning_rate": 1.1513210732632165e-05,
+      "loss": 0.1794,
+      "step": 1483
+    },
+    {
+      "epoch": 3.3765642775881686,
+      "grad_norm": 0.38897925073848516,
+      "learning_rate": 1.1484414168698547e-05,
+      "loss": 0.2945,
+      "step": 1484
+    },
+    {
+      "epoch": 3.378839590443686,
+      "grad_norm": 0.3795025010628272,
+      "learning_rate": 1.145563915142622e-05,
+      "loss": 0.2186,
+      "step": 1485
+    },
+    {
+      "epoch": 3.3811149032992036,
+      "grad_norm": 0.3985729069914676,
+      "learning_rate": 1.14268857536235e-05,
+      "loss": 0.2607,
+      "step": 1486
+    },
+    {
+      "epoch": 3.383390216154721,
+      "grad_norm": 0.3437196809151993,
+      "learning_rate": 1.1398154048044022e-05,
+      "loss": 0.2123,
+      "step": 1487
+    },
+    {
+      "epoch": 3.385665529010239,
+      "grad_norm": 0.3289103296082285,
+      "learning_rate": 1.1369444107386505e-05,
+      "loss": 0.2253,
+      "step": 1488
+    },
+    {
+      "epoch": 3.3879408418657566,
+      "grad_norm": 0.38051780478879027,
+      "learning_rate": 1.1340756004294626e-05,
+      "loss": 0.2673,
+      "step": 1489
+    },
+    {
+      "epoch": 3.390216154721274,
+      "grad_norm": 0.3060080157181269,
+      "learning_rate": 1.1312089811356803e-05,
+      "loss": 0.2191,
+      "step": 1490
+    },
+    {
+      "epoch": 3.3924914675767917,
+      "grad_norm": 0.31770551000551334,
+      "learning_rate": 1.1283445601105988e-05,
+      "loss": 0.2286,
+      "step": 1491
+    },
+    {
+      "epoch": 3.3947667804323096,
+      "grad_norm": 0.36426252820361144,
+      "learning_rate": 1.1254823446019544e-05,
+      "loss": 0.2482,
+      "step": 1492
+    },
+    {
+      "epoch": 3.397042093287827,
+      "grad_norm": 0.33824561418896854,
+      "learning_rate": 1.1226223418519024e-05,
+      "loss": 0.2309,
+      "step": 1493
+    },
+    {
+      "epoch": 3.3993174061433447,
+      "grad_norm": 0.34268422073166216,
+      "learning_rate": 1.1197645590969965e-05,
+      "loss": 0.2557,
+      "step": 1494
+    },
+    {
+      "epoch": 3.401592718998862,
+      "grad_norm": 0.33591817107983135,
+      "learning_rate": 1.1169090035681772e-05,
+      "loss": 0.2784,
+      "step": 1495
+    },
+    {
+      "epoch": 3.40386803185438,
+      "grad_norm": 0.3114963695621851,
+      "learning_rate": 1.114055682490745e-05,
+      "loss": 0.22,
+      "step": 1496
+    },
+    {
+      "epoch": 3.4061433447098977,
+      "grad_norm": 0.35637555406825994,
+      "learning_rate": 1.111204603084351e-05,
+      "loss": 0.2992,
+      "step": 1497
+    },
+    {
+      "epoch": 3.408418657565415,
+      "grad_norm": 0.30938826214986764,
+      "learning_rate": 1.1083557725629723e-05,
+      "loss": 0.2179,
+      "step": 1498
+    },
+    {
+      "epoch": 3.4106939704209327,
+      "grad_norm": 0.33999235461053073,
+      "learning_rate": 1.105509198134895e-05,
+      "loss": 0.2293,
+      "step": 1499
+    },
+    {
+      "epoch": 3.4129692832764507,
+      "grad_norm": 0.3346137284958368,
+      "learning_rate": 1.102664887002698e-05,
+      "loss": 0.2321,
+      "step": 1500
+    },
+    {
+      "epoch": 3.415244596131968,
+      "grad_norm": 0.35961820157939406,
+      "learning_rate": 1.099822846363234e-05,
+      "loss": 0.2671,
+      "step": 1501
+    },
+    {
+      "epoch": 3.4175199089874857,
+      "grad_norm": 0.3032687812693232,
+      "learning_rate": 1.0969830834076079e-05,
+      "loss": 0.2435,
+      "step": 1502
+    },
+    {
+      "epoch": 3.419795221843003,
+      "grad_norm": 0.3299085763321149,
+      "learning_rate": 1.0941456053211643e-05,
+      "loss": 0.2227,
+      "step": 1503
+    },
+    {
+      "epoch": 3.422070534698521,
+      "grad_norm": 0.35062052767564167,
+      "learning_rate": 1.0913104192834665e-05,
+      "loss": 0.2577,
+      "step": 1504
+    },
+    {
+      "epoch": 3.4243458475540387,
+      "grad_norm": 0.3171905569045706,
+      "learning_rate": 1.0884775324682755e-05,
+      "loss": 0.2714,
+      "step": 1505
+    },
+    {
+      "epoch": 3.426621160409556,
+      "grad_norm": 0.3050583814158292,
+      "learning_rate": 1.085646952043538e-05,
+      "loss": 0.2076,
+      "step": 1506
+    },
+    {
+      "epoch": 3.428896473265074,
+      "grad_norm": 0.31824113240968743,
+      "learning_rate": 1.0828186851713625e-05,
+      "loss": 0.2311,
+      "step": 1507
+    },
+    {
+      "epoch": 3.4311717861205917,
+      "grad_norm": 0.3258202780170239,
+      "learning_rate": 1.0799927390080045e-05,
+      "loss": 0.2339,
+      "step": 1508
+    },
+    {
+      "epoch": 3.4334470989761092,
+      "grad_norm": 0.32618685833040023,
+      "learning_rate": 1.0771691207038489e-05,
+      "loss": 0.2507,
+      "step": 1509
+    },
+    {
+      "epoch": 3.4357224118316267,
+      "grad_norm": 0.3469821193351217,
+      "learning_rate": 1.074347837403387e-05,
+      "loss": 0.2596,
+      "step": 1510
+    },
+    {
+      "epoch": 3.4379977246871443,
+      "grad_norm": 0.32569711419010833,
+      "learning_rate": 1.0715288962452055e-05,
+      "loss": 0.2089,
+      "step": 1511
+    },
+    {
+      "epoch": 3.4402730375426622,
+      "grad_norm": 0.3561430085177732,
+      "learning_rate": 1.0687123043619643e-05,
+      "loss": 0.3069,
+      "step": 1512
+    },
+    {
+      "epoch": 3.4425483503981797,
+      "grad_norm": 0.33589253239111483,
+      "learning_rate": 1.065898068880376e-05,
+      "loss": 0.2299,
+      "step": 1513
+    },
+    {
+      "epoch": 3.4448236632536973,
+      "grad_norm": 0.3289052416730758,
+      "learning_rate": 1.0630861969211948e-05,
+      "loss": 0.2578,
+      "step": 1514
+    },
+    {
+      "epoch": 3.4470989761092152,
+      "grad_norm": 0.29329682255766953,
+      "learning_rate": 1.0602766955991913e-05,
+      "loss": 0.2069,
+      "step": 1515
+    },
+    {
+      "epoch": 3.4493742889647327,
+      "grad_norm": 0.34344762978311627,
+      "learning_rate": 1.0574695720231402e-05,
+      "loss": 0.2597,
+      "step": 1516
+    },
+    {
+      "epoch": 3.4516496018202503,
+      "grad_norm": 0.33783465976482363,
+      "learning_rate": 1.0546648332957999e-05,
+      "loss": 0.2619,
+      "step": 1517
+    },
+    {
+      "epoch": 3.453924914675768,
+      "grad_norm": 0.3038166722942614,
+      "learning_rate": 1.0518624865138915e-05,
+      "loss": 0.2057,
+      "step": 1518
+    },
+    {
+      "epoch": 3.4562002275312853,
+      "grad_norm": 0.34704986882471467,
+      "learning_rate": 1.0490625387680872e-05,
+      "loss": 0.2656,
+      "step": 1519
+    },
+    {
+      "epoch": 3.4584755403868033,
+      "grad_norm": 0.28763258753357995,
+      "learning_rate": 1.0462649971429884e-05,
+      "loss": 0.2081,
+      "step": 1520
+    },
+    {
+      "epoch": 3.460750853242321,
+      "grad_norm": 0.3539165701148113,
+      "learning_rate": 1.0434698687171059e-05,
+      "loss": 0.2499,
+      "step": 1521
+    },
+    {
+      "epoch": 3.4630261660978383,
+      "grad_norm": 0.2983963903646923,
+      "learning_rate": 1.040677160562848e-05,
+      "loss": 0.2042,
+      "step": 1522
+    },
+    {
+      "epoch": 3.4653014789533563,
+      "grad_norm": 0.33886381721871905,
+      "learning_rate": 1.0378868797464961e-05,
+      "loss": 0.2539,
+      "step": 1523
+    },
+    {
+      "epoch": 3.467576791808874,
+      "grad_norm": 0.33131937800817474,
+      "learning_rate": 1.0350990333281915e-05,
+      "loss": 0.183,
+      "step": 1524
+    },
+    {
+      "epoch": 3.4698521046643913,
+      "grad_norm": 0.3334157312213503,
+      "learning_rate": 1.0323136283619167e-05,
+      "loss": 0.2291,
+      "step": 1525
+    },
+    {
+      "epoch": 3.472127417519909,
+      "grad_norm": 0.36557655183280097,
+      "learning_rate": 1.029530671895473e-05,
+      "loss": 0.2592,
+      "step": 1526
+    },
+    {
+      "epoch": 3.474402730375427,
+      "grad_norm": 0.3098710582941038,
+      "learning_rate": 1.02675017097047e-05,
+      "loss": 0.2216,
+      "step": 1527
+    },
+    {
+      "epoch": 3.4766780432309443,
+      "grad_norm": 0.33798309183034975,
+      "learning_rate": 1.023972132622304e-05,
+      "loss": 0.2486,
+      "step": 1528
+    },
+    {
+      "epoch": 3.478953356086462,
+      "grad_norm": 0.29037782431225956,
+      "learning_rate": 1.0211965638801374e-05,
+      "loss": 0.1917,
+      "step": 1529
+    },
+    {
+      "epoch": 3.4812286689419794,
+      "grad_norm": 0.33565460827318755,
+      "learning_rate": 1.0184234717668867e-05,
+      "loss": 0.2528,
+      "step": 1530
+    },
+    {
+      "epoch": 3.4835039817974973,
+      "grad_norm": 0.33527755448329066,
+      "learning_rate": 1.0156528632992004e-05,
+      "loss": 0.2312,
+      "step": 1531
+    },
+    {
+      "epoch": 3.485779294653015,
+      "grad_norm": 0.34743533992096876,
+      "learning_rate": 1.012884745487443e-05,
+      "loss": 0.2828,
+      "step": 1532
+    },
+    {
+      "epoch": 3.4880546075085324,
+      "grad_norm": 0.3029170690482156,
+      "learning_rate": 1.0101191253356783e-05,
+      "loss": 0.2052,
+      "step": 1533
+    },
+    {
+      "epoch": 3.49032992036405,
+      "grad_norm": 0.31040382592429244,
+      "learning_rate": 1.0073560098416479e-05,
+      "loss": 0.2323,
+      "step": 1534
+    },
+    {
+      "epoch": 3.492605233219568,
+      "grad_norm": 0.3291743879414811,
+      "learning_rate": 1.0045954059967577e-05,
+      "loss": 0.2515,
+      "step": 1535
+    },
+    {
+      "epoch": 3.4948805460750854,
+      "grad_norm": 0.3030205392683074,
+      "learning_rate": 1.0018373207860593e-05,
+      "loss": 0.2239,
+      "step": 1536
+    },
+    {
+      "epoch": 3.497155858930603,
+      "grad_norm": 0.34698271644365286,
+      "learning_rate": 9.990817611882284e-06,
+      "loss": 0.2902,
+      "step": 1537
+    },
+    {
+      "epoch": 3.4994311717861204,
+      "grad_norm": 0.3131439916281456,
+      "learning_rate": 9.963287341755537e-06,
+      "loss": 0.1915,
+      "step": 1538
+    },
+    {
+      "epoch": 3.5017064846416384,
+      "grad_norm": 0.36945832587752353,
+      "learning_rate": 9.935782467139126e-06,
+      "loss": 0.306,
+      "step": 1539
+    },
+    {
+      "epoch": 3.503981797497156,
+      "grad_norm": 0.3339534854776466,
+      "learning_rate": 9.908303057627591e-06,
+      "loss": 0.2465,
+      "step": 1540
+    },
+    {
+      "epoch": 3.5062571103526734,
+      "grad_norm": 0.2856094747480189,
+      "learning_rate": 9.880849182751042e-06,
+      "loss": 0.2048,
+      "step": 1541
+    },
+    {
+      "epoch": 3.508532423208191,
+      "grad_norm": 0.3074938623502293,
+      "learning_rate": 9.85342091197495e-06,
+      "loss": 0.2517,
+      "step": 1542
+    },
+    {
+      "epoch": 3.510807736063709,
+      "grad_norm": 0.3341468790362437,
+      "learning_rate": 9.826018314700033e-06,
+      "loss": 0.2492,
+      "step": 1543
+    },
+    {
+      "epoch": 3.5130830489192264,
+      "grad_norm": 0.33344807328931014,
+      "learning_rate": 9.798641460262037e-06,
+      "loss": 0.2431,
+      "step": 1544
+    },
+    {
+      "epoch": 3.515358361774744,
+      "grad_norm": 0.31979928441877864,
+      "learning_rate": 9.771290417931559e-06,
+      "loss": 0.2353,
+      "step": 1545
+    },
+    {
+      "epoch": 3.517633674630262,
+      "grad_norm": 0.319428974933737,
+      "learning_rate": 9.743965256913908e-06,
+      "loss": 0.2378,
+      "step": 1546
+    },
+    {
+      "epoch": 3.5199089874857794,
+      "grad_norm": 0.3283945511768955,
+      "learning_rate": 9.716666046348872e-06,
+      "loss": 0.2313,
+      "step": 1547
+    },
+    {
+      "epoch": 3.522184300341297,
+      "grad_norm": 0.34531998696589944,
+      "learning_rate": 9.689392855310633e-06,
+      "loss": 0.2617,
+      "step": 1548
+    },
+    {
+      "epoch": 3.5244596131968144,
+      "grad_norm": 0.31043476202292597,
+      "learning_rate": 9.662145752807488e-06,
+      "loss": 0.2135,
+      "step": 1549
+    },
+    {
+      "epoch": 3.526734926052332,
+      "grad_norm": 0.33406451563957484,
+      "learning_rate": 9.634924807781729e-06,
+      "loss": 0.226,
+      "step": 1550
+    },
+    {
+      "epoch": 3.52901023890785,
+      "grad_norm": 0.3365790815539411,
+      "learning_rate": 9.607730089109479e-06,
+      "loss": 0.2166,
+      "step": 1551
+    },
+    {
+      "epoch": 3.5312855517633674,
+      "grad_norm": 0.34723707276145616,
+      "learning_rate": 9.580561665600507e-06,
+      "loss": 0.2038,
+      "step": 1552
+    },
+    {
+      "epoch": 3.533560864618885,
+      "grad_norm": 0.3537401937413241,
+      "learning_rate": 9.55341960599802e-06,
+      "loss": 0.2231,
+      "step": 1553
+    },
+    {
+      "epoch": 3.535836177474403,
+      "grad_norm": 0.33721120314003905,
+      "learning_rate": 9.526303978978546e-06,
+      "loss": 0.2631,
+      "step": 1554
+    },
+    {
+      "epoch": 3.5381114903299204,
+      "grad_norm": 0.3302043329828681,
+      "learning_rate": 9.499214853151699e-06,
+      "loss": 0.2437,
+      "step": 1555
+    },
+    {
+      "epoch": 3.540386803185438,
+      "grad_norm": 0.3826580704046,
+      "learning_rate": 9.472152297060088e-06,
+      "loss": 0.2351,
+      "step": 1556
+    },
+    {
+      "epoch": 3.5426621160409555,
+      "grad_norm": 0.3489687845389196,
+      "learning_rate": 9.445116379179052e-06,
+      "loss": 0.2463,
+      "step": 1557
+    },
+    {
+      "epoch": 3.544937428896473,
+      "grad_norm": 0.3270717426834041,
+      "learning_rate": 9.418107167916526e-06,
+      "loss": 0.2356,
+      "step": 1558
+    },
+    {
+      "epoch": 3.547212741751991,
+      "grad_norm": 0.35109088802465127,
+      "learning_rate": 9.3911247316129e-06,
+      "loss": 0.2142,
+      "step": 1559
+    },
+    {
+      "epoch": 3.5494880546075085,
+      "grad_norm": 0.4163412510419317,
+      "learning_rate": 9.364169138540805e-06,
+      "loss": 0.2296,
+      "step": 1560
+    },
+    {
+      "epoch": 3.551763367463026,
+      "grad_norm": 0.3808250685074194,
+      "learning_rate": 9.33724045690493e-06,
+      "loss": 0.2646,
+      "step": 1561
+    },
+    {
+      "epoch": 3.554038680318544,
+      "grad_norm": 0.3357674933739998,
+      "learning_rate": 9.310338754841907e-06,
+      "loss": 0.2462,
+      "step": 1562
+    },
+    {
+      "epoch": 3.5563139931740615,
+      "grad_norm": 0.31994156406414276,
+      "learning_rate": 9.283464100420064e-06,
+      "loss": 0.2236,
+      "step": 1563
+    },
+    {
+      "epoch": 3.558589306029579,
+      "grad_norm": 0.35903123661267816,
+      "learning_rate": 9.256616561639337e-06,
+      "loss": 0.2279,
+      "step": 1564
+    },
+    {
+      "epoch": 3.5608646188850965,
+      "grad_norm": 0.35744852928869253,
+      "learning_rate": 9.229796206431015e-06,
+      "loss": 0.233,
+      "step": 1565
+    },
+    {
+      "epoch": 3.5631399317406145,
+      "grad_norm": 0.34937225764885005,
+      "learning_rate": 9.203003102657608e-06,
+      "loss": 0.2473,
+      "step": 1566
+    },
+    {
+      "epoch": 3.565415244596132,
+      "grad_norm": 0.32826327293032775,
+      "learning_rate": 9.176237318112691e-06,
+      "loss": 0.2095,
+      "step": 1567
+    },
+    {
+      "epoch": 3.5676905574516495,
+      "grad_norm": 0.34233653160590005,
+      "learning_rate": 9.149498920520712e-06,
+      "loss": 0.2295,
+      "step": 1568
+    },
+    {
+      "epoch": 3.5699658703071675,
+      "grad_norm": 0.40351720299178445,
+      "learning_rate": 9.122787977536802e-06,
+      "loss": 0.2963,
+      "step": 1569
+    },
+    {
+      "epoch": 3.572241183162685,
+      "grad_norm": 0.3003693453078882,
+      "learning_rate": 9.096104556746654e-06,
+      "loss": 0.2107,
+      "step": 1570
+    },
+    {
+      "epoch": 3.5745164960182025,
+      "grad_norm": 0.39003272892477975,
+      "learning_rate": 9.069448725666285e-06,
+      "loss": 0.2816,
+      "step": 1571
+    },
+    {
+      "epoch": 3.57679180887372,
+      "grad_norm": 0.3273815427887145,
+      "learning_rate": 9.042820551741955e-06,
+      "loss": 0.2348,
+      "step": 1572
+    },
+    {
+      "epoch": 3.5790671217292376,
+      "grad_norm": 0.3629427404955351,
+      "learning_rate": 9.0162201023499e-06,
+      "loss": 0.2429,
+      "step": 1573
+    },
+    {
+      "epoch": 3.5813424345847555,
+      "grad_norm": 0.33919168356346674,
+      "learning_rate": 8.98964744479621e-06,
+      "loss": 0.2513,
+      "step": 1574
+    },
+    {
+      "epoch": 3.583617747440273,
+      "grad_norm": 0.33498714294273985,
+      "learning_rate": 8.963102646316677e-06,
+      "loss": 0.2161,
+      "step": 1575
+    },
+    {
+      "epoch": 3.5858930602957906,
+      "grad_norm": 0.40300529209904284,
+      "learning_rate": 8.936585774076594e-06,
+      "loss": 0.2622,
+      "step": 1576
+    },
+    {
+      "epoch": 3.5881683731513085,
+      "grad_norm": 0.324525449514623,
+      "learning_rate": 8.910096895170572e-06,
+      "loss": 0.2033,
+      "step": 1577
+    },
+    {
+      "epoch": 3.590443686006826,
+      "grad_norm": 0.3418244524758691,
+      "learning_rate": 8.88363607662243e-06,
+      "loss": 0.2596,
+      "step": 1578
+    },
+    {
+      "epoch": 3.5927189988623436,
+      "grad_norm": 0.3813231405164915,
+      "learning_rate": 8.857203385384938e-06,
+      "loss": 0.2493,
+      "step": 1579
+    },
+    {
+      "epoch": 3.594994311717861,
+      "grad_norm": 0.3544543624874025,
+      "learning_rate": 8.830798888339756e-06,
+      "loss": 0.2406,
+      "step": 1580
+    },
+    {
+      "epoch": 3.5972696245733786,
+      "grad_norm": 0.3679759598835115,
+      "learning_rate": 8.804422652297157e-06,
+      "loss": 0.2328,
+      "step": 1581
+    },
+    {
+      "epoch": 3.5995449374288966,
+      "grad_norm": 0.3523367839674939,
+      "learning_rate": 8.778074743995916e-06,
+      "loss": 0.2556,
+      "step": 1582
+    },
+    {
+      "epoch": 3.601820250284414,
+      "grad_norm": 0.3183567073970889,
+      "learning_rate": 8.75175523010315e-06,
+      "loss": 0.1958,
+      "step": 1583
+    },
+    {
+      "epoch": 3.6040955631399316,
+      "grad_norm": 0.37485870319862513,
+      "learning_rate": 8.725464177214119e-06,
+      "loss": 0.2611,
+      "step": 1584
+    },
+    {
+      "epoch": 3.6063708759954496,
+      "grad_norm": 0.35712252822034435,
+      "learning_rate": 8.699201651852056e-06,
+      "loss": 0.2352,
+      "step": 1585
+    },
+    {
+      "epoch": 3.608646188850967,
+      "grad_norm": 0.37283944559977766,
+      "learning_rate": 8.672967720468044e-06,
+      "loss": 0.2844,
+      "step": 1586
+    },
+    {
+      "epoch": 3.6109215017064846,
+      "grad_norm": 0.311035508065256,
+      "learning_rate": 8.646762449440768e-06,
+      "loss": 0.2201,
+      "step": 1587
+    },
+    {
+      "epoch": 3.613196814562002,
+      "grad_norm": 0.2988780921453698,
+      "learning_rate": 8.620585905076462e-06,
+      "loss": 0.2186,
+      "step": 1588
+    },
+    {
+      "epoch": 3.6154721274175197,
+      "grad_norm": 0.4240158000247216,
+      "learning_rate": 8.594438153608611e-06,
+      "loss": 0.3004,
+      "step": 1589
+    },
+    {
+      "epoch": 3.6177474402730376,
+      "grad_norm": 0.32568214186763933,
+      "learning_rate": 8.56831926119787e-06,
+      "loss": 0.2174,
+      "step": 1590
+    },
+    {
+      "epoch": 3.620022753128555,
+      "grad_norm": 0.30055241402531796,
+      "learning_rate": 8.542229293931878e-06,
+      "loss": 0.2077,
+      "step": 1591
+    },
+    {
+      "epoch": 3.6222980659840727,
+      "grad_norm": 0.3528261607678874,
+      "learning_rate": 8.51616831782509e-06,
+      "loss": 0.2505,
+      "step": 1592
+    },
+    {
+      "epoch": 3.6245733788395906,
+      "grad_norm": 0.3366924619649751,
+      "learning_rate": 8.490136398818584e-06,
+      "loss": 0.2288,
+      "step": 1593
+    },
+    {
+      "epoch": 3.626848691695108,
+      "grad_norm": 0.32639722472827926,
+      "learning_rate": 8.464133602779937e-06,
+      "loss": 0.2159,
+      "step": 1594
+    },
+    {
+      "epoch": 3.6291240045506257,
+      "grad_norm": 0.335990588461995,
+      "learning_rate": 8.43815999550303e-06,
+      "loss": 0.2318,
+      "step": 1595
+    },
+    {
+      "epoch": 3.631399317406143,
+      "grad_norm": 0.3170153764669358,
+      "learning_rate": 8.412215642707892e-06,
+      "loss": 0.2235,
+      "step": 1596
+    },
+    {
+      "epoch": 3.6336746302616607,
+      "grad_norm": 0.3499875495179629,
+      "learning_rate": 8.386300610040523e-06,
+      "loss": 0.2622,
+      "step": 1597
+    },
+    {
+      "epoch": 3.6359499431171787,
+      "grad_norm": 0.31930183238712134,
+      "learning_rate": 8.360414963072731e-06,
+      "loss": 0.2304,
+      "step": 1598
+    },
+    {
+      "epoch": 3.638225255972696,
+      "grad_norm": 0.3062417178184185,
+      "learning_rate": 8.33455876730199e-06,
+      "loss": 0.2244,
+      "step": 1599
+    },
+    {
+      "epoch": 3.640500568828214,
+      "grad_norm": 0.3128935206442823,
+      "learning_rate": 8.308732088151245e-06,
+      "loss": 0.2047,
+      "step": 1600
+    },
+    {
+      "epoch": 3.6427758816837317,
+      "grad_norm": 0.315330561097062,
+      "learning_rate": 8.282934990968742e-06,
+      "loss": 0.2143,
+      "step": 1601
+    },
+    {
+      "epoch": 3.645051194539249,
+      "grad_norm": 0.3649351233173286,
+      "learning_rate": 8.257167541027901e-06,
+      "loss": 0.2447,
+      "step": 1602
+    },
+    {
+      "epoch": 3.6473265073947667,
+      "grad_norm": 0.3590062283162223,
+      "learning_rate": 8.231429803527114e-06,
+      "loss": 0.2331,
+      "step": 1603
+    },
+    {
+      "epoch": 3.6496018202502842,
+      "grad_norm": 0.3070921237937883,
+      "learning_rate": 8.205721843589597e-06,
+      "loss": 0.2066,
+      "step": 1604
+    },
+    {
+      "epoch": 3.651877133105802,
+      "grad_norm": 0.3182414091477948,
+      "learning_rate": 8.180043726263216e-06,
+      "loss": 0.2238,
+      "step": 1605
+    },
+    {
+      "epoch": 3.6541524459613197,
+      "grad_norm": 0.2883644587362991,
+      "learning_rate": 8.15439551652032e-06,
+      "loss": 0.1942,
+      "step": 1606
+    },
+    {
+      "epoch": 3.6564277588168372,
+      "grad_norm": 0.35235761141896277,
+      "learning_rate": 8.1287772792576e-06,
+      "loss": 0.285,
+      "step": 1607
+    },
+    {
+      "epoch": 3.658703071672355,
+      "grad_norm": 0.3066178498743668,
+      "learning_rate": 8.103189079295911e-06,
+      "loss": 0.2116,
+      "step": 1608
+    },
+    {
+      "epoch": 3.6609783845278727,
+      "grad_norm": 0.31320618542805817,
+      "learning_rate": 8.077630981380087e-06,
+      "loss": 0.2585,
+      "step": 1609
+    },
+    {
+      "epoch": 3.6632536973833902,
+      "grad_norm": 0.30635364571870777,
+      "learning_rate": 8.052103050178806e-06,
+      "loss": 0.2032,
+      "step": 1610
+    },
+    {
+      "epoch": 3.6655290102389078,
+      "grad_norm": 0.34555526775527085,
+      "learning_rate": 8.02660535028442e-06,
+      "loss": 0.2762,
+      "step": 1611
+    },
+    {
+      "epoch": 3.6678043230944253,
+      "grad_norm": 0.3478375071781361,
+      "learning_rate": 8.001137946212792e-06,
+      "loss": 0.264,
+      "step": 1612
+    },
+    {
+      "epoch": 3.6700796359499432,
+      "grad_norm": 0.32259979937173106,
+      "learning_rate": 7.975700902403112e-06,
+      "loss": 0.2322,
+      "step": 1613
+    },
+    {
+      "epoch": 3.6723549488054608,
+      "grad_norm": 0.30054960499588285,
+      "learning_rate": 7.950294283217752e-06,
+      "loss": 0.2176,
+      "step": 1614
+    },
+    {
+      "epoch": 3.6746302616609783,
+      "grad_norm": 0.334280726764928,
+      "learning_rate": 7.924918152942117e-06,
+      "loss": 0.2472,
+      "step": 1615
+    },
+    {
+      "epoch": 3.6769055745164962,
+      "grad_norm": 0.3334232858921928,
+      "learning_rate": 7.89957257578447e-06,
+      "loss": 0.2476,
+      "step": 1616
+    },
+    {
+      "epoch": 3.6791808873720138,
+      "grad_norm": 0.3350423889743071,
+      "learning_rate": 7.874257615875737e-06,
+      "loss": 0.249,
+      "step": 1617
+    },
+    {
+      "epoch": 3.6814562002275313,
+      "grad_norm": 0.31257388107369793,
+      "learning_rate": 7.8489733372694e-06,
+      "loss": 0.2477,
+      "step": 1618
+    },
+    {
+      "epoch": 3.683731513083049,
+      "grad_norm": 0.3241796725065618,
+      "learning_rate": 7.823719803941308e-06,
+      "loss": 0.207,
+      "step": 1619
+    },
+    {
+      "epoch": 3.6860068259385663,
+      "grad_norm": 0.3678066881401646,
+      "learning_rate": 7.798497079789513e-06,
+      "loss": 0.239,
+      "step": 1620
+    },
+    {
+      "epoch": 3.6882821387940843,
+      "grad_norm": 0.3072202237563532,
+      "learning_rate": 7.773305228634105e-06,
+      "loss": 0.241,
+      "step": 1621
+    },
+    {
+      "epoch": 3.690557451649602,
+      "grad_norm": 0.3171033622346416,
+      "learning_rate": 7.748144314217055e-06,
+      "loss": 0.2508,
+      "step": 1622
+    },
+    {
+      "epoch": 3.6928327645051193,
+      "grad_norm": 0.35675741351576,
+      "learning_rate": 7.723014400202067e-06,
+      "loss": 0.2364,
+      "step": 1623
+    },
+    {
+      "epoch": 3.6951080773606373,
+      "grad_norm": 0.32339638486968514,
+      "learning_rate": 7.697915550174414e-06,
+      "loss": 0.2606,
+      "step": 1624
+    },
+    {
+      "epoch": 3.697383390216155,
+      "grad_norm": 0.3277199881918983,
+      "learning_rate": 7.672847827640735e-06,
+      "loss": 0.2524,
+      "step": 1625
+    },
+    {
+      "epoch": 3.6996587030716723,
+      "grad_norm": 0.3325613261956722,
+      "learning_rate": 7.647811296028943e-06,
+      "loss": 0.2434,
+      "step": 1626
+    },
+    {
+      "epoch": 3.70193401592719,
+      "grad_norm": 0.3235937045048768,
+      "learning_rate": 7.622806018688016e-06,
+      "loss": 0.2226,
+      "step": 1627
+    },
+    {
+      "epoch": 3.7042093287827074,
+      "grad_norm": 0.31644979052423133,
+      "learning_rate": 7.5978320588878595e-06,
+      "loss": 0.2689,
+      "step": 1628
+    },
+    {
+      "epoch": 3.7064846416382253,
+      "grad_norm": 0.30401544252127216,
+      "learning_rate": 7.572889479819117e-06,
+      "loss": 0.2134,
+      "step": 1629
+    },
+    {
+      "epoch": 3.708759954493743,
+      "grad_norm": 0.3247293577698998,
+      "learning_rate": 7.5479783445930414e-06,
+      "loss": 0.2431,
+      "step": 1630
+    },
+    {
+      "epoch": 3.7110352673492604,
+      "grad_norm": 0.3343087436377371,
+      "learning_rate": 7.523098716241331e-06,
+      "loss": 0.256,
+      "step": 1631
+    },
+    {
+      "epoch": 3.7133105802047783,
+      "grad_norm": 0.29014471473935294,
+      "learning_rate": 7.498250657715962e-06,
+      "loss": 0.2437,
+      "step": 1632
+    },
+    {
+      "epoch": 3.715585893060296,
+      "grad_norm": 0.30810596917320315,
+      "learning_rate": 7.473434231889018e-06,
+      "loss": 0.2602,
+      "step": 1633
+    },
+    {
+      "epoch": 3.7178612059158134,
+      "grad_norm": 0.3193602943821451,
+      "learning_rate": 7.448649501552559e-06,
+      "loss": 0.2353,
+      "step": 1634
+    },
+    {
+      "epoch": 3.720136518771331,
+      "grad_norm": 0.34473063758045885,
+      "learning_rate": 7.4238965294184374e-06,
+      "loss": 0.2464,
+      "step": 1635
+    },
+    {
+      "epoch": 3.722411831626849,
+      "grad_norm": 0.2870067699187315,
+      "learning_rate": 7.39917537811816e-06,
+      "loss": 0.2133,
+      "step": 1636
+    },
+    {
+      "epoch": 3.7246871444823664,
+      "grad_norm": 0.3509161389314907,
+      "learning_rate": 7.374486110202705e-06,
+      "loss": 0.2372,
+      "step": 1637
+    },
+    {
+      "epoch": 3.726962457337884,
+      "grad_norm": 0.3382312400661472,
+      "learning_rate": 7.349828788142377e-06,
+      "loss": 0.2763,
+      "step": 1638
+    },
+    {
+      "epoch": 3.729237770193402,
+      "grad_norm": 0.31679908169558146,
+      "learning_rate": 7.325203474326663e-06,
+      "loss": 0.2134,
+      "step": 1639
+    },
+    {
+      "epoch": 3.7315130830489194,
+      "grad_norm": 0.2817679233954677,
+      "learning_rate": 7.300610231064056e-06,
+      "loss": 0.2025,
+      "step": 1640
+    },
+    {
+      "epoch": 3.733788395904437,
+      "grad_norm": 0.3301349553684187,
+      "learning_rate": 7.276049120581889e-06,
+      "loss": 0.2618,
+      "step": 1641
+    },
+    {
+      "epoch": 3.7360637087599544,
+      "grad_norm": 0.3047477589901083,
+      "learning_rate": 7.251520205026206e-06,
+      "loss": 0.2176,
+      "step": 1642
+    },
+    {
+      "epoch": 3.738339021615472,
+      "grad_norm": 0.3243528934414049,
+      "learning_rate": 7.2270235464615865e-06,
+      "loss": 0.2796,
+      "step": 1643
+    },
+    {
+      "epoch": 3.74061433447099,
+      "grad_norm": 0.31843745378436267,
+      "learning_rate": 7.202559206870994e-06,
+      "loss": 0.2259,
+      "step": 1644
+    },
+    {
+      "epoch": 3.7428896473265074,
+      "grad_norm": 0.3455587389606911,
+      "learning_rate": 7.178127248155604e-06,
+      "loss": 0.2699,
+      "step": 1645
+    },
+    {
+      "epoch": 3.745164960182025,
+      "grad_norm": 0.31124739072799124,
+      "learning_rate": 7.153727732134656e-06,
+      "loss": 0.2582,
+      "step": 1646
+    },
+    {
+      "epoch": 3.747440273037543,
+      "grad_norm": 0.3294380869999005,
+      "learning_rate": 7.12936072054534e-06,
+      "loss": 0.236,
+      "step": 1647
+    },
+    {
+      "epoch": 3.7497155858930604,
+      "grad_norm": 0.291163440361664,
+      "learning_rate": 7.1050262750425595e-06,
+      "loss": 0.2117,
+      "step": 1648
+    },
+    {
+      "epoch": 3.751990898748578,
+      "grad_norm": 0.3114191632293124,
+      "learning_rate": 7.080724457198824e-06,
+      "loss": 0.2324,
+      "step": 1649
+    },
+    {
+      "epoch": 3.7542662116040955,
+      "grad_norm": 0.3552526236860238,
+      "learning_rate": 7.056455328504104e-06,
+      "loss": 0.2511,
+      "step": 1650
+    },
+    {
+      "epoch": 3.756541524459613,
+      "grad_norm": 0.3400004570124406,
+      "learning_rate": 7.0322189503656435e-06,
+      "loss": 0.2529,
+      "step": 1651
+    },
+    {
+      "epoch": 3.758816837315131,
+      "grad_norm": 0.3310169176404239,
+      "learning_rate": 7.008015384107836e-06,
+      "loss": 0.2114,
+      "step": 1652
+    },
+    {
+      "epoch": 3.7610921501706485,
+      "grad_norm": 0.3413549271745786,
+      "learning_rate": 6.983844690972026e-06,
+      "loss": 0.2544,
+      "step": 1653
+    },
+    {
+      "epoch": 3.763367463026166,
+      "grad_norm": 0.3167657265637229,
+      "learning_rate": 6.959706932116388e-06,
+      "loss": 0.2192,
+      "step": 1654
+    },
+    {
+      "epoch": 3.765642775881684,
+      "grad_norm": 0.32633386745815,
+      "learning_rate": 6.935602168615792e-06,
+      "loss": 0.2766,
+      "step": 1655
+    },
+    {
+      "epoch": 3.7679180887372015,
+      "grad_norm": 0.3072647148220716,
+      "learning_rate": 6.911530461461588e-06,
+      "loss": 0.2063,
+      "step": 1656
+    },
+    {
+      "epoch": 3.770193401592719,
+      "grad_norm": 0.34161134198142956,
+      "learning_rate": 6.887491871561492e-06,
+      "loss": 0.2454,
+      "step": 1657
+    },
+    {
+      "epoch": 3.7724687144482365,
+      "grad_norm": 0.32664499006330927,
+      "learning_rate": 6.863486459739432e-06,
+      "loss": 0.2649,
+      "step": 1658
+    },
+    {
+      "epoch": 3.774744027303754,
+      "grad_norm": 0.2910768160518619,
+      "learning_rate": 6.839514286735387e-06,
+      "loss": 0.2164,
+      "step": 1659
+    },
+    {
+      "epoch": 3.777019340159272,
+      "grad_norm": 0.3188256092091695,
+      "learning_rate": 6.815575413205235e-06,
+      "loss": 0.2289,
+      "step": 1660
+    },
+    {
+      "epoch": 3.7792946530147895,
+      "grad_norm": 0.3218510980676637,
+      "learning_rate": 6.791669899720585e-06,
+      "loss": 0.244,
+      "step": 1661
+    },
+    {
+      "epoch": 3.781569965870307,
+      "grad_norm": 0.3166561361853286,
+      "learning_rate": 6.767797806768632e-06,
+      "loss": 0.2283,
+      "step": 1662
+    },
+    {
+      "epoch": 3.783845278725825,
+      "grad_norm": 0.30145917906713504,
+      "learning_rate": 6.743959194752045e-06,
+      "loss": 0.2155,
+      "step": 1663
+    },
+    {
+      "epoch": 3.7861205915813425,
+      "grad_norm": 0.2983439047380337,
+      "learning_rate": 6.720154123988736e-06,
+      "loss": 0.2092,
+      "step": 1664
+    },
+    {
+      "epoch": 3.78839590443686,
+      "grad_norm": 0.36687788457985926,
+      "learning_rate": 6.696382654711777e-06,
+      "loss": 0.2517,
+      "step": 1665
+    },
+    {
+      "epoch": 3.7906712172923775,
+      "grad_norm": 0.3666781923402803,
+      "learning_rate": 6.672644847069194e-06,
+      "loss": 0.2592,
+      "step": 1666
+    },
+    {
+      "epoch": 3.792946530147895,
+      "grad_norm": 0.29020154536513587,
+      "learning_rate": 6.648940761123861e-06,
+      "loss": 0.2371,
+      "step": 1667
+    },
+    {
+      "epoch": 3.795221843003413,
+      "grad_norm": 0.3113324715095254,
+      "learning_rate": 6.6252704568533276e-06,
+      "loss": 0.2205,
+      "step": 1668
+    },
+    {
+      "epoch": 3.7974971558589306,
+      "grad_norm": 0.31286782876307456,
+      "learning_rate": 6.6016339941496545e-06,
+      "loss": 0.1956,
+      "step": 1669
+    },
+    {
+      "epoch": 3.799772468714448,
+      "grad_norm": 0.34208468192794533,
+      "learning_rate": 6.578031432819263e-06,
+      "loss": 0.2232,
+      "step": 1670
+    },
+    {
+      "epoch": 3.802047781569966,
+      "grad_norm": 0.32683834732407896,
+      "learning_rate": 6.5544628325828395e-06,
+      "loss": 0.2547,
+      "step": 1671
+    },
+    {
+      "epoch": 3.8043230944254836,
+      "grad_norm": 0.31381099568132315,
+      "learning_rate": 6.530928253075095e-06,
+      "loss": 0.2558,
+      "step": 1672
+    },
+    {
+      "epoch": 3.806598407281001,
+      "grad_norm": 0.32223668259824584,
+      "learning_rate": 6.507427753844686e-06,
+      "loss": 0.2058,
+      "step": 1673
+    },
+    {
+      "epoch": 3.8088737201365186,
+      "grad_norm": 0.349394908203502,
+      "learning_rate": 6.483961394354017e-06,
+      "loss": 0.2578,
+      "step": 1674
+    },
+    {
+      "epoch": 3.8111490329920366,
+      "grad_norm": 0.30275090569381213,
+      "learning_rate": 6.460529233979127e-06,
+      "loss": 0.2182,
+      "step": 1675
+    },
+    {
+      "epoch": 3.813424345847554,
+      "grad_norm": 0.6374668756457942,
+      "learning_rate": 6.4371313320095184e-06,
+      "loss": 0.2717,
+      "step": 1676
+    },
+    {
+      "epoch": 3.8156996587030716,
+      "grad_norm": 0.3346189406584071,
+      "learning_rate": 6.413767747648003e-06,
+      "loss": 0.2278,
+      "step": 1677
+    },
+    {
+      "epoch": 3.8179749715585896,
+      "grad_norm": 0.29711987988698924,
+      "learning_rate": 6.390438540010555e-06,
+      "loss": 0.2525,
+      "step": 1678
+    },
+    {
+      "epoch": 3.820250284414107,
+      "grad_norm": 0.3191024033166926,
+      "learning_rate": 6.3671437681262e-06,
+      "loss": 0.2334,
+      "step": 1679
+    },
+    {
+      "epoch": 3.8225255972696246,
+      "grad_norm": 0.32524532366123116,
+      "learning_rate": 6.343883490936791e-06,
+      "loss": 0.243,
+      "step": 1680
+    },
+    {
+      "epoch": 3.824800910125142,
+      "grad_norm": 0.3062541748746329,
+      "learning_rate": 6.3206577672969275e-06,
+      "loss": 0.2294,
+      "step": 1681
+    },
+    {
+      "epoch": 3.8270762229806596,
+      "grad_norm": 0.2989301976941865,
+      "learning_rate": 6.297466655973763e-06,
+      "loss": 0.2118,
+      "step": 1682
+    },
+    {
+      "epoch": 3.8293515358361776,
+      "grad_norm": 0.32480587573609554,
+      "learning_rate": 6.27431021564688e-06,
+      "loss": 0.2532,
+      "step": 1683
+    },
+    {
+      "epoch": 3.831626848691695,
+      "grad_norm": 0.2927691847070395,
+      "learning_rate": 6.251188504908146e-06,
+      "loss": 0.2378,
+      "step": 1684
+    },
+    {
+      "epoch": 3.8339021615472126,
+      "grad_norm": 0.3225685587581803,
+      "learning_rate": 6.228101582261532e-06,
+      "loss": 0.252,
+      "step": 1685
+    },
+    {
+      "epoch": 3.8361774744027306,
+      "grad_norm": 0.3120723897409664,
+      "learning_rate": 6.205049506122982e-06,
+      "loss": 0.2455,
+      "step": 1686
+    },
+    {
+      "epoch": 3.838452787258248,
+      "grad_norm": 0.305763622169105,
+      "learning_rate": 6.182032334820313e-06,
+      "loss": 0.2243,
+      "step": 1687
+    },
+    {
+      "epoch": 3.8407281001137656,
+      "grad_norm": 0.30591586026809525,
+      "learning_rate": 6.159050126592971e-06,
+      "loss": 0.2241,
+      "step": 1688
+    },
+    {
+      "epoch": 3.843003412969283,
+      "grad_norm": 0.3254019572609186,
+      "learning_rate": 6.1361029395919746e-06,
+      "loss": 0.2408,
+      "step": 1689
+    },
+    {
+      "epoch": 3.8452787258248007,
+      "grad_norm": 0.299444349239614,
+      "learning_rate": 6.113190831879698e-06,
+      "loss": 0.204,
+      "step": 1690
+    },
+    {
+      "epoch": 3.8475540386803186,
+      "grad_norm": 0.32388225987773905,
+      "learning_rate": 6.09031386142978e-06,
+      "loss": 0.2135,
+      "step": 1691
+    },
+    {
+      "epoch": 3.849829351535836,
+      "grad_norm": 0.29890541047871094,
+      "learning_rate": 6.067472086126946e-06,
+      "loss": 0.2041,
+      "step": 1692
+    },
+    {
+      "epoch": 3.8521046643913537,
+      "grad_norm": 0.37306649437408845,
+      "learning_rate": 6.044665563766863e-06,
+      "loss": 0.2873,
+      "step": 1693
+    },
+    {
+      "epoch": 3.8543799772468716,
+      "grad_norm": 0.31341717595458574,
+      "learning_rate": 6.021894352055999e-06,
+      "loss": 0.2161,
+      "step": 1694
+    },
+    {
+      "epoch": 3.856655290102389,
+      "grad_norm": 0.30672897543736083,
+      "learning_rate": 5.999158508611496e-06,
+      "loss": 0.215,
+      "step": 1695
+    },
+    {
+      "epoch": 3.8589306029579067,
+      "grad_norm": 0.3314305934662205,
+      "learning_rate": 5.976458090960968e-06,
+      "loss": 0.2289,
+      "step": 1696
+    },
+    {
+      "epoch": 3.861205915813424,
+      "grad_norm": 0.30488419979629056,
+      "learning_rate": 5.9537931565424355e-06,
+      "loss": 0.2384,
+      "step": 1697
+    },
+    {
+      "epoch": 3.8634812286689417,
+      "grad_norm": 0.33073897968252053,
+      "learning_rate": 5.931163762704095e-06,
+      "loss": 0.243,
+      "step": 1698
+    },
+    {
+      "epoch": 3.8657565415244597,
+      "grad_norm": 0.30130774200700605,
+      "learning_rate": 5.908569966704247e-06,
+      "loss": 0.234,
+      "step": 1699
+    },
+    {
+      "epoch": 3.868031854379977,
+      "grad_norm": 0.2781595694315222,
+      "learning_rate": 5.886011825711117e-06,
+      "loss": 0.2349,
+      "step": 1700
+    },
+    {
+      "epoch": 3.8703071672354947,
+      "grad_norm": 0.3306004490674625,
+      "learning_rate": 5.863489396802693e-06,
+      "loss": 0.2772,
+      "step": 1701
+    },
+    {
+      "epoch": 3.8725824800910127,
+      "grad_norm": 0.29953193615493356,
+      "learning_rate": 5.8410027369666215e-06,
+      "loss": 0.2272,
+      "step": 1702
+    },
+    {
+      "epoch": 3.87485779294653,
+      "grad_norm": 0.31428588424946236,
+      "learning_rate": 5.818551903100045e-06,
+      "loss": 0.2528,
+      "step": 1703
+    },
+    {
+      "epoch": 3.8771331058020477,
+      "grad_norm": 0.30490740383671394,
+      "learning_rate": 5.7961369520094365e-06,
+      "loss": 0.174,
+      "step": 1704
+    },
+    {
+      "epoch": 3.8794084186575652,
+      "grad_norm": 0.34149110657492043,
+      "learning_rate": 5.773757940410503e-06,
+      "loss": 0.2552,
+      "step": 1705
+    },
+    {
+      "epoch": 3.881683731513083,
+      "grad_norm": 0.31442759218870003,
+      "learning_rate": 5.7514149249279915e-06,
+      "loss": 0.2149,
+      "step": 1706
+    },
+    {
+      "epoch": 3.8839590443686007,
+      "grad_norm": 0.3054708293276751,
+      "learning_rate": 5.729107962095583e-06,
+      "loss": 0.2143,
+      "step": 1707
+    },
+    {
+      "epoch": 3.8862343572241183,
+      "grad_norm": 0.32096087445485844,
+      "learning_rate": 5.7068371083557425e-06,
+      "loss": 0.2439,
+      "step": 1708
+    },
+    {
+      "epoch": 3.888509670079636,
+      "grad_norm": 0.3315098330191521,
+      "learning_rate": 5.684602420059544e-06,
+      "loss": 0.2381,
+      "step": 1709
+    },
+    {
+      "epoch": 3.8907849829351537,
+      "grad_norm": 0.3298488700402563,
+      "learning_rate": 5.6624039534665775e-06,
+      "loss": 0.2254,
+      "step": 1710
+    },
+    {
+      "epoch": 3.8930602957906713,
+      "grad_norm": 0.3071846393117354,
+      "learning_rate": 5.640241764744778e-06,
+      "loss": 0.2169,
+      "step": 1711
+    },
+    {
+      "epoch": 3.8953356086461888,
+      "grad_norm": 0.32783291817800214,
+      "learning_rate": 5.618115909970277e-06,
+      "loss": 0.2764,
+      "step": 1712
+    },
+    {
+      "epoch": 3.8976109215017063,
+      "grad_norm": 0.3445664041400783,
+      "learning_rate": 5.5960264451272874e-06,
+      "loss": 0.2304,
+      "step": 1713
+    },
+    {
+      "epoch": 3.8998862343572243,
+      "grad_norm": 0.30215319207944386,
+      "learning_rate": 5.573973426107926e-06,
+      "loss": 0.1803,
+      "step": 1714
+    },
+    {
+      "epoch": 3.9021615472127418,
+      "grad_norm": 0.30959760486419285,
+      "learning_rate": 5.55195690871211e-06,
+      "loss": 0.2265,
+      "step": 1715
+    },
+    {
+      "epoch": 3.9044368600682593,
+      "grad_norm": 0.32219954567635933,
+      "learning_rate": 5.529976948647399e-06,
+      "loss": 0.2654,
+      "step": 1716
+    },
+    {
+      "epoch": 3.9067121729237773,
+      "grad_norm": 0.2952888972067496,
+      "learning_rate": 5.5080336015288305e-06,
+      "loss": 0.2169,
+      "step": 1717
+    },
+    {
+      "epoch": 3.908987485779295,
+      "grad_norm": 0.33608629315045474,
+      "learning_rate": 5.486126922878827e-06,
+      "loss": 0.2257,
+      "step": 1718
+    },
+    {
+      "epoch": 3.9112627986348123,
+      "grad_norm": 0.38313895035840695,
+      "learning_rate": 5.464256968127024e-06,
+      "loss": 0.2548,
+      "step": 1719
+    },
+    {
+      "epoch": 3.91353811149033,
+      "grad_norm": 0.29776975666195865,
+      "learning_rate": 5.442423792610118e-06,
+      "loss": 0.2036,
+      "step": 1720
+    },
+    {
+      "epoch": 3.9158134243458473,
+      "grad_norm": 6.2431277746955445,
+      "learning_rate": 5.4206274515717735e-06,
+      "loss": 0.3037,
+      "step": 1721
+    },
+    {
+      "epoch": 3.9180887372013653,
+      "grad_norm": 0.28750011822850136,
+      "learning_rate": 5.3988680001624246e-06,
+      "loss": 0.2043,
+      "step": 1722
+    },
+    {
+      "epoch": 3.920364050056883,
+      "grad_norm": 0.3534671305395048,
+      "learning_rate": 5.377145493439189e-06,
+      "loss": 0.2488,
+      "step": 1723
+    },
+    {
+      "epoch": 3.9226393629124003,
+      "grad_norm": 0.3387306762933544,
+      "learning_rate": 5.3554599863656964e-06,
+      "loss": 0.2459,
+      "step": 1724
+    },
+    {
+      "epoch": 3.9249146757679183,
+      "grad_norm": 0.3380266671759454,
+      "learning_rate": 5.333811533811945e-06,
+      "loss": 0.252,
+      "step": 1725
+    },
+    {
+      "epoch": 3.927189988623436,
+      "grad_norm": 0.3245287829380904,
+      "learning_rate": 5.312200190554197e-06,
+      "loss": 0.2136,
+      "step": 1726
+    },
+    {
+      "epoch": 3.9294653014789533,
+      "grad_norm": 0.30177056610765146,
+      "learning_rate": 5.290626011274811e-06,
+      "loss": 0.2157,
+      "step": 1727
+    },
+    {
+      "epoch": 3.931740614334471,
+      "grad_norm": 0.3184248640213574,
+      "learning_rate": 5.269089050562095e-06,
+      "loss": 0.2413,
+      "step": 1728
+    },
+    {
+      "epoch": 3.9340159271899884,
+      "grad_norm": 0.3449914811634588,
+      "learning_rate": 5.24758936291021e-06,
+      "loss": 0.2774,
+      "step": 1729
+    },
+    {
+      "epoch": 3.9362912400455063,
+      "grad_norm": 0.29415305944975656,
+      "learning_rate": 5.226127002718984e-06,
+      "loss": 0.2211,
+      "step": 1730
+    },
+    {
+      "epoch": 3.938566552901024,
+      "grad_norm": 0.3063318019988386,
+      "learning_rate": 5.204702024293809e-06,
+      "loss": 0.2348,
+      "step": 1731
+    },
+    {
+      "epoch": 3.9408418657565414,
+      "grad_norm": 0.31018938854217054,
+      "learning_rate": 5.1833144818454984e-06,
+      "loss": 0.2326,
+      "step": 1732
+    },
+    {
+      "epoch": 3.9431171786120593,
+      "grad_norm": 0.31877450093183923,
+      "learning_rate": 5.161964429490118e-06,
+      "loss": 0.2323,
+      "step": 1733
+    },
+    {
+      "epoch": 3.945392491467577,
+      "grad_norm": 0.3129394494579511,
+      "learning_rate": 5.140651921248896e-06,
+      "loss": 0.2156,
+      "step": 1734
+    },
+    {
+      "epoch": 3.9476678043230944,
+      "grad_norm": 0.31451145749768966,
+      "learning_rate": 5.119377011048066e-06,
+      "loss": 0.205,
+      "step": 1735
+    },
+    {
+      "epoch": 3.949943117178612,
+      "grad_norm": 0.31972587932927893,
+      "learning_rate": 5.098139752718705e-06,
+      "loss": 0.2147,
+      "step": 1736
+    },
+    {
+      "epoch": 3.9522184300341294,
+      "grad_norm": 0.3097361980800767,
+      "learning_rate": 5.07694019999665e-06,
+      "loss": 0.2493,
+      "step": 1737
+    },
+    {
+      "epoch": 3.9544937428896474,
+      "grad_norm": 0.28718295253995646,
+      "learning_rate": 5.055778406522302e-06,
+      "loss": 0.2076,
+      "step": 1738
+    },
+    {
+      "epoch": 3.956769055745165,
+      "grad_norm": 0.3304463564819114,
+      "learning_rate": 5.03465442584055e-06,
+      "loss": 0.2384,
+      "step": 1739
+    },
+    {
+      "epoch": 3.9590443686006824,
+      "grad_norm": 0.2894887585174249,
+      "learning_rate": 5.013568311400599e-06,
+      "loss": 0.205,
+      "step": 1740
+    },
+    {
+      "epoch": 3.9613196814562004,
+      "grad_norm": 0.355332008884822,
+      "learning_rate": 4.992520116555824e-06,
+      "loss": 0.2662,
+      "step": 1741
+    },
+    {
+      "epoch": 3.963594994311718,
+      "grad_norm": 0.317299732708024,
+      "learning_rate": 4.971509894563675e-06,
+      "loss": 0.214,
+      "step": 1742
+    },
+    {
+      "epoch": 3.9658703071672354,
+      "grad_norm": 0.3022323304304223,
+      "learning_rate": 4.9505376985855205e-06,
+      "loss": 0.2071,
+      "step": 1743
+    },
+    {
+      "epoch": 3.968145620022753,
+      "grad_norm": 0.3240147150371773,
+      "learning_rate": 4.929603581686493e-06,
+      "loss": 0.2562,
+      "step": 1744
+    },
+    {
+      "epoch": 3.970420932878271,
+      "grad_norm": 0.2920458914250409,
+      "learning_rate": 4.908707596835396e-06,
+      "loss": 0.2288,
+      "step": 1745
+    },
+    {
+      "epoch": 3.9726962457337884,
+      "grad_norm": 0.34721554596739446,
+      "learning_rate": 4.887849796904545e-06,
+      "loss": 0.2738,
+      "step": 1746
+    },
+    {
+      "epoch": 3.974971558589306,
+      "grad_norm": 0.2790951894936184,
+      "learning_rate": 4.867030234669623e-06,
+      "loss": 0.1989,
+      "step": 1747
+    },
+    {
+      "epoch": 3.977246871444824,
+      "grad_norm": 0.30847017475707605,
+      "learning_rate": 4.846248962809583e-06,
+      "loss": 0.223,
+      "step": 1748
+    },
+    {
+      "epoch": 3.9795221843003414,
+      "grad_norm": 0.31553898501131733,
+      "learning_rate": 4.825506033906473e-06,
+      "loss": 0.2499,
+      "step": 1749
+    },
+    {
+      "epoch": 3.981797497155859,
+      "grad_norm": 0.30621288235785427,
+      "learning_rate": 4.804801500445338e-06,
+      "loss": 0.2343,
+      "step": 1750
+    },
+    {
+      "epoch": 3.9840728100113765,
+      "grad_norm": 0.31119300827944296,
+      "learning_rate": 4.784135414814073e-06,
+      "loss": 0.2293,
+      "step": 1751
+    },
+    {
+      "epoch": 3.986348122866894,
+      "grad_norm": 0.315135314468061,
+      "learning_rate": 4.763507829303273e-06,
+      "loss": 0.2426,
+      "step": 1752
+    },
+    {
+      "epoch": 3.988623435722412,
+      "grad_norm": 0.30300641894384645,
+      "learning_rate": 4.742918796106133e-06,
+      "loss": 0.2327,
+      "step": 1753
+    },
+    {
+      "epoch": 3.9908987485779295,
+      "grad_norm": 0.28620975306710267,
+      "learning_rate": 4.7223683673183065e-06,
+      "loss": 0.221,
+      "step": 1754
+    },
+    {
+      "epoch": 3.993174061433447,
+      "grad_norm": 0.35135584656251523,
+      "learning_rate": 4.701856594937744e-06,
+      "loss": 0.2737,
+      "step": 1755
+    },
+    {
+      "epoch": 3.995449374288965,
+      "grad_norm": 0.30154724781040915,
+      "learning_rate": 4.68138353086461e-06,
+      "loss": 0.2218,
+      "step": 1756
+    },
+    {
+      "epoch": 3.9977246871444825,
+      "grad_norm": 0.3270433705829516,
+      "learning_rate": 4.6609492269011035e-06,
+      "loss": 0.2238,
+      "step": 1757
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.45117243369326687,
+      "learning_rate": 4.640553734751372e-06,
+      "loss": 0.3146,
+      "step": 1758
+    },
+    {
+      "epoch": 4.0022753128555175,
+      "grad_norm": 0.4769075032451865,
+      "learning_rate": 4.620197106021352e-06,
+      "loss": 0.2012,
+      "step": 1759
+    },
+    {
+      "epoch": 4.004550625711035,
+      "grad_norm": 0.3976079400523734,
+      "learning_rate": 4.5998793922186315e-06,
+      "loss": 0.1877,
+      "step": 1760
+    },
+    {
+      "epoch": 4.006825938566553,
+      "grad_norm": 0.3093737942856265,
+      "learning_rate": 4.579600644752351e-06,
+      "loss": 0.1623,
+      "step": 1761
+    },
+    {
+      "epoch": 4.009101251422071,
+      "grad_norm": 0.3078795723720559,
+      "learning_rate": 4.5593609149330595e-06,
+      "loss": 0.2004,
+      "step": 1762
+    },
+    {
+      "epoch": 4.0113765642775885,
+      "grad_norm": 0.3945750519655033,
+      "learning_rate": 4.539160253972554e-06,
+      "loss": 0.1868,
+      "step": 1763
+    },
+    {
+      "epoch": 4.013651877133106,
+      "grad_norm": 0.4965258610062153,
+      "learning_rate": 4.518998712983813e-06,
+      "loss": 0.2092,
+      "step": 1764
+    },
+    {
+      "epoch": 4.0159271899886235,
+      "grad_norm": 0.38386284945859933,
+      "learning_rate": 4.498876342980796e-06,
+      "loss": 0.168,
+      "step": 1765
+    },
+    {
+      "epoch": 4.018202502844141,
+      "grad_norm": 0.3478458205726723,
+      "learning_rate": 4.478793194878372e-06,
+      "loss": 0.1832,
+      "step": 1766
+    },
+    {
+      "epoch": 4.020477815699659,
+      "grad_norm": 0.31275256183291067,
+      "learning_rate": 4.458749319492173e-06,
+      "loss": 0.1595,
+      "step": 1767
+    },
+    {
+      "epoch": 4.022753128555176,
+      "grad_norm": 0.37246833572422183,
+      "learning_rate": 4.438744767538437e-06,
+      "loss": 0.174,
+      "step": 1768
+    },
+    {
+      "epoch": 4.025028441410694,
+      "grad_norm": 0.3836746699164272,
+      "learning_rate": 4.4187795896339215e-06,
+      "loss": 0.1718,
+      "step": 1769
+    },
+    {
+      "epoch": 4.027303754266212,
+      "grad_norm": 0.3799208669671832,
+      "learning_rate": 4.39885383629576e-06,
+      "loss": 0.1911,
+      "step": 1770
+    },
+    {
+      "epoch": 4.0295790671217295,
+      "grad_norm": 0.3362384313237527,
+      "learning_rate": 4.378967557941312e-06,
+      "loss": 0.1724,
+      "step": 1771
+    },
+    {
+      "epoch": 4.031854379977247,
+      "grad_norm": 0.2893434745348152,
+      "learning_rate": 4.359120804888079e-06,
+      "loss": 0.1408,
+      "step": 1772
+    },
+    {
+      "epoch": 4.034129692832765,
+      "grad_norm": 0.29531089760755574,
+      "learning_rate": 4.339313627353534e-06,
+      "loss": 0.1492,
+      "step": 1773
+    },
+    {
+      "epoch": 4.036405005688282,
+      "grad_norm": 0.35813511824881245,
+      "learning_rate": 4.319546075455021e-06,
+      "loss": 0.1926,
+      "step": 1774
+    },
+    {
+      "epoch": 4.0386803185438,
+      "grad_norm": 0.3569884122339448,
+      "learning_rate": 4.299818199209629e-06,
+      "loss": 0.1869,
+      "step": 1775
+    },
+    {
+      "epoch": 4.040955631399317,
+      "grad_norm": 0.33200820587055785,
+      "learning_rate": 4.280130048534037e-06,
+      "loss": 0.1765,
+      "step": 1776
+    },
+    {
+      "epoch": 4.043230944254835,
+      "grad_norm": 0.29622461605612205,
+      "learning_rate": 4.2604816732444275e-06,
+      "loss": 0.1566,
+      "step": 1777
+    },
+    {
+      "epoch": 4.045506257110353,
+      "grad_norm": 0.292514545715088,
+      "learning_rate": 4.240873123056337e-06,
+      "loss": 0.1563,
+      "step": 1778
+    },
+    {
+      "epoch": 4.047781569965871,
+      "grad_norm": 0.31986845280561066,
+      "learning_rate": 4.2213044475845225e-06,
+      "loss": 0.1746,
+      "step": 1779
+    },
+    {
+      "epoch": 4.050056882821388,
+      "grad_norm": 0.33983512793590304,
+      "learning_rate": 4.201775696342862e-06,
+      "loss": 0.1592,
+      "step": 1780
+    },
+    {
+      "epoch": 4.052332195676906,
+      "grad_norm": 0.31379639034080037,
+      "learning_rate": 4.182286918744198e-06,
+      "loss": 0.2042,
+      "step": 1781
+    },
+    {
+      "epoch": 4.054607508532423,
+      "grad_norm": 0.3023478945890041,
+      "learning_rate": 4.162838164100249e-06,
+      "loss": 0.1579,
+      "step": 1782
+    },
+    {
+      "epoch": 4.056882821387941,
+      "grad_norm": 0.2969213560175581,
+      "learning_rate": 4.143429481621457e-06,
+      "loss": 0.1743,
+      "step": 1783
+    },
+    {
+      "epoch": 4.059158134243458,
+      "grad_norm": 0.3119541374701204,
+      "learning_rate": 4.124060920416859e-06,
+      "loss": 0.1659,
+      "step": 1784
+    },
+    {
+      "epoch": 4.061433447098976,
+      "grad_norm": 0.31438726822031443,
+      "learning_rate": 4.104732529493991e-06,
+      "loss": 0.1768,
+      "step": 1785
+    },
+    {
+      "epoch": 4.063708759954494,
+      "grad_norm": 0.32453073231477475,
+      "learning_rate": 4.085444357758747e-06,
+      "loss": 0.1684,
+      "step": 1786
+    },
+    {
+      "epoch": 4.065984072810012,
+      "grad_norm": 0.3181510377112917,
+      "learning_rate": 4.066196454015241e-06,
+      "loss": 0.174,
+      "step": 1787
+    },
+    {
+      "epoch": 4.068259385665529,
+      "grad_norm": 0.285844082778199,
+      "learning_rate": 4.046988866965715e-06,
+      "loss": 0.1685,
+      "step": 1788
+    },
+    {
+      "epoch": 4.070534698521047,
+      "grad_norm": 0.32331948067270616,
+      "learning_rate": 4.0278216452103855e-06,
+      "loss": 0.1844,
+      "step": 1789
+    },
+    {
+      "epoch": 4.072810011376564,
+      "grad_norm": 0.2862103052814747,
+      "learning_rate": 4.008694837247345e-06,
+      "loss": 0.1526,
+      "step": 1790
+    },
+    {
+      "epoch": 4.075085324232082,
+      "grad_norm": 0.28875087074361716,
+      "learning_rate": 3.9896084914724295e-06,
+      "loss": 0.1742,
+      "step": 1791
+    },
+    {
+      "epoch": 4.077360637087599,
+      "grad_norm": 0.2702349719704281,
+      "learning_rate": 3.970562656179078e-06,
+      "loss": 0.1647,
+      "step": 1792
+    },
+    {
+      "epoch": 4.079635949943118,
+      "grad_norm": 0.2773168310451333,
+      "learning_rate": 3.951557379558248e-06,
+      "loss": 0.1684,
+      "step": 1793
+    },
+    {
+      "epoch": 4.081911262798635,
+      "grad_norm": 0.30690349747480467,
+      "learning_rate": 3.932592709698266e-06,
+      "loss": 0.1742,
+      "step": 1794
+    },
+    {
+      "epoch": 4.084186575654153,
+      "grad_norm": 0.3017798594775822,
+      "learning_rate": 3.913668694584705e-06,
+      "loss": 0.1889,
+      "step": 1795
+    },
+    {
+      "epoch": 4.08646188850967,
+      "grad_norm": 0.2956134096062524,
+      "learning_rate": 3.894785382100284e-06,
+      "loss": 0.2176,
+      "step": 1796
+    },
+    {
+      "epoch": 4.088737201365188,
+      "grad_norm": 0.3051750160939217,
+      "learning_rate": 3.875942820024716e-06,
+      "loss": 0.1815,
+      "step": 1797
+    },
+    {
+      "epoch": 4.091012514220705,
+      "grad_norm": 0.3130918040476627,
+      "learning_rate": 3.857141056034623e-06,
+      "loss": 0.1689,
+      "step": 1798
+    },
+    {
+      "epoch": 4.093287827076223,
+      "grad_norm": 0.2842894915597038,
+      "learning_rate": 3.838380137703395e-06,
+      "loss": 0.1655,
+      "step": 1799
+    },
+    {
+      "epoch": 4.09556313993174,
+      "grad_norm": 0.29715175234366936,
+      "learning_rate": 3.819660112501053e-06,
+      "loss": 0.1796,
+      "step": 1800
+    },
+    {
+      "epoch": 4.097838452787259,
+      "grad_norm": 0.29377940866244734,
+      "learning_rate": 3.8009810277941684e-06,
+      "loss": 0.1639,
+      "step": 1801
+    },
+    {
+      "epoch": 4.100113765642776,
+      "grad_norm": 0.29276927104650885,
+      "learning_rate": 3.7823429308457216e-06,
+      "loss": 0.1973,
+      "step": 1802
+    },
+    {
+      "epoch": 4.102389078498294,
+      "grad_norm": 0.2994435046602159,
+      "learning_rate": 3.763745868814963e-06,
+      "loss": 0.1804,
+      "step": 1803
+    },
+    {
+      "epoch": 4.104664391353811,
+      "grad_norm": 0.28482583960056823,
+      "learning_rate": 3.7451898887573414e-06,
+      "loss": 0.1523,
+      "step": 1804
+    },
+    {
+      "epoch": 4.106939704209329,
+      "grad_norm": 0.28838680832452307,
+      "learning_rate": 3.72667503762433e-06,
+      "loss": 0.1579,
+      "step": 1805
+    },
+    {
+      "epoch": 4.109215017064846,
+      "grad_norm": 0.270794150814169,
+      "learning_rate": 3.7082013622633574e-06,
+      "loss": 0.1802,
+      "step": 1806
+    },
+    {
+      "epoch": 4.111490329920364,
+      "grad_norm": 0.28536534956566173,
+      "learning_rate": 3.6897689094176614e-06,
+      "loss": 0.155,
+      "step": 1807
+    },
+    {
+      "epoch": 4.113765642775881,
+      "grad_norm": 0.2943625378053496,
+      "learning_rate": 3.671377725726164e-06,
+      "loss": 0.1482,
+      "step": 1808
+    },
+    {
+      "epoch": 4.1160409556314,
+      "grad_norm": 0.30861999050570416,
+      "learning_rate": 3.653027857723379e-06,
+      "loss": 0.1741,
+      "step": 1809
+    },
+    {
+      "epoch": 4.118316268486917,
+      "grad_norm": 0.2923682813079667,
+      "learning_rate": 3.6347193518392776e-06,
+      "loss": 0.1523,
+      "step": 1810
+    },
+    {
+      "epoch": 4.120591581342435,
+      "grad_norm": 0.2895713734766889,
+      "learning_rate": 3.6164522543991675e-06,
+      "loss": 0.1847,
+      "step": 1811
+    },
+    {
+      "epoch": 4.122866894197952,
+      "grad_norm": 0.2977731114241597,
+      "learning_rate": 3.5982266116235918e-06,
+      "loss": 0.1849,
+      "step": 1812
+    },
+    {
+      "epoch": 4.12514220705347,
+      "grad_norm": 0.2915909276442138,
+      "learning_rate": 3.5800424696281933e-06,
+      "loss": 0.1835,
+      "step": 1813
+    },
+    {
+      "epoch": 4.127417519908987,
+      "grad_norm": 0.2964235964857184,
+      "learning_rate": 3.5618998744236112e-06,
+      "loss": 0.1877,
+      "step": 1814
+    },
+    {
+      "epoch": 4.129692832764505,
+      "grad_norm": 0.2744018510234478,
+      "learning_rate": 3.543798871915367e-06,
+      "loss": 0.1762,
+      "step": 1815
+    },
+    {
+      "epoch": 4.131968145620022,
+      "grad_norm": 0.28501466540089593,
+      "learning_rate": 3.5257395079037293e-06,
+      "loss": 0.1727,
+      "step": 1816
+    },
+    {
+      "epoch": 4.134243458475541,
+      "grad_norm": 0.3124331135487368,
+      "learning_rate": 3.5077218280836143e-06,
+      "loss": 0.1898,
+      "step": 1817
+    },
+    {
+      "epoch": 4.136518771331058,
+      "grad_norm": 0.27779228301613235,
+      "learning_rate": 3.4897458780444815e-06,
+      "loss": 0.1827,
+      "step": 1818
+    },
+    {
+      "epoch": 4.138794084186576,
+      "grad_norm": 0.29143118659774375,
+      "learning_rate": 3.4718117032701736e-06,
+      "loss": 0.1855,
+      "step": 1819
+    },
+    {
+      "epoch": 4.141069397042093,
+      "grad_norm": 0.2955170890916771,
+      "learning_rate": 3.453919349138859e-06,
+      "loss": 0.1892,
+      "step": 1820
+    },
+    {
+      "epoch": 4.143344709897611,
+      "grad_norm": 0.2960059703618523,
+      "learning_rate": 3.4360688609228696e-06,
+      "loss": 0.1992,
+      "step": 1821
+    },
+    {
+      "epoch": 4.145620022753128,
+      "grad_norm": 0.3018565283140866,
+      "learning_rate": 3.418260283788617e-06,
+      "loss": 0.1778,
+      "step": 1822
+    },
+    {
+      "epoch": 4.147895335608646,
+      "grad_norm": 0.29023871017690334,
+      "learning_rate": 3.4004936627964667e-06,
+      "loss": 0.1688,
+      "step": 1823
+    },
+    {
+      "epoch": 4.150170648464163,
+      "grad_norm": 0.2761367828407954,
+      "learning_rate": 3.382769042900613e-06,
+      "loss": 0.1893,
+      "step": 1824
+    },
+    {
+      "epoch": 4.152445961319682,
+      "grad_norm": 0.30011109570647254,
+      "learning_rate": 3.365086468948988e-06,
+      "loss": 0.1524,
+      "step": 1825
+    },
+    {
+      "epoch": 4.154721274175199,
+      "grad_norm": 0.30881472569270824,
+      "learning_rate": 3.3474459856831352e-06,
+      "loss": 0.1697,
+      "step": 1826
+    },
+    {
+      "epoch": 4.156996587030717,
+      "grad_norm": 0.2948790323073478,
+      "learning_rate": 3.3298476377380907e-06,
+      "loss": 0.1788,
+      "step": 1827
+    },
+    {
+      "epoch": 4.159271899886234,
+      "grad_norm": 0.28948403698960046,
+      "learning_rate": 3.312291469642286e-06,
+      "loss": 0.1775,
+      "step": 1828
+    },
+    {
+      "epoch": 4.161547212741752,
+      "grad_norm": 0.44070767588943566,
+      "learning_rate": 3.294777525817414e-06,
+      "loss": 0.2049,
+      "step": 1829
+    },
+    {
+      "epoch": 4.163822525597269,
+      "grad_norm": 0.2852153227163277,
+      "learning_rate": 3.277305850578345e-06,
+      "loss": 0.19,
+      "step": 1830
+    },
+    {
+      "epoch": 4.166097838452787,
+      "grad_norm": 0.299312680649051,
+      "learning_rate": 3.2598764881329915e-06,
+      "loss": 0.1784,
+      "step": 1831
+    },
+    {
+      "epoch": 4.168373151308305,
+      "grad_norm": 0.2771732420759297,
+      "learning_rate": 3.242489482582194e-06,
+      "loss": 0.1529,
+      "step": 1832
+    },
+    {
+      "epoch": 4.170648464163823,
+      "grad_norm": 0.28136004468362263,
+      "learning_rate": 3.225144877919635e-06,
+      "loss": 0.1675,
+      "step": 1833
+    },
+    {
+      "epoch": 4.17292377701934,
+      "grad_norm": 0.26408170571926015,
+      "learning_rate": 3.2078427180317084e-06,
+      "loss": 0.1751,
+      "step": 1834
+    },
+    {
+      "epoch": 4.175199089874858,
+      "grad_norm": 0.2945328639456804,
+      "learning_rate": 3.1905830466973975e-06,
+      "loss": 0.1822,
+      "step": 1835
+    },
+    {
+      "epoch": 4.177474402730375,
+      "grad_norm": 0.2943639900150499,
+      "learning_rate": 3.173365907588195e-06,
+      "loss": 0.1674,
+      "step": 1836
+    },
+    {
+      "epoch": 4.179749715585893,
+      "grad_norm": 0.2995630964889057,
+      "learning_rate": 3.1561913442679694e-06,
+      "loss": 0.1787,
+      "step": 1837
+    },
+    {
+      "epoch": 4.18202502844141,
+      "grad_norm": 0.28199152495272156,
+      "learning_rate": 3.1390594001928654e-06,
+      "loss": 0.1761,
+      "step": 1838
+    },
+    {
+      "epoch": 4.184300341296928,
+      "grad_norm": 0.29468647972746453,
+      "learning_rate": 3.1219701187111817e-06,
+      "loss": 0.179,
+      "step": 1839
+    },
+    {
+      "epoch": 4.186575654152446,
+      "grad_norm": 0.29082483162051903,
+      "learning_rate": 3.1049235430632696e-06,
+      "loss": 0.1955,
+      "step": 1840
+    },
+    {
+      "epoch": 4.188850967007964,
+      "grad_norm": 0.29879641184857786,
+      "learning_rate": 3.0879197163814334e-06,
+      "loss": 0.1724,
+      "step": 1841
+    },
+    {
+      "epoch": 4.191126279863481,
+      "grad_norm": 0.31145266853383446,
+      "learning_rate": 3.0709586816898086e-06,
+      "loss": 0.2368,
+      "step": 1842
+    },
+    {
+      "epoch": 4.193401592718999,
+      "grad_norm": 0.2721191961469994,
+      "learning_rate": 3.054040481904246e-06,
+      "loss": 0.134,
+      "step": 1843
+    },
+    {
+      "epoch": 4.1956769055745164,
+      "grad_norm": 0.27510529396254574,
+      "learning_rate": 3.037165159832223e-06,
+      "loss": 0.1499,
+      "step": 1844
+    },
+    {
+      "epoch": 4.197952218430034,
+      "grad_norm": 0.2797481261232522,
+      "learning_rate": 3.0203327581727195e-06,
+      "loss": 0.1666,
+      "step": 1845
+    },
+    {
+      "epoch": 4.2002275312855515,
+      "grad_norm": 0.306155624592477,
+      "learning_rate": 3.0035433195161266e-06,
+      "loss": 0.17,
+      "step": 1846
+    },
+    {
+      "epoch": 4.202502844141069,
+      "grad_norm": 0.28488195762459484,
+      "learning_rate": 2.986796886344112e-06,
+      "loss": 0.1831,
+      "step": 1847
+    },
+    {
+      "epoch": 4.204778156996587,
+      "grad_norm": 0.28418092083664215,
+      "learning_rate": 2.9700935010295316e-06,
+      "loss": 0.1528,
+      "step": 1848
+    },
+    {
+      "epoch": 4.207053469852105,
+      "grad_norm": 0.2945372190049564,
+      "learning_rate": 2.953433205836327e-06,
+      "loss": 0.1585,
+      "step": 1849
+    },
+    {
+      "epoch": 4.2093287827076225,
+      "grad_norm": 0.27504913273286286,
+      "learning_rate": 2.9368160429194127e-06,
+      "loss": 0.1718,
+      "step": 1850
+    },
+    {
+      "epoch": 4.21160409556314,
+      "grad_norm": 0.2880017693010912,
+      "learning_rate": 2.920242054324547e-06,
+      "loss": 0.1598,
+      "step": 1851
+    },
+    {
+      "epoch": 4.2138794084186575,
+      "grad_norm": 0.2983460026372298,
+      "learning_rate": 2.9037112819882685e-06,
+      "loss": 0.1737,
+      "step": 1852
+    },
+    {
+      "epoch": 4.216154721274175,
+      "grad_norm": 0.2711852394125882,
+      "learning_rate": 2.887223767737759e-06,
+      "loss": 0.1713,
+      "step": 1853
+    },
+    {
+      "epoch": 4.2184300341296925,
+      "grad_norm": 0.28858227659794555,
+      "learning_rate": 2.8707795532907457e-06,
+      "loss": 0.1474,
+      "step": 1854
+    },
+    {
+      "epoch": 4.22070534698521,
+      "grad_norm": 0.27749228116646407,
+      "learning_rate": 2.8543786802553943e-06,
+      "loss": 0.1627,
+      "step": 1855
+    },
+    {
+      "epoch": 4.2229806598407285,
+      "grad_norm": 0.29530178216415215,
+      "learning_rate": 2.8380211901302e-06,
+      "loss": 0.2047,
+      "step": 1856
+    },
+    {
+      "epoch": 4.225255972696246,
+      "grad_norm": 0.3023235137542973,
+      "learning_rate": 2.8217071243039027e-06,
+      "loss": 0.2001,
+      "step": 1857
+    },
+    {
+      "epoch": 4.2275312855517635,
+      "grad_norm": 0.29111559436168455,
+      "learning_rate": 2.8054365240553627e-06,
+      "loss": 0.1691,
+      "step": 1858
+    },
+    {
+      "epoch": 4.229806598407281,
+      "grad_norm": 0.28642161220324286,
+      "learning_rate": 2.789209430553448e-06,
+      "loss": 0.1825,
+      "step": 1859
+    },
+    {
+      "epoch": 4.2320819112627985,
+      "grad_norm": 0.28236465353986595,
+      "learning_rate": 2.773025884856957e-06,
+      "loss": 0.187,
+      "step": 1860
+    },
+    {
+      "epoch": 4.234357224118316,
+      "grad_norm": 0.2898860096952971,
+      "learning_rate": 2.756885927914499e-06,
+      "loss": 0.1411,
+      "step": 1861
+    },
+    {
+      "epoch": 4.236632536973834,
+      "grad_norm": 0.2799644334188099,
+      "learning_rate": 2.740789600564391e-06,
+      "loss": 0.1544,
+      "step": 1862
+    },
+    {
+      "epoch": 4.238907849829351,
+      "grad_norm": 0.283940319165611,
+      "learning_rate": 2.724736943534554e-06,
+      "loss": 0.1562,
+      "step": 1863
+    },
+    {
+      "epoch": 4.2411831626848695,
+      "grad_norm": 0.29108290944445414,
+      "learning_rate": 2.7087279974424064e-06,
+      "loss": 0.1814,
+      "step": 1864
+    },
+    {
+      "epoch": 4.243458475540387,
+      "grad_norm": 0.2881703274709376,
+      "learning_rate": 2.692762802794775e-06,
+      "loss": 0.1733,
+      "step": 1865
+    },
+    {
+      "epoch": 4.2457337883959045,
+      "grad_norm": 0.2866066755967616,
+      "learning_rate": 2.6768413999877885e-06,
+      "loss": 0.1772,
+      "step": 1866
+    },
+    {
+      "epoch": 4.248009101251422,
+      "grad_norm": 0.2871068941071818,
+      "learning_rate": 2.660963829306753e-06,
+      "loss": 0.1868,
+      "step": 1867
+    },
+    {
+      "epoch": 4.25028441410694,
+      "grad_norm": 0.28265703450221186,
+      "learning_rate": 2.645130130926086e-06,
+      "loss": 0.1845,
+      "step": 1868
+    },
+    {
+      "epoch": 4.252559726962457,
+      "grad_norm": 0.28517829274735246,
+      "learning_rate": 2.629340344909186e-06,
+      "loss": 0.159,
+      "step": 1869
+    },
+    {
+      "epoch": 4.254835039817975,
+      "grad_norm": 0.28641183845888735,
+      "learning_rate": 2.6135945112083506e-06,
+      "loss": 0.1842,
+      "step": 1870
+    },
+    {
+      "epoch": 4.257110352673493,
+      "grad_norm": 0.2959083940515715,
+      "learning_rate": 2.5978926696646567e-06,
+      "loss": 0.1846,
+      "step": 1871
+    },
+    {
+      "epoch": 4.2593856655290105,
+      "grad_norm": 0.3125252382625625,
+      "learning_rate": 2.5822348600078684e-06,
+      "loss": 0.2053,
+      "step": 1872
+    },
+    {
+      "epoch": 4.261660978384528,
+      "grad_norm": 0.29353899348561113,
+      "learning_rate": 2.56662112185635e-06,
+      "loss": 0.1565,
+      "step": 1873
+    },
+    {
+      "epoch": 4.263936291240046,
+      "grad_norm": 0.30462610206364243,
+      "learning_rate": 2.5510514947169474e-06,
+      "loss": 0.1805,
+      "step": 1874
+    },
+    {
+      "epoch": 4.266211604095563,
+      "grad_norm": 0.27048278700046385,
+      "learning_rate": 2.535526017984884e-06,
+      "loss": 0.1596,
+      "step": 1875
+    },
+    {
+      "epoch": 4.268486916951081,
+      "grad_norm": 0.3074253140038664,
+      "learning_rate": 2.520044730943687e-06,
+      "loss": 0.1681,
+      "step": 1876
+    },
+    {
+      "epoch": 4.270762229806598,
+      "grad_norm": 0.2965521216090326,
+      "learning_rate": 2.5046076727650647e-06,
+      "loss": 0.1677,
+      "step": 1877
+    },
+    {
+      "epoch": 4.273037542662116,
+      "grad_norm": 0.28241542497027616,
+      "learning_rate": 2.4892148825088167e-06,
+      "loss": 0.1736,
+      "step": 1878
+    },
+    {
+      "epoch": 4.275312855517634,
+      "grad_norm": 0.2910565396763097,
+      "learning_rate": 2.473866399122733e-06,
+      "loss": 0.2,
+      "step": 1879
+    },
+    {
+      "epoch": 4.277588168373152,
+      "grad_norm": 0.26706170621906083,
+      "learning_rate": 2.458562261442483e-06,
+      "loss": 0.1633,
+      "step": 1880
+    },
+    {
+      "epoch": 4.279863481228669,
+      "grad_norm": 0.337322390050979,
+      "learning_rate": 2.4433025081915516e-06,
+      "loss": 0.1909,
+      "step": 1881
+    },
+    {
+      "epoch": 4.282138794084187,
+      "grad_norm": 0.29527568345775845,
+      "learning_rate": 2.428087177981109e-06,
+      "loss": 0.1621,
+      "step": 1882
+    },
+    {
+      "epoch": 4.284414106939704,
+      "grad_norm": 0.2772630547807417,
+      "learning_rate": 2.4129163093099183e-06,
+      "loss": 0.1722,
+      "step": 1883
+    },
+    {
+      "epoch": 4.286689419795222,
+      "grad_norm": 0.29035733460933594,
+      "learning_rate": 2.3977899405642502e-06,
+      "loss": 0.1331,
+      "step": 1884
+    },
+    {
+      "epoch": 4.288964732650739,
+      "grad_norm": 0.28924669219995647,
+      "learning_rate": 2.3827081100177797e-06,
+      "loss": 0.1552,
+      "step": 1885
+    },
+    {
+      "epoch": 4.291240045506257,
+      "grad_norm": 0.293749628581807,
+      "learning_rate": 2.36767085583149e-06,
+      "loss": 0.1877,
+      "step": 1886
+    },
+    {
+      "epoch": 4.293515358361775,
+      "grad_norm": 0.29530983346945033,
+      "learning_rate": 2.352678216053563e-06,
+      "loss": 0.171,
+      "step": 1887
+    },
+    {
+      "epoch": 4.295790671217293,
+      "grad_norm": 0.29070568416114295,
+      "learning_rate": 2.337730228619297e-06,
+      "loss": 0.2007,
+      "step": 1888
+    },
+    {
+      "epoch": 4.29806598407281,
+      "grad_norm": 0.28420095387571376,
+      "learning_rate": 2.322826931351028e-06,
+      "loss": 0.1502,
+      "step": 1889
+    },
+    {
+      "epoch": 4.300341296928328,
+      "grad_norm": 0.3047011945298094,
+      "learning_rate": 2.307968361957993e-06,
+      "loss": 0.1593,
+      "step": 1890
+    },
+    {
+      "epoch": 4.302616609783845,
+      "grad_norm": 0.2735842227975423,
+      "learning_rate": 2.293154558036257e-06,
+      "loss": 0.1705,
+      "step": 1891
+    },
+    {
+      "epoch": 4.304891922639363,
+      "grad_norm": 0.3006310279420523,
+      "learning_rate": 2.278385557068623e-06,
+      "loss": 0.1888,
+      "step": 1892
+    },
+    {
+      "epoch": 4.30716723549488,
+      "grad_norm": 0.2839083760791352,
+      "learning_rate": 2.2636613964245348e-06,
+      "loss": 0.1623,
+      "step": 1893
+    },
+    {
+      "epoch": 4.309442548350399,
+      "grad_norm": 0.2773286092711842,
+      "learning_rate": 2.2489821133599744e-06,
+      "loss": 0.1467,
+      "step": 1894
+    },
+    {
+      "epoch": 4.311717861205916,
+      "grad_norm": 0.2830527267043027,
+      "learning_rate": 2.2343477450173665e-06,
+      "loss": 0.1524,
+      "step": 1895
+    },
+    {
+      "epoch": 4.313993174061434,
+      "grad_norm": 0.29987047576137166,
+      "learning_rate": 2.2197583284254853e-06,
+      "loss": 0.1648,
+      "step": 1896
+    },
+    {
+      "epoch": 4.316268486916951,
+      "grad_norm": 0.29598096279832914,
+      "learning_rate": 2.205213900499392e-06,
+      "loss": 0.1902,
+      "step": 1897
+    },
+    {
+      "epoch": 4.318543799772469,
+      "grad_norm": 0.28814294396320594,
+      "learning_rate": 2.1907144980402852e-06,
+      "loss": 0.1644,
+      "step": 1898
+    },
+    {
+      "epoch": 4.320819112627986,
+      "grad_norm": 0.301351812170094,
+      "learning_rate": 2.1762601577354482e-06,
+      "loss": 0.189,
+      "step": 1899
+    },
+    {
+      "epoch": 4.323094425483504,
+      "grad_norm": 0.3018885038085118,
+      "learning_rate": 2.161850916158148e-06,
+      "loss": 0.1699,
+      "step": 1900
+    },
+    {
+      "epoch": 4.325369738339021,
+      "grad_norm": 0.2869854738978282,
+      "learning_rate": 2.147486809767538e-06,
+      "loss": 0.1773,
+      "step": 1901
+    },
+    {
+      "epoch": 4.327645051194539,
+      "grad_norm": 0.28405930271661783,
+      "learning_rate": 2.1331678749085706e-06,
+      "loss": 0.173,
+      "step": 1902
+    },
+    {
+      "epoch": 4.329920364050057,
+      "grad_norm": 0.28386739272398476,
+      "learning_rate": 2.118894147811896e-06,
+      "loss": 0.2033,
+      "step": 1903
+    },
+    {
+      "epoch": 4.332195676905575,
+      "grad_norm": 0.285225964317923,
+      "learning_rate": 2.1046656645937725e-06,
+      "loss": 0.1947,
+      "step": 1904
+    },
+    {
+      "epoch": 4.334470989761092,
+      "grad_norm": 0.2730586271912445,
+      "learning_rate": 2.0904824612560046e-06,
+      "loss": 0.1643,
+      "step": 1905
+    },
+    {
+      "epoch": 4.33674630261661,
+      "grad_norm": 0.28470372889885964,
+      "learning_rate": 2.076344573685798e-06,
+      "loss": 0.1842,
+      "step": 1906
+    },
+    {
+      "epoch": 4.339021615472127,
+      "grad_norm": 0.2896583206658402,
+      "learning_rate": 2.0622520376557074e-06,
+      "loss": 0.2125,
+      "step": 1907
+    },
+    {
+      "epoch": 4.341296928327645,
+      "grad_norm": 0.28492206440633683,
+      "learning_rate": 2.048204888823542e-06,
+      "loss": 0.1766,
+      "step": 1908
+    },
+    {
+      "epoch": 4.343572241183162,
+      "grad_norm": 0.2776716747062078,
+      "learning_rate": 2.0342031627322644e-06,
+      "loss": 0.1571,
+      "step": 1909
+    },
+    {
+      "epoch": 4.345847554038681,
+      "grad_norm": 0.30420878779082033,
+      "learning_rate": 2.020246894809912e-06,
+      "loss": 0.1809,
+      "step": 1910
+    },
+    {
+      "epoch": 4.348122866894198,
+      "grad_norm": 0.29233315451399383,
+      "learning_rate": 2.0063361203694897e-06,
+      "loss": 0.192,
+      "step": 1911
+    },
+    {
+      "epoch": 4.350398179749716,
+      "grad_norm": 0.29427195621230223,
+      "learning_rate": 1.992470874608894e-06,
+      "loss": 0.1832,
+      "step": 1912
+    },
+    {
+      "epoch": 4.352673492605233,
+      "grad_norm": 0.28092217612998305,
+      "learning_rate": 1.978651192610841e-06,
+      "loss": 0.1687,
+      "step": 1913
+    },
+    {
+      "epoch": 4.354948805460751,
+      "grad_norm": 0.3008121187898929,
+      "learning_rate": 1.964877109342733e-06,
+      "loss": 0.1919,
+      "step": 1914
+    },
+    {
+      "epoch": 4.357224118316268,
+      "grad_norm": 0.285257260726303,
+      "learning_rate": 1.9511486596566054e-06,
+      "loss": 0.1717,
+      "step": 1915
+    },
+    {
+      "epoch": 4.359499431171786,
+      "grad_norm": 0.28989930245491863,
+      "learning_rate": 1.9374658782890332e-06,
+      "loss": 0.1732,
+      "step": 1916
+    },
+    {
+      "epoch": 4.361774744027303,
+      "grad_norm": 0.2903557658703925,
+      "learning_rate": 1.923828799861034e-06,
+      "loss": 0.1571,
+      "step": 1917
+    },
+    {
+      "epoch": 4.364050056882822,
+      "grad_norm": 0.2800855565851557,
+      "learning_rate": 1.9102374588779883e-06,
+      "loss": 0.1478,
+      "step": 1918
+    },
+    {
+      "epoch": 4.366325369738339,
+      "grad_norm": 0.27784082913006314,
+      "learning_rate": 1.8966918897295472e-06,
+      "loss": 0.1556,
+      "step": 1919
+    },
+    {
+      "epoch": 4.368600682593857,
+      "grad_norm": 0.2765254308847868,
+      "learning_rate": 1.8831921266895348e-06,
+      "loss": 0.2131,
+      "step": 1920
+    },
+    {
+      "epoch": 4.370875995449374,
+      "grad_norm": 0.3047570532309943,
+      "learning_rate": 1.869738203915903e-06,
+      "loss": 0.1765,
+      "step": 1921
+    },
+    {
+      "epoch": 4.373151308304892,
+      "grad_norm": 0.28003907661446176,
+      "learning_rate": 1.8563301554505876e-06,
+      "loss": 0.1629,
+      "step": 1922
+    },
+    {
+      "epoch": 4.375426621160409,
+      "grad_norm": 0.2970382076063424,
+      "learning_rate": 1.8429680152194619e-06,
+      "loss": 0.1729,
+      "step": 1923
+    },
+    {
+      "epoch": 4.377701934015927,
+      "grad_norm": 0.2925407037838341,
+      "learning_rate": 1.8296518170322409e-06,
+      "loss": 0.1792,
+      "step": 1924
+    },
+    {
+      "epoch": 4.379977246871444,
+      "grad_norm": 0.3026444999079674,
+      "learning_rate": 1.8163815945823881e-06,
+      "loss": 0.1783,
+      "step": 1925
+    },
+    {
+      "epoch": 4.382252559726963,
+      "grad_norm": 0.2778803748916032,
+      "learning_rate": 1.803157381447047e-06,
+      "loss": 0.1694,
+      "step": 1926
+    },
+    {
+      "epoch": 4.38452787258248,
+      "grad_norm": 0.28888117266682606,
+      "learning_rate": 1.7899792110869364e-06,
+      "loss": 0.1285,
+      "step": 1927
+    },
+    {
+      "epoch": 4.386803185437998,
+      "grad_norm": 0.2885464253312428,
+      "learning_rate": 1.7768471168462652e-06,
+      "loss": 0.1823,
+      "step": 1928
+    },
+    {
+      "epoch": 4.389078498293515,
+      "grad_norm": 0.28847188145684455,
+      "learning_rate": 1.7637611319526881e-06,
+      "loss": 0.2026,
+      "step": 1929
+    },
+    {
+      "epoch": 4.391353811149033,
+      "grad_norm": 0.2793182449132348,
+      "learning_rate": 1.7507212895171632e-06,
+      "loss": 0.1746,
+      "step": 1930
+    },
+    {
+      "epoch": 4.39362912400455,
+      "grad_norm": 0.29292601622964687,
+      "learning_rate": 1.7377276225339e-06,
+      "loss": 0.1806,
+      "step": 1931
+    },
+    {
+      "epoch": 4.395904436860068,
+      "grad_norm": 0.28209452442347205,
+      "learning_rate": 1.7247801638802842e-06,
+      "loss": 0.1795,
+      "step": 1932
+    },
+    {
+      "epoch": 4.398179749715586,
+      "grad_norm": 0.28464106504951486,
+      "learning_rate": 1.711878946316774e-06,
+      "loss": 0.1968,
+      "step": 1933
+    },
+    {
+      "epoch": 4.400455062571104,
+      "grad_norm": 0.2772080474047025,
+      "learning_rate": 1.699024002486831e-06,
+      "loss": 0.1856,
+      "step": 1934
+    },
+    {
+      "epoch": 4.402730375426621,
+      "grad_norm": 0.265831728673992,
+      "learning_rate": 1.6862153649168211e-06,
+      "loss": 0.1552,
+      "step": 1935
+    },
+    {
+      "epoch": 4.405005688282139,
+      "grad_norm": 0.2787073981628462,
+      "learning_rate": 1.6734530660159532e-06,
+      "loss": 0.1398,
+      "step": 1936
+    },
+    {
+      "epoch": 4.407281001137656,
+      "grad_norm": 0.2841532952793566,
+      "learning_rate": 1.660737138076187e-06,
+      "loss": 0.1535,
+      "step": 1937
+    },
+    {
+      "epoch": 4.409556313993174,
+      "grad_norm": 0.30484118224343415,
+      "learning_rate": 1.6480676132721484e-06,
+      "loss": 0.1788,
+      "step": 1938
+    },
+    {
+      "epoch": 4.4118316268486915,
+      "grad_norm": 0.2780969954617563,
+      "learning_rate": 1.635444523661045e-06,
+      "loss": 0.1804,
+      "step": 1939
+    },
+    {
+      "epoch": 4.414106939704209,
+      "grad_norm": 0.2805039981184527,
+      "learning_rate": 1.6228679011826032e-06,
+      "loss": 0.1641,
+      "step": 1940
+    },
+    {
+      "epoch": 4.4163822525597265,
+      "grad_norm": 0.2955433059424568,
+      "learning_rate": 1.610337777658968e-06,
+      "loss": 0.1741,
+      "step": 1941
+    },
+    {
+      "epoch": 4.418657565415245,
+      "grad_norm": 0.29651803614171446,
+      "learning_rate": 1.5978541847946383e-06,
+      "loss": 0.1738,
+      "step": 1942
+    },
+    {
+      "epoch": 4.420932878270762,
+      "grad_norm": 0.2865596574026191,
+      "learning_rate": 1.585417154176363e-06,
+      "loss": 0.168,
+      "step": 1943
+    },
+    {
+      "epoch": 4.42320819112628,
+      "grad_norm": 0.2715688446862325,
+      "learning_rate": 1.5730267172730873e-06,
+      "loss": 0.1518,
+      "step": 1944
+    },
+    {
+      "epoch": 4.4254835039817975,
+      "grad_norm": 0.27806911710441523,
+      "learning_rate": 1.5606829054358686e-06,
+      "loss": 0.1952,
+      "step": 1945
+    },
+    {
+      "epoch": 4.427758816837315,
+      "grad_norm": 0.29787549728839613,
+      "learning_rate": 1.548385749897774e-06,
+      "loss": 0.1902,
+      "step": 1946
+    },
+    {
+      "epoch": 4.4300341296928325,
+      "grad_norm": 0.28023191525733876,
+      "learning_rate": 1.5361352817738228e-06,
+      "loss": 0.1722,
+      "step": 1947
+    },
+    {
+      "epoch": 4.43230944254835,
+      "grad_norm": 0.28967761875456094,
+      "learning_rate": 1.5239315320609116e-06,
+      "loss": 0.1619,
+      "step": 1948
+    },
+    {
+      "epoch": 4.434584755403868,
+      "grad_norm": 0.2721067778091864,
+      "learning_rate": 1.511774531637722e-06,
+      "loss": 0.172,
+      "step": 1949
+    },
+    {
+      "epoch": 4.436860068259386,
+      "grad_norm": 0.305432121933115,
+      "learning_rate": 1.499664311264648e-06,
+      "loss": 0.1689,
+      "step": 1950
+    },
+    {
+      "epoch": 4.4391353811149035,
+      "grad_norm": 0.26894904879069875,
+      "learning_rate": 1.4876009015837124e-06,
+      "loss": 0.2014,
+      "step": 1951
+    },
+    {
+      "epoch": 4.441410693970421,
+      "grad_norm": 0.27691927726903726,
+      "learning_rate": 1.475584333118505e-06,
+      "loss": 0.1716,
+      "step": 1952
+    },
+    {
+      "epoch": 4.4436860068259385,
+      "grad_norm": 0.27467787878431865,
+      "learning_rate": 1.4636146362740888e-06,
+      "loss": 0.1682,
+      "step": 1953
+    },
+    {
+      "epoch": 4.445961319681456,
+      "grad_norm": 0.2872238149705631,
+      "learning_rate": 1.4516918413369285e-06,
+      "loss": 0.1541,
+      "step": 1954
+    },
+    {
+      "epoch": 4.4482366325369735,
+      "grad_norm": 0.29846788798446056,
+      "learning_rate": 1.4398159784748144e-06,
+      "loss": 0.1537,
+      "step": 1955
+    },
+    {
+      "epoch": 4.450511945392491,
+      "grad_norm": 0.2908559333263553,
+      "learning_rate": 1.4279870777367899e-06,
+      "loss": 0.2251,
+      "step": 1956
+    },
+    {
+      "epoch": 4.4527872582480095,
+      "grad_norm": 0.29080353198439524,
+      "learning_rate": 1.416205169053071e-06,
+      "loss": 0.1673,
+      "step": 1957
+    },
+    {
+      "epoch": 4.455062571103527,
+      "grad_norm": 0.2783122256879772,
+      "learning_rate": 1.4044702822349731e-06,
+      "loss": 0.1829,
+      "step": 1958
+    },
+    {
+      "epoch": 4.4573378839590445,
+      "grad_norm": 0.29610239322941884,
+      "learning_rate": 1.3927824469748274e-06,
+      "loss": 0.181,
+      "step": 1959
+    },
+    {
+      "epoch": 4.459613196814562,
+      "grad_norm": 0.29321444810574243,
+      "learning_rate": 1.3811416928459177e-06,
+      "loss": 0.178,
+      "step": 1960
+    },
+    {
+      "epoch": 4.4618885096700796,
+      "grad_norm": 0.2792011932790621,
+      "learning_rate": 1.3695480493024027e-06,
+      "loss": 0.1658,
+      "step": 1961
+    },
+    {
+      "epoch": 4.464163822525597,
+      "grad_norm": 0.2808332313261231,
+      "learning_rate": 1.358001545679235e-06,
+      "loss": 0.1803,
+      "step": 1962
+    },
+    {
+      "epoch": 4.466439135381115,
+      "grad_norm": 0.3074530335022185,
+      "learning_rate": 1.3465022111920823e-06,
+      "loss": 0.1732,
+      "step": 1963
+    },
+    {
+      "epoch": 4.468714448236632,
+      "grad_norm": 0.28964736313004563,
+      "learning_rate": 1.3350500749372808e-06,
+      "loss": 0.141,
+      "step": 1964
+    },
+    {
+      "epoch": 4.4709897610921505,
+      "grad_norm": 0.2719958918001761,
+      "learning_rate": 1.3236451658917293e-06,
+      "loss": 0.1976,
+      "step": 1965
+    },
+    {
+      "epoch": 4.473265073947668,
+      "grad_norm": 0.2924179782646019,
+      "learning_rate": 1.3122875129128355e-06,
+      "loss": 0.1745,
+      "step": 1966
+    },
+    {
+      "epoch": 4.4755403868031856,
+      "grad_norm": 0.2996967323657901,
+      "learning_rate": 1.3009771447384313e-06,
+      "loss": 0.1666,
+      "step": 1967
+    },
+    {
+      "epoch": 4.477815699658703,
+      "grad_norm": 0.25889338362216696,
+      "learning_rate": 1.289714089986711e-06,
+      "loss": 0.1612,
+      "step": 1968
+    },
+    {
+      "epoch": 4.480091012514221,
+      "grad_norm": 0.27753924235718963,
+      "learning_rate": 1.2784983771561544e-06,
+      "loss": 0.186,
+      "step": 1969
+    },
+    {
+      "epoch": 4.482366325369738,
+      "grad_norm": 0.29480688552138207,
+      "learning_rate": 1.2673300346254447e-06,
+      "loss": 0.1824,
+      "step": 1970
+    },
+    {
+      "epoch": 4.484641638225256,
+      "grad_norm": 0.29040361553018085,
+      "learning_rate": 1.2562090906534153e-06,
+      "loss": 0.2132,
+      "step": 1971
+    },
+    {
+      "epoch": 4.486916951080774,
+      "grad_norm": 0.27954190765759973,
+      "learning_rate": 1.2451355733789616e-06,
+      "loss": 0.1869,
+      "step": 1972
+    },
+    {
+      "epoch": 4.489192263936292,
+      "grad_norm": 0.2805230943656468,
+      "learning_rate": 1.2341095108209866e-06,
+      "loss": 0.1753,
+      "step": 1973
+    },
+    {
+      "epoch": 4.491467576791809,
+      "grad_norm": 0.29244508925798035,
+      "learning_rate": 1.2231309308783157e-06,
+      "loss": 0.1506,
+      "step": 1974
+    },
+    {
+      "epoch": 4.493742889647327,
+      "grad_norm": 0.2787002830771308,
+      "learning_rate": 1.2121998613296259e-06,
+      "loss": 0.1879,
+      "step": 1975
+    },
+    {
+      "epoch": 4.496018202502844,
+      "grad_norm": 0.2941140207001762,
+      "learning_rate": 1.2013163298333842e-06,
+      "loss": 0.1797,
+      "step": 1976
+    },
+    {
+      "epoch": 4.498293515358362,
+      "grad_norm": 0.27528088765213415,
+      "learning_rate": 1.1904803639277773e-06,
+      "loss": 0.1939,
+      "step": 1977
+    },
+    {
+      "epoch": 4.500568828213879,
+      "grad_norm": 0.27949573204598344,
+      "learning_rate": 1.1796919910306359e-06,
+      "loss": 0.1935,
+      "step": 1978
+    },
+    {
+      "epoch": 4.502844141069397,
+      "grad_norm": 0.29184060626111324,
+      "learning_rate": 1.1689512384393597e-06,
+      "loss": 0.172,
+      "step": 1979
+    },
+    {
+      "epoch": 4.505119453924914,
+      "grad_norm": 0.28235052123789983,
+      "learning_rate": 1.1582581333308784e-06,
+      "loss": 0.171,
+      "step": 1980
+    },
+    {
+      "epoch": 4.507394766780433,
+      "grad_norm": 0.294521787307336,
+      "learning_rate": 1.1476127027615336e-06,
+      "loss": 0.1598,
+      "step": 1981
+    },
+    {
+      "epoch": 4.50967007963595,
+      "grad_norm": 0.2849206624146031,
+      "learning_rate": 1.1370149736670589e-06,
+      "loss": 0.1771,
+      "step": 1982
+    },
+    {
+      "epoch": 4.511945392491468,
+      "grad_norm": 0.2819395265085714,
+      "learning_rate": 1.12646497286248e-06,
+      "loss": 0.1443,
+      "step": 1983
+    },
+    {
+      "epoch": 4.514220705346985,
+      "grad_norm": 0.2839388508662871,
+      "learning_rate": 1.1159627270420615e-06,
+      "loss": 0.172,
+      "step": 1984
+    },
+    {
+      "epoch": 4.516496018202503,
+      "grad_norm": 0.28864243658579875,
+      "learning_rate": 1.1055082627792357e-06,
+      "loss": 0.1547,
+      "step": 1985
+    },
+    {
+      "epoch": 4.51877133105802,
+      "grad_norm": 0.30765219362613433,
+      "learning_rate": 1.0951016065265341e-06,
+      "loss": 0.1648,
+      "step": 1986
+    },
+    {
+      "epoch": 4.521046643913538,
+      "grad_norm": 0.26711896993626416,
+      "learning_rate": 1.0847427846155157e-06,
+      "loss": 0.1841,
+      "step": 1987
+    },
+    {
+      "epoch": 4.523321956769056,
+      "grad_norm": 0.27032220055464995,
+      "learning_rate": 1.0744318232567185e-06,
+      "loss": 0.1601,
+      "step": 1988
+    },
+    {
+      "epoch": 4.525597269624574,
+      "grad_norm": 0.2821744851476248,
+      "learning_rate": 1.0641687485395691e-06,
+      "loss": 0.1883,
+      "step": 1989
+    },
+    {
+      "epoch": 4.527872582480091,
+      "grad_norm": 0.7871524095136547,
+      "learning_rate": 1.0539535864323391e-06,
+      "loss": 0.1749,
+      "step": 1990
+    },
+    {
+      "epoch": 4.530147895335609,
+      "grad_norm": 0.29128322927257533,
+      "learning_rate": 1.0437863627820555e-06,
+      "loss": 0.1494,
+      "step": 1991
+    },
+    {
+      "epoch": 4.532423208191126,
+      "grad_norm": 0.2817846298238865,
+      "learning_rate": 1.0336671033144573e-06,
+      "loss": 0.1558,
+      "step": 1992
+    },
+    {
+      "epoch": 4.534698521046644,
+      "grad_norm": 0.2787263682512981,
+      "learning_rate": 1.0235958336339259e-06,
+      "loss": 0.1503,
+      "step": 1993
+    },
+    {
+      "epoch": 4.536973833902161,
+      "grad_norm": 0.30180138304813886,
+      "learning_rate": 1.0135725792234052e-06,
+      "loss": 0.1737,
+      "step": 1994
+    },
+    {
+      "epoch": 4.53924914675768,
+      "grad_norm": 0.3026589362256529,
+      "learning_rate": 1.0035973654443466e-06,
+      "loss": 0.1546,
+      "step": 1995
+    },
+    {
+      "epoch": 4.541524459613197,
+      "grad_norm": 0.2808793452470622,
+      "learning_rate": 9.936702175366642e-07,
+      "loss": 0.1769,
+      "step": 1996
+    },
+    {
+      "epoch": 4.543799772468715,
+      "grad_norm": 0.28062446297994786,
+      "learning_rate": 9.837911606186323e-07,
+      "loss": 0.1908,
+      "step": 1997
+    },
+    {
+      "epoch": 4.546075085324232,
+      "grad_norm": 0.283034769731874,
+      "learning_rate": 9.739602196868558e-07,
+      "loss": 0.1672,
+      "step": 1998
+    },
+    {
+      "epoch": 4.54835039817975,
+      "grad_norm": 0.26885302611911016,
+      "learning_rate": 9.641774196161836e-07,
+      "loss": 0.1747,
+      "step": 1999
+    },
+    {
+      "epoch": 4.550625711035267,
+      "grad_norm": 0.28787602733944306,
+      "learning_rate": 9.544427851596661e-07,
+      "loss": 0.163,
+      "step": 2000
+    },
+    {
+      "epoch": 4.552901023890785,
+      "grad_norm": 0.2763668454393897,
+      "learning_rate": 9.447563409484739e-07,
+      "loss": 0.2064,
+      "step": 2001
+    },
+    {
+      "epoch": 4.555176336746302,
+      "grad_norm": 0.2788164498517762,
+      "learning_rate": 9.351181114918506e-07,
+      "loss": 0.1724,
+      "step": 2002
+    },
+    {
+      "epoch": 4.55745164960182,
+      "grad_norm": 0.28335342939031766,
+      "learning_rate": 9.255281211770284e-07,
+      "loss": 0.1791,
+      "step": 2003
+    },
+    {
+      "epoch": 4.559726962457338,
+      "grad_norm": 0.2650324119047187,
+      "learning_rate": 9.15986394269206e-07,
+      "loss": 0.1311,
+      "step": 2004
+    },
+    {
+      "epoch": 4.562002275312856,
+      "grad_norm": 0.2765044538049791,
+      "learning_rate": 9.064929549114421e-07,
+      "loss": 0.1539,
+      "step": 2005
+    },
+    {
+      "epoch": 4.564277588168373,
+      "grad_norm": 0.2821192439754218,
+      "learning_rate": 8.970478271246307e-07,
+      "loss": 0.1905,
+      "step": 2006
+    },
+    {
+      "epoch": 4.566552901023891,
+      "grad_norm": 0.2902969512335743,
+      "learning_rate": 8.876510348074108e-07,
+      "loss": 0.1787,
+      "step": 2007
+    },
+    {
+      "epoch": 4.568828213879408,
+      "grad_norm": 0.288015139227802,
+      "learning_rate": 8.783026017361274e-07,
+      "loss": 0.1836,
+      "step": 2008
+    },
+    {
+      "epoch": 4.571103526734926,
+      "grad_norm": 0.29975893178963964,
+      "learning_rate": 8.690025515647682e-07,
+      "loss": 0.1718,
+      "step": 2009
+    },
+    {
+      "epoch": 4.573378839590443,
+      "grad_norm": 0.2762337937692315,
+      "learning_rate": 8.597509078248923e-07,
+      "loss": 0.1584,
+      "step": 2010
+    },
+    {
+      "epoch": 4.575654152445962,
+      "grad_norm": 0.27817524325145293,
+      "learning_rate": 8.505476939255741e-07,
+      "loss": 0.1739,
+      "step": 2011
+    },
+    {
+      "epoch": 4.577929465301479,
+      "grad_norm": 0.29507089156780547,
+      "learning_rate": 8.413929331533643e-07,
+      "loss": 0.1726,
+      "step": 2012
+    },
+    {
+      "epoch": 4.580204778156997,
+      "grad_norm": 0.279622887867422,
+      "learning_rate": 8.322866486721959e-07,
+      "loss": 0.1659,
+      "step": 2013
+    },
+    {
+      "epoch": 4.582480091012514,
+      "grad_norm": 0.2752351248049167,
+      "learning_rate": 8.232288635233598e-07,
+      "loss": 0.2172,
+      "step": 2014
+    },
+    {
+      "epoch": 4.584755403868032,
+      "grad_norm": 0.3000052844155821,
+      "learning_rate": 8.142196006254144e-07,
+      "loss": 0.1732,
+      "step": 2015
+    },
+    {
+      "epoch": 4.587030716723549,
+      "grad_norm": 0.2846970904267727,
+      "learning_rate": 8.052588827741603e-07,
+      "loss": 0.156,
+      "step": 2016
+    },
+    {
+      "epoch": 4.589306029579067,
+      "grad_norm": 0.28172746152505274,
+      "learning_rate": 7.963467326425567e-07,
+      "loss": 0.2044,
+      "step": 2017
+    },
+    {
+      "epoch": 4.591581342434584,
+      "grad_norm": 0.29195791070167354,
+      "learning_rate": 7.874831727806764e-07,
+      "loss": 0.1706,
+      "step": 2018
+    },
+    {
+      "epoch": 4.593856655290102,
+      "grad_norm": 0.2659143548975755,
+      "learning_rate": 7.786682256156464e-07,
+      "loss": 0.1806,
+      "step": 2019
+    },
+    {
+      "epoch": 4.59613196814562,
+      "grad_norm": 0.2650641816698844,
+      "learning_rate": 7.699019134515917e-07,
+      "loss": 0.1478,
+      "step": 2020
+    },
+    {
+      "epoch": 4.598407281001138,
+      "grad_norm": 0.2705433799082718,
+      "learning_rate": 7.611842584695739e-07,
+      "loss": 0.1766,
+      "step": 2021
+    },
+    {
+      "epoch": 4.600682593856655,
+      "grad_norm": 0.2933000010470503,
+      "learning_rate": 7.525152827275462e-07,
+      "loss": 0.1501,
+      "step": 2022
+    },
+    {
+      "epoch": 4.602957906712173,
+      "grad_norm": 0.27973283445352826,
+      "learning_rate": 7.438950081602824e-07,
+      "loss": 0.1658,
+      "step": 2023
+    },
+    {
+      "epoch": 4.60523321956769,
+      "grad_norm": 0.28510624819401426,
+      "learning_rate": 7.353234565793332e-07,
+      "loss": 0.1804,
+      "step": 2024
+    },
+    {
+      "epoch": 4.607508532423208,
+      "grad_norm": 0.28306854477557686,
+      "learning_rate": 7.268006496729762e-07,
+      "loss": 0.1963,
+      "step": 2025
+    },
+    {
+      "epoch": 4.609783845278725,
+      "grad_norm": 0.295611257023681,
+      "learning_rate": 7.183266090061324e-07,
+      "loss": 0.1549,
+      "step": 2026
+    },
+    {
+      "epoch": 4.612059158134244,
+      "grad_norm": 0.28424148773989993,
+      "learning_rate": 7.099013560203505e-07,
+      "loss": 0.1492,
+      "step": 2027
+    },
+    {
+      "epoch": 4.614334470989761,
+      "grad_norm": 0.28015465791620475,
+      "learning_rate": 7.015249120337242e-07,
+      "loss": 0.1923,
+      "step": 2028
+    },
+    {
+      "epoch": 4.616609783845279,
+      "grad_norm": 0.28384048033951154,
+      "learning_rate": 6.931972982408486e-07,
+      "loss": 0.1859,
+      "step": 2029
+    },
+    {
+      "epoch": 4.618885096700796,
+      "grad_norm": 0.2722259762819922,
+      "learning_rate": 6.849185357127686e-07,
+      "loss": 0.1628,
+      "step": 2030
+    },
+    {
+      "epoch": 4.621160409556314,
+      "grad_norm": 0.3079090242998302,
+      "learning_rate": 6.766886453969168e-07,
+      "loss": 0.1741,
+      "step": 2031
+    },
+    {
+      "epoch": 4.623435722411831,
+      "grad_norm": 0.27119843108967856,
+      "learning_rate": 6.685076481170715e-07,
+      "loss": 0.1691,
+      "step": 2032
+    },
+    {
+      "epoch": 4.625711035267349,
+      "grad_norm": 0.28656343897509196,
+      "learning_rate": 6.603755645733012e-07,
+      "loss": 0.1762,
+      "step": 2033
+    },
+    {
+      "epoch": 4.627986348122867,
+      "grad_norm": 0.28516364364971947,
+      "learning_rate": 6.522924153419019e-07,
+      "loss": 0.1544,
+      "step": 2034
+    },
+    {
+      "epoch": 4.630261660978385,
+      "grad_norm": 0.2678058152557827,
+      "learning_rate": 6.442582208753578e-07,
+      "loss": 0.1607,
+      "step": 2035
+    },
+    {
+      "epoch": 4.632536973833902,
+      "grad_norm": 0.2818467878057201,
+      "learning_rate": 6.362730015022855e-07,
+      "loss": 0.1781,
+      "step": 2036
+    },
+    {
+      "epoch": 4.63481228668942,
+      "grad_norm": 0.28705706508095613,
+      "learning_rate": 6.283367774273785e-07,
+      "loss": 0.1478,
+      "step": 2037
+    },
+    {
+      "epoch": 4.637087599544937,
+      "grad_norm": 0.2731517221285001,
+      "learning_rate": 6.204495687313627e-07,
+      "loss": 0.1695,
+      "step": 2038
+    },
+    {
+      "epoch": 4.639362912400455,
+      "grad_norm": 0.28683037296597896,
+      "learning_rate": 6.126113953709389e-07,
+      "loss": 0.1646,
+      "step": 2039
+    },
+    {
+      "epoch": 4.6416382252559725,
+      "grad_norm": 0.27722178521797003,
+      "learning_rate": 6.048222771787382e-07,
+      "loss": 0.1819,
+      "step": 2040
+    },
+    {
+      "epoch": 4.64391353811149,
+      "grad_norm": 0.2903974402660376,
+      "learning_rate": 5.970822338632709e-07,
+      "loss": 0.1662,
+      "step": 2041
+    },
+    {
+      "epoch": 4.6461888509670075,
+      "grad_norm": 0.2868561185405596,
+      "learning_rate": 5.893912850088668e-07,
+      "loss": 0.1593,
+      "step": 2042
+    },
+    {
+      "epoch": 4.648464163822526,
+      "grad_norm": 0.27903422660884797,
+      "learning_rate": 5.81749450075646e-07,
+      "loss": 0.1709,
+      "step": 2043
+    },
+    {
+      "epoch": 4.650739476678043,
+      "grad_norm": 0.2846468286225907,
+      "learning_rate": 5.741567483994503e-07,
+      "loss": 0.1894,
+      "step": 2044
+    },
+    {
+      "epoch": 4.653014789533561,
+      "grad_norm": 0.29921985053486083,
+      "learning_rate": 5.666131991917989e-07,
+      "loss": 0.183,
+      "step": 2045
+    },
+    {
+      "epoch": 4.6552901023890785,
+      "grad_norm": 0.2705631695784153,
+      "learning_rate": 5.591188215398524e-07,
+      "loss": 0.1637,
+      "step": 2046
+    },
+    {
+      "epoch": 4.657565415244596,
+      "grad_norm": 0.2804359500444224,
+      "learning_rate": 5.5167363440634e-07,
+      "loss": 0.1795,
+      "step": 2047
+    },
+    {
+      "epoch": 4.6598407281001135,
+      "grad_norm": 0.27327279446058655,
+      "learning_rate": 5.442776566295393e-07,
+      "loss": 0.1446,
+      "step": 2048
+    },
+    {
+      "epoch": 4.662116040955631,
+      "grad_norm": 0.2766472513272786,
+      "learning_rate": 5.369309069232098e-07,
+      "loss": 0.1671,
+      "step": 2049
+    },
+    {
+      "epoch": 4.664391353811149,
+      "grad_norm": 0.27251263166270223,
+      "learning_rate": 5.296334038765483e-07,
+      "loss": 0.164,
+      "step": 2050
+    },
+    {
+      "epoch": 4.666666666666667,
+      "grad_norm": 0.28798270256549147,
+      "learning_rate": 5.22385165954149e-07,
+      "loss": 0.172,
+      "step": 2051
+    },
+    {
+      "epoch": 4.6689419795221845,
+      "grad_norm": 0.29135865846989056,
+      "learning_rate": 5.151862114959572e-07,
+      "loss": 0.1911,
+      "step": 2052
+    },
+    {
+      "epoch": 4.671217292377702,
+      "grad_norm": 0.30326455275234954,
+      "learning_rate": 5.080365587172021e-07,
+      "loss": 0.1675,
+      "step": 2053
+    },
+    {
+      "epoch": 4.6734926052332195,
+      "grad_norm": 0.30492590779764683,
+      "learning_rate": 5.009362257083861e-07,
+      "loss": 0.1655,
+      "step": 2054
+    },
+    {
+      "epoch": 4.675767918088737,
+      "grad_norm": 0.27971636539782796,
+      "learning_rate": 4.938852304352026e-07,
+      "loss": 0.1789,
+      "step": 2055
+    },
+    {
+      "epoch": 4.678043230944255,
+      "grad_norm": 0.27888012850239485,
+      "learning_rate": 4.868835907385183e-07,
+      "loss": 0.1639,
+      "step": 2056
+    },
+    {
+      "epoch": 4.680318543799773,
+      "grad_norm": 0.2852215033386593,
+      "learning_rate": 4.799313243343195e-07,
+      "loss": 0.1571,
+      "step": 2057
+    },
+    {
+      "epoch": 4.6825938566552905,
+      "grad_norm": 0.2542474079594607,
+      "learning_rate": 4.7302844881365273e-07,
+      "loss": 0.1581,
+      "step": 2058
+    },
+    {
+      "epoch": 4.684869169510808,
+      "grad_norm": 0.29933681216203895,
+      "learning_rate": 4.66174981642602e-07,
+      "loss": 0.1774,
+      "step": 2059
+    },
+    {
+      "epoch": 4.6871444823663255,
+      "grad_norm": 0.29791808335229336,
+      "learning_rate": 4.593709401622359e-07,
+      "loss": 0.1626,
+      "step": 2060
+    },
+    {
+      "epoch": 4.689419795221843,
+      "grad_norm": 0.29723177433851855,
+      "learning_rate": 4.5261634158855873e-07,
+      "loss": 0.1765,
+      "step": 2061
+    },
+    {
+      "epoch": 4.691695108077361,
+      "grad_norm": 0.2890583967251704,
+      "learning_rate": 4.459112030124746e-07,
+      "loss": 0.1906,
+      "step": 2062
+    },
+    {
+      "epoch": 4.693970420932878,
+      "grad_norm": 0.28742325857083567,
+      "learning_rate": 4.392555413997346e-07,
+      "loss": 0.1664,
+      "step": 2063
+    },
+    {
+      "epoch": 4.696245733788396,
+      "grad_norm": 0.2792348986230206,
+      "learning_rate": 4.326493735909099e-07,
+      "loss": 0.1908,
+      "step": 2064
+    },
+    {
+      "epoch": 4.698521046643913,
+      "grad_norm": 0.2813521785281826,
+      "learning_rate": 4.2609271630133174e-07,
+      "loss": 0.1688,
+      "step": 2065
+    },
+    {
+      "epoch": 4.7007963594994315,
+      "grad_norm": 0.26469530740883623,
+      "learning_rate": 4.195855861210607e-07,
+      "loss": 0.1797,
+      "step": 2066
+    },
+    {
+      "epoch": 4.703071672354949,
+      "grad_norm": 0.264192160968152,
+      "learning_rate": 4.1312799951483964e-07,
+      "loss": 0.1848,
+      "step": 2067
+    },
+    {
+      "epoch": 4.705346985210467,
+      "grad_norm": 0.2813125672200287,
+      "learning_rate": 4.0671997282205165e-07,
+      "loss": 0.1631,
+      "step": 2068
+    },
+    {
+      "epoch": 4.707622298065984,
+      "grad_norm": 0.2790958711410099,
+      "learning_rate": 4.0036152225668257e-07,
+      "loss": 0.1688,
+      "step": 2069
+    },
+    {
+      "epoch": 4.709897610921502,
+      "grad_norm": 0.2765276490546693,
+      "learning_rate": 3.9405266390727836e-07,
+      "loss": 0.1647,
+      "step": 2070
+    },
+    {
+      "epoch": 4.712172923777019,
+      "grad_norm": 0.29796867137025485,
+      "learning_rate": 3.877934137368988e-07,
+      "loss": 0.1652,
+      "step": 2071
+    },
+    {
+      "epoch": 4.714448236632537,
+      "grad_norm": 0.2806553335956702,
+      "learning_rate": 3.8158378758308634e-07,
+      "loss": 0.1722,
+      "step": 2072
+    },
+    {
+      "epoch": 4.716723549488055,
+      "grad_norm": 0.2821445087667387,
+      "learning_rate": 3.754238011578237e-07,
+      "loss": 0.1754,
+      "step": 2073
+    },
+    {
+      "epoch": 4.718998862343573,
+      "grad_norm": 0.29472595215090663,
+      "learning_rate": 3.693134700474854e-07,
+      "loss": 0.183,
+      "step": 2074
+    },
+    {
+      "epoch": 4.72127417519909,
+      "grad_norm": 0.2840198880867471,
+      "learning_rate": 3.632528097128085e-07,
+      "loss": 0.1758,
+      "step": 2075
+    },
+    {
+      "epoch": 4.723549488054608,
+      "grad_norm": 0.2694843262307711,
+      "learning_rate": 3.5724183548885514e-07,
+      "loss": 0.2065,
+      "step": 2076
+    },
+    {
+      "epoch": 4.725824800910125,
+      "grad_norm": 0.28265532948248634,
+      "learning_rate": 3.512805625849569e-07,
+      "loss": 0.197,
+      "step": 2077
+    },
+    {
+      "epoch": 4.728100113765643,
+      "grad_norm": 0.29936116017536235,
+      "learning_rate": 3.4536900608469924e-07,
+      "loss": 0.1521,
+      "step": 2078
+    },
+    {
+      "epoch": 4.73037542662116,
+      "grad_norm": 0.2788020784215453,
+      "learning_rate": 3.395071809458661e-07,
+      "loss": 0.161,
+      "step": 2079
+    },
+    {
+      "epoch": 4.732650739476678,
+      "grad_norm": 0.2753131318262618,
+      "learning_rate": 3.336951020004087e-07,
+      "loss": 0.1875,
+      "step": 2080
+    },
+    {
+      "epoch": 4.734926052332195,
+      "grad_norm": 0.27164247082837706,
+      "learning_rate": 3.279327839544122e-07,
+      "loss": 0.1714,
+      "step": 2081
+    },
+    {
+      "epoch": 4.737201365187714,
+      "grad_norm": 0.2859412688523052,
+      "learning_rate": 3.2222024138804264e-07,
+      "loss": 0.1543,
+      "step": 2082
+    },
+    {
+      "epoch": 4.739476678043231,
+      "grad_norm": 0.2803506336195304,
+      "learning_rate": 3.165574887555334e-07,
+      "loss": 0.1461,
+      "step": 2083
+    },
+    {
+      "epoch": 4.741751990898749,
+      "grad_norm": 0.27359098725884834,
+      "learning_rate": 3.109445403851319e-07,
+      "loss": 0.1654,
+      "step": 2084
+    },
+    {
+      "epoch": 4.744027303754266,
+      "grad_norm": 0.2768874816031207,
+      "learning_rate": 3.05381410479062e-07,
+      "loss": 0.1632,
+      "step": 2085
+    },
+    {
+      "epoch": 4.746302616609784,
+      "grad_norm": 0.27909644994464433,
+      "learning_rate": 2.998681131134995e-07,
+      "loss": 0.1925,
+      "step": 2086
+    },
+    {
+      "epoch": 4.748577929465301,
+      "grad_norm": 0.27589370596456175,
+      "learning_rate": 2.9440466223853656e-07,
+      "loss": 0.153,
+      "step": 2087
+    },
+    {
+      "epoch": 4.750853242320819,
+      "grad_norm": 0.28659963483934087,
+      "learning_rate": 2.8899107167812634e-07,
+      "loss": 0.1794,
+      "step": 2088
+    },
+    {
+      "epoch": 4.753128555176337,
+      "grad_norm": 0.2802335924025782,
+      "learning_rate": 2.8362735513007387e-07,
+      "loss": 0.2002,
+      "step": 2089
+    },
+    {
+      "epoch": 4.755403868031855,
+      "grad_norm": 0.2837564569638717,
+      "learning_rate": 2.783135261659831e-07,
+      "loss": 0.1426,
+      "step": 2090
+    },
+    {
+      "epoch": 4.757679180887372,
+      "grad_norm": 0.2724450386394914,
+      "learning_rate": 2.730495982312342e-07,
+      "loss": 0.1363,
+      "step": 2091
+    },
+    {
+      "epoch": 4.75995449374289,
+      "grad_norm": 0.2736902453279952,
+      "learning_rate": 2.678355846449465e-07,
+      "loss": 0.161,
+      "step": 2092
+    },
+    {
+      "epoch": 4.762229806598407,
+      "grad_norm": 0.2821596651544852,
+      "learning_rate": 2.6267149859993347e-07,
+      "loss": 0.1815,
+      "step": 2093
+    },
+    {
+      "epoch": 4.764505119453925,
+      "grad_norm": 0.2764937025615303,
+      "learning_rate": 2.5755735316268514e-07,
+      "loss": 0.1806,
+      "step": 2094
+    },
+    {
+      "epoch": 4.766780432309442,
+      "grad_norm": 0.2810702288129479,
+      "learning_rate": 2.524931612733328e-07,
+      "loss": 0.1332,
+      "step": 2095
+    },
+    {
+      "epoch": 4.769055745164961,
+      "grad_norm": 0.29044871796679766,
+      "learning_rate": 2.47478935745602e-07,
+      "loss": 0.1612,
+      "step": 2096
+    },
+    {
+      "epoch": 4.771331058020478,
+      "grad_norm": 0.2683096871030952,
+      "learning_rate": 2.4251468926680175e-07,
+      "loss": 0.1661,
+      "step": 2097
+    },
+    {
+      "epoch": 4.773606370875996,
+      "grad_norm": 0.2945643148140594,
+      "learning_rate": 2.3760043439776892e-07,
+      "loss": 0.1813,
+      "step": 2098
+    },
+    {
+      "epoch": 4.775881683731513,
+      "grad_norm": 0.2893737837113469,
+      "learning_rate": 2.327361835728592e-07,
+      "loss": 0.1498,
+      "step": 2099
+    },
+    {
+      "epoch": 4.778156996587031,
+      "grad_norm": 0.286127100981063,
+      "learning_rate": 2.279219490998985e-07,
+      "loss": 0.2069,
+      "step": 2100
+    },
+    {
+      "epoch": 4.780432309442548,
+      "grad_norm": 0.2760250960457435,
+      "learning_rate": 2.231577431601606e-07,
+      "loss": 0.1459,
+      "step": 2101
+    },
+    {
+      "epoch": 4.782707622298066,
+      "grad_norm": 0.28125460188963713,
+      "learning_rate": 2.184435778083338e-07,
+      "loss": 0.1778,
+      "step": 2102
+    },
+    {
+      "epoch": 4.784982935153583,
+      "grad_norm": 0.26720575727894325,
+      "learning_rate": 2.1377946497249225e-07,
+      "loss": 0.1476,
+      "step": 2103
+    },
+    {
+      "epoch": 4.787258248009101,
+      "grad_norm": 0.28726572485784596,
+      "learning_rate": 2.091654164540602e-07,
+      "loss": 0.1678,
+      "step": 2104
+    },
+    {
+      "epoch": 4.789533560864619,
+      "grad_norm": 0.2960648414045368,
+      "learning_rate": 2.0460144392778768e-07,
+      "loss": 0.1665,
+      "step": 2105
+    },
+    {
+      "epoch": 4.791808873720137,
+      "grad_norm": 0.2851934467857292,
+      "learning_rate": 2.0008755894172172e-07,
+      "loss": 0.1666,
+      "step": 2106
+    },
+    {
+      "epoch": 4.794084186575654,
+      "grad_norm": 0.29867023568951107,
+      "learning_rate": 1.956237729171706e-07,
+      "loss": 0.1609,
+      "step": 2107
+    },
+    {
+      "epoch": 4.796359499431172,
+      "grad_norm": 0.27203307996084625,
+      "learning_rate": 1.912100971486841e-07,
+      "loss": 0.1728,
+      "step": 2108
+    },
+    {
+      "epoch": 4.798634812286689,
+      "grad_norm": 0.29093396081678047,
+      "learning_rate": 1.868465428040156e-07,
+      "loss": 0.1785,
+      "step": 2109
+    },
+    {
+      "epoch": 4.800910125142207,
+      "grad_norm": 0.2775746820409794,
+      "learning_rate": 1.8253312092409992e-07,
+      "loss": 0.1677,
+      "step": 2110
+    },
+    {
+      "epoch": 4.803185437997724,
+      "grad_norm": 0.27453488428691564,
+      "learning_rate": 1.782698424230267e-07,
+      "loss": 0.1854,
+      "step": 2111
+    },
+    {
+      "epoch": 4.805460750853243,
+      "grad_norm": 0.27937845048345333,
+      "learning_rate": 1.740567180880004e-07,
+      "loss": 0.1774,
+      "step": 2112
+    },
+    {
+      "epoch": 4.80773606370876,
+      "grad_norm": 0.2982003035523139,
+      "learning_rate": 1.6989375857933588e-07,
+      "loss": 0.1588,
+      "step": 2113
+    },
+    {
+      "epoch": 4.810011376564278,
+      "grad_norm": 0.27122001773024823,
+      "learning_rate": 1.6578097443040286e-07,
+      "loss": 0.1639,
+      "step": 2114
+    },
+    {
+      "epoch": 4.812286689419795,
+      "grad_norm": 0.2910168540094759,
+      "learning_rate": 1.6171837604762597e-07,
+      "loss": 0.1812,
+      "step": 2115
+    },
+    {
+      "epoch": 4.814562002275313,
+      "grad_norm": 0.2847303758330056,
+      "learning_rate": 1.577059737104447e-07,
+      "loss": 0.1667,
+      "step": 2116
+    },
+    {
+      "epoch": 4.81683731513083,
+      "grad_norm": 0.26816369266991275,
+      "learning_rate": 1.5374377757128245e-07,
+      "loss": 0.1707,
+      "step": 2117
+    },
+    {
+      "epoch": 4.819112627986348,
+      "grad_norm": 0.2848979545208066,
+      "learning_rate": 1.4983179765553523e-07,
+      "loss": 0.1886,
+      "step": 2118
+    },
+    {
+      "epoch": 4.821387940841865,
+      "grad_norm": 0.2619835505950323,
+      "learning_rate": 1.4597004386153635e-07,
+      "loss": 0.1701,
+      "step": 2119
+    },
+    {
+      "epoch": 4.823663253697383,
+      "grad_norm": 0.2808499371759947,
+      "learning_rate": 1.421585259605318e-07,
+      "loss": 0.1677,
+      "step": 2120
+    },
+    {
+      "epoch": 4.825938566552901,
+      "grad_norm": 0.2916709025066011,
+      "learning_rate": 1.3839725359666267e-07,
+      "loss": 0.1836,
+      "step": 2121
+    },
+    {
+      "epoch": 4.828213879408419,
+      "grad_norm": 0.2754848446765079,
+      "learning_rate": 1.3468623628692945e-07,
+      "loss": 0.1433,
+      "step": 2122
+    },
+    {
+      "epoch": 4.830489192263936,
+      "grad_norm": 0.30298090295810787,
+      "learning_rate": 1.3102548342118105e-07,
+      "loss": 0.1909,
+      "step": 2123
+    },
+    {
+      "epoch": 4.832764505119454,
+      "grad_norm": 0.28860241950399845,
+      "learning_rate": 1.274150042620792e-07,
+      "loss": 0.2022,
+      "step": 2124
+    },
+    {
+      "epoch": 4.835039817974971,
+      "grad_norm": 0.26800108458504196,
+      "learning_rate": 1.2385480794507853e-07,
+      "loss": 0.16,
+      "step": 2125
+    },
+    {
+      "epoch": 4.837315130830489,
+      "grad_norm": 0.2853323557437661,
+      "learning_rate": 1.20344903478411e-07,
+      "loss": 0.202,
+      "step": 2126
+    },
+    {
+      "epoch": 4.839590443686006,
+      "grad_norm": 0.2733948236798397,
+      "learning_rate": 1.1688529974305251e-07,
+      "loss": 0.1737,
+      "step": 2127
+    },
+    {
+      "epoch": 4.841865756541525,
+      "grad_norm": 0.28961179550636856,
+      "learning_rate": 1.1347600549270532e-07,
+      "loss": 0.1734,
+      "step": 2128
+    },
+    {
+      "epoch": 4.844141069397042,
+      "grad_norm": 0.28913141216509214,
+      "learning_rate": 1.1011702935377789e-07,
+      "loss": 0.1769,
+      "step": 2129
+    },
+    {
+      "epoch": 4.84641638225256,
+      "grad_norm": 0.27173575099170844,
+      "learning_rate": 1.0680837982535607e-07,
+      "loss": 0.1619,
+      "step": 2130
+    },
+    {
+      "epoch": 4.848691695108077,
+      "grad_norm": 0.2737763078214319,
+      "learning_rate": 1.0355006527919209e-07,
+      "loss": 0.1705,
+      "step": 2131
+    },
+    {
+      "epoch": 4.850967007963595,
+      "grad_norm": 0.26823155637529944,
+      "learning_rate": 1.0034209395967554e-07,
+      "loss": 0.1719,
+      "step": 2132
+    },
+    {
+      "epoch": 4.853242320819112,
+      "grad_norm": 0.27890526953507444,
+      "learning_rate": 9.718447398381348e-08,
+      "loss": 0.1477,
+      "step": 2133
+    },
+    {
+      "epoch": 4.85551763367463,
+      "grad_norm": 0.26588733805101394,
+      "learning_rate": 9.407721334120601e-08,
+      "loss": 0.1825,
+      "step": 2134
+    },
+    {
+      "epoch": 4.857792946530148,
+      "grad_norm": 0.29092629682676213,
+      "learning_rate": 9.102031989404403e-08,
+      "loss": 0.1523,
+      "step": 2135
+    },
+    {
+      "epoch": 4.860068259385666,
+      "grad_norm": 0.29047716670487517,
+      "learning_rate": 8.80138013770604e-08,
+      "loss": 0.1752,
+      "step": 2136
+    },
+    {
+      "epoch": 4.862343572241183,
+      "grad_norm": 0.26357179549459897,
+      "learning_rate": 8.505766539753879e-08,
+      "loss": 0.1711,
+      "step": 2137
+    },
+    {
+      "epoch": 4.864618885096701,
+      "grad_norm": 0.2767793820851537,
+      "learning_rate": 8.215191943527378e-08,
+      "loss": 0.1839,
+      "step": 2138
+    },
+    {
+      "epoch": 4.8668941979522184,
+      "grad_norm": 0.2752011773302518,
+      "learning_rate": 7.929657084256858e-08,
+      "loss": 0.1387,
+      "step": 2139
+    },
+    {
+      "epoch": 4.869169510807736,
+      "grad_norm": 0.29282854702207456,
+      "learning_rate": 7.649162684419731e-08,
+      "loss": 0.1856,
+      "step": 2140
+    },
+    {
+      "epoch": 4.8714448236632535,
+      "grad_norm": 0.29086067102522595,
+      "learning_rate": 7.373709453740497e-08,
+      "loss": 0.1738,
+      "step": 2141
+    },
+    {
+      "epoch": 4.873720136518771,
+      "grad_norm": 0.27826866182859905,
+      "learning_rate": 7.10329808918786e-08,
+      "loss": 0.1433,
+      "step": 2142
+    },
+    {
+      "epoch": 4.8759954493742885,
+      "grad_norm": 0.27848071371945615,
+      "learning_rate": 6.837929274973399e-08,
+      "loss": 0.2025,
+      "step": 2143
+    },
+    {
+      "epoch": 4.878270762229807,
+      "grad_norm": 0.28174616730635366,
+      "learning_rate": 6.57760368254956e-08,
+      "loss": 0.169,
+      "step": 2144
+    },
+    {
+      "epoch": 4.8805460750853245,
+      "grad_norm": 0.28216646768926074,
+      "learning_rate": 6.322321970608337e-08,
+      "loss": 0.1998,
+      "step": 2145
+    },
+    {
+      "epoch": 4.882821387940842,
+      "grad_norm": 0.2839738511495771,
+      "learning_rate": 6.072084785079257e-08,
+      "loss": 0.1881,
+      "step": 2146
+    },
+    {
+      "epoch": 4.8850967007963595,
+      "grad_norm": 0.2719170431364124,
+      "learning_rate": 5.826892759128066e-08,
+      "loss": 0.1637,
+      "step": 2147
+    },
+    {
+      "epoch": 4.887372013651877,
+      "grad_norm": 0.2900758080734818,
+      "learning_rate": 5.5867465131547174e-08,
+      "loss": 0.2036,
+      "step": 2148
+    },
+    {
+      "epoch": 4.8896473265073945,
+      "grad_norm": 0.27675763159719285,
+      "learning_rate": 5.35164665479182e-08,
+      "loss": 0.1753,
+      "step": 2149
+    },
+    {
+      "epoch": 4.891922639362912,
+      "grad_norm": 0.2682400680433344,
+      "learning_rate": 5.121593778903755e-08,
+      "loss": 0.1654,
+      "step": 2150
+    },
+    {
+      "epoch": 4.8941979522184305,
+      "grad_norm": 0.2834066342426588,
+      "learning_rate": 4.896588467584451e-08,
+      "loss": 0.1727,
+      "step": 2151
+    },
+    {
+      "epoch": 4.896473265073948,
+      "grad_norm": 0.28432645164270115,
+      "learning_rate": 4.676631290156497e-08,
+      "loss": 0.1694,
+      "step": 2152
+    },
+    {
+      "epoch": 4.8987485779294655,
+      "grad_norm": 0.2705477739865789,
+      "learning_rate": 4.4617228031687e-08,
+      "loss": 0.1916,
+      "step": 2153
+    },
+    {
+      "epoch": 4.901023890784983,
+      "grad_norm": 0.28281818023901295,
+      "learning_rate": 4.2518635503960845e-08,
+      "loss": 0.1919,
+      "step": 2154
+    },
+    {
+      "epoch": 4.9032992036405005,
+      "grad_norm": 0.275802639540135,
+      "learning_rate": 4.047054062837452e-08,
+      "loss": 0.1598,
+      "step": 2155
+    },
+    {
+      "epoch": 4.905574516496018,
+      "grad_norm": 0.2716919729625155,
+      "learning_rate": 3.8472948587144895e-08,
+      "loss": 0.1694,
+      "step": 2156
+    },
+    {
+      "epoch": 4.907849829351536,
+      "grad_norm": 0.28826008139527803,
+      "learning_rate": 3.652586443470219e-08,
+      "loss": 0.1645,
+      "step": 2157
+    },
+    {
+      "epoch": 4.910125142207053,
+      "grad_norm": 0.27948590042199084,
+      "learning_rate": 3.46292930976766e-08,
+      "loss": 0.1624,
+      "step": 2158
+    },
+    {
+      "epoch": 4.912400455062571,
+      "grad_norm": 0.26789167961996163,
+      "learning_rate": 3.2783239374893914e-08,
+      "loss": 0.1697,
+      "step": 2159
+    },
+    {
+      "epoch": 4.914675767918089,
+      "grad_norm": 0.2670318829401166,
+      "learning_rate": 3.0987707937351066e-08,
+      "loss": 0.1837,
+      "step": 2160
+    },
+    {
+      "epoch": 4.9169510807736065,
+      "grad_norm": 0.27546556570733666,
+      "learning_rate": 2.9242703328211662e-08,
+      "loss": 0.1556,
+      "step": 2161
+    },
+    {
+      "epoch": 4.919226393629124,
+      "grad_norm": 0.2725902589660314,
+      "learning_rate": 2.7548229962794938e-08,
+      "loss": 0.1507,
+      "step": 2162
+    },
+    {
+      "epoch": 4.921501706484642,
+      "grad_norm": 0.2747503934785236,
+      "learning_rate": 2.59042921285646e-08,
+      "loss": 0.1849,
+      "step": 2163
+    },
+    {
+      "epoch": 4.923777019340159,
+      "grad_norm": 0.29811239655450217,
+      "learning_rate": 2.4310893985113328e-08,
+      "loss": 0.1761,
+      "step": 2164
+    },
+    {
+      "epoch": 4.926052332195677,
+      "grad_norm": 0.29286564186734176,
+      "learning_rate": 2.2768039564151635e-08,
+      "loss": 0.1631,
+      "step": 2165
+    },
+    {
+      "epoch": 4.928327645051194,
+      "grad_norm": 0.2824169805110718,
+      "learning_rate": 2.1275732769505675e-08,
+      "loss": 0.1669,
+      "step": 2166
+    },
+    {
+      "epoch": 4.9306029579067125,
+      "grad_norm": 0.29684890488399457,
+      "learning_rate": 1.983397737710613e-08,
+      "loss": 0.1758,
+      "step": 2167
+    },
+    {
+      "epoch": 4.93287827076223,
+      "grad_norm": 0.2653707831593049,
+      "learning_rate": 1.8442777034970437e-08,
+      "loss": 0.1806,
+      "step": 2168
+    },
+    {
+      "epoch": 4.935153583617748,
+      "grad_norm": 0.27223470203609895,
+      "learning_rate": 1.7102135263198372e-08,
+      "loss": 0.2211,
+      "step": 2169
+    },
+    {
+      "epoch": 4.937428896473265,
+      "grad_norm": 0.2864468681497177,
+      "learning_rate": 1.5812055453963136e-08,
+      "loss": 0.166,
+      "step": 2170
+    },
+    {
+      "epoch": 4.939704209328783,
+      "grad_norm": 0.2797238676132464,
+      "learning_rate": 1.4572540871506947e-08,
+      "loss": 0.168,
+      "step": 2171
+    },
+    {
+      "epoch": 4.9419795221843,
+      "grad_norm": 0.2848478065778815,
+      "learning_rate": 1.3383594652125464e-08,
+      "loss": 0.2005,
+      "step": 2172
+    },
+    {
+      "epoch": 4.944254835039818,
+      "grad_norm": 0.2931191481258792,
+      "learning_rate": 1.2245219804163377e-08,
+      "loss": 0.2171,
+      "step": 2173
+    },
+    {
+      "epoch": 4.946530147895336,
+      "grad_norm": 0.2709247922397299,
+      "learning_rate": 1.1157419208005505e-08,
+      "loss": 0.2141,
+      "step": 2174
+    },
+    {
+      "epoch": 4.948805460750854,
+      "grad_norm": 0.269577290444941,
+      "learning_rate": 1.01201956160768e-08,
+      "loss": 0.1655,
+      "step": 2175
+    },
+    {
+      "epoch": 4.951080773606371,
+      "grad_norm": 0.28366020700511824,
+      "learning_rate": 9.13355165282015e-09,
+      "loss": 0.1655,
+      "step": 2176
+    },
+    {
+      "epoch": 4.953356086461889,
+      "grad_norm": 0.2752590720590919,
+      "learning_rate": 8.197489814705251e-09,
+      "loss": 0.1724,
+      "step": 2177
+    },
+    {
+      "epoch": 4.955631399317406,
+      "grad_norm": 0.2751081587825422,
+      "learning_rate": 7.31201247021085e-09,
+      "loss": 0.186,
+      "step": 2178
+    },
+    {
+      "epoch": 4.957906712172924,
+      "grad_norm": 0.29165567713398455,
+      "learning_rate": 6.477121859831404e-09,
+      "loss": 0.14,
+      "step": 2179
+    },
+    {
+      "epoch": 4.960182025028441,
+      "grad_norm": 0.26807798272654737,
+      "learning_rate": 5.692820096054874e-09,
+      "loss": 0.1915,
+      "step": 2180
+    },
+    {
+      "epoch": 4.962457337883959,
+      "grad_norm": 0.307986343441169,
+      "learning_rate": 4.959109163369391e-09,
+      "loss": 0.17,
+      "step": 2181
+    },
+    {
+      "epoch": 4.964732650739476,
+      "grad_norm": 0.28646812677908806,
+      "learning_rate": 4.275990918256589e-09,
+      "loss": 0.1604,
+      "step": 2182
+    },
+    {
+      "epoch": 4.967007963594995,
+      "grad_norm": 0.28434276408514697,
+      "learning_rate": 3.6434670891871694e-09,
+      "loss": 0.1653,
+      "step": 2183
+    },
+    {
+      "epoch": 4.969283276450512,
+      "grad_norm": 0.28547418520789625,
+      "learning_rate": 3.061539276609793e-09,
+      "loss": 0.1963,
+      "step": 2184
+    },
+    {
+      "epoch": 4.97155858930603,
+      "grad_norm": 0.2858738833930566,
+      "learning_rate": 2.530208952953306e-09,
+      "loss": 0.1919,
+      "step": 2185
+    },
+    {
+      "epoch": 4.973833902161547,
+      "grad_norm": 0.3116700983501146,
+      "learning_rate": 2.049477462622296e-09,
+      "loss": 0.154,
+      "step": 2186
+    },
+    {
+      "epoch": 4.976109215017065,
+      "grad_norm": 0.27812011451558866,
+      "learning_rate": 1.6193460219970924e-09,
+      "loss": 0.1861,
+      "step": 2187
+    },
+    {
+      "epoch": 4.978384527872582,
+      "grad_norm": 0.28680689327047326,
+      "learning_rate": 1.2398157194182248e-09,
+      "loss": 0.1636,
+      "step": 2188
+    },
+    {
+      "epoch": 4.9806598407281,
+      "grad_norm": 0.27175734346430375,
+      "learning_rate": 9.108875151975227e-10,
+      "loss": 0.1453,
+      "step": 2189
+    },
+    {
+      "epoch": 4.982935153583618,
+      "grad_norm": 0.28732427721080184,
+      "learning_rate": 6.325622416136767e-10,
+      "loss": 0.1984,
+      "step": 2190
+    },
+    {
+      "epoch": 4.985210466439136,
+      "grad_norm": 0.2795942358884219,
+      "learning_rate": 4.048406028966945e-10,
+      "loss": 0.1735,
+      "step": 2191
+    },
+    {
+      "epoch": 4.987485779294653,
+      "grad_norm": 0.2901912120404457,
+      "learning_rate": 2.2772317524566434e-10,
+      "loss": 0.1743,
+      "step": 2192
+    },
+    {
+      "epoch": 4.989761092150171,
+      "grad_norm": 0.2822809573112233,
+      "learning_rate": 1.0121040681321249e-10,
+      "loss": 0.1844,
+      "step": 2193
+    },
+    {
+      "epoch": 4.992036405005688,
+      "grad_norm": 0.29485501566563027,
+      "learning_rate": 2.530261770772313e-11,
+      "loss": 0.1665,
+      "step": 2194
+    },
+    {
+      "epoch": 4.994311717861206,
+      "grad_norm": 0.2749861414237793,
+      "learning_rate": 0.0,
+      "loss": 0.2018,
+      "step": 2195
+    },
+    {
+      "epoch": 4.994311717861206,
+      "step": 2195,
+      "total_flos": 3.168901699486089e+18,
+      "train_loss": 0.34217921912262816,
+      "train_runtime": 81219.8428,
+      "train_samples_per_second": 3.463,
+      "train_steps_per_second": 0.027
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2195,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "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": 3.168901699486089e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}