diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,45192 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.999534955820803,
+  "eval_steps": 500,
+  "global_step": 6450,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004650441791970237,
+      "grad_norm": 6.0558061599731445,
+      "learning_rate": 1.550387596899225e-08,
+      "loss": 0.8113,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009300883583940474,
+      "grad_norm": 6.12605619430542,
+      "learning_rate": 3.10077519379845e-08,
+      "loss": 0.825,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013951325375910712,
+      "grad_norm": 6.057300567626953,
+      "learning_rate": 4.6511627906976744e-08,
+      "loss": 0.8147,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018601767167880949,
+      "grad_norm": 6.23735237121582,
+      "learning_rate": 6.2015503875969e-08,
+      "loss": 0.8659,
+      "step": 4
+    },
+    {
+      "epoch": 0.002325220895985119,
+      "grad_norm": 5.970339298248291,
+      "learning_rate": 7.751937984496124e-08,
+      "loss": 0.7994,
+      "step": 5
+    },
+    {
+      "epoch": 0.0027902650751821423,
+      "grad_norm": 6.014130592346191,
+      "learning_rate": 9.302325581395349e-08,
+      "loss": 0.7927,
+      "step": 6
+    },
+    {
+      "epoch": 0.0032553092543791662,
+      "grad_norm": 6.405108451843262,
+      "learning_rate": 1.0852713178294575e-07,
+      "loss": 0.8445,
+      "step": 7
+    },
+    {
+      "epoch": 0.0037203534335761897,
+      "grad_norm": 5.915679931640625,
+      "learning_rate": 1.24031007751938e-07,
+      "loss": 0.7874,
+      "step": 8
+    },
+    {
+      "epoch": 0.004185397612773214,
+      "grad_norm": 6.164096355438232,
+      "learning_rate": 1.3953488372093024e-07,
+      "loss": 0.841,
+      "step": 9
+    },
+    {
+      "epoch": 0.004650441791970238,
+      "grad_norm": 6.06751012802124,
+      "learning_rate": 1.5503875968992249e-07,
+      "loss": 0.8139,
+      "step": 10
+    },
+    {
+      "epoch": 0.005115485971167261,
+      "grad_norm": 6.07175350189209,
+      "learning_rate": 1.7054263565891473e-07,
+      "loss": 0.8137,
+      "step": 11
+    },
+    {
+      "epoch": 0.005580530150364285,
+      "grad_norm": 6.381070613861084,
+      "learning_rate": 1.8604651162790698e-07,
+      "loss": 0.8233,
+      "step": 12
+    },
+    {
+      "epoch": 0.0060455743295613085,
+      "grad_norm": 6.221768379211426,
+      "learning_rate": 2.0155038759689925e-07,
+      "loss": 0.8265,
+      "step": 13
+    },
+    {
+      "epoch": 0.0065106185087583325,
+      "grad_norm": 5.883142471313477,
+      "learning_rate": 2.170542635658915e-07,
+      "loss": 0.7789,
+      "step": 14
+    },
+    {
+      "epoch": 0.0069756626879553555,
+      "grad_norm": 5.981773853302002,
+      "learning_rate": 2.3255813953488374e-07,
+      "loss": 0.7948,
+      "step": 15
+    },
+    {
+      "epoch": 0.0074407068671523795,
+      "grad_norm": 5.9482035636901855,
+      "learning_rate": 2.48062015503876e-07,
+      "loss": 0.782,
+      "step": 16
+    },
+    {
+      "epoch": 0.007905751046349403,
+      "grad_norm": 6.413148403167725,
+      "learning_rate": 2.6356589147286826e-07,
+      "loss": 0.8528,
+      "step": 17
+    },
+    {
+      "epoch": 0.008370795225546427,
+      "grad_norm": 5.644394397735596,
+      "learning_rate": 2.790697674418605e-07,
+      "loss": 0.8007,
+      "step": 18
+    },
+    {
+      "epoch": 0.00883583940474345,
+      "grad_norm": 5.608232021331787,
+      "learning_rate": 2.9457364341085275e-07,
+      "loss": 0.7642,
+      "step": 19
+    },
+    {
+      "epoch": 0.009300883583940475,
+      "grad_norm": 5.738401889801025,
+      "learning_rate": 3.1007751937984497e-07,
+      "loss": 0.8086,
+      "step": 20
+    },
+    {
+      "epoch": 0.009765927763137498,
+      "grad_norm": 5.834144115447998,
+      "learning_rate": 3.2558139534883724e-07,
+      "loss": 0.8128,
+      "step": 21
+    },
+    {
+      "epoch": 0.010230971942334521,
+      "grad_norm": 5.556512355804443,
+      "learning_rate": 3.4108527131782946e-07,
+      "loss": 0.7919,
+      "step": 22
+    },
+    {
+      "epoch": 0.010696016121531546,
+      "grad_norm": 5.6250200271606445,
+      "learning_rate": 3.5658914728682174e-07,
+      "loss": 0.7924,
+      "step": 23
+    },
+    {
+      "epoch": 0.01116106030072857,
+      "grad_norm": 4.8199992179870605,
+      "learning_rate": 3.7209302325581396e-07,
+      "loss": 0.7686,
+      "step": 24
+    },
+    {
+      "epoch": 0.011626104479925592,
+      "grad_norm": 4.794100761413574,
+      "learning_rate": 3.8759689922480623e-07,
+      "loss": 0.7635,
+      "step": 25
+    },
+    {
+      "epoch": 0.012091148659122617,
+      "grad_norm": 4.6750874519348145,
+      "learning_rate": 4.031007751937985e-07,
+      "loss": 0.748,
+      "step": 26
+    },
+    {
+      "epoch": 0.01255619283831964,
+      "grad_norm": 4.775611877441406,
+      "learning_rate": 4.186046511627907e-07,
+      "loss": 0.7594,
+      "step": 27
+    },
+    {
+      "epoch": 0.013021237017516665,
+      "grad_norm": 4.403256893157959,
+      "learning_rate": 4.34108527131783e-07,
+      "loss": 0.7302,
+      "step": 28
+    },
+    {
+      "epoch": 0.013486281196713688,
+      "grad_norm": 4.565180778503418,
+      "learning_rate": 4.496124031007752e-07,
+      "loss": 0.7729,
+      "step": 29
+    },
+    {
+      "epoch": 0.013951325375910711,
+      "grad_norm": 4.50689697265625,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 0.7524,
+      "step": 30
+    },
+    {
+      "epoch": 0.014416369555107736,
+      "grad_norm": 4.400961875915527,
+      "learning_rate": 4.806201550387598e-07,
+      "loss": 0.7466,
+      "step": 31
+    },
+    {
+      "epoch": 0.014881413734304759,
+      "grad_norm": 4.230681419372559,
+      "learning_rate": 4.96124031007752e-07,
+      "loss": 0.7352,
+      "step": 32
+    },
+    {
+      "epoch": 0.015346457913501782,
+      "grad_norm": 3.3713021278381348,
+      "learning_rate": 5.116279069767442e-07,
+      "loss": 0.7095,
+      "step": 33
+    },
+    {
+      "epoch": 0.015811502092698805,
+      "grad_norm": 2.6536169052124023,
+      "learning_rate": 5.271317829457365e-07,
+      "loss": 0.7099,
+      "step": 34
+    },
+    {
+      "epoch": 0.01627654627189583,
+      "grad_norm": 2.5625736713409424,
+      "learning_rate": 5.426356589147287e-07,
+      "loss": 0.7209,
+      "step": 35
+    },
+    {
+      "epoch": 0.016741590451092855,
+      "grad_norm": 2.7283709049224854,
+      "learning_rate": 5.58139534883721e-07,
+      "loss": 0.7052,
+      "step": 36
+    },
+    {
+      "epoch": 0.017206634630289878,
+      "grad_norm": 2.3967738151550293,
+      "learning_rate": 5.736434108527132e-07,
+      "loss": 0.6705,
+      "step": 37
+    },
+    {
+      "epoch": 0.0176716788094869,
+      "grad_norm": 2.436204195022583,
+      "learning_rate": 5.891472868217055e-07,
+      "loss": 0.726,
+      "step": 38
+    },
+    {
+      "epoch": 0.018136722988683924,
+      "grad_norm": 2.3361918926239014,
+      "learning_rate": 6.046511627906977e-07,
+      "loss": 0.6694,
+      "step": 39
+    },
+    {
+      "epoch": 0.01860176716788095,
+      "grad_norm": 2.221872568130493,
+      "learning_rate": 6.201550387596899e-07,
+      "loss": 0.6728,
+      "step": 40
+    },
+    {
+      "epoch": 0.019066811347077973,
+      "grad_norm": 2.210745096206665,
+      "learning_rate": 6.356589147286822e-07,
+      "loss": 0.6902,
+      "step": 41
+    },
+    {
+      "epoch": 0.019531855526274997,
+      "grad_norm": 2.1230080127716064,
+      "learning_rate": 6.511627906976745e-07,
+      "loss": 0.697,
+      "step": 42
+    },
+    {
+      "epoch": 0.01999689970547202,
+      "grad_norm": 2.061232089996338,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 0.6746,
+      "step": 43
+    },
+    {
+      "epoch": 0.020461943884669043,
+      "grad_norm": 1.6226105690002441,
+      "learning_rate": 6.821705426356589e-07,
+      "loss": 0.6732,
+      "step": 44
+    },
+    {
+      "epoch": 0.020926988063866066,
+      "grad_norm": 1.4576574563980103,
+      "learning_rate": 6.976744186046513e-07,
+      "loss": 0.6821,
+      "step": 45
+    },
+    {
+      "epoch": 0.021392032243063092,
+      "grad_norm": 1.4970301389694214,
+      "learning_rate": 7.131782945736435e-07,
+      "loss": 0.6661,
+      "step": 46
+    },
+    {
+      "epoch": 0.021857076422260115,
+      "grad_norm": 1.6816171407699585,
+      "learning_rate": 7.286821705426357e-07,
+      "loss": 0.671,
+      "step": 47
+    },
+    {
+      "epoch": 0.02232212060145714,
+      "grad_norm": 1.7516964673995972,
+      "learning_rate": 7.441860465116279e-07,
+      "loss": 0.5983,
+      "step": 48
+    },
+    {
+      "epoch": 0.02278716478065416,
+      "grad_norm": 1.8953578472137451,
+      "learning_rate": 7.596899224806202e-07,
+      "loss": 0.6441,
+      "step": 49
+    },
+    {
+      "epoch": 0.023252208959851185,
+      "grad_norm": 1.8705769777297974,
+      "learning_rate": 7.751937984496125e-07,
+      "loss": 0.6332,
+      "step": 50
+    },
+    {
+      "epoch": 0.02371725313904821,
+      "grad_norm": 1.9592753648757935,
+      "learning_rate": 7.906976744186047e-07,
+      "loss": 0.6445,
+      "step": 51
+    },
+    {
+      "epoch": 0.024182297318245234,
+      "grad_norm": 1.7867804765701294,
+      "learning_rate": 8.06201550387597e-07,
+      "loss": 0.6449,
+      "step": 52
+    },
+    {
+      "epoch": 0.024647341497442257,
+      "grad_norm": 1.7160060405731201,
+      "learning_rate": 8.217054263565892e-07,
+      "loss": 0.6299,
+      "step": 53
+    },
+    {
+      "epoch": 0.02511238567663928,
+      "grad_norm": 1.3905798196792603,
+      "learning_rate": 8.372093023255814e-07,
+      "loss": 0.5903,
+      "step": 54
+    },
+    {
+      "epoch": 0.025577429855836303,
+      "grad_norm": 1.2955713272094727,
+      "learning_rate": 8.527131782945737e-07,
+      "loss": 0.6332,
+      "step": 55
+    },
+    {
+      "epoch": 0.02604247403503333,
+      "grad_norm": 1.2509273290634155,
+      "learning_rate": 8.68217054263566e-07,
+      "loss": 0.6233,
+      "step": 56
+    },
+    {
+      "epoch": 0.026507518214230353,
+      "grad_norm": 1.0419048070907593,
+      "learning_rate": 8.837209302325582e-07,
+      "loss": 0.5845,
+      "step": 57
+    },
+    {
+      "epoch": 0.026972562393427376,
+      "grad_norm": 1.083308458328247,
+      "learning_rate": 8.992248062015504e-07,
+      "loss": 0.6179,
+      "step": 58
+    },
+    {
+      "epoch": 0.0274376065726244,
+      "grad_norm": 1.011205792427063,
+      "learning_rate": 9.147286821705427e-07,
+      "loss": 0.6055,
+      "step": 59
+    },
+    {
+      "epoch": 0.027902650751821422,
+      "grad_norm": 0.9679791927337646,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 0.549,
+      "step": 60
+    },
+    {
+      "epoch": 0.028367694931018445,
+      "grad_norm": 0.9358096718788147,
+      "learning_rate": 9.457364341085272e-07,
+      "loss": 0.5544,
+      "step": 61
+    },
+    {
+      "epoch": 0.028832739110215472,
+      "grad_norm": 0.8703131079673767,
+      "learning_rate": 9.612403100775195e-07,
+      "loss": 0.5859,
+      "step": 62
+    },
+    {
+      "epoch": 0.029297783289412495,
+      "grad_norm": 0.7809782028198242,
+      "learning_rate": 9.767441860465117e-07,
+      "loss": 0.5716,
+      "step": 63
+    },
+    {
+      "epoch": 0.029762827468609518,
+      "grad_norm": 0.7614837884902954,
+      "learning_rate": 9.92248062015504e-07,
+      "loss": 0.5784,
+      "step": 64
+    },
+    {
+      "epoch": 0.03022787164780654,
+      "grad_norm": 0.7258252501487732,
+      "learning_rate": 1.0077519379844962e-06,
+      "loss": 0.5985,
+      "step": 65
+    },
+    {
+      "epoch": 0.030692915827003564,
+      "grad_norm": 0.7365714907646179,
+      "learning_rate": 1.0232558139534884e-06,
+      "loss": 0.5962,
+      "step": 66
+    },
+    {
+      "epoch": 0.03115796000620059,
+      "grad_norm": 0.7005903124809265,
+      "learning_rate": 1.0387596899224806e-06,
+      "loss": 0.5533,
+      "step": 67
+    },
+    {
+      "epoch": 0.03162300418539761,
+      "grad_norm": 0.7548089623451233,
+      "learning_rate": 1.054263565891473e-06,
+      "loss": 0.6215,
+      "step": 68
+    },
+    {
+      "epoch": 0.03208804836459463,
+      "grad_norm": 0.6190081238746643,
+      "learning_rate": 1.0697674418604653e-06,
+      "loss": 0.5489,
+      "step": 69
+    },
+    {
+      "epoch": 0.03255309254379166,
+      "grad_norm": 0.7042703628540039,
+      "learning_rate": 1.0852713178294575e-06,
+      "loss": 0.5756,
+      "step": 70
+    },
+    {
+      "epoch": 0.033018136722988686,
+      "grad_norm": 0.6290520429611206,
+      "learning_rate": 1.1007751937984497e-06,
+      "loss": 0.5925,
+      "step": 71
+    },
+    {
+      "epoch": 0.03348318090218571,
+      "grad_norm": 0.5906243920326233,
+      "learning_rate": 1.116279069767442e-06,
+      "loss": 0.5618,
+      "step": 72
+    },
+    {
+      "epoch": 0.03394822508138273,
+      "grad_norm": 0.5369678735733032,
+      "learning_rate": 1.1317829457364341e-06,
+      "loss": 0.555,
+      "step": 73
+    },
+    {
+      "epoch": 0.034413269260579755,
+      "grad_norm": 0.5198186039924622,
+      "learning_rate": 1.1472868217054264e-06,
+      "loss": 0.534,
+      "step": 74
+    },
+    {
+      "epoch": 0.03487831343977678,
+      "grad_norm": 0.5380486845970154,
+      "learning_rate": 1.1627906976744188e-06,
+      "loss": 0.5528,
+      "step": 75
+    },
+    {
+      "epoch": 0.0353433576189738,
+      "grad_norm": 0.5507330298423767,
+      "learning_rate": 1.178294573643411e-06,
+      "loss": 0.5608,
+      "step": 76
+    },
+    {
+      "epoch": 0.035808401798170825,
+      "grad_norm": 0.5470713376998901,
+      "learning_rate": 1.1937984496124032e-06,
+      "loss": 0.549,
+      "step": 77
+    },
+    {
+      "epoch": 0.03627344597736785,
+      "grad_norm": 0.5553979277610779,
+      "learning_rate": 1.2093023255813954e-06,
+      "loss": 0.5203,
+      "step": 78
+    },
+    {
+      "epoch": 0.03673849015656487,
+      "grad_norm": 0.5427577495574951,
+      "learning_rate": 1.2248062015503877e-06,
+      "loss": 0.5348,
+      "step": 79
+    },
+    {
+      "epoch": 0.0372035343357619,
+      "grad_norm": 0.5340808629989624,
+      "learning_rate": 1.2403100775193799e-06,
+      "loss": 0.5373,
+      "step": 80
+    },
+    {
+      "epoch": 0.037668578514958924,
+      "grad_norm": 0.5276907682418823,
+      "learning_rate": 1.2558139534883723e-06,
+      "loss": 0.5403,
+      "step": 81
+    },
+    {
+      "epoch": 0.03813362269415595,
+      "grad_norm": 0.4982495605945587,
+      "learning_rate": 1.2713178294573643e-06,
+      "loss": 0.5283,
+      "step": 82
+    },
+    {
+      "epoch": 0.03859866687335297,
+      "grad_norm": 0.48218581080436707,
+      "learning_rate": 1.2868217054263568e-06,
+      "loss": 0.5499,
+      "step": 83
+    },
+    {
+      "epoch": 0.03906371105254999,
+      "grad_norm": 0.4612601399421692,
+      "learning_rate": 1.302325581395349e-06,
+      "loss": 0.5276,
+      "step": 84
+    },
+    {
+      "epoch": 0.039528755231747016,
+      "grad_norm": 0.48117271065711975,
+      "learning_rate": 1.3178294573643414e-06,
+      "loss": 0.5462,
+      "step": 85
+    },
+    {
+      "epoch": 0.03999379941094404,
+      "grad_norm": 0.464334636926651,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 0.5255,
+      "step": 86
+    },
+    {
+      "epoch": 0.04045884359014106,
+      "grad_norm": 0.5120902061462402,
+      "learning_rate": 1.3488372093023258e-06,
+      "loss": 0.553,
+      "step": 87
+    },
+    {
+      "epoch": 0.040923887769338085,
+      "grad_norm": 0.5096696019172668,
+      "learning_rate": 1.3643410852713179e-06,
+      "loss": 0.5242,
+      "step": 88
+    },
+    {
+      "epoch": 0.04138893194853511,
+      "grad_norm": 0.44623059034347534,
+      "learning_rate": 1.3798449612403103e-06,
+      "loss": 0.523,
+      "step": 89
+    },
+    {
+      "epoch": 0.04185397612773213,
+      "grad_norm": 0.47654956579208374,
+      "learning_rate": 1.3953488372093025e-06,
+      "loss": 0.5335,
+      "step": 90
+    },
+    {
+      "epoch": 0.04231902030692916,
+      "grad_norm": 0.4882866144180298,
+      "learning_rate": 1.4108527131782947e-06,
+      "loss": 0.5176,
+      "step": 91
+    },
+    {
+      "epoch": 0.042784064486126185,
+      "grad_norm": 0.46192339062690735,
+      "learning_rate": 1.426356589147287e-06,
+      "loss": 0.545,
+      "step": 92
+    },
+    {
+      "epoch": 0.04324910866532321,
+      "grad_norm": 0.4416024684906006,
+      "learning_rate": 1.4418604651162794e-06,
+      "loss": 0.526,
+      "step": 93
+    },
+    {
+      "epoch": 0.04371415284452023,
+      "grad_norm": 0.45274990797042847,
+      "learning_rate": 1.4573643410852714e-06,
+      "loss": 0.5233,
+      "step": 94
+    },
+    {
+      "epoch": 0.044179197023717254,
+      "grad_norm": 0.4544108808040619,
+      "learning_rate": 1.4728682170542638e-06,
+      "loss": 0.5118,
+      "step": 95
+    },
+    {
+      "epoch": 0.04464424120291428,
+      "grad_norm": 0.4827730655670166,
+      "learning_rate": 1.4883720930232558e-06,
+      "loss": 0.5087,
+      "step": 96
+    },
+    {
+      "epoch": 0.0451092853821113,
+      "grad_norm": 0.45739415287971497,
+      "learning_rate": 1.5038759689922483e-06,
+      "loss": 0.5058,
+      "step": 97
+    },
+    {
+      "epoch": 0.04557432956130832,
+      "grad_norm": 0.4581778645515442,
+      "learning_rate": 1.5193798449612405e-06,
+      "loss": 0.5272,
+      "step": 98
+    },
+    {
+      "epoch": 0.046039373740505346,
+      "grad_norm": 0.43706098198890686,
+      "learning_rate": 1.534883720930233e-06,
+      "loss": 0.5205,
+      "step": 99
+    },
+    {
+      "epoch": 0.04650441791970237,
+      "grad_norm": 0.43774181604385376,
+      "learning_rate": 1.550387596899225e-06,
+      "loss": 0.5033,
+      "step": 100
+    },
+    {
+      "epoch": 0.04696946209889939,
+      "grad_norm": 0.42946186661720276,
+      "learning_rate": 1.5658914728682173e-06,
+      "loss": 0.5567,
+      "step": 101
+    },
+    {
+      "epoch": 0.04743450627809642,
+      "grad_norm": 0.4620288908481598,
+      "learning_rate": 1.5813953488372093e-06,
+      "loss": 0.4879,
+      "step": 102
+    },
+    {
+      "epoch": 0.047899550457293445,
+      "grad_norm": 0.4660142958164215,
+      "learning_rate": 1.5968992248062018e-06,
+      "loss": 0.5373,
+      "step": 103
+    },
+    {
+      "epoch": 0.04836459463649047,
+      "grad_norm": 0.4969952404499054,
+      "learning_rate": 1.612403100775194e-06,
+      "loss": 0.5156,
+      "step": 104
+    },
+    {
+      "epoch": 0.04882963881568749,
+      "grad_norm": 0.4132183790206909,
+      "learning_rate": 1.6279069767441862e-06,
+      "loss": 0.496,
+      "step": 105
+    },
+    {
+      "epoch": 0.049294682994884514,
+      "grad_norm": 0.433496356010437,
+      "learning_rate": 1.6434108527131784e-06,
+      "loss": 0.5313,
+      "step": 106
+    },
+    {
+      "epoch": 0.04975972717408154,
+      "grad_norm": 0.4335452914237976,
+      "learning_rate": 1.6589147286821709e-06,
+      "loss": 0.5167,
+      "step": 107
+    },
+    {
+      "epoch": 0.05022477135327856,
+      "grad_norm": 0.43790504336357117,
+      "learning_rate": 1.6744186046511629e-06,
+      "loss": 0.4973,
+      "step": 108
+    },
+    {
+      "epoch": 0.050689815532475584,
+      "grad_norm": 0.4186980724334717,
+      "learning_rate": 1.6899224806201553e-06,
+      "loss": 0.4954,
+      "step": 109
+    },
+    {
+      "epoch": 0.05115485971167261,
+      "grad_norm": 0.40744200348854065,
+      "learning_rate": 1.7054263565891473e-06,
+      "loss": 0.4927,
+      "step": 110
+    },
+    {
+      "epoch": 0.05161990389086963,
+      "grad_norm": 0.43091249465942383,
+      "learning_rate": 1.7209302325581397e-06,
+      "loss": 0.5062,
+      "step": 111
+    },
+    {
+      "epoch": 0.05208494807006666,
+      "grad_norm": 0.4677499830722809,
+      "learning_rate": 1.736434108527132e-06,
+      "loss": 0.5107,
+      "step": 112
+    },
+    {
+      "epoch": 0.05254999224926368,
+      "grad_norm": 0.42137694358825684,
+      "learning_rate": 1.7519379844961242e-06,
+      "loss": 0.5132,
+      "step": 113
+    },
+    {
+      "epoch": 0.053015036428460706,
+      "grad_norm": 0.4344700872898102,
+      "learning_rate": 1.7674418604651164e-06,
+      "loss": 0.4793,
+      "step": 114
+    },
+    {
+      "epoch": 0.05348008060765773,
+      "grad_norm": 0.4326712489128113,
+      "learning_rate": 1.7829457364341088e-06,
+      "loss": 0.4715,
+      "step": 115
+    },
+    {
+      "epoch": 0.05394512478685475,
+      "grad_norm": 0.4487822949886322,
+      "learning_rate": 1.7984496124031008e-06,
+      "loss": 0.5165,
+      "step": 116
+    },
+    {
+      "epoch": 0.054410168966051775,
+      "grad_norm": 0.46336719393730164,
+      "learning_rate": 1.8139534883720933e-06,
+      "loss": 0.512,
+      "step": 117
+    },
+    {
+      "epoch": 0.0548752131452488,
+      "grad_norm": 0.4590426981449127,
+      "learning_rate": 1.8294573643410855e-06,
+      "loss": 0.5036,
+      "step": 118
+    },
+    {
+      "epoch": 0.05534025732444582,
+      "grad_norm": 0.4003012478351593,
+      "learning_rate": 1.8449612403100777e-06,
+      "loss": 0.4894,
+      "step": 119
+    },
+    {
+      "epoch": 0.055805301503642844,
+      "grad_norm": 0.45950761437416077,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 0.5109,
+      "step": 120
+    },
+    {
+      "epoch": 0.05627034568283987,
+      "grad_norm": 0.43410784006118774,
+      "learning_rate": 1.8759689922480624e-06,
+      "loss": 0.4966,
+      "step": 121
+    },
+    {
+      "epoch": 0.05673538986203689,
+      "grad_norm": 0.45534461736679077,
+      "learning_rate": 1.8914728682170544e-06,
+      "loss": 0.5311,
+      "step": 122
+    },
+    {
+      "epoch": 0.05720043404123392,
+      "grad_norm": 0.44469308853149414,
+      "learning_rate": 1.9069767441860468e-06,
+      "loss": 0.4967,
+      "step": 123
+    },
+    {
+      "epoch": 0.057665478220430944,
+      "grad_norm": 0.4573187232017517,
+      "learning_rate": 1.922480620155039e-06,
+      "loss": 0.4864,
+      "step": 124
+    },
+    {
+      "epoch": 0.05813052239962797,
+      "grad_norm": 0.45671388506889343,
+      "learning_rate": 1.9379844961240315e-06,
+      "loss": 0.5046,
+      "step": 125
+    },
+    {
+      "epoch": 0.05859556657882499,
+      "grad_norm": 0.4924532473087311,
+      "learning_rate": 1.9534883720930235e-06,
+      "loss": 0.5474,
+      "step": 126
+    },
+    {
+      "epoch": 0.05906061075802201,
+      "grad_norm": 0.45947206020355225,
+      "learning_rate": 1.968992248062016e-06,
+      "loss": 0.463,
+      "step": 127
+    },
+    {
+      "epoch": 0.059525654937219036,
+      "grad_norm": 0.4114924967288971,
+      "learning_rate": 1.984496124031008e-06,
+      "loss": 0.5045,
+      "step": 128
+    },
+    {
+      "epoch": 0.05999069911641606,
+      "grad_norm": 0.424427330493927,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.4865,
+      "step": 129
+    },
+    {
+      "epoch": 0.06045574329561308,
+      "grad_norm": 0.457966685295105,
+      "learning_rate": 2.0155038759689923e-06,
+      "loss": 0.492,
+      "step": 130
+    },
+    {
+      "epoch": 0.060920787474810105,
+      "grad_norm": 0.4255077540874481,
+      "learning_rate": 2.0310077519379848e-06,
+      "loss": 0.4961,
+      "step": 131
+    },
+    {
+      "epoch": 0.06138583165400713,
+      "grad_norm": 0.39936286211013794,
+      "learning_rate": 2.0465116279069768e-06,
+      "loss": 0.4595,
+      "step": 132
+    },
+    {
+      "epoch": 0.06185087583320415,
+      "grad_norm": 0.39685145020484924,
+      "learning_rate": 2.062015503875969e-06,
+      "loss": 0.4645,
+      "step": 133
+    },
+    {
+      "epoch": 0.06231592001240118,
+      "grad_norm": 0.44578003883361816,
+      "learning_rate": 2.0775193798449612e-06,
+      "loss": 0.4848,
+      "step": 134
+    },
+    {
+      "epoch": 0.0627809641915982,
+      "grad_norm": 0.4608939290046692,
+      "learning_rate": 2.0930232558139536e-06,
+      "loss": 0.444,
+      "step": 135
+    },
+    {
+      "epoch": 0.06324600837079522,
+      "grad_norm": 0.43158158659935,
+      "learning_rate": 2.108527131782946e-06,
+      "loss": 0.4781,
+      "step": 136
+    },
+    {
+      "epoch": 0.06371105254999225,
+      "grad_norm": 0.43560659885406494,
+      "learning_rate": 2.124031007751938e-06,
+      "loss": 0.4768,
+      "step": 137
+    },
+    {
+      "epoch": 0.06417609672918927,
+      "grad_norm": 0.39161673188209534,
+      "learning_rate": 2.1395348837209305e-06,
+      "loss": 0.4824,
+      "step": 138
+    },
+    {
+      "epoch": 0.0646411409083863,
+      "grad_norm": 0.492993026971817,
+      "learning_rate": 2.155038759689923e-06,
+      "loss": 0.4708,
+      "step": 139
+    },
+    {
+      "epoch": 0.06510618508758333,
+      "grad_norm": 0.4824112057685852,
+      "learning_rate": 2.170542635658915e-06,
+      "loss": 0.5064,
+      "step": 140
+    },
+    {
+      "epoch": 0.06557122926678034,
+      "grad_norm": 0.42813244462013245,
+      "learning_rate": 2.1860465116279074e-06,
+      "loss": 0.4815,
+      "step": 141
+    },
+    {
+      "epoch": 0.06603627344597737,
+      "grad_norm": 0.43404629826545715,
+      "learning_rate": 2.2015503875968994e-06,
+      "loss": 0.4808,
+      "step": 142
+    },
+    {
+      "epoch": 0.06650131762517439,
+      "grad_norm": 0.44562390446662903,
+      "learning_rate": 2.217054263565892e-06,
+      "loss": 0.495,
+      "step": 143
+    },
+    {
+      "epoch": 0.06696636180437142,
+      "grad_norm": 0.43140050768852234,
+      "learning_rate": 2.232558139534884e-06,
+      "loss": 0.4634,
+      "step": 144
+    },
+    {
+      "epoch": 0.06743140598356843,
+      "grad_norm": 0.43824145197868347,
+      "learning_rate": 2.2480620155038763e-06,
+      "loss": 0.5077,
+      "step": 145
+    },
+    {
+      "epoch": 0.06789645016276546,
+      "grad_norm": 0.45454517006874084,
+      "learning_rate": 2.2635658914728683e-06,
+      "loss": 0.4605,
+      "step": 146
+    },
+    {
+      "epoch": 0.06836149434196248,
+      "grad_norm": 0.4356607496738434,
+      "learning_rate": 2.2790697674418607e-06,
+      "loss": 0.5132,
+      "step": 147
+    },
+    {
+      "epoch": 0.06882653852115951,
+      "grad_norm": 0.4299354553222656,
+      "learning_rate": 2.2945736434108527e-06,
+      "loss": 0.4709,
+      "step": 148
+    },
+    {
+      "epoch": 0.06929158270035653,
+      "grad_norm": 0.4432096481323242,
+      "learning_rate": 2.310077519379845e-06,
+      "loss": 0.4743,
+      "step": 149
+    },
+    {
+      "epoch": 0.06975662687955356,
+      "grad_norm": 0.4662824273109436,
+      "learning_rate": 2.3255813953488376e-06,
+      "loss": 0.5184,
+      "step": 150
+    },
+    {
+      "epoch": 0.07022167105875059,
+      "grad_norm": 0.4367971420288086,
+      "learning_rate": 2.3410852713178296e-06,
+      "loss": 0.4778,
+      "step": 151
+    },
+    {
+      "epoch": 0.0706867152379476,
+      "grad_norm": 0.4500759243965149,
+      "learning_rate": 2.356589147286822e-06,
+      "loss": 0.4534,
+      "step": 152
+    },
+    {
+      "epoch": 0.07115175941714463,
+      "grad_norm": 0.4312923848628998,
+      "learning_rate": 2.3720930232558144e-06,
+      "loss": 0.5134,
+      "step": 153
+    },
+    {
+      "epoch": 0.07161680359634165,
+      "grad_norm": 0.4794353246688843,
+      "learning_rate": 2.3875968992248065e-06,
+      "loss": 0.4641,
+      "step": 154
+    },
+    {
+      "epoch": 0.07208184777553868,
+      "grad_norm": 0.4282940924167633,
+      "learning_rate": 2.403100775193799e-06,
+      "loss": 0.4693,
+      "step": 155
+    },
+    {
+      "epoch": 0.0725468919547357,
+      "grad_norm": 0.4426284730434418,
+      "learning_rate": 2.418604651162791e-06,
+      "loss": 0.4785,
+      "step": 156
+    },
+    {
+      "epoch": 0.07301193613393273,
+      "grad_norm": 0.45532897114753723,
+      "learning_rate": 2.4341085271317833e-06,
+      "loss": 0.4934,
+      "step": 157
+    },
+    {
+      "epoch": 0.07347698031312974,
+      "grad_norm": 0.5064568519592285,
+      "learning_rate": 2.4496124031007753e-06,
+      "loss": 0.495,
+      "step": 158
+    },
+    {
+      "epoch": 0.07394202449232677,
+      "grad_norm": 0.4576449692249298,
+      "learning_rate": 2.4651162790697678e-06,
+      "loss": 0.4551,
+      "step": 159
+    },
+    {
+      "epoch": 0.0744070686715238,
+      "grad_norm": 0.4424917697906494,
+      "learning_rate": 2.4806201550387598e-06,
+      "loss": 0.4861,
+      "step": 160
+    },
+    {
+      "epoch": 0.07487211285072082,
+      "grad_norm": 0.48563066124916077,
+      "learning_rate": 2.496124031007752e-06,
+      "loss": 0.4614,
+      "step": 161
+    },
+    {
+      "epoch": 0.07533715702991785,
+      "grad_norm": 0.43007153272628784,
+      "learning_rate": 2.5116279069767446e-06,
+      "loss": 0.4822,
+      "step": 162
+    },
+    {
+      "epoch": 0.07580220120911486,
+      "grad_norm": 0.41486161947250366,
+      "learning_rate": 2.5271317829457366e-06,
+      "loss": 0.4713,
+      "step": 163
+    },
+    {
+      "epoch": 0.0762672453883119,
+      "grad_norm": 0.39590543508529663,
+      "learning_rate": 2.5426356589147286e-06,
+      "loss": 0.5051,
+      "step": 164
+    },
+    {
+      "epoch": 0.07673228956750891,
+      "grad_norm": 0.4512157738208771,
+      "learning_rate": 2.558139534883721e-06,
+      "loss": 0.4597,
+      "step": 165
+    },
+    {
+      "epoch": 0.07719733374670594,
+      "grad_norm": 0.5571392178535461,
+      "learning_rate": 2.5736434108527135e-06,
+      "loss": 0.4783,
+      "step": 166
+    },
+    {
+      "epoch": 0.07766237792590296,
+      "grad_norm": 0.41649627685546875,
+      "learning_rate": 2.5891472868217055e-06,
+      "loss": 0.4648,
+      "step": 167
+    },
+    {
+      "epoch": 0.07812742210509999,
+      "grad_norm": 0.45857080817222595,
+      "learning_rate": 2.604651162790698e-06,
+      "loss": 0.4793,
+      "step": 168
+    },
+    {
+      "epoch": 0.078592466284297,
+      "grad_norm": 0.42742466926574707,
+      "learning_rate": 2.6201550387596904e-06,
+      "loss": 0.4894,
+      "step": 169
+    },
+    {
+      "epoch": 0.07905751046349403,
+      "grad_norm": 0.48390820622444153,
+      "learning_rate": 2.635658914728683e-06,
+      "loss": 0.4707,
+      "step": 170
+    },
+    {
+      "epoch": 0.07952255464269106,
+      "grad_norm": 0.5031663775444031,
+      "learning_rate": 2.6511627906976744e-06,
+      "loss": 0.4591,
+      "step": 171
+    },
+    {
+      "epoch": 0.07998759882188808,
+      "grad_norm": 0.4769449830055237,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 0.4735,
+      "step": 172
+    },
+    {
+      "epoch": 0.08045264300108511,
+      "grad_norm": 0.4525051712989807,
+      "learning_rate": 2.6821705426356593e-06,
+      "loss": 0.4655,
+      "step": 173
+    },
+    {
+      "epoch": 0.08091768718028212,
+      "grad_norm": 0.4477654695510864,
+      "learning_rate": 2.6976744186046517e-06,
+      "loss": 0.461,
+      "step": 174
+    },
+    {
+      "epoch": 0.08138273135947915,
+      "grad_norm": 0.45438140630722046,
+      "learning_rate": 2.7131782945736433e-06,
+      "loss": 0.4807,
+      "step": 175
+    },
+    {
+      "epoch": 0.08184777553867617,
+      "grad_norm": 0.47995445132255554,
+      "learning_rate": 2.7286821705426357e-06,
+      "loss": 0.4768,
+      "step": 176
+    },
+    {
+      "epoch": 0.0823128197178732,
+      "grad_norm": 0.40142083168029785,
+      "learning_rate": 2.744186046511628e-06,
+      "loss": 0.4814,
+      "step": 177
+    },
+    {
+      "epoch": 0.08277786389707022,
+      "grad_norm": 0.4523696005344391,
+      "learning_rate": 2.7596899224806206e-06,
+      "loss": 0.4561,
+      "step": 178
+    },
+    {
+      "epoch": 0.08324290807626725,
+      "grad_norm": 0.4552527368068695,
+      "learning_rate": 2.7751937984496126e-06,
+      "loss": 0.4796,
+      "step": 179
+    },
+    {
+      "epoch": 0.08370795225546426,
+      "grad_norm": 0.4601724445819855,
+      "learning_rate": 2.790697674418605e-06,
+      "loss": 0.4733,
+      "step": 180
+    },
+    {
+      "epoch": 0.08417299643466129,
+      "grad_norm": 0.4687198996543884,
+      "learning_rate": 2.8062015503875974e-06,
+      "loss": 0.4648,
+      "step": 181
+    },
+    {
+      "epoch": 0.08463804061385832,
+      "grad_norm": 0.4488019347190857,
+      "learning_rate": 2.8217054263565894e-06,
+      "loss": 0.4794,
+      "step": 182
+    },
+    {
+      "epoch": 0.08510308479305534,
+      "grad_norm": 0.462455153465271,
+      "learning_rate": 2.8372093023255815e-06,
+      "loss": 0.4699,
+      "step": 183
+    },
+    {
+      "epoch": 0.08556812897225237,
+      "grad_norm": 0.43786558508872986,
+      "learning_rate": 2.852713178294574e-06,
+      "loss": 0.4807,
+      "step": 184
+    },
+    {
+      "epoch": 0.08603317315144939,
+      "grad_norm": 0.4396842122077942,
+      "learning_rate": 2.8682170542635663e-06,
+      "loss": 0.4744,
+      "step": 185
+    },
+    {
+      "epoch": 0.08649821733064642,
+      "grad_norm": 0.40669193863868713,
+      "learning_rate": 2.8837209302325587e-06,
+      "loss": 0.5,
+      "step": 186
+    },
+    {
+      "epoch": 0.08696326150984343,
+      "grad_norm": 0.43069615960121155,
+      "learning_rate": 2.8992248062015503e-06,
+      "loss": 0.4666,
+      "step": 187
+    },
+    {
+      "epoch": 0.08742830568904046,
+      "grad_norm": 0.3929974138736725,
+      "learning_rate": 2.9147286821705428e-06,
+      "loss": 0.4599,
+      "step": 188
+    },
+    {
+      "epoch": 0.08789334986823748,
+      "grad_norm": 0.44300031661987305,
+      "learning_rate": 2.930232558139535e-06,
+      "loss": 0.4441,
+      "step": 189
+    },
+    {
+      "epoch": 0.08835839404743451,
+      "grad_norm": 0.40551915764808655,
+      "learning_rate": 2.9457364341085276e-06,
+      "loss": 0.4536,
+      "step": 190
+    },
+    {
+      "epoch": 0.08882343822663152,
+      "grad_norm": 0.45988717675209045,
+      "learning_rate": 2.9612403100775196e-06,
+      "loss": 0.4442,
+      "step": 191
+    },
+    {
+      "epoch": 0.08928848240582855,
+      "grad_norm": 0.41964244842529297,
+      "learning_rate": 2.9767441860465116e-06,
+      "loss": 0.4294,
+      "step": 192
+    },
+    {
+      "epoch": 0.08975352658502558,
+      "grad_norm": 0.4278603196144104,
+      "learning_rate": 2.992248062015504e-06,
+      "loss": 0.4512,
+      "step": 193
+    },
+    {
+      "epoch": 0.0902185707642226,
+      "grad_norm": 0.4759548008441925,
+      "learning_rate": 3.0077519379844965e-06,
+      "loss": 0.4856,
+      "step": 194
+    },
+    {
+      "epoch": 0.09068361494341963,
+      "grad_norm": 0.44720789790153503,
+      "learning_rate": 3.0232558139534885e-06,
+      "loss": 0.439,
+      "step": 195
+    },
+    {
+      "epoch": 0.09114865912261665,
+      "grad_norm": 0.4375242590904236,
+      "learning_rate": 3.038759689922481e-06,
+      "loss": 0.4559,
+      "step": 196
+    },
+    {
+      "epoch": 0.09161370330181368,
+      "grad_norm": 0.4382531940937042,
+      "learning_rate": 3.0542635658914734e-06,
+      "loss": 0.4651,
+      "step": 197
+    },
+    {
+      "epoch": 0.09207874748101069,
+      "grad_norm": 0.46348533034324646,
+      "learning_rate": 3.069767441860466e-06,
+      "loss": 0.4594,
+      "step": 198
+    },
+    {
+      "epoch": 0.09254379166020772,
+      "grad_norm": 0.5095227360725403,
+      "learning_rate": 3.0852713178294574e-06,
+      "loss": 0.4621,
+      "step": 199
+    },
+    {
+      "epoch": 0.09300883583940474,
+      "grad_norm": 0.5287566781044006,
+      "learning_rate": 3.10077519379845e-06,
+      "loss": 0.4805,
+      "step": 200
+    },
+    {
+      "epoch": 0.09347388001860177,
+      "grad_norm": 0.49240946769714355,
+      "learning_rate": 3.1162790697674423e-06,
+      "loss": 0.4836,
+      "step": 201
+    },
+    {
+      "epoch": 0.09393892419779878,
+      "grad_norm": 0.4462595283985138,
+      "learning_rate": 3.1317829457364347e-06,
+      "loss": 0.4466,
+      "step": 202
+    },
+    {
+      "epoch": 0.09440396837699581,
+      "grad_norm": 0.5251316428184509,
+      "learning_rate": 3.1472868217054263e-06,
+      "loss": 0.4754,
+      "step": 203
+    },
+    {
+      "epoch": 0.09486901255619284,
+      "grad_norm": 0.516490638256073,
+      "learning_rate": 3.1627906976744187e-06,
+      "loss": 0.4768,
+      "step": 204
+    },
+    {
+      "epoch": 0.09533405673538986,
+      "grad_norm": 0.455197274684906,
+      "learning_rate": 3.178294573643411e-06,
+      "loss": 0.449,
+      "step": 205
+    },
+    {
+      "epoch": 0.09579910091458689,
+      "grad_norm": 0.4805566966533661,
+      "learning_rate": 3.1937984496124036e-06,
+      "loss": 0.4775,
+      "step": 206
+    },
+    {
+      "epoch": 0.0962641450937839,
+      "grad_norm": 0.49571681022644043,
+      "learning_rate": 3.2093023255813956e-06,
+      "loss": 0.4297,
+      "step": 207
+    },
+    {
+      "epoch": 0.09672918927298094,
+      "grad_norm": 0.43034252524375916,
+      "learning_rate": 3.224806201550388e-06,
+      "loss": 0.4941,
+      "step": 208
+    },
+    {
+      "epoch": 0.09719423345217795,
+      "grad_norm": 0.539557933807373,
+      "learning_rate": 3.24031007751938e-06,
+      "loss": 0.4412,
+      "step": 209
+    },
+    {
+      "epoch": 0.09765927763137498,
+      "grad_norm": 0.5281848907470703,
+      "learning_rate": 3.2558139534883724e-06,
+      "loss": 0.468,
+      "step": 210
+    },
+    {
+      "epoch": 0.098124321810572,
+      "grad_norm": 0.4832194745540619,
+      "learning_rate": 3.2713178294573644e-06,
+      "loss": 0.4691,
+      "step": 211
+    },
+    {
+      "epoch": 0.09858936598976903,
+      "grad_norm": 0.4533843696117401,
+      "learning_rate": 3.286821705426357e-06,
+      "loss": 0.4746,
+      "step": 212
+    },
+    {
+      "epoch": 0.09905441016896605,
+      "grad_norm": 0.509421169757843,
+      "learning_rate": 3.3023255813953493e-06,
+      "loss": 0.4646,
+      "step": 213
+    },
+    {
+      "epoch": 0.09951945434816308,
+      "grad_norm": 0.4463287889957428,
+      "learning_rate": 3.3178294573643417e-06,
+      "loss": 0.4763,
+      "step": 214
+    },
+    {
+      "epoch": 0.0999844985273601,
+      "grad_norm": 0.443873792886734,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.443,
+      "step": 215
+    },
+    {
+      "epoch": 0.10044954270655712,
+      "grad_norm": 0.5233954191207886,
+      "learning_rate": 3.3488372093023258e-06,
+      "loss": 0.4574,
+      "step": 216
+    },
+    {
+      "epoch": 0.10091458688575415,
+      "grad_norm": 0.4142616093158722,
+      "learning_rate": 3.364341085271318e-06,
+      "loss": 0.4304,
+      "step": 217
+    },
+    {
+      "epoch": 0.10137963106495117,
+      "grad_norm": 0.4118039309978485,
+      "learning_rate": 3.3798449612403106e-06,
+      "loss": 0.4272,
+      "step": 218
+    },
+    {
+      "epoch": 0.1018446752441482,
+      "grad_norm": 0.4856218099594116,
+      "learning_rate": 3.3953488372093026e-06,
+      "loss": 0.4834,
+      "step": 219
+    },
+    {
+      "epoch": 0.10230971942334521,
+      "grad_norm": 0.4576835036277771,
+      "learning_rate": 3.4108527131782946e-06,
+      "loss": 0.4216,
+      "step": 220
+    },
+    {
+      "epoch": 0.10277476360254224,
+      "grad_norm": 0.5164148211479187,
+      "learning_rate": 3.426356589147287e-06,
+      "loss": 0.4545,
+      "step": 221
+    },
+    {
+      "epoch": 0.10323980778173926,
+      "grad_norm": 0.4515243470668793,
+      "learning_rate": 3.4418604651162795e-06,
+      "loss": 0.4515,
+      "step": 222
+    },
+    {
+      "epoch": 0.10370485196093629,
+      "grad_norm": 0.4391280710697174,
+      "learning_rate": 3.4573643410852715e-06,
+      "loss": 0.4466,
+      "step": 223
+    },
+    {
+      "epoch": 0.10416989614013332,
+      "grad_norm": 0.41174498200416565,
+      "learning_rate": 3.472868217054264e-06,
+      "loss": 0.4424,
+      "step": 224
+    },
+    {
+      "epoch": 0.10463494031933034,
+      "grad_norm": 0.45546606183052063,
+      "learning_rate": 3.4883720930232564e-06,
+      "loss": 0.481,
+      "step": 225
+    },
+    {
+      "epoch": 0.10509998449852737,
+      "grad_norm": 0.4433421194553375,
+      "learning_rate": 3.5038759689922484e-06,
+      "loss": 0.4586,
+      "step": 226
+    },
+    {
+      "epoch": 0.10556502867772438,
+      "grad_norm": 0.46707531809806824,
+      "learning_rate": 3.5193798449612404e-06,
+      "loss": 0.4627,
+      "step": 227
+    },
+    {
+      "epoch": 0.10603007285692141,
+      "grad_norm": 0.48355457186698914,
+      "learning_rate": 3.534883720930233e-06,
+      "loss": 0.4834,
+      "step": 228
+    },
+    {
+      "epoch": 0.10649511703611843,
+      "grad_norm": 0.5134936571121216,
+      "learning_rate": 3.5503875968992252e-06,
+      "loss": 0.4496,
+      "step": 229
+    },
+    {
+      "epoch": 0.10696016121531546,
+      "grad_norm": 0.4306149184703827,
+      "learning_rate": 3.5658914728682177e-06,
+      "loss": 0.4689,
+      "step": 230
+    },
+    {
+      "epoch": 0.10742520539451247,
+      "grad_norm": 0.4603032171726227,
+      "learning_rate": 3.5813953488372093e-06,
+      "loss": 0.4489,
+      "step": 231
+    },
+    {
+      "epoch": 0.1078902495737095,
+      "grad_norm": 0.46613025665283203,
+      "learning_rate": 3.5968992248062017e-06,
+      "loss": 0.4648,
+      "step": 232
+    },
+    {
+      "epoch": 0.10835529375290652,
+      "grad_norm": 0.497821182012558,
+      "learning_rate": 3.612403100775194e-06,
+      "loss": 0.4454,
+      "step": 233
+    },
+    {
+      "epoch": 0.10882033793210355,
+      "grad_norm": 0.44097408652305603,
+      "learning_rate": 3.6279069767441866e-06,
+      "loss": 0.4443,
+      "step": 234
+    },
+    {
+      "epoch": 0.10928538211130058,
+      "grad_norm": 0.4849492013454437,
+      "learning_rate": 3.6434108527131786e-06,
+      "loss": 0.4546,
+      "step": 235
+    },
+    {
+      "epoch": 0.1097504262904976,
+      "grad_norm": 0.4576839506626129,
+      "learning_rate": 3.658914728682171e-06,
+      "loss": 0.44,
+      "step": 236
+    },
+    {
+      "epoch": 0.11021547046969463,
+      "grad_norm": 0.44836199283599854,
+      "learning_rate": 3.674418604651163e-06,
+      "loss": 0.4779,
+      "step": 237
+    },
+    {
+      "epoch": 0.11068051464889164,
+      "grad_norm": 0.46997904777526855,
+      "learning_rate": 3.6899224806201554e-06,
+      "loss": 0.4755,
+      "step": 238
+    },
+    {
+      "epoch": 0.11114555882808867,
+      "grad_norm": 0.4699132740497589,
+      "learning_rate": 3.7054263565891474e-06,
+      "loss": 0.4279,
+      "step": 239
+    },
+    {
+      "epoch": 0.11161060300728569,
+      "grad_norm": 0.4158969819545746,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 0.4462,
+      "step": 240
+    },
+    {
+      "epoch": 0.11207564718648272,
+      "grad_norm": 0.4518444240093231,
+      "learning_rate": 3.7364341085271323e-06,
+      "loss": 0.4419,
+      "step": 241
+    },
+    {
+      "epoch": 0.11254069136567973,
+      "grad_norm": 0.43959298729896545,
+      "learning_rate": 3.7519379844961247e-06,
+      "loss": 0.4277,
+      "step": 242
+    },
+    {
+      "epoch": 0.11300573554487676,
+      "grad_norm": 0.46462079882621765,
+      "learning_rate": 3.7674418604651163e-06,
+      "loss": 0.4666,
+      "step": 243
+    },
+    {
+      "epoch": 0.11347077972407378,
+      "grad_norm": 0.44643181562423706,
+      "learning_rate": 3.7829457364341087e-06,
+      "loss": 0.4542,
+      "step": 244
+    },
+    {
+      "epoch": 0.11393582390327081,
+      "grad_norm": 0.4474817216396332,
+      "learning_rate": 3.798449612403101e-06,
+      "loss": 0.4457,
+      "step": 245
+    },
+    {
+      "epoch": 0.11440086808246784,
+      "grad_norm": 0.440893292427063,
+      "learning_rate": 3.8139534883720936e-06,
+      "loss": 0.4567,
+      "step": 246
+    },
+    {
+      "epoch": 0.11486591226166486,
+      "grad_norm": 0.5157288908958435,
+      "learning_rate": 3.829457364341085e-06,
+      "loss": 0.4668,
+      "step": 247
+    },
+    {
+      "epoch": 0.11533095644086189,
+      "grad_norm": 0.41228559613227844,
+      "learning_rate": 3.844961240310078e-06,
+      "loss": 0.4532,
+      "step": 248
+    },
+    {
+      "epoch": 0.1157960006200589,
+      "grad_norm": 0.49492067098617554,
+      "learning_rate": 3.86046511627907e-06,
+      "loss": 0.4717,
+      "step": 249
+    },
+    {
+      "epoch": 0.11626104479925593,
+      "grad_norm": 0.3908706605434418,
+      "learning_rate": 3.875968992248063e-06,
+      "loss": 0.4363,
+      "step": 250
+    },
+    {
+      "epoch": 0.11672608897845295,
+      "grad_norm": 0.4512355625629425,
+      "learning_rate": 3.891472868217054e-06,
+      "loss": 0.461,
+      "step": 251
+    },
+    {
+      "epoch": 0.11719113315764998,
+      "grad_norm": 0.45557087659835815,
+      "learning_rate": 3.906976744186047e-06,
+      "loss": 0.4479,
+      "step": 252
+    },
+    {
+      "epoch": 0.117656177336847,
+      "grad_norm": 0.42376312613487244,
+      "learning_rate": 3.922480620155039e-06,
+      "loss": 0.4614,
+      "step": 253
+    },
+    {
+      "epoch": 0.11812122151604403,
+      "grad_norm": 0.494684636592865,
+      "learning_rate": 3.937984496124032e-06,
+      "loss": 0.4528,
+      "step": 254
+    },
+    {
+      "epoch": 0.11858626569524104,
+      "grad_norm": 0.46942609548568726,
+      "learning_rate": 3.953488372093024e-06,
+      "loss": 0.4363,
+      "step": 255
+    },
+    {
+      "epoch": 0.11905130987443807,
+      "grad_norm": 0.4552278518676758,
+      "learning_rate": 3.968992248062016e-06,
+      "loss": 0.4462,
+      "step": 256
+    },
+    {
+      "epoch": 0.1195163540536351,
+      "grad_norm": 0.4802659749984741,
+      "learning_rate": 3.984496124031008e-06,
+      "loss": 0.4694,
+      "step": 257
+    },
+    {
+      "epoch": 0.11998139823283212,
+      "grad_norm": 0.49204587936401367,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.449,
+      "step": 258
+    },
+    {
+      "epoch": 0.12044644241202915,
+      "grad_norm": 0.4670722484588623,
+      "learning_rate": 4.015503875968993e-06,
+      "loss": 0.4624,
+      "step": 259
+    },
+    {
+      "epoch": 0.12091148659122616,
+      "grad_norm": 0.4534282386302948,
+      "learning_rate": 4.031007751937985e-06,
+      "loss": 0.4338,
+      "step": 260
+    },
+    {
+      "epoch": 0.1213765307704232,
+      "grad_norm": 0.4467426538467407,
+      "learning_rate": 4.0465116279069775e-06,
+      "loss": 0.4513,
+      "step": 261
+    },
+    {
+      "epoch": 0.12184157494962021,
+      "grad_norm": 0.4626162648200989,
+      "learning_rate": 4.0620155038759695e-06,
+      "loss": 0.4714,
+      "step": 262
+    },
+    {
+      "epoch": 0.12230661912881724,
+      "grad_norm": 0.45661380887031555,
+      "learning_rate": 4.0775193798449616e-06,
+      "loss": 0.4559,
+      "step": 263
+    },
+    {
+      "epoch": 0.12277166330801426,
+      "grad_norm": 0.5042157769203186,
+      "learning_rate": 4.0930232558139536e-06,
+      "loss": 0.4717,
+      "step": 264
+    },
+    {
+      "epoch": 0.12323670748721129,
+      "grad_norm": 0.5059670805931091,
+      "learning_rate": 4.108527131782946e-06,
+      "loss": 0.4517,
+      "step": 265
+    },
+    {
+      "epoch": 0.1237017516664083,
+      "grad_norm": 0.4609099328517914,
+      "learning_rate": 4.124031007751938e-06,
+      "loss": 0.4458,
+      "step": 266
+    },
+    {
+      "epoch": 0.12416679584560533,
+      "grad_norm": 0.5168102979660034,
+      "learning_rate": 4.1395348837209304e-06,
+      "loss": 0.4068,
+      "step": 267
+    },
+    {
+      "epoch": 0.12463184002480236,
+      "grad_norm": 0.4813331961631775,
+      "learning_rate": 4.1550387596899224e-06,
+      "loss": 0.4364,
+      "step": 268
+    },
+    {
+      "epoch": 0.1250968842039994,
+      "grad_norm": 0.4824882745742798,
+      "learning_rate": 4.170542635658915e-06,
+      "loss": 0.4726,
+      "step": 269
+    },
+    {
+      "epoch": 0.1255619283831964,
+      "grad_norm": 0.4692811965942383,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 0.462,
+      "step": 270
+    },
+    {
+      "epoch": 0.12602697256239342,
+      "grad_norm": 0.473664253950119,
+      "learning_rate": 4.201550387596899e-06,
+      "loss": 0.4681,
+      "step": 271
+    },
+    {
+      "epoch": 0.12649201674159044,
+      "grad_norm": 0.42069149017333984,
+      "learning_rate": 4.217054263565892e-06,
+      "loss": 0.4428,
+      "step": 272
+    },
+    {
+      "epoch": 0.12695706092078748,
+      "grad_norm": 0.4619064927101135,
+      "learning_rate": 4.232558139534884e-06,
+      "loss": 0.4677,
+      "step": 273
+    },
+    {
+      "epoch": 0.1274221050999845,
+      "grad_norm": 0.5394362807273865,
+      "learning_rate": 4.248062015503876e-06,
+      "loss": 0.4262,
+      "step": 274
+    },
+    {
+      "epoch": 0.12788714927918152,
+      "grad_norm": 0.44251418113708496,
+      "learning_rate": 4.263565891472868e-06,
+      "loss": 0.4826,
+      "step": 275
+    },
+    {
+      "epoch": 0.12835219345837853,
+      "grad_norm": 0.40653759241104126,
+      "learning_rate": 4.279069767441861e-06,
+      "loss": 0.4229,
+      "step": 276
+    },
+    {
+      "epoch": 0.12881723763757558,
+      "grad_norm": 0.5286734104156494,
+      "learning_rate": 4.294573643410853e-06,
+      "loss": 0.4283,
+      "step": 277
+    },
+    {
+      "epoch": 0.1292822818167726,
+      "grad_norm": 0.47520795464515686,
+      "learning_rate": 4.310077519379846e-06,
+      "loss": 0.4462,
+      "step": 278
+    },
+    {
+      "epoch": 0.1297473259959696,
+      "grad_norm": 0.40976548194885254,
+      "learning_rate": 4.325581395348837e-06,
+      "loss": 0.4408,
+      "step": 279
+    },
+    {
+      "epoch": 0.13021237017516665,
+      "grad_norm": 0.46813562512397766,
+      "learning_rate": 4.34108527131783e-06,
+      "loss": 0.4555,
+      "step": 280
+    },
+    {
+      "epoch": 0.13067741435436367,
+      "grad_norm": 0.4154888689517975,
+      "learning_rate": 4.356589147286822e-06,
+      "loss": 0.4492,
+      "step": 281
+    },
+    {
+      "epoch": 0.13114245853356069,
+      "grad_norm": 0.5551090240478516,
+      "learning_rate": 4.372093023255815e-06,
+      "loss": 0.4811,
+      "step": 282
+    },
+    {
+      "epoch": 0.1316075027127577,
+      "grad_norm": 0.47705549001693726,
+      "learning_rate": 4.387596899224806e-06,
+      "loss": 0.436,
+      "step": 283
+    },
+    {
+      "epoch": 0.13207254689195475,
+      "grad_norm": 0.4338889718055725,
+      "learning_rate": 4.403100775193799e-06,
+      "loss": 0.4636,
+      "step": 284
+    },
+    {
+      "epoch": 0.13253759107115176,
+      "grad_norm": 0.49717769026756287,
+      "learning_rate": 4.418604651162791e-06,
+      "loss": 0.4257,
+      "step": 285
+    },
+    {
+      "epoch": 0.13300263525034878,
+      "grad_norm": 0.4574042558670044,
+      "learning_rate": 4.434108527131784e-06,
+      "loss": 0.4135,
+      "step": 286
+    },
+    {
+      "epoch": 0.1334676794295458,
+      "grad_norm": 0.4649648368358612,
+      "learning_rate": 4.449612403100776e-06,
+      "loss": 0.4557,
+      "step": 287
+    },
+    {
+      "epoch": 0.13393272360874284,
+      "grad_norm": 0.43360039591789246,
+      "learning_rate": 4.465116279069768e-06,
+      "loss": 0.423,
+      "step": 288
+    },
+    {
+      "epoch": 0.13439776778793985,
+      "grad_norm": 0.43359556794166565,
+      "learning_rate": 4.4806201550387605e-06,
+      "loss": 0.4346,
+      "step": 289
+    },
+    {
+      "epoch": 0.13486281196713687,
+      "grad_norm": 0.4726753532886505,
+      "learning_rate": 4.4961240310077525e-06,
+      "loss": 0.4376,
+      "step": 290
+    },
+    {
+      "epoch": 0.1353278561463339,
+      "grad_norm": 0.43972247838974,
+      "learning_rate": 4.5116279069767445e-06,
+      "loss": 0.4422,
+      "step": 291
+    },
+    {
+      "epoch": 0.13579290032553093,
+      "grad_norm": 0.4216838479042053,
+      "learning_rate": 4.5271317829457366e-06,
+      "loss": 0.4275,
+      "step": 292
+    },
+    {
+      "epoch": 0.13625794450472795,
+      "grad_norm": 0.4523601830005646,
+      "learning_rate": 4.542635658914729e-06,
+      "loss": 0.4524,
+      "step": 293
+    },
+    {
+      "epoch": 0.13672298868392496,
+      "grad_norm": 0.4428309500217438,
+      "learning_rate": 4.558139534883721e-06,
+      "loss": 0.4247,
+      "step": 294
+    },
+    {
+      "epoch": 0.137188032863122,
+      "grad_norm": 0.42991116642951965,
+      "learning_rate": 4.573643410852713e-06,
+      "loss": 0.4232,
+      "step": 295
+    },
+    {
+      "epoch": 0.13765307704231902,
+      "grad_norm": 0.40712928771972656,
+      "learning_rate": 4.5891472868217054e-06,
+      "loss": 0.4426,
+      "step": 296
+    },
+    {
+      "epoch": 0.13811812122151604,
+      "grad_norm": 0.45721739530563354,
+      "learning_rate": 4.604651162790698e-06,
+      "loss": 0.4666,
+      "step": 297
+    },
+    {
+      "epoch": 0.13858316540071305,
+      "grad_norm": 0.41279730200767517,
+      "learning_rate": 4.62015503875969e-06,
+      "loss": 0.4258,
+      "step": 298
+    },
+    {
+      "epoch": 0.1390482095799101,
+      "grad_norm": 0.47948822379112244,
+      "learning_rate": 4.635658914728682e-06,
+      "loss": 0.4285,
+      "step": 299
+    },
+    {
+      "epoch": 0.13951325375910711,
+      "grad_norm": 0.45233485102653503,
+      "learning_rate": 4.651162790697675e-06,
+      "loss": 0.4522,
+      "step": 300
+    },
+    {
+      "epoch": 0.13997829793830413,
+      "grad_norm": 0.4481063783168793,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.4332,
+      "step": 301
+    },
+    {
+      "epoch": 0.14044334211750117,
+      "grad_norm": 0.4825620651245117,
+      "learning_rate": 4.682170542635659e-06,
+      "loss": 0.4643,
+      "step": 302
+    },
+    {
+      "epoch": 0.1409083862966982,
+      "grad_norm": 0.4766821265220642,
+      "learning_rate": 4.697674418604651e-06,
+      "loss": 0.4433,
+      "step": 303
+    },
+    {
+      "epoch": 0.1413734304758952,
+      "grad_norm": 0.46264582872390747,
+      "learning_rate": 4.713178294573644e-06,
+      "loss": 0.4409,
+      "step": 304
+    },
+    {
+      "epoch": 0.14183847465509222,
+      "grad_norm": 0.4554003179073334,
+      "learning_rate": 4.728682170542636e-06,
+      "loss": 0.4215,
+      "step": 305
+    },
+    {
+      "epoch": 0.14230351883428927,
+      "grad_norm": 0.45970794558525085,
+      "learning_rate": 4.744186046511629e-06,
+      "loss": 0.4259,
+      "step": 306
+    },
+    {
+      "epoch": 0.14276856301348628,
+      "grad_norm": 0.4238308370113373,
+      "learning_rate": 4.75968992248062e-06,
+      "loss": 0.4723,
+      "step": 307
+    },
+    {
+      "epoch": 0.1432336071926833,
+      "grad_norm": 0.4310249090194702,
+      "learning_rate": 4.775193798449613e-06,
+      "loss": 0.4526,
+      "step": 308
+    },
+    {
+      "epoch": 0.14369865137188031,
+      "grad_norm": 0.4667305648326874,
+      "learning_rate": 4.790697674418605e-06,
+      "loss": 0.4631,
+      "step": 309
+    },
+    {
+      "epoch": 0.14416369555107736,
+      "grad_norm": 0.465160071849823,
+      "learning_rate": 4.806201550387598e-06,
+      "loss": 0.4399,
+      "step": 310
+    },
+    {
+      "epoch": 0.14462873973027437,
+      "grad_norm": 0.44416457414627075,
+      "learning_rate": 4.821705426356589e-06,
+      "loss": 0.4398,
+      "step": 311
+    },
+    {
+      "epoch": 0.1450937839094714,
+      "grad_norm": 0.44952088594436646,
+      "learning_rate": 4.837209302325582e-06,
+      "loss": 0.4311,
+      "step": 312
+    },
+    {
+      "epoch": 0.14555882808866843,
+      "grad_norm": 0.44717299938201904,
+      "learning_rate": 4.852713178294574e-06,
+      "loss": 0.4541,
+      "step": 313
+    },
+    {
+      "epoch": 0.14602387226786545,
+      "grad_norm": 0.4240880310535431,
+      "learning_rate": 4.868217054263567e-06,
+      "loss": 0.4496,
+      "step": 314
+    },
+    {
+      "epoch": 0.14648891644706247,
+      "grad_norm": 0.46114540100097656,
+      "learning_rate": 4.883720930232559e-06,
+      "loss": 0.4493,
+      "step": 315
+    },
+    {
+      "epoch": 0.14695396062625948,
+      "grad_norm": 0.4437052309513092,
+      "learning_rate": 4.899224806201551e-06,
+      "loss": 0.435,
+      "step": 316
+    },
+    {
+      "epoch": 0.14741900480545653,
+      "grad_norm": 0.458609938621521,
+      "learning_rate": 4.9147286821705435e-06,
+      "loss": 0.4451,
+      "step": 317
+    },
+    {
+      "epoch": 0.14788404898465354,
+      "grad_norm": 0.4837147295475006,
+      "learning_rate": 4.9302325581395355e-06,
+      "loss": 0.4383,
+      "step": 318
+    },
+    {
+      "epoch": 0.14834909316385056,
+      "grad_norm": 0.4129205048084259,
+      "learning_rate": 4.9457364341085275e-06,
+      "loss": 0.4556,
+      "step": 319
+    },
+    {
+      "epoch": 0.1488141373430476,
+      "grad_norm": 0.5567089319229126,
+      "learning_rate": 4.9612403100775195e-06,
+      "loss": 0.4308,
+      "step": 320
+    },
+    {
+      "epoch": 0.14927918152224462,
+      "grad_norm": 0.47763490676879883,
+      "learning_rate": 4.976744186046512e-06,
+      "loss": 0.4492,
+      "step": 321
+    },
+    {
+      "epoch": 0.14974422570144164,
+      "grad_norm": 0.4176934063434601,
+      "learning_rate": 4.992248062015504e-06,
+      "loss": 0.4239,
+      "step": 322
+    },
+    {
+      "epoch": 0.15020926988063865,
+      "grad_norm": 0.5204504728317261,
+      "learning_rate": 5.007751937984496e-06,
+      "loss": 0.4357,
+      "step": 323
+    },
+    {
+      "epoch": 0.1506743140598357,
+      "grad_norm": 0.4915800988674164,
+      "learning_rate": 5.023255813953489e-06,
+      "loss": 0.4803,
+      "step": 324
+    },
+    {
+      "epoch": 0.1511393582390327,
+      "grad_norm": 0.4427052438259125,
+      "learning_rate": 5.038759689922481e-06,
+      "loss": 0.4279,
+      "step": 325
+    },
+    {
+      "epoch": 0.15160440241822973,
+      "grad_norm": 0.5173365473747253,
+      "learning_rate": 5.054263565891473e-06,
+      "loss": 0.4304,
+      "step": 326
+    },
+    {
+      "epoch": 0.15206944659742674,
+      "grad_norm": 0.4465036988258362,
+      "learning_rate": 5.069767441860466e-06,
+      "loss": 0.4288,
+      "step": 327
+    },
+    {
+      "epoch": 0.1525344907766238,
+      "grad_norm": 0.4534618556499481,
+      "learning_rate": 5.085271317829457e-06,
+      "loss": 0.4308,
+      "step": 328
+    },
+    {
+      "epoch": 0.1529995349558208,
+      "grad_norm": 0.42236584424972534,
+      "learning_rate": 5.100775193798449e-06,
+      "loss": 0.4221,
+      "step": 329
+    },
+    {
+      "epoch": 0.15346457913501782,
+      "grad_norm": 0.44343680143356323,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 0.4384,
+      "step": 330
+    },
+    {
+      "epoch": 0.15392962331421486,
+      "grad_norm": 0.45353367924690247,
+      "learning_rate": 5.131782945736434e-06,
+      "loss": 0.4425,
+      "step": 331
+    },
+    {
+      "epoch": 0.15439466749341188,
+      "grad_norm": 0.4323454797267914,
+      "learning_rate": 5.147286821705427e-06,
+      "loss": 0.4398,
+      "step": 332
+    },
+    {
+      "epoch": 0.1548597116726089,
+      "grad_norm": 0.4849368929862976,
+      "learning_rate": 5.162790697674419e-06,
+      "loss": 0.4586,
+      "step": 333
+    },
+    {
+      "epoch": 0.1553247558518059,
+      "grad_norm": 0.430549681186676,
+      "learning_rate": 5.178294573643411e-06,
+      "loss": 0.4308,
+      "step": 334
+    },
+    {
+      "epoch": 0.15578980003100296,
+      "grad_norm": 0.4729780852794647,
+      "learning_rate": 5.193798449612404e-06,
+      "loss": 0.4485,
+      "step": 335
+    },
+    {
+      "epoch": 0.15625484421019997,
+      "grad_norm": 0.45753175020217896,
+      "learning_rate": 5.209302325581396e-06,
+      "loss": 0.4205,
+      "step": 336
+    },
+    {
+      "epoch": 0.156719888389397,
+      "grad_norm": 0.44031214714050293,
+      "learning_rate": 5.224806201550388e-06,
+      "loss": 0.4309,
+      "step": 337
+    },
+    {
+      "epoch": 0.157184932568594,
+      "grad_norm": 0.5231649279594421,
+      "learning_rate": 5.240310077519381e-06,
+      "loss": 0.429,
+      "step": 338
+    },
+    {
+      "epoch": 0.15764997674779105,
+      "grad_norm": 0.48316752910614014,
+      "learning_rate": 5.255813953488372e-06,
+      "loss": 0.4651,
+      "step": 339
+    },
+    {
+      "epoch": 0.15811502092698806,
+      "grad_norm": 0.4905528426170349,
+      "learning_rate": 5.271317829457366e-06,
+      "loss": 0.4233,
+      "step": 340
+    },
+    {
+      "epoch": 0.15858006510618508,
+      "grad_norm": 0.44905000925064087,
+      "learning_rate": 5.286821705426357e-06,
+      "loss": 0.4741,
+      "step": 341
+    },
+    {
+      "epoch": 0.15904510928538212,
+      "grad_norm": 0.4897783696651459,
+      "learning_rate": 5.302325581395349e-06,
+      "loss": 0.4568,
+      "step": 342
+    },
+    {
+      "epoch": 0.15951015346457914,
+      "grad_norm": 0.4445964992046356,
+      "learning_rate": 5.317829457364342e-06,
+      "loss": 0.4294,
+      "step": 343
+    },
+    {
+      "epoch": 0.15997519764377616,
+      "grad_norm": 0.48060405254364014,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.4612,
+      "step": 344
+    },
+    {
+      "epoch": 0.16044024182297317,
+      "grad_norm": 0.4548921585083008,
+      "learning_rate": 5.348837209302326e-06,
+      "loss": 0.4569,
+      "step": 345
+    },
+    {
+      "epoch": 0.16090528600217022,
+      "grad_norm": 0.47072821855545044,
+      "learning_rate": 5.3643410852713185e-06,
+      "loss": 0.4403,
+      "step": 346
+    },
+    {
+      "epoch": 0.16137033018136723,
+      "grad_norm": 0.40528589487075806,
+      "learning_rate": 5.3798449612403105e-06,
+      "loss": 0.4524,
+      "step": 347
+    },
+    {
+      "epoch": 0.16183537436056425,
+      "grad_norm": 0.4871981739997864,
+      "learning_rate": 5.395348837209303e-06,
+      "loss": 0.4402,
+      "step": 348
+    },
+    {
+      "epoch": 0.16230041853976127,
+      "grad_norm": 0.4808438718318939,
+      "learning_rate": 5.410852713178295e-06,
+      "loss": 0.475,
+      "step": 349
+    },
+    {
+      "epoch": 0.1627654627189583,
+      "grad_norm": 0.4704870283603668,
+      "learning_rate": 5.4263565891472865e-06,
+      "loss": 0.432,
+      "step": 350
+    },
+    {
+      "epoch": 0.16323050689815533,
+      "grad_norm": 0.5283604860305786,
+      "learning_rate": 5.44186046511628e-06,
+      "loss": 0.4388,
+      "step": 351
+    },
+    {
+      "epoch": 0.16369555107735234,
+      "grad_norm": 0.4891282618045807,
+      "learning_rate": 5.457364341085271e-06,
+      "loss": 0.4561,
+      "step": 352
+    },
+    {
+      "epoch": 0.16416059525654939,
+      "grad_norm": 0.5148476362228394,
+      "learning_rate": 5.472868217054263e-06,
+      "loss": 0.4419,
+      "step": 353
+    },
+    {
+      "epoch": 0.1646256394357464,
+      "grad_norm": 0.43338412046432495,
+      "learning_rate": 5.488372093023256e-06,
+      "loss": 0.4295,
+      "step": 354
+    },
+    {
+      "epoch": 0.16509068361494342,
+      "grad_norm": 0.5159112811088562,
+      "learning_rate": 5.503875968992248e-06,
+      "loss": 0.4322,
+      "step": 355
+    },
+    {
+      "epoch": 0.16555572779414043,
+      "grad_norm": 0.4641515910625458,
+      "learning_rate": 5.519379844961241e-06,
+      "loss": 0.4303,
+      "step": 356
+    },
+    {
+      "epoch": 0.16602077197333748,
+      "grad_norm": 0.47173258662223816,
+      "learning_rate": 5.534883720930233e-06,
+      "loss": 0.4437,
+      "step": 357
+    },
+    {
+      "epoch": 0.1664858161525345,
+      "grad_norm": 0.5051222443580627,
+      "learning_rate": 5.550387596899225e-06,
+      "loss": 0.4265,
+      "step": 358
+    },
+    {
+      "epoch": 0.1669508603317315,
+      "grad_norm": 0.4815625250339508,
+      "learning_rate": 5.565891472868218e-06,
+      "loss": 0.4446,
+      "step": 359
+    },
+    {
+      "epoch": 0.16741590451092853,
+      "grad_norm": 0.5002831816673279,
+      "learning_rate": 5.58139534883721e-06,
+      "loss": 0.4428,
+      "step": 360
+    },
+    {
+      "epoch": 0.16788094869012557,
+      "grad_norm": 0.47330793738365173,
+      "learning_rate": 5.596899224806201e-06,
+      "loss": 0.4442,
+      "step": 361
+    },
+    {
+      "epoch": 0.16834599286932259,
+      "grad_norm": 0.4468974173069,
+      "learning_rate": 5.612403100775195e-06,
+      "loss": 0.4482,
+      "step": 362
+    },
+    {
+      "epoch": 0.1688110370485196,
+      "grad_norm": 0.46247512102127075,
+      "learning_rate": 5.627906976744186e-06,
+      "loss": 0.4534,
+      "step": 363
+    },
+    {
+      "epoch": 0.16927608122771665,
+      "grad_norm": 0.4856452941894531,
+      "learning_rate": 5.643410852713179e-06,
+      "loss": 0.4304,
+      "step": 364
+    },
+    {
+      "epoch": 0.16974112540691366,
+      "grad_norm": 0.48045817017555237,
+      "learning_rate": 5.658914728682171e-06,
+      "loss": 0.4347,
+      "step": 365
+    },
+    {
+      "epoch": 0.17020616958611068,
+      "grad_norm": 0.4817154109477997,
+      "learning_rate": 5.674418604651163e-06,
+      "loss": 0.455,
+      "step": 366
+    },
+    {
+      "epoch": 0.1706712137653077,
+      "grad_norm": 0.4779435396194458,
+      "learning_rate": 5.689922480620156e-06,
+      "loss": 0.398,
+      "step": 367
+    },
+    {
+      "epoch": 0.17113625794450474,
+      "grad_norm": 0.5880154371261597,
+      "learning_rate": 5.705426356589148e-06,
+      "loss": 0.4537,
+      "step": 368
+    },
+    {
+      "epoch": 0.17160130212370175,
+      "grad_norm": 0.53252112865448,
+      "learning_rate": 5.72093023255814e-06,
+      "loss": 0.446,
+      "step": 369
+    },
+    {
+      "epoch": 0.17206634630289877,
+      "grad_norm": 0.4977021813392639,
+      "learning_rate": 5.736434108527133e-06,
+      "loss": 0.4353,
+      "step": 370
+    },
+    {
+      "epoch": 0.1725313904820958,
+      "grad_norm": 0.5705452561378479,
+      "learning_rate": 5.751937984496125e-06,
+      "loss": 0.444,
+      "step": 371
+    },
+    {
+      "epoch": 0.17299643466129283,
+      "grad_norm": 0.4966931641101837,
+      "learning_rate": 5.7674418604651175e-06,
+      "loss": 0.4264,
+      "step": 372
+    },
+    {
+      "epoch": 0.17346147884048985,
+      "grad_norm": 0.5850646495819092,
+      "learning_rate": 5.782945736434109e-06,
+      "loss": 0.4223,
+      "step": 373
+    },
+    {
+      "epoch": 0.17392652301968686,
+      "grad_norm": 0.469243586063385,
+      "learning_rate": 5.798449612403101e-06,
+      "loss": 0.4206,
+      "step": 374
+    },
+    {
+      "epoch": 0.1743915671988839,
+      "grad_norm": 0.5241683125495911,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 0.4422,
+      "step": 375
+    },
+    {
+      "epoch": 0.17485661137808092,
+      "grad_norm": 0.5510844588279724,
+      "learning_rate": 5.8294573643410855e-06,
+      "loss": 0.4335,
+      "step": 376
+    },
+    {
+      "epoch": 0.17532165555727794,
+      "grad_norm": 0.4422820508480072,
+      "learning_rate": 5.8449612403100775e-06,
+      "loss": 0.444,
+      "step": 377
+    },
+    {
+      "epoch": 0.17578669973647496,
+      "grad_norm": 0.4676722586154938,
+      "learning_rate": 5.86046511627907e-06,
+      "loss": 0.4477,
+      "step": 378
+    },
+    {
+      "epoch": 0.176251743915672,
+      "grad_norm": 0.41679903864860535,
+      "learning_rate": 5.875968992248062e-06,
+      "loss": 0.4085,
+      "step": 379
+    },
+    {
+      "epoch": 0.17671678809486902,
+      "grad_norm": 0.4864595830440521,
+      "learning_rate": 5.891472868217055e-06,
+      "loss": 0.4067,
+      "step": 380
+    },
+    {
+      "epoch": 0.17718183227406603,
+      "grad_norm": 0.47440460324287415,
+      "learning_rate": 5.906976744186047e-06,
+      "loss": 0.4496,
+      "step": 381
+    },
+    {
+      "epoch": 0.17764687645326305,
+      "grad_norm": 0.49285659193992615,
+      "learning_rate": 5.922480620155039e-06,
+      "loss": 0.4476,
+      "step": 382
+    },
+    {
+      "epoch": 0.1781119206324601,
+      "grad_norm": 0.4602559804916382,
+      "learning_rate": 5.937984496124032e-06,
+      "loss": 0.4111,
+      "step": 383
+    },
+    {
+      "epoch": 0.1785769648116571,
+      "grad_norm": 0.4376397728919983,
+      "learning_rate": 5.953488372093023e-06,
+      "loss": 0.4232,
+      "step": 384
+    },
+    {
+      "epoch": 0.17904200899085412,
+      "grad_norm": 0.4615304470062256,
+      "learning_rate": 5.968992248062015e-06,
+      "loss": 0.4413,
+      "step": 385
+    },
+    {
+      "epoch": 0.17950705317005117,
+      "grad_norm": 0.49313583970069885,
+      "learning_rate": 5.984496124031008e-06,
+      "loss": 0.4426,
+      "step": 386
+    },
+    {
+      "epoch": 0.17997209734924818,
+      "grad_norm": 0.4725257158279419,
+      "learning_rate": 6e-06,
+      "loss": 0.4515,
+      "step": 387
+    },
+    {
+      "epoch": 0.1804371415284452,
+      "grad_norm": 0.44713687896728516,
+      "learning_rate": 6.015503875968993e-06,
+      "loss": 0.4232,
+      "step": 388
+    },
+    {
+      "epoch": 0.18090218570764222,
+      "grad_norm": 0.444200724363327,
+      "learning_rate": 6.031007751937985e-06,
+      "loss": 0.4353,
+      "step": 389
+    },
+    {
+      "epoch": 0.18136722988683926,
+      "grad_norm": 0.4371640980243683,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 0.4257,
+      "step": 390
+    },
+    {
+      "epoch": 0.18183227406603628,
+      "grad_norm": 0.5189304947853088,
+      "learning_rate": 6.06201550387597e-06,
+      "loss": 0.4365,
+      "step": 391
+    },
+    {
+      "epoch": 0.1822973182452333,
+      "grad_norm": 0.4122354984283447,
+      "learning_rate": 6.077519379844962e-06,
+      "loss": 0.4226,
+      "step": 392
+    },
+    {
+      "epoch": 0.1827623624244303,
+      "grad_norm": 0.47240152955055237,
+      "learning_rate": 6.093023255813954e-06,
+      "loss": 0.4387,
+      "step": 393
+    },
+    {
+      "epoch": 0.18322740660362735,
+      "grad_norm": 0.46310725808143616,
+      "learning_rate": 6.108527131782947e-06,
+      "loss": 0.4572,
+      "step": 394
+    },
+    {
+      "epoch": 0.18369245078282437,
+      "grad_norm": 0.4549012780189514,
+      "learning_rate": 6.124031007751938e-06,
+      "loss": 0.4321,
+      "step": 395
+    },
+    {
+      "epoch": 0.18415749496202138,
+      "grad_norm": 0.5056280493736267,
+      "learning_rate": 6.139534883720932e-06,
+      "loss": 0.4233,
+      "step": 396
+    },
+    {
+      "epoch": 0.18462253914121843,
+      "grad_norm": 0.4270991086959839,
+      "learning_rate": 6.155038759689923e-06,
+      "loss": 0.4204,
+      "step": 397
+    },
+    {
+      "epoch": 0.18508758332041544,
+      "grad_norm": 0.47168463468551636,
+      "learning_rate": 6.170542635658915e-06,
+      "loss": 0.3967,
+      "step": 398
+    },
+    {
+      "epoch": 0.18555262749961246,
+      "grad_norm": 0.5067258477210999,
+      "learning_rate": 6.186046511627908e-06,
+      "loss": 0.4273,
+      "step": 399
+    },
+    {
+      "epoch": 0.18601767167880948,
+      "grad_norm": 0.46395808458328247,
+      "learning_rate": 6.2015503875969e-06,
+      "loss": 0.4256,
+      "step": 400
+    },
+    {
+      "epoch": 0.18648271585800652,
+      "grad_norm": 0.46003836393356323,
+      "learning_rate": 6.217054263565892e-06,
+      "loss": 0.41,
+      "step": 401
+    },
+    {
+      "epoch": 0.18694776003720354,
+      "grad_norm": 0.5008445382118225,
+      "learning_rate": 6.2325581395348845e-06,
+      "loss": 0.3952,
+      "step": 402
+    },
+    {
+      "epoch": 0.18741280421640055,
+      "grad_norm": 0.5617318153381348,
+      "learning_rate": 6.2480620155038765e-06,
+      "loss": 0.4104,
+      "step": 403
+    },
+    {
+      "epoch": 0.18787784839559757,
+      "grad_norm": 0.49167153239250183,
+      "learning_rate": 6.263565891472869e-06,
+      "loss": 0.4252,
+      "step": 404
+    },
+    {
+      "epoch": 0.1883428925747946,
+      "grad_norm": 0.4648562967777252,
+      "learning_rate": 6.279069767441861e-06,
+      "loss": 0.4485,
+      "step": 405
+    },
+    {
+      "epoch": 0.18880793675399163,
+      "grad_norm": 0.48742571473121643,
+      "learning_rate": 6.2945736434108525e-06,
+      "loss": 0.4088,
+      "step": 406
+    },
+    {
+      "epoch": 0.18927298093318864,
+      "grad_norm": 0.4988233745098114,
+      "learning_rate": 6.310077519379845e-06,
+      "loss": 0.4159,
+      "step": 407
+    },
+    {
+      "epoch": 0.1897380251123857,
+      "grad_norm": 0.45389029383659363,
+      "learning_rate": 6.325581395348837e-06,
+      "loss": 0.45,
+      "step": 408
+    },
+    {
+      "epoch": 0.1902030692915827,
+      "grad_norm": 0.48481884598731995,
+      "learning_rate": 6.341085271317829e-06,
+      "loss": 0.4327,
+      "step": 409
+    },
+    {
+      "epoch": 0.19066811347077972,
+      "grad_norm": 0.5068842768669128,
+      "learning_rate": 6.356589147286822e-06,
+      "loss": 0.4068,
+      "step": 410
+    },
+    {
+      "epoch": 0.19113315764997674,
+      "grad_norm": 0.5590021014213562,
+      "learning_rate": 6.372093023255814e-06,
+      "loss": 0.4622,
+      "step": 411
+    },
+    {
+      "epoch": 0.19159820182917378,
+      "grad_norm": 0.4978815019130707,
+      "learning_rate": 6.387596899224807e-06,
+      "loss": 0.4285,
+      "step": 412
+    },
+    {
+      "epoch": 0.1920632460083708,
+      "grad_norm": 0.48937058448791504,
+      "learning_rate": 6.403100775193799e-06,
+      "loss": 0.432,
+      "step": 413
+    },
+    {
+      "epoch": 0.1925282901875678,
+      "grad_norm": 0.47953492403030396,
+      "learning_rate": 6.418604651162791e-06,
+      "loss": 0.428,
+      "step": 414
+    },
+    {
+      "epoch": 0.19299333436676483,
+      "grad_norm": 0.5264999270439148,
+      "learning_rate": 6.434108527131784e-06,
+      "loss": 0.4338,
+      "step": 415
+    },
+    {
+      "epoch": 0.19345837854596187,
+      "grad_norm": 0.5225833058357239,
+      "learning_rate": 6.449612403100776e-06,
+      "loss": 0.4423,
+      "step": 416
+    },
+    {
+      "epoch": 0.1939234227251589,
+      "grad_norm": 0.43435758352279663,
+      "learning_rate": 6.465116279069767e-06,
+      "loss": 0.4034,
+      "step": 417
+    },
+    {
+      "epoch": 0.1943884669043559,
+      "grad_norm": 0.4764443337917328,
+      "learning_rate": 6.48062015503876e-06,
+      "loss": 0.4623,
+      "step": 418
+    },
+    {
+      "epoch": 0.19485351108355295,
+      "grad_norm": 0.49896514415740967,
+      "learning_rate": 6.496124031007752e-06,
+      "loss": 0.4413,
+      "step": 419
+    },
+    {
+      "epoch": 0.19531855526274997,
+      "grad_norm": 0.46910813450813293,
+      "learning_rate": 6.511627906976745e-06,
+      "loss": 0.4089,
+      "step": 420
+    },
+    {
+      "epoch": 0.19578359944194698,
+      "grad_norm": 0.46757951378822327,
+      "learning_rate": 6.527131782945737e-06,
+      "loss": 0.42,
+      "step": 421
+    },
+    {
+      "epoch": 0.196248643621144,
+      "grad_norm": 0.4977341890335083,
+      "learning_rate": 6.542635658914729e-06,
+      "loss": 0.4167,
+      "step": 422
+    },
+    {
+      "epoch": 0.19671368780034104,
+      "grad_norm": 0.4749394357204437,
+      "learning_rate": 6.558139534883722e-06,
+      "loss": 0.4044,
+      "step": 423
+    },
+    {
+      "epoch": 0.19717873197953806,
+      "grad_norm": 0.461539626121521,
+      "learning_rate": 6.573643410852714e-06,
+      "loss": 0.4022,
+      "step": 424
+    },
+    {
+      "epoch": 0.19764377615873507,
+      "grad_norm": 0.5062358379364014,
+      "learning_rate": 6.589147286821706e-06,
+      "loss": 0.435,
+      "step": 425
+    },
+    {
+      "epoch": 0.1981088203379321,
+      "grad_norm": 0.42745712399482727,
+      "learning_rate": 6.604651162790699e-06,
+      "loss": 0.4393,
+      "step": 426
+    },
+    {
+      "epoch": 0.19857386451712913,
+      "grad_norm": 0.43482205271720886,
+      "learning_rate": 6.620155038759691e-06,
+      "loss": 0.4132,
+      "step": 427
+    },
+    {
+      "epoch": 0.19903890869632615,
+      "grad_norm": 0.452120840549469,
+      "learning_rate": 6.6356589147286835e-06,
+      "loss": 0.4207,
+      "step": 428
+    },
+    {
+      "epoch": 0.19950395287552317,
+      "grad_norm": 0.453553169965744,
+      "learning_rate": 6.651162790697675e-06,
+      "loss": 0.4211,
+      "step": 429
+    },
+    {
+      "epoch": 0.1999689970547202,
+      "grad_norm": 0.49294397234916687,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.4354,
+      "step": 430
+    },
+    {
+      "epoch": 0.20043404123391723,
+      "grad_norm": 0.46980249881744385,
+      "learning_rate": 6.6821705426356595e-06,
+      "loss": 0.4315,
+      "step": 431
+    },
+    {
+      "epoch": 0.20089908541311424,
+      "grad_norm": 0.48995155096054077,
+      "learning_rate": 6.6976744186046515e-06,
+      "loss": 0.4428,
+      "step": 432
+    },
+    {
+      "epoch": 0.20136412959231126,
+      "grad_norm": 0.4816022515296936,
+      "learning_rate": 6.7131782945736435e-06,
+      "loss": 0.4411,
+      "step": 433
+    },
+    {
+      "epoch": 0.2018291737715083,
+      "grad_norm": 0.49030253291130066,
+      "learning_rate": 6.728682170542636e-06,
+      "loss": 0.409,
+      "step": 434
+    },
+    {
+      "epoch": 0.20229421795070532,
+      "grad_norm": 0.5009063482284546,
+      "learning_rate": 6.744186046511628e-06,
+      "loss": 0.4306,
+      "step": 435
+    },
+    {
+      "epoch": 0.20275926212990233,
+      "grad_norm": 0.49570536613464355,
+      "learning_rate": 6.759689922480621e-06,
+      "loss": 0.4166,
+      "step": 436
+    },
+    {
+      "epoch": 0.20322430630909935,
+      "grad_norm": 0.5443177223205566,
+      "learning_rate": 6.775193798449613e-06,
+      "loss": 0.4188,
+      "step": 437
+    },
+    {
+      "epoch": 0.2036893504882964,
+      "grad_norm": 0.46004390716552734,
+      "learning_rate": 6.790697674418605e-06,
+      "loss": 0.4149,
+      "step": 438
+    },
+    {
+      "epoch": 0.2041543946674934,
+      "grad_norm": 0.5454478859901428,
+      "learning_rate": 6.806201550387598e-06,
+      "loss": 0.4153,
+      "step": 439
+    },
+    {
+      "epoch": 0.20461943884669043,
+      "grad_norm": 0.6096868515014648,
+      "learning_rate": 6.821705426356589e-06,
+      "loss": 0.4373,
+      "step": 440
+    },
+    {
+      "epoch": 0.20508448302588747,
+      "grad_norm": 0.4331457018852234,
+      "learning_rate": 6.837209302325581e-06,
+      "loss": 0.4273,
+      "step": 441
+    },
+    {
+      "epoch": 0.2055495272050845,
+      "grad_norm": 0.5101227760314941,
+      "learning_rate": 6.852713178294574e-06,
+      "loss": 0.4174,
+      "step": 442
+    },
+    {
+      "epoch": 0.2060145713842815,
+      "grad_norm": 0.47638994455337524,
+      "learning_rate": 6.868217054263566e-06,
+      "loss": 0.4332,
+      "step": 443
+    },
+    {
+      "epoch": 0.20647961556347852,
+      "grad_norm": 0.48688259720802307,
+      "learning_rate": 6.883720930232559e-06,
+      "loss": 0.4269,
+      "step": 444
+    },
+    {
+      "epoch": 0.20694465974267556,
+      "grad_norm": 0.5044205188751221,
+      "learning_rate": 6.899224806201551e-06,
+      "loss": 0.4269,
+      "step": 445
+    },
+    {
+      "epoch": 0.20740970392187258,
+      "grad_norm": 0.5487541556358337,
+      "learning_rate": 6.914728682170543e-06,
+      "loss": 0.4411,
+      "step": 446
+    },
+    {
+      "epoch": 0.2078747481010696,
+      "grad_norm": 0.47763657569885254,
+      "learning_rate": 6.930232558139536e-06,
+      "loss": 0.4382,
+      "step": 447
+    },
+    {
+      "epoch": 0.20833979228026664,
+      "grad_norm": 0.49321773648262024,
+      "learning_rate": 6.945736434108528e-06,
+      "loss": 0.4441,
+      "step": 448
+    },
+    {
+      "epoch": 0.20880483645946366,
+      "grad_norm": 0.4817297160625458,
+      "learning_rate": 6.961240310077519e-06,
+      "loss": 0.4286,
+      "step": 449
+    },
+    {
+      "epoch": 0.20926988063866067,
+      "grad_norm": 0.5427500605583191,
+      "learning_rate": 6.976744186046513e-06,
+      "loss": 0.4021,
+      "step": 450
+    },
+    {
+      "epoch": 0.2097349248178577,
+      "grad_norm": 0.5691028237342834,
+      "learning_rate": 6.992248062015504e-06,
+      "loss": 0.4075,
+      "step": 451
+    },
+    {
+      "epoch": 0.21019996899705473,
+      "grad_norm": 0.4561224579811096,
+      "learning_rate": 7.007751937984497e-06,
+      "loss": 0.4075,
+      "step": 452
+    },
+    {
+      "epoch": 0.21066501317625175,
+      "grad_norm": 0.5992524027824402,
+      "learning_rate": 7.023255813953489e-06,
+      "loss": 0.4214,
+      "step": 453
+    },
+    {
+      "epoch": 0.21113005735544876,
+      "grad_norm": 0.6167622208595276,
+      "learning_rate": 7.038759689922481e-06,
+      "loss": 0.3898,
+      "step": 454
+    },
+    {
+      "epoch": 0.21159510153464578,
+      "grad_norm": 0.5301486253738403,
+      "learning_rate": 7.054263565891474e-06,
+      "loss": 0.4306,
+      "step": 455
+    },
+    {
+      "epoch": 0.21206014571384282,
+      "grad_norm": 0.6576060652732849,
+      "learning_rate": 7.069767441860466e-06,
+      "loss": 0.4162,
+      "step": 456
+    },
+    {
+      "epoch": 0.21252518989303984,
+      "grad_norm": 0.5479761958122253,
+      "learning_rate": 7.085271317829458e-06,
+      "loss": 0.4585,
+      "step": 457
+    },
+    {
+      "epoch": 0.21299023407223686,
+      "grad_norm": 0.47175103425979614,
+      "learning_rate": 7.1007751937984505e-06,
+      "loss": 0.4103,
+      "step": 458
+    },
+    {
+      "epoch": 0.2134552782514339,
+      "grad_norm": 0.5980085730552673,
+      "learning_rate": 7.1162790697674425e-06,
+      "loss": 0.4308,
+      "step": 459
+    },
+    {
+      "epoch": 0.21392032243063092,
+      "grad_norm": 0.49691030383110046,
+      "learning_rate": 7.131782945736435e-06,
+      "loss": 0.4177,
+      "step": 460
+    },
+    {
+      "epoch": 0.21438536660982793,
+      "grad_norm": 0.4692418575286865,
+      "learning_rate": 7.147286821705427e-06,
+      "loss": 0.4184,
+      "step": 461
+    },
+    {
+      "epoch": 0.21485041078902495,
+      "grad_norm": 0.4812504053115845,
+      "learning_rate": 7.1627906976744185e-06,
+      "loss": 0.4077,
+      "step": 462
+    },
+    {
+      "epoch": 0.215315454968222,
+      "grad_norm": 0.47342851758003235,
+      "learning_rate": 7.178294573643411e-06,
+      "loss": 0.4117,
+      "step": 463
+    },
+    {
+      "epoch": 0.215780499147419,
+      "grad_norm": 0.5251373648643494,
+      "learning_rate": 7.193798449612403e-06,
+      "loss": 0.451,
+      "step": 464
+    },
+    {
+      "epoch": 0.21624554332661602,
+      "grad_norm": 0.4466230869293213,
+      "learning_rate": 7.209302325581395e-06,
+      "loss": 0.4364,
+      "step": 465
+    },
+    {
+      "epoch": 0.21671058750581304,
+      "grad_norm": 0.502120852470398,
+      "learning_rate": 7.224806201550388e-06,
+      "loss": 0.451,
+      "step": 466
+    },
+    {
+      "epoch": 0.21717563168501008,
+      "grad_norm": 0.5081861615180969,
+      "learning_rate": 7.24031007751938e-06,
+      "loss": 0.4521,
+      "step": 467
+    },
+    {
+      "epoch": 0.2176406758642071,
+      "grad_norm": 0.5037353038787842,
+      "learning_rate": 7.255813953488373e-06,
+      "loss": 0.4055,
+      "step": 468
+    },
+    {
+      "epoch": 0.21810572004340412,
+      "grad_norm": 0.48872482776641846,
+      "learning_rate": 7.271317829457365e-06,
+      "loss": 0.4353,
+      "step": 469
+    },
+    {
+      "epoch": 0.21857076422260116,
+      "grad_norm": 0.4901602268218994,
+      "learning_rate": 7.286821705426357e-06,
+      "loss": 0.4176,
+      "step": 470
+    },
+    {
+      "epoch": 0.21903580840179818,
+      "grad_norm": 0.4682338535785675,
+      "learning_rate": 7.30232558139535e-06,
+      "loss": 0.394,
+      "step": 471
+    },
+    {
+      "epoch": 0.2195008525809952,
+      "grad_norm": 0.5010263919830322,
+      "learning_rate": 7.317829457364342e-06,
+      "loss": 0.4417,
+      "step": 472
+    },
+    {
+      "epoch": 0.2199658967601922,
+      "grad_norm": 0.4669908285140991,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.395,
+      "step": 473
+    },
+    {
+      "epoch": 0.22043094093938925,
+      "grad_norm": 0.47183507680892944,
+      "learning_rate": 7.348837209302326e-06,
+      "loss": 0.4388,
+      "step": 474
+    },
+    {
+      "epoch": 0.22089598511858627,
+      "grad_norm": 0.49007242918014526,
+      "learning_rate": 7.364341085271318e-06,
+      "loss": 0.4062,
+      "step": 475
+    },
+    {
+      "epoch": 0.22136102929778328,
+      "grad_norm": 0.5357131361961365,
+      "learning_rate": 7.379844961240311e-06,
+      "loss": 0.4414,
+      "step": 476
+    },
+    {
+      "epoch": 0.2218260734769803,
+      "grad_norm": 0.5643349289894104,
+      "learning_rate": 7.395348837209303e-06,
+      "loss": 0.4158,
+      "step": 477
+    },
+    {
+      "epoch": 0.22229111765617734,
+      "grad_norm": 0.4729050099849701,
+      "learning_rate": 7.410852713178295e-06,
+      "loss": 0.4273,
+      "step": 478
+    },
+    {
+      "epoch": 0.22275616183537436,
+      "grad_norm": 0.485498309135437,
+      "learning_rate": 7.426356589147288e-06,
+      "loss": 0.408,
+      "step": 479
+    },
+    {
+      "epoch": 0.22322120601457138,
+      "grad_norm": 0.5941533446311951,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 0.4191,
+      "step": 480
+    },
+    {
+      "epoch": 0.22368625019376842,
+      "grad_norm": 0.4895523488521576,
+      "learning_rate": 7.457364341085272e-06,
+      "loss": 0.4311,
+      "step": 481
+    },
+    {
+      "epoch": 0.22415129437296544,
+      "grad_norm": 0.6243600249290466,
+      "learning_rate": 7.472868217054265e-06,
+      "loss": 0.4214,
+      "step": 482
+    },
+    {
+      "epoch": 0.22461633855216245,
+      "grad_norm": 0.45703691244125366,
+      "learning_rate": 7.488372093023256e-06,
+      "loss": 0.4056,
+      "step": 483
+    },
+    {
+      "epoch": 0.22508138273135947,
+      "grad_norm": 0.5641500353813171,
+      "learning_rate": 7.5038759689922495e-06,
+      "loss": 0.4293,
+      "step": 484
+    },
+    {
+      "epoch": 0.2255464269105565,
+      "grad_norm": 0.5574773550033569,
+      "learning_rate": 7.519379844961241e-06,
+      "loss": 0.4195,
+      "step": 485
+    },
+    {
+      "epoch": 0.22601147108975353,
+      "grad_norm": 0.5667125582695007,
+      "learning_rate": 7.534883720930233e-06,
+      "loss": 0.4228,
+      "step": 486
+    },
+    {
+      "epoch": 0.22647651526895055,
+      "grad_norm": 0.4847676455974579,
+      "learning_rate": 7.5503875968992255e-06,
+      "loss": 0.4179,
+      "step": 487
+    },
+    {
+      "epoch": 0.22694155944814756,
+      "grad_norm": 0.51683509349823,
+      "learning_rate": 7.5658914728682175e-06,
+      "loss": 0.4035,
+      "step": 488
+    },
+    {
+      "epoch": 0.2274066036273446,
+      "grad_norm": 0.5952682495117188,
+      "learning_rate": 7.5813953488372095e-06,
+      "loss": 0.4308,
+      "step": 489
+    },
+    {
+      "epoch": 0.22787164780654162,
+      "grad_norm": 0.5058827996253967,
+      "learning_rate": 7.596899224806202e-06,
+      "loss": 0.4276,
+      "step": 490
+    },
+    {
+      "epoch": 0.22833669198573864,
+      "grad_norm": 0.7033687233924866,
+      "learning_rate": 7.612403100775194e-06,
+      "loss": 0.4345,
+      "step": 491
+    },
+    {
+      "epoch": 0.22880173616493568,
+      "grad_norm": 0.5743710398674011,
+      "learning_rate": 7.627906976744187e-06,
+      "loss": 0.4216,
+      "step": 492
+    },
+    {
+      "epoch": 0.2292667803441327,
+      "grad_norm": 0.637058675289154,
+      "learning_rate": 7.643410852713178e-06,
+      "loss": 0.439,
+      "step": 493
+    },
+    {
+      "epoch": 0.22973182452332971,
+      "grad_norm": 0.49991941452026367,
+      "learning_rate": 7.65891472868217e-06,
+      "loss": 0.4383,
+      "step": 494
+    },
+    {
+      "epoch": 0.23019686870252673,
+      "grad_norm": 0.6461667418479919,
+      "learning_rate": 7.674418604651164e-06,
+      "loss": 0.4542,
+      "step": 495
+    },
+    {
+      "epoch": 0.23066191288172377,
+      "grad_norm": 0.5306419134140015,
+      "learning_rate": 7.689922480620156e-06,
+      "loss": 0.4266,
+      "step": 496
+    },
+    {
+      "epoch": 0.2311269570609208,
+      "grad_norm": 0.5668182969093323,
+      "learning_rate": 7.705426356589148e-06,
+      "loss": 0.4369,
+      "step": 497
+    },
+    {
+      "epoch": 0.2315920012401178,
+      "grad_norm": 0.546444833278656,
+      "learning_rate": 7.72093023255814e-06,
+      "loss": 0.4281,
+      "step": 498
+    },
+    {
+      "epoch": 0.23205704541931482,
+      "grad_norm": 0.5376360416412354,
+      "learning_rate": 7.736434108527132e-06,
+      "loss": 0.4138,
+      "step": 499
+    },
+    {
+      "epoch": 0.23252208959851187,
+      "grad_norm": 0.6407905220985413,
+      "learning_rate": 7.751937984496126e-06,
+      "loss": 0.4204,
+      "step": 500
+    },
+    {
+      "epoch": 0.23298713377770888,
+      "grad_norm": 0.5134121179580688,
+      "learning_rate": 7.767441860465116e-06,
+      "loss": 0.4251,
+      "step": 501
+    },
+    {
+      "epoch": 0.2334521779569059,
+      "grad_norm": 0.6588038206100464,
+      "learning_rate": 7.782945736434108e-06,
+      "loss": 0.423,
+      "step": 502
+    },
+    {
+      "epoch": 0.23391722213610294,
+      "grad_norm": 0.5844888687133789,
+      "learning_rate": 7.798449612403102e-06,
+      "loss": 0.4362,
+      "step": 503
+    },
+    {
+      "epoch": 0.23438226631529996,
+      "grad_norm": 0.5324265956878662,
+      "learning_rate": 7.813953488372094e-06,
+      "loss": 0.4184,
+      "step": 504
+    },
+    {
+      "epoch": 0.23484731049449697,
+      "grad_norm": 0.6014962196350098,
+      "learning_rate": 7.829457364341086e-06,
+      "loss": 0.4314,
+      "step": 505
+    },
+    {
+      "epoch": 0.235312354673694,
+      "grad_norm": 0.51766037940979,
+      "learning_rate": 7.844961240310078e-06,
+      "loss": 0.405,
+      "step": 506
+    },
+    {
+      "epoch": 0.23577739885289103,
+      "grad_norm": 0.5190640091896057,
+      "learning_rate": 7.86046511627907e-06,
+      "loss": 0.4085,
+      "step": 507
+    },
+    {
+      "epoch": 0.23624244303208805,
+      "grad_norm": 0.5045374631881714,
+      "learning_rate": 7.875968992248064e-06,
+      "loss": 0.4219,
+      "step": 508
+    },
+    {
+      "epoch": 0.23670748721128507,
+      "grad_norm": 0.557181179523468,
+      "learning_rate": 7.891472868217056e-06,
+      "loss": 0.4026,
+      "step": 509
+    },
+    {
+      "epoch": 0.23717253139048208,
+      "grad_norm": 0.5194858908653259,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 0.4232,
+      "step": 510
+    },
+    {
+      "epoch": 0.23763757556967913,
+      "grad_norm": 0.5417995452880859,
+      "learning_rate": 7.92248062015504e-06,
+      "loss": 0.416,
+      "step": 511
+    },
+    {
+      "epoch": 0.23810261974887614,
+      "grad_norm": 0.48320630192756653,
+      "learning_rate": 7.937984496124032e-06,
+      "loss": 0.3952,
+      "step": 512
+    },
+    {
+      "epoch": 0.23856766392807316,
+      "grad_norm": 0.5941821336746216,
+      "learning_rate": 7.953488372093024e-06,
+      "loss": 0.4101,
+      "step": 513
+    },
+    {
+      "epoch": 0.2390327081072702,
+      "grad_norm": 0.5255295634269714,
+      "learning_rate": 7.968992248062016e-06,
+      "loss": 0.4091,
+      "step": 514
+    },
+    {
+      "epoch": 0.23949775228646722,
+      "grad_norm": 0.4808565378189087,
+      "learning_rate": 7.984496124031008e-06,
+      "loss": 0.4147,
+      "step": 515
+    },
+    {
+      "epoch": 0.23996279646566424,
+      "grad_norm": 0.6180495023727417,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.4193,
+      "step": 516
+    },
+    {
+      "epoch": 0.24042784064486125,
+      "grad_norm": 0.6234874725341797,
+      "learning_rate": 8.015503875968993e-06,
+      "loss": 0.4413,
+      "step": 517
+    },
+    {
+      "epoch": 0.2408928848240583,
+      "grad_norm": 0.49461182951927185,
+      "learning_rate": 8.031007751937985e-06,
+      "loss": 0.3924,
+      "step": 518
+    },
+    {
+      "epoch": 0.2413579290032553,
+      "grad_norm": 0.6341349482536316,
+      "learning_rate": 8.046511627906977e-06,
+      "loss": 0.4284,
+      "step": 519
+    },
+    {
+      "epoch": 0.24182297318245233,
+      "grad_norm": 0.5244817137718201,
+      "learning_rate": 8.06201550387597e-06,
+      "loss": 0.4037,
+      "step": 520
+    },
+    {
+      "epoch": 0.24228801736164934,
+      "grad_norm": 0.5345357656478882,
+      "learning_rate": 8.077519379844961e-06,
+      "loss": 0.4568,
+      "step": 521
+    },
+    {
+      "epoch": 0.2427530615408464,
+      "grad_norm": 0.685580849647522,
+      "learning_rate": 8.093023255813955e-06,
+      "loss": 0.4431,
+      "step": 522
+    },
+    {
+      "epoch": 0.2432181057200434,
+      "grad_norm": 0.5502837896347046,
+      "learning_rate": 8.108527131782945e-06,
+      "loss": 0.4197,
+      "step": 523
+    },
+    {
+      "epoch": 0.24368314989924042,
+      "grad_norm": 0.5113924741744995,
+      "learning_rate": 8.124031007751939e-06,
+      "loss": 0.4108,
+      "step": 524
+    },
+    {
+      "epoch": 0.24414819407843746,
+      "grad_norm": 0.5599501729011536,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 0.4374,
+      "step": 525
+    },
+    {
+      "epoch": 0.24461323825763448,
+      "grad_norm": 0.5084413886070251,
+      "learning_rate": 8.155038759689923e-06,
+      "loss": 0.4296,
+      "step": 526
+    },
+    {
+      "epoch": 0.2450782824368315,
+      "grad_norm": 0.49575719237327576,
+      "learning_rate": 8.170542635658915e-06,
+      "loss": 0.4252,
+      "step": 527
+    },
+    {
+      "epoch": 0.2455433266160285,
+      "grad_norm": 0.4909055233001709,
+      "learning_rate": 8.186046511627907e-06,
+      "loss": 0.4206,
+      "step": 528
+    },
+    {
+      "epoch": 0.24600837079522556,
+      "grad_norm": 0.5348622798919678,
+      "learning_rate": 8.201550387596899e-06,
+      "loss": 0.4373,
+      "step": 529
+    },
+    {
+      "epoch": 0.24647341497442257,
+      "grad_norm": 0.41662895679473877,
+      "learning_rate": 8.217054263565893e-06,
+      "loss": 0.4021,
+      "step": 530
+    },
+    {
+      "epoch": 0.2469384591536196,
+      "grad_norm": 0.5400405526161194,
+      "learning_rate": 8.232558139534885e-06,
+      "loss": 0.4104,
+      "step": 531
+    },
+    {
+      "epoch": 0.2474035033328166,
+      "grad_norm": 0.4757499694824219,
+      "learning_rate": 8.248062015503877e-06,
+      "loss": 0.4034,
+      "step": 532
+    },
+    {
+      "epoch": 0.24786854751201365,
+      "grad_norm": 0.49683964252471924,
+      "learning_rate": 8.263565891472869e-06,
+      "loss": 0.4145,
+      "step": 533
+    },
+    {
+      "epoch": 0.24833359169121066,
+      "grad_norm": 0.529530942440033,
+      "learning_rate": 8.279069767441861e-06,
+      "loss": 0.4188,
+      "step": 534
+    },
+    {
+      "epoch": 0.24879863587040768,
+      "grad_norm": 0.5073703527450562,
+      "learning_rate": 8.294573643410853e-06,
+      "loss": 0.4071,
+      "step": 535
+    },
+    {
+      "epoch": 0.24926368004960472,
+      "grad_norm": 0.5136784315109253,
+      "learning_rate": 8.310077519379845e-06,
+      "loss": 0.4178,
+      "step": 536
+    },
+    {
+      "epoch": 0.24972872422880174,
+      "grad_norm": 0.4692307412624359,
+      "learning_rate": 8.325581395348837e-06,
+      "loss": 0.4155,
+      "step": 537
+    },
+    {
+      "epoch": 0.2501937684079988,
+      "grad_norm": 0.5519499778747559,
+      "learning_rate": 8.34108527131783e-06,
+      "loss": 0.4088,
+      "step": 538
+    },
+    {
+      "epoch": 0.2506588125871958,
+      "grad_norm": 0.5323079228401184,
+      "learning_rate": 8.356589147286823e-06,
+      "loss": 0.3887,
+      "step": 539
+    },
+    {
+      "epoch": 0.2511238567663928,
+      "grad_norm": 0.5130917429924011,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 0.4473,
+      "step": 540
+    },
+    {
+      "epoch": 0.25158890094558983,
+      "grad_norm": 0.5379475951194763,
+      "learning_rate": 8.387596899224807e-06,
+      "loss": 0.4092,
+      "step": 541
+    },
+    {
+      "epoch": 0.25205394512478685,
+      "grad_norm": 0.4682493507862091,
+      "learning_rate": 8.403100775193799e-06,
+      "loss": 0.407,
+      "step": 542
+    },
+    {
+      "epoch": 0.25251898930398387,
+      "grad_norm": 0.5762583017349243,
+      "learning_rate": 8.418604651162792e-06,
+      "loss": 0.4402,
+      "step": 543
+    },
+    {
+      "epoch": 0.2529840334831809,
+      "grad_norm": 0.4663183093070984,
+      "learning_rate": 8.434108527131784e-06,
+      "loss": 0.4079,
+      "step": 544
+    },
+    {
+      "epoch": 0.25344907766237795,
+      "grad_norm": 0.5599426627159119,
+      "learning_rate": 8.449612403100775e-06,
+      "loss": 0.4041,
+      "step": 545
+    },
+    {
+      "epoch": 0.25391412184157497,
+      "grad_norm": 0.4949679374694824,
+      "learning_rate": 8.465116279069768e-06,
+      "loss": 0.4055,
+      "step": 546
+    },
+    {
+      "epoch": 0.254379166020772,
+      "grad_norm": 0.4728296995162964,
+      "learning_rate": 8.48062015503876e-06,
+      "loss": 0.4044,
+      "step": 547
+    },
+    {
+      "epoch": 0.254844210199969,
+      "grad_norm": 0.5103499889373779,
+      "learning_rate": 8.496124031007752e-06,
+      "loss": 0.4158,
+      "step": 548
+    },
+    {
+      "epoch": 0.255309254379166,
+      "grad_norm": 0.5350103974342346,
+      "learning_rate": 8.511627906976744e-06,
+      "loss": 0.4333,
+      "step": 549
+    },
+    {
+      "epoch": 0.25577429855836303,
+      "grad_norm": 0.47756704688072205,
+      "learning_rate": 8.527131782945736e-06,
+      "loss": 0.4273,
+      "step": 550
+    },
+    {
+      "epoch": 0.25623934273756005,
+      "grad_norm": 0.6039769649505615,
+      "learning_rate": 8.54263565891473e-06,
+      "loss": 0.447,
+      "step": 551
+    },
+    {
+      "epoch": 0.25670438691675707,
+      "grad_norm": 0.5659769177436829,
+      "learning_rate": 8.558139534883722e-06,
+      "loss": 0.4298,
+      "step": 552
+    },
+    {
+      "epoch": 0.25716943109595414,
+      "grad_norm": 0.46599307656288147,
+      "learning_rate": 8.573643410852714e-06,
+      "loss": 0.3993,
+      "step": 553
+    },
+    {
+      "epoch": 0.25763447527515115,
+      "grad_norm": 0.5073681473731995,
+      "learning_rate": 8.589147286821706e-06,
+      "loss": 0.426,
+      "step": 554
+    },
+    {
+      "epoch": 0.25809951945434817,
+      "grad_norm": 0.4540097117424011,
+      "learning_rate": 8.604651162790698e-06,
+      "loss": 0.4197,
+      "step": 555
+    },
+    {
+      "epoch": 0.2585645636335452,
+      "grad_norm": 0.46219104528427124,
+      "learning_rate": 8.620155038759692e-06,
+      "loss": 0.4181,
+      "step": 556
+    },
+    {
+      "epoch": 0.2590296078127422,
+      "grad_norm": 0.5334506034851074,
+      "learning_rate": 8.635658914728682e-06,
+      "loss": 0.4244,
+      "step": 557
+    },
+    {
+      "epoch": 0.2594946519919392,
+      "grad_norm": 0.6203073859214783,
+      "learning_rate": 8.651162790697674e-06,
+      "loss": 0.4125,
+      "step": 558
+    },
+    {
+      "epoch": 0.25995969617113623,
+      "grad_norm": 0.5533168315887451,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.4251,
+      "step": 559
+    },
+    {
+      "epoch": 0.2604247403503333,
+      "grad_norm": 0.5441173911094666,
+      "learning_rate": 8.68217054263566e-06,
+      "loss": 0.4062,
+      "step": 560
+    },
+    {
+      "epoch": 0.2608897845295303,
+      "grad_norm": 0.6109142899513245,
+      "learning_rate": 8.697674418604652e-06,
+      "loss": 0.454,
+      "step": 561
+    },
+    {
+      "epoch": 0.26135482870872734,
+      "grad_norm": 0.5405747890472412,
+      "learning_rate": 8.713178294573644e-06,
+      "loss": 0.418,
+      "step": 562
+    },
+    {
+      "epoch": 0.26181987288792435,
+      "grad_norm": 0.48734837770462036,
+      "learning_rate": 8.728682170542636e-06,
+      "loss": 0.417,
+      "step": 563
+    },
+    {
+      "epoch": 0.26228491706712137,
+      "grad_norm": 0.6338199973106384,
+      "learning_rate": 8.74418604651163e-06,
+      "loss": 0.3976,
+      "step": 564
+    },
+    {
+      "epoch": 0.2627499612463184,
+      "grad_norm": 0.5382621884346008,
+      "learning_rate": 8.759689922480622e-06,
+      "loss": 0.446,
+      "step": 565
+    },
+    {
+      "epoch": 0.2632150054255154,
+      "grad_norm": 0.4548230767250061,
+      "learning_rate": 8.775193798449612e-06,
+      "loss": 0.4227,
+      "step": 566
+    },
+    {
+      "epoch": 0.2636800496047125,
+      "grad_norm": 0.6364620327949524,
+      "learning_rate": 8.790697674418606e-06,
+      "loss": 0.4114,
+      "step": 567
+    },
+    {
+      "epoch": 0.2641450937839095,
+      "grad_norm": 0.6153837442398071,
+      "learning_rate": 8.806201550387598e-06,
+      "loss": 0.4259,
+      "step": 568
+    },
+    {
+      "epoch": 0.2646101379631065,
+      "grad_norm": 0.5064429044723511,
+      "learning_rate": 8.82170542635659e-06,
+      "loss": 0.4256,
+      "step": 569
+    },
+    {
+      "epoch": 0.2650751821423035,
+      "grad_norm": 0.6010052561759949,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 0.411,
+      "step": 570
+    },
+    {
+      "epoch": 0.26554022632150054,
+      "grad_norm": 0.659740686416626,
+      "learning_rate": 8.852713178294574e-06,
+      "loss": 0.3959,
+      "step": 571
+    },
+    {
+      "epoch": 0.26600527050069755,
+      "grad_norm": 0.548054039478302,
+      "learning_rate": 8.868217054263567e-06,
+      "loss": 0.4324,
+      "step": 572
+    },
+    {
+      "epoch": 0.26647031467989457,
+      "grad_norm": 0.5880405306816101,
+      "learning_rate": 8.88372093023256e-06,
+      "loss": 0.3957,
+      "step": 573
+    },
+    {
+      "epoch": 0.2669353588590916,
+      "grad_norm": 0.5269092321395874,
+      "learning_rate": 8.899224806201551e-06,
+      "loss": 0.4376,
+      "step": 574
+    },
+    {
+      "epoch": 0.26740040303828866,
+      "grad_norm": 0.5422910451889038,
+      "learning_rate": 8.914728682170543e-06,
+      "loss": 0.3812,
+      "step": 575
+    },
+    {
+      "epoch": 0.2678654472174857,
+      "grad_norm": 0.5918363332748413,
+      "learning_rate": 8.930232558139535e-06,
+      "loss": 0.4285,
+      "step": 576
+    },
+    {
+      "epoch": 0.2683304913966827,
+      "grad_norm": 0.5592337250709534,
+      "learning_rate": 8.945736434108527e-06,
+      "loss": 0.3972,
+      "step": 577
+    },
+    {
+      "epoch": 0.2687955355758797,
+      "grad_norm": 0.5338684916496277,
+      "learning_rate": 8.961240310077521e-06,
+      "loss": 0.4245,
+      "step": 578
+    },
+    {
+      "epoch": 0.2692605797550767,
+      "grad_norm": 0.46233999729156494,
+      "learning_rate": 8.976744186046511e-06,
+      "loss": 0.416,
+      "step": 579
+    },
+    {
+      "epoch": 0.26972562393427374,
+      "grad_norm": 0.6775352954864502,
+      "learning_rate": 8.992248062015505e-06,
+      "loss": 0.4152,
+      "step": 580
+    },
+    {
+      "epoch": 0.27019066811347076,
+      "grad_norm": 0.6111975312232971,
+      "learning_rate": 9.007751937984497e-06,
+      "loss": 0.4496,
+      "step": 581
+    },
+    {
+      "epoch": 0.2706557122926678,
+      "grad_norm": 0.4907594621181488,
+      "learning_rate": 9.023255813953489e-06,
+      "loss": 0.4145,
+      "step": 582
+    },
+    {
+      "epoch": 0.27112075647186484,
+      "grad_norm": 0.6536867022514343,
+      "learning_rate": 9.038759689922481e-06,
+      "loss": 0.4421,
+      "step": 583
+    },
+    {
+      "epoch": 0.27158580065106186,
+      "grad_norm": 0.4851461946964264,
+      "learning_rate": 9.054263565891473e-06,
+      "loss": 0.3856,
+      "step": 584
+    },
+    {
+      "epoch": 0.2720508448302589,
+      "grad_norm": 0.552401602268219,
+      "learning_rate": 9.069767441860465e-06,
+      "loss": 0.4085,
+      "step": 585
+    },
+    {
+      "epoch": 0.2725158890094559,
+      "grad_norm": 0.5714566111564636,
+      "learning_rate": 9.085271317829459e-06,
+      "loss": 0.3893,
+      "step": 586
+    },
+    {
+      "epoch": 0.2729809331886529,
+      "grad_norm": 0.5839101076126099,
+      "learning_rate": 9.10077519379845e-06,
+      "loss": 0.4145,
+      "step": 587
+    },
+    {
+      "epoch": 0.2734459773678499,
+      "grad_norm": 0.5246545076370239,
+      "learning_rate": 9.116279069767443e-06,
+      "loss": 0.4316,
+      "step": 588
+    },
+    {
+      "epoch": 0.273911021547047,
+      "grad_norm": 0.49281787872314453,
+      "learning_rate": 9.131782945736435e-06,
+      "loss": 0.4272,
+      "step": 589
+    },
+    {
+      "epoch": 0.274376065726244,
+      "grad_norm": 0.5275765657424927,
+      "learning_rate": 9.147286821705427e-06,
+      "loss": 0.4141,
+      "step": 590
+    },
+    {
+      "epoch": 0.27484110990544103,
+      "grad_norm": 0.4734654724597931,
+      "learning_rate": 9.162790697674419e-06,
+      "loss": 0.4031,
+      "step": 591
+    },
+    {
+      "epoch": 0.27530615408463804,
+      "grad_norm": 0.5400969386100769,
+      "learning_rate": 9.178294573643411e-06,
+      "loss": 0.4517,
+      "step": 592
+    },
+    {
+      "epoch": 0.27577119826383506,
+      "grad_norm": 0.4679478406906128,
+      "learning_rate": 9.193798449612403e-06,
+      "loss": 0.4014,
+      "step": 593
+    },
+    {
+      "epoch": 0.2762362424430321,
+      "grad_norm": 0.6086196899414062,
+      "learning_rate": 9.209302325581397e-06,
+      "loss": 0.4185,
+      "step": 594
+    },
+    {
+      "epoch": 0.2767012866222291,
+      "grad_norm": 0.5482898950576782,
+      "learning_rate": 9.224806201550389e-06,
+      "loss": 0.4391,
+      "step": 595
+    },
+    {
+      "epoch": 0.2771663308014261,
+      "grad_norm": 0.5058215856552124,
+      "learning_rate": 9.24031007751938e-06,
+      "loss": 0.3826,
+      "step": 596
+    },
+    {
+      "epoch": 0.2776313749806232,
+      "grad_norm": 0.5932140350341797,
+      "learning_rate": 9.255813953488373e-06,
+      "loss": 0.4105,
+      "step": 597
+    },
+    {
+      "epoch": 0.2780964191598202,
+      "grad_norm": 0.4306205213069916,
+      "learning_rate": 9.271317829457365e-06,
+      "loss": 0.4083,
+      "step": 598
+    },
+    {
+      "epoch": 0.2785614633390172,
+      "grad_norm": 0.4987516701221466,
+      "learning_rate": 9.286821705426358e-06,
+      "loss": 0.4078,
+      "step": 599
+    },
+    {
+      "epoch": 0.27902650751821423,
+      "grad_norm": 0.5451740026473999,
+      "learning_rate": 9.30232558139535e-06,
+      "loss": 0.4184,
+      "step": 600
+    },
+    {
+      "epoch": 0.27949155169741124,
+      "grad_norm": 0.52000892162323,
+      "learning_rate": 9.31782945736434e-06,
+      "loss": 0.422,
+      "step": 601
+    },
+    {
+      "epoch": 0.27995659587660826,
+      "grad_norm": 0.524783194065094,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.4121,
+      "step": 602
+    },
+    {
+      "epoch": 0.2804216400558053,
+      "grad_norm": 0.5225769877433777,
+      "learning_rate": 9.348837209302326e-06,
+      "loss": 0.4339,
+      "step": 603
+    },
+    {
+      "epoch": 0.28088668423500235,
+      "grad_norm": 0.563931405544281,
+      "learning_rate": 9.364341085271318e-06,
+      "loss": 0.4212,
+      "step": 604
+    },
+    {
+      "epoch": 0.28135172841419936,
+      "grad_norm": 0.6426302790641785,
+      "learning_rate": 9.37984496124031e-06,
+      "loss": 0.4,
+      "step": 605
+    },
+    {
+      "epoch": 0.2818167725933964,
+      "grad_norm": 0.476382851600647,
+      "learning_rate": 9.395348837209302e-06,
+      "loss": 0.399,
+      "step": 606
+    },
+    {
+      "epoch": 0.2822818167725934,
+      "grad_norm": 0.6564604640007019,
+      "learning_rate": 9.410852713178296e-06,
+      "loss": 0.4106,
+      "step": 607
+    },
+    {
+      "epoch": 0.2827468609517904,
+      "grad_norm": 0.5618086457252502,
+      "learning_rate": 9.426356589147288e-06,
+      "loss": 0.4193,
+      "step": 608
+    },
+    {
+      "epoch": 0.28321190513098743,
+      "grad_norm": 0.6813092231750488,
+      "learning_rate": 9.44186046511628e-06,
+      "loss": 0.4184,
+      "step": 609
+    },
+    {
+      "epoch": 0.28367694931018445,
+      "grad_norm": 0.5115700960159302,
+      "learning_rate": 9.457364341085272e-06,
+      "loss": 0.4251,
+      "step": 610
+    },
+    {
+      "epoch": 0.2841419934893815,
+      "grad_norm": 0.620830774307251,
+      "learning_rate": 9.472868217054264e-06,
+      "loss": 0.3938,
+      "step": 611
+    },
+    {
+      "epoch": 0.28460703766857853,
+      "grad_norm": 0.4752020239830017,
+      "learning_rate": 9.488372093023258e-06,
+      "loss": 0.4005,
+      "step": 612
+    },
+    {
+      "epoch": 0.28507208184777555,
+      "grad_norm": 0.5442911386489868,
+      "learning_rate": 9.503875968992248e-06,
+      "loss": 0.4185,
+      "step": 613
+    },
+    {
+      "epoch": 0.28553712602697257,
+      "grad_norm": 0.5141138434410095,
+      "learning_rate": 9.51937984496124e-06,
+      "loss": 0.4421,
+      "step": 614
+    },
+    {
+      "epoch": 0.2860021702061696,
+      "grad_norm": 0.5211554169654846,
+      "learning_rate": 9.534883720930234e-06,
+      "loss": 0.402,
+      "step": 615
+    },
+    {
+      "epoch": 0.2864672143853666,
+      "grad_norm": 0.5350854396820068,
+      "learning_rate": 9.550387596899226e-06,
+      "loss": 0.4229,
+      "step": 616
+    },
+    {
+      "epoch": 0.2869322585645636,
+      "grad_norm": 0.5418005585670471,
+      "learning_rate": 9.565891472868218e-06,
+      "loss": 0.4264,
+      "step": 617
+    },
+    {
+      "epoch": 0.28739730274376063,
+      "grad_norm": 0.5651183128356934,
+      "learning_rate": 9.58139534883721e-06,
+      "loss": 0.412,
+      "step": 618
+    },
+    {
+      "epoch": 0.2878623469229577,
+      "grad_norm": 0.4958003759384155,
+      "learning_rate": 9.596899224806202e-06,
+      "loss": 0.3929,
+      "step": 619
+    },
+    {
+      "epoch": 0.2883273911021547,
+      "grad_norm": 0.6447129845619202,
+      "learning_rate": 9.612403100775196e-06,
+      "loss": 0.4177,
+      "step": 620
+    },
+    {
+      "epoch": 0.28879243528135173,
+      "grad_norm": 0.5777575969696045,
+      "learning_rate": 9.627906976744188e-06,
+      "loss": 0.4193,
+      "step": 621
+    },
+    {
+      "epoch": 0.28925747946054875,
+      "grad_norm": 0.7008224129676819,
+      "learning_rate": 9.643410852713178e-06,
+      "loss": 0.423,
+      "step": 622
+    },
+    {
+      "epoch": 0.28972252363974577,
+      "grad_norm": 0.4426723122596741,
+      "learning_rate": 9.658914728682172e-06,
+      "loss": 0.4038,
+      "step": 623
+    },
+    {
+      "epoch": 0.2901875678189428,
+      "grad_norm": 0.6383287310600281,
+      "learning_rate": 9.674418604651164e-06,
+      "loss": 0.4213,
+      "step": 624
+    },
+    {
+      "epoch": 0.2906526119981398,
+      "grad_norm": 0.4844573438167572,
+      "learning_rate": 9.689922480620156e-06,
+      "loss": 0.4355,
+      "step": 625
+    },
+    {
+      "epoch": 0.29111765617733687,
+      "grad_norm": 0.6763868927955627,
+      "learning_rate": 9.705426356589148e-06,
+      "loss": 0.4122,
+      "step": 626
+    },
+    {
+      "epoch": 0.2915827003565339,
+      "grad_norm": 0.508306622505188,
+      "learning_rate": 9.72093023255814e-06,
+      "loss": 0.4067,
+      "step": 627
+    },
+    {
+      "epoch": 0.2920477445357309,
+      "grad_norm": 0.5674280524253845,
+      "learning_rate": 9.736434108527133e-06,
+      "loss": 0.4359,
+      "step": 628
+    },
+    {
+      "epoch": 0.2925127887149279,
+      "grad_norm": 0.5861987471580505,
+      "learning_rate": 9.751937984496125e-06,
+      "loss": 0.3948,
+      "step": 629
+    },
+    {
+      "epoch": 0.29297783289412493,
+      "grad_norm": 0.5261598825454712,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 0.4339,
+      "step": 630
+    },
+    {
+      "epoch": 0.29344287707332195,
+      "grad_norm": 0.4901929199695587,
+      "learning_rate": 9.78294573643411e-06,
+      "loss": 0.3916,
+      "step": 631
+    },
+    {
+      "epoch": 0.29390792125251897,
+      "grad_norm": 0.5431584715843201,
+      "learning_rate": 9.798449612403101e-06,
+      "loss": 0.4372,
+      "step": 632
+    },
+    {
+      "epoch": 0.29437296543171604,
+      "grad_norm": 0.4933854043483734,
+      "learning_rate": 9.813953488372093e-06,
+      "loss": 0.4107,
+      "step": 633
+    },
+    {
+      "epoch": 0.29483800961091305,
+      "grad_norm": 0.5128764510154724,
+      "learning_rate": 9.829457364341087e-06,
+      "loss": 0.4222,
+      "step": 634
+    },
+    {
+      "epoch": 0.29530305379011007,
+      "grad_norm": 0.577727198600769,
+      "learning_rate": 9.844961240310077e-06,
+      "loss": 0.4417,
+      "step": 635
+    },
+    {
+      "epoch": 0.2957680979693071,
+      "grad_norm": 0.5306776762008667,
+      "learning_rate": 9.860465116279071e-06,
+      "loss": 0.4031,
+      "step": 636
+    },
+    {
+      "epoch": 0.2962331421485041,
+      "grad_norm": 0.5434820055961609,
+      "learning_rate": 9.875968992248063e-06,
+      "loss": 0.4148,
+      "step": 637
+    },
+    {
+      "epoch": 0.2966981863277011,
+      "grad_norm": 0.5308932065963745,
+      "learning_rate": 9.891472868217055e-06,
+      "loss": 0.4017,
+      "step": 638
+    },
+    {
+      "epoch": 0.29716323050689814,
+      "grad_norm": 0.6192681193351746,
+      "learning_rate": 9.906976744186047e-06,
+      "loss": 0.4236,
+      "step": 639
+    },
+    {
+      "epoch": 0.2976282746860952,
+      "grad_norm": 0.5245199799537659,
+      "learning_rate": 9.922480620155039e-06,
+      "loss": 0.4206,
+      "step": 640
+    },
+    {
+      "epoch": 0.2980933188652922,
+      "grad_norm": 0.6172415018081665,
+      "learning_rate": 9.937984496124031e-06,
+      "loss": 0.4147,
+      "step": 641
+    },
+    {
+      "epoch": 0.29855836304448924,
+      "grad_norm": 0.49070969223976135,
+      "learning_rate": 9.953488372093025e-06,
+      "loss": 0.4104,
+      "step": 642
+    },
+    {
+      "epoch": 0.29902340722368626,
+      "grad_norm": 0.5000258684158325,
+      "learning_rate": 9.968992248062017e-06,
+      "loss": 0.4026,
+      "step": 643
+    },
+    {
+      "epoch": 0.29948845140288327,
+      "grad_norm": 0.601202130317688,
+      "learning_rate": 9.984496124031009e-06,
+      "loss": 0.4187,
+      "step": 644
+    },
+    {
+      "epoch": 0.2999534955820803,
+      "grad_norm": 0.5487759113311768,
+      "learning_rate": 1e-05,
+      "loss": 0.3904,
+      "step": 645
+    },
+    {
+      "epoch": 0.3004185397612773,
+      "grad_norm": 0.4648313522338867,
+      "learning_rate": 9.99999926779061e-06,
+      "loss": 0.4065,
+      "step": 646
+    },
+    {
+      "epoch": 0.3008835839404743,
+      "grad_norm": 0.542242705821991,
+      "learning_rate": 9.999997071162647e-06,
+      "loss": 0.403,
+      "step": 647
+    },
+    {
+      "epoch": 0.3013486281196714,
+      "grad_norm": 0.5556986331939697,
+      "learning_rate": 9.999993410116758e-06,
+      "loss": 0.4012,
+      "step": 648
+    },
+    {
+      "epoch": 0.3018136722988684,
+      "grad_norm": 0.4553111493587494,
+      "learning_rate": 9.999988284654016e-06,
+      "loss": 0.423,
+      "step": 649
+    },
+    {
+      "epoch": 0.3022787164780654,
+      "grad_norm": 0.5509373545646667,
+      "learning_rate": 9.999981694775921e-06,
+      "loss": 0.4202,
+      "step": 650
+    },
+    {
+      "epoch": 0.30274376065726244,
+      "grad_norm": 0.4689308702945709,
+      "learning_rate": 9.999973640484402e-06,
+      "loss": 0.3928,
+      "step": 651
+    },
+    {
+      "epoch": 0.30320880483645946,
+      "grad_norm": 0.4984801411628723,
+      "learning_rate": 9.99996412178182e-06,
+      "loss": 0.4171,
+      "step": 652
+    },
+    {
+      "epoch": 0.30367384901565647,
+      "grad_norm": 0.5575225949287415,
+      "learning_rate": 9.999953138670961e-06,
+      "loss": 0.4134,
+      "step": 653
+    },
+    {
+      "epoch": 0.3041388931948535,
+      "grad_norm": 0.5620846748352051,
+      "learning_rate": 9.999940691155043e-06,
+      "loss": 0.4043,
+      "step": 654
+    },
+    {
+      "epoch": 0.30460393737405056,
+      "grad_norm": 0.5309882760047913,
+      "learning_rate": 9.999926779237713e-06,
+      "loss": 0.4286,
+      "step": 655
+    },
+    {
+      "epoch": 0.3050689815532476,
+      "grad_norm": 0.5248701572418213,
+      "learning_rate": 9.999911402923043e-06,
+      "loss": 0.4163,
+      "step": 656
+    },
+    {
+      "epoch": 0.3055340257324446,
+      "grad_norm": 0.5015712380409241,
+      "learning_rate": 9.999894562215538e-06,
+      "loss": 0.4136,
+      "step": 657
+    },
+    {
+      "epoch": 0.3059990699116416,
+      "grad_norm": 0.546489953994751,
+      "learning_rate": 9.999876257120127e-06,
+      "loss": 0.4217,
+      "step": 658
+    },
+    {
+      "epoch": 0.3064641140908386,
+      "grad_norm": 0.4728710651397705,
+      "learning_rate": 9.999856487642177e-06,
+      "loss": 0.3868,
+      "step": 659
+    },
+    {
+      "epoch": 0.30692915827003564,
+      "grad_norm": 0.5765323638916016,
+      "learning_rate": 9.999835253787472e-06,
+      "loss": 0.4518,
+      "step": 660
+    },
+    {
+      "epoch": 0.30739420244923266,
+      "grad_norm": 0.5118497014045715,
+      "learning_rate": 9.999812555562239e-06,
+      "loss": 0.432,
+      "step": 661
+    },
+    {
+      "epoch": 0.30785924662842973,
+      "grad_norm": 0.5273929238319397,
+      "learning_rate": 9.999788392973117e-06,
+      "loss": 0.3831,
+      "step": 662
+    },
+    {
+      "epoch": 0.30832429080762674,
+      "grad_norm": 0.5201013088226318,
+      "learning_rate": 9.99976276602719e-06,
+      "loss": 0.4127,
+      "step": 663
+    },
+    {
+      "epoch": 0.30878933498682376,
+      "grad_norm": 0.6540030837059021,
+      "learning_rate": 9.999735674731959e-06,
+      "loss": 0.4359,
+      "step": 664
+    },
+    {
+      "epoch": 0.3092543791660208,
+      "grad_norm": 0.5233065485954285,
+      "learning_rate": 9.999707119095361e-06,
+      "loss": 0.428,
+      "step": 665
+    },
+    {
+      "epoch": 0.3097194233452178,
+      "grad_norm": 0.5521329641342163,
+      "learning_rate": 9.99967709912576e-06,
+      "loss": 0.4054,
+      "step": 666
+    },
+    {
+      "epoch": 0.3101844675244148,
+      "grad_norm": 0.6145597696304321,
+      "learning_rate": 9.999645614831946e-06,
+      "loss": 0.3983,
+      "step": 667
+    },
+    {
+      "epoch": 0.3106495117036118,
+      "grad_norm": 0.6071444153785706,
+      "learning_rate": 9.999612666223139e-06,
+      "loss": 0.4286,
+      "step": 668
+    },
+    {
+      "epoch": 0.31111455588280884,
+      "grad_norm": 0.5950531363487244,
+      "learning_rate": 9.999578253308994e-06,
+      "loss": 0.4325,
+      "step": 669
+    },
+    {
+      "epoch": 0.3115796000620059,
+      "grad_norm": 0.608697772026062,
+      "learning_rate": 9.99954237609959e-06,
+      "loss": 0.3965,
+      "step": 670
+    },
+    {
+      "epoch": 0.31204464424120293,
+      "grad_norm": 0.5764162540435791,
+      "learning_rate": 9.999505034605428e-06,
+      "loss": 0.4166,
+      "step": 671
+    },
+    {
+      "epoch": 0.31250968842039994,
+      "grad_norm": 0.6588258147239685,
+      "learning_rate": 9.999466228837452e-06,
+      "loss": 0.4122,
+      "step": 672
+    },
+    {
+      "epoch": 0.31297473259959696,
+      "grad_norm": 0.5878240466117859,
+      "learning_rate": 9.999425958807023e-06,
+      "loss": 0.4289,
+      "step": 673
+    },
+    {
+      "epoch": 0.313439776778794,
+      "grad_norm": 0.6487173438072205,
+      "learning_rate": 9.999384224525938e-06,
+      "loss": 0.4348,
+      "step": 674
+    },
+    {
+      "epoch": 0.313904820957991,
+      "grad_norm": 0.5158074498176575,
+      "learning_rate": 9.99934102600642e-06,
+      "loss": 0.402,
+      "step": 675
+    },
+    {
+      "epoch": 0.314369865137188,
+      "grad_norm": 0.6373001933097839,
+      "learning_rate": 9.999296363261118e-06,
+      "loss": 0.4098,
+      "step": 676
+    },
+    {
+      "epoch": 0.3148349093163851,
+      "grad_norm": 0.5033225417137146,
+      "learning_rate": 9.999250236303118e-06,
+      "loss": 0.4035,
+      "step": 677
+    },
+    {
+      "epoch": 0.3152999534955821,
+      "grad_norm": 0.4822746217250824,
+      "learning_rate": 9.999202645145927e-06,
+      "loss": 0.4164,
+      "step": 678
+    },
+    {
+      "epoch": 0.3157649976747791,
+      "grad_norm": 0.4495120644569397,
+      "learning_rate": 9.999153589803484e-06,
+      "loss": 0.3747,
+      "step": 679
+    },
+    {
+      "epoch": 0.31623004185397613,
+      "grad_norm": 0.5534586310386658,
+      "learning_rate": 9.999103070290155e-06,
+      "loss": 0.4011,
+      "step": 680
+    },
+    {
+      "epoch": 0.31669508603317315,
+      "grad_norm": 0.4584212601184845,
+      "learning_rate": 9.999051086620738e-06,
+      "loss": 0.4116,
+      "step": 681
+    },
+    {
+      "epoch": 0.31716013021237016,
+      "grad_norm": 0.6077730655670166,
+      "learning_rate": 9.998997638810462e-06,
+      "loss": 0.4163,
+      "step": 682
+    },
+    {
+      "epoch": 0.3176251743915672,
+      "grad_norm": 0.4868007004261017,
+      "learning_rate": 9.998942726874974e-06,
+      "loss": 0.4075,
+      "step": 683
+    },
+    {
+      "epoch": 0.31809021857076425,
+      "grad_norm": 0.5131934285163879,
+      "learning_rate": 9.99888635083036e-06,
+      "loss": 0.4178,
+      "step": 684
+    },
+    {
+      "epoch": 0.31855526274996127,
+      "grad_norm": 0.5666374564170837,
+      "learning_rate": 9.998828510693133e-06,
+      "loss": 0.4147,
+      "step": 685
+    },
+    {
+      "epoch": 0.3190203069291583,
+      "grad_norm": 0.6439737677574158,
+      "learning_rate": 9.99876920648023e-06,
+      "loss": 0.4148,
+      "step": 686
+    },
+    {
+      "epoch": 0.3194853511083553,
+      "grad_norm": 0.5670461058616638,
+      "learning_rate": 9.998708438209022e-06,
+      "loss": 0.4088,
+      "step": 687
+    },
+    {
+      "epoch": 0.3199503952875523,
+      "grad_norm": 0.6633548736572266,
+      "learning_rate": 9.99864620589731e-06,
+      "loss": 0.3982,
+      "step": 688
+    },
+    {
+      "epoch": 0.32041543946674933,
+      "grad_norm": 0.6031489968299866,
+      "learning_rate": 9.998582509563315e-06,
+      "loss": 0.4095,
+      "step": 689
+    },
+    {
+      "epoch": 0.32088048364594635,
+      "grad_norm": 0.49488380551338196,
+      "learning_rate": 9.998517349225698e-06,
+      "loss": 0.3967,
+      "step": 690
+    },
+    {
+      "epoch": 0.32134552782514336,
+      "grad_norm": 0.6333311200141907,
+      "learning_rate": 9.99845072490354e-06,
+      "loss": 0.4425,
+      "step": 691
+    },
+    {
+      "epoch": 0.32181057200434043,
+      "grad_norm": 0.5096359252929688,
+      "learning_rate": 9.998382636616355e-06,
+      "loss": 0.4033,
+      "step": 692
+    },
+    {
+      "epoch": 0.32227561618353745,
+      "grad_norm": 0.502375602722168,
+      "learning_rate": 9.998313084384086e-06,
+      "loss": 0.3965,
+      "step": 693
+    },
+    {
+      "epoch": 0.32274066036273447,
+      "grad_norm": 0.6580330729484558,
+      "learning_rate": 9.998242068227103e-06,
+      "loss": 0.4052,
+      "step": 694
+    },
+    {
+      "epoch": 0.3232057045419315,
+      "grad_norm": 0.5036015510559082,
+      "learning_rate": 9.998169588166204e-06,
+      "loss": 0.4154,
+      "step": 695
+    },
+    {
+      "epoch": 0.3236707487211285,
+      "grad_norm": 0.5556278228759766,
+      "learning_rate": 9.99809564422262e-06,
+      "loss": 0.4296,
+      "step": 696
+    },
+    {
+      "epoch": 0.3241357929003255,
+      "grad_norm": 0.635485053062439,
+      "learning_rate": 9.998020236418008e-06,
+      "loss": 0.4379,
+      "step": 697
+    },
+    {
+      "epoch": 0.32460083707952253,
+      "grad_norm": 0.48889610171318054,
+      "learning_rate": 9.997943364774451e-06,
+      "loss": 0.414,
+      "step": 698
+    },
+    {
+      "epoch": 0.3250658812587196,
+      "grad_norm": 0.4967147707939148,
+      "learning_rate": 9.997865029314464e-06,
+      "loss": 0.4011,
+      "step": 699
+    },
+    {
+      "epoch": 0.3255309254379166,
+      "grad_norm": 0.595470666885376,
+      "learning_rate": 9.997785230060993e-06,
+      "loss": 0.438,
+      "step": 700
+    },
+    {
+      "epoch": 0.32599596961711363,
+      "grad_norm": 0.533093273639679,
+      "learning_rate": 9.997703967037406e-06,
+      "loss": 0.4152,
+      "step": 701
+    },
+    {
+      "epoch": 0.32646101379631065,
+      "grad_norm": 0.5301925539970398,
+      "learning_rate": 9.997621240267505e-06,
+      "loss": 0.4096,
+      "step": 702
+    },
+    {
+      "epoch": 0.32692605797550767,
+      "grad_norm": 0.5256016254425049,
+      "learning_rate": 9.997537049775522e-06,
+      "loss": 0.4057,
+      "step": 703
+    },
+    {
+      "epoch": 0.3273911021547047,
+      "grad_norm": 0.6002234816551208,
+      "learning_rate": 9.997451395586112e-06,
+      "loss": 0.4221,
+      "step": 704
+    },
+    {
+      "epoch": 0.3278561463339017,
+      "grad_norm": 0.5671415328979492,
+      "learning_rate": 9.997364277724362e-06,
+      "loss": 0.4038,
+      "step": 705
+    },
+    {
+      "epoch": 0.32832119051309877,
+      "grad_norm": 0.5628596544265747,
+      "learning_rate": 9.997275696215788e-06,
+      "loss": 0.3987,
+      "step": 706
+    },
+    {
+      "epoch": 0.3287862346922958,
+      "grad_norm": 0.6205607652664185,
+      "learning_rate": 9.997185651086336e-06,
+      "loss": 0.4259,
+      "step": 707
+    },
+    {
+      "epoch": 0.3292512788714928,
+      "grad_norm": 0.5086963176727295,
+      "learning_rate": 9.997094142362376e-06,
+      "loss": 0.4084,
+      "step": 708
+    },
+    {
+      "epoch": 0.3297163230506898,
+      "grad_norm": 0.5511040091514587,
+      "learning_rate": 9.99700117007071e-06,
+      "loss": 0.4085,
+      "step": 709
+    },
+    {
+      "epoch": 0.33018136722988684,
+      "grad_norm": 0.5336337089538574,
+      "learning_rate": 9.996906734238568e-06,
+      "loss": 0.4397,
+      "step": 710
+    },
+    {
+      "epoch": 0.33064641140908385,
+      "grad_norm": 0.5333123207092285,
+      "learning_rate": 9.99681083489361e-06,
+      "loss": 0.4234,
+      "step": 711
+    },
+    {
+      "epoch": 0.33111145558828087,
+      "grad_norm": 0.5140913724899292,
+      "learning_rate": 9.99671347206392e-06,
+      "loss": 0.4133,
+      "step": 712
+    },
+    {
+      "epoch": 0.3315764997674779,
+      "grad_norm": 0.5568680167198181,
+      "learning_rate": 9.996614645778019e-06,
+      "loss": 0.4144,
+      "step": 713
+    },
+    {
+      "epoch": 0.33204154394667496,
+      "grad_norm": 0.4299158751964569,
+      "learning_rate": 9.996514356064848e-06,
+      "loss": 0.4325,
+      "step": 714
+    },
+    {
+      "epoch": 0.33250658812587197,
+      "grad_norm": 0.5466892123222351,
+      "learning_rate": 9.996412602953782e-06,
+      "loss": 0.4002,
+      "step": 715
+    },
+    {
+      "epoch": 0.332971632305069,
+      "grad_norm": 0.5300124287605286,
+      "learning_rate": 9.99630938647462e-06,
+      "loss": 0.4113,
+      "step": 716
+    },
+    {
+      "epoch": 0.333436676484266,
+      "grad_norm": 0.4665888845920563,
+      "learning_rate": 9.996204706657597e-06,
+      "loss": 0.4264,
+      "step": 717
+    },
+    {
+      "epoch": 0.333901720663463,
+      "grad_norm": 0.6327192187309265,
+      "learning_rate": 9.99609856353337e-06,
+      "loss": 0.4139,
+      "step": 718
+    },
+    {
+      "epoch": 0.33436676484266004,
+      "grad_norm": 0.5009886622428894,
+      "learning_rate": 9.995990957133024e-06,
+      "loss": 0.3818,
+      "step": 719
+    },
+    {
+      "epoch": 0.33483180902185705,
+      "grad_norm": 0.5028725266456604,
+      "learning_rate": 9.99588188748808e-06,
+      "loss": 0.3934,
+      "step": 720
+    },
+    {
+      "epoch": 0.3352968532010541,
+      "grad_norm": 0.6734301447868347,
+      "learning_rate": 9.995771354630476e-06,
+      "loss": 0.4213,
+      "step": 721
+    },
+    {
+      "epoch": 0.33576189738025114,
+      "grad_norm": 0.5146905183792114,
+      "learning_rate": 9.995659358592592e-06,
+      "loss": 0.387,
+      "step": 722
+    },
+    {
+      "epoch": 0.33622694155944816,
+      "grad_norm": 0.4680449664592743,
+      "learning_rate": 9.995545899407226e-06,
+      "loss": 0.4296,
+      "step": 723
+    },
+    {
+      "epoch": 0.33669198573864517,
+      "grad_norm": 0.5770846605300903,
+      "learning_rate": 9.995430977107612e-06,
+      "loss": 0.4088,
+      "step": 724
+    },
+    {
+      "epoch": 0.3371570299178422,
+      "grad_norm": 0.5241069793701172,
+      "learning_rate": 9.995314591727404e-06,
+      "loss": 0.4065,
+      "step": 725
+    },
+    {
+      "epoch": 0.3376220740970392,
+      "grad_norm": 0.46266067028045654,
+      "learning_rate": 9.995196743300693e-06,
+      "loss": 0.4014,
+      "step": 726
+    },
+    {
+      "epoch": 0.3380871182762362,
+      "grad_norm": 0.4953598380088806,
+      "learning_rate": 9.995077431861992e-06,
+      "loss": 0.3707,
+      "step": 727
+    },
+    {
+      "epoch": 0.3385521624554333,
+      "grad_norm": 0.47356897592544556,
+      "learning_rate": 9.994956657446248e-06,
+      "loss": 0.4058,
+      "step": 728
+    },
+    {
+      "epoch": 0.3390172066346303,
+      "grad_norm": 0.47377198934555054,
+      "learning_rate": 9.994834420088832e-06,
+      "loss": 0.4042,
+      "step": 729
+    },
+    {
+      "epoch": 0.3394822508138273,
+      "grad_norm": 0.5034839510917664,
+      "learning_rate": 9.99471071982555e-06,
+      "loss": 0.4001,
+      "step": 730
+    },
+    {
+      "epoch": 0.33994729499302434,
+      "grad_norm": 0.462456613779068,
+      "learning_rate": 9.994585556692624e-06,
+      "loss": 0.3885,
+      "step": 731
+    },
+    {
+      "epoch": 0.34041233917222136,
+      "grad_norm": 0.5245911478996277,
+      "learning_rate": 9.994458930726717e-06,
+      "loss": 0.419,
+      "step": 732
+    },
+    {
+      "epoch": 0.3408773833514184,
+      "grad_norm": 0.4660412669181824,
+      "learning_rate": 9.994330841964916e-06,
+      "loss": 0.4144,
+      "step": 733
+    },
+    {
+      "epoch": 0.3413424275306154,
+      "grad_norm": 0.54203861951828,
+      "learning_rate": 9.994201290444734e-06,
+      "loss": 0.3974,
+      "step": 734
+    },
+    {
+      "epoch": 0.3418074717098124,
+      "grad_norm": 0.5267844200134277,
+      "learning_rate": 9.994070276204115e-06,
+      "loss": 0.428,
+      "step": 735
+    },
+    {
+      "epoch": 0.3422725158890095,
+      "grad_norm": 0.5129567384719849,
+      "learning_rate": 9.993937799281435e-06,
+      "loss": 0.4135,
+      "step": 736
+    },
+    {
+      "epoch": 0.3427375600682065,
+      "grad_norm": 0.5932128429412842,
+      "learning_rate": 9.993803859715488e-06,
+      "loss": 0.4103,
+      "step": 737
+    },
+    {
+      "epoch": 0.3432026042474035,
+      "grad_norm": 0.48848265409469604,
+      "learning_rate": 9.993668457545505e-06,
+      "loss": 0.4022,
+      "step": 738
+    },
+    {
+      "epoch": 0.3436676484266005,
+      "grad_norm": 0.5327843427658081,
+      "learning_rate": 9.993531592811146e-06,
+      "loss": 0.4113,
+      "step": 739
+    },
+    {
+      "epoch": 0.34413269260579754,
+      "grad_norm": 0.571151077747345,
+      "learning_rate": 9.993393265552494e-06,
+      "loss": 0.4191,
+      "step": 740
+    },
+    {
+      "epoch": 0.34459773678499456,
+      "grad_norm": 0.5364315509796143,
+      "learning_rate": 9.993253475810061e-06,
+      "loss": 0.4134,
+      "step": 741
+    },
+    {
+      "epoch": 0.3450627809641916,
+      "grad_norm": 0.5578524470329285,
+      "learning_rate": 9.993112223624793e-06,
+      "loss": 0.4269,
+      "step": 742
+    },
+    {
+      "epoch": 0.34552782514338864,
+      "grad_norm": 0.599855899810791,
+      "learning_rate": 9.992969509038057e-06,
+      "loss": 0.4095,
+      "step": 743
+    },
+    {
+      "epoch": 0.34599286932258566,
+      "grad_norm": 0.48594430088996887,
+      "learning_rate": 9.992825332091654e-06,
+      "loss": 0.4119,
+      "step": 744
+    },
+    {
+      "epoch": 0.3464579135017827,
+      "grad_norm": 0.6377593874931335,
+      "learning_rate": 9.992679692827812e-06,
+      "loss": 0.4411,
+      "step": 745
+    },
+    {
+      "epoch": 0.3469229576809797,
+      "grad_norm": 0.5200589895248413,
+      "learning_rate": 9.992532591289183e-06,
+      "loss": 0.4241,
+      "step": 746
+    },
+    {
+      "epoch": 0.3473880018601767,
+      "grad_norm": 0.5711746215820312,
+      "learning_rate": 9.992384027518853e-06,
+      "loss": 0.422,
+      "step": 747
+    },
+    {
+      "epoch": 0.3478530460393737,
+      "grad_norm": 0.5109982490539551,
+      "learning_rate": 9.992234001560333e-06,
+      "loss": 0.3909,
+      "step": 748
+    },
+    {
+      "epoch": 0.34831809021857074,
+      "grad_norm": 0.6026611328125,
+      "learning_rate": 9.992082513457564e-06,
+      "loss": 0.4277,
+      "step": 749
+    },
+    {
+      "epoch": 0.3487831343977678,
+      "grad_norm": 0.5200704336166382,
+      "learning_rate": 9.991929563254913e-06,
+      "loss": 0.4007,
+      "step": 750
+    },
+    {
+      "epoch": 0.34924817857696483,
+      "grad_norm": 0.5779603123664856,
+      "learning_rate": 9.99177515099718e-06,
+      "loss": 0.3991,
+      "step": 751
+    },
+    {
+      "epoch": 0.34971322275616185,
+      "grad_norm": 0.5946140885353088,
+      "learning_rate": 9.991619276729585e-06,
+      "loss": 0.439,
+      "step": 752
+    },
+    {
+      "epoch": 0.35017826693535886,
+      "grad_norm": 0.5260143280029297,
+      "learning_rate": 9.991461940497786e-06,
+      "loss": 0.4202,
+      "step": 753
+    },
+    {
+      "epoch": 0.3506433111145559,
+      "grad_norm": 0.5628990530967712,
+      "learning_rate": 9.99130314234786e-06,
+      "loss": 0.3997,
+      "step": 754
+    },
+    {
+      "epoch": 0.3511083552937529,
+      "grad_norm": 0.6147162318229675,
+      "learning_rate": 9.99114288232632e-06,
+      "loss": 0.4137,
+      "step": 755
+    },
+    {
+      "epoch": 0.3515733994729499,
+      "grad_norm": 0.5871340036392212,
+      "learning_rate": 9.990981160480098e-06,
+      "loss": 0.4044,
+      "step": 756
+    },
+    {
+      "epoch": 0.3520384436521469,
+      "grad_norm": 0.5283817648887634,
+      "learning_rate": 9.990817976856566e-06,
+      "loss": 0.3815,
+      "step": 757
+    },
+    {
+      "epoch": 0.352503487831344,
+      "grad_norm": 0.5750249624252319,
+      "learning_rate": 9.990653331503515e-06,
+      "loss": 0.4088,
+      "step": 758
+    },
+    {
+      "epoch": 0.352968532010541,
+      "grad_norm": 0.5326053500175476,
+      "learning_rate": 9.990487224469167e-06,
+      "loss": 0.4239,
+      "step": 759
+    },
+    {
+      "epoch": 0.35343357618973803,
+      "grad_norm": 0.50933837890625,
+      "learning_rate": 9.990319655802171e-06,
+      "loss": 0.4022,
+      "step": 760
+    },
+    {
+      "epoch": 0.35389862036893505,
+      "grad_norm": 0.4641541838645935,
+      "learning_rate": 9.990150625551609e-06,
+      "loss": 0.3847,
+      "step": 761
+    },
+    {
+      "epoch": 0.35436366454813206,
+      "grad_norm": 0.49445950984954834,
+      "learning_rate": 9.989980133766983e-06,
+      "loss": 0.395,
+      "step": 762
+    },
+    {
+      "epoch": 0.3548287087273291,
+      "grad_norm": 0.5578946471214294,
+      "learning_rate": 9.989808180498229e-06,
+      "loss": 0.4031,
+      "step": 763
+    },
+    {
+      "epoch": 0.3552937529065261,
+      "grad_norm": 0.47986650466918945,
+      "learning_rate": 9.98963476579571e-06,
+      "loss": 0.4222,
+      "step": 764
+    },
+    {
+      "epoch": 0.35575879708572317,
+      "grad_norm": 0.5039169788360596,
+      "learning_rate": 9.989459889710214e-06,
+      "loss": 0.3973,
+      "step": 765
+    },
+    {
+      "epoch": 0.3562238412649202,
+      "grad_norm": 0.47546157240867615,
+      "learning_rate": 9.98928355229296e-06,
+      "loss": 0.3891,
+      "step": 766
+    },
+    {
+      "epoch": 0.3566888854441172,
+      "grad_norm": 0.4875746965408325,
+      "learning_rate": 9.989105753595599e-06,
+      "loss": 0.3952,
+      "step": 767
+    },
+    {
+      "epoch": 0.3571539296233142,
+      "grad_norm": 0.5113381743431091,
+      "learning_rate": 9.988926493670198e-06,
+      "loss": 0.3624,
+      "step": 768
+    },
+    {
+      "epoch": 0.35761897380251123,
+      "grad_norm": 0.4663374722003937,
+      "learning_rate": 9.988745772569266e-06,
+      "loss": 0.417,
+      "step": 769
+    },
+    {
+      "epoch": 0.35808401798170825,
+      "grad_norm": 0.5435273051261902,
+      "learning_rate": 9.988563590345728e-06,
+      "loss": 0.4112,
+      "step": 770
+    },
+    {
+      "epoch": 0.35854906216090526,
+      "grad_norm": 0.46149757504463196,
+      "learning_rate": 9.988379947052944e-06,
+      "loss": 0.3915,
+      "step": 771
+    },
+    {
+      "epoch": 0.35901410634010233,
+      "grad_norm": 0.5259143710136414,
+      "learning_rate": 9.988194842744701e-06,
+      "loss": 0.4186,
+      "step": 772
+    },
+    {
+      "epoch": 0.35947915051929935,
+      "grad_norm": 0.4799407422542572,
+      "learning_rate": 9.988008277475214e-06,
+      "loss": 0.4059,
+      "step": 773
+    },
+    {
+      "epoch": 0.35994419469849637,
+      "grad_norm": 0.5095851421356201,
+      "learning_rate": 9.987820251299121e-06,
+      "loss": 0.4179,
+      "step": 774
+    },
+    {
+      "epoch": 0.3604092388776934,
+      "grad_norm": 0.4954157769680023,
+      "learning_rate": 9.987630764271497e-06,
+      "loss": 0.3825,
+      "step": 775
+    },
+    {
+      "epoch": 0.3608742830568904,
+      "grad_norm": 0.5038895010948181,
+      "learning_rate": 9.987439816447836e-06,
+      "loss": 0.4293,
+      "step": 776
+    },
+    {
+      "epoch": 0.3613393272360874,
+      "grad_norm": 0.5088568925857544,
+      "learning_rate": 9.987247407884064e-06,
+      "loss": 0.4113,
+      "step": 777
+    },
+    {
+      "epoch": 0.36180437141528443,
+      "grad_norm": 0.5863829851150513,
+      "learning_rate": 9.987053538636535e-06,
+      "loss": 0.3904,
+      "step": 778
+    },
+    {
+      "epoch": 0.3622694155944815,
+      "grad_norm": 0.44420042634010315,
+      "learning_rate": 9.986858208762032e-06,
+      "loss": 0.4165,
+      "step": 779
+    },
+    {
+      "epoch": 0.3627344597736785,
+      "grad_norm": 0.5741595029830933,
+      "learning_rate": 9.986661418317759e-06,
+      "loss": 0.4164,
+      "step": 780
+    },
+    {
+      "epoch": 0.36319950395287554,
+      "grad_norm": 0.5108224749565125,
+      "learning_rate": 9.986463167361358e-06,
+      "loss": 0.4025,
+      "step": 781
+    },
+    {
+      "epoch": 0.36366454813207255,
+      "grad_norm": 0.6276097893714905,
+      "learning_rate": 9.986263455950888e-06,
+      "loss": 0.4264,
+      "step": 782
+    },
+    {
+      "epoch": 0.36412959231126957,
+      "grad_norm": 0.5375624895095825,
+      "learning_rate": 9.986062284144848e-06,
+      "loss": 0.4184,
+      "step": 783
+    },
+    {
+      "epoch": 0.3645946364904666,
+      "grad_norm": 0.6147474050521851,
+      "learning_rate": 9.985859652002152e-06,
+      "loss": 0.3896,
+      "step": 784
+    },
+    {
+      "epoch": 0.3650596806696636,
+      "grad_norm": 0.586616039276123,
+      "learning_rate": 9.985655559582152e-06,
+      "loss": 0.4102,
+      "step": 785
+    },
+    {
+      "epoch": 0.3655247248488606,
+      "grad_norm": 0.5644299387931824,
+      "learning_rate": 9.985450006944621e-06,
+      "loss": 0.3909,
+      "step": 786
+    },
+    {
+      "epoch": 0.3659897690280577,
+      "grad_norm": 0.5941767692565918,
+      "learning_rate": 9.98524299414976e-06,
+      "loss": 0.4318,
+      "step": 787
+    },
+    {
+      "epoch": 0.3664548132072547,
+      "grad_norm": 0.5848608613014221,
+      "learning_rate": 9.985034521258206e-06,
+      "loss": 0.4178,
+      "step": 788
+    },
+    {
+      "epoch": 0.3669198573864517,
+      "grad_norm": 0.5143994688987732,
+      "learning_rate": 9.98482458833101e-06,
+      "loss": 0.395,
+      "step": 789
+    },
+    {
+      "epoch": 0.36738490156564874,
+      "grad_norm": 0.560663104057312,
+      "learning_rate": 9.984613195429662e-06,
+      "loss": 0.4118,
+      "step": 790
+    },
+    {
+      "epoch": 0.36784994574484575,
+      "grad_norm": 0.520053505897522,
+      "learning_rate": 9.984400342616076e-06,
+      "loss": 0.3849,
+      "step": 791
+    },
+    {
+      "epoch": 0.36831498992404277,
+      "grad_norm": 0.5158041715621948,
+      "learning_rate": 9.984186029952591e-06,
+      "loss": 0.4338,
+      "step": 792
+    },
+    {
+      "epoch": 0.3687800341032398,
+      "grad_norm": 0.5674142837524414,
+      "learning_rate": 9.983970257501978e-06,
+      "loss": 0.3954,
+      "step": 793
+    },
+    {
+      "epoch": 0.36924507828243686,
+      "grad_norm": 0.4826439321041107,
+      "learning_rate": 9.983753025327431e-06,
+      "loss": 0.3815,
+      "step": 794
+    },
+    {
+      "epoch": 0.36971012246163387,
+      "grad_norm": 0.4967092275619507,
+      "learning_rate": 9.983534333492575e-06,
+      "loss": 0.41,
+      "step": 795
+    },
+    {
+      "epoch": 0.3701751666408309,
+      "grad_norm": 0.5571979284286499,
+      "learning_rate": 9.983314182061461e-06,
+      "loss": 0.4217,
+      "step": 796
+    },
+    {
+      "epoch": 0.3706402108200279,
+      "grad_norm": 0.5594668984413147,
+      "learning_rate": 9.983092571098569e-06,
+      "loss": 0.3869,
+      "step": 797
+    },
+    {
+      "epoch": 0.3711052549992249,
+      "grad_norm": 0.50605309009552,
+      "learning_rate": 9.982869500668804e-06,
+      "loss": 0.435,
+      "step": 798
+    },
+    {
+      "epoch": 0.37157029917842194,
+      "grad_norm": 0.6257908344268799,
+      "learning_rate": 9.982644970837499e-06,
+      "loss": 0.408,
+      "step": 799
+    },
+    {
+      "epoch": 0.37203534335761895,
+      "grad_norm": 0.5521469712257385,
+      "learning_rate": 9.982418981670414e-06,
+      "loss": 0.4362,
+      "step": 800
+    },
+    {
+      "epoch": 0.372500387536816,
+      "grad_norm": 0.6889628171920776,
+      "learning_rate": 9.982191533233742e-06,
+      "loss": 0.4091,
+      "step": 801
+    },
+    {
+      "epoch": 0.37296543171601304,
+      "grad_norm": 0.5120216608047485,
+      "learning_rate": 9.981962625594094e-06,
+      "loss": 0.4025,
+      "step": 802
+    },
+    {
+      "epoch": 0.37343047589521006,
+      "grad_norm": 0.6625473499298096,
+      "learning_rate": 9.981732258818519e-06,
+      "loss": 0.4212,
+      "step": 803
+    },
+    {
+      "epoch": 0.3738955200744071,
+      "grad_norm": 0.5750365853309631,
+      "learning_rate": 9.981500432974482e-06,
+      "loss": 0.4146,
+      "step": 804
+    },
+    {
+      "epoch": 0.3743605642536041,
+      "grad_norm": 0.5535095930099487,
+      "learning_rate": 9.981267148129884e-06,
+      "loss": 0.4175,
+      "step": 805
+    },
+    {
+      "epoch": 0.3748256084328011,
+      "grad_norm": 0.6257491111755371,
+      "learning_rate": 9.981032404353052e-06,
+      "loss": 0.3994,
+      "step": 806
+    },
+    {
+      "epoch": 0.3752906526119981,
+      "grad_norm": 0.48537763953208923,
+      "learning_rate": 9.980796201712734e-06,
+      "loss": 0.4095,
+      "step": 807
+    },
+    {
+      "epoch": 0.37575569679119514,
+      "grad_norm": 0.5952275395393372,
+      "learning_rate": 9.980558540278113e-06,
+      "loss": 0.4032,
+      "step": 808
+    },
+    {
+      "epoch": 0.3762207409703922,
+      "grad_norm": 0.5576409101486206,
+      "learning_rate": 9.980319420118796e-06,
+      "loss": 0.4195,
+      "step": 809
+    },
+    {
+      "epoch": 0.3766857851495892,
+      "grad_norm": 0.5643303990364075,
+      "learning_rate": 9.980078841304817e-06,
+      "loss": 0.426,
+      "step": 810
+    },
+    {
+      "epoch": 0.37715082932878624,
+      "grad_norm": 0.5821006894111633,
+      "learning_rate": 9.979836803906636e-06,
+      "loss": 0.4227,
+      "step": 811
+    },
+    {
+      "epoch": 0.37761587350798326,
+      "grad_norm": 0.46532806754112244,
+      "learning_rate": 9.979593307995145e-06,
+      "loss": 0.3683,
+      "step": 812
+    },
+    {
+      "epoch": 0.3780809176871803,
+      "grad_norm": 0.534476637840271,
+      "learning_rate": 9.979348353641659e-06,
+      "loss": 0.4112,
+      "step": 813
+    },
+    {
+      "epoch": 0.3785459618663773,
+      "grad_norm": 0.5859639644622803,
+      "learning_rate": 9.979101940917918e-06,
+      "loss": 0.4257,
+      "step": 814
+    },
+    {
+      "epoch": 0.3790110060455743,
+      "grad_norm": 0.4468208849430084,
+      "learning_rate": 9.978854069896096e-06,
+      "loss": 0.4216,
+      "step": 815
+    },
+    {
+      "epoch": 0.3794760502247714,
+      "grad_norm": 0.597046434879303,
+      "learning_rate": 9.97860474064879e-06,
+      "loss": 0.4121,
+      "step": 816
+    },
+    {
+      "epoch": 0.3799410944039684,
+      "grad_norm": 0.5563921332359314,
+      "learning_rate": 9.978353953249023e-06,
+      "loss": 0.4184,
+      "step": 817
+    },
+    {
+      "epoch": 0.3804061385831654,
+      "grad_norm": 0.49227768182754517,
+      "learning_rate": 9.978101707770247e-06,
+      "loss": 0.4037,
+      "step": 818
+    },
+    {
+      "epoch": 0.3808711827623624,
+      "grad_norm": 0.48585137724876404,
+      "learning_rate": 9.977848004286342e-06,
+      "loss": 0.3985,
+      "step": 819
+    },
+    {
+      "epoch": 0.38133622694155944,
+      "grad_norm": 0.500493586063385,
+      "learning_rate": 9.977592842871612e-06,
+      "loss": 0.4261,
+      "step": 820
+    },
+    {
+      "epoch": 0.38180127112075646,
+      "grad_norm": 0.5106586217880249,
+      "learning_rate": 9.97733622360079e-06,
+      "loss": 0.4301,
+      "step": 821
+    },
+    {
+      "epoch": 0.3822663152999535,
+      "grad_norm": 0.4543416500091553,
+      "learning_rate": 9.977078146549036e-06,
+      "loss": 0.4127,
+      "step": 822
+    },
+    {
+      "epoch": 0.38273135947915055,
+      "grad_norm": 0.6294241547584534,
+      "learning_rate": 9.976818611791937e-06,
+      "loss": 0.4152,
+      "step": 823
+    },
+    {
+      "epoch": 0.38319640365834756,
+      "grad_norm": 0.5176417827606201,
+      "learning_rate": 9.976557619405503e-06,
+      "loss": 0.4029,
+      "step": 824
+    },
+    {
+      "epoch": 0.3836614478375446,
+      "grad_norm": 0.527394711971283,
+      "learning_rate": 9.97629516946618e-06,
+      "loss": 0.4007,
+      "step": 825
+    },
+    {
+      "epoch": 0.3841264920167416,
+      "grad_norm": 0.5030791163444519,
+      "learning_rate": 9.976031262050832e-06,
+      "loss": 0.3884,
+      "step": 826
+    },
+    {
+      "epoch": 0.3845915361959386,
+      "grad_norm": 0.5006067156791687,
+      "learning_rate": 9.975765897236754e-06,
+      "loss": 0.4097,
+      "step": 827
+    },
+    {
+      "epoch": 0.3850565803751356,
+      "grad_norm": 0.5824684500694275,
+      "learning_rate": 9.975499075101667e-06,
+      "loss": 0.4203,
+      "step": 828
+    },
+    {
+      "epoch": 0.38552162455433264,
+      "grad_norm": 0.48303505778312683,
+      "learning_rate": 9.975230795723717e-06,
+      "loss": 0.3836,
+      "step": 829
+    },
+    {
+      "epoch": 0.38598666873352966,
+      "grad_norm": 0.5456617474555969,
+      "learning_rate": 9.974961059181482e-06,
+      "loss": 0.427,
+      "step": 830
+    },
+    {
+      "epoch": 0.38645171291272673,
+      "grad_norm": 0.4804321229457855,
+      "learning_rate": 9.97468986555396e-06,
+      "loss": 0.409,
+      "step": 831
+    },
+    {
+      "epoch": 0.38691675709192375,
+      "grad_norm": 0.4723159372806549,
+      "learning_rate": 9.974417214920584e-06,
+      "loss": 0.4049,
+      "step": 832
+    },
+    {
+      "epoch": 0.38738180127112076,
+      "grad_norm": 0.5413050651550293,
+      "learning_rate": 9.974143107361205e-06,
+      "loss": 0.4057,
+      "step": 833
+    },
+    {
+      "epoch": 0.3878468454503178,
+      "grad_norm": 0.5091968774795532,
+      "learning_rate": 9.973867542956104e-06,
+      "loss": 0.4015,
+      "step": 834
+    },
+    {
+      "epoch": 0.3883118896295148,
+      "grad_norm": 0.4703305959701538,
+      "learning_rate": 9.973590521785992e-06,
+      "loss": 0.4009,
+      "step": 835
+    },
+    {
+      "epoch": 0.3887769338087118,
+      "grad_norm": 0.5493623614311218,
+      "learning_rate": 9.973312043932004e-06,
+      "loss": 0.424,
+      "step": 836
+    },
+    {
+      "epoch": 0.3892419779879088,
+      "grad_norm": 0.590499997138977,
+      "learning_rate": 9.9730321094757e-06,
+      "loss": 0.4294,
+      "step": 837
+    },
+    {
+      "epoch": 0.3897070221671059,
+      "grad_norm": 0.4965055584907532,
+      "learning_rate": 9.972750718499067e-06,
+      "loss": 0.3899,
+      "step": 838
+    },
+    {
+      "epoch": 0.3901720663463029,
+      "grad_norm": 0.5021485686302185,
+      "learning_rate": 9.972467871084524e-06,
+      "loss": 0.4507,
+      "step": 839
+    },
+    {
+      "epoch": 0.39063711052549993,
+      "grad_norm": 0.5755704045295715,
+      "learning_rate": 9.97218356731491e-06,
+      "loss": 0.3978,
+      "step": 840
+    },
+    {
+      "epoch": 0.39110215470469695,
+      "grad_norm": 0.5343739986419678,
+      "learning_rate": 9.971897807273492e-06,
+      "loss": 0.3725,
+      "step": 841
+    },
+    {
+      "epoch": 0.39156719888389396,
+      "grad_norm": 0.5048686265945435,
+      "learning_rate": 9.971610591043966e-06,
+      "loss": 0.4254,
+      "step": 842
+    },
+    {
+      "epoch": 0.392032243063091,
+      "grad_norm": 0.5324344038963318,
+      "learning_rate": 9.971321918710452e-06,
+      "loss": 0.3935,
+      "step": 843
+    },
+    {
+      "epoch": 0.392497287242288,
+      "grad_norm": 0.6189853549003601,
+      "learning_rate": 9.9710317903575e-06,
+      "loss": 0.4212,
+      "step": 844
+    },
+    {
+      "epoch": 0.39296233142148507,
+      "grad_norm": 0.48477646708488464,
+      "learning_rate": 9.97074020607008e-06,
+      "loss": 0.4003,
+      "step": 845
+    },
+    {
+      "epoch": 0.3934273756006821,
+      "grad_norm": 0.548075795173645,
+      "learning_rate": 9.970447165933594e-06,
+      "loss": 0.4232,
+      "step": 846
+    },
+    {
+      "epoch": 0.3938924197798791,
+      "grad_norm": 0.5070050954818726,
+      "learning_rate": 9.97015267003387e-06,
+      "loss": 0.4053,
+      "step": 847
+    },
+    {
+      "epoch": 0.3943574639590761,
+      "grad_norm": 0.5388278365135193,
+      "learning_rate": 9.96985671845716e-06,
+      "loss": 0.4132,
+      "step": 848
+    },
+    {
+      "epoch": 0.39482250813827313,
+      "grad_norm": 0.5621599555015564,
+      "learning_rate": 9.969559311290142e-06,
+      "loss": 0.4126,
+      "step": 849
+    },
+    {
+      "epoch": 0.39528755231747015,
+      "grad_norm": 0.500983476638794,
+      "learning_rate": 9.969260448619925e-06,
+      "loss": 0.4072,
+      "step": 850
+    },
+    {
+      "epoch": 0.39575259649666716,
+      "grad_norm": 0.5636142492294312,
+      "learning_rate": 9.968960130534036e-06,
+      "loss": 0.3949,
+      "step": 851
+    },
+    {
+      "epoch": 0.3962176406758642,
+      "grad_norm": 0.5257579684257507,
+      "learning_rate": 9.96865835712044e-06,
+      "loss": 0.3993,
+      "step": 852
+    },
+    {
+      "epoch": 0.39668268485506125,
+      "grad_norm": 0.6351518630981445,
+      "learning_rate": 9.968355128467515e-06,
+      "loss": 0.4175,
+      "step": 853
+    },
+    {
+      "epoch": 0.39714772903425827,
+      "grad_norm": 0.4981622099876404,
+      "learning_rate": 9.968050444664074e-06,
+      "loss": 0.4217,
+      "step": 854
+    },
+    {
+      "epoch": 0.3976127732134553,
+      "grad_norm": 0.5462387800216675,
+      "learning_rate": 9.967744305799358e-06,
+      "loss": 0.4094,
+      "step": 855
+    },
+    {
+      "epoch": 0.3980778173926523,
+      "grad_norm": 0.599955141544342,
+      "learning_rate": 9.967436711963026e-06,
+      "loss": 0.3897,
+      "step": 856
+    },
+    {
+      "epoch": 0.3985428615718493,
+      "grad_norm": 0.4772481918334961,
+      "learning_rate": 9.967127663245167e-06,
+      "loss": 0.426,
+      "step": 857
+    },
+    {
+      "epoch": 0.39900790575104633,
+      "grad_norm": 0.6691299676895142,
+      "learning_rate": 9.966817159736295e-06,
+      "loss": 0.3885,
+      "step": 858
+    },
+    {
+      "epoch": 0.39947294993024335,
+      "grad_norm": 0.6346452832221985,
+      "learning_rate": 9.966505201527357e-06,
+      "loss": 0.4217,
+      "step": 859
+    },
+    {
+      "epoch": 0.3999379941094404,
+      "grad_norm": 0.47586584091186523,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 0.4165,
+      "step": 860
+    },
+    {
+      "epoch": 0.40040303828863744,
+      "grad_norm": 0.6300751566886902,
+      "learning_rate": 9.965876921375165e-06,
+      "loss": 0.4087,
+      "step": 861
+    },
+    {
+      "epoch": 0.40086808246783445,
+      "grad_norm": 0.5367612242698669,
+      "learning_rate": 9.965560599615928e-06,
+      "loss": 0.3919,
+      "step": 862
+    },
+    {
+      "epoch": 0.40133312664703147,
+      "grad_norm": 0.4811550974845886,
+      "learning_rate": 9.965242823524648e-06,
+      "loss": 0.4112,
+      "step": 863
+    },
+    {
+      "epoch": 0.4017981708262285,
+      "grad_norm": 0.6218395233154297,
+      "learning_rate": 9.964923593194394e-06,
+      "loss": 0.3741,
+      "step": 864
+    },
+    {
+      "epoch": 0.4022632150054255,
+      "grad_norm": 0.5514636635780334,
+      "learning_rate": 9.964602908718667e-06,
+      "loss": 0.4253,
+      "step": 865
+    },
+    {
+      "epoch": 0.4027282591846225,
+      "grad_norm": 0.5390025973320007,
+      "learning_rate": 9.964280770191388e-06,
+      "loss": 0.3988,
+      "step": 866
+    },
+    {
+      "epoch": 0.4031933033638196,
+      "grad_norm": 0.64164799451828,
+      "learning_rate": 9.963957177706908e-06,
+      "loss": 0.4129,
+      "step": 867
+    },
+    {
+      "epoch": 0.4036583475430166,
+      "grad_norm": 0.4808669984340668,
+      "learning_rate": 9.96363213136e-06,
+      "loss": 0.4205,
+      "step": 868
+    },
+    {
+      "epoch": 0.4041233917222136,
+      "grad_norm": 0.5149733424186707,
+      "learning_rate": 9.963305631245866e-06,
+      "loss": 0.4049,
+      "step": 869
+    },
+    {
+      "epoch": 0.40458843590141064,
+      "grad_norm": 0.5427307486534119,
+      "learning_rate": 9.962977677460132e-06,
+      "loss": 0.402,
+      "step": 870
+    },
+    {
+      "epoch": 0.40505348008060765,
+      "grad_norm": 0.5210288763046265,
+      "learning_rate": 9.96264827009885e-06,
+      "loss": 0.4063,
+      "step": 871
+    },
+    {
+      "epoch": 0.40551852425980467,
+      "grad_norm": 0.5169207453727722,
+      "learning_rate": 9.962317409258501e-06,
+      "loss": 0.4032,
+      "step": 872
+    },
+    {
+      "epoch": 0.4059835684390017,
+      "grad_norm": 0.5519936084747314,
+      "learning_rate": 9.961985095035987e-06,
+      "loss": 0.3884,
+      "step": 873
+    },
+    {
+      "epoch": 0.4064486126181987,
+      "grad_norm": 0.5302849411964417,
+      "learning_rate": 9.961651327528636e-06,
+      "loss": 0.406,
+      "step": 874
+    },
+    {
+      "epoch": 0.4069136567973958,
+      "grad_norm": 0.5135921835899353,
+      "learning_rate": 9.961316106834202e-06,
+      "loss": 0.3885,
+      "step": 875
+    },
+    {
+      "epoch": 0.4073787009765928,
+      "grad_norm": 0.4829360842704773,
+      "learning_rate": 9.96097943305087e-06,
+      "loss": 0.4054,
+      "step": 876
+    },
+    {
+      "epoch": 0.4078437451557898,
+      "grad_norm": 0.48682668805122375,
+      "learning_rate": 9.960641306277244e-06,
+      "loss": 0.3872,
+      "step": 877
+    },
+    {
+      "epoch": 0.4083087893349868,
+      "grad_norm": 0.5358735918998718,
+      "learning_rate": 9.960301726612355e-06,
+      "loss": 0.4106,
+      "step": 878
+    },
+    {
+      "epoch": 0.40877383351418384,
+      "grad_norm": 0.491878867149353,
+      "learning_rate": 9.959960694155662e-06,
+      "loss": 0.4054,
+      "step": 879
+    },
+    {
+      "epoch": 0.40923887769338085,
+      "grad_norm": 0.4896703362464905,
+      "learning_rate": 9.959618209007045e-06,
+      "loss": 0.3981,
+      "step": 880
+    },
+    {
+      "epoch": 0.40970392187257787,
+      "grad_norm": 0.5020521283149719,
+      "learning_rate": 9.959274271266816e-06,
+      "loss": 0.3881,
+      "step": 881
+    },
+    {
+      "epoch": 0.41016896605177494,
+      "grad_norm": 0.5467438101768494,
+      "learning_rate": 9.958928881035708e-06,
+      "loss": 0.4188,
+      "step": 882
+    },
+    {
+      "epoch": 0.41063401023097196,
+      "grad_norm": 0.455102801322937,
+      "learning_rate": 9.958582038414878e-06,
+      "loss": 0.407,
+      "step": 883
+    },
+    {
+      "epoch": 0.411099054410169,
+      "grad_norm": 0.4519009292125702,
+      "learning_rate": 9.958233743505912e-06,
+      "loss": 0.4008,
+      "step": 884
+    },
+    {
+      "epoch": 0.411564098589366,
+      "grad_norm": 0.5280046463012695,
+      "learning_rate": 9.957883996410821e-06,
+      "loss": 0.3925,
+      "step": 885
+    },
+    {
+      "epoch": 0.412029142768563,
+      "grad_norm": 0.4995780885219574,
+      "learning_rate": 9.95753279723204e-06,
+      "loss": 0.397,
+      "step": 886
+    },
+    {
+      "epoch": 0.41249418694776,
+      "grad_norm": 0.4849075675010681,
+      "learning_rate": 9.957180146072426e-06,
+      "loss": 0.4214,
+      "step": 887
+    },
+    {
+      "epoch": 0.41295923112695704,
+      "grad_norm": 0.548945426940918,
+      "learning_rate": 9.956826043035268e-06,
+      "loss": 0.4044,
+      "step": 888
+    },
+    {
+      "epoch": 0.4134242753061541,
+      "grad_norm": 0.5245530009269714,
+      "learning_rate": 9.956470488224278e-06,
+      "loss": 0.4133,
+      "step": 889
+    },
+    {
+      "epoch": 0.4138893194853511,
+      "grad_norm": 0.5736802816390991,
+      "learning_rate": 9.95611348174359e-06,
+      "loss": 0.3939,
+      "step": 890
+    },
+    {
+      "epoch": 0.41435436366454814,
+      "grad_norm": 0.5248284935951233,
+      "learning_rate": 9.955755023697767e-06,
+      "loss": 0.4058,
+      "step": 891
+    },
+    {
+      "epoch": 0.41481940784374516,
+      "grad_norm": 0.4462113678455353,
+      "learning_rate": 9.955395114191792e-06,
+      "loss": 0.4168,
+      "step": 892
+    },
+    {
+      "epoch": 0.4152844520229422,
+      "grad_norm": 0.6396908164024353,
+      "learning_rate": 9.955033753331082e-06,
+      "loss": 0.4049,
+      "step": 893
+    },
+    {
+      "epoch": 0.4157494962021392,
+      "grad_norm": 0.5714189410209656,
+      "learning_rate": 9.954670941221469e-06,
+      "loss": 0.41,
+      "step": 894
+    },
+    {
+      "epoch": 0.4162145403813362,
+      "grad_norm": 0.5378727316856384,
+      "learning_rate": 9.954306677969218e-06,
+      "loss": 0.3946,
+      "step": 895
+    },
+    {
+      "epoch": 0.4166795845605333,
+      "grad_norm": 0.5027511715888977,
+      "learning_rate": 9.953940963681015e-06,
+      "loss": 0.3757,
+      "step": 896
+    },
+    {
+      "epoch": 0.4171446287397303,
+      "grad_norm": 0.5424126386642456,
+      "learning_rate": 9.953573798463972e-06,
+      "loss": 0.417,
+      "step": 897
+    },
+    {
+      "epoch": 0.4176096729189273,
+      "grad_norm": 0.5236470103263855,
+      "learning_rate": 9.953205182425623e-06,
+      "loss": 0.415,
+      "step": 898
+    },
+    {
+      "epoch": 0.4180747170981243,
+      "grad_norm": 0.5600041747093201,
+      "learning_rate": 9.952835115673933e-06,
+      "loss": 0.4227,
+      "step": 899
+    },
+    {
+      "epoch": 0.41853976127732134,
+      "grad_norm": 0.44945916533470154,
+      "learning_rate": 9.952463598317286e-06,
+      "loss": 0.405,
+      "step": 900
+    },
+    {
+      "epoch": 0.41900480545651836,
+      "grad_norm": 0.5260853171348572,
+      "learning_rate": 9.952090630464495e-06,
+      "loss": 0.4051,
+      "step": 901
+    },
+    {
+      "epoch": 0.4194698496357154,
+      "grad_norm": 0.5023328065872192,
+      "learning_rate": 9.951716212224798e-06,
+      "loss": 0.408,
+      "step": 902
+    },
+    {
+      "epoch": 0.4199348938149124,
+      "grad_norm": 0.5364269018173218,
+      "learning_rate": 9.951340343707852e-06,
+      "loss": 0.4036,
+      "step": 903
+    },
+    {
+      "epoch": 0.42039993799410946,
+      "grad_norm": 0.5100162625312805,
+      "learning_rate": 9.950963025023746e-06,
+      "loss": 0.4104,
+      "step": 904
+    },
+    {
+      "epoch": 0.4208649821733065,
+      "grad_norm": 0.5659197568893433,
+      "learning_rate": 9.950584256282988e-06,
+      "loss": 0.4216,
+      "step": 905
+    },
+    {
+      "epoch": 0.4213300263525035,
+      "grad_norm": 0.5025163292884827,
+      "learning_rate": 9.950204037596516e-06,
+      "loss": 0.3989,
+      "step": 906
+    },
+    {
+      "epoch": 0.4217950705317005,
+      "grad_norm": 0.6001001596450806,
+      "learning_rate": 9.949822369075687e-06,
+      "loss": 0.389,
+      "step": 907
+    },
+    {
+      "epoch": 0.4222601147108975,
+      "grad_norm": 0.5134057998657227,
+      "learning_rate": 9.949439250832287e-06,
+      "loss": 0.3975,
+      "step": 908
+    },
+    {
+      "epoch": 0.42272515889009454,
+      "grad_norm": 0.5745139122009277,
+      "learning_rate": 9.949054682978525e-06,
+      "loss": 0.4186,
+      "step": 909
+    },
+    {
+      "epoch": 0.42319020306929156,
+      "grad_norm": 0.6479323506355286,
+      "learning_rate": 9.948668665627034e-06,
+      "loss": 0.4201,
+      "step": 910
+    },
+    {
+      "epoch": 0.42365524724848863,
+      "grad_norm": 0.5262349843978882,
+      "learning_rate": 9.948281198890875e-06,
+      "loss": 0.4016,
+      "step": 911
+    },
+    {
+      "epoch": 0.42412029142768565,
+      "grad_norm": 0.6013730764389038,
+      "learning_rate": 9.947892282883527e-06,
+      "loss": 0.3814,
+      "step": 912
+    },
+    {
+      "epoch": 0.42458533560688266,
+      "grad_norm": 0.6294665336608887,
+      "learning_rate": 9.947501917718897e-06,
+      "loss": 0.391,
+      "step": 913
+    },
+    {
+      "epoch": 0.4250503797860797,
+      "grad_norm": 0.5530297756195068,
+      "learning_rate": 9.947110103511322e-06,
+      "loss": 0.3957,
+      "step": 914
+    },
+    {
+      "epoch": 0.4255154239652767,
+      "grad_norm": 0.5969755053520203,
+      "learning_rate": 9.946716840375552e-06,
+      "loss": 0.3933,
+      "step": 915
+    },
+    {
+      "epoch": 0.4259804681444737,
+      "grad_norm": 0.6482888460159302,
+      "learning_rate": 9.946322128426771e-06,
+      "loss": 0.4384,
+      "step": 916
+    },
+    {
+      "epoch": 0.42644551232367073,
+      "grad_norm": 0.6915735602378845,
+      "learning_rate": 9.945925967780581e-06,
+      "loss": 0.4165,
+      "step": 917
+    },
+    {
+      "epoch": 0.4269105565028678,
+      "grad_norm": 0.5215064883232117,
+      "learning_rate": 9.945528358553014e-06,
+      "loss": 0.4123,
+      "step": 918
+    },
+    {
+      "epoch": 0.4273756006820648,
+      "grad_norm": 0.6709214448928833,
+      "learning_rate": 9.945129300860521e-06,
+      "loss": 0.4214,
+      "step": 919
+    },
+    {
+      "epoch": 0.42784064486126183,
+      "grad_norm": 0.5739949345588684,
+      "learning_rate": 9.94472879481998e-06,
+      "loss": 0.3663,
+      "step": 920
+    },
+    {
+      "epoch": 0.42830568904045885,
+      "grad_norm": 0.5512825846672058,
+      "learning_rate": 9.944326840548693e-06,
+      "loss": 0.386,
+      "step": 921
+    },
+    {
+      "epoch": 0.42877073321965586,
+      "grad_norm": 0.6003051400184631,
+      "learning_rate": 9.943923438164387e-06,
+      "loss": 0.4002,
+      "step": 922
+    },
+    {
+      "epoch": 0.4292357773988529,
+      "grad_norm": 0.5427140593528748,
+      "learning_rate": 9.943518587785208e-06,
+      "loss": 0.3789,
+      "step": 923
+    },
+    {
+      "epoch": 0.4297008215780499,
+      "grad_norm": 0.4838401675224304,
+      "learning_rate": 9.943112289529737e-06,
+      "loss": 0.4048,
+      "step": 924
+    },
+    {
+      "epoch": 0.4301658657572469,
+      "grad_norm": 0.487448126077652,
+      "learning_rate": 9.942704543516966e-06,
+      "loss": 0.4082,
+      "step": 925
+    },
+    {
+      "epoch": 0.430630909936444,
+      "grad_norm": 0.5517868995666504,
+      "learning_rate": 9.94229534986632e-06,
+      "loss": 0.415,
+      "step": 926
+    },
+    {
+      "epoch": 0.431095954115641,
+      "grad_norm": 0.5240123867988586,
+      "learning_rate": 9.941884708697644e-06,
+      "loss": 0.3909,
+      "step": 927
+    },
+    {
+      "epoch": 0.431560998294838,
+      "grad_norm": 0.5013271570205688,
+      "learning_rate": 9.941472620131208e-06,
+      "loss": 0.3914,
+      "step": 928
+    },
+    {
+      "epoch": 0.43202604247403503,
+      "grad_norm": 0.49305182695388794,
+      "learning_rate": 9.941059084287708e-06,
+      "loss": 0.3725,
+      "step": 929
+    },
+    {
+      "epoch": 0.43249108665323205,
+      "grad_norm": 0.5340526103973389,
+      "learning_rate": 9.940644101288259e-06,
+      "loss": 0.3961,
+      "step": 930
+    },
+    {
+      "epoch": 0.43295613083242906,
+      "grad_norm": 0.5031039118766785,
+      "learning_rate": 9.940227671254406e-06,
+      "loss": 0.4108,
+      "step": 931
+    },
+    {
+      "epoch": 0.4334211750116261,
+      "grad_norm": 0.5576218962669373,
+      "learning_rate": 9.939809794308111e-06,
+      "loss": 0.3985,
+      "step": 932
+    },
+    {
+      "epoch": 0.43388621919082315,
+      "grad_norm": 0.48340535163879395,
+      "learning_rate": 9.939390470571769e-06,
+      "loss": 0.4029,
+      "step": 933
+    },
+    {
+      "epoch": 0.43435126337002017,
+      "grad_norm": 0.5440919399261475,
+      "learning_rate": 9.938969700168186e-06,
+      "loss": 0.3998,
+      "step": 934
+    },
+    {
+      "epoch": 0.4348163075492172,
+      "grad_norm": 0.44848132133483887,
+      "learning_rate": 9.938547483220602e-06,
+      "loss": 0.3841,
+      "step": 935
+    },
+    {
+      "epoch": 0.4352813517284142,
+      "grad_norm": 0.5585298538208008,
+      "learning_rate": 9.93812381985268e-06,
+      "loss": 0.4138,
+      "step": 936
+    },
+    {
+      "epoch": 0.4357463959076112,
+      "grad_norm": 0.5106562376022339,
+      "learning_rate": 9.9376987101885e-06,
+      "loss": 0.3974,
+      "step": 937
+    },
+    {
+      "epoch": 0.43621144008680823,
+      "grad_norm": 0.5067592263221741,
+      "learning_rate": 9.937272154352573e-06,
+      "loss": 0.3889,
+      "step": 938
+    },
+    {
+      "epoch": 0.43667648426600525,
+      "grad_norm": 0.5147528052330017,
+      "learning_rate": 9.936844152469828e-06,
+      "loss": 0.3935,
+      "step": 939
+    },
+    {
+      "epoch": 0.4371415284452023,
+      "grad_norm": 0.48401129245758057,
+      "learning_rate": 9.936414704665622e-06,
+      "loss": 0.3967,
+      "step": 940
+    },
+    {
+      "epoch": 0.43760657262439934,
+      "grad_norm": 0.6152265071868896,
+      "learning_rate": 9.935983811065732e-06,
+      "loss": 0.4419,
+      "step": 941
+    },
+    {
+      "epoch": 0.43807161680359635,
+      "grad_norm": 0.4920692443847656,
+      "learning_rate": 9.935551471796358e-06,
+      "loss": 0.3831,
+      "step": 942
+    },
+    {
+      "epoch": 0.43853666098279337,
+      "grad_norm": 0.4826490581035614,
+      "learning_rate": 9.935117686984128e-06,
+      "loss": 0.3911,
+      "step": 943
+    },
+    {
+      "epoch": 0.4390017051619904,
+      "grad_norm": 0.5583831667900085,
+      "learning_rate": 9.93468245675609e-06,
+      "loss": 0.4013,
+      "step": 944
+    },
+    {
+      "epoch": 0.4394667493411874,
+      "grad_norm": 0.5026276111602783,
+      "learning_rate": 9.934245781239714e-06,
+      "loss": 0.3954,
+      "step": 945
+    },
+    {
+      "epoch": 0.4399317935203844,
+      "grad_norm": 0.6956951022148132,
+      "learning_rate": 9.933807660562898e-06,
+      "loss": 0.4083,
+      "step": 946
+    },
+    {
+      "epoch": 0.44039683769958143,
+      "grad_norm": 0.43628400564193726,
+      "learning_rate": 9.933368094853958e-06,
+      "loss": 0.3529,
+      "step": 947
+    },
+    {
+      "epoch": 0.4408618818787785,
+      "grad_norm": 0.5511034727096558,
+      "learning_rate": 9.932927084241635e-06,
+      "loss": 0.4084,
+      "step": 948
+    },
+    {
+      "epoch": 0.4413269260579755,
+      "grad_norm": 0.5073305368423462,
+      "learning_rate": 9.932484628855097e-06,
+      "loss": 0.3879,
+      "step": 949
+    },
+    {
+      "epoch": 0.44179197023717254,
+      "grad_norm": 0.5978685617446899,
+      "learning_rate": 9.932040728823928e-06,
+      "loss": 0.4192,
+      "step": 950
+    },
+    {
+      "epoch": 0.44225701441636955,
+      "grad_norm": 0.5503034591674805,
+      "learning_rate": 9.931595384278143e-06,
+      "loss": 0.4073,
+      "step": 951
+    },
+    {
+      "epoch": 0.44272205859556657,
+      "grad_norm": 0.5104547739028931,
+      "learning_rate": 9.931148595348176e-06,
+      "loss": 0.3689,
+      "step": 952
+    },
+    {
+      "epoch": 0.4431871027747636,
+      "grad_norm": 0.594742476940155,
+      "learning_rate": 9.93070036216488e-06,
+      "loss": 0.418,
+      "step": 953
+    },
+    {
+      "epoch": 0.4436521469539606,
+      "grad_norm": 0.4962625205516815,
+      "learning_rate": 9.930250684859542e-06,
+      "loss": 0.4008,
+      "step": 954
+    },
+    {
+      "epoch": 0.4441171911331577,
+      "grad_norm": 0.5209368467330933,
+      "learning_rate": 9.929799563563858e-06,
+      "loss": 0.3777,
+      "step": 955
+    },
+    {
+      "epoch": 0.4445822353123547,
+      "grad_norm": 0.5389686226844788,
+      "learning_rate": 9.929346998409958e-06,
+      "loss": 0.4047,
+      "step": 956
+    },
+    {
+      "epoch": 0.4450472794915517,
+      "grad_norm": 0.5484662652015686,
+      "learning_rate": 9.92889298953039e-06,
+      "loss": 0.42,
+      "step": 957
+    },
+    {
+      "epoch": 0.4455123236707487,
+      "grad_norm": 0.5775797963142395,
+      "learning_rate": 9.928437537058126e-06,
+      "loss": 0.4016,
+      "step": 958
+    },
+    {
+      "epoch": 0.44597736784994574,
+      "grad_norm": 0.48525333404541016,
+      "learning_rate": 9.927980641126562e-06,
+      "loss": 0.385,
+      "step": 959
+    },
+    {
+      "epoch": 0.44644241202914275,
+      "grad_norm": 0.6398960947990417,
+      "learning_rate": 9.927522301869515e-06,
+      "loss": 0.4008,
+      "step": 960
+    },
+    {
+      "epoch": 0.44690745620833977,
+      "grad_norm": 0.5224960446357727,
+      "learning_rate": 9.927062519421223e-06,
+      "loss": 0.4061,
+      "step": 961
+    },
+    {
+      "epoch": 0.44737250038753684,
+      "grad_norm": 0.5396710634231567,
+      "learning_rate": 9.926601293916349e-06,
+      "loss": 0.4128,
+      "step": 962
+    },
+    {
+      "epoch": 0.44783754456673386,
+      "grad_norm": 0.540110170841217,
+      "learning_rate": 9.926138625489981e-06,
+      "loss": 0.4037,
+      "step": 963
+    },
+    {
+      "epoch": 0.4483025887459309,
+      "grad_norm": 0.5284585952758789,
+      "learning_rate": 9.925674514277625e-06,
+      "loss": 0.3929,
+      "step": 964
+    },
+    {
+      "epoch": 0.4487676329251279,
+      "grad_norm": 0.5301303863525391,
+      "learning_rate": 9.925208960415214e-06,
+      "loss": 0.4064,
+      "step": 965
+    },
+    {
+      "epoch": 0.4492326771043249,
+      "grad_norm": 0.5405663847923279,
+      "learning_rate": 9.924741964039098e-06,
+      "loss": 0.3813,
+      "step": 966
+    },
+    {
+      "epoch": 0.4496977212835219,
+      "grad_norm": 0.567683756351471,
+      "learning_rate": 9.924273525286053e-06,
+      "loss": 0.3801,
+      "step": 967
+    },
+    {
+      "epoch": 0.45016276546271894,
+      "grad_norm": 0.5406811237335205,
+      "learning_rate": 9.92380364429328e-06,
+      "loss": 0.394,
+      "step": 968
+    },
+    {
+      "epoch": 0.45062780964191596,
+      "grad_norm": 0.6923626065254211,
+      "learning_rate": 9.923332321198396e-06,
+      "loss": 0.389,
+      "step": 969
+    },
+    {
+      "epoch": 0.451092853821113,
+      "grad_norm": 0.5620383620262146,
+      "learning_rate": 9.922859556139447e-06,
+      "loss": 0.393,
+      "step": 970
+    },
+    {
+      "epoch": 0.45155789800031004,
+      "grad_norm": 0.5418940782546997,
+      "learning_rate": 9.922385349254895e-06,
+      "loss": 0.4071,
+      "step": 971
+    },
+    {
+      "epoch": 0.45202294217950706,
+      "grad_norm": 0.5649824738502502,
+      "learning_rate": 9.921909700683632e-06,
+      "loss": 0.3882,
+      "step": 972
+    },
+    {
+      "epoch": 0.4524879863587041,
+      "grad_norm": 0.5053907036781311,
+      "learning_rate": 9.921432610564962e-06,
+      "loss": 0.3893,
+      "step": 973
+    },
+    {
+      "epoch": 0.4529530305379011,
+      "grad_norm": 0.5486416816711426,
+      "learning_rate": 9.920954079038623e-06,
+      "loss": 0.4207,
+      "step": 974
+    },
+    {
+      "epoch": 0.4534180747170981,
+      "grad_norm": 0.4447232186794281,
+      "learning_rate": 9.920474106244764e-06,
+      "loss": 0.3837,
+      "step": 975
+    },
+    {
+      "epoch": 0.4538831188962951,
+      "grad_norm": 0.610753059387207,
+      "learning_rate": 9.919992692323965e-06,
+      "loss": 0.3905,
+      "step": 976
+    },
+    {
+      "epoch": 0.4543481630754922,
+      "grad_norm": 0.5709927678108215,
+      "learning_rate": 9.919509837417221e-06,
+      "loss": 0.3985,
+      "step": 977
+    },
+    {
+      "epoch": 0.4548132072546892,
+      "grad_norm": 0.6064723134040833,
+      "learning_rate": 9.919025541665955e-06,
+      "loss": 0.4376,
+      "step": 978
+    },
+    {
+      "epoch": 0.4552782514338862,
+      "grad_norm": 0.6571378111839294,
+      "learning_rate": 9.918539805212008e-06,
+      "loss": 0.3945,
+      "step": 979
+    },
+    {
+      "epoch": 0.45574329561308324,
+      "grad_norm": 0.528195321559906,
+      "learning_rate": 9.918052628197645e-06,
+      "loss": 0.406,
+      "step": 980
+    },
+    {
+      "epoch": 0.45620833979228026,
+      "grad_norm": 0.6316954493522644,
+      "learning_rate": 9.917564010765551e-06,
+      "loss": 0.3779,
+      "step": 981
+    },
+    {
+      "epoch": 0.4566733839714773,
+      "grad_norm": 0.5615218281745911,
+      "learning_rate": 9.917073953058836e-06,
+      "loss": 0.4308,
+      "step": 982
+    },
+    {
+      "epoch": 0.4571384281506743,
+      "grad_norm": 0.5495420098304749,
+      "learning_rate": 9.916582455221029e-06,
+      "loss": 0.4006,
+      "step": 983
+    },
+    {
+      "epoch": 0.45760347232987136,
+      "grad_norm": 0.5130931735038757,
+      "learning_rate": 9.916089517396081e-06,
+      "loss": 0.3881,
+      "step": 984
+    },
+    {
+      "epoch": 0.4580685165090684,
+      "grad_norm": 0.6450378894805908,
+      "learning_rate": 9.915595139728366e-06,
+      "loss": 0.4132,
+      "step": 985
+    },
+    {
+      "epoch": 0.4585335606882654,
+      "grad_norm": 0.5673415064811707,
+      "learning_rate": 9.915099322362681e-06,
+      "loss": 0.4007,
+      "step": 986
+    },
+    {
+      "epoch": 0.4589986048674624,
+      "grad_norm": 0.5420086979866028,
+      "learning_rate": 9.91460206544424e-06,
+      "loss": 0.4066,
+      "step": 987
+    },
+    {
+      "epoch": 0.45946364904665943,
+      "grad_norm": 0.5733571648597717,
+      "learning_rate": 9.914103369118682e-06,
+      "loss": 0.371,
+      "step": 988
+    },
+    {
+      "epoch": 0.45992869322585644,
+      "grad_norm": 0.5512123107910156,
+      "learning_rate": 9.913603233532067e-06,
+      "loss": 0.3742,
+      "step": 989
+    },
+    {
+      "epoch": 0.46039373740505346,
+      "grad_norm": 0.5796908140182495,
+      "learning_rate": 9.913101658830879e-06,
+      "loss": 0.4115,
+      "step": 990
+    },
+    {
+      "epoch": 0.4608587815842505,
+      "grad_norm": 0.5767021179199219,
+      "learning_rate": 9.91259864516202e-06,
+      "loss": 0.4203,
+      "step": 991
+    },
+    {
+      "epoch": 0.46132382576344755,
+      "grad_norm": 0.5332590937614441,
+      "learning_rate": 9.912094192672812e-06,
+      "loss": 0.3986,
+      "step": 992
+    },
+    {
+      "epoch": 0.46178886994264456,
+      "grad_norm": 0.4922131597995758,
+      "learning_rate": 9.911588301511004e-06,
+      "loss": 0.4085,
+      "step": 993
+    },
+    {
+      "epoch": 0.4622539141218416,
+      "grad_norm": 0.5280721783638,
+      "learning_rate": 9.911080971824762e-06,
+      "loss": 0.4059,
+      "step": 994
+    },
+    {
+      "epoch": 0.4627189583010386,
+      "grad_norm": 0.5713456869125366,
+      "learning_rate": 9.910572203762676e-06,
+      "loss": 0.4249,
+      "step": 995
+    },
+    {
+      "epoch": 0.4631840024802356,
+      "grad_norm": 0.5211708545684814,
+      "learning_rate": 9.910061997473753e-06,
+      "loss": 0.3874,
+      "step": 996
+    },
+    {
+      "epoch": 0.46364904665943263,
+      "grad_norm": 0.5675033330917358,
+      "learning_rate": 9.909550353107426e-06,
+      "loss": 0.4086,
+      "step": 997
+    },
+    {
+      "epoch": 0.46411409083862964,
+      "grad_norm": 0.6571573615074158,
+      "learning_rate": 9.909037270813547e-06,
+      "loss": 0.4148,
+      "step": 998
+    },
+    {
+      "epoch": 0.4645791350178267,
+      "grad_norm": 0.6051549911499023,
+      "learning_rate": 9.908522750742391e-06,
+      "loss": 0.4061,
+      "step": 999
+    },
+    {
+      "epoch": 0.46504417919702373,
+      "grad_norm": 0.5821381211280823,
+      "learning_rate": 9.90800679304465e-06,
+      "loss": 0.4033,
+      "step": 1000
+    },
+    {
+      "epoch": 0.46550922337622075,
+      "grad_norm": 0.6213275194168091,
+      "learning_rate": 9.907489397871441e-06,
+      "loss": 0.4018,
+      "step": 1001
+    },
+    {
+      "epoch": 0.46597426755541776,
+      "grad_norm": 0.5967691540718079,
+      "learning_rate": 9.9069705653743e-06,
+      "loss": 0.376,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4664393117346148,
+      "grad_norm": 0.4699487090110779,
+      "learning_rate": 9.906450295705188e-06,
+      "loss": 0.4172,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4669043559138118,
+      "grad_norm": 0.6863399147987366,
+      "learning_rate": 9.905928589016479e-06,
+      "loss": 0.4105,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4673694000930088,
+      "grad_norm": 0.5886141657829285,
+      "learning_rate": 9.905405445460972e-06,
+      "loss": 0.3957,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4678344442722059,
+      "grad_norm": 0.4847775399684906,
+      "learning_rate": 9.90488086519189e-06,
+      "loss": 0.4155,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4682994884514029,
+      "grad_norm": 0.6540789604187012,
+      "learning_rate": 9.904354848362876e-06,
+      "loss": 0.4178,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4687645326305999,
+      "grad_norm": 0.5532976388931274,
+      "learning_rate": 9.903827395127987e-06,
+      "loss": 0.3777,
+      "step": 1008
+    },
+    {
+      "epoch": 0.46922957680979693,
+      "grad_norm": 0.44451895356178284,
+      "learning_rate": 9.903298505641707e-06,
+      "loss": 0.4003,
+      "step": 1009
+    },
+    {
+      "epoch": 0.46969462098899395,
+      "grad_norm": 0.5525593757629395,
+      "learning_rate": 9.902768180058942e-06,
+      "loss": 0.3992,
+      "step": 1010
+    },
+    {
+      "epoch": 0.47015966516819097,
+      "grad_norm": 0.480257123708725,
+      "learning_rate": 9.902236418535012e-06,
+      "loss": 0.3906,
+      "step": 1011
+    },
+    {
+      "epoch": 0.470624709347388,
+      "grad_norm": 0.5119292140007019,
+      "learning_rate": 9.901703221225663e-06,
+      "loss": 0.42,
+      "step": 1012
+    },
+    {
+      "epoch": 0.47108975352658505,
+      "grad_norm": 0.500453531742096,
+      "learning_rate": 9.901168588287057e-06,
+      "loss": 0.4131,
+      "step": 1013
+    },
+    {
+      "epoch": 0.47155479770578207,
+      "grad_norm": 0.5501608848571777,
+      "learning_rate": 9.900632519875786e-06,
+      "loss": 0.3866,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4720198418849791,
+      "grad_norm": 0.4848511219024658,
+      "learning_rate": 9.900095016148849e-06,
+      "loss": 0.4138,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4724848860641761,
+      "grad_norm": 0.5268184542655945,
+      "learning_rate": 9.899556077263676e-06,
+      "loss": 0.3947,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4729499302433731,
+      "grad_norm": 0.555121660232544,
+      "learning_rate": 9.899015703378115e-06,
+      "loss": 0.4129,
+      "step": 1017
+    },
+    {
+      "epoch": 0.47341497442257013,
+      "grad_norm": 0.5795762538909912,
+      "learning_rate": 9.898473894650425e-06,
+      "loss": 0.4121,
+      "step": 1018
+    },
+    {
+      "epoch": 0.47388001860176715,
+      "grad_norm": 0.515906810760498,
+      "learning_rate": 9.8979306512393e-06,
+      "loss": 0.3946,
+      "step": 1019
+    },
+    {
+      "epoch": 0.47434506278096417,
+      "grad_norm": 0.5714851021766663,
+      "learning_rate": 9.897385973303845e-06,
+      "loss": 0.3858,
+      "step": 1020
+    },
+    {
+      "epoch": 0.47481010696016124,
+      "grad_norm": 0.6117233037948608,
+      "learning_rate": 9.896839861003588e-06,
+      "loss": 0.411,
+      "step": 1021
+    },
+    {
+      "epoch": 0.47527515113935825,
+      "grad_norm": 0.5009344220161438,
+      "learning_rate": 9.896292314498475e-06,
+      "loss": 0.3939,
+      "step": 1022
+    },
+    {
+      "epoch": 0.47574019531855527,
+      "grad_norm": 0.46868664026260376,
+      "learning_rate": 9.895743333948875e-06,
+      "loss": 0.412,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4762052394977523,
+      "grad_norm": 0.6705508828163147,
+      "learning_rate": 9.895192919515575e-06,
+      "loss": 0.3981,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4766702836769493,
+      "grad_norm": 0.535101056098938,
+      "learning_rate": 9.894641071359784e-06,
+      "loss": 0.4057,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4771353278561463,
+      "grad_norm": 0.4733811616897583,
+      "learning_rate": 9.894087789643123e-06,
+      "loss": 0.4066,
+      "step": 1026
+    },
+    {
+      "epoch": 0.47760037203534333,
+      "grad_norm": 0.514105498790741,
+      "learning_rate": 9.893533074527647e-06,
+      "loss": 0.3957,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4780654162145404,
+      "grad_norm": 0.50275719165802,
+      "learning_rate": 9.892976926175819e-06,
+      "loss": 0.386,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4785304603937374,
+      "grad_norm": 0.4467027187347412,
+      "learning_rate": 9.892419344750528e-06,
+      "loss": 0.3965,
+      "step": 1029
+    },
+    {
+      "epoch": 0.47899550457293444,
+      "grad_norm": 0.48868438601493835,
+      "learning_rate": 9.891860330415077e-06,
+      "loss": 0.408,
+      "step": 1030
+    },
+    {
+      "epoch": 0.47946054875213145,
+      "grad_norm": 0.4734288156032562,
+      "learning_rate": 9.891299883333197e-06,
+      "loss": 0.3778,
+      "step": 1031
+    },
+    {
+      "epoch": 0.47992559293132847,
+      "grad_norm": 0.49330592155456543,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.3766,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4803906371105255,
+      "grad_norm": 0.5068243145942688,
+      "learning_rate": 9.890174691587142e-06,
+      "loss": 0.3832,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4808556812897225,
+      "grad_norm": 0.4672917127609253,
+      "learning_rate": 9.889609947252519e-06,
+      "loss": 0.3933,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4813207254689196,
+      "grad_norm": 0.5433964729309082,
+      "learning_rate": 9.889043770830566e-06,
+      "loss": 0.378,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4817857696481166,
+      "grad_norm": 0.4645717740058899,
+      "learning_rate": 9.888476162487106e-06,
+      "loss": 0.3803,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4822508138273136,
+      "grad_norm": 0.5179428458213806,
+      "learning_rate": 9.887907122388382e-06,
+      "loss": 0.4093,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4827158580065106,
+      "grad_norm": 0.5166782140731812,
+      "learning_rate": 9.887336650701055e-06,
+      "loss": 0.4107,
+      "step": 1038
+    },
+    {
+      "epoch": 0.48318090218570764,
+      "grad_norm": 0.4443417489528656,
+      "learning_rate": 9.886764747592212e-06,
+      "loss": 0.3626,
+      "step": 1039
+    },
+    {
+      "epoch": 0.48364594636490466,
+      "grad_norm": 0.5102760195732117,
+      "learning_rate": 9.886191413229349e-06,
+      "loss": 0.4014,
+      "step": 1040
+    },
+    {
+      "epoch": 0.48411099054410167,
+      "grad_norm": 0.5018773674964905,
+      "learning_rate": 9.885616647780389e-06,
+      "loss": 0.3895,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4845760347232987,
+      "grad_norm": 0.5214076638221741,
+      "learning_rate": 9.88504045141367e-06,
+      "loss": 0.389,
+      "step": 1042
+    },
+    {
+      "epoch": 0.48504107890249576,
+      "grad_norm": 0.4620245397090912,
+      "learning_rate": 9.884462824297952e-06,
+      "loss": 0.389,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4855061230816928,
+      "grad_norm": 0.4827862083911896,
+      "learning_rate": 9.88388376660241e-06,
+      "loss": 0.4153,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4859711672608898,
+      "grad_norm": 0.5498597025871277,
+      "learning_rate": 9.883303278496646e-06,
+      "loss": 0.4076,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4864362114400868,
+      "grad_norm": 0.49216824769973755,
+      "learning_rate": 9.88272136015067e-06,
+      "loss": 0.4254,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4869012556192838,
+      "grad_norm": 0.534773588180542,
+      "learning_rate": 9.882138011734919e-06,
+      "loss": 0.3948,
+      "step": 1047
+    },
+    {
+      "epoch": 0.48736629979848084,
+      "grad_norm": 0.5610633492469788,
+      "learning_rate": 9.881553233420244e-06,
+      "loss": 0.41,
+      "step": 1048
+    },
+    {
+      "epoch": 0.48783134397767786,
+      "grad_norm": 0.48247793316841125,
+      "learning_rate": 9.88096702537792e-06,
+      "loss": 0.394,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4882963881568749,
+      "grad_norm": 0.5383715033531189,
+      "learning_rate": 9.880379387779637e-06,
+      "loss": 0.3893,
+      "step": 1050
+    },
+    {
+      "epoch": 0.48876143233607194,
+      "grad_norm": 0.5967577695846558,
+      "learning_rate": 9.879790320797504e-06,
+      "loss": 0.373,
+      "step": 1051
+    },
+    {
+      "epoch": 0.48922647651526896,
+      "grad_norm": 0.4835599660873413,
+      "learning_rate": 9.879199824604048e-06,
+      "loss": 0.4082,
+      "step": 1052
+    },
+    {
+      "epoch": 0.489691520694466,
+      "grad_norm": 0.5845022797584534,
+      "learning_rate": 9.878607899372217e-06,
+      "loss": 0.3954,
+      "step": 1053
+    },
+    {
+      "epoch": 0.490156564873663,
+      "grad_norm": 0.49885934591293335,
+      "learning_rate": 9.878014545275379e-06,
+      "loss": 0.4067,
+      "step": 1054
+    },
+    {
+      "epoch": 0.49062160905286,
+      "grad_norm": 0.5852560997009277,
+      "learning_rate": 9.877419762487312e-06,
+      "loss": 0.3932,
+      "step": 1055
+    },
+    {
+      "epoch": 0.491086653232057,
+      "grad_norm": 0.48509690165519714,
+      "learning_rate": 9.876823551182223e-06,
+      "loss": 0.433,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4915516974112541,
+      "grad_norm": 0.44763508439064026,
+      "learning_rate": 9.876225911534729e-06,
+      "loss": 0.3651,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4920167415904511,
+      "grad_norm": 0.5484299063682556,
+      "learning_rate": 9.875626843719871e-06,
+      "loss": 0.3959,
+      "step": 1058
+    },
+    {
+      "epoch": 0.49248178576964813,
+      "grad_norm": 0.481046199798584,
+      "learning_rate": 9.875026347913109e-06,
+      "loss": 0.3765,
+      "step": 1059
+    },
+    {
+      "epoch": 0.49294682994884514,
+      "grad_norm": 0.5341517925262451,
+      "learning_rate": 9.874424424290313e-06,
+      "loss": 0.3963,
+      "step": 1060
+    },
+    {
+      "epoch": 0.49341187412804216,
+      "grad_norm": 0.469200998544693,
+      "learning_rate": 9.87382107302778e-06,
+      "loss": 0.3995,
+      "step": 1061
+    },
+    {
+      "epoch": 0.4938769183072392,
+      "grad_norm": 0.5132390260696411,
+      "learning_rate": 9.87321629430222e-06,
+      "loss": 0.3837,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4943419624864362,
+      "grad_norm": 0.5703272223472595,
+      "learning_rate": 9.872610088290766e-06,
+      "loss": 0.4025,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4948070066656332,
+      "grad_norm": 0.5358856916427612,
+      "learning_rate": 9.87200245517096e-06,
+      "loss": 0.4287,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4952720508448303,
+      "grad_norm": 0.41558536887168884,
+      "learning_rate": 9.871393395120774e-06,
+      "loss": 0.3621,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4957370950240273,
+      "grad_norm": 0.5119854807853699,
+      "learning_rate": 9.870782908318591e-06,
+      "loss": 0.3781,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4962021392032243,
+      "grad_norm": 0.4697974622249603,
+      "learning_rate": 9.87017099494321e-06,
+      "loss": 0.4138,
+      "step": 1067
+    },
+    {
+      "epoch": 0.49666718338242133,
+      "grad_norm": 0.4632170796394348,
+      "learning_rate": 9.869557655173849e-06,
+      "loss": 0.3952,
+      "step": 1068
+    },
+    {
+      "epoch": 0.49713222756161835,
+      "grad_norm": 0.4312477111816406,
+      "learning_rate": 9.86894288919015e-06,
+      "loss": 0.396,
+      "step": 1069
+    },
+    {
+      "epoch": 0.49759727174081536,
+      "grad_norm": 0.47128409147262573,
+      "learning_rate": 9.868326697172164e-06,
+      "loss": 0.4233,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4980623159200124,
+      "grad_norm": 0.5045310258865356,
+      "learning_rate": 9.867709079300366e-06,
+      "loss": 0.3823,
+      "step": 1071
+    },
+    {
+      "epoch": 0.49852736009920945,
+      "grad_norm": 0.486145943403244,
+      "learning_rate": 9.867090035755648e-06,
+      "loss": 0.3714,
+      "step": 1072
+    },
+    {
+      "epoch": 0.49899240427840647,
+      "grad_norm": 0.4772482216358185,
+      "learning_rate": 9.866469566719314e-06,
+      "loss": 0.3799,
+      "step": 1073
+    },
+    {
+      "epoch": 0.4994574484576035,
+      "grad_norm": 0.546289324760437,
+      "learning_rate": 9.86584767237309e-06,
+      "loss": 0.4263,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4999224926368005,
+      "grad_norm": 0.5823156833648682,
+      "learning_rate": 9.86522435289912e-06,
+      "loss": 0.4101,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5003875368159976,
+      "grad_norm": 0.5479342341423035,
+      "learning_rate": 9.864599608479963e-06,
+      "loss": 0.3717,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5008525809951946,
+      "grad_norm": 0.5319987535476685,
+      "learning_rate": 9.863973439298597e-06,
+      "loss": 0.3993,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5013176251743916,
+      "grad_norm": 0.5460782647132874,
+      "learning_rate": 9.86334584553842e-06,
+      "loss": 0.4069,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5017826693535886,
+      "grad_norm": 0.6070926189422607,
+      "learning_rate": 9.862716827383238e-06,
+      "loss": 0.3999,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5022477135327856,
+      "grad_norm": 0.530796229839325,
+      "learning_rate": 9.862086385017283e-06,
+      "loss": 0.3975,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5027127577119826,
+      "grad_norm": 0.4481984078884125,
+      "learning_rate": 9.861454518625202e-06,
+      "loss": 0.3962,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5031778018911797,
+      "grad_norm": 0.4798201024532318,
+      "learning_rate": 9.86082122839206e-06,
+      "loss": 0.3859,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5036428460703767,
+      "grad_norm": 0.5404430627822876,
+      "learning_rate": 9.86018651450333e-06,
+      "loss": 0.3999,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5041078902495737,
+      "grad_norm": 0.4909595549106598,
+      "learning_rate": 9.85955037714492e-06,
+      "loss": 0.3951,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5045729344287707,
+      "grad_norm": 0.5814231634140015,
+      "learning_rate": 9.858912816503136e-06,
+      "loss": 0.4005,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5050379786079677,
+      "grad_norm": 0.5166906118392944,
+      "learning_rate": 9.858273832764712e-06,
+      "loss": 0.4184,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5055030227871647,
+      "grad_norm": 0.5268685221672058,
+      "learning_rate": 9.8576334261168e-06,
+      "loss": 0.4257,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5059680669663618,
+      "grad_norm": 0.4857880473136902,
+      "learning_rate": 9.856991596746957e-06,
+      "loss": 0.3995,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5064331111455588,
+      "grad_norm": 0.6406244039535522,
+      "learning_rate": 9.85634834484317e-06,
+      "loss": 0.4282,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5068981553247559,
+      "grad_norm": 0.5598058104515076,
+      "learning_rate": 9.855703670593834e-06,
+      "loss": 0.3901,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5073631995039529,
+      "grad_norm": 0.5567111372947693,
+      "learning_rate": 9.855057574187766e-06,
+      "loss": 0.4043,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5078282436831499,
+      "grad_norm": 0.534076452255249,
+      "learning_rate": 9.854410055814195e-06,
+      "loss": 0.4011,
+      "step": 1092
+    },
+    {
+      "epoch": 0.508293287862347,
+      "grad_norm": 0.42775899171829224,
+      "learning_rate": 9.85376111566277e-06,
+      "loss": 0.3986,
+      "step": 1093
+    },
+    {
+      "epoch": 0.508758332041544,
+      "grad_norm": 0.5820823907852173,
+      "learning_rate": 9.853110753923553e-06,
+      "loss": 0.4143,
+      "step": 1094
+    },
+    {
+      "epoch": 0.509223376220741,
+      "grad_norm": 0.46398279070854187,
+      "learning_rate": 9.852458970787027e-06,
+      "loss": 0.3816,
+      "step": 1095
+    },
+    {
+      "epoch": 0.509688420399938,
+      "grad_norm": 0.4692523777484894,
+      "learning_rate": 9.85180576644409e-06,
+      "loss": 0.3855,
+      "step": 1096
+    },
+    {
+      "epoch": 0.510153464579135,
+      "grad_norm": 0.45275288820266724,
+      "learning_rate": 9.851151141086049e-06,
+      "loss": 0.3736,
+      "step": 1097
+    },
+    {
+      "epoch": 0.510618508758332,
+      "grad_norm": 0.6101323962211609,
+      "learning_rate": 9.850495094904639e-06,
+      "loss": 0.3844,
+      "step": 1098
+    },
+    {
+      "epoch": 0.511083552937529,
+      "grad_norm": 0.5054144859313965,
+      "learning_rate": 9.849837628092003e-06,
+      "loss": 0.4187,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5115485971167261,
+      "grad_norm": 0.4948999285697937,
+      "learning_rate": 9.849178740840701e-06,
+      "loss": 0.3826,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5120136412959231,
+      "grad_norm": 0.6334494948387146,
+      "learning_rate": 9.848518433343714e-06,
+      "loss": 0.3986,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5124786854751201,
+      "grad_norm": 0.5066748857498169,
+      "learning_rate": 9.847856705794432e-06,
+      "loss": 0.4185,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5129437296543171,
+      "grad_norm": 0.5187157392501831,
+      "learning_rate": 9.847193558386666e-06,
+      "loss": 0.385,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5134087738335141,
+      "grad_norm": 0.5060462951660156,
+      "learning_rate": 9.846528991314638e-06,
+      "loss": 0.415,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5138738180127113,
+      "grad_norm": 0.418716162443161,
+      "learning_rate": 9.845863004772994e-06,
+      "loss": 0.3735,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5143388621919083,
+      "grad_norm": 0.45739948749542236,
+      "learning_rate": 9.845195598956787e-06,
+      "loss": 0.3532,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5148039063711053,
+      "grad_norm": 0.42475655674934387,
+      "learning_rate": 9.84452677406149e-06,
+      "loss": 0.3871,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5152689505503023,
+      "grad_norm": 0.5018404722213745,
+      "learning_rate": 9.843856530282992e-06,
+      "loss": 0.396,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5157339947294993,
+      "grad_norm": 0.4738996624946594,
+      "learning_rate": 9.843184867817596e-06,
+      "loss": 0.4255,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5161990389086963,
+      "grad_norm": 0.47539880871772766,
+      "learning_rate": 9.842511786862018e-06,
+      "loss": 0.4214,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5166640830878934,
+      "grad_norm": 0.43585407733917236,
+      "learning_rate": 9.841837287613399e-06,
+      "loss": 0.3895,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5171291272670904,
+      "grad_norm": 0.4282119870185852,
+      "learning_rate": 9.841161370269284e-06,
+      "loss": 0.3917,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5175941714462874,
+      "grad_norm": 0.42898568511009216,
+      "learning_rate": 9.84048403502764e-06,
+      "loss": 0.3847,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5180592156254844,
+      "grad_norm": 0.5623050928115845,
+      "learning_rate": 9.839805282086844e-06,
+      "loss": 0.415,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5185242598046814,
+      "grad_norm": 0.4873126447200775,
+      "learning_rate": 9.839125111645699e-06,
+      "loss": 0.4168,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5189893039838784,
+      "grad_norm": 0.4611026644706726,
+      "learning_rate": 9.83844352390341e-06,
+      "loss": 0.3892,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5194543481630755,
+      "grad_norm": 0.5437391996383667,
+      "learning_rate": 9.837760519059603e-06,
+      "loss": 0.4169,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5199193923422725,
+      "grad_norm": 0.47111251950263977,
+      "learning_rate": 9.83707609731432e-06,
+      "loss": 0.398,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5203844365214695,
+      "grad_norm": 0.4614557921886444,
+      "learning_rate": 9.83639025886802e-06,
+      "loss": 0.3906,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5208494807006666,
+      "grad_norm": 0.5393845438957214,
+      "learning_rate": 9.835703003921569e-06,
+      "loss": 0.3886,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5213145248798636,
+      "grad_norm": 0.4539320468902588,
+      "learning_rate": 9.835014332676256e-06,
+      "loss": 0.4078,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5217795690590606,
+      "grad_norm": 0.5154919028282166,
+      "learning_rate": 9.834324245333782e-06,
+      "loss": 0.3883,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5222446132382577,
+      "grad_norm": 0.5002031326293945,
+      "learning_rate": 9.833632742096259e-06,
+      "loss": 0.3834,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5227096574174547,
+      "grad_norm": 0.551112949848175,
+      "learning_rate": 9.83293982316622e-06,
+      "loss": 0.4026,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5231747015966517,
+      "grad_norm": 0.5185533761978149,
+      "learning_rate": 9.832245488746612e-06,
+      "loss": 0.3856,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5236397457758487,
+      "grad_norm": 0.4748762249946594,
+      "learning_rate": 9.831549739040788e-06,
+      "loss": 0.3975,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5241047899550457,
+      "grad_norm": 0.5362114906311035,
+      "learning_rate": 9.830852574252525e-06,
+      "loss": 0.4028,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5245698341342427,
+      "grad_norm": 0.47415444254875183,
+      "learning_rate": 9.830153994586013e-06,
+      "loss": 0.3969,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5250348783134398,
+      "grad_norm": 0.6344930529594421,
+      "learning_rate": 9.829454000245854e-06,
+      "loss": 0.4292,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5254999224926368,
+      "grad_norm": 0.48250138759613037,
+      "learning_rate": 9.82875259143706e-06,
+      "loss": 0.3995,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5259649666718338,
+      "grad_norm": 0.5570687055587769,
+      "learning_rate": 9.82804976836507e-06,
+      "loss": 0.3855,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5264300108510308,
+      "grad_norm": 0.554617166519165,
+      "learning_rate": 9.827345531235722e-06,
+      "loss": 0.4045,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5268950550302278,
+      "grad_norm": 0.544276773929596,
+      "learning_rate": 9.826639880255282e-06,
+      "loss": 0.4065,
+      "step": 1133
+    },
+    {
+      "epoch": 0.527360099209425,
+      "grad_norm": 0.5573837161064148,
+      "learning_rate": 9.825932815630418e-06,
+      "loss": 0.4231,
+      "step": 1134
+    },
+    {
+      "epoch": 0.527825143388622,
+      "grad_norm": 0.4598175883293152,
+      "learning_rate": 9.825224337568224e-06,
+      "loss": 0.4033,
+      "step": 1135
+    },
+    {
+      "epoch": 0.528290187567819,
+      "grad_norm": 0.5561431050300598,
+      "learning_rate": 9.824514446276197e-06,
+      "loss": 0.3891,
+      "step": 1136
+    },
+    {
+      "epoch": 0.528755231747016,
+      "grad_norm": 0.46380776166915894,
+      "learning_rate": 9.823803141962253e-06,
+      "loss": 0.3938,
+      "step": 1137
+    },
+    {
+      "epoch": 0.529220275926213,
+      "grad_norm": 0.5651960968971252,
+      "learning_rate": 9.823090424834725e-06,
+      "loss": 0.3797,
+      "step": 1138
+    },
+    {
+      "epoch": 0.52968532010541,
+      "grad_norm": 0.5517980456352234,
+      "learning_rate": 9.822376295102352e-06,
+      "loss": 0.3812,
+      "step": 1139
+    },
+    {
+      "epoch": 0.530150364284607,
+      "grad_norm": 0.5638952851295471,
+      "learning_rate": 9.821660752974294e-06,
+      "loss": 0.3835,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5306154084638041,
+      "grad_norm": 0.5966677665710449,
+      "learning_rate": 9.82094379866012e-06,
+      "loss": 0.4018,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5310804526430011,
+      "grad_norm": 0.5064060091972351,
+      "learning_rate": 9.820225432369814e-06,
+      "loss": 0.3956,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5315454968221981,
+      "grad_norm": 0.5523810386657715,
+      "learning_rate": 9.819505654313775e-06,
+      "loss": 0.3872,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5320105410013951,
+      "grad_norm": 0.5287060141563416,
+      "learning_rate": 9.818784464702813e-06,
+      "loss": 0.4018,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5324755851805921,
+      "grad_norm": 0.5854330062866211,
+      "learning_rate": 9.818061863748153e-06,
+      "loss": 0.4165,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5329406293597891,
+      "grad_norm": 0.4986901581287384,
+      "learning_rate": 9.817337851661436e-06,
+      "loss": 0.3944,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5334056735389862,
+      "grad_norm": 0.51901775598526,
+      "learning_rate": 9.81661242865471e-06,
+      "loss": 0.4012,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5338707177181832,
+      "grad_norm": 0.4892278015613556,
+      "learning_rate": 9.815885594940442e-06,
+      "loss": 0.3884,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5343357618973803,
+      "grad_norm": 0.4971807599067688,
+      "learning_rate": 9.815157350731506e-06,
+      "loss": 0.3971,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5348008060765773,
+      "grad_norm": 0.49850407242774963,
+      "learning_rate": 9.814427696241197e-06,
+      "loss": 0.3975,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5352658502557743,
+      "grad_norm": 0.49632036685943604,
+      "learning_rate": 9.813696631683216e-06,
+      "loss": 0.3923,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5357308944349713,
+      "grad_norm": 0.5661031007766724,
+      "learning_rate": 9.812964157271683e-06,
+      "loss": 0.376,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5361959386141684,
+      "grad_norm": 0.48331817984580994,
+      "learning_rate": 9.812230273221124e-06,
+      "loss": 0.41,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5366609827933654,
+      "grad_norm": 0.4824008047580719,
+      "learning_rate": 9.811494979746486e-06,
+      "loss": 0.397,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5371260269725624,
+      "grad_norm": 0.47880852222442627,
+      "learning_rate": 9.81075827706312e-06,
+      "loss": 0.3718,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5375910711517594,
+      "grad_norm": 0.4963204562664032,
+      "learning_rate": 9.810020165386797e-06,
+      "loss": 0.4233,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5380561153309564,
+      "grad_norm": 0.5019689202308655,
+      "learning_rate": 9.809280644933698e-06,
+      "loss": 0.4058,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5385211595101534,
+      "grad_norm": 0.5196408033370972,
+      "learning_rate": 9.808539715920415e-06,
+      "loss": 0.4112,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5389862036893505,
+      "grad_norm": 0.48643895983695984,
+      "learning_rate": 9.807797378563957e-06,
+      "loss": 0.397,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5394512478685475,
+      "grad_norm": 0.49545997381210327,
+      "learning_rate": 9.80705363308174e-06,
+      "loss": 0.3896,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5399162920477445,
+      "grad_norm": 0.5133934617042542,
+      "learning_rate": 9.806308479691595e-06,
+      "loss": 0.3979,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5403813362269415,
+      "grad_norm": 0.47291284799575806,
+      "learning_rate": 9.805561918611766e-06,
+      "loss": 0.3839,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5408463804061386,
+      "grad_norm": 0.4970063865184784,
+      "learning_rate": 9.804813950060909e-06,
+      "loss": 0.383,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5413114245853357,
+      "grad_norm": 0.46584850549697876,
+      "learning_rate": 9.804064574258092e-06,
+      "loss": 0.3885,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5417764687645327,
+      "grad_norm": 0.5352959632873535,
+      "learning_rate": 9.803313791422793e-06,
+      "loss": 0.4041,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5422415129437297,
+      "grad_norm": 0.5488241910934448,
+      "learning_rate": 9.802561601774905e-06,
+      "loss": 0.3859,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5427065571229267,
+      "grad_norm": 0.45431554317474365,
+      "learning_rate": 9.801808005534734e-06,
+      "loss": 0.4022,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5431716013021237,
+      "grad_norm": 0.5047218203544617,
+      "learning_rate": 9.801053002922994e-06,
+      "loss": 0.3866,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5436366454813207,
+      "grad_norm": 0.5306795239448547,
+      "learning_rate": 9.800296594160814e-06,
+      "loss": 0.3917,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5441016896605178,
+      "grad_norm": 0.48350954055786133,
+      "learning_rate": 9.799538779469734e-06,
+      "loss": 0.4058,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5445667338397148,
+      "grad_norm": 0.5609748959541321,
+      "learning_rate": 9.798779559071706e-06,
+      "loss": 0.3784,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5450317780189118,
+      "grad_norm": 0.4680517315864563,
+      "learning_rate": 9.798018933189089e-06,
+      "loss": 0.4218,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5454968221981088,
+      "grad_norm": 0.4936869442462921,
+      "learning_rate": 9.797256902044666e-06,
+      "loss": 0.3879,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5459618663773058,
+      "grad_norm": 0.48776108026504517,
+      "learning_rate": 9.796493465861613e-06,
+      "loss": 0.4006,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5464269105565028,
+      "grad_norm": 0.49075883626937866,
+      "learning_rate": 9.795728624863539e-06,
+      "loss": 0.4062,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5468919547356998,
+      "grad_norm": 0.4694967269897461,
+      "learning_rate": 9.794962379274448e-06,
+      "loss": 0.3955,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5473569989148969,
+      "grad_norm": 0.5107150077819824,
+      "learning_rate": 9.79419472931876e-06,
+      "loss": 0.4148,
+      "step": 1177
+    },
+    {
+      "epoch": 0.547822043094094,
+      "grad_norm": 0.4907507002353668,
+      "learning_rate": 9.793425675221308e-06,
+      "loss": 0.3937,
+      "step": 1178
+    },
+    {
+      "epoch": 0.548287087273291,
+      "grad_norm": 0.6052386164665222,
+      "learning_rate": 9.79265521720734e-06,
+      "loss": 0.3849,
+      "step": 1179
+    },
+    {
+      "epoch": 0.548752131452488,
+      "grad_norm": 0.5238395929336548,
+      "learning_rate": 9.791883355502503e-06,
+      "loss": 0.4057,
+      "step": 1180
+    },
+    {
+      "epoch": 0.549217175631685,
+      "grad_norm": 0.5540690422058105,
+      "learning_rate": 9.791110090332866e-06,
+      "loss": 0.3914,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5496822198108821,
+      "grad_norm": 0.5288317203521729,
+      "learning_rate": 9.79033542192491e-06,
+      "loss": 0.3932,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5501472639900791,
+      "grad_norm": 0.5038856863975525,
+      "learning_rate": 9.789559350505515e-06,
+      "loss": 0.398,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5506123081692761,
+      "grad_norm": 0.45208224654197693,
+      "learning_rate": 9.788781876301988e-06,
+      "loss": 0.4266,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5510773523484731,
+      "grad_norm": 0.5108579993247986,
+      "learning_rate": 9.78800299954203e-06,
+      "loss": 0.4126,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5515423965276701,
+      "grad_norm": 0.45797398686408997,
+      "learning_rate": 9.787222720453769e-06,
+      "loss": 0.4055,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5520074407068671,
+      "grad_norm": 0.43965256214141846,
+      "learning_rate": 9.78644103926573e-06,
+      "loss": 0.3847,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5524724848860642,
+      "grad_norm": 0.5082615613937378,
+      "learning_rate": 9.78565795620686e-06,
+      "loss": 0.3791,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5529375290652612,
+      "grad_norm": 0.5475904941558838,
+      "learning_rate": 9.784873471506509e-06,
+      "loss": 0.379,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5534025732444582,
+      "grad_norm": 0.5068783760070801,
+      "learning_rate": 9.784087585394437e-06,
+      "loss": 0.422,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5538676174236552,
+      "grad_norm": 0.48961615562438965,
+      "learning_rate": 9.783300298100822e-06,
+      "loss": 0.4116,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5543326616028522,
+      "grad_norm": 0.46741366386413574,
+      "learning_rate": 9.782511609856244e-06,
+      "loss": 0.4024,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5547977057820493,
+      "grad_norm": 0.4937583804130554,
+      "learning_rate": 9.7817215208917e-06,
+      "loss": 0.3889,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5552627499612464,
+      "grad_norm": 0.5000990033149719,
+      "learning_rate": 9.780930031438594e-06,
+      "loss": 0.4062,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5557277941404434,
+      "grad_norm": 0.5129031538963318,
+      "learning_rate": 9.780137141728737e-06,
+      "loss": 0.3849,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5561928383196404,
+      "grad_norm": 0.46762484312057495,
+      "learning_rate": 9.779342851994356e-06,
+      "loss": 0.3877,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5566578824988374,
+      "grad_norm": 0.4732123911380768,
+      "learning_rate": 9.778547162468087e-06,
+      "loss": 0.3746,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5571229266780344,
+      "grad_norm": 0.44343671202659607,
+      "learning_rate": 9.77775007338297e-06,
+      "loss": 0.3878,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5575879708572314,
+      "grad_norm": 0.45024973154067993,
+      "learning_rate": 9.776951584972464e-06,
+      "loss": 0.3907,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5580530150364285,
+      "grad_norm": 0.4816182255744934,
+      "learning_rate": 9.776151697470431e-06,
+      "loss": 0.4158,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5585180592156255,
+      "grad_norm": 0.4887128472328186,
+      "learning_rate": 9.775350411111145e-06,
+      "loss": 0.4127,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5589831033948225,
+      "grad_norm": 0.520602285861969,
+      "learning_rate": 9.77454772612929e-06,
+      "loss": 0.4024,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5594481475740195,
+      "grad_norm": 0.4444476366043091,
+      "learning_rate": 9.773743642759961e-06,
+      "loss": 0.4007,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5599131917532165,
+      "grad_norm": 0.5314255356788635,
+      "learning_rate": 9.77293816123866e-06,
+      "loss": 0.4009,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5603782359324135,
+      "grad_norm": 0.5080925226211548,
+      "learning_rate": 9.7721312818013e-06,
+      "loss": 0.3741,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5608432801116106,
+      "grad_norm": 0.4945761263370514,
+      "learning_rate": 9.7713230046842e-06,
+      "loss": 0.4204,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5613083242908077,
+      "grad_norm": 0.5242900848388672,
+      "learning_rate": 9.770513330124094e-06,
+      "loss": 0.3796,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5617733684700047,
+      "grad_norm": 0.48071667551994324,
+      "learning_rate": 9.769702258358123e-06,
+      "loss": 0.4133,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5622384126492017,
+      "grad_norm": 0.497071772813797,
+      "learning_rate": 9.768889789623833e-06,
+      "loss": 0.3845,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5627034568283987,
+      "grad_norm": 0.5164023041725159,
+      "learning_rate": 9.768075924159185e-06,
+      "loss": 0.3954,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5631685010075957,
+      "grad_norm": 0.46989476680755615,
+      "learning_rate": 9.767260662202551e-06,
+      "loss": 0.4029,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5636335451867928,
+      "grad_norm": 0.5252077579498291,
+      "learning_rate": 9.766444003992704e-06,
+      "loss": 0.3907,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5640985893659898,
+      "grad_norm": 0.5021181106567383,
+      "learning_rate": 9.765625949768828e-06,
+      "loss": 0.3868,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5645636335451868,
+      "grad_norm": 0.4650791585445404,
+      "learning_rate": 9.764806499770521e-06,
+      "loss": 0.3805,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5650286777243838,
+      "grad_norm": 0.47574383020401,
+      "learning_rate": 9.763985654237785e-06,
+      "loss": 0.3812,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5654937219035808,
+      "grad_norm": 0.44989126920700073,
+      "learning_rate": 9.763163413411034e-06,
+      "loss": 0.3736,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5659587660827778,
+      "grad_norm": 0.5040990710258484,
+      "learning_rate": 9.762339777531088e-06,
+      "loss": 0.3744,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5664238102619749,
+      "grad_norm": 0.5459234118461609,
+      "learning_rate": 9.761514746839176e-06,
+      "loss": 0.3916,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5668888544411719,
+      "grad_norm": 0.48448291420936584,
+      "learning_rate": 9.760688321576938e-06,
+      "loss": 0.4425,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5673538986203689,
+      "grad_norm": 0.5275092720985413,
+      "learning_rate": 9.759860501986417e-06,
+      "loss": 0.3998,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5678189427995659,
+      "grad_norm": 0.6015114188194275,
+      "learning_rate": 9.759031288310072e-06,
+      "loss": 0.4155,
+      "step": 1221
+    },
+    {
+      "epoch": 0.568283986978763,
+      "grad_norm": 0.47165900468826294,
+      "learning_rate": 9.758200680790764e-06,
+      "loss": 0.387,
+      "step": 1222
+    },
+    {
+      "epoch": 0.56874903115796,
+      "grad_norm": 0.592989444732666,
+      "learning_rate": 9.757368679671764e-06,
+      "loss": 0.3973,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5692140753371571,
+      "grad_norm": 0.4669657349586487,
+      "learning_rate": 9.756535285196754e-06,
+      "loss": 0.3762,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5696791195163541,
+      "grad_norm": 0.48349782824516296,
+      "learning_rate": 9.755700497609819e-06,
+      "loss": 0.4043,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5701441636955511,
+      "grad_norm": 0.5364663600921631,
+      "learning_rate": 9.754864317155455e-06,
+      "loss": 0.4216,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5706092078747481,
+      "grad_norm": 0.5064231753349304,
+      "learning_rate": 9.754026744078569e-06,
+      "loss": 0.4004,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5710742520539451,
+      "grad_norm": 0.48567453026771545,
+      "learning_rate": 9.753187778624467e-06,
+      "loss": 0.3985,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5715392962331421,
+      "grad_norm": 0.5714061856269836,
+      "learning_rate": 9.752347421038873e-06,
+      "loss": 0.3836,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5720043404123392,
+      "grad_norm": 0.4642956256866455,
+      "learning_rate": 9.751505671567914e-06,
+      "loss": 0.3919,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5724693845915362,
+      "grad_norm": 0.4755660891532898,
+      "learning_rate": 9.750662530458121e-06,
+      "loss": 0.3881,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5729344287707332,
+      "grad_norm": 0.48824259638786316,
+      "learning_rate": 9.749817997956438e-06,
+      "loss": 0.3871,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5733994729499302,
+      "grad_norm": 0.49661746621131897,
+      "learning_rate": 9.748972074310217e-06,
+      "loss": 0.3709,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5738645171291272,
+      "grad_norm": 0.4752178490161896,
+      "learning_rate": 9.748124759767215e-06,
+      "loss": 0.3907,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5743295613083242,
+      "grad_norm": 0.47395604848861694,
+      "learning_rate": 9.747276054575593e-06,
+      "loss": 0.416,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5747946054875213,
+      "grad_norm": 0.4589254558086395,
+      "learning_rate": 9.746425958983925e-06,
+      "loss": 0.4059,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5752596496667184,
+      "grad_norm": 0.42140620946884155,
+      "learning_rate": 9.745574473241193e-06,
+      "loss": 0.3705,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5757246938459154,
+      "grad_norm": 0.45815083384513855,
+      "learning_rate": 9.744721597596778e-06,
+      "loss": 0.4068,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5761897380251124,
+      "grad_norm": 0.45956507325172424,
+      "learning_rate": 9.743867332300478e-06,
+      "loss": 0.3637,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5766547822043094,
+      "grad_norm": 0.49071380496025085,
+      "learning_rate": 9.743011677602493e-06,
+      "loss": 0.375,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5771198263835065,
+      "grad_norm": 0.4949245750904083,
+      "learning_rate": 9.742154633753428e-06,
+      "loss": 0.378,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5775848705627035,
+      "grad_norm": 0.5535846948623657,
+      "learning_rate": 9.741296201004298e-06,
+      "loss": 0.3901,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5780499147419005,
+      "grad_norm": 0.44817158579826355,
+      "learning_rate": 9.740436379606524e-06,
+      "loss": 0.4016,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5785149589210975,
+      "grad_norm": 0.5253201723098755,
+      "learning_rate": 9.739575169811934e-06,
+      "loss": 0.38,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5789800031002945,
+      "grad_norm": 0.6024049520492554,
+      "learning_rate": 9.738712571872765e-06,
+      "loss": 0.3948,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5794450472794915,
+      "grad_norm": 0.41917502880096436,
+      "learning_rate": 9.737848586041652e-06,
+      "loss": 0.3869,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5799100914586885,
+      "grad_norm": 0.5947693586349487,
+      "learning_rate": 9.736983212571646e-06,
+      "loss": 0.3991,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5803751356378856,
+      "grad_norm": 0.6152756810188293,
+      "learning_rate": 9.736116451716203e-06,
+      "loss": 0.4075,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5808401798170826,
+      "grad_norm": 0.48533540964126587,
+      "learning_rate": 9.735248303729178e-06,
+      "loss": 0.3881,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5813052239962796,
+      "grad_norm": 0.6273863315582275,
+      "learning_rate": 9.734378768864843e-06,
+      "loss": 0.3723,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5817702681754767,
+      "grad_norm": 0.5250769257545471,
+      "learning_rate": 9.733507847377866e-06,
+      "loss": 0.3853,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5822353123546737,
+      "grad_norm": 0.49252286553382874,
+      "learning_rate": 9.73263553952333e-06,
+      "loss": 0.3889,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5827003565338708,
+      "grad_norm": 0.5308325886726379,
+      "learning_rate": 9.731761845556713e-06,
+      "loss": 0.3936,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5831654007130678,
+      "grad_norm": 0.5030363202095032,
+      "learning_rate": 9.730886765733914e-06,
+      "loss": 0.3786,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5836304448922648,
+      "grad_norm": 0.46710026264190674,
+      "learning_rate": 9.730010300311226e-06,
+      "loss": 0.4055,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5840954890714618,
+      "grad_norm": 0.5640740394592285,
+      "learning_rate": 9.72913244954535e-06,
+      "loss": 0.4313,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5845605332506588,
+      "grad_norm": 0.43550801277160645,
+      "learning_rate": 9.728253213693395e-06,
+      "loss": 0.3745,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5850255774298558,
+      "grad_norm": 0.5071004629135132,
+      "learning_rate": 9.727372593012875e-06,
+      "loss": 0.4029,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5854906216090529,
+      "grad_norm": 0.5142107605934143,
+      "learning_rate": 9.72649058776171e-06,
+      "loss": 0.3912,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5859556657882499,
+      "grad_norm": 0.4210042655467987,
+      "learning_rate": 9.725607198198227e-06,
+      "loss": 0.3884,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5864207099674469,
+      "grad_norm": 0.47974395751953125,
+      "learning_rate": 9.724722424581154e-06,
+      "loss": 0.3834,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5868857541466439,
+      "grad_norm": 0.49919867515563965,
+      "learning_rate": 9.723836267169626e-06,
+      "loss": 0.394,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5873507983258409,
+      "grad_norm": 0.46105697751045227,
+      "learning_rate": 9.722948726223185e-06,
+      "loss": 0.383,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5878158425050379,
+      "grad_norm": 0.5934450030326843,
+      "learning_rate": 9.72205980200178e-06,
+      "loss": 0.3716,
+      "step": 1264
+    },
+    {
+      "epoch": 0.588280886684235,
+      "grad_norm": 0.5127496719360352,
+      "learning_rate": 9.72116949476576e-06,
+      "loss": 0.3684,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5887459308634321,
+      "grad_norm": 0.4598689675331116,
+      "learning_rate": 9.720277804775879e-06,
+      "loss": 0.4068,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5892109750426291,
+      "grad_norm": 0.5856637358665466,
+      "learning_rate": 9.719384732293302e-06,
+      "loss": 0.3637,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5896760192218261,
+      "grad_norm": 0.5395975708961487,
+      "learning_rate": 9.718490277579595e-06,
+      "loss": 0.3902,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5901410634010231,
+      "grad_norm": 0.46271106600761414,
+      "learning_rate": 9.71759444089673e-06,
+      "loss": 0.374,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5906061075802201,
+      "grad_norm": 0.5440768003463745,
+      "learning_rate": 9.716697222507081e-06,
+      "loss": 0.3957,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5910711517594172,
+      "grad_norm": 0.5980409979820251,
+      "learning_rate": 9.715798622673429e-06,
+      "loss": 0.3879,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5915361959386142,
+      "grad_norm": 0.5140041708946228,
+      "learning_rate": 9.71489864165896e-06,
+      "loss": 0.3859,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5920012401178112,
+      "grad_norm": 0.6424972414970398,
+      "learning_rate": 9.713997279727265e-06,
+      "loss": 0.4169,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5924662842970082,
+      "grad_norm": 0.5433067679405212,
+      "learning_rate": 9.713094537142336e-06,
+      "loss": 0.3957,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5929313284762052,
+      "grad_norm": 0.5382039546966553,
+      "learning_rate": 9.712190414168573e-06,
+      "loss": 0.3878,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5933963726554022,
+      "grad_norm": 0.5087778568267822,
+      "learning_rate": 9.711284911070777e-06,
+      "loss": 0.4115,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5938614168345993,
+      "grad_norm": 0.5524824261665344,
+      "learning_rate": 9.71037802811416e-06,
+      "loss": 0.3898,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5943264610137963,
+      "grad_norm": 0.579130232334137,
+      "learning_rate": 9.709469765564328e-06,
+      "loss": 0.3868,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5947915051929933,
+      "grad_norm": 0.5298312306404114,
+      "learning_rate": 9.708560123687298e-06,
+      "loss": 0.4156,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5952565493721904,
+      "grad_norm": 0.5209439396858215,
+      "learning_rate": 9.707649102749488e-06,
+      "loss": 0.384,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5957215935513874,
+      "grad_norm": 0.5347442030906677,
+      "learning_rate": 9.706736703017725e-06,
+      "loss": 0.3886,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5961866377305844,
+      "grad_norm": 0.5219368934631348,
+      "learning_rate": 9.705822924759235e-06,
+      "loss": 0.388,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5966516819097815,
+      "grad_norm": 0.618898868560791,
+      "learning_rate": 9.704907768241648e-06,
+      "loss": 0.3944,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5971167260889785,
+      "grad_norm": 0.5122677087783813,
+      "learning_rate": 9.703991233732995e-06,
+      "loss": 0.3902,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5975817702681755,
+      "grad_norm": 0.5251182317733765,
+      "learning_rate": 9.70307332150172e-06,
+      "loss": 0.4144,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5980468144473725,
+      "grad_norm": 0.5196467041969299,
+      "learning_rate": 9.702154031816659e-06,
+      "loss": 0.3837,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5985118586265695,
+      "grad_norm": 0.5128290057182312,
+      "learning_rate": 9.701233364947062e-06,
+      "loss": 0.3989,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5989769028057665,
+      "grad_norm": 0.5336976051330566,
+      "learning_rate": 9.700311321162577e-06,
+      "loss": 0.3915,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5994419469849636,
+      "grad_norm": 0.6457362771034241,
+      "learning_rate": 9.69938790073325e-06,
+      "loss": 0.3883,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5999069911641606,
+      "grad_norm": 0.5426515340805054,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.4055,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6003720353433576,
+      "grad_norm": 0.49356475472450256,
+      "learning_rate": 9.697536931022308e-06,
+      "loss": 0.3999,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6008370795225546,
+      "grad_norm": 0.5430606007575989,
+      "learning_rate": 9.69660938228281e-06,
+      "loss": 0.3807,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6013021237017516,
+      "grad_norm": 0.5589854121208191,
+      "learning_rate": 9.695680457982713e-06,
+      "loss": 0.4153,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6017671678809486,
+      "grad_norm": 0.6247249841690063,
+      "learning_rate": 9.694750158394081e-06,
+      "loss": 0.4118,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6022322120601458,
+      "grad_norm": 0.6207286715507507,
+      "learning_rate": 9.693818483789386e-06,
+      "loss": 0.3833,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6026972562393428,
+      "grad_norm": 0.5082651376724243,
+      "learning_rate": 9.692885434441498e-06,
+      "loss": 0.4056,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6031623004185398,
+      "grad_norm": 0.5024476051330566,
+      "learning_rate": 9.691951010623696e-06,
+      "loss": 0.3943,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6036273445977368,
+      "grad_norm": 0.5954005718231201,
+      "learning_rate": 9.691015212609654e-06,
+      "loss": 0.3939,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6040923887769338,
+      "grad_norm": 0.6106828451156616,
+      "learning_rate": 9.690078040673454e-06,
+      "loss": 0.387,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6045574329561308,
+      "grad_norm": 0.5513212084770203,
+      "learning_rate": 9.689139495089575e-06,
+      "loss": 0.3904,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6050224771353279,
+      "grad_norm": 0.6389715671539307,
+      "learning_rate": 9.688199576132905e-06,
+      "loss": 0.3904,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6054875213145249,
+      "grad_norm": 0.4722382426261902,
+      "learning_rate": 9.687258284078733e-06,
+      "loss": 0.393,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6059525654937219,
+      "grad_norm": 0.5727136731147766,
+      "learning_rate": 9.686315619202743e-06,
+      "loss": 0.394,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6064176096729189,
+      "grad_norm": 0.6134669780731201,
+      "learning_rate": 9.685371581781029e-06,
+      "loss": 0.3903,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6068826538521159,
+      "grad_norm": 0.5710572600364685,
+      "learning_rate": 9.684426172090084e-06,
+      "loss": 0.4115,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6073476980313129,
+      "grad_norm": 0.5541101098060608,
+      "learning_rate": 9.683479390406803e-06,
+      "loss": 0.3935,
+      "step": 1306
+    },
+    {
+      "epoch": 0.60781274221051,
+      "grad_norm": 0.5224672555923462,
+      "learning_rate": 9.682531237008483e-06,
+      "loss": 0.3989,
+      "step": 1307
+    },
+    {
+      "epoch": 0.608277786389707,
+      "grad_norm": 0.5431579351425171,
+      "learning_rate": 9.681581712172824e-06,
+      "loss": 0.3995,
+      "step": 1308
+    },
+    {
+      "epoch": 0.608742830568904,
+      "grad_norm": 0.5996741056442261,
+      "learning_rate": 9.680630816177924e-06,
+      "loss": 0.4012,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6092078747481011,
+      "grad_norm": 0.5371108055114746,
+      "learning_rate": 9.679678549302287e-06,
+      "loss": 0.404,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6096729189272981,
+      "grad_norm": 0.4972769320011139,
+      "learning_rate": 9.678724911824815e-06,
+      "loss": 0.3939,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6101379631064952,
+      "grad_norm": 0.7065192461013794,
+      "learning_rate": 9.677769904024815e-06,
+      "loss": 0.3968,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6106030072856922,
+      "grad_norm": 0.5725729465484619,
+      "learning_rate": 9.676813526181989e-06,
+      "loss": 0.4079,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6110680514648892,
+      "grad_norm": 0.5286259651184082,
+      "learning_rate": 9.675855778576448e-06,
+      "loss": 0.3818,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6115330956440862,
+      "grad_norm": 0.4823831617832184,
+      "learning_rate": 9.674896661488702e-06,
+      "loss": 0.3856,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6119981398232832,
+      "grad_norm": 0.5325484871864319,
+      "learning_rate": 9.673936175199657e-06,
+      "loss": 0.3887,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6124631840024802,
+      "grad_norm": 0.5285559296607971,
+      "learning_rate": 9.672974319990627e-06,
+      "loss": 0.4181,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6129282281816772,
+      "grad_norm": 0.5194781422615051,
+      "learning_rate": 9.672011096143323e-06,
+      "loss": 0.4084,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6133932723608743,
+      "grad_norm": 0.5809126496315002,
+      "learning_rate": 9.671046503939857e-06,
+      "loss": 0.3985,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6138583165400713,
+      "grad_norm": 0.443920373916626,
+      "learning_rate": 9.670080543662742e-06,
+      "loss": 0.386,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6143233607192683,
+      "grad_norm": 0.5228665471076965,
+      "learning_rate": 9.669113215594892e-06,
+      "loss": 0.391,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6147884048984653,
+      "grad_norm": 0.504928469657898,
+      "learning_rate": 9.668144520019622e-06,
+      "loss": 0.402,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6152534490776623,
+      "grad_norm": 0.5686684250831604,
+      "learning_rate": 9.667174457220648e-06,
+      "loss": 0.3837,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6157184932568595,
+      "grad_norm": 0.5331105589866638,
+      "learning_rate": 9.666203027482086e-06,
+      "loss": 0.3919,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6161835374360565,
+      "grad_norm": 0.5707592368125916,
+      "learning_rate": 9.665230231088451e-06,
+      "loss": 0.3985,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6166485816152535,
+      "grad_norm": 0.5226622223854065,
+      "learning_rate": 9.664256068324657e-06,
+      "loss": 0.399,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6171136257944505,
+      "grad_norm": 0.5462534427642822,
+      "learning_rate": 9.663280539476026e-06,
+      "loss": 0.399,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6175786699736475,
+      "grad_norm": 0.4789917469024658,
+      "learning_rate": 9.66230364482827e-06,
+      "loss": 0.3907,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6180437141528445,
+      "grad_norm": 0.45184510946273804,
+      "learning_rate": 9.661325384667508e-06,
+      "loss": 0.3753,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6185087583320416,
+      "grad_norm": 0.5181789398193359,
+      "learning_rate": 9.660345759280254e-06,
+      "loss": 0.3864,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6189738025112386,
+      "grad_norm": 0.504653811454773,
+      "learning_rate": 9.659364768953426e-06,
+      "loss": 0.3747,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6194388466904356,
+      "grad_norm": 0.47670918703079224,
+      "learning_rate": 9.65838241397434e-06,
+      "loss": 0.3764,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6199038908696326,
+      "grad_norm": 0.5914029479026794,
+      "learning_rate": 9.657398694630713e-06,
+      "loss": 0.4123,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6203689350488296,
+      "grad_norm": 0.443935751914978,
+      "learning_rate": 9.656413611210657e-06,
+      "loss": 0.4088,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6208339792280266,
+      "grad_norm": 0.4935535788536072,
+      "learning_rate": 9.655427164002692e-06,
+      "loss": 0.4064,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6212990234072236,
+      "grad_norm": 0.41665172576904297,
+      "learning_rate": 9.654439353295728e-06,
+      "loss": 0.3886,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6217640675864207,
+      "grad_norm": 0.524173378944397,
+      "learning_rate": 9.653450179379081e-06,
+      "loss": 0.4056,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6222291117656177,
+      "grad_norm": 0.497357040643692,
+      "learning_rate": 9.65245964254246e-06,
+      "loss": 0.4009,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6226941559448148,
+      "grad_norm": 0.46005746722221375,
+      "learning_rate": 9.651467743075984e-06,
+      "loss": 0.3666,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6231592001240118,
+      "grad_norm": 0.5369196534156799,
+      "learning_rate": 9.650474481270159e-06,
+      "loss": 0.3952,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6236242443032088,
+      "grad_norm": 0.5053161978721619,
+      "learning_rate": 9.649479857415896e-06,
+      "loss": 0.4066,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6240892884824059,
+      "grad_norm": 0.50250643491745,
+      "learning_rate": 9.648483871804506e-06,
+      "loss": 0.3863,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6245543326616029,
+      "grad_norm": 0.540744960308075,
+      "learning_rate": 9.647486524727696e-06,
+      "loss": 0.3592,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6250193768407999,
+      "grad_norm": 0.4549930989742279,
+      "learning_rate": 9.646487816477575e-06,
+      "loss": 0.4106,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6254844210199969,
+      "grad_norm": 0.5376303791999817,
+      "learning_rate": 9.645487747346643e-06,
+      "loss": 0.3856,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6259494651991939,
+      "grad_norm": 0.5555692911148071,
+      "learning_rate": 9.644486317627808e-06,
+      "loss": 0.3952,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6264145093783909,
+      "grad_norm": 0.48616769909858704,
+      "learning_rate": 9.643483527614372e-06,
+      "loss": 0.3639,
+      "step": 1347
+    },
+    {
+      "epoch": 0.626879553557588,
+      "grad_norm": 0.5379546284675598,
+      "learning_rate": 9.642479377600036e-06,
+      "loss": 0.3779,
+      "step": 1348
+    },
+    {
+      "epoch": 0.627344597736785,
+      "grad_norm": 0.5338390469551086,
+      "learning_rate": 9.641473867878898e-06,
+      "loss": 0.3819,
+      "step": 1349
+    },
+    {
+      "epoch": 0.627809641915982,
+      "grad_norm": 0.4833146929740906,
+      "learning_rate": 9.640466998745456e-06,
+      "loss": 0.38,
+      "step": 1350
+    },
+    {
+      "epoch": 0.628274686095179,
+      "grad_norm": 0.5547218918800354,
+      "learning_rate": 9.639458770494608e-06,
+      "loss": 0.3975,
+      "step": 1351
+    },
+    {
+      "epoch": 0.628739730274376,
+      "grad_norm": 0.4861772060394287,
+      "learning_rate": 9.638449183421644e-06,
+      "loss": 0.3833,
+      "step": 1352
+    },
+    {
+      "epoch": 0.629204774453573,
+      "grad_norm": 0.48040562868118286,
+      "learning_rate": 9.637438237822256e-06,
+      "loss": 0.3903,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6296698186327702,
+      "grad_norm": 0.49921414256095886,
+      "learning_rate": 9.636425933992536e-06,
+      "loss": 0.4055,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6301348628119672,
+      "grad_norm": 0.5730396509170532,
+      "learning_rate": 9.63541227222897e-06,
+      "loss": 0.3813,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6305999069911642,
+      "grad_norm": 0.49284279346466064,
+      "learning_rate": 9.634397252828444e-06,
+      "loss": 0.3643,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6310649511703612,
+      "grad_norm": 0.5882490277290344,
+      "learning_rate": 9.63338087608824e-06,
+      "loss": 0.3976,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6315299953495582,
+      "grad_norm": 0.5045767426490784,
+      "learning_rate": 9.632363142306036e-06,
+      "loss": 0.4051,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6319950395287552,
+      "grad_norm": 0.5159674882888794,
+      "learning_rate": 9.631344051779914e-06,
+      "loss": 0.3642,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6324600837079523,
+      "grad_norm": 0.6393707990646362,
+      "learning_rate": 9.630323604808344e-06,
+      "loss": 0.4197,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6329251278871493,
+      "grad_norm": 0.46538159251213074,
+      "learning_rate": 9.629301801690205e-06,
+      "loss": 0.3911,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6333901720663463,
+      "grad_norm": 0.5735601782798767,
+      "learning_rate": 9.62827864272476e-06,
+      "loss": 0.401,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6338552162455433,
+      "grad_norm": 0.44864973425865173,
+      "learning_rate": 9.62725412821168e-06,
+      "loss": 0.3685,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6343202604247403,
+      "grad_norm": 0.46898770332336426,
+      "learning_rate": 9.626228258451027e-06,
+      "loss": 0.3885,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6347853046039373,
+      "grad_norm": 0.4860677421092987,
+      "learning_rate": 9.625201033743262e-06,
+      "loss": 0.3992,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6352503487831344,
+      "grad_norm": 0.47478920221328735,
+      "learning_rate": 9.62417245438924e-06,
+      "loss": 0.4055,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6357153929623314,
+      "grad_norm": 0.5337481498718262,
+      "learning_rate": 9.623142520690219e-06,
+      "loss": 0.3892,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6361804371415285,
+      "grad_norm": 0.5208576321601868,
+      "learning_rate": 9.622111232947847e-06,
+      "loss": 0.3763,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6366454813207255,
+      "grad_norm": 0.490197092294693,
+      "learning_rate": 9.621078591464174e-06,
+      "loss": 0.4076,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6371105254999225,
+      "grad_norm": 0.5205878615379333,
+      "learning_rate": 9.620044596541642e-06,
+      "loss": 0.3825,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6375755696791195,
+      "grad_norm": 0.4583769142627716,
+      "learning_rate": 9.61900924848309e-06,
+      "loss": 0.3924,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6380406138583166,
+      "grad_norm": 0.4924391806125641,
+      "learning_rate": 9.617972547591759e-06,
+      "loss": 0.4019,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6385056580375136,
+      "grad_norm": 0.5124062895774841,
+      "learning_rate": 9.616934494171277e-06,
+      "loss": 0.3913,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6389707022167106,
+      "grad_norm": 0.5022979974746704,
+      "learning_rate": 9.615895088525677e-06,
+      "loss": 0.3889,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6394357463959076,
+      "grad_norm": 0.42661193013191223,
+      "learning_rate": 9.614854330959382e-06,
+      "loss": 0.3649,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6399007905751046,
+      "grad_norm": 0.6055207848548889,
+      "learning_rate": 9.613812221777212e-06,
+      "loss": 0.3975,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6403658347543016,
+      "grad_norm": 0.6047768592834473,
+      "learning_rate": 9.612768761284386e-06,
+      "loss": 0.3953,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6408308789334987,
+      "grad_norm": 0.4532669186592102,
+      "learning_rate": 9.611723949786517e-06,
+      "loss": 0.3992,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6412959231126957,
+      "grad_norm": 0.45877280831336975,
+      "learning_rate": 9.610677787589611e-06,
+      "loss": 0.4096,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6417609672918927,
+      "grad_norm": 0.599404513835907,
+      "learning_rate": 9.609630275000072e-06,
+      "loss": 0.4022,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6422260114710897,
+      "grad_norm": 0.5159070491790771,
+      "learning_rate": 9.608581412324701e-06,
+      "loss": 0.3832,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6426910556502867,
+      "grad_norm": 0.45905065536499023,
+      "learning_rate": 9.607531199870692e-06,
+      "loss": 0.3963,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6431560998294839,
+      "grad_norm": 0.6395869851112366,
+      "learning_rate": 9.606479637945635e-06,
+      "loss": 0.4031,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6436211440086809,
+      "grad_norm": 0.5056988000869751,
+      "learning_rate": 9.60542672685752e-06,
+      "loss": 0.3899,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6440861881878779,
+      "grad_norm": 0.56319260597229,
+      "learning_rate": 9.604372466914717e-06,
+      "loss": 0.388,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6445512323670749,
+      "grad_norm": 0.6158375144004822,
+      "learning_rate": 9.603316858426014e-06,
+      "loss": 0.3957,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6450162765462719,
+      "grad_norm": 0.4532202184200287,
+      "learning_rate": 9.60225990170057e-06,
+      "loss": 0.3801,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6454813207254689,
+      "grad_norm": 0.6414798498153687,
+      "learning_rate": 9.60120159704796e-06,
+      "loss": 0.3984,
+      "step": 1388
+    },
+    {
+      "epoch": 0.645946364904666,
+      "grad_norm": 0.4911268651485443,
+      "learning_rate": 9.600141944778139e-06,
+      "loss": 0.3868,
+      "step": 1389
+    },
+    {
+      "epoch": 0.646411409083863,
+      "grad_norm": 0.4856494963169098,
+      "learning_rate": 9.599080945201462e-06,
+      "loss": 0.421,
+      "step": 1390
+    },
+    {
+      "epoch": 0.64687645326306,
+      "grad_norm": 0.49976852536201477,
+      "learning_rate": 9.598018598628682e-06,
+      "loss": 0.4045,
+      "step": 1391
+    },
+    {
+      "epoch": 0.647341497442257,
+      "grad_norm": 0.4958864748477936,
+      "learning_rate": 9.59695490537094e-06,
+      "loss": 0.3759,
+      "step": 1392
+    },
+    {
+      "epoch": 0.647806541621454,
+      "grad_norm": 0.4640512764453888,
+      "learning_rate": 9.595889865739774e-06,
+      "loss": 0.3725,
+      "step": 1393
+    },
+    {
+      "epoch": 0.648271585800651,
+      "grad_norm": 0.4767007529735565,
+      "learning_rate": 9.594823480047118e-06,
+      "loss": 0.3764,
+      "step": 1394
+    },
+    {
+      "epoch": 0.648736629979848,
+      "grad_norm": 0.509242594242096,
+      "learning_rate": 9.5937557486053e-06,
+      "loss": 0.374,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6492016741590451,
+      "grad_norm": 0.49067145586013794,
+      "learning_rate": 9.59268667172704e-06,
+      "loss": 0.3969,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6496667183382422,
+      "grad_norm": 0.5714993476867676,
+      "learning_rate": 9.591616249725456e-06,
+      "loss": 0.3665,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6501317625174392,
+      "grad_norm": 0.5996623635292053,
+      "learning_rate": 9.590544482914052e-06,
+      "loss": 0.3722,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6505968066966362,
+      "grad_norm": 0.4260701835155487,
+      "learning_rate": 9.589471371606735e-06,
+      "loss": 0.3707,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6510618508758332,
+      "grad_norm": 0.6318243741989136,
+      "learning_rate": 9.5883969161178e-06,
+      "loss": 0.4004,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6515268950550303,
+      "grad_norm": 0.5253607630729675,
+      "learning_rate": 9.587321116761938e-06,
+      "loss": 0.3856,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6519919392342273,
+      "grad_norm": 0.4526579976081848,
+      "learning_rate": 9.586243973854234e-06,
+      "loss": 0.4028,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6524569834134243,
+      "grad_norm": 0.5988115668296814,
+      "learning_rate": 9.585165487710167e-06,
+      "loss": 0.3829,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6529220275926213,
+      "grad_norm": 0.47281304001808167,
+      "learning_rate": 9.584085658645604e-06,
+      "loss": 0.3911,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6533870717718183,
+      "grad_norm": 0.48429980874061584,
+      "learning_rate": 9.583004486976813e-06,
+      "loss": 0.3721,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6538521159510153,
+      "grad_norm": 0.5094161033630371,
+      "learning_rate": 9.58192197302045e-06,
+      "loss": 0.3976,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6543171601302123,
+      "grad_norm": 0.41321924328804016,
+      "learning_rate": 9.580838117093564e-06,
+      "loss": 0.3871,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6547822043094094,
+      "grad_norm": 0.5384671092033386,
+      "learning_rate": 9.579752919513602e-06,
+      "loss": 0.4158,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6552472484886064,
+      "grad_norm": 0.5002989768981934,
+      "learning_rate": 9.5786663805984e-06,
+      "loss": 0.3833,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6557122926678034,
+      "grad_norm": 0.4137031137943268,
+      "learning_rate": 9.577578500666187e-06,
+      "loss": 0.3849,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6561773368470004,
+      "grad_norm": 0.5422747731208801,
+      "learning_rate": 9.576489280035584e-06,
+      "loss": 0.4073,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6566423810261975,
+      "grad_norm": 0.49949032068252563,
+      "learning_rate": 9.57539871902561e-06,
+      "loss": 0.3724,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6571074252053946,
+      "grad_norm": 0.4298678934574127,
+      "learning_rate": 9.574306817955669e-06,
+      "loss": 0.4176,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6575724693845916,
+      "grad_norm": 0.48693087697029114,
+      "learning_rate": 9.57321357714556e-06,
+      "loss": 0.404,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6580375135637886,
+      "grad_norm": 0.5552796721458435,
+      "learning_rate": 9.572118996915482e-06,
+      "loss": 0.4036,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6585025577429856,
+      "grad_norm": 0.4444204866886139,
+      "learning_rate": 9.571023077586012e-06,
+      "loss": 0.391,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6589676019221826,
+      "grad_norm": 0.46437713503837585,
+      "learning_rate": 9.569925819478132e-06,
+      "loss": 0.4061,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6594326461013796,
+      "grad_norm": 0.47912266850471497,
+      "learning_rate": 9.56882722291321e-06,
+      "loss": 0.3957,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6598976902805767,
+      "grad_norm": 0.4714813530445099,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.4099,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6603627344597737,
+      "grad_norm": 0.4540778398513794,
+      "learning_rate": 9.566626015699673e-06,
+      "loss": 0.3901,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6608277786389707,
+      "grad_norm": 0.47948190569877625,
+      "learning_rate": 9.565523405695756e-06,
+      "loss": 0.3996,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6612928228181677,
+      "grad_norm": 0.4244531989097595,
+      "learning_rate": 9.564419458524193e-06,
+      "loss": 0.3919,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6617578669973647,
+      "grad_norm": 0.4253584146499634,
+      "learning_rate": 9.563314174508312e-06,
+      "loss": 0.3817,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6622229111765617,
+      "grad_norm": 0.4504033923149109,
+      "learning_rate": 9.56220755397183e-06,
+      "loss": 0.3648,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6626879553557588,
+      "grad_norm": 0.4773210883140564,
+      "learning_rate": 9.561099597238862e-06,
+      "loss": 0.385,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6631529995349558,
+      "grad_norm": 0.4333602488040924,
+      "learning_rate": 9.559990304633906e-06,
+      "loss": 0.3873,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6636180437141529,
+      "grad_norm": 0.48123425245285034,
+      "learning_rate": 9.55887967648186e-06,
+      "loss": 0.3683,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6640830878933499,
+      "grad_norm": 0.5246723890304565,
+      "learning_rate": 9.557767713108009e-06,
+      "loss": 0.3702,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6645481320725469,
+      "grad_norm": 0.4679216146469116,
+      "learning_rate": 9.556654414838025e-06,
+      "loss": 0.3839,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6650131762517439,
+      "grad_norm": 0.5367023944854736,
+      "learning_rate": 9.555539781997978e-06,
+      "loss": 0.3777,
+      "step": 1430
+    },
+    {
+      "epoch": 0.665478220430941,
+      "grad_norm": 0.48515602946281433,
+      "learning_rate": 9.554423814914324e-06,
+      "loss": 0.3749,
+      "step": 1431
+    },
+    {
+      "epoch": 0.665943264610138,
+      "grad_norm": 0.46489590406417847,
+      "learning_rate": 9.553306513913915e-06,
+      "loss": 0.3707,
+      "step": 1432
+    },
+    {
+      "epoch": 0.666408308789335,
+      "grad_norm": 0.4412256181240082,
+      "learning_rate": 9.552187879323987e-06,
+      "loss": 0.3971,
+      "step": 1433
+    },
+    {
+      "epoch": 0.666873352968532,
+      "grad_norm": 0.4211561977863312,
+      "learning_rate": 9.551067911472172e-06,
+      "loss": 0.39,
+      "step": 1434
+    },
+    {
+      "epoch": 0.667338397147729,
+      "grad_norm": 0.48854532837867737,
+      "learning_rate": 9.549946610686488e-06,
+      "loss": 0.4125,
+      "step": 1435
+    },
+    {
+      "epoch": 0.667803441326926,
+      "grad_norm": 0.47784653306007385,
+      "learning_rate": 9.548823977295348e-06,
+      "loss": 0.3656,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6682684855061231,
+      "grad_norm": 0.4391965866088867,
+      "learning_rate": 9.547700011627552e-06,
+      "loss": 0.3929,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6687335296853201,
+      "grad_norm": 0.4571762979030609,
+      "learning_rate": 9.546574714012291e-06,
+      "loss": 0.4102,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6691985738645171,
+      "grad_norm": 0.4999452531337738,
+      "learning_rate": 9.545448084779148e-06,
+      "loss": 0.3953,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6696636180437141,
+      "grad_norm": 0.5316603779792786,
+      "learning_rate": 9.544320124258093e-06,
+      "loss": 0.4101,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6701286622229112,
+      "grad_norm": 0.502018928527832,
+      "learning_rate": 9.543190832779488e-06,
+      "loss": 0.394,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6705937064021082,
+      "grad_norm": 0.49598726630210876,
+      "learning_rate": 9.542060210674084e-06,
+      "loss": 0.4067,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6710587505813053,
+      "grad_norm": 0.5268638134002686,
+      "learning_rate": 9.540928258273021e-06,
+      "loss": 0.3724,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6715237947605023,
+      "grad_norm": 0.63950514793396,
+      "learning_rate": 9.539794975907831e-06,
+      "loss": 0.393,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6719888389396993,
+      "grad_norm": 0.45709770917892456,
+      "learning_rate": 9.538660363910433e-06,
+      "loss": 0.3779,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6724538831188963,
+      "grad_norm": 0.6071939468383789,
+      "learning_rate": 9.537524422613135e-06,
+      "loss": 0.3989,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6729189272980933,
+      "grad_norm": 0.5442386865615845,
+      "learning_rate": 9.53638715234864e-06,
+      "loss": 0.3666,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6733839714772903,
+      "grad_norm": 0.5082783102989197,
+      "learning_rate": 9.535248553450031e-06,
+      "loss": 0.4001,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6738490156564874,
+      "grad_norm": 0.5492592453956604,
+      "learning_rate": 9.53410862625079e-06,
+      "loss": 0.3681,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6743140598356844,
+      "grad_norm": 0.5500537157058716,
+      "learning_rate": 9.532967371084778e-06,
+      "loss": 0.4094,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6747791040148814,
+      "grad_norm": 0.528328001499176,
+      "learning_rate": 9.531824788286255e-06,
+      "loss": 0.4036,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6752441481940784,
+      "grad_norm": 0.49432677030563354,
+      "learning_rate": 9.53068087818986e-06,
+      "loss": 0.4007,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6757091923732754,
+      "grad_norm": 0.48525556921958923,
+      "learning_rate": 9.52953564113063e-06,
+      "loss": 0.3855,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6761742365524724,
+      "grad_norm": 0.5153108835220337,
+      "learning_rate": 9.528389077443985e-06,
+      "loss": 0.3863,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6766392807316695,
+      "grad_norm": 0.48874610662460327,
+      "learning_rate": 9.527241187465735e-06,
+      "loss": 0.3963,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6771043249108666,
+      "grad_norm": 0.5944570302963257,
+      "learning_rate": 9.526091971532075e-06,
+      "loss": 0.3698,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6775693690900636,
+      "grad_norm": 0.4809949994087219,
+      "learning_rate": 9.524941429979597e-06,
+      "loss": 0.3884,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6780344132692606,
+      "grad_norm": 0.5726730227470398,
+      "learning_rate": 9.523789563145274e-06,
+      "loss": 0.3955,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6784994574484576,
+      "grad_norm": 0.5116811394691467,
+      "learning_rate": 9.522636371366467e-06,
+      "loss": 0.3916,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6789645016276546,
+      "grad_norm": 0.4806547462940216,
+      "learning_rate": 9.521481854980928e-06,
+      "loss": 0.3909,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6794295458068517,
+      "grad_norm": 0.5462357997894287,
+      "learning_rate": 9.520326014326799e-06,
+      "loss": 0.3862,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6798945899860487,
+      "grad_norm": 0.44127699732780457,
+      "learning_rate": 9.519168849742603e-06,
+      "loss": 0.395,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6803596341652457,
+      "grad_norm": 0.5247107148170471,
+      "learning_rate": 9.518010361567259e-06,
+      "loss": 0.3771,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6808246783444427,
+      "grad_norm": 0.49513915181159973,
+      "learning_rate": 9.516850550140064e-06,
+      "loss": 0.3896,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6812897225236397,
+      "grad_norm": 0.4648475646972656,
+      "learning_rate": 9.515689415800713e-06,
+      "loss": 0.3866,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6817547667028367,
+      "grad_norm": 0.5785578489303589,
+      "learning_rate": 9.514526958889279e-06,
+      "loss": 0.4013,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6822198108820338,
+      "grad_norm": 0.5234367847442627,
+      "learning_rate": 9.51336317974623e-06,
+      "loss": 0.398,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6826848550612308,
+      "grad_norm": 0.5856658220291138,
+      "learning_rate": 9.512198078712417e-06,
+      "loss": 0.3752,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6831498992404278,
+      "grad_norm": 0.5868757367134094,
+      "learning_rate": 9.511031656129079e-06,
+      "loss": 0.3974,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6836149434196248,
+      "grad_norm": 0.5716127157211304,
+      "learning_rate": 9.509863912337843e-06,
+      "loss": 0.4253,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6840799875988219,
+      "grad_norm": 0.49978068470954895,
+      "learning_rate": 9.50869484768072e-06,
+      "loss": 0.4068,
+      "step": 1471
+    },
+    {
+      "epoch": 0.684545031778019,
+      "grad_norm": 0.5351032614707947,
+      "learning_rate": 9.507524462500112e-06,
+      "loss": 0.3921,
+      "step": 1472
+    },
+    {
+      "epoch": 0.685010075957216,
+      "grad_norm": 0.5618773102760315,
+      "learning_rate": 9.506352757138806e-06,
+      "loss": 0.3945,
+      "step": 1473
+    },
+    {
+      "epoch": 0.685475120136413,
+      "grad_norm": 0.6486148834228516,
+      "learning_rate": 9.505179731939975e-06,
+      "loss": 0.4109,
+      "step": 1474
+    },
+    {
+      "epoch": 0.68594016431561,
+      "grad_norm": 0.5250218510627747,
+      "learning_rate": 9.504005387247178e-06,
+      "loss": 0.413,
+      "step": 1475
+    },
+    {
+      "epoch": 0.686405208494807,
+      "grad_norm": 0.5226433277130127,
+      "learning_rate": 9.502829723404363e-06,
+      "loss": 0.3724,
+      "step": 1476
+    },
+    {
+      "epoch": 0.686870252674004,
+      "grad_norm": 0.5525023341178894,
+      "learning_rate": 9.50165274075586e-06,
+      "loss": 0.3952,
+      "step": 1477
+    },
+    {
+      "epoch": 0.687335296853201,
+      "grad_norm": 0.5429186820983887,
+      "learning_rate": 9.500474439646394e-06,
+      "loss": 0.3966,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6878003410323981,
+      "grad_norm": 0.47291281819343567,
+      "learning_rate": 9.499294820421064e-06,
+      "loss": 0.415,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6882653852115951,
+      "grad_norm": 0.6125563383102417,
+      "learning_rate": 9.498113883425364e-06,
+      "loss": 0.3928,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6887304293907921,
+      "grad_norm": 0.48280781507492065,
+      "learning_rate": 9.496931629005171e-06,
+      "loss": 0.3946,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6891954735699891,
+      "grad_norm": 0.5750291347503662,
+      "learning_rate": 9.49574805750675e-06,
+      "loss": 0.3947,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6896605177491861,
+      "grad_norm": 0.5800142288208008,
+      "learning_rate": 9.494563169276747e-06,
+      "loss": 0.4352,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6901255619283831,
+      "grad_norm": 0.469500869512558,
+      "learning_rate": 9.493376964662197e-06,
+      "loss": 0.3659,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6905906061075803,
+      "grad_norm": 0.5604742765426636,
+      "learning_rate": 9.492189444010522e-06,
+      "loss": 0.3868,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6910556502867773,
+      "grad_norm": 0.5544446706771851,
+      "learning_rate": 9.491000607669525e-06,
+      "loss": 0.4125,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6915206944659743,
+      "grad_norm": 0.5978662371635437,
+      "learning_rate": 9.489810455987398e-06,
+      "loss": 0.374,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6919857386451713,
+      "grad_norm": 0.4866108298301697,
+      "learning_rate": 9.488618989312719e-06,
+      "loss": 0.3797,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6924507828243683,
+      "grad_norm": 0.5635210275650024,
+      "learning_rate": 9.487426207994445e-06,
+      "loss": 0.3627,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6929158270035654,
+      "grad_norm": 0.534838855266571,
+      "learning_rate": 9.486232112381926e-06,
+      "loss": 0.397,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6933808711827624,
+      "grad_norm": 0.5099875926971436,
+      "learning_rate": 9.485036702824892e-06,
+      "loss": 0.397,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6938459153619594,
+      "grad_norm": 0.6172494888305664,
+      "learning_rate": 9.483839979673459e-06,
+      "loss": 0.4001,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6943109595411564,
+      "grad_norm": 0.5031498670578003,
+      "learning_rate": 9.482641943278127e-06,
+      "loss": 0.395,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6947760037203534,
+      "grad_norm": 0.5278390645980835,
+      "learning_rate": 9.481442593989781e-06,
+      "loss": 0.3882,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6952410478995504,
+      "grad_norm": 0.564165472984314,
+      "learning_rate": 9.480241932159692e-06,
+      "loss": 0.3947,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6957060920787475,
+      "grad_norm": 0.47055506706237793,
+      "learning_rate": 9.479039958139516e-06,
+      "loss": 0.3843,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6961711362579445,
+      "grad_norm": 0.6169893741607666,
+      "learning_rate": 9.477836672281291e-06,
+      "loss": 0.3824,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6966361804371415,
+      "grad_norm": 0.4662085473537445,
+      "learning_rate": 9.476632074937438e-06,
+      "loss": 0.4037,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6971012246163385,
+      "grad_norm": 0.5007343888282776,
+      "learning_rate": 9.475426166460763e-06,
+      "loss": 0.3552,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6975662687955356,
+      "grad_norm": 0.5537876486778259,
+      "learning_rate": 9.47421894720446e-06,
+      "loss": 0.3963,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6980313129747326,
+      "grad_norm": 0.49283310770988464,
+      "learning_rate": 9.473010417522104e-06,
+      "loss": 0.3937,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6984963571539297,
+      "grad_norm": 0.5946938395500183,
+      "learning_rate": 9.471800577767651e-06,
+      "loss": 0.3875,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6989614013331267,
+      "grad_norm": 0.4660334885120392,
+      "learning_rate": 9.470589428295444e-06,
+      "loss": 0.3862,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6994264455123237,
+      "grad_norm": 0.5642980337142944,
+      "learning_rate": 9.469376969460212e-06,
+      "loss": 0.3944,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6998914896915207,
+      "grad_norm": 0.5256946682929993,
+      "learning_rate": 9.468163201617063e-06,
+      "loss": 0.3991,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7003565338707177,
+      "grad_norm": 0.5281638503074646,
+      "learning_rate": 9.466948125121486e-06,
+      "loss": 0.4028,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7008215780499147,
+      "grad_norm": 0.5681607723236084,
+      "learning_rate": 9.465731740329364e-06,
+      "loss": 0.3888,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7012866222291118,
+      "grad_norm": 0.5455275177955627,
+      "learning_rate": 9.46451404759695e-06,
+      "loss": 0.3982,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7017516664083088,
+      "grad_norm": 0.49493181705474854,
+      "learning_rate": 9.463295047280892e-06,
+      "loss": 0.4045,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7022167105875058,
+      "grad_norm": 0.574228048324585,
+      "learning_rate": 9.462074739738212e-06,
+      "loss": 0.3897,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7026817547667028,
+      "grad_norm": 0.5371845364570618,
+      "learning_rate": 9.460853125326317e-06,
+      "loss": 0.3884,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7031467989458998,
+      "grad_norm": 0.5111741423606873,
+      "learning_rate": 9.459630204403001e-06,
+      "loss": 0.3557,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7036118431250968,
+      "grad_norm": 0.5982129573822021,
+      "learning_rate": 9.458405977326436e-06,
+      "loss": 0.405,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7040768873042939,
+      "grad_norm": 0.48484864830970764,
+      "learning_rate": 9.45718044445518e-06,
+      "loss": 0.3856,
+      "step": 1514
+    },
+    {
+      "epoch": 0.704541931483491,
+      "grad_norm": 0.4435729682445526,
+      "learning_rate": 9.455953606148172e-06,
+      "loss": 0.3944,
+      "step": 1515
+    },
+    {
+      "epoch": 0.705006975662688,
+      "grad_norm": 0.5365220308303833,
+      "learning_rate": 9.454725462764729e-06,
+      "loss": 0.3986,
+      "step": 1516
+    },
+    {
+      "epoch": 0.705472019841885,
+      "grad_norm": 0.5052123069763184,
+      "learning_rate": 9.453496014664557e-06,
+      "loss": 0.3782,
+      "step": 1517
+    },
+    {
+      "epoch": 0.705937064021082,
+      "grad_norm": 0.4664888381958008,
+      "learning_rate": 9.452265262207741e-06,
+      "loss": 0.3802,
+      "step": 1518
+    },
+    {
+      "epoch": 0.706402108200279,
+      "grad_norm": 0.5673739314079285,
+      "learning_rate": 9.451033205754749e-06,
+      "loss": 0.4167,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7068671523794761,
+      "grad_norm": 0.5380971431732178,
+      "learning_rate": 9.44979984566643e-06,
+      "loss": 0.3781,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7073321965586731,
+      "grad_norm": 0.4242795407772064,
+      "learning_rate": 9.448565182304015e-06,
+      "loss": 0.3824,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7077972407378701,
+      "grad_norm": 0.6041889786720276,
+      "learning_rate": 9.447329216029117e-06,
+      "loss": 0.394,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7082622849170671,
+      "grad_norm": 0.4233911633491516,
+      "learning_rate": 9.44609194720373e-06,
+      "loss": 0.3657,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7087273290962641,
+      "grad_norm": 0.542805552482605,
+      "learning_rate": 9.44485337619023e-06,
+      "loss": 0.3966,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7091923732754611,
+      "grad_norm": 0.5100137591362,
+      "learning_rate": 9.443613503351375e-06,
+      "loss": 0.3737,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7096574174546582,
+      "grad_norm": 0.4733894169330597,
+      "learning_rate": 9.442372329050304e-06,
+      "loss": 0.3776,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7101224616338552,
+      "grad_norm": 0.5760788917541504,
+      "learning_rate": 9.441129853650534e-06,
+      "loss": 0.3838,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7105875058130522,
+      "grad_norm": 0.5053048729896545,
+      "learning_rate": 9.43988607751597e-06,
+      "loss": 0.3741,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7110525499922493,
+      "grad_norm": 0.6125737428665161,
+      "learning_rate": 9.43864100101089e-06,
+      "loss": 0.3941,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7115175941714463,
+      "grad_norm": 0.6082503795623779,
+      "learning_rate": 9.437394624499957e-06,
+      "loss": 0.3769,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7119826383506433,
+      "grad_norm": 0.5114914178848267,
+      "learning_rate": 9.43614694834822e-06,
+      "loss": 0.3972,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7124476825298404,
+      "grad_norm": 0.5447007417678833,
+      "learning_rate": 9.434897972921095e-06,
+      "loss": 0.3758,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7129127267090374,
+      "grad_norm": 0.4863920509815216,
+      "learning_rate": 9.433647698584393e-06,
+      "loss": 0.385,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7133777708882344,
+      "grad_norm": 0.5778424739837646,
+      "learning_rate": 9.432396125704294e-06,
+      "loss": 0.3576,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7138428150674314,
+      "grad_norm": 0.541344940662384,
+      "learning_rate": 9.431143254647368e-06,
+      "loss": 0.4063,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7143078592466284,
+      "grad_norm": 0.5227439999580383,
+      "learning_rate": 9.429889085780559e-06,
+      "loss": 0.4045,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7147729034258254,
+      "grad_norm": 0.5893809795379639,
+      "learning_rate": 9.42863361947119e-06,
+      "loss": 0.3556,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7152379476050225,
+      "grad_norm": 0.45113706588745117,
+      "learning_rate": 9.42737685608697e-06,
+      "loss": 0.3984,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7157029917842195,
+      "grad_norm": 0.6279287338256836,
+      "learning_rate": 9.426118795995984e-06,
+      "loss": 0.368,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7161680359634165,
+      "grad_norm": 0.45758605003356934,
+      "learning_rate": 9.424859439566696e-06,
+      "loss": 0.378,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7166330801426135,
+      "grad_norm": 0.5116802453994751,
+      "learning_rate": 9.423598787167952e-06,
+      "loss": 0.3775,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7170981243218105,
+      "grad_norm": 0.5365654826164246,
+      "learning_rate": 9.422336839168974e-06,
+      "loss": 0.3701,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7175631685010075,
+      "grad_norm": 0.4998980760574341,
+      "learning_rate": 9.421073595939373e-06,
+      "loss": 0.3778,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7180282126802047,
+      "grad_norm": 0.5392967462539673,
+      "learning_rate": 9.419809057849125e-06,
+      "loss": 0.4003,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7184932568594017,
+      "grad_norm": 0.5534929633140564,
+      "learning_rate": 9.418543225268598e-06,
+      "loss": 0.3992,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7189583010385987,
+      "grad_norm": 0.45041725039482117,
+      "learning_rate": 9.41727609856853e-06,
+      "loss": 0.3951,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7194233452177957,
+      "grad_norm": 0.5060805082321167,
+      "learning_rate": 9.416007678120041e-06,
+      "loss": 0.3657,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7198883893969927,
+      "grad_norm": 0.4890674352645874,
+      "learning_rate": 9.414737964294636e-06,
+      "loss": 0.3763,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7203534335761898,
+      "grad_norm": 0.48126208782196045,
+      "learning_rate": 9.41346695746419e-06,
+      "loss": 0.403,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7208184777553868,
+      "grad_norm": 0.42710211873054504,
+      "learning_rate": 9.41219465800096e-06,
+      "loss": 0.408,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7212835219345838,
+      "grad_norm": 0.5024427771568298,
+      "learning_rate": 9.410921066277583e-06,
+      "loss": 0.3946,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7217485661137808,
+      "grad_norm": 0.5658032298088074,
+      "learning_rate": 9.409646182667073e-06,
+      "loss": 0.4095,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7222136102929778,
+      "grad_norm": 0.46872782707214355,
+      "learning_rate": 9.408370007542822e-06,
+      "loss": 0.3693,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7226786544721748,
+      "grad_norm": 0.5501477718353271,
+      "learning_rate": 9.407092541278602e-06,
+      "loss": 0.3939,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7231436986513718,
+      "grad_norm": 0.5214014649391174,
+      "learning_rate": 9.405813784248562e-06,
+      "loss": 0.3936,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7236087428305689,
+      "grad_norm": 0.47471073269844055,
+      "learning_rate": 9.40453373682723e-06,
+      "loss": 0.3869,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7240737870097659,
+      "grad_norm": 0.5597638487815857,
+      "learning_rate": 9.403252399389508e-06,
+      "loss": 0.3925,
+      "step": 1557
+    },
+    {
+      "epoch": 0.724538831188963,
+      "grad_norm": 0.43586617708206177,
+      "learning_rate": 9.401969772310681e-06,
+      "loss": 0.3636,
+      "step": 1558
+    },
+    {
+      "epoch": 0.72500387536816,
+      "grad_norm": 0.5254845023155212,
+      "learning_rate": 9.400685855966411e-06,
+      "loss": 0.3647,
+      "step": 1559
+    },
+    {
+      "epoch": 0.725468919547357,
+      "grad_norm": 0.5308626294136047,
+      "learning_rate": 9.399400650732735e-06,
+      "loss": 0.3843,
+      "step": 1560
+    },
+    {
+      "epoch": 0.725933963726554,
+      "grad_norm": 0.4976446032524109,
+      "learning_rate": 9.398114156986068e-06,
+      "loss": 0.4107,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7263990079057511,
+      "grad_norm": 0.47500959038734436,
+      "learning_rate": 9.396826375103203e-06,
+      "loss": 0.3745,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7268640520849481,
+      "grad_norm": 0.5488372445106506,
+      "learning_rate": 9.395537305461312e-06,
+      "loss": 0.3783,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7273290962641451,
+      "grad_norm": 0.48007380962371826,
+      "learning_rate": 9.394246948437943e-06,
+      "loss": 0.3835,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7277941404433421,
+      "grad_norm": 0.5635498762130737,
+      "learning_rate": 9.392955304411015e-06,
+      "loss": 0.38,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7282591846225391,
+      "grad_norm": 0.48347046971321106,
+      "learning_rate": 9.391662373758836e-06,
+      "loss": 0.4036,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7287242288017362,
+      "grad_norm": 0.448574423789978,
+      "learning_rate": 9.390368156860083e-06,
+      "loss": 0.3928,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7291892729809332,
+      "grad_norm": 0.4473694860935211,
+      "learning_rate": 9.389072654093809e-06,
+      "loss": 0.3579,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7296543171601302,
+      "grad_norm": 0.48732516169548035,
+      "learning_rate": 9.387775865839449e-06,
+      "loss": 0.3803,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7301193613393272,
+      "grad_norm": 0.46211230754852295,
+      "learning_rate": 9.386477792476806e-06,
+      "loss": 0.4012,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7305844055185242,
+      "grad_norm": 0.4076167643070221,
+      "learning_rate": 9.38517843438607e-06,
+      "loss": 0.3788,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7310494496977212,
+      "grad_norm": 0.5187703967094421,
+      "learning_rate": 9.383877791947802e-06,
+      "loss": 0.4111,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7315144938769184,
+      "grad_norm": 0.5166497230529785,
+      "learning_rate": 9.382575865542933e-06,
+      "loss": 0.3982,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7319795380561154,
+      "grad_norm": 0.4667831063270569,
+      "learning_rate": 9.38127265555278e-06,
+      "loss": 0.4002,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7324445822353124,
+      "grad_norm": 0.4515557289123535,
+      "learning_rate": 9.379968162359034e-06,
+      "loss": 0.3646,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7329096264145094,
+      "grad_norm": 0.48344147205352783,
+      "learning_rate": 9.378662386343758e-06,
+      "loss": 0.4046,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7333746705937064,
+      "grad_norm": 0.505967915058136,
+      "learning_rate": 9.377355327889391e-06,
+      "loss": 0.3664,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7338397147729034,
+      "grad_norm": 0.43122488260269165,
+      "learning_rate": 9.37604698737875e-06,
+      "loss": 0.3962,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7343047589521005,
+      "grad_norm": 0.4845030903816223,
+      "learning_rate": 9.374737365195028e-06,
+      "loss": 0.3937,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7347698031312975,
+      "grad_norm": 0.4910258948802948,
+      "learning_rate": 9.37342646172179e-06,
+      "loss": 0.3892,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7352348473104945,
+      "grad_norm": 0.4109004735946655,
+      "learning_rate": 9.372114277342981e-06,
+      "loss": 0.3735,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7356998914896915,
+      "grad_norm": 0.4791964888572693,
+      "learning_rate": 9.370800812442917e-06,
+      "loss": 0.3956,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7361649356688885,
+      "grad_norm": 0.4479603171348572,
+      "learning_rate": 9.36948606740629e-06,
+      "loss": 0.3587,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7366299798480855,
+      "grad_norm": 0.49391624331474304,
+      "learning_rate": 9.36817004261817e-06,
+      "loss": 0.3785,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7370950240272826,
+      "grad_norm": 0.4788481593132019,
+      "learning_rate": 9.366852738463995e-06,
+      "loss": 0.395,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7375600682064796,
+      "grad_norm": 0.4550861120223999,
+      "learning_rate": 9.365534155329585e-06,
+      "loss": 0.3826,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7380251123856766,
+      "grad_norm": 0.45728829503059387,
+      "learning_rate": 9.364214293601133e-06,
+      "loss": 0.3662,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7384901565648737,
+      "grad_norm": 0.46696847677230835,
+      "learning_rate": 9.3628931536652e-06,
+      "loss": 0.3751,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7389552007440707,
+      "grad_norm": 0.48484426736831665,
+      "learning_rate": 9.361570735908731e-06,
+      "loss": 0.3764,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7394202449232677,
+      "grad_norm": 0.4298027455806732,
+      "learning_rate": 9.36024704071904e-06,
+      "loss": 0.3879,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7398852891024648,
+      "grad_norm": 0.5281281471252441,
+      "learning_rate": 9.358922068483813e-06,
+      "loss": 0.3742,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7403503332816618,
+      "grad_norm": 0.39582934975624084,
+      "learning_rate": 9.357595819591116e-06,
+      "loss": 0.3886,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7408153774608588,
+      "grad_norm": 0.4835725724697113,
+      "learning_rate": 9.356268294429384e-06,
+      "loss": 0.3904,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7412804216400558,
+      "grad_norm": 0.49166139960289,
+      "learning_rate": 9.354939493387428e-06,
+      "loss": 0.4038,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7417454658192528,
+      "grad_norm": 0.4498700499534607,
+      "learning_rate": 9.353609416854432e-06,
+      "loss": 0.3906,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7422105099984498,
+      "grad_norm": 0.5919501781463623,
+      "learning_rate": 9.352278065219955e-06,
+      "loss": 0.3753,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7426755541776469,
+      "grad_norm": 0.5172655582427979,
+      "learning_rate": 9.350945438873927e-06,
+      "loss": 0.3858,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7431405983568439,
+      "grad_norm": 0.4497489333152771,
+      "learning_rate": 9.349611538206654e-06,
+      "loss": 0.3979,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7436056425360409,
+      "grad_norm": 0.4453744888305664,
+      "learning_rate": 9.348276363608812e-06,
+      "loss": 0.3854,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7440706867152379,
+      "grad_norm": 0.5212786793708801,
+      "learning_rate": 9.346939915471453e-06,
+      "loss": 0.3838,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7445357308944349,
+      "grad_norm": 0.5124901533126831,
+      "learning_rate": 9.345602194186001e-06,
+      "loss": 0.3859,
+      "step": 1601
+    },
+    {
+      "epoch": 0.745000775073632,
+      "grad_norm": 0.5541741847991943,
+      "learning_rate": 9.344263200144253e-06,
+      "loss": 0.3653,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7454658192528291,
+      "grad_norm": 0.5633660554885864,
+      "learning_rate": 9.342922933738377e-06,
+      "loss": 0.3784,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7459308634320261,
+      "grad_norm": 0.4769454300403595,
+      "learning_rate": 9.341581395360917e-06,
+      "loss": 0.389,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7463959076112231,
+      "grad_norm": 0.48664960265159607,
+      "learning_rate": 9.340238585404787e-06,
+      "loss": 0.3782,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7468609517904201,
+      "grad_norm": 0.45407751202583313,
+      "learning_rate": 9.338894504263276e-06,
+      "loss": 0.38,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7473259959696171,
+      "grad_norm": 0.47645387053489685,
+      "learning_rate": 9.33754915233004e-06,
+      "loss": 0.3541,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7477910401488141,
+      "grad_norm": 0.5349624752998352,
+      "learning_rate": 9.336202529999114e-06,
+      "loss": 0.3867,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7482560843280112,
+      "grad_norm": 0.49598512053489685,
+      "learning_rate": 9.3348546376649e-06,
+      "loss": 0.3848,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7487211285072082,
+      "grad_norm": 0.5270540714263916,
+      "learning_rate": 9.333505475722175e-06,
+      "loss": 0.3953,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7491861726864052,
+      "grad_norm": 0.4383714199066162,
+      "learning_rate": 9.332155044566085e-06,
+      "loss": 0.3936,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7496512168656022,
+      "grad_norm": 0.4920140206813812,
+      "learning_rate": 9.330803344592151e-06,
+      "loss": 0.3805,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7501162610447992,
+      "grad_norm": 0.5141205787658691,
+      "learning_rate": 9.329450376196264e-06,
+      "loss": 0.3821,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7505813052239962,
+      "grad_norm": 0.5176628828048706,
+      "learning_rate": 9.328096139774686e-06,
+      "loss": 0.3903,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7510463494031933,
+      "grad_norm": 0.5027486085891724,
+      "learning_rate": 9.326740635724047e-06,
+      "loss": 0.3814,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7515113935823903,
+      "grad_norm": 0.504041314125061,
+      "learning_rate": 9.32538386444136e-06,
+      "loss": 0.3701,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7519764377615874,
+      "grad_norm": 0.4461967647075653,
+      "learning_rate": 9.324025826323995e-06,
+      "loss": 0.3603,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7524414819407844,
+      "grad_norm": 0.4477754831314087,
+      "learning_rate": 9.3226665217697e-06,
+      "loss": 0.3585,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7529065261199814,
+      "grad_norm": 0.48952674865722656,
+      "learning_rate": 9.321305951176597e-06,
+      "loss": 0.3974,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7533715702991785,
+      "grad_norm": 0.4486616253852844,
+      "learning_rate": 9.319944114943171e-06,
+      "loss": 0.3745,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7538366144783755,
+      "grad_norm": 0.5282906293869019,
+      "learning_rate": 9.318581013468285e-06,
+      "loss": 0.3743,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7543016586575725,
+      "grad_norm": 0.5025021433830261,
+      "learning_rate": 9.317216647151166e-06,
+      "loss": 0.385,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7547667028367695,
+      "grad_norm": 0.48723548650741577,
+      "learning_rate": 9.315851016391417e-06,
+      "loss": 0.3891,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7552317470159665,
+      "grad_norm": 0.5467572212219238,
+      "learning_rate": 9.31448412158901e-06,
+      "loss": 0.3931,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7556967911951635,
+      "grad_norm": 0.503555178642273,
+      "learning_rate": 9.313115963144281e-06,
+      "loss": 0.3466,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7561618353743605,
+      "grad_norm": 0.570252537727356,
+      "learning_rate": 9.311746541457946e-06,
+      "loss": 0.3588,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7566268795535576,
+      "grad_norm": 0.49849367141723633,
+      "learning_rate": 9.310375856931086e-06,
+      "loss": 0.4005,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7570919237327546,
+      "grad_norm": 0.4657578766345978,
+      "learning_rate": 9.309003909965152e-06,
+      "loss": 0.3811,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7575569679119516,
+      "grad_norm": 0.5395064353942871,
+      "learning_rate": 9.307630700961966e-06,
+      "loss": 0.4136,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7580220120911486,
+      "grad_norm": 0.43396592140197754,
+      "learning_rate": 9.306256230323714e-06,
+      "loss": 0.3834,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7584870562703456,
+      "grad_norm": 0.4718068838119507,
+      "learning_rate": 9.304880498452962e-06,
+      "loss": 0.3759,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7589521004495428,
+      "grad_norm": 0.49737459421157837,
+      "learning_rate": 9.303503505752636e-06,
+      "loss": 0.3928,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7594171446287398,
+      "grad_norm": 0.4730997681617737,
+      "learning_rate": 9.302125252626035e-06,
+      "loss": 0.3916,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7598821888079368,
+      "grad_norm": 0.49301835894584656,
+      "learning_rate": 9.30074573947683e-06,
+      "loss": 0.3821,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7603472329871338,
+      "grad_norm": 0.5187510251998901,
+      "learning_rate": 9.299364966709051e-06,
+      "loss": 0.3527,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7608122771663308,
+      "grad_norm": 0.5241015553474426,
+      "learning_rate": 9.29798293472711e-06,
+      "loss": 0.378,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7612773213455278,
+      "grad_norm": 0.4729224443435669,
+      "learning_rate": 9.296599643935782e-06,
+      "loss": 0.4126,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7617423655247249,
+      "grad_norm": 0.6277045011520386,
+      "learning_rate": 9.295215094740208e-06,
+      "loss": 0.4036,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7622074097039219,
+      "grad_norm": 0.4564441740512848,
+      "learning_rate": 9.293829287545902e-06,
+      "loss": 0.3894,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7626724538831189,
+      "grad_norm": 0.48436111211776733,
+      "learning_rate": 9.292442222758741e-06,
+      "loss": 0.3695,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7631374980623159,
+      "grad_norm": 0.5133222341537476,
+      "learning_rate": 9.291053900784977e-06,
+      "loss": 0.3843,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7636025422415129,
+      "grad_norm": 0.4502391517162323,
+      "learning_rate": 9.289664322031225e-06,
+      "loss": 0.3745,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7640675864207099,
+      "grad_norm": 0.5892338156700134,
+      "learning_rate": 9.28827348690447e-06,
+      "loss": 0.3855,
+      "step": 1643
+    },
+    {
+      "epoch": 0.764532630599907,
+      "grad_norm": 0.4681336283683777,
+      "learning_rate": 9.286881395812066e-06,
+      "loss": 0.3817,
+      "step": 1644
+    },
+    {
+      "epoch": 0.764997674779104,
+      "grad_norm": 0.5281332731246948,
+      "learning_rate": 9.285488049161735e-06,
+      "loss": 0.3903,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7654627189583011,
+      "grad_norm": 0.5051922798156738,
+      "learning_rate": 9.284093447361563e-06,
+      "loss": 0.4035,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7659277631374981,
+      "grad_norm": 0.4733637273311615,
+      "learning_rate": 9.282697590820008e-06,
+      "loss": 0.3933,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7663928073166951,
+      "grad_norm": 0.43565672636032104,
+      "learning_rate": 9.281300479945894e-06,
+      "loss": 0.3654,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7668578514958921,
+      "grad_norm": 0.49669885635375977,
+      "learning_rate": 9.27990211514841e-06,
+      "loss": 0.3514,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7673228956750892,
+      "grad_norm": 0.4607134163379669,
+      "learning_rate": 9.278502496837116e-06,
+      "loss": 0.3954,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7677879398542862,
+      "grad_norm": 0.5022844672203064,
+      "learning_rate": 9.277101625421938e-06,
+      "loss": 0.3917,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7682529840334832,
+      "grad_norm": 0.5731377601623535,
+      "learning_rate": 9.275699501313164e-06,
+      "loss": 0.4014,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7687180282126802,
+      "grad_norm": 0.44675296545028687,
+      "learning_rate": 9.27429612492146e-06,
+      "loss": 0.3754,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7691830723918772,
+      "grad_norm": 0.6787543296813965,
+      "learning_rate": 9.27289149665785e-06,
+      "loss": 0.3779,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7696481165710742,
+      "grad_norm": 0.6024219989776611,
+      "learning_rate": 9.271485616933725e-06,
+      "loss": 0.3736,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7701131607502713,
+      "grad_norm": 0.49429091811180115,
+      "learning_rate": 9.270078486160843e-06,
+      "loss": 0.3944,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7705782049294683,
+      "grad_norm": 0.6208419799804688,
+      "learning_rate": 9.268670104751334e-06,
+      "loss": 0.3796,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7710432491086653,
+      "grad_norm": 0.5730566382408142,
+      "learning_rate": 9.267260473117687e-06,
+      "loss": 0.3936,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7715082932878623,
+      "grad_norm": 0.4064892828464508,
+      "learning_rate": 9.265849591672762e-06,
+      "loss": 0.3665,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7719733374670593,
+      "grad_norm": 0.6317027807235718,
+      "learning_rate": 9.264437460829783e-06,
+      "loss": 0.3973,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7724383816462564,
+      "grad_norm": 0.4885081350803375,
+      "learning_rate": 9.263024081002338e-06,
+      "loss": 0.401,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7729034258254535,
+      "grad_norm": 0.4672366976737976,
+      "learning_rate": 9.261609452604387e-06,
+      "loss": 0.3705,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7733684700046505,
+      "grad_norm": 0.5605355501174927,
+      "learning_rate": 9.260193576050247e-06,
+      "loss": 0.4115,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7738335141838475,
+      "grad_norm": 0.5084188580513,
+      "learning_rate": 9.25877645175461e-06,
+      "loss": 0.3905,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7742985583630445,
+      "grad_norm": 0.4235016107559204,
+      "learning_rate": 9.257358080132524e-06,
+      "loss": 0.3808,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7747636025422415,
+      "grad_norm": 0.5295476913452148,
+      "learning_rate": 9.25593846159941e-06,
+      "loss": 0.3704,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7752286467214385,
+      "grad_norm": 0.4431585669517517,
+      "learning_rate": 9.25451759657105e-06,
+      "loss": 0.3784,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7756936909006356,
+      "grad_norm": 0.43355417251586914,
+      "learning_rate": 9.253095485463594e-06,
+      "loss": 0.3708,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7761587350798326,
+      "grad_norm": 0.5399253368377686,
+      "learning_rate": 9.251672128693553e-06,
+      "loss": 0.3887,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7766237792590296,
+      "grad_norm": 0.47674787044525146,
+      "learning_rate": 9.250247526677806e-06,
+      "loss": 0.4136,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7770888234382266,
+      "grad_norm": 0.39808011054992676,
+      "learning_rate": 9.248821679833596e-06,
+      "loss": 0.3894,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7775538676174236,
+      "grad_norm": 0.5415627360343933,
+      "learning_rate": 9.24739458857853e-06,
+      "loss": 0.3698,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7780189117966206,
+      "grad_norm": 0.4286542534828186,
+      "learning_rate": 9.245966253330581e-06,
+      "loss": 0.3781,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7784839559758177,
+      "grad_norm": 0.42257702350616455,
+      "learning_rate": 9.244536674508085e-06,
+      "loss": 0.3868,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7789490001550148,
+      "grad_norm": 0.41875237226486206,
+      "learning_rate": 9.243105852529739e-06,
+      "loss": 0.3594,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7794140443342118,
+      "grad_norm": 0.45857277512550354,
+      "learning_rate": 9.241673787814612e-06,
+      "loss": 0.4015,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7798790885134088,
+      "grad_norm": 0.4585971236228943,
+      "learning_rate": 9.24024048078213e-06,
+      "loss": 0.3895,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7803441326926058,
+      "grad_norm": 0.4482724666595459,
+      "learning_rate": 9.238805931852088e-06,
+      "loss": 0.3787,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7808091768718028,
+      "grad_norm": 0.4379400312900543,
+      "learning_rate": 9.237370141444636e-06,
+      "loss": 0.3712,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7812742210509999,
+      "grad_norm": 0.452578604221344,
+      "learning_rate": 9.235933109980302e-06,
+      "loss": 0.4232,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7817392652301969,
+      "grad_norm": 0.4528006613254547,
+      "learning_rate": 9.234494837879963e-06,
+      "loss": 0.3626,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7822043094093939,
+      "grad_norm": 0.4689446985721588,
+      "learning_rate": 9.233055325564869e-06,
+      "loss": 0.386,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7826693535885909,
+      "grad_norm": 0.4909113645553589,
+      "learning_rate": 9.231614573456628e-06,
+      "loss": 0.3899,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7831343977677879,
+      "grad_norm": 0.45796748995780945,
+      "learning_rate": 9.230172581977212e-06,
+      "loss": 0.3907,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7835994419469849,
+      "grad_norm": 0.4643401801586151,
+      "learning_rate": 9.22872935154896e-06,
+      "loss": 0.3749,
+      "step": 1685
+    },
+    {
+      "epoch": 0.784064486126182,
+      "grad_norm": 0.5391439199447632,
+      "learning_rate": 9.227284882594567e-06,
+      "loss": 0.3901,
+      "step": 1686
+    },
+    {
+      "epoch": 0.784529530305379,
+      "grad_norm": 0.4336642920970917,
+      "learning_rate": 9.225839175537096e-06,
+      "loss": 0.3952,
+      "step": 1687
+    },
+    {
+      "epoch": 0.784994574484576,
+      "grad_norm": 0.4696858823299408,
+      "learning_rate": 9.224392230799972e-06,
+      "loss": 0.3925,
+      "step": 1688
+    },
+    {
+      "epoch": 0.785459618663773,
+      "grad_norm": 0.43825915455818176,
+      "learning_rate": 9.222944048806982e-06,
+      "loss": 0.3605,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7859246628429701,
+      "grad_norm": 0.5373666286468506,
+      "learning_rate": 9.221494629982274e-06,
+      "loss": 0.3634,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7863897070221672,
+      "grad_norm": 0.4807939827442169,
+      "learning_rate": 9.22004397475036e-06,
+      "loss": 0.3908,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7868547512013642,
+      "grad_norm": 0.462829053401947,
+      "learning_rate": 9.21859208353611e-06,
+      "loss": 0.351,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7873197953805612,
+      "grad_norm": 0.49336567521095276,
+      "learning_rate": 9.217138956764764e-06,
+      "loss": 0.3722,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7877848395597582,
+      "grad_norm": 0.5379830002784729,
+      "learning_rate": 9.215684594861915e-06,
+      "loss": 0.3967,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7882498837389552,
+      "grad_norm": 0.4929734766483307,
+      "learning_rate": 9.214228998253526e-06,
+      "loss": 0.3666,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7887149279181522,
+      "grad_norm": 0.5308467149734497,
+      "learning_rate": 9.212772167365915e-06,
+      "loss": 0.3863,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7891799720973492,
+      "grad_norm": 0.43540507555007935,
+      "learning_rate": 9.211314102625768e-06,
+      "loss": 0.4,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7896450162765463,
+      "grad_norm": 0.5050954222679138,
+      "learning_rate": 9.209854804460121e-06,
+      "loss": 0.3893,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7901100604557433,
+      "grad_norm": 0.5224504470825195,
+      "learning_rate": 9.208394273296387e-06,
+      "loss": 0.3957,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7905751046349403,
+      "grad_norm": 0.5353153944015503,
+      "learning_rate": 9.206932509562325e-06,
+      "loss": 0.3839,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7910401488141373,
+      "grad_norm": 0.5350932478904724,
+      "learning_rate": 9.205469513686065e-06,
+      "loss": 0.4252,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7915051929933343,
+      "grad_norm": 0.5694266557693481,
+      "learning_rate": 9.204005286096095e-06,
+      "loss": 0.3813,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7919702371725313,
+      "grad_norm": 0.5682549476623535,
+      "learning_rate": 9.202539827221264e-06,
+      "loss": 0.3519,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7924352813517284,
+      "grad_norm": 0.4279302656650543,
+      "learning_rate": 9.20107313749078e-06,
+      "loss": 0.3863,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7929003255309255,
+      "grad_norm": 0.5173167586326599,
+      "learning_rate": 9.19960521733421e-06,
+      "loss": 0.3747,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7933653697101225,
+      "grad_norm": 0.5256390571594238,
+      "learning_rate": 9.198136067181491e-06,
+      "loss": 0.3963,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7938304138893195,
+      "grad_norm": 0.5231375098228455,
+      "learning_rate": 9.196665687462906e-06,
+      "loss": 0.3706,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7942954580685165,
+      "grad_norm": 0.4618844985961914,
+      "learning_rate": 9.19519407860911e-06,
+      "loss": 0.3796,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7947605022477136,
+      "grad_norm": 0.49417373538017273,
+      "learning_rate": 9.193721241051108e-06,
+      "loss": 0.3847,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7952255464269106,
+      "grad_norm": 0.5576480031013489,
+      "learning_rate": 9.192247175220276e-06,
+      "loss": 0.3747,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7956905906061076,
+      "grad_norm": 0.5018484592437744,
+      "learning_rate": 9.190771881548343e-06,
+      "loss": 0.4008,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7961556347853046,
+      "grad_norm": 0.48867353796958923,
+      "learning_rate": 9.189295360467397e-06,
+      "loss": 0.3806,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7966206789645016,
+      "grad_norm": 0.5056695342063904,
+      "learning_rate": 9.187817612409886e-06,
+      "loss": 0.3718,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7970857231436986,
+      "grad_norm": 0.45721235871315,
+      "learning_rate": 9.18633863780862e-06,
+      "loss": 0.3794,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7975507673228956,
+      "grad_norm": 0.487956702709198,
+      "learning_rate": 9.184858437096766e-06,
+      "loss": 0.3905,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7980158115020927,
+      "grad_norm": 0.5465155243873596,
+      "learning_rate": 9.183377010707853e-06,
+      "loss": 0.3968,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7984808556812897,
+      "grad_norm": 0.5411269068717957,
+      "learning_rate": 9.181894359075763e-06,
+      "loss": 0.3826,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7989458998604867,
+      "grad_norm": 0.511201798915863,
+      "learning_rate": 9.180410482634744e-06,
+      "loss": 0.3873,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7994109440396838,
+      "grad_norm": 0.5773457288742065,
+      "learning_rate": 9.178925381819396e-06,
+      "loss": 0.4048,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7998759882188808,
+      "grad_norm": 0.5756568312644958,
+      "learning_rate": 9.177439057064684e-06,
+      "loss": 0.3772,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8003410323980779,
+      "grad_norm": 0.5717048048973083,
+      "learning_rate": 9.175951508805924e-06,
+      "loss": 0.3796,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8008060765772749,
+      "grad_norm": 0.5515778064727783,
+      "learning_rate": 9.174462737478801e-06,
+      "loss": 0.3781,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8012711207564719,
+      "grad_norm": 0.6067471504211426,
+      "learning_rate": 9.172972743519348e-06,
+      "loss": 0.3922,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8017361649356689,
+      "grad_norm": 0.4658670127391815,
+      "learning_rate": 9.17148152736396e-06,
+      "loss": 0.3948,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8022012091148659,
+      "grad_norm": 0.5393667817115784,
+      "learning_rate": 9.16998908944939e-06,
+      "loss": 0.3883,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8026662532940629,
+      "grad_norm": 0.5518912076950073,
+      "learning_rate": 9.168495430212752e-06,
+      "loss": 0.4006,
+      "step": 1726
+    },
+    {
+      "epoch": 0.80313129747326,
+      "grad_norm": 0.4799440801143646,
+      "learning_rate": 9.16700055009151e-06,
+      "loss": 0.3985,
+      "step": 1727
+    },
+    {
+      "epoch": 0.803596341652457,
+      "grad_norm": 0.48832663893699646,
+      "learning_rate": 9.165504449523492e-06,
+      "loss": 0.3831,
+      "step": 1728
+    },
+    {
+      "epoch": 0.804061385831654,
+      "grad_norm": 0.4758247435092926,
+      "learning_rate": 9.164007128946881e-06,
+      "loss": 0.3625,
+      "step": 1729
+    },
+    {
+      "epoch": 0.804526430010851,
+      "grad_norm": 0.4202597439289093,
+      "learning_rate": 9.162508588800221e-06,
+      "loss": 0.3416,
+      "step": 1730
+    },
+    {
+      "epoch": 0.804991474190048,
+      "grad_norm": 0.44846394658088684,
+      "learning_rate": 9.161008829522406e-06,
+      "loss": 0.3712,
+      "step": 1731
+    },
+    {
+      "epoch": 0.805456518369245,
+      "grad_norm": 0.4517316222190857,
+      "learning_rate": 9.159507851552693e-06,
+      "loss": 0.3893,
+      "step": 1732
+    },
+    {
+      "epoch": 0.805921562548442,
+      "grad_norm": 0.38894450664520264,
+      "learning_rate": 9.158005655330694e-06,
+      "loss": 0.3726,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8063866067276392,
+      "grad_norm": 0.45114031434059143,
+      "learning_rate": 9.156502241296376e-06,
+      "loss": 0.3862,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8068516509068362,
+      "grad_norm": 0.4183841943740845,
+      "learning_rate": 9.154997609890068e-06,
+      "loss": 0.3567,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8073166950860332,
+      "grad_norm": 0.4706588387489319,
+      "learning_rate": 9.15349176155245e-06,
+      "loss": 0.3892,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8077817392652302,
+      "grad_norm": 0.4675155282020569,
+      "learning_rate": 9.151984696724563e-06,
+      "loss": 0.4099,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8082467834444272,
+      "grad_norm": 0.46057257056236267,
+      "learning_rate": 9.150476415847797e-06,
+      "loss": 0.3915,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8087118276236243,
+      "grad_norm": 0.41681477427482605,
+      "learning_rate": 9.148966919363906e-06,
+      "loss": 0.3839,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8091768718028213,
+      "grad_norm": 0.4783117473125458,
+      "learning_rate": 9.147456207714998e-06,
+      "loss": 0.3541,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8096419159820183,
+      "grad_norm": 0.5050014853477478,
+      "learning_rate": 9.145944281343534e-06,
+      "loss": 0.36,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8101069601612153,
+      "grad_norm": 0.4405074715614319,
+      "learning_rate": 9.144431140692332e-06,
+      "loss": 0.4016,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8105720043404123,
+      "grad_norm": 0.41197729110717773,
+      "learning_rate": 9.142916786204568e-06,
+      "loss": 0.3787,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8110370485196093,
+      "grad_norm": 0.4868341386318207,
+      "learning_rate": 9.141401218323772e-06,
+      "loss": 0.3678,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8115020926988064,
+      "grad_norm": 0.4434308409690857,
+      "learning_rate": 9.139884437493828e-06,
+      "loss": 0.3633,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8119671368780034,
+      "grad_norm": 0.4891102612018585,
+      "learning_rate": 9.138366444158977e-06,
+      "loss": 0.3761,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8124321810572004,
+      "grad_norm": 0.5441362261772156,
+      "learning_rate": 9.136847238763814e-06,
+      "loss": 0.3794,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8128972252363974,
+      "grad_norm": 0.44566887617111206,
+      "learning_rate": 9.135326821753291e-06,
+      "loss": 0.3944,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8133622694155945,
+      "grad_norm": 0.4834020733833313,
+      "learning_rate": 9.133805193572713e-06,
+      "loss": 0.3902,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8138273135947915,
+      "grad_norm": 0.46305882930755615,
+      "learning_rate": 9.132282354667741e-06,
+      "loss": 0.3725,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8142923577739886,
+      "grad_norm": 0.4824235141277313,
+      "learning_rate": 9.130758305484387e-06,
+      "loss": 0.4076,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8147574019531856,
+      "grad_norm": 0.5479027032852173,
+      "learning_rate": 9.129233046469021e-06,
+      "loss": 0.3881,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8152224461323826,
+      "grad_norm": 0.42961257696151733,
+      "learning_rate": 9.127706578068369e-06,
+      "loss": 0.3726,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8156874903115796,
+      "grad_norm": 0.4997014105319977,
+      "learning_rate": 9.126178900729507e-06,
+      "loss": 0.3848,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8161525344907766,
+      "grad_norm": 0.5122492909431458,
+      "learning_rate": 9.124650014899868e-06,
+      "loss": 0.398,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8166175786699736,
+      "grad_norm": 0.5152801275253296,
+      "learning_rate": 9.123119921027234e-06,
+      "loss": 0.383,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8170826228491707,
+      "grad_norm": 0.4216311275959015,
+      "learning_rate": 9.121588619559752e-06,
+      "loss": 0.3729,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8175476670283677,
+      "grad_norm": 0.5480169057846069,
+      "learning_rate": 9.120056110945907e-06,
+      "loss": 0.4093,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8180127112075647,
+      "grad_norm": 0.6168447732925415,
+      "learning_rate": 9.118522395634552e-06,
+      "loss": 0.3907,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8184777553867617,
+      "grad_norm": 0.45844393968582153,
+      "learning_rate": 9.116987474074885e-06,
+      "loss": 0.3886,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8189427995659587,
+      "grad_norm": 0.543725311756134,
+      "learning_rate": 9.115451346716459e-06,
+      "loss": 0.3968,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8194078437451557,
+      "grad_norm": 0.48329514265060425,
+      "learning_rate": 9.113914014009182e-06,
+      "loss": 0.3837,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8198728879243529,
+      "grad_norm": 0.4946017265319824,
+      "learning_rate": 9.112375476403313e-06,
+      "loss": 0.3877,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8203379321035499,
+      "grad_norm": 0.5260226130485535,
+      "learning_rate": 9.110835734349464e-06,
+      "loss": 0.3735,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8208029762827469,
+      "grad_norm": 0.49095699191093445,
+      "learning_rate": 9.109294788298601e-06,
+      "loss": 0.3832,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8212680204619439,
+      "grad_norm": 0.5328912138938904,
+      "learning_rate": 9.107752638702046e-06,
+      "loss": 0.3704,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8217330646411409,
+      "grad_norm": 0.5146880745887756,
+      "learning_rate": 9.106209286011463e-06,
+      "loss": 0.3679,
+      "step": 1767
+    },
+    {
+      "epoch": 0.822198108820338,
+      "grad_norm": 0.6965378522872925,
+      "learning_rate": 9.104664730678878e-06,
+      "loss": 0.3939,
+      "step": 1768
+    },
+    {
+      "epoch": 0.822663152999535,
+      "grad_norm": 0.4589329659938812,
+      "learning_rate": 9.103118973156667e-06,
+      "loss": 0.3813,
+      "step": 1769
+    },
+    {
+      "epoch": 0.823128197178732,
+      "grad_norm": 0.4879869222640991,
+      "learning_rate": 9.101572013897555e-06,
+      "loss": 0.3608,
+      "step": 1770
+    },
+    {
+      "epoch": 0.823593241357929,
+      "grad_norm": 0.5170464515686035,
+      "learning_rate": 9.100023853354624e-06,
+      "loss": 0.3932,
+      "step": 1771
+    },
+    {
+      "epoch": 0.824058285537126,
+      "grad_norm": 0.48912927508354187,
+      "learning_rate": 9.098474491981305e-06,
+      "loss": 0.3588,
+      "step": 1772
+    },
+    {
+      "epoch": 0.824523329716323,
+      "grad_norm": 0.4621887803077698,
+      "learning_rate": 9.096923930231377e-06,
+      "loss": 0.3501,
+      "step": 1773
+    },
+    {
+      "epoch": 0.82498837389552,
+      "grad_norm": 0.6245998740196228,
+      "learning_rate": 9.095372168558977e-06,
+      "loss": 0.3937,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8254534180747171,
+      "grad_norm": 0.43584007024765015,
+      "learning_rate": 9.09381920741859e-06,
+      "loss": 0.3947,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8259184622539141,
+      "grad_norm": 0.5115625262260437,
+      "learning_rate": 9.092265047265057e-06,
+      "loss": 0.3878,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8263835064331111,
+      "grad_norm": 0.5149562954902649,
+      "learning_rate": 9.090709688553561e-06,
+      "loss": 0.4063,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8268485506123082,
+      "grad_norm": 0.4879224896430969,
+      "learning_rate": 9.089153131739642e-06,
+      "loss": 0.3646,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8273135947915052,
+      "grad_norm": 0.4819486737251282,
+      "learning_rate": 9.087595377279192e-06,
+      "loss": 0.3915,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8277786389707023,
+      "grad_norm": 0.5894208550453186,
+      "learning_rate": 9.086036425628453e-06,
+      "loss": 0.395,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8282436831498993,
+      "grad_norm": 0.45492544770240784,
+      "learning_rate": 9.084476277244013e-06,
+      "loss": 0.3976,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8287087273290963,
+      "grad_norm": 0.4946926236152649,
+      "learning_rate": 9.082914932582818e-06,
+      "loss": 0.3795,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8291737715082933,
+      "grad_norm": 0.5285559892654419,
+      "learning_rate": 9.081352392102159e-06,
+      "loss": 0.3766,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8296388156874903,
+      "grad_norm": 0.45280373096466064,
+      "learning_rate": 9.079788656259677e-06,
+      "loss": 0.3862,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8301038598666873,
+      "grad_norm": 0.5158466100692749,
+      "learning_rate": 9.078223725513366e-06,
+      "loss": 0.3877,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8305689040458843,
+      "grad_norm": 0.5172114968299866,
+      "learning_rate": 9.076657600321569e-06,
+      "loss": 0.395,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8310339482250814,
+      "grad_norm": 0.4534607529640198,
+      "learning_rate": 9.07509028114298e-06,
+      "loss": 0.3724,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8314989924042784,
+      "grad_norm": 0.45312753319740295,
+      "learning_rate": 9.073521768436638e-06,
+      "loss": 0.3771,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8319640365834754,
+      "grad_norm": 0.4360067844390869,
+      "learning_rate": 9.071952062661938e-06,
+      "loss": 0.3927,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8324290807626724,
+      "grad_norm": 0.4952300190925598,
+      "learning_rate": 9.070381164278622e-06,
+      "loss": 0.3784,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8328941249418694,
+      "grad_norm": 0.4036773145198822,
+      "learning_rate": 9.068809073746776e-06,
+      "loss": 0.3878,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8333591691210666,
+      "grad_norm": 0.4290243089199066,
+      "learning_rate": 9.067235791526844e-06,
+      "loss": 0.3467,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8338242133002636,
+      "grad_norm": 0.49693888425827026,
+      "learning_rate": 9.065661318079613e-06,
+      "loss": 0.3583,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8342892574794606,
+      "grad_norm": 0.5083548426628113,
+      "learning_rate": 9.064085653866222e-06,
+      "loss": 0.3691,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8347543016586576,
+      "grad_norm": 0.47161993384361267,
+      "learning_rate": 9.062508799348155e-06,
+      "loss": 0.3803,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8352193458378546,
+      "grad_norm": 0.4640964865684509,
+      "learning_rate": 9.06093075498725e-06,
+      "loss": 0.3636,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8356843900170516,
+      "grad_norm": 0.4891905188560486,
+      "learning_rate": 9.059351521245688e-06,
+      "loss": 0.3904,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8361494341962487,
+      "grad_norm": 0.5055262446403503,
+      "learning_rate": 9.057771098586003e-06,
+      "loss": 0.4009,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8366144783754457,
+      "grad_norm": 0.44761812686920166,
+      "learning_rate": 9.056189487471074e-06,
+      "loss": 0.3927,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8370795225546427,
+      "grad_norm": 0.49880966544151306,
+      "learning_rate": 9.05460668836413e-06,
+      "loss": 0.4054,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8375445667338397,
+      "grad_norm": 0.427062451839447,
+      "learning_rate": 9.053022701728744e-06,
+      "loss": 0.3694,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8380096109130367,
+      "grad_norm": 0.48414304852485657,
+      "learning_rate": 9.051437528028846e-06,
+      "loss": 0.3839,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8384746550922337,
+      "grad_norm": 0.5623500347137451,
+      "learning_rate": 9.049851167728702e-06,
+      "loss": 0.405,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8389396992714307,
+      "grad_norm": 0.4517265558242798,
+      "learning_rate": 9.048263621292934e-06,
+      "loss": 0.381,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8394047434506278,
+      "grad_norm": 0.6233772039413452,
+      "learning_rate": 9.046674889186509e-06,
+      "loss": 0.4002,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8398697876298248,
+      "grad_norm": 0.5412980914115906,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.3837,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8403348318090219,
+      "grad_norm": 0.6208202838897705,
+      "learning_rate": 9.043493869823283e-06,
+      "loss": 0.4094,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8407998759882189,
+      "grad_norm": 0.5141086578369141,
+      "learning_rate": 9.041901583498156e-06,
+      "loss": 0.4012,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8412649201674159,
+      "grad_norm": 0.579380452632904,
+      "learning_rate": 9.040308113365706e-06,
+      "loss": 0.4096,
+      "step": 1809
+    },
+    {
+      "epoch": 0.841729964346613,
+      "grad_norm": 0.5328314304351807,
+      "learning_rate": 9.038713459892637e-06,
+      "loss": 0.3613,
+      "step": 1810
+    },
+    {
+      "epoch": 0.84219500852581,
+      "grad_norm": 0.4614507853984833,
+      "learning_rate": 9.037117623545998e-06,
+      "loss": 0.4189,
+      "step": 1811
+    },
+    {
+      "epoch": 0.842660052705007,
+      "grad_norm": 0.6282251477241516,
+      "learning_rate": 9.035520604793183e-06,
+      "loss": 0.389,
+      "step": 1812
+    },
+    {
+      "epoch": 0.843125096884204,
+      "grad_norm": 0.4729471206665039,
+      "learning_rate": 9.03392240410193e-06,
+      "loss": 0.3953,
+      "step": 1813
+    },
+    {
+      "epoch": 0.843590141063401,
+      "grad_norm": 0.5759819149971008,
+      "learning_rate": 9.03232302194033e-06,
+      "loss": 0.3916,
+      "step": 1814
+    },
+    {
+      "epoch": 0.844055185242598,
+      "grad_norm": 0.5700781345367432,
+      "learning_rate": 9.030722458776815e-06,
+      "loss": 0.4019,
+      "step": 1815
+    },
+    {
+      "epoch": 0.844520229421795,
+      "grad_norm": 0.429929256439209,
+      "learning_rate": 9.029120715080162e-06,
+      "loss": 0.3942,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8449852736009921,
+      "grad_norm": 0.6056925654411316,
+      "learning_rate": 9.027517791319499e-06,
+      "loss": 0.3985,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8454503177801891,
+      "grad_norm": 0.5345470905303955,
+      "learning_rate": 9.025913687964293e-06,
+      "loss": 0.3734,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8459153619593861,
+      "grad_norm": 0.5462255477905273,
+      "learning_rate": 9.024308405484363e-06,
+      "loss": 0.3822,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8463804061385831,
+      "grad_norm": 0.5566556453704834,
+      "learning_rate": 9.022701944349867e-06,
+      "loss": 0.3878,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8468454503177801,
+      "grad_norm": 0.5047350525856018,
+      "learning_rate": 9.021094305031314e-06,
+      "loss": 0.3838,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8473104944969773,
+      "grad_norm": 0.5499956011772156,
+      "learning_rate": 9.019485487999553e-06,
+      "loss": 0.4,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8477755386761743,
+      "grad_norm": 0.49966034293174744,
+      "learning_rate": 9.017875493725783e-06,
+      "loss": 0.3779,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8482405828553713,
+      "grad_norm": 0.5475969910621643,
+      "learning_rate": 9.016264322681543e-06,
+      "loss": 0.3703,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8487056270345683,
+      "grad_norm": 0.5151212811470032,
+      "learning_rate": 9.01465197533872e-06,
+      "loss": 0.3803,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8491706712137653,
+      "grad_norm": 0.5074899196624756,
+      "learning_rate": 9.013038452169544e-06,
+      "loss": 0.3724,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8496357153929623,
+      "grad_norm": 0.49190855026245117,
+      "learning_rate": 9.01142375364659e-06,
+      "loss": 0.4028,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8501007595721594,
+      "grad_norm": 0.48248571157455444,
+      "learning_rate": 9.009807880242777e-06,
+      "loss": 0.3861,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8505658037513564,
+      "grad_norm": 0.5088501572608948,
+      "learning_rate": 9.008190832431367e-06,
+      "loss": 0.3925,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8510308479305534,
+      "grad_norm": 0.5075740218162537,
+      "learning_rate": 9.006572610685969e-06,
+      "loss": 0.4187,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8514958921097504,
+      "grad_norm": 0.6194763779640198,
+      "learning_rate": 9.004953215480532e-06,
+      "loss": 0.3997,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8519609362889474,
+      "grad_norm": 0.526176929473877,
+      "learning_rate": 9.003332647289351e-06,
+      "loss": 0.3969,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8524259804681444,
+      "grad_norm": 0.578385055065155,
+      "learning_rate": 9.001710906587064e-06,
+      "loss": 0.3834,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8528910246473415,
+      "grad_norm": 0.47598031163215637,
+      "learning_rate": 9.000087993848655e-06,
+      "loss": 0.3935,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8533560688265385,
+      "grad_norm": 0.4297819137573242,
+      "learning_rate": 8.998463909549445e-06,
+      "loss": 0.3873,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8538211130057356,
+      "grad_norm": 0.49305081367492676,
+      "learning_rate": 8.996838654165103e-06,
+      "loss": 0.3863,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8542861571849326,
+      "grad_norm": 0.5345546007156372,
+      "learning_rate": 8.99521222817164e-06,
+      "loss": 0.3716,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8547512013641296,
+      "grad_norm": 0.43835559487342834,
+      "learning_rate": 8.993584632045412e-06,
+      "loss": 0.3649,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8552162455433266,
+      "grad_norm": 0.5586456656455994,
+      "learning_rate": 8.991955866263112e-06,
+      "loss": 0.3804,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8556812897225237,
+      "grad_norm": 0.5465202331542969,
+      "learning_rate": 8.990325931301783e-06,
+      "loss": 0.3674,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8561463339017207,
+      "grad_norm": 0.40788722038269043,
+      "learning_rate": 8.988694827638803e-06,
+      "loss": 0.3849,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8566113780809177,
+      "grad_norm": 0.47962188720703125,
+      "learning_rate": 8.987062555751896e-06,
+      "loss": 0.3805,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8570764222601147,
+      "grad_norm": 0.4847969114780426,
+      "learning_rate": 8.985429116119132e-06,
+      "loss": 0.3694,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8575414664393117,
+      "grad_norm": 0.4671127200126648,
+      "learning_rate": 8.983794509218912e-06,
+      "loss": 0.3918,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8580065106185087,
+      "grad_norm": 0.473143070936203,
+      "learning_rate": 8.982158735529991e-06,
+      "loss": 0.382,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8584715547977058,
+      "grad_norm": 0.4718664586544037,
+      "learning_rate": 8.980521795531461e-06,
+      "loss": 0.3806,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8589365989769028,
+      "grad_norm": 0.4485674500465393,
+      "learning_rate": 8.97888368970275e-06,
+      "loss": 0.3918,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8594016431560998,
+      "grad_norm": 0.4322403073310852,
+      "learning_rate": 8.977244418523638e-06,
+      "loss": 0.3745,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8598666873352968,
+      "grad_norm": 0.44763296842575073,
+      "learning_rate": 8.97560398247424e-06,
+      "loss": 0.3607,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8603317315144938,
+      "grad_norm": 0.5826919674873352,
+      "learning_rate": 8.97396238203501e-06,
+      "loss": 0.395,
+      "step": 1850
+    },
+    {
+      "epoch": 0.860796775693691,
+      "grad_norm": 0.42212069034576416,
+      "learning_rate": 8.97231961768675e-06,
+      "loss": 0.3603,
+      "step": 1851
+    },
+    {
+      "epoch": 0.861261819872888,
+      "grad_norm": 0.49117767810821533,
+      "learning_rate": 8.970675689910596e-06,
+      "loss": 0.3789,
+      "step": 1852
+    },
+    {
+      "epoch": 0.861726864052085,
+      "grad_norm": 0.5760953426361084,
+      "learning_rate": 8.969030599188027e-06,
+      "loss": 0.3962,
+      "step": 1853
+    },
+    {
+      "epoch": 0.862191908231282,
+      "grad_norm": 0.4554479420185089,
+      "learning_rate": 8.967384346000866e-06,
+      "loss": 0.3854,
+      "step": 1854
+    },
+    {
+      "epoch": 0.862656952410479,
+      "grad_norm": 0.4693755507469177,
+      "learning_rate": 8.965736930831272e-06,
+      "loss": 0.383,
+      "step": 1855
+    },
+    {
+      "epoch": 0.863121996589676,
+      "grad_norm": 0.487821102142334,
+      "learning_rate": 8.964088354161748e-06,
+      "loss": 0.3819,
+      "step": 1856
+    },
+    {
+      "epoch": 0.863587040768873,
+      "grad_norm": 0.46892300248146057,
+      "learning_rate": 8.962438616475136e-06,
+      "loss": 0.3768,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8640520849480701,
+      "grad_norm": 0.44840800762176514,
+      "learning_rate": 8.960787718254615e-06,
+      "loss": 0.3779,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8645171291272671,
+      "grad_norm": 0.4931188225746155,
+      "learning_rate": 8.959135659983706e-06,
+      "loss": 0.3753,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8649821733064641,
+      "grad_norm": 0.4526351988315582,
+      "learning_rate": 8.957482442146271e-06,
+      "loss": 0.3664,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8654472174856611,
+      "grad_norm": 0.46007290482521057,
+      "learning_rate": 8.955828065226512e-06,
+      "loss": 0.3647,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8659122616648581,
+      "grad_norm": 0.4510989487171173,
+      "learning_rate": 8.954172529708967e-06,
+      "loss": 0.3839,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8663773058440551,
+      "grad_norm": 0.46228083968162537,
+      "learning_rate": 8.952515836078516e-06,
+      "loss": 0.3681,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8668423500232522,
+      "grad_norm": 0.5067192912101746,
+      "learning_rate": 8.950857984820378e-06,
+      "loss": 0.3969,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8673073942024492,
+      "grad_norm": 0.5053953528404236,
+      "learning_rate": 8.94919897642011e-06,
+      "loss": 0.3643,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8677724383816463,
+      "grad_norm": 0.5040225386619568,
+      "learning_rate": 8.947538811363612e-06,
+      "loss": 0.4229,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8682374825608433,
+      "grad_norm": 0.5020115971565247,
+      "learning_rate": 8.945877490137113e-06,
+      "loss": 0.3804,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8687025267400403,
+      "grad_norm": 0.518023669719696,
+      "learning_rate": 8.944215013227193e-06,
+      "loss": 0.4007,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8691675709192374,
+      "grad_norm": 0.4483552873134613,
+      "learning_rate": 8.942551381120763e-06,
+      "loss": 0.3953,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8696326150984344,
+      "grad_norm": 0.5058504939079285,
+      "learning_rate": 8.94088659430507e-06,
+      "loss": 0.3782,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8700976592776314,
+      "grad_norm": 0.46518880128860474,
+      "learning_rate": 8.939220653267708e-06,
+      "loss": 0.3827,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8705627034568284,
+      "grad_norm": 0.4913824200630188,
+      "learning_rate": 8.937553558496602e-06,
+      "loss": 0.3815,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8710277476360254,
+      "grad_norm": 0.5250646471977234,
+      "learning_rate": 8.935885310480018e-06,
+      "loss": 0.3706,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8714927918152224,
+      "grad_norm": 0.44793933629989624,
+      "learning_rate": 8.934215909706554e-06,
+      "loss": 0.3856,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8719578359944195,
+      "grad_norm": 0.5750781297683716,
+      "learning_rate": 8.932545356665157e-06,
+      "loss": 0.3723,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8724228801736165,
+      "grad_norm": 0.46956464648246765,
+      "learning_rate": 8.930873651845101e-06,
+      "loss": 0.3936,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8728879243528135,
+      "grad_norm": 0.4354320168495178,
+      "learning_rate": 8.929200795736003e-06,
+      "loss": 0.3627,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8733529685320105,
+      "grad_norm": 0.5778939127922058,
+      "learning_rate": 8.927526788827814e-06,
+      "loss": 0.3856,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8738180127112075,
+      "grad_norm": 0.4760781526565552,
+      "learning_rate": 8.925851631610825e-06,
+      "loss": 0.3778,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8742830568904046,
+      "grad_norm": 0.48736944794654846,
+      "learning_rate": 8.92417532457566e-06,
+      "loss": 0.3975,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8747481010696017,
+      "grad_norm": 0.4641439914703369,
+      "learning_rate": 8.922497868213284e-06,
+      "loss": 0.3756,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8752131452487987,
+      "grad_norm": 0.5082623362541199,
+      "learning_rate": 8.920819263014995e-06,
+      "loss": 0.3836,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8756781894279957,
+      "grad_norm": 0.5371342301368713,
+      "learning_rate": 8.919139509472433e-06,
+      "loss": 0.3863,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8761432336071927,
+      "grad_norm": 0.44863003492355347,
+      "learning_rate": 8.917458608077566e-06,
+      "loss": 0.3654,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8766082777863897,
+      "grad_norm": 0.48676854372024536,
+      "learning_rate": 8.915776559322704e-06,
+      "loss": 0.3949,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8770733219655867,
+      "grad_norm": 0.4752778708934784,
+      "learning_rate": 8.914093363700493e-06,
+      "loss": 0.3897,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8775383661447838,
+      "grad_norm": 0.4733431935310364,
+      "learning_rate": 8.912409021703914e-06,
+      "loss": 0.3728,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8780034103239808,
+      "grad_norm": 0.44890302419662476,
+      "learning_rate": 8.910723533826281e-06,
+      "loss": 0.3891,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8784684545031778,
+      "grad_norm": 0.574249804019928,
+      "learning_rate": 8.909036900561248e-06,
+      "loss": 0.3874,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8789334986823748,
+      "grad_norm": 0.5202791094779968,
+      "learning_rate": 8.907349122402803e-06,
+      "loss": 0.3842,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8793985428615718,
+      "grad_norm": 0.46931275725364685,
+      "learning_rate": 8.905660199845265e-06,
+      "loss": 0.3718,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8798635870407688,
+      "grad_norm": 0.5120000243186951,
+      "learning_rate": 8.903970133383297e-06,
+      "loss": 0.3927,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8803286312199659,
+      "grad_norm": 0.44593966007232666,
+      "learning_rate": 8.902278923511888e-06,
+      "loss": 0.388,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8807936753991629,
+      "grad_norm": 0.4737493097782135,
+      "learning_rate": 8.900586570726369e-06,
+      "loss": 0.3728,
+      "step": 1894
+    },
+    {
+      "epoch": 0.88125871957836,
+      "grad_norm": 0.4651558995246887,
+      "learning_rate": 8.8988930755224e-06,
+      "loss": 0.3921,
+      "step": 1895
+    },
+    {
+      "epoch": 0.881723763757557,
+      "grad_norm": 0.5006027817726135,
+      "learning_rate": 8.897198438395983e-06,
+      "loss": 0.3782,
+      "step": 1896
+    },
+    {
+      "epoch": 0.882188807936754,
+      "grad_norm": 0.444982647895813,
+      "learning_rate": 8.895502659843442e-06,
+      "loss": 0.3813,
+      "step": 1897
+    },
+    {
+      "epoch": 0.882653852115951,
+      "grad_norm": 0.4869557023048401,
+      "learning_rate": 8.89380574036145e-06,
+      "loss": 0.4084,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8831188962951481,
+      "grad_norm": 0.496622771024704,
+      "learning_rate": 8.892107680447005e-06,
+      "loss": 0.392,
+      "step": 1899
+    },
+    {
+      "epoch": 0.8835839404743451,
+      "grad_norm": 0.49721208214759827,
+      "learning_rate": 8.890408480597437e-06,
+      "loss": 0.3591,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8840489846535421,
+      "grad_norm": 0.5105478763580322,
+      "learning_rate": 8.88870814131042e-06,
+      "loss": 0.387,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8845140288327391,
+      "grad_norm": 0.5221246480941772,
+      "learning_rate": 8.887006663083952e-06,
+      "loss": 0.373,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8849790730119361,
+      "grad_norm": 0.4854063391685486,
+      "learning_rate": 8.885304046416369e-06,
+      "loss": 0.3778,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8854441171911331,
+      "grad_norm": 0.4377305209636688,
+      "learning_rate": 8.883600291806344e-06,
+      "loss": 0.4064,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8859091613703302,
+      "grad_norm": 0.4939655065536499,
+      "learning_rate": 8.881895399752873e-06,
+      "loss": 0.3839,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8863742055495272,
+      "grad_norm": 0.5449526906013489,
+      "learning_rate": 8.880189370755293e-06,
+      "loss": 0.3848,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8868392497287242,
+      "grad_norm": 0.46097972989082336,
+      "learning_rate": 8.878482205313275e-06,
+      "loss": 0.3829,
+      "step": 1907
+    },
+    {
+      "epoch": 0.8873042939079212,
+      "grad_norm": 0.5087864995002747,
+      "learning_rate": 8.876773903926816e-06,
+      "loss": 0.3827,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8877693380871183,
+      "grad_norm": 0.5424960255622864,
+      "learning_rate": 8.875064467096252e-06,
+      "loss": 0.37,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8882343822663153,
+      "grad_norm": 0.4373340904712677,
+      "learning_rate": 8.873353895322248e-06,
+      "loss": 0.3723,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8886994264455124,
+      "grad_norm": 0.5504875183105469,
+      "learning_rate": 8.871642189105804e-06,
+      "loss": 0.4028,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8891644706247094,
+      "grad_norm": 0.5143681168556213,
+      "learning_rate": 8.869929348948252e-06,
+      "loss": 0.3809,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8896295148039064,
+      "grad_norm": 0.4348513185977936,
+      "learning_rate": 8.868215375351251e-06,
+      "loss": 0.3806,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8900945589831034,
+      "grad_norm": 0.5529665350914001,
+      "learning_rate": 8.866500268816803e-06,
+      "loss": 0.3994,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8905596031623004,
+      "grad_norm": 0.5090189576148987,
+      "learning_rate": 8.864784029847227e-06,
+      "loss": 0.3922,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8910246473414974,
+      "grad_norm": 0.4400188624858856,
+      "learning_rate": 8.863066658945185e-06,
+      "loss": 0.3955,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8914896915206945,
+      "grad_norm": 0.5315966606140137,
+      "learning_rate": 8.861348156613667e-06,
+      "loss": 0.3723,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8919547356998915,
+      "grad_norm": 0.5052797198295593,
+      "learning_rate": 8.859628523355995e-06,
+      "loss": 0.3741,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8924197798790885,
+      "grad_norm": 0.4738040864467621,
+      "learning_rate": 8.857907759675822e-06,
+      "loss": 0.368,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8928848240582855,
+      "grad_norm": 0.5382641553878784,
+      "learning_rate": 8.85618586607713e-06,
+      "loss": 0.3906,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8933498682374825,
+      "grad_norm": 0.5306910276412964,
+      "learning_rate": 8.854462843064233e-06,
+      "loss": 0.3685,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8938149124166795,
+      "grad_norm": 0.4780616760253906,
+      "learning_rate": 8.85273869114178e-06,
+      "loss": 0.3719,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8942799565958766,
+      "grad_norm": 0.5241276621818542,
+      "learning_rate": 8.851013410814745e-06,
+      "loss": 0.3668,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8947450007750737,
+      "grad_norm": 0.547930896282196,
+      "learning_rate": 8.84928700258843e-06,
+      "loss": 0.3678,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8952100449542707,
+      "grad_norm": 0.5266174077987671,
+      "learning_rate": 8.847559466968482e-06,
+      "loss": 0.3808,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8956750891334677,
+      "grad_norm": 0.4367474317550659,
+      "learning_rate": 8.845830804460861e-06,
+      "loss": 0.3975,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8961401333126647,
+      "grad_norm": 0.4753339886665344,
+      "learning_rate": 8.844101015571867e-06,
+      "loss": 0.3894,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8966051774918617,
+      "grad_norm": 0.5285556316375732,
+      "learning_rate": 8.842370100808123e-06,
+      "loss": 0.3891,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8970702216710588,
+      "grad_norm": 0.48880597949028015,
+      "learning_rate": 8.84063806067659e-06,
+      "loss": 0.3949,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8975352658502558,
+      "grad_norm": 0.46564510464668274,
+      "learning_rate": 8.838904895684555e-06,
+      "loss": 0.4002,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8980003100294528,
+      "grad_norm": 0.4695758521556854,
+      "learning_rate": 8.837170606339628e-06,
+      "loss": 0.3763,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8984653542086498,
+      "grad_norm": 0.5208839178085327,
+      "learning_rate": 8.835435193149762e-06,
+      "loss": 0.3677,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8989303983878468,
+      "grad_norm": 0.5239862203598022,
+      "learning_rate": 8.833698656623227e-06,
+      "loss": 0.3733,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8993954425670438,
+      "grad_norm": 0.4721670150756836,
+      "learning_rate": 8.831960997268625e-06,
+      "loss": 0.3773,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8998604867462409,
+      "grad_norm": 0.5257996320724487,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.3535,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9003255309254379,
+      "grad_norm": 0.5932273268699646,
+      "learning_rate": 8.828482312111285e-06,
+      "loss": 0.3875,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9007905751046349,
+      "grad_norm": 0.4747033417224884,
+      "learning_rate": 8.826741287327396e-06,
+      "loss": 0.3726,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9012556192838319,
+      "grad_norm": 0.4879438877105713,
+      "learning_rate": 8.824999141753144e-06,
+      "loss": 0.3816,
+      "step": 1938
+    },
+    {
+      "epoch": 0.901720663463029,
+      "grad_norm": 0.5382803678512573,
+      "learning_rate": 8.82325587589877e-06,
+      "loss": 0.4059,
+      "step": 1939
+    },
+    {
+      "epoch": 0.902185707642226,
+      "grad_norm": 0.5141168236732483,
+      "learning_rate": 8.821511490274854e-06,
+      "loss": 0.3807,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9026507518214231,
+      "grad_norm": 0.5135001540184021,
+      "learning_rate": 8.819765985392297e-06,
+      "loss": 0.4041,
+      "step": 1941
+    },
+    {
+      "epoch": 0.9031157960006201,
+      "grad_norm": 0.5456048250198364,
+      "learning_rate": 8.818019361762325e-06,
+      "loss": 0.4146,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9035808401798171,
+      "grad_norm": 0.5352010726928711,
+      "learning_rate": 8.816271619896502e-06,
+      "loss": 0.385,
+      "step": 1943
+    },
+    {
+      "epoch": 0.9040458843590141,
+      "grad_norm": 0.43585771322250366,
+      "learning_rate": 8.814522760306708e-06,
+      "loss": 0.3692,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9045109285382111,
+      "grad_norm": 0.47126439213752747,
+      "learning_rate": 8.812772783505158e-06,
+      "loss": 0.3653,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9049759727174082,
+      "grad_norm": 0.5107633471488953,
+      "learning_rate": 8.811021690004389e-06,
+      "loss": 0.3952,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9054410168966052,
+      "grad_norm": 0.482185423374176,
+      "learning_rate": 8.80926948031727e-06,
+      "loss": 0.3861,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9059060610758022,
+      "grad_norm": 0.4702143669128418,
+      "learning_rate": 8.807516154956997e-06,
+      "loss": 0.3701,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9063711052549992,
+      "grad_norm": 0.5154238343238831,
+      "learning_rate": 8.80576171443709e-06,
+      "loss": 0.3809,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9068361494341962,
+      "grad_norm": 0.49696359038352966,
+      "learning_rate": 8.80400615927139e-06,
+      "loss": 0.3743,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9073011936133932,
+      "grad_norm": 0.49918052554130554,
+      "learning_rate": 8.802249489974078e-06,
+      "loss": 0.4045,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9077662377925902,
+      "grad_norm": 0.45246899127960205,
+      "learning_rate": 8.80049170705965e-06,
+      "loss": 0.3693,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9082312819717874,
+      "grad_norm": 0.4851575791835785,
+      "learning_rate": 8.798732811042934e-06,
+      "loss": 0.4069,
+      "step": 1953
+    },
+    {
+      "epoch": 0.9086963261509844,
+      "grad_norm": 0.5256077647209167,
+      "learning_rate": 8.796972802439079e-06,
+      "loss": 0.3695,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9091613703301814,
+      "grad_norm": 0.4606427848339081,
+      "learning_rate": 8.795211681763565e-06,
+      "loss": 0.3749,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9096264145093784,
+      "grad_norm": 0.4951748847961426,
+      "learning_rate": 8.793449449532198e-06,
+      "loss": 0.4129,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9100914586885754,
+      "grad_norm": 0.4839325547218323,
+      "learning_rate": 8.791686106261104e-06,
+      "loss": 0.3883,
+      "step": 1957
+    },
+    {
+      "epoch": 0.9105565028677725,
+      "grad_norm": 0.4781914949417114,
+      "learning_rate": 8.789921652466738e-06,
+      "loss": 0.3883,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9110215470469695,
+      "grad_norm": 0.4980541467666626,
+      "learning_rate": 8.78815608866588e-06,
+      "loss": 0.3689,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9114865912261665,
+      "grad_norm": 0.523271381855011,
+      "learning_rate": 8.786389415375636e-06,
+      "loss": 0.3942,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9119516354053635,
+      "grad_norm": 0.5033456683158875,
+      "learning_rate": 8.784621633113434e-06,
+      "loss": 0.3624,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9124166795845605,
+      "grad_norm": 0.441829651594162,
+      "learning_rate": 8.78285274239703e-06,
+      "loss": 0.371,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9128817237637575,
+      "grad_norm": 0.49837011098861694,
+      "learning_rate": 8.781082743744505e-06,
+      "loss": 0.4078,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9133467679429546,
+      "grad_norm": 0.46190333366394043,
+      "learning_rate": 8.779311637674259e-06,
+      "loss": 0.397,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9138118121221516,
+      "grad_norm": 0.5549447536468506,
+      "learning_rate": 8.777539424705022e-06,
+      "loss": 0.3938,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9142768563013486,
+      "grad_norm": 0.4479234516620636,
+      "learning_rate": 8.775766105355849e-06,
+      "loss": 0.3767,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9147419004805456,
+      "grad_norm": 0.4744507372379303,
+      "learning_rate": 8.773991680146113e-06,
+      "loss": 0.3853,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9152069446597427,
+      "grad_norm": 0.4301864206790924,
+      "learning_rate": 8.772216149595515e-06,
+      "loss": 0.3836,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9156719888389397,
+      "grad_norm": 0.4343145489692688,
+      "learning_rate": 8.77043951422408e-06,
+      "loss": 0.3789,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9161370330181368,
+      "grad_norm": 0.43467792868614197,
+      "learning_rate": 8.768661774552155e-06,
+      "loss": 0.3736,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9166020771973338,
+      "grad_norm": 0.43103089928627014,
+      "learning_rate": 8.766882931100411e-06,
+      "loss": 0.3627,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9170671213765308,
+      "grad_norm": 0.43862855434417725,
+      "learning_rate": 8.765102984389842e-06,
+      "loss": 0.3811,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9175321655557278,
+      "grad_norm": 0.48002341389656067,
+      "learning_rate": 8.763321934941766e-06,
+      "loss": 0.4119,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9179972097349248,
+      "grad_norm": 0.43228980898857117,
+      "learning_rate": 8.761539783277825e-06,
+      "loss": 0.3808,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9184622539141218,
+      "grad_norm": 0.43084922432899475,
+      "learning_rate": 8.75975652991998e-06,
+      "loss": 0.3998,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9189272980933189,
+      "grad_norm": 0.4248911738395691,
+      "learning_rate": 8.757972175390516e-06,
+      "loss": 0.4031,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9193923422725159,
+      "grad_norm": 0.4406249523162842,
+      "learning_rate": 8.756186720212045e-06,
+      "loss": 0.3819,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9198573864517129,
+      "grad_norm": 0.4917864501476288,
+      "learning_rate": 8.754400164907496e-06,
+      "loss": 0.3958,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9203224306309099,
+      "grad_norm": 0.44568273425102234,
+      "learning_rate": 8.752612510000123e-06,
+      "loss": 0.3707,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9207874748101069,
+      "grad_norm": 0.5774033069610596,
+      "learning_rate": 8.750823756013498e-06,
+      "loss": 0.3855,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9212525189893039,
+      "grad_norm": 0.4670921564102173,
+      "learning_rate": 8.749033903471522e-06,
+      "loss": 0.4012,
+      "step": 1981
+    },
+    {
+      "epoch": 0.921717563168501,
+      "grad_norm": 0.45798271894454956,
+      "learning_rate": 8.74724295289841e-06,
+      "loss": 0.3566,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9221826073476981,
+      "grad_norm": 0.47769033908843994,
+      "learning_rate": 8.745450904818705e-06,
+      "loss": 0.3623,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9226476515268951,
+      "grad_norm": 0.42745545506477356,
+      "learning_rate": 8.743657759757267e-06,
+      "loss": 0.3977,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9231126957060921,
+      "grad_norm": 0.4700676500797272,
+      "learning_rate": 8.741863518239283e-06,
+      "loss": 0.3815,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9235777398852891,
+      "grad_norm": 0.4442138671875,
+      "learning_rate": 8.740068180790252e-06,
+      "loss": 0.3957,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9240427840644861,
+      "grad_norm": 0.42371588945388794,
+      "learning_rate": 8.738271747936001e-06,
+      "loss": 0.3715,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9245078282436832,
+      "grad_norm": 0.476406067609787,
+      "learning_rate": 8.736474220202675e-06,
+      "loss": 0.3946,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9249728724228802,
+      "grad_norm": 0.4362446367740631,
+      "learning_rate": 8.734675598116743e-06,
+      "loss": 0.3665,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9254379166020772,
+      "grad_norm": 0.4169759154319763,
+      "learning_rate": 8.732875882204993e-06,
+      "loss": 0.3747,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9259029607812742,
+      "grad_norm": 0.4496033191680908,
+      "learning_rate": 8.73107507299453e-06,
+      "loss": 0.3798,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9263680049604712,
+      "grad_norm": 0.47536760568618774,
+      "learning_rate": 8.729273171012782e-06,
+      "loss": 0.3914,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9268330491396682,
+      "grad_norm": 0.4569942355155945,
+      "learning_rate": 8.727470176787498e-06,
+      "loss": 0.3665,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9272980933188653,
+      "grad_norm": 0.46827104687690735,
+      "learning_rate": 8.725666090846746e-06,
+      "loss": 0.3673,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9277631374980623,
+      "grad_norm": 0.5109233856201172,
+      "learning_rate": 8.72386091371891e-06,
+      "loss": 0.3717,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9282281816772593,
+      "grad_norm": 0.4976223409175873,
+      "learning_rate": 8.7220546459327e-06,
+      "loss": 0.3748,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9286932258564564,
+      "grad_norm": 0.42419055104255676,
+      "learning_rate": 8.720247288017143e-06,
+      "loss": 0.3478,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9291582700356534,
+      "grad_norm": 0.5530095100402832,
+      "learning_rate": 8.718438840501585e-06,
+      "loss": 0.3746,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9296233142148504,
+      "grad_norm": 0.44324877858161926,
+      "learning_rate": 8.716629303915689e-06,
+      "loss": 0.3711,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9300883583940475,
+      "grad_norm": 0.4611654281616211,
+      "learning_rate": 8.71481867878944e-06,
+      "loss": 0.3642,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9305534025732445,
+      "grad_norm": 0.5441577434539795,
+      "learning_rate": 8.71300696565314e-06,
+      "loss": 0.373,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9310184467524415,
+      "grad_norm": 0.5150545239448547,
+      "learning_rate": 8.71119416503741e-06,
+      "loss": 0.366,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9314834909316385,
+      "grad_norm": 0.5382184386253357,
+      "learning_rate": 8.709380277473191e-06,
+      "loss": 0.3885,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9319485351108355,
+      "grad_norm": 0.5589922070503235,
+      "learning_rate": 8.707565303491741e-06,
+      "loss": 0.3871,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9324135792900325,
+      "grad_norm": 0.5470889210700989,
+      "learning_rate": 8.705749243624635e-06,
+      "loss": 0.4045,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9328786234692296,
+      "grad_norm": 0.5791128873825073,
+      "learning_rate": 8.70393209840377e-06,
+      "loss": 0.3931,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9333436676484266,
+      "grad_norm": 0.4637211859226227,
+      "learning_rate": 8.702113868361357e-06,
+      "loss": 0.3823,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9338087118276236,
+      "grad_norm": 0.5710762143135071,
+      "learning_rate": 8.700294554029926e-06,
+      "loss": 0.381,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9342737560068206,
+      "grad_norm": 0.454738050699234,
+      "learning_rate": 8.698474155942325e-06,
+      "loss": 0.3765,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9347388001860176,
+      "grad_norm": 0.5094452500343323,
+      "learning_rate": 8.696652674631716e-06,
+      "loss": 0.4076,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9352038443652146,
+      "grad_norm": 0.46923717856407166,
+      "learning_rate": 8.694830110631587e-06,
+      "loss": 0.3796,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9356688885444118,
+      "grad_norm": 0.5491566061973572,
+      "learning_rate": 8.693006464475732e-06,
+      "loss": 0.3716,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9361339327236088,
+      "grad_norm": 0.5266678929328918,
+      "learning_rate": 8.691181736698272e-06,
+      "loss": 0.3616,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9365989769028058,
+      "grad_norm": 0.4861200153827667,
+      "learning_rate": 8.689355927833636e-06,
+      "loss": 0.3638,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9370640210820028,
+      "grad_norm": 0.5367874503135681,
+      "learning_rate": 8.687529038416575e-06,
+      "loss": 0.3787,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9375290652611998,
+      "grad_norm": 0.47685980796813965,
+      "learning_rate": 8.685701068982158e-06,
+      "loss": 0.3904,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9379941094403969,
+      "grad_norm": 0.4525419771671295,
+      "learning_rate": 8.683872020065763e-06,
+      "loss": 0.4035,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9384591536195939,
+      "grad_norm": 0.45574548840522766,
+      "learning_rate": 8.682041892203093e-06,
+      "loss": 0.4004,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9389241977987909,
+      "grad_norm": 0.4963834285736084,
+      "learning_rate": 8.68021068593016e-06,
+      "loss": 0.3706,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9393892419779879,
+      "grad_norm": 0.46662190556526184,
+      "learning_rate": 8.678378401783293e-06,
+      "loss": 0.4034,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9398542861571849,
+      "grad_norm": 0.45267191529273987,
+      "learning_rate": 8.676545040299145e-06,
+      "loss": 0.3866,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9403193303363819,
+      "grad_norm": 0.4719865620136261,
+      "learning_rate": 8.674710602014672e-06,
+      "loss": 0.3831,
+      "step": 2022
+    },
+    {
+      "epoch": 0.940784374515579,
+      "grad_norm": 0.42862021923065186,
+      "learning_rate": 8.67287508746715e-06,
+      "loss": 0.3466,
+      "step": 2023
+    },
+    {
+      "epoch": 0.941249418694776,
+      "grad_norm": 0.48866480588912964,
+      "learning_rate": 8.671038497194175e-06,
+      "loss": 0.3776,
+      "step": 2024
+    },
+    {
+      "epoch": 0.941714462873973,
+      "grad_norm": 0.4378620684146881,
+      "learning_rate": 8.669200831733655e-06,
+      "loss": 0.3899,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9421795070531701,
+      "grad_norm": 0.45231664180755615,
+      "learning_rate": 8.66736209162381e-06,
+      "loss": 0.3469,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9426445512323671,
+      "grad_norm": 0.560242235660553,
+      "learning_rate": 8.665522277403177e-06,
+      "loss": 0.3633,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9431095954115641,
+      "grad_norm": 0.431618332862854,
+      "learning_rate": 8.66368138961061e-06,
+      "loss": 0.3988,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9435746395907612,
+      "grad_norm": 0.46893802285194397,
+      "learning_rate": 8.661839428785273e-06,
+      "loss": 0.3829,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9440396837699582,
+      "grad_norm": 0.506652295589447,
+      "learning_rate": 8.659996395466648e-06,
+      "loss": 0.3853,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9445047279491552,
+      "grad_norm": 0.5342637896537781,
+      "learning_rate": 8.658152290194526e-06,
+      "loss": 0.4173,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9449697721283522,
+      "grad_norm": 0.43193793296813965,
+      "learning_rate": 8.656307113509021e-06,
+      "loss": 0.367,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9454348163075492,
+      "grad_norm": 0.43902087211608887,
+      "learning_rate": 8.654460865950551e-06,
+      "loss": 0.3629,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9458998604867462,
+      "grad_norm": 0.5017895698547363,
+      "learning_rate": 8.652613548059854e-06,
+      "loss": 0.3706,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9463649046659433,
+      "grad_norm": 0.45130717754364014,
+      "learning_rate": 8.650765160377978e-06,
+      "loss": 0.3811,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9468299488451403,
+      "grad_norm": 0.4452032744884491,
+      "learning_rate": 8.648915703446287e-06,
+      "loss": 0.3904,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9472949930243373,
+      "grad_norm": 0.5498278141021729,
+      "learning_rate": 8.647065177806457e-06,
+      "loss": 0.3611,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9477600372035343,
+      "grad_norm": 0.41174963116645813,
+      "learning_rate": 8.645213584000476e-06,
+      "loss": 0.3793,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9482250813827313,
+      "grad_norm": 0.5462541580200195,
+      "learning_rate": 8.643360922570646e-06,
+      "loss": 0.3601,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9486901255619283,
+      "grad_norm": 0.5303115844726562,
+      "learning_rate": 8.64150719405958e-06,
+      "loss": 0.3918,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9491551697411255,
+      "grad_norm": 0.45511001348495483,
+      "learning_rate": 8.639652399010208e-06,
+      "loss": 0.4027,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9496202139203225,
+      "grad_norm": 0.5173806548118591,
+      "learning_rate": 8.637796537965768e-06,
+      "loss": 0.372,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9500852580995195,
+      "grad_norm": 0.45395195484161377,
+      "learning_rate": 8.63593961146981e-06,
+      "loss": 0.3794,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9505503022787165,
+      "grad_norm": 0.5278088450431824,
+      "learning_rate": 8.634081620066199e-06,
+      "loss": 0.388,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9510153464579135,
+      "grad_norm": 0.4131206274032593,
+      "learning_rate": 8.632222564299111e-06,
+      "loss": 0.3784,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9514803906371105,
+      "grad_norm": 0.4544680118560791,
+      "learning_rate": 8.630362444713033e-06,
+      "loss": 0.3816,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9519454348163076,
+      "grad_norm": 0.534386396408081,
+      "learning_rate": 8.628501261852765e-06,
+      "loss": 0.3543,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9524104789955046,
+      "grad_norm": 0.44800546765327454,
+      "learning_rate": 8.626639016263413e-06,
+      "loss": 0.3832,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9528755231747016,
+      "grad_norm": 0.5088453888893127,
+      "learning_rate": 8.624775708490403e-06,
+      "loss": 0.3755,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9533405673538986,
+      "grad_norm": 0.5941973328590393,
+      "learning_rate": 8.622911339079464e-06,
+      "loss": 0.3948,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9538056115330956,
+      "grad_norm": 0.45851653814315796,
+      "learning_rate": 8.621045908576642e-06,
+      "loss": 0.3968,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9542706557122926,
+      "grad_norm": 0.5592479109764099,
+      "learning_rate": 8.619179417528293e-06,
+      "loss": 0.3982,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9547356998914897,
+      "grad_norm": 0.5447689294815063,
+      "learning_rate": 8.617311866481076e-06,
+      "loss": 0.3877,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9552007440706867,
+      "grad_norm": 0.4597291052341461,
+      "learning_rate": 8.61544325598197e-06,
+      "loss": 0.3644,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9556657882498837,
+      "grad_norm": 0.5028977394104004,
+      "learning_rate": 8.613573586578262e-06,
+      "loss": 0.3604,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9561308324290808,
+      "grad_norm": 0.5437200665473938,
+      "learning_rate": 8.611702858817545e-06,
+      "loss": 0.3916,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9565958766082778,
+      "grad_norm": 0.4929006099700928,
+      "learning_rate": 8.609831073247728e-06,
+      "loss": 0.3633,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9570609207874748,
+      "grad_norm": 0.4636727273464203,
+      "learning_rate": 8.607958230417024e-06,
+      "loss": 0.3587,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9575259649666719,
+      "grad_norm": 0.39754369854927063,
+      "learning_rate": 8.606084330873958e-06,
+      "loss": 0.39,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9579910091458689,
+      "grad_norm": 0.47713038325309753,
+      "learning_rate": 8.604209375167366e-06,
+      "loss": 0.3848,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9584560533250659,
+      "grad_norm": 0.4594153165817261,
+      "learning_rate": 8.602333363846393e-06,
+      "loss": 0.3805,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9589210975042629,
+      "grad_norm": 0.4612623453140259,
+      "learning_rate": 8.600456297460491e-06,
+      "loss": 0.3809,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9593861416834599,
+      "grad_norm": 0.4625248312950134,
+      "learning_rate": 8.598578176559423e-06,
+      "loss": 0.3529,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9598511858626569,
+      "grad_norm": 0.4562045633792877,
+      "learning_rate": 8.596699001693257e-06,
+      "loss": 0.367,
+      "step": 2064
+    },
+    {
+      "epoch": 0.960316230041854,
+      "grad_norm": 0.4903472065925598,
+      "learning_rate": 8.594818773412376e-06,
+      "loss": 0.3759,
+      "step": 2065
+    },
+    {
+      "epoch": 0.960781274221051,
+      "grad_norm": 0.4430171251296997,
+      "learning_rate": 8.592937492267466e-06,
+      "loss": 0.3993,
+      "step": 2066
+    },
+    {
+      "epoch": 0.961246318400248,
+      "grad_norm": 0.4540698826313019,
+      "learning_rate": 8.591055158809526e-06,
+      "loss": 0.3555,
+      "step": 2067
+    },
+    {
+      "epoch": 0.961711362579445,
+      "grad_norm": 0.4455225169658661,
+      "learning_rate": 8.589171773589861e-06,
+      "loss": 0.3937,
+      "step": 2068
+    },
+    {
+      "epoch": 0.962176406758642,
+      "grad_norm": 0.45158520340919495,
+      "learning_rate": 8.587287337160083e-06,
+      "loss": 0.3694,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9626414509378391,
+      "grad_norm": 0.44573891162872314,
+      "learning_rate": 8.585401850072114e-06,
+      "loss": 0.364,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9631064951170362,
+      "grad_norm": 0.44659656286239624,
+      "learning_rate": 8.58351531287818e-06,
+      "loss": 0.393,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9635715392962332,
+      "grad_norm": 0.4550364911556244,
+      "learning_rate": 8.581627726130817e-06,
+      "loss": 0.3919,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9640365834754302,
+      "grad_norm": 0.4669656455516815,
+      "learning_rate": 8.579739090382873e-06,
+      "loss": 0.4129,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9645016276546272,
+      "grad_norm": 0.45073822140693665,
+      "learning_rate": 8.577849406187493e-06,
+      "loss": 0.3963,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9649666718338242,
+      "grad_norm": 0.4639911949634552,
+      "learning_rate": 8.575958674098138e-06,
+      "loss": 0.3857,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9654317160130212,
+      "grad_norm": 0.4226040244102478,
+      "learning_rate": 8.574066894668573e-06,
+      "loss": 0.3762,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9658967601922183,
+      "grad_norm": 0.42542764544487,
+      "learning_rate": 8.572174068452867e-06,
+      "loss": 0.3502,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9663618043714153,
+      "grad_norm": 0.46741747856140137,
+      "learning_rate": 8.570280196005403e-06,
+      "loss": 0.3717,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9668268485506123,
+      "grad_norm": 0.4867299199104309,
+      "learning_rate": 8.568385277880859e-06,
+      "loss": 0.3835,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9672918927298093,
+      "grad_norm": 0.4860565662384033,
+      "learning_rate": 8.56648931463423e-06,
+      "loss": 0.3879,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9677569369090063,
+      "grad_norm": 0.44712382555007935,
+      "learning_rate": 8.564592306820813e-06,
+      "loss": 0.3715,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9682219810882033,
+      "grad_norm": 0.4493529498577118,
+      "learning_rate": 8.562694254996208e-06,
+      "loss": 0.3594,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9686870252674004,
+      "grad_norm": 0.44625014066696167,
+      "learning_rate": 8.560795159716327e-06,
+      "loss": 0.3816,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9691520694465974,
+      "grad_norm": 0.44804733991622925,
+      "learning_rate": 8.55889502153738e-06,
+      "loss": 0.3942,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9696171136257945,
+      "grad_norm": 0.43751975893974304,
+      "learning_rate": 8.55699384101589e-06,
+      "loss": 0.3754,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9700821578049915,
+      "grad_norm": 0.5086488723754883,
+      "learning_rate": 8.555091618708681e-06,
+      "loss": 0.3861,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9705472019841885,
+      "grad_norm": 0.4142938554286957,
+      "learning_rate": 8.553188355172882e-06,
+      "loss": 0.382,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9710122461633856,
+      "grad_norm": 0.4268786311149597,
+      "learning_rate": 8.551284050965929e-06,
+      "loss": 0.3612,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9714772903425826,
+      "grad_norm": 0.40908634662628174,
+      "learning_rate": 8.54937870664556e-06,
+      "loss": 0.3853,
+      "step": 2089
+    },
+    {
+      "epoch": 0.9719423345217796,
+      "grad_norm": 0.4633747935295105,
+      "learning_rate": 8.547472322769825e-06,
+      "loss": 0.3561,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9724073787009766,
+      "grad_norm": 0.45280638337135315,
+      "learning_rate": 8.545564899897066e-06,
+      "loss": 0.4023,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9728724228801736,
+      "grad_norm": 0.5002387166023254,
+      "learning_rate": 8.54365643858594e-06,
+      "loss": 0.3743,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9733374670593706,
+      "grad_norm": 0.49016910791397095,
+      "learning_rate": 8.541746939395403e-06,
+      "loss": 0.3708,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9738025112385676,
+      "grad_norm": 0.4835710823535919,
+      "learning_rate": 8.539836402884715e-06,
+      "loss": 0.3901,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9742675554177647,
+      "grad_norm": 0.5034793615341187,
+      "learning_rate": 8.537924829613444e-06,
+      "loss": 0.3716,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9747325995969617,
+      "grad_norm": 0.5456627011299133,
+      "learning_rate": 8.536012220141458e-06,
+      "loss": 0.3759,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9751976437761587,
+      "grad_norm": 0.4535893499851227,
+      "learning_rate": 8.534098575028928e-06,
+      "loss": 0.3818,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9756626879553557,
+      "grad_norm": 0.40453192591667175,
+      "learning_rate": 8.53218389483633e-06,
+      "loss": 0.3686,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9761277321345527,
+      "grad_norm": 0.5098088979721069,
+      "learning_rate": 8.530268180124444e-06,
+      "loss": 0.385,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9765927763137499,
+      "grad_norm": 0.47901231050491333,
+      "learning_rate": 8.528351431454352e-06,
+      "loss": 0.3932,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9770578204929469,
+      "grad_norm": 0.4342050850391388,
+      "learning_rate": 8.526433649387435e-06,
+      "loss": 0.3771,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9775228646721439,
+      "grad_norm": 0.424624502658844,
+      "learning_rate": 8.524514834485382e-06,
+      "loss": 0.3976,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9779879088513409,
+      "grad_norm": 0.4495830237865448,
+      "learning_rate": 8.522594987310184e-06,
+      "loss": 0.3901,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9784529530305379,
+      "grad_norm": 0.46972575783729553,
+      "learning_rate": 8.520674108424134e-06,
+      "loss": 0.4025,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9789179972097349,
+      "grad_norm": 0.46081069111824036,
+      "learning_rate": 8.518752198389823e-06,
+      "loss": 0.3794,
+      "step": 2105
+    },
+    {
+      "epoch": 0.979383041388932,
+      "grad_norm": 0.47898101806640625,
+      "learning_rate": 8.51682925777015e-06,
+      "loss": 0.3701,
+      "step": 2106
+    },
+    {
+      "epoch": 0.979848085568129,
+      "grad_norm": 0.4954424500465393,
+      "learning_rate": 8.51490528712831e-06,
+      "loss": 0.3656,
+      "step": 2107
+    },
+    {
+      "epoch": 0.980313129747326,
+      "grad_norm": 0.5420515537261963,
+      "learning_rate": 8.512980287027805e-06,
+      "loss": 0.37,
+      "step": 2108
+    },
+    {
+      "epoch": 0.980778173926523,
+      "grad_norm": 0.42437291145324707,
+      "learning_rate": 8.511054258032436e-06,
+      "loss": 0.395,
+      "step": 2109
+    },
+    {
+      "epoch": 0.98124321810572,
+      "grad_norm": 0.4806736409664154,
+      "learning_rate": 8.509127200706305e-06,
+      "loss": 0.3452,
+      "step": 2110
+    },
+    {
+      "epoch": 0.981708262284917,
+      "grad_norm": 0.4276758134365082,
+      "learning_rate": 8.507199115613818e-06,
+      "loss": 0.3351,
+      "step": 2111
+    },
+    {
+      "epoch": 0.982173306464114,
+      "grad_norm": 0.484733909368515,
+      "learning_rate": 8.505270003319676e-06,
+      "loss": 0.3501,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9826383506433111,
+      "grad_norm": 0.5105139017105103,
+      "learning_rate": 8.503339864388887e-06,
+      "loss": 0.4193,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9831033948225082,
+      "grad_norm": 0.5405691266059875,
+      "learning_rate": 8.501408699386758e-06,
+      "loss": 0.3912,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9835684390017052,
+      "grad_norm": 0.438722163438797,
+      "learning_rate": 8.499476508878894e-06,
+      "loss": 0.3935,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9840334831809022,
+      "grad_norm": 0.5080367922782898,
+      "learning_rate": 8.497543293431202e-06,
+      "loss": 0.3764,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9844985273600992,
+      "grad_norm": 0.49082526564598083,
+      "learning_rate": 8.495609053609893e-06,
+      "loss": 0.3737,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9849635715392963,
+      "grad_norm": 0.4634253978729248,
+      "learning_rate": 8.49367378998147e-06,
+      "loss": 0.3979,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9854286157184933,
+      "grad_norm": 0.4806537926197052,
+      "learning_rate": 8.491737503112744e-06,
+      "loss": 0.3773,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9858936598976903,
+      "grad_norm": 0.5054664611816406,
+      "learning_rate": 8.489800193570818e-06,
+      "loss": 0.3704,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9863587040768873,
+      "grad_norm": 0.47105222940444946,
+      "learning_rate": 8.487861861923103e-06,
+      "loss": 0.3684,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9868237482560843,
+      "grad_norm": 0.4627211391925812,
+      "learning_rate": 8.485922508737302e-06,
+      "loss": 0.3776,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9872887924352813,
+      "grad_norm": 0.4847998321056366,
+      "learning_rate": 8.483982134581419e-06,
+      "loss": 0.3859,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9877538366144784,
+      "grad_norm": 0.5242019295692444,
+      "learning_rate": 8.48204074002376e-06,
+      "loss": 0.3859,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9882188807936754,
+      "grad_norm": 0.48214638233184814,
+      "learning_rate": 8.480098325632928e-06,
+      "loss": 0.364,
+      "step": 2125
+    },
+    {
+      "epoch": 0.9886839249728724,
+      "grad_norm": 0.4689680337905884,
+      "learning_rate": 8.478154891977825e-06,
+      "loss": 0.3843,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9891489691520694,
+      "grad_norm": 0.609089195728302,
+      "learning_rate": 8.47621043962765e-06,
+      "loss": 0.4067,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9896140133312664,
+      "grad_norm": 0.4945160150527954,
+      "learning_rate": 8.474264969151902e-06,
+      "loss": 0.3718,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9900790575104635,
+      "grad_norm": 0.5477257370948792,
+      "learning_rate": 8.472318481120377e-06,
+      "loss": 0.3698,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9905441016896606,
+      "grad_norm": 0.5951197743415833,
+      "learning_rate": 8.470370976103171e-06,
+      "loss": 0.3752,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9910091458688576,
+      "grad_norm": 0.4970504343509674,
+      "learning_rate": 8.468422454670674e-06,
+      "loss": 0.3917,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9914741900480546,
+      "grad_norm": 0.48202067613601685,
+      "learning_rate": 8.46647291739358e-06,
+      "loss": 0.3855,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9919392342272516,
+      "grad_norm": 0.49319443106651306,
+      "learning_rate": 8.464522364842874e-06,
+      "loss": 0.357,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9924042784064486,
+      "grad_norm": 0.47641077637672424,
+      "learning_rate": 8.462570797589842e-06,
+      "loss": 0.3823,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9928693225856456,
+      "grad_norm": 0.4628462791442871,
+      "learning_rate": 8.460618216206069e-06,
+      "loss": 0.36,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9933343667648427,
+      "grad_norm": 0.4999821186065674,
+      "learning_rate": 8.458664621263428e-06,
+      "loss": 0.402,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9937994109440397,
+      "grad_norm": 0.4732115864753723,
+      "learning_rate": 8.456710013334102e-06,
+      "loss": 0.376,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9942644551232367,
+      "grad_norm": 0.42666521668434143,
+      "learning_rate": 8.45475439299056e-06,
+      "loss": 0.3759,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9947294993024337,
+      "grad_norm": 0.512298583984375,
+      "learning_rate": 8.452797760805572e-06,
+      "loss": 0.3646,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9951945434816307,
+      "grad_norm": 0.5457414984703064,
+      "learning_rate": 8.450840117352203e-06,
+      "loss": 0.3637,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9956595876608277,
+      "grad_norm": 0.4568066895008087,
+      "learning_rate": 8.448881463203819e-06,
+      "loss": 0.4164,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9961246318400248,
+      "grad_norm": 0.4378894865512848,
+      "learning_rate": 8.446921798934074e-06,
+      "loss": 0.3666,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9965896760192218,
+      "grad_norm": 0.5727525949478149,
+      "learning_rate": 8.444961125116924e-06,
+      "loss": 0.376,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9970547201984189,
+      "grad_norm": 0.5255966782569885,
+      "learning_rate": 8.442999442326617e-06,
+      "loss": 0.3902,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9975197643776159,
+      "grad_norm": 0.4558255076408386,
+      "learning_rate": 8.441036751137697e-06,
+      "loss": 0.3963,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9979848085568129,
+      "grad_norm": 0.49802669882774353,
+      "learning_rate": 8.439073052125006e-06,
+      "loss": 0.3579,
+      "step": 2146
+    },
+    {
+      "epoch": 0.99844985273601,
+      "grad_norm": 0.5205224752426147,
+      "learning_rate": 8.43710834586368e-06,
+      "loss": 0.3696,
+      "step": 2147
+    },
+    {
+      "epoch": 0.998914896915207,
+      "grad_norm": 0.4866522252559662,
+      "learning_rate": 8.435142632929149e-06,
+      "loss": 0.3896,
+      "step": 2148
+    },
+    {
+      "epoch": 0.999379941094404,
+      "grad_norm": 0.4538208842277527,
+      "learning_rate": 8.43317591389714e-06,
+      "loss": 0.3607,
+      "step": 2149
+    },
+    {
+      "epoch": 0.999844985273601,
+      "grad_norm": 0.5789209604263306,
+      "learning_rate": 8.43120818934367e-06,
+      "loss": 0.379,
+      "step": 2150
+    },
+    {
+      "epoch": 1.000310029452798,
+      "grad_norm": 0.8949771523475647,
+      "learning_rate": 8.429239459845053e-06,
+      "loss": 0.5959,
+      "step": 2151
+    },
+    {
+      "epoch": 1.0007750736319951,
+      "grad_norm": 0.4994892477989197,
+      "learning_rate": 8.427269725977902e-06,
+      "loss": 0.3726,
+      "step": 2152
+    },
+    {
+      "epoch": 1.001240117811192,
+      "grad_norm": 0.49586233496665955,
+      "learning_rate": 8.425298988319119e-06,
+      "loss": 0.3367,
+      "step": 2153
+    },
+    {
+      "epoch": 1.0017051619903892,
+      "grad_norm": 0.4944624900817871,
+      "learning_rate": 8.423327247445898e-06,
+      "loss": 0.3802,
+      "step": 2154
+    },
+    {
+      "epoch": 1.002170206169586,
+      "grad_norm": 0.45939287543296814,
+      "learning_rate": 8.421354503935733e-06,
+      "loss": 0.3271,
+      "step": 2155
+    },
+    {
+      "epoch": 1.0026352503487832,
+      "grad_norm": 0.5419012308120728,
+      "learning_rate": 8.419380758366407e-06,
+      "loss": 0.3778,
+      "step": 2156
+    },
+    {
+      "epoch": 1.00310029452798,
+      "grad_norm": 0.43736547231674194,
+      "learning_rate": 8.417406011316e-06,
+      "loss": 0.3852,
+      "step": 2157
+    },
+    {
+      "epoch": 1.0035653387071772,
+      "grad_norm": 0.5195603966712952,
+      "learning_rate": 8.415430263362878e-06,
+      "loss": 0.3837,
+      "step": 2158
+    },
+    {
+      "epoch": 1.0040303828863741,
+      "grad_norm": 0.5915653109550476,
+      "learning_rate": 8.413453515085712e-06,
+      "loss": 0.3953,
+      "step": 2159
+    },
+    {
+      "epoch": 1.0044954270655713,
+      "grad_norm": 0.4842241406440735,
+      "learning_rate": 8.411475767063454e-06,
+      "loss": 0.3372,
+      "step": 2160
+    },
+    {
+      "epoch": 1.0049604712447682,
+      "grad_norm": 0.46653223037719727,
+      "learning_rate": 8.409497019875362e-06,
+      "loss": 0.3541,
+      "step": 2161
+    },
+    {
+      "epoch": 1.0054255154239653,
+      "grad_norm": 0.4800226390361786,
+      "learning_rate": 8.40751727410097e-06,
+      "loss": 0.3579,
+      "step": 2162
+    },
+    {
+      "epoch": 1.0058905596031622,
+      "grad_norm": 0.4519990086555481,
+      "learning_rate": 8.405536530320118e-06,
+      "loss": 0.3372,
+      "step": 2163
+    },
+    {
+      "epoch": 1.0063556037823593,
+      "grad_norm": 0.4979685842990875,
+      "learning_rate": 8.403554789112934e-06,
+      "loss": 0.343,
+      "step": 2164
+    },
+    {
+      "epoch": 1.0068206479615562,
+      "grad_norm": 0.4970398545265198,
+      "learning_rate": 8.401572051059835e-06,
+      "loss": 0.3701,
+      "step": 2165
+    },
+    {
+      "epoch": 1.0072856921407534,
+      "grad_norm": 0.5320008993148804,
+      "learning_rate": 8.399588316741535e-06,
+      "loss": 0.3512,
+      "step": 2166
+    },
+    {
+      "epoch": 1.0077507363199505,
+      "grad_norm": 0.4425172209739685,
+      "learning_rate": 8.397603586739039e-06,
+      "loss": 0.3165,
+      "step": 2167
+    },
+    {
+      "epoch": 1.0082157804991474,
+      "grad_norm": 0.4745187759399414,
+      "learning_rate": 8.395617861633637e-06,
+      "loss": 0.3919,
+      "step": 2168
+    },
+    {
+      "epoch": 1.0086808246783445,
+      "grad_norm": 0.5007448196411133,
+      "learning_rate": 8.393631142006922e-06,
+      "loss": 0.3564,
+      "step": 2169
+    },
+    {
+      "epoch": 1.0091458688575414,
+      "grad_norm": 0.4433797001838684,
+      "learning_rate": 8.391643428440766e-06,
+      "loss": 0.367,
+      "step": 2170
+    },
+    {
+      "epoch": 1.0096109130367386,
+      "grad_norm": 0.37245190143585205,
+      "learning_rate": 8.389654721517341e-06,
+      "loss": 0.3361,
+      "step": 2171
+    },
+    {
+      "epoch": 1.0100759572159355,
+      "grad_norm": 0.5664768815040588,
+      "learning_rate": 8.38766502181911e-06,
+      "loss": 0.382,
+      "step": 2172
+    },
+    {
+      "epoch": 1.0105410013951326,
+      "grad_norm": 0.4073793888092041,
+      "learning_rate": 8.385674329928819e-06,
+      "loss": 0.3291,
+      "step": 2173
+    },
+    {
+      "epoch": 1.0110060455743295,
+      "grad_norm": 0.509306788444519,
+      "learning_rate": 8.383682646429509e-06,
+      "loss": 0.4038,
+      "step": 2174
+    },
+    {
+      "epoch": 1.0114710897535266,
+      "grad_norm": 0.39777812361717224,
+      "learning_rate": 8.381689971904514e-06,
+      "loss": 0.3413,
+      "step": 2175
+    },
+    {
+      "epoch": 1.0119361339327235,
+      "grad_norm": 0.5179665088653564,
+      "learning_rate": 8.379696306937457e-06,
+      "loss": 0.3944,
+      "step": 2176
+    },
+    {
+      "epoch": 1.0124011781119207,
+      "grad_norm": 0.4474855661392212,
+      "learning_rate": 8.377701652112249e-06,
+      "loss": 0.3663,
+      "step": 2177
+    },
+    {
+      "epoch": 1.0128662222911176,
+      "grad_norm": 0.4197114408016205,
+      "learning_rate": 8.37570600801309e-06,
+      "loss": 0.325,
+      "step": 2178
+    },
+    {
+      "epoch": 1.0133312664703147,
+      "grad_norm": 0.48934638500213623,
+      "learning_rate": 8.373709375224475e-06,
+      "loss": 0.3751,
+      "step": 2179
+    },
+    {
+      "epoch": 1.0137963106495118,
+      "grad_norm": 0.4234052002429962,
+      "learning_rate": 8.371711754331181e-06,
+      "loss": 0.3585,
+      "step": 2180
+    },
+    {
+      "epoch": 1.0142613548287087,
+      "grad_norm": 0.511944055557251,
+      "learning_rate": 8.369713145918284e-06,
+      "loss": 0.3847,
+      "step": 2181
+    },
+    {
+      "epoch": 1.0147263990079058,
+      "grad_norm": 0.46818193793296814,
+      "learning_rate": 8.36771355057114e-06,
+      "loss": 0.38,
+      "step": 2182
+    },
+    {
+      "epoch": 1.0151914431871027,
+      "grad_norm": 0.41124582290649414,
+      "learning_rate": 8.365712968875399e-06,
+      "loss": 0.3481,
+      "step": 2183
+    },
+    {
+      "epoch": 1.0156564873662999,
+      "grad_norm": 0.42587098479270935,
+      "learning_rate": 8.363711401417e-06,
+      "loss": 0.3544,
+      "step": 2184
+    },
+    {
+      "epoch": 1.0161215315454968,
+      "grad_norm": 0.41804370284080505,
+      "learning_rate": 8.36170884878217e-06,
+      "loss": 0.3613,
+      "step": 2185
+    },
+    {
+      "epoch": 1.016586575724694,
+      "grad_norm": 0.41786617040634155,
+      "learning_rate": 8.359705311557421e-06,
+      "loss": 0.3388,
+      "step": 2186
+    },
+    {
+      "epoch": 1.0170516199038908,
+      "grad_norm": 0.4178462028503418,
+      "learning_rate": 8.35770079032956e-06,
+      "loss": 0.3509,
+      "step": 2187
+    },
+    {
+      "epoch": 1.017516664083088,
+      "grad_norm": 0.3867464065551758,
+      "learning_rate": 8.355695285685675e-06,
+      "loss": 0.3443,
+      "step": 2188
+    },
+    {
+      "epoch": 1.0179817082622848,
+      "grad_norm": 0.5491018891334534,
+      "learning_rate": 8.35368879821315e-06,
+      "loss": 0.3935,
+      "step": 2189
+    },
+    {
+      "epoch": 1.018446752441482,
+      "grad_norm": 0.4546567499637604,
+      "learning_rate": 8.35168132849965e-06,
+      "loss": 0.3451,
+      "step": 2190
+    },
+    {
+      "epoch": 1.0189117966206789,
+      "grad_norm": 0.426011323928833,
+      "learning_rate": 8.349672877133131e-06,
+      "loss": 0.315,
+      "step": 2191
+    },
+    {
+      "epoch": 1.019376840799876,
+      "grad_norm": 0.43346622586250305,
+      "learning_rate": 8.347663444701835e-06,
+      "loss": 0.3507,
+      "step": 2192
+    },
+    {
+      "epoch": 1.019841884979073,
+      "grad_norm": 0.4720572233200073,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 0.3445,
+      "step": 2193
+    },
+    {
+      "epoch": 1.02030692915827,
+      "grad_norm": 0.4907239079475403,
+      "learning_rate": 8.34364163899932e-06,
+      "loss": 0.336,
+      "step": 2194
+    },
+    {
+      "epoch": 1.0207719733374672,
+      "grad_norm": 0.47017717361450195,
+      "learning_rate": 8.341629266906024e-06,
+      "loss": 0.4029,
+      "step": 2195
+    },
+    {
+      "epoch": 1.021237017516664,
+      "grad_norm": 0.4325951933860779,
+      "learning_rate": 8.339615916103795e-06,
+      "loss": 0.3129,
+      "step": 2196
+    },
+    {
+      "epoch": 1.0217020616958612,
+      "grad_norm": 0.4916847050189972,
+      "learning_rate": 8.33760158718231e-06,
+      "loss": 0.3647,
+      "step": 2197
+    },
+    {
+      "epoch": 1.022167105875058,
+      "grad_norm": 0.43246665596961975,
+      "learning_rate": 8.335586280731532e-06,
+      "loss": 0.3216,
+      "step": 2198
+    },
+    {
+      "epoch": 1.0226321500542552,
+      "grad_norm": 0.5200701355934143,
+      "learning_rate": 8.333569997341713e-06,
+      "loss": 0.355,
+      "step": 2199
+    },
+    {
+      "epoch": 1.0230971942334521,
+      "grad_norm": 0.4282546639442444,
+      "learning_rate": 8.33155273760339e-06,
+      "loss": 0.3829,
+      "step": 2200
+    },
+    {
+      "epoch": 1.0235622384126493,
+      "grad_norm": 0.42160195112228394,
+      "learning_rate": 8.329534502107386e-06,
+      "loss": 0.3202,
+      "step": 2201
+    },
+    {
+      "epoch": 1.0240272825918462,
+      "grad_norm": 0.42160677909851074,
+      "learning_rate": 8.327515291444807e-06,
+      "loss": 0.3368,
+      "step": 2202
+    },
+    {
+      "epoch": 1.0244923267710433,
+      "grad_norm": 0.4635396897792816,
+      "learning_rate": 8.325495106207049e-06,
+      "loss": 0.3864,
+      "step": 2203
+    },
+    {
+      "epoch": 1.0249573709502402,
+      "grad_norm": 0.4313600957393646,
+      "learning_rate": 8.32347394698579e-06,
+      "loss": 0.3538,
+      "step": 2204
+    },
+    {
+      "epoch": 1.0254224151294373,
+      "grad_norm": 0.4247414767742157,
+      "learning_rate": 8.321451814372998e-06,
+      "loss": 0.3499,
+      "step": 2205
+    },
+    {
+      "epoch": 1.0258874593086342,
+      "grad_norm": 0.5032163262367249,
+      "learning_rate": 8.319428708960917e-06,
+      "loss": 0.3768,
+      "step": 2206
+    },
+    {
+      "epoch": 1.0263525034878314,
+      "grad_norm": 0.44325828552246094,
+      "learning_rate": 8.317404631342088e-06,
+      "loss": 0.3473,
+      "step": 2207
+    },
+    {
+      "epoch": 1.0268175476670283,
+      "grad_norm": 0.3861435353755951,
+      "learning_rate": 8.315379582109326e-06,
+      "loss": 0.3206,
+      "step": 2208
+    },
+    {
+      "epoch": 1.0272825918462254,
+      "grad_norm": 0.4646325707435608,
+      "learning_rate": 8.313353561855737e-06,
+      "loss": 0.3673,
+      "step": 2209
+    },
+    {
+      "epoch": 1.0277476360254225,
+      "grad_norm": 0.4774647653102875,
+      "learning_rate": 8.31132657117471e-06,
+      "loss": 0.3534,
+      "step": 2210
+    },
+    {
+      "epoch": 1.0282126802046194,
+      "grad_norm": 0.4477936029434204,
+      "learning_rate": 8.309298610659917e-06,
+      "loss": 0.3531,
+      "step": 2211
+    },
+    {
+      "epoch": 1.0286777243838165,
+      "grad_norm": 0.5256711840629578,
+      "learning_rate": 8.307269680905312e-06,
+      "loss": 0.3825,
+      "step": 2212
+    },
+    {
+      "epoch": 1.0291427685630135,
+      "grad_norm": 0.43662089109420776,
+      "learning_rate": 8.305239782505142e-06,
+      "loss": 0.3622,
+      "step": 2213
+    },
+    {
+      "epoch": 1.0296078127422106,
+      "grad_norm": 0.4843900501728058,
+      "learning_rate": 8.303208916053924e-06,
+      "loss": 0.3555,
+      "step": 2214
+    },
+    {
+      "epoch": 1.0300728569214075,
+      "grad_norm": 0.5311547517776489,
+      "learning_rate": 8.30117708214647e-06,
+      "loss": 0.3647,
+      "step": 2215
+    },
+    {
+      "epoch": 1.0305379011006046,
+      "grad_norm": 0.4405294954776764,
+      "learning_rate": 8.299144281377869e-06,
+      "loss": 0.308,
+      "step": 2216
+    },
+    {
+      "epoch": 1.0310029452798015,
+      "grad_norm": 0.5497903823852539,
+      "learning_rate": 8.297110514343498e-06,
+      "loss": 0.3571,
+      "step": 2217
+    },
+    {
+      "epoch": 1.0314679894589986,
+      "grad_norm": 0.44999319314956665,
+      "learning_rate": 8.295075781639013e-06,
+      "loss": 0.3713,
+      "step": 2218
+    },
+    {
+      "epoch": 1.0319330336381956,
+      "grad_norm": 0.5399612188339233,
+      "learning_rate": 8.293040083860352e-06,
+      "loss": 0.3285,
+      "step": 2219
+    },
+    {
+      "epoch": 1.0323980778173927,
+      "grad_norm": 0.5000502467155457,
+      "learning_rate": 8.29100342160374e-06,
+      "loss": 0.351,
+      "step": 2220
+    },
+    {
+      "epoch": 1.0328631219965896,
+      "grad_norm": 0.5676979422569275,
+      "learning_rate": 8.288965795465684e-06,
+      "loss": 0.3698,
+      "step": 2221
+    },
+    {
+      "epoch": 1.0333281661757867,
+      "grad_norm": 0.5297927260398865,
+      "learning_rate": 8.28692720604297e-06,
+      "loss": 0.3449,
+      "step": 2222
+    },
+    {
+      "epoch": 1.0337932103549836,
+      "grad_norm": 0.5518194437026978,
+      "learning_rate": 8.284887653932665e-06,
+      "loss": 0.3581,
+      "step": 2223
+    },
+    {
+      "epoch": 1.0342582545341807,
+      "grad_norm": 0.49941563606262207,
+      "learning_rate": 8.282847139732125e-06,
+      "loss": 0.3357,
+      "step": 2224
+    },
+    {
+      "epoch": 1.0347232987133779,
+      "grad_norm": 0.5394763946533203,
+      "learning_rate": 8.28080566403898e-06,
+      "loss": 0.3671,
+      "step": 2225
+    },
+    {
+      "epoch": 1.0351883428925748,
+      "grad_norm": 0.5620566010475159,
+      "learning_rate": 8.278763227451148e-06,
+      "loss": 0.3658,
+      "step": 2226
+    },
+    {
+      "epoch": 1.035653387071772,
+      "grad_norm": 0.5896021723747253,
+      "learning_rate": 8.276719830566823e-06,
+      "loss": 0.3925,
+      "step": 2227
+    },
+    {
+      "epoch": 1.0361184312509688,
+      "grad_norm": 0.5375012159347534,
+      "learning_rate": 8.274675473984486e-06,
+      "loss": 0.3573,
+      "step": 2228
+    },
+    {
+      "epoch": 1.036583475430166,
+      "grad_norm": 0.48443183302879333,
+      "learning_rate": 8.272630158302892e-06,
+      "loss": 0.3385,
+      "step": 2229
+    },
+    {
+      "epoch": 1.0370485196093628,
+      "grad_norm": 0.5011869072914124,
+      "learning_rate": 8.270583884121083e-06,
+      "loss": 0.356,
+      "step": 2230
+    },
+    {
+      "epoch": 1.03751356378856,
+      "grad_norm": 0.5027603507041931,
+      "learning_rate": 8.268536652038379e-06,
+      "loss": 0.3276,
+      "step": 2231
+    },
+    {
+      "epoch": 1.0379786079677569,
+      "grad_norm": 0.593014657497406,
+      "learning_rate": 8.266488462654381e-06,
+      "loss": 0.3687,
+      "step": 2232
+    },
+    {
+      "epoch": 1.038443652146954,
+      "grad_norm": 0.6836045384407043,
+      "learning_rate": 8.264439316568969e-06,
+      "loss": 0.3779,
+      "step": 2233
+    },
+    {
+      "epoch": 1.038908696326151,
+      "grad_norm": 0.47743064165115356,
+      "learning_rate": 8.262389214382307e-06,
+      "loss": 0.3386,
+      "step": 2234
+    },
+    {
+      "epoch": 1.039373740505348,
+      "grad_norm": 0.7030071020126343,
+      "learning_rate": 8.260338156694836e-06,
+      "loss": 0.3694,
+      "step": 2235
+    },
+    {
+      "epoch": 1.039838784684545,
+      "grad_norm": 0.468199223279953,
+      "learning_rate": 8.258286144107277e-06,
+      "loss": 0.3337,
+      "step": 2236
+    },
+    {
+      "epoch": 1.040303828863742,
+      "grad_norm": 0.43109989166259766,
+      "learning_rate": 8.256233177220632e-06,
+      "loss": 0.3538,
+      "step": 2237
+    },
+    {
+      "epoch": 1.0407688730429392,
+      "grad_norm": 0.5127283930778503,
+      "learning_rate": 8.25417925663618e-06,
+      "loss": 0.3447,
+      "step": 2238
+    },
+    {
+      "epoch": 1.041233917222136,
+      "grad_norm": 0.5324409604072571,
+      "learning_rate": 8.25212438295548e-06,
+      "loss": 0.3476,
+      "step": 2239
+    },
+    {
+      "epoch": 1.0416989614013332,
+      "grad_norm": 0.4684849679470062,
+      "learning_rate": 8.250068556780376e-06,
+      "loss": 0.348,
+      "step": 2240
+    },
+    {
+      "epoch": 1.0421640055805301,
+      "grad_norm": 0.5129745006561279,
+      "learning_rate": 8.24801177871298e-06,
+      "loss": 0.3528,
+      "step": 2241
+    },
+    {
+      "epoch": 1.0426290497597273,
+      "grad_norm": 0.5319026112556458,
+      "learning_rate": 8.245954049355696e-06,
+      "loss": 0.3404,
+      "step": 2242
+    },
+    {
+      "epoch": 1.0430940939389242,
+      "grad_norm": 0.4605168104171753,
+      "learning_rate": 8.243895369311192e-06,
+      "loss": 0.3319,
+      "step": 2243
+    },
+    {
+      "epoch": 1.0435591381181213,
+      "grad_norm": 0.4982357621192932,
+      "learning_rate": 8.241835739182426e-06,
+      "loss": 0.3557,
+      "step": 2244
+    },
+    {
+      "epoch": 1.0440241822973182,
+      "grad_norm": 0.5314902067184448,
+      "learning_rate": 8.239775159572632e-06,
+      "loss": 0.352,
+      "step": 2245
+    },
+    {
+      "epoch": 1.0444892264765153,
+      "grad_norm": 0.4709586799144745,
+      "learning_rate": 8.237713631085316e-06,
+      "loss": 0.3741,
+      "step": 2246
+    },
+    {
+      "epoch": 1.0449542706557122,
+      "grad_norm": 0.4382578134536743,
+      "learning_rate": 8.235651154324269e-06,
+      "loss": 0.3414,
+      "step": 2247
+    },
+    {
+      "epoch": 1.0454193148349094,
+      "grad_norm": 0.43601369857788086,
+      "learning_rate": 8.233587729893555e-06,
+      "loss": 0.3409,
+      "step": 2248
+    },
+    {
+      "epoch": 1.0458843590141063,
+      "grad_norm": 0.4836721122264862,
+      "learning_rate": 8.23152335839752e-06,
+      "loss": 0.3506,
+      "step": 2249
+    },
+    {
+      "epoch": 1.0463494031933034,
+      "grad_norm": 0.5360286235809326,
+      "learning_rate": 8.229458040440783e-06,
+      "loss": 0.3817,
+      "step": 2250
+    },
+    {
+      "epoch": 1.0468144473725003,
+      "grad_norm": 0.4707900583744049,
+      "learning_rate": 8.227391776628242e-06,
+      "loss": 0.3439,
+      "step": 2251
+    },
+    {
+      "epoch": 1.0472794915516974,
+      "grad_norm": 0.48144766688346863,
+      "learning_rate": 8.225324567565071e-06,
+      "loss": 0.3562,
+      "step": 2252
+    },
+    {
+      "epoch": 1.0477445357308945,
+      "grad_norm": 0.5262990593910217,
+      "learning_rate": 8.223256413856726e-06,
+      "loss": 0.365,
+      "step": 2253
+    },
+    {
+      "epoch": 1.0482095799100914,
+      "grad_norm": 0.5710676312446594,
+      "learning_rate": 8.221187316108935e-06,
+      "loss": 0.3493,
+      "step": 2254
+    },
+    {
+      "epoch": 1.0486746240892886,
+      "grad_norm": 0.44150006771087646,
+      "learning_rate": 8.219117274927696e-06,
+      "loss": 0.3598,
+      "step": 2255
+    },
+    {
+      "epoch": 1.0491396682684855,
+      "grad_norm": 0.43997588753700256,
+      "learning_rate": 8.2170462909193e-06,
+      "loss": 0.307,
+      "step": 2256
+    },
+    {
+      "epoch": 1.0496047124476826,
+      "grad_norm": 0.582988440990448,
+      "learning_rate": 8.2149743646903e-06,
+      "loss": 0.3721,
+      "step": 2257
+    },
+    {
+      "epoch": 1.0500697566268795,
+      "grad_norm": 0.4914950728416443,
+      "learning_rate": 8.212901496847528e-06,
+      "loss": 0.3651,
+      "step": 2258
+    },
+    {
+      "epoch": 1.0505348008060766,
+      "grad_norm": 0.4819602966308594,
+      "learning_rate": 8.210827687998098e-06,
+      "loss": 0.4051,
+      "step": 2259
+    },
+    {
+      "epoch": 1.0509998449852735,
+      "grad_norm": 0.47203314304351807,
+      "learning_rate": 8.208752938749389e-06,
+      "loss": 0.3389,
+      "step": 2260
+    },
+    {
+      "epoch": 1.0514648891644707,
+      "grad_norm": 0.5789709687232971,
+      "learning_rate": 8.206677249709066e-06,
+      "loss": 0.3793,
+      "step": 2261
+    },
+    {
+      "epoch": 1.0519299333436676,
+      "grad_norm": 0.5130416750907898,
+      "learning_rate": 8.204600621485064e-06,
+      "loss": 0.3352,
+      "step": 2262
+    },
+    {
+      "epoch": 1.0523949775228647,
+      "grad_norm": 0.4451179802417755,
+      "learning_rate": 8.202523054685592e-06,
+      "loss": 0.3723,
+      "step": 2263
+    },
+    {
+      "epoch": 1.0528600217020616,
+      "grad_norm": 0.5486533641815186,
+      "learning_rate": 8.200444549919135e-06,
+      "loss": 0.3537,
+      "step": 2264
+    },
+    {
+      "epoch": 1.0533250658812587,
+      "grad_norm": 0.4916934072971344,
+      "learning_rate": 8.198365107794457e-06,
+      "loss": 0.3705,
+      "step": 2265
+    },
+    {
+      "epoch": 1.0537901100604556,
+      "grad_norm": 0.46391212940216064,
+      "learning_rate": 8.196284728920589e-06,
+      "loss": 0.3469,
+      "step": 2266
+    },
+    {
+      "epoch": 1.0542551542396528,
+      "grad_norm": 0.5530695915222168,
+      "learning_rate": 8.194203413906843e-06,
+      "loss": 0.3614,
+      "step": 2267
+    },
+    {
+      "epoch": 1.05472019841885,
+      "grad_norm": 0.4474685490131378,
+      "learning_rate": 8.1921211633628e-06,
+      "loss": 0.3722,
+      "step": 2268
+    },
+    {
+      "epoch": 1.0551852425980468,
+      "grad_norm": 0.4766155779361725,
+      "learning_rate": 8.190037977898319e-06,
+      "loss": 0.3416,
+      "step": 2269
+    },
+    {
+      "epoch": 1.055650286777244,
+      "grad_norm": 0.541581392288208,
+      "learning_rate": 8.187953858123529e-06,
+      "loss": 0.3799,
+      "step": 2270
+    },
+    {
+      "epoch": 1.0561153309564408,
+      "grad_norm": 0.5254855751991272,
+      "learning_rate": 8.185868804648838e-06,
+      "loss": 0.3685,
+      "step": 2271
+    },
+    {
+      "epoch": 1.056580375135638,
+      "grad_norm": 0.4539804458618164,
+      "learning_rate": 8.183782818084922e-06,
+      "loss": 0.3398,
+      "step": 2272
+    },
+    {
+      "epoch": 1.0570454193148349,
+      "grad_norm": 0.4451497197151184,
+      "learning_rate": 8.181695899042733e-06,
+      "loss": 0.35,
+      "step": 2273
+    },
+    {
+      "epoch": 1.057510463494032,
+      "grad_norm": 0.4841856360435486,
+      "learning_rate": 8.179608048133497e-06,
+      "loss": 0.3849,
+      "step": 2274
+    },
+    {
+      "epoch": 1.057975507673229,
+      "grad_norm": 0.3748435080051422,
+      "learning_rate": 8.17751926596871e-06,
+      "loss": 0.3153,
+      "step": 2275
+    },
+    {
+      "epoch": 1.058440551852426,
+      "grad_norm": 0.4471512734889984,
+      "learning_rate": 8.175429553160142e-06,
+      "loss": 0.3684,
+      "step": 2276
+    },
+    {
+      "epoch": 1.058905596031623,
+      "grad_norm": 0.47942283749580383,
+      "learning_rate": 8.17333891031984e-06,
+      "loss": 0.3909,
+      "step": 2277
+    },
+    {
+      "epoch": 1.05937064021082,
+      "grad_norm": 0.3868976831436157,
+      "learning_rate": 8.171247338060113e-06,
+      "loss": 0.3071,
+      "step": 2278
+    },
+    {
+      "epoch": 1.059835684390017,
+      "grad_norm": 0.42615851759910583,
+      "learning_rate": 8.16915483699355e-06,
+      "loss": 0.375,
+      "step": 2279
+    },
+    {
+      "epoch": 1.060300728569214,
+      "grad_norm": 0.42659878730773926,
+      "learning_rate": 8.167061407733018e-06,
+      "loss": 0.3678,
+      "step": 2280
+    },
+    {
+      "epoch": 1.060765772748411,
+      "grad_norm": 0.4303070306777954,
+      "learning_rate": 8.164967050891639e-06,
+      "loss": 0.3455,
+      "step": 2281
+    },
+    {
+      "epoch": 1.0612308169276081,
+      "grad_norm": 0.4534488618373871,
+      "learning_rate": 8.16287176708282e-06,
+      "loss": 0.3615,
+      "step": 2282
+    },
+    {
+      "epoch": 1.0616958611068052,
+      "grad_norm": 0.5386179685592651,
+      "learning_rate": 8.160775556920236e-06,
+      "loss": 0.3719,
+      "step": 2283
+    },
+    {
+      "epoch": 1.0621609052860022,
+      "grad_norm": 0.48542487621307373,
+      "learning_rate": 8.158678421017833e-06,
+      "loss": 0.34,
+      "step": 2284
+    },
+    {
+      "epoch": 1.0626259494651993,
+      "grad_norm": 0.43062224984169006,
+      "learning_rate": 8.156580359989827e-06,
+      "loss": 0.3546,
+      "step": 2285
+    },
+    {
+      "epoch": 1.0630909936443962,
+      "grad_norm": 0.5932109355926514,
+      "learning_rate": 8.154481374450707e-06,
+      "loss": 0.3409,
+      "step": 2286
+    },
+    {
+      "epoch": 1.0635560378235933,
+      "grad_norm": 0.5103998184204102,
+      "learning_rate": 8.15238146501523e-06,
+      "loss": 0.3665,
+      "step": 2287
+    },
+    {
+      "epoch": 1.0640210820027902,
+      "grad_norm": 0.41876837611198425,
+      "learning_rate": 8.150280632298426e-06,
+      "loss": 0.313,
+      "step": 2288
+    },
+    {
+      "epoch": 1.0644861261819873,
+      "grad_norm": 0.5669198036193848,
+      "learning_rate": 8.148178876915598e-06,
+      "loss": 0.3474,
+      "step": 2289
+    },
+    {
+      "epoch": 1.0649511703611843,
+      "grad_norm": 0.5304418206214905,
+      "learning_rate": 8.14607619948231e-06,
+      "loss": 0.3581,
+      "step": 2290
+    },
+    {
+      "epoch": 1.0654162145403814,
+      "grad_norm": 0.48082804679870605,
+      "learning_rate": 8.143972600614407e-06,
+      "loss": 0.3727,
+      "step": 2291
+    },
+    {
+      "epoch": 1.0658812587195783,
+      "grad_norm": 0.5468645691871643,
+      "learning_rate": 8.141868080927998e-06,
+      "loss": 0.3748,
+      "step": 2292
+    },
+    {
+      "epoch": 1.0663463028987754,
+      "grad_norm": 0.46219947934150696,
+      "learning_rate": 8.13976264103946e-06,
+      "loss": 0.392,
+      "step": 2293
+    },
+    {
+      "epoch": 1.0668113470779723,
+      "grad_norm": 0.508678674697876,
+      "learning_rate": 8.137656281565445e-06,
+      "loss": 0.345,
+      "step": 2294
+    },
+    {
+      "epoch": 1.0672763912571694,
+      "grad_norm": 0.5073127746582031,
+      "learning_rate": 8.135549003122871e-06,
+      "loss": 0.3621,
+      "step": 2295
+    },
+    {
+      "epoch": 1.0677414354363663,
+      "grad_norm": 0.4401457905769348,
+      "learning_rate": 8.133440806328925e-06,
+      "loss": 0.3351,
+      "step": 2296
+    },
+    {
+      "epoch": 1.0682064796155635,
+      "grad_norm": 0.4791904091835022,
+      "learning_rate": 8.131331691801066e-06,
+      "loss": 0.3336,
+      "step": 2297
+    },
+    {
+      "epoch": 1.0686715237947606,
+      "grad_norm": 0.4278305470943451,
+      "learning_rate": 8.129221660157014e-06,
+      "loss": 0.3651,
+      "step": 2298
+    },
+    {
+      "epoch": 1.0691365679739575,
+      "grad_norm": 0.47636422514915466,
+      "learning_rate": 8.127110712014767e-06,
+      "loss": 0.3022,
+      "step": 2299
+    },
+    {
+      "epoch": 1.0696016121531546,
+      "grad_norm": 0.4815133512020111,
+      "learning_rate": 8.124998847992587e-06,
+      "loss": 0.335,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0700666563323515,
+      "grad_norm": 0.4794626235961914,
+      "learning_rate": 8.122886068709003e-06,
+      "loss": 0.374,
+      "step": 2301
+    },
+    {
+      "epoch": 1.0705317005115487,
+      "grad_norm": 0.4852800667285919,
+      "learning_rate": 8.120772374782818e-06,
+      "loss": 0.3387,
+      "step": 2302
+    },
+    {
+      "epoch": 1.0709967446907456,
+      "grad_norm": 0.4669995903968811,
+      "learning_rate": 8.118657766833093e-06,
+      "loss": 0.332,
+      "step": 2303
+    },
+    {
+      "epoch": 1.0714617888699427,
+      "grad_norm": 0.4429568946361542,
+      "learning_rate": 8.116542245479165e-06,
+      "loss": 0.3507,
+      "step": 2304
+    },
+    {
+      "epoch": 1.0719268330491396,
+      "grad_norm": 0.49815958738327026,
+      "learning_rate": 8.114425811340635e-06,
+      "loss": 0.3344,
+      "step": 2305
+    },
+    {
+      "epoch": 1.0723918772283367,
+      "grad_norm": 0.41795018315315247,
+      "learning_rate": 8.112308465037375e-06,
+      "loss": 0.3721,
+      "step": 2306
+    },
+    {
+      "epoch": 1.0728569214075336,
+      "grad_norm": 0.5189369916915894,
+      "learning_rate": 8.110190207189519e-06,
+      "loss": 0.3475,
+      "step": 2307
+    },
+    {
+      "epoch": 1.0733219655867308,
+      "grad_norm": 0.5295676589012146,
+      "learning_rate": 8.108071038417471e-06,
+      "loss": 0.378,
+      "step": 2308
+    },
+    {
+      "epoch": 1.0737870097659277,
+      "grad_norm": 0.4301398694515228,
+      "learning_rate": 8.1059509593419e-06,
+      "loss": 0.3511,
+      "step": 2309
+    },
+    {
+      "epoch": 1.0742520539451248,
+      "grad_norm": 0.47263363003730774,
+      "learning_rate": 8.103829970583742e-06,
+      "loss": 0.3233,
+      "step": 2310
+    },
+    {
+      "epoch": 1.0747170981243217,
+      "grad_norm": 0.519551157951355,
+      "learning_rate": 8.101708072764204e-06,
+      "loss": 0.3633,
+      "step": 2311
+    },
+    {
+      "epoch": 1.0751821423035188,
+      "grad_norm": 0.40797778964042664,
+      "learning_rate": 8.099585266504753e-06,
+      "loss": 0.3504,
+      "step": 2312
+    },
+    {
+      "epoch": 1.075647186482716,
+      "grad_norm": 0.4842034578323364,
+      "learning_rate": 8.097461552427123e-06,
+      "loss": 0.3747,
+      "step": 2313
+    },
+    {
+      "epoch": 1.0761122306619129,
+      "grad_norm": 0.5274269580841064,
+      "learning_rate": 8.095336931153318e-06,
+      "loss": 0.3811,
+      "step": 2314
+    },
+    {
+      "epoch": 1.07657727484111,
+      "grad_norm": 0.4279079735279083,
+      "learning_rate": 8.093211403305603e-06,
+      "loss": 0.339,
+      "step": 2315
+    },
+    {
+      "epoch": 1.077042319020307,
+      "grad_norm": 0.4614260792732239,
+      "learning_rate": 8.09108496950651e-06,
+      "loss": 0.3576,
+      "step": 2316
+    },
+    {
+      "epoch": 1.077507363199504,
+      "grad_norm": 0.5641934275627136,
+      "learning_rate": 8.088957630378842e-06,
+      "loss": 0.353,
+      "step": 2317
+    },
+    {
+      "epoch": 1.077972407378701,
+      "grad_norm": 0.5266045928001404,
+      "learning_rate": 8.086829386545655e-06,
+      "loss": 0.3438,
+      "step": 2318
+    },
+    {
+      "epoch": 1.078437451557898,
+      "grad_norm": 0.39795106649398804,
+      "learning_rate": 8.084700238630283e-06,
+      "loss": 0.2865,
+      "step": 2319
+    },
+    {
+      "epoch": 1.078902495737095,
+      "grad_norm": 0.47763675451278687,
+      "learning_rate": 8.082570187256315e-06,
+      "loss": 0.3863,
+      "step": 2320
+    },
+    {
+      "epoch": 1.079367539916292,
+      "grad_norm": 0.43691733479499817,
+      "learning_rate": 8.080439233047612e-06,
+      "loss": 0.3406,
+      "step": 2321
+    },
+    {
+      "epoch": 1.079832584095489,
+      "grad_norm": 0.42221641540527344,
+      "learning_rate": 8.078307376628292e-06,
+      "loss": 0.3743,
+      "step": 2322
+    },
+    {
+      "epoch": 1.0802976282746861,
+      "grad_norm": 0.4146184027194977,
+      "learning_rate": 8.076174618622744e-06,
+      "loss": 0.3584,
+      "step": 2323
+    },
+    {
+      "epoch": 1.080762672453883,
+      "grad_norm": 0.4470728039741516,
+      "learning_rate": 8.074040959655616e-06,
+      "loss": 0.379,
+      "step": 2324
+    },
+    {
+      "epoch": 1.0812277166330801,
+      "grad_norm": 0.512160062789917,
+      "learning_rate": 8.071906400351823e-06,
+      "loss": 0.3425,
+      "step": 2325
+    },
+    {
+      "epoch": 1.081692760812277,
+      "grad_norm": 0.39240655303001404,
+      "learning_rate": 8.069770941336542e-06,
+      "loss": 0.35,
+      "step": 2326
+    },
+    {
+      "epoch": 1.0821578049914742,
+      "grad_norm": 0.4476917088031769,
+      "learning_rate": 8.067634583235215e-06,
+      "loss": 0.3455,
+      "step": 2327
+    },
+    {
+      "epoch": 1.0826228491706713,
+      "grad_norm": 0.5166687965393066,
+      "learning_rate": 8.065497326673548e-06,
+      "loss": 0.4265,
+      "step": 2328
+    },
+    {
+      "epoch": 1.0830878933498682,
+      "grad_norm": 0.4078577160835266,
+      "learning_rate": 8.063359172277507e-06,
+      "loss": 0.3306,
+      "step": 2329
+    },
+    {
+      "epoch": 1.0835529375290653,
+      "grad_norm": 0.4202713072299957,
+      "learning_rate": 8.061220120673323e-06,
+      "loss": 0.3573,
+      "step": 2330
+    },
+    {
+      "epoch": 1.0840179817082622,
+      "grad_norm": 0.4990968108177185,
+      "learning_rate": 8.05908017248749e-06,
+      "loss": 0.3774,
+      "step": 2331
+    },
+    {
+      "epoch": 1.0844830258874594,
+      "grad_norm": 0.4540191888809204,
+      "learning_rate": 8.056939328346763e-06,
+      "loss": 0.3521,
+      "step": 2332
+    },
+    {
+      "epoch": 1.0849480700666563,
+      "grad_norm": 0.3652085065841675,
+      "learning_rate": 8.05479758887816e-06,
+      "loss": 0.2742,
+      "step": 2333
+    },
+    {
+      "epoch": 1.0854131142458534,
+      "grad_norm": 0.5675728917121887,
+      "learning_rate": 8.052654954708966e-06,
+      "loss": 0.3901,
+      "step": 2334
+    },
+    {
+      "epoch": 1.0858781584250503,
+      "grad_norm": 0.46835145354270935,
+      "learning_rate": 8.050511426466717e-06,
+      "loss": 0.3319,
+      "step": 2335
+    },
+    {
+      "epoch": 1.0863432026042474,
+      "grad_norm": 0.4239199459552765,
+      "learning_rate": 8.048367004779223e-06,
+      "loss": 0.3336,
+      "step": 2336
+    },
+    {
+      "epoch": 1.0868082467834443,
+      "grad_norm": 0.5129764080047607,
+      "learning_rate": 8.046221690274547e-06,
+      "loss": 0.3662,
+      "step": 2337
+    },
+    {
+      "epoch": 1.0872732909626415,
+      "grad_norm": 0.46359536051750183,
+      "learning_rate": 8.04407548358102e-06,
+      "loss": 0.3276,
+      "step": 2338
+    },
+    {
+      "epoch": 1.0877383351418384,
+      "grad_norm": 0.4142586290836334,
+      "learning_rate": 8.041928385327229e-06,
+      "loss": 0.3574,
+      "step": 2339
+    },
+    {
+      "epoch": 1.0882033793210355,
+      "grad_norm": 0.4258030951023102,
+      "learning_rate": 8.039780396142023e-06,
+      "loss": 0.334,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0886684235002324,
+      "grad_norm": 0.43229323625564575,
+      "learning_rate": 8.037631516654516e-06,
+      "loss": 0.3329,
+      "step": 2341
+    },
+    {
+      "epoch": 1.0891334676794295,
+      "grad_norm": 0.4798569977283478,
+      "learning_rate": 8.035481747494078e-06,
+      "loss": 0.3859,
+      "step": 2342
+    },
+    {
+      "epoch": 1.0895985118586267,
+      "grad_norm": 0.4725603759288788,
+      "learning_rate": 8.033331089290342e-06,
+      "loss": 0.3613,
+      "step": 2343
+    },
+    {
+      "epoch": 1.0900635560378236,
+      "grad_norm": 0.4565368890762329,
+      "learning_rate": 8.0311795426732e-06,
+      "loss": 0.3214,
+      "step": 2344
+    },
+    {
+      "epoch": 1.0905286002170207,
+      "grad_norm": 0.436138778924942,
+      "learning_rate": 8.029027108272806e-06,
+      "loss": 0.3499,
+      "step": 2345
+    },
+    {
+      "epoch": 1.0909936443962176,
+      "grad_norm": 0.4970560371875763,
+      "learning_rate": 8.026873786719574e-06,
+      "loss": 0.3763,
+      "step": 2346
+    },
+    {
+      "epoch": 1.0914586885754147,
+      "grad_norm": 0.47304481267929077,
+      "learning_rate": 8.024719578644176e-06,
+      "loss": 0.3628,
+      "step": 2347
+    },
+    {
+      "epoch": 1.0919237327546116,
+      "grad_norm": 0.4886552393436432,
+      "learning_rate": 8.022564484677545e-06,
+      "loss": 0.3415,
+      "step": 2348
+    },
+    {
+      "epoch": 1.0923887769338088,
+      "grad_norm": 0.4249359965324402,
+      "learning_rate": 8.020408505450869e-06,
+      "loss": 0.3737,
+      "step": 2349
+    },
+    {
+      "epoch": 1.0928538211130057,
+      "grad_norm": 0.43461641669273376,
+      "learning_rate": 8.018251641595604e-06,
+      "loss": 0.3466,
+      "step": 2350
+    },
+    {
+      "epoch": 1.0933188652922028,
+      "grad_norm": 0.49044400453567505,
+      "learning_rate": 8.016093893743462e-06,
+      "loss": 0.3437,
+      "step": 2351
+    },
+    {
+      "epoch": 1.0937839094713997,
+      "grad_norm": 0.5193617343902588,
+      "learning_rate": 8.013935262526407e-06,
+      "loss": 0.4029,
+      "step": 2352
+    },
+    {
+      "epoch": 1.0942489536505968,
+      "grad_norm": 0.4232480823993683,
+      "learning_rate": 8.01177574857667e-06,
+      "loss": 0.3253,
+      "step": 2353
+    },
+    {
+      "epoch": 1.0947139978297937,
+      "grad_norm": 0.5477461814880371,
+      "learning_rate": 8.009615352526737e-06,
+      "loss": 0.3584,
+      "step": 2354
+    },
+    {
+      "epoch": 1.0951790420089909,
+      "grad_norm": 0.4962575137615204,
+      "learning_rate": 8.007454075009352e-06,
+      "loss": 0.387,
+      "step": 2355
+    },
+    {
+      "epoch": 1.0956440861881878,
+      "grad_norm": 0.4006698727607727,
+      "learning_rate": 8.00529191665752e-06,
+      "loss": 0.3545,
+      "step": 2356
+    },
+    {
+      "epoch": 1.0961091303673849,
+      "grad_norm": 0.44313761591911316,
+      "learning_rate": 8.0031288781045e-06,
+      "loss": 0.3291,
+      "step": 2357
+    },
+    {
+      "epoch": 1.096574174546582,
+      "grad_norm": 0.525131344795227,
+      "learning_rate": 8.000964959983815e-06,
+      "loss": 0.3665,
+      "step": 2358
+    },
+    {
+      "epoch": 1.097039218725779,
+      "grad_norm": 0.44491708278656006,
+      "learning_rate": 7.998800162929236e-06,
+      "loss": 0.3512,
+      "step": 2359
+    },
+    {
+      "epoch": 1.097504262904976,
+      "grad_norm": 0.5135394334793091,
+      "learning_rate": 7.9966344875748e-06,
+      "loss": 0.3638,
+      "step": 2360
+    },
+    {
+      "epoch": 1.097969307084173,
+      "grad_norm": 0.5230858325958252,
+      "learning_rate": 7.994467934554794e-06,
+      "loss": 0.3372,
+      "step": 2361
+    },
+    {
+      "epoch": 1.09843435126337,
+      "grad_norm": 0.5185748934745789,
+      "learning_rate": 7.992300504503774e-06,
+      "loss": 0.395,
+      "step": 2362
+    },
+    {
+      "epoch": 1.098899395442567,
+      "grad_norm": 0.440336138010025,
+      "learning_rate": 7.990132198056538e-06,
+      "loss": 0.3397,
+      "step": 2363
+    },
+    {
+      "epoch": 1.0993644396217641,
+      "grad_norm": 0.539335310459137,
+      "learning_rate": 7.987963015848152e-06,
+      "loss": 0.3714,
+      "step": 2364
+    },
+    {
+      "epoch": 1.099829483800961,
+      "grad_norm": 0.4946892559528351,
+      "learning_rate": 7.985792958513932e-06,
+      "loss": 0.3258,
+      "step": 2365
+    },
+    {
+      "epoch": 1.1002945279801581,
+      "grad_norm": 0.4519208073616028,
+      "learning_rate": 7.983622026689452e-06,
+      "loss": 0.3841,
+      "step": 2366
+    },
+    {
+      "epoch": 1.100759572159355,
+      "grad_norm": 0.4441494345664978,
+      "learning_rate": 7.981450221010547e-06,
+      "loss": 0.3603,
+      "step": 2367
+    },
+    {
+      "epoch": 1.1012246163385522,
+      "grad_norm": 0.545825183391571,
+      "learning_rate": 7.979277542113297e-06,
+      "loss": 0.3385,
+      "step": 2368
+    },
+    {
+      "epoch": 1.1016896605177493,
+      "grad_norm": 0.46227794885635376,
+      "learning_rate": 7.97710399063405e-06,
+      "loss": 0.3456,
+      "step": 2369
+    },
+    {
+      "epoch": 1.1021547046969462,
+      "grad_norm": 0.405261754989624,
+      "learning_rate": 7.974929567209399e-06,
+      "loss": 0.3543,
+      "step": 2370
+    },
+    {
+      "epoch": 1.1026197488761433,
+      "grad_norm": 0.5703674554824829,
+      "learning_rate": 7.972754272476203e-06,
+      "loss": 0.3407,
+      "step": 2371
+    },
+    {
+      "epoch": 1.1030847930553402,
+      "grad_norm": 0.5644888877868652,
+      "learning_rate": 7.970578107071566e-06,
+      "loss": 0.4069,
+      "step": 2372
+    },
+    {
+      "epoch": 1.1035498372345374,
+      "grad_norm": 0.391647070646286,
+      "learning_rate": 7.968401071632854e-06,
+      "loss": 0.3582,
+      "step": 2373
+    },
+    {
+      "epoch": 1.1040148814137343,
+      "grad_norm": 0.6647116541862488,
+      "learning_rate": 7.966223166797684e-06,
+      "loss": 0.3614,
+      "step": 2374
+    },
+    {
+      "epoch": 1.1044799255929314,
+      "grad_norm": 0.4719175696372986,
+      "learning_rate": 7.964044393203928e-06,
+      "loss": 0.3655,
+      "step": 2375
+    },
+    {
+      "epoch": 1.1049449697721283,
+      "grad_norm": 0.43519163131713867,
+      "learning_rate": 7.961864751489717e-06,
+      "loss": 0.3422,
+      "step": 2376
+    },
+    {
+      "epoch": 1.1054100139513254,
+      "grad_norm": 0.5003102421760559,
+      "learning_rate": 7.959684242293428e-06,
+      "loss": 0.326,
+      "step": 2377
+    },
+    {
+      "epoch": 1.1058750581305223,
+      "grad_norm": 0.5152011513710022,
+      "learning_rate": 7.957502866253699e-06,
+      "loss": 0.313,
+      "step": 2378
+    },
+    {
+      "epoch": 1.1063401023097195,
+      "grad_norm": 0.46767157316207886,
+      "learning_rate": 7.955320624009421e-06,
+      "loss": 0.374,
+      "step": 2379
+    },
+    {
+      "epoch": 1.1068051464889164,
+      "grad_norm": 0.4796226918697357,
+      "learning_rate": 7.953137516199737e-06,
+      "loss": 0.394,
+      "step": 2380
+    },
+    {
+      "epoch": 1.1072701906681135,
+      "grad_norm": 0.47181200981140137,
+      "learning_rate": 7.950953543464039e-06,
+      "loss": 0.3253,
+      "step": 2381
+    },
+    {
+      "epoch": 1.1077352348473104,
+      "grad_norm": 0.48946288228034973,
+      "learning_rate": 7.948768706441985e-06,
+      "loss": 0.3536,
+      "step": 2382
+    },
+    {
+      "epoch": 1.1082002790265075,
+      "grad_norm": 0.43955692648887634,
+      "learning_rate": 7.946583005773471e-06,
+      "loss": 0.3505,
+      "step": 2383
+    },
+    {
+      "epoch": 1.1086653232057047,
+      "grad_norm": 0.5506858825683594,
+      "learning_rate": 7.944396442098659e-06,
+      "loss": 0.3846,
+      "step": 2384
+    },
+    {
+      "epoch": 1.1091303673849016,
+      "grad_norm": 0.38817930221557617,
+      "learning_rate": 7.942209016057954e-06,
+      "loss": 0.3297,
+      "step": 2385
+    },
+    {
+      "epoch": 1.1095954115640987,
+      "grad_norm": 0.4331190288066864,
+      "learning_rate": 7.94002072829202e-06,
+      "loss": 0.3632,
+      "step": 2386
+    },
+    {
+      "epoch": 1.1100604557432956,
+      "grad_norm": 0.4732077419757843,
+      "learning_rate": 7.937831579441768e-06,
+      "loss": 0.3593,
+      "step": 2387
+    },
+    {
+      "epoch": 1.1105254999224927,
+      "grad_norm": 0.3843960464000702,
+      "learning_rate": 7.935641570148368e-06,
+      "loss": 0.3042,
+      "step": 2388
+    },
+    {
+      "epoch": 1.1109905441016896,
+      "grad_norm": 0.43816637992858887,
+      "learning_rate": 7.933450701053235e-06,
+      "loss": 0.3798,
+      "step": 2389
+    },
+    {
+      "epoch": 1.1114555882808868,
+      "grad_norm": 0.5748831033706665,
+      "learning_rate": 7.931258972798041e-06,
+      "loss": 0.3534,
+      "step": 2390
+    },
+    {
+      "epoch": 1.1119206324600837,
+      "grad_norm": 0.48593536019325256,
+      "learning_rate": 7.929066386024707e-06,
+      "loss": 0.3994,
+      "step": 2391
+    },
+    {
+      "epoch": 1.1123856766392808,
+      "grad_norm": 0.5419626832008362,
+      "learning_rate": 7.926872941375404e-06,
+      "loss": 0.331,
+      "step": 2392
+    },
+    {
+      "epoch": 1.1128507208184777,
+      "grad_norm": 0.5766515731811523,
+      "learning_rate": 7.924678639492559e-06,
+      "loss": 0.3805,
+      "step": 2393
+    },
+    {
+      "epoch": 1.1133157649976748,
+      "grad_norm": 0.4355737268924713,
+      "learning_rate": 7.922483481018848e-06,
+      "loss": 0.3525,
+      "step": 2394
+    },
+    {
+      "epoch": 1.1137808091768717,
+      "grad_norm": 0.5494896173477173,
+      "learning_rate": 7.920287466597193e-06,
+      "loss": 0.3321,
+      "step": 2395
+    },
+    {
+      "epoch": 1.1142458533560688,
+      "grad_norm": 0.5438748598098755,
+      "learning_rate": 7.918090596870776e-06,
+      "loss": 0.3795,
+      "step": 2396
+    },
+    {
+      "epoch": 1.1147108975352658,
+      "grad_norm": 0.4230845868587494,
+      "learning_rate": 7.915892872483023e-06,
+      "loss": 0.3506,
+      "step": 2397
+    },
+    {
+      "epoch": 1.1151759417144629,
+      "grad_norm": 0.6087936162948608,
+      "learning_rate": 7.913694294077607e-06,
+      "loss": 0.3583,
+      "step": 2398
+    },
+    {
+      "epoch": 1.11564098589366,
+      "grad_norm": 0.5080538988113403,
+      "learning_rate": 7.911494862298464e-06,
+      "loss": 0.3579,
+      "step": 2399
+    },
+    {
+      "epoch": 1.116106030072857,
+      "grad_norm": 0.5257863402366638,
+      "learning_rate": 7.909294577789765e-06,
+      "loss": 0.3751,
+      "step": 2400
+    },
+    {
+      "epoch": 1.116571074252054,
+      "grad_norm": 0.4901667833328247,
+      "learning_rate": 7.90709344119594e-06,
+      "loss": 0.3494,
+      "step": 2401
+    },
+    {
+      "epoch": 1.117036118431251,
+      "grad_norm": 0.5573136210441589,
+      "learning_rate": 7.90489145316167e-06,
+      "loss": 0.3318,
+      "step": 2402
+    },
+    {
+      "epoch": 1.117501162610448,
+      "grad_norm": 0.5585957765579224,
+      "learning_rate": 7.902688614331875e-06,
+      "loss": 0.3737,
+      "step": 2403
+    },
+    {
+      "epoch": 1.117966206789645,
+      "grad_norm": 0.5125075578689575,
+      "learning_rate": 7.900484925351734e-06,
+      "loss": 0.4047,
+      "step": 2404
+    },
+    {
+      "epoch": 1.118431250968842,
+      "grad_norm": 0.45521363615989685,
+      "learning_rate": 7.898280386866673e-06,
+      "loss": 0.321,
+      "step": 2405
+    },
+    {
+      "epoch": 1.118896295148039,
+      "grad_norm": 0.5232580304145813,
+      "learning_rate": 7.896074999522362e-06,
+      "loss": 0.3717,
+      "step": 2406
+    },
+    {
+      "epoch": 1.1193613393272361,
+      "grad_norm": 0.395935982465744,
+      "learning_rate": 7.893868763964724e-06,
+      "loss": 0.3292,
+      "step": 2407
+    },
+    {
+      "epoch": 1.119826383506433,
+      "grad_norm": 0.5463009476661682,
+      "learning_rate": 7.891661680839932e-06,
+      "loss": 0.373,
+      "step": 2408
+    },
+    {
+      "epoch": 1.1202914276856302,
+      "grad_norm": 0.44592776894569397,
+      "learning_rate": 7.889453750794405e-06,
+      "loss": 0.3653,
+      "step": 2409
+    },
+    {
+      "epoch": 1.120756471864827,
+      "grad_norm": 0.45497268438339233,
+      "learning_rate": 7.887244974474807e-06,
+      "loss": 0.3252,
+      "step": 2410
+    },
+    {
+      "epoch": 1.1212215160440242,
+      "grad_norm": 0.4712156653404236,
+      "learning_rate": 7.885035352528054e-06,
+      "loss": 0.3514,
+      "step": 2411
+    },
+    {
+      "epoch": 1.121686560223221,
+      "grad_norm": 0.4327942430973053,
+      "learning_rate": 7.882824885601308e-06,
+      "loss": 0.3591,
+      "step": 2412
+    },
+    {
+      "epoch": 1.1221516044024182,
+      "grad_norm": 0.3762139678001404,
+      "learning_rate": 7.88061357434198e-06,
+      "loss": 0.3263,
+      "step": 2413
+    },
+    {
+      "epoch": 1.1226166485816154,
+      "grad_norm": 0.5221520662307739,
+      "learning_rate": 7.878401419397725e-06,
+      "loss": 0.3907,
+      "step": 2414
+    },
+    {
+      "epoch": 1.1230816927608123,
+      "grad_norm": 0.41757580637931824,
+      "learning_rate": 7.87618842141645e-06,
+      "loss": 0.3556,
+      "step": 2415
+    },
+    {
+      "epoch": 1.1235467369400094,
+      "grad_norm": 0.4390140771865845,
+      "learning_rate": 7.873974581046303e-06,
+      "loss": 0.3543,
+      "step": 2416
+    },
+    {
+      "epoch": 1.1240117811192063,
+      "grad_norm": 0.40529686212539673,
+      "learning_rate": 7.871759898935685e-06,
+      "loss": 0.3468,
+      "step": 2417
+    },
+    {
+      "epoch": 1.1244768252984034,
+      "grad_norm": 0.4217506945133209,
+      "learning_rate": 7.86954437573324e-06,
+      "loss": 0.3467,
+      "step": 2418
+    },
+    {
+      "epoch": 1.1249418694776003,
+      "grad_norm": 0.418305367231369,
+      "learning_rate": 7.867328012087856e-06,
+      "loss": 0.3951,
+      "step": 2419
+    },
+    {
+      "epoch": 1.1254069136567975,
+      "grad_norm": 0.42547518014907837,
+      "learning_rate": 7.865110808648671e-06,
+      "loss": 0.3369,
+      "step": 2420
+    },
+    {
+      "epoch": 1.1258719578359944,
+      "grad_norm": 0.41389068961143494,
+      "learning_rate": 7.862892766065072e-06,
+      "loss": 0.3423,
+      "step": 2421
+    },
+    {
+      "epoch": 1.1263370020151915,
+      "grad_norm": 0.47316908836364746,
+      "learning_rate": 7.86067388498668e-06,
+      "loss": 0.375,
+      "step": 2422
+    },
+    {
+      "epoch": 1.1268020461943884,
+      "grad_norm": 0.4010957181453705,
+      "learning_rate": 7.858454166063376e-06,
+      "loss": 0.3348,
+      "step": 2423
+    },
+    {
+      "epoch": 1.1272670903735855,
+      "grad_norm": 0.4541621804237366,
+      "learning_rate": 7.856233609945276e-06,
+      "loss": 0.3627,
+      "step": 2424
+    },
+    {
+      "epoch": 1.1277321345527824,
+      "grad_norm": 0.4253348410129547,
+      "learning_rate": 7.854012217282747e-06,
+      "loss": 0.3391,
+      "step": 2425
+    },
+    {
+      "epoch": 1.1281971787319796,
+      "grad_norm": 0.4224770963191986,
+      "learning_rate": 7.851789988726397e-06,
+      "loss": 0.3227,
+      "step": 2426
+    },
+    {
+      "epoch": 1.1286622229111765,
+      "grad_norm": 0.4292997121810913,
+      "learning_rate": 7.849566924927082e-06,
+      "loss": 0.3815,
+      "step": 2427
+    },
+    {
+      "epoch": 1.1291272670903736,
+      "grad_norm": 0.39106887578964233,
+      "learning_rate": 7.8473430265359e-06,
+      "loss": 0.3313,
+      "step": 2428
+    },
+    {
+      "epoch": 1.1295923112695707,
+      "grad_norm": 0.41027796268463135,
+      "learning_rate": 7.845118294204195e-06,
+      "loss": 0.3478,
+      "step": 2429
+    },
+    {
+      "epoch": 1.1300573554487676,
+      "grad_norm": 0.454773873090744,
+      "learning_rate": 7.842892728583557e-06,
+      "loss": 0.3566,
+      "step": 2430
+    },
+    {
+      "epoch": 1.1305223996279647,
+      "grad_norm": 0.4108147919178009,
+      "learning_rate": 7.840666330325815e-06,
+      "loss": 0.338,
+      "step": 2431
+    },
+    {
+      "epoch": 1.1309874438071617,
+      "grad_norm": 0.41285791993141174,
+      "learning_rate": 7.838439100083048e-06,
+      "loss": 0.372,
+      "step": 2432
+    },
+    {
+      "epoch": 1.1314524879863588,
+      "grad_norm": 0.44633156061172485,
+      "learning_rate": 7.836211038507571e-06,
+      "loss": 0.341,
+      "step": 2433
+    },
+    {
+      "epoch": 1.1319175321655557,
+      "grad_norm": 0.3967339098453522,
+      "learning_rate": 7.833982146251952e-06,
+      "loss": 0.3189,
+      "step": 2434
+    },
+    {
+      "epoch": 1.1323825763447528,
+      "grad_norm": 0.4380015432834625,
+      "learning_rate": 7.831752423968995e-06,
+      "loss": 0.3506,
+      "step": 2435
+    },
+    {
+      "epoch": 1.1328476205239497,
+      "grad_norm": 0.5487619638442993,
+      "learning_rate": 7.829521872311747e-06,
+      "loss": 0.3675,
+      "step": 2436
+    },
+    {
+      "epoch": 1.1333126647031468,
+      "grad_norm": 0.4468625783920288,
+      "learning_rate": 7.827290491933506e-06,
+      "loss": 0.3792,
+      "step": 2437
+    },
+    {
+      "epoch": 1.1337777088823437,
+      "grad_norm": 0.3934735059738159,
+      "learning_rate": 7.825058283487803e-06,
+      "loss": 0.3421,
+      "step": 2438
+    },
+    {
+      "epoch": 1.1342427530615409,
+      "grad_norm": 0.49053192138671875,
+      "learning_rate": 7.822825247628416e-06,
+      "loss": 0.3482,
+      "step": 2439
+    },
+    {
+      "epoch": 1.1347077972407378,
+      "grad_norm": 0.512285053730011,
+      "learning_rate": 7.820591385009366e-06,
+      "loss": 0.3616,
+      "step": 2440
+    },
+    {
+      "epoch": 1.135172841419935,
+      "grad_norm": 0.49586769938468933,
+      "learning_rate": 7.818356696284916e-06,
+      "loss": 0.3796,
+      "step": 2441
+    },
+    {
+      "epoch": 1.1356378855991318,
+      "grad_norm": 0.4258926808834076,
+      "learning_rate": 7.816121182109567e-06,
+      "loss": 0.332,
+      "step": 2442
+    },
+    {
+      "epoch": 1.136102929778329,
+      "grad_norm": 0.45421749353408813,
+      "learning_rate": 7.813884843138067e-06,
+      "loss": 0.3504,
+      "step": 2443
+    },
+    {
+      "epoch": 1.136567973957526,
+      "grad_norm": 0.45513275265693665,
+      "learning_rate": 7.811647680025403e-06,
+      "loss": 0.3345,
+      "step": 2444
+    },
+    {
+      "epoch": 1.137033018136723,
+      "grad_norm": 0.4567432403564453,
+      "learning_rate": 7.809409693426803e-06,
+      "loss": 0.3494,
+      "step": 2445
+    },
+    {
+      "epoch": 1.13749806231592,
+      "grad_norm": 0.4656786620616913,
+      "learning_rate": 7.807170883997738e-06,
+      "loss": 0.3428,
+      "step": 2446
+    },
+    {
+      "epoch": 1.137963106495117,
+      "grad_norm": 0.44131651520729065,
+      "learning_rate": 7.804931252393918e-06,
+      "loss": 0.3475,
+      "step": 2447
+    },
+    {
+      "epoch": 1.1384281506743141,
+      "grad_norm": 0.42925527691841125,
+      "learning_rate": 7.802690799271295e-06,
+      "loss": 0.3882,
+      "step": 2448
+    },
+    {
+      "epoch": 1.138893194853511,
+      "grad_norm": 0.4417758584022522,
+      "learning_rate": 7.800449525286062e-06,
+      "loss": 0.3428,
+      "step": 2449
+    },
+    {
+      "epoch": 1.1393582390327082,
+      "grad_norm": 0.4203170835971832,
+      "learning_rate": 7.79820743109465e-06,
+      "loss": 0.3425,
+      "step": 2450
+    },
+    {
+      "epoch": 1.139823283211905,
+      "grad_norm": 0.45581138134002686,
+      "learning_rate": 7.795964517353734e-06,
+      "loss": 0.3596,
+      "step": 2451
+    },
+    {
+      "epoch": 1.1402883273911022,
+      "grad_norm": 0.44257980585098267,
+      "learning_rate": 7.793720784720227e-06,
+      "loss": 0.3394,
+      "step": 2452
+    },
+    {
+      "epoch": 1.140753371570299,
+      "grad_norm": 0.40145984292030334,
+      "learning_rate": 7.791476233851281e-06,
+      "loss": 0.3502,
+      "step": 2453
+    },
+    {
+      "epoch": 1.1412184157494962,
+      "grad_norm": 0.48211121559143066,
+      "learning_rate": 7.789230865404287e-06,
+      "loss": 0.3394,
+      "step": 2454
+    },
+    {
+      "epoch": 1.1416834599286931,
+      "grad_norm": 0.4565165340900421,
+      "learning_rate": 7.78698468003688e-06,
+      "loss": 0.3687,
+      "step": 2455
+    },
+    {
+      "epoch": 1.1421485041078903,
+      "grad_norm": 0.4464455842971802,
+      "learning_rate": 7.784737678406929e-06,
+      "loss": 0.3255,
+      "step": 2456
+    },
+    {
+      "epoch": 1.1426135482870872,
+      "grad_norm": 0.6133958101272583,
+      "learning_rate": 7.782489861172545e-06,
+      "loss": 0.3996,
+      "step": 2457
+    },
+    {
+      "epoch": 1.1430785924662843,
+      "grad_norm": 0.3757973909378052,
+      "learning_rate": 7.780241228992075e-06,
+      "loss": 0.3129,
+      "step": 2458
+    },
+    {
+      "epoch": 1.1435436366454814,
+      "grad_norm": 0.5780712366104126,
+      "learning_rate": 7.777991782524112e-06,
+      "loss": 0.3765,
+      "step": 2459
+    },
+    {
+      "epoch": 1.1440086808246783,
+      "grad_norm": 0.5257070660591125,
+      "learning_rate": 7.775741522427477e-06,
+      "loss": 0.3453,
+      "step": 2460
+    },
+    {
+      "epoch": 1.1444737250038755,
+      "grad_norm": 0.49410927295684814,
+      "learning_rate": 7.773490449361238e-06,
+      "loss": 0.3747,
+      "step": 2461
+    },
+    {
+      "epoch": 1.1449387691830724,
+      "grad_norm": 0.46261581778526306,
+      "learning_rate": 7.771238563984696e-06,
+      "loss": 0.3363,
+      "step": 2462
+    },
+    {
+      "epoch": 1.1454038133622695,
+      "grad_norm": 0.6092031598091125,
+      "learning_rate": 7.768985866957392e-06,
+      "loss": 0.3472,
+      "step": 2463
+    },
+    {
+      "epoch": 1.1458688575414664,
+      "grad_norm": 0.42954081296920776,
+      "learning_rate": 7.766732358939106e-06,
+      "loss": 0.3336,
+      "step": 2464
+    },
+    {
+      "epoch": 1.1463339017206635,
+      "grad_norm": 0.4083530604839325,
+      "learning_rate": 7.764478040589854e-06,
+      "loss": 0.3561,
+      "step": 2465
+    },
+    {
+      "epoch": 1.1467989458998604,
+      "grad_norm": 0.46050816774368286,
+      "learning_rate": 7.762222912569886e-06,
+      "loss": 0.3711,
+      "step": 2466
+    },
+    {
+      "epoch": 1.1472639900790575,
+      "grad_norm": 0.460467129945755,
+      "learning_rate": 7.759966975539693e-06,
+      "loss": 0.3107,
+      "step": 2467
+    },
+    {
+      "epoch": 1.1477290342582545,
+      "grad_norm": 0.5141720771789551,
+      "learning_rate": 7.757710230160003e-06,
+      "loss": 0.3846,
+      "step": 2468
+    },
+    {
+      "epoch": 1.1481940784374516,
+      "grad_norm": 0.45427727699279785,
+      "learning_rate": 7.755452677091783e-06,
+      "loss": 0.3422,
+      "step": 2469
+    },
+    {
+      "epoch": 1.1486591226166485,
+      "grad_norm": 0.47798460721969604,
+      "learning_rate": 7.75319431699623e-06,
+      "loss": 0.3462,
+      "step": 2470
+    },
+    {
+      "epoch": 1.1491241667958456,
+      "grad_norm": 0.4749840497970581,
+      "learning_rate": 7.750935150534781e-06,
+      "loss": 0.3547,
+      "step": 2471
+    },
+    {
+      "epoch": 1.1495892109750425,
+      "grad_norm": 0.46182170510292053,
+      "learning_rate": 7.748675178369112e-06,
+      "loss": 0.395,
+      "step": 2472
+    },
+    {
+      "epoch": 1.1500542551542396,
+      "grad_norm": 0.504657506942749,
+      "learning_rate": 7.74641440116113e-06,
+      "loss": 0.3707,
+      "step": 2473
+    },
+    {
+      "epoch": 1.1505192993334368,
+      "grad_norm": 0.4313875436782837,
+      "learning_rate": 7.74415281957298e-06,
+      "loss": 0.3357,
+      "step": 2474
+    },
+    {
+      "epoch": 1.1509843435126337,
+      "grad_norm": 0.5049738883972168,
+      "learning_rate": 7.741890434267043e-06,
+      "loss": 0.3556,
+      "step": 2475
+    },
+    {
+      "epoch": 1.1514493876918308,
+      "grad_norm": 0.4582679569721222,
+      "learning_rate": 7.739627245905935e-06,
+      "loss": 0.3233,
+      "step": 2476
+    },
+    {
+      "epoch": 1.1519144318710277,
+      "grad_norm": 0.463096559047699,
+      "learning_rate": 7.737363255152506e-06,
+      "loss": 0.3661,
+      "step": 2477
+    },
+    {
+      "epoch": 1.1523794760502248,
+      "grad_norm": 0.4227137267589569,
+      "learning_rate": 7.735098462669843e-06,
+      "loss": 0.3238,
+      "step": 2478
+    },
+    {
+      "epoch": 1.1528445202294217,
+      "grad_norm": 0.5194125771522522,
+      "learning_rate": 7.732832869121267e-06,
+      "loss": 0.3713,
+      "step": 2479
+    },
+    {
+      "epoch": 1.1533095644086189,
+      "grad_norm": 0.5065972805023193,
+      "learning_rate": 7.730566475170334e-06,
+      "loss": 0.3743,
+      "step": 2480
+    },
+    {
+      "epoch": 1.1537746085878158,
+      "grad_norm": 0.4301440715789795,
+      "learning_rate": 7.728299281480833e-06,
+      "loss": 0.3142,
+      "step": 2481
+    },
+    {
+      "epoch": 1.154239652767013,
+      "grad_norm": 0.4768223464488983,
+      "learning_rate": 7.726031288716789e-06,
+      "loss": 0.3369,
+      "step": 2482
+    },
+    {
+      "epoch": 1.1547046969462098,
+      "grad_norm": 0.5396889448165894,
+      "learning_rate": 7.723762497542459e-06,
+      "loss": 0.383,
+      "step": 2483
+    },
+    {
+      "epoch": 1.155169741125407,
+      "grad_norm": 0.4935142993927002,
+      "learning_rate": 7.72149290862234e-06,
+      "loss": 0.3244,
+      "step": 2484
+    },
+    {
+      "epoch": 1.155634785304604,
+      "grad_norm": 0.428690642118454,
+      "learning_rate": 7.719222522621149e-06,
+      "loss": 0.3658,
+      "step": 2485
+    },
+    {
+      "epoch": 1.156099829483801,
+      "grad_norm": 0.4382781982421875,
+      "learning_rate": 7.716951340203851e-06,
+      "loss": 0.3524,
+      "step": 2486
+    },
+    {
+      "epoch": 1.1565648736629979,
+      "grad_norm": 0.594252347946167,
+      "learning_rate": 7.714679362035638e-06,
+      "loss": 0.3808,
+      "step": 2487
+    },
+    {
+      "epoch": 1.157029917842195,
+      "grad_norm": 0.4175407290458679,
+      "learning_rate": 7.712406588781935e-06,
+      "loss": 0.3164,
+      "step": 2488
+    },
+    {
+      "epoch": 1.1574949620213921,
+      "grad_norm": 0.4126685559749603,
+      "learning_rate": 7.7101330211084e-06,
+      "loss": 0.3416,
+      "step": 2489
+    },
+    {
+      "epoch": 1.157960006200589,
+      "grad_norm": 0.5235841274261475,
+      "learning_rate": 7.707858659680924e-06,
+      "loss": 0.3892,
+      "step": 2490
+    },
+    {
+      "epoch": 1.1584250503797862,
+      "grad_norm": 0.3950689733028412,
+      "learning_rate": 7.70558350516563e-06,
+      "loss": 0.3272,
+      "step": 2491
+    },
+    {
+      "epoch": 1.158890094558983,
+      "grad_norm": 0.4256301820278168,
+      "learning_rate": 7.703307558228875e-06,
+      "loss": 0.3423,
+      "step": 2492
+    },
+    {
+      "epoch": 1.1593551387381802,
+      "grad_norm": 0.4690292775630951,
+      "learning_rate": 7.701030819537248e-06,
+      "loss": 0.333,
+      "step": 2493
+    },
+    {
+      "epoch": 1.159820182917377,
+      "grad_norm": 0.4405955672264099,
+      "learning_rate": 7.698753289757565e-06,
+      "loss": 0.3981,
+      "step": 2494
+    },
+    {
+      "epoch": 1.1602852270965742,
+      "grad_norm": 0.5139094591140747,
+      "learning_rate": 7.69647496955688e-06,
+      "loss": 0.35,
+      "step": 2495
+    },
+    {
+      "epoch": 1.1607502712757711,
+      "grad_norm": 0.39744675159454346,
+      "learning_rate": 7.694195859602475e-06,
+      "loss": 0.3465,
+      "step": 2496
+    },
+    {
+      "epoch": 1.1612153154549683,
+      "grad_norm": 0.4022337794303894,
+      "learning_rate": 7.691915960561869e-06,
+      "loss": 0.3574,
+      "step": 2497
+    },
+    {
+      "epoch": 1.1616803596341652,
+      "grad_norm": 0.3752380907535553,
+      "learning_rate": 7.6896352731028e-06,
+      "loss": 0.3333,
+      "step": 2498
+    },
+    {
+      "epoch": 1.1621454038133623,
+      "grad_norm": 0.48939090967178345,
+      "learning_rate": 7.687353797893249e-06,
+      "loss": 0.3831,
+      "step": 2499
+    },
+    {
+      "epoch": 1.1626104479925594,
+      "grad_norm": 0.4494510591030121,
+      "learning_rate": 7.68507153560142e-06,
+      "loss": 0.3882,
+      "step": 2500
+    },
+    {
+      "epoch": 1.1630754921717563,
+      "grad_norm": 0.38291284441947937,
+      "learning_rate": 7.682788486895754e-06,
+      "loss": 0.3589,
+      "step": 2501
+    },
+    {
+      "epoch": 1.1635405363509532,
+      "grad_norm": 0.383709192276001,
+      "learning_rate": 7.680504652444917e-06,
+      "loss": 0.37,
+      "step": 2502
+    },
+    {
+      "epoch": 1.1640055805301504,
+      "grad_norm": 0.507161557674408,
+      "learning_rate": 7.678220032917806e-06,
+      "loss": 0.3843,
+      "step": 2503
+    },
+    {
+      "epoch": 1.1644706247093475,
+      "grad_norm": 0.3985297977924347,
+      "learning_rate": 7.675934628983551e-06,
+      "loss": 0.3142,
+      "step": 2504
+    },
+    {
+      "epoch": 1.1649356688885444,
+      "grad_norm": 0.41284704208374023,
+      "learning_rate": 7.67364844131151e-06,
+      "loss": 0.3457,
+      "step": 2505
+    },
+    {
+      "epoch": 1.1654007130677415,
+      "grad_norm": 0.49253928661346436,
+      "learning_rate": 7.671361470571265e-06,
+      "loss": 0.3974,
+      "step": 2506
+    },
+    {
+      "epoch": 1.1658657572469384,
+      "grad_norm": 0.41742268204689026,
+      "learning_rate": 7.669073717432641e-06,
+      "loss": 0.346,
+      "step": 2507
+    },
+    {
+      "epoch": 1.1663308014261355,
+      "grad_norm": 0.4668549597263336,
+      "learning_rate": 7.666785182565676e-06,
+      "loss": 0.3568,
+      "step": 2508
+    },
+    {
+      "epoch": 1.1667958456053324,
+      "grad_norm": 0.43478819727897644,
+      "learning_rate": 7.66449586664065e-06,
+      "loss": 0.2967,
+      "step": 2509
+    },
+    {
+      "epoch": 1.1672608897845296,
+      "grad_norm": 0.5247063636779785,
+      "learning_rate": 7.662205770328064e-06,
+      "loss": 0.4326,
+      "step": 2510
+    },
+    {
+      "epoch": 1.1677259339637265,
+      "grad_norm": 0.3936578333377838,
+      "learning_rate": 7.65991489429865e-06,
+      "loss": 0.3177,
+      "step": 2511
+    },
+    {
+      "epoch": 1.1681909781429236,
+      "grad_norm": 0.43789342045783997,
+      "learning_rate": 7.65762323922337e-06,
+      "loss": 0.3701,
+      "step": 2512
+    },
+    {
+      "epoch": 1.1686560223221205,
+      "grad_norm": 0.46906614303588867,
+      "learning_rate": 7.655330805773411e-06,
+      "loss": 0.3368,
+      "step": 2513
+    },
+    {
+      "epoch": 1.1691210665013176,
+      "grad_norm": 0.4493185877799988,
+      "learning_rate": 7.653037594620189e-06,
+      "loss": 0.3654,
+      "step": 2514
+    },
+    {
+      "epoch": 1.1695861106805148,
+      "grad_norm": 0.40783825516700745,
+      "learning_rate": 7.650743606435352e-06,
+      "loss": 0.3143,
+      "step": 2515
+    },
+    {
+      "epoch": 1.1700511548597117,
+      "grad_norm": 0.5146582722663879,
+      "learning_rate": 7.648448841890765e-06,
+      "loss": 0.3812,
+      "step": 2516
+    },
+    {
+      "epoch": 1.1705161990389086,
+      "grad_norm": 0.4294111430644989,
+      "learning_rate": 7.646153301658534e-06,
+      "loss": 0.3192,
+      "step": 2517
+    },
+    {
+      "epoch": 1.1709812432181057,
+      "grad_norm": 0.4016132354736328,
+      "learning_rate": 7.643856986410983e-06,
+      "loss": 0.3582,
+      "step": 2518
+    },
+    {
+      "epoch": 1.1714462873973028,
+      "grad_norm": 0.4334622919559479,
+      "learning_rate": 7.641559896820664e-06,
+      "loss": 0.3611,
+      "step": 2519
+    },
+    {
+      "epoch": 1.1719113315764997,
+      "grad_norm": 0.4894271492958069,
+      "learning_rate": 7.63926203356036e-06,
+      "loss": 0.3673,
+      "step": 2520
+    },
+    {
+      "epoch": 1.1723763757556969,
+      "grad_norm": 0.45334476232528687,
+      "learning_rate": 7.636963397303074e-06,
+      "loss": 0.3622,
+      "step": 2521
+    },
+    {
+      "epoch": 1.1728414199348938,
+      "grad_norm": 0.4029448926448822,
+      "learning_rate": 7.634663988722044e-06,
+      "loss": 0.3332,
+      "step": 2522
+    },
+    {
+      "epoch": 1.173306464114091,
+      "grad_norm": 0.4322480261325836,
+      "learning_rate": 7.632363808490726e-06,
+      "loss": 0.3174,
+      "step": 2523
+    },
+    {
+      "epoch": 1.1737715082932878,
+      "grad_norm": 0.4665205180644989,
+      "learning_rate": 7.630062857282806e-06,
+      "loss": 0.3799,
+      "step": 2524
+    },
+    {
+      "epoch": 1.174236552472485,
+      "grad_norm": 0.438932329416275,
+      "learning_rate": 7.627761135772196e-06,
+      "loss": 0.3446,
+      "step": 2525
+    },
+    {
+      "epoch": 1.1747015966516818,
+      "grad_norm": 0.40907543897628784,
+      "learning_rate": 7.625458644633032e-06,
+      "loss": 0.3606,
+      "step": 2526
+    },
+    {
+      "epoch": 1.175166640830879,
+      "grad_norm": 0.39875200390815735,
+      "learning_rate": 7.623155384539678e-06,
+      "loss": 0.315,
+      "step": 2527
+    },
+    {
+      "epoch": 1.1756316850100759,
+      "grad_norm": 0.39767640829086304,
+      "learning_rate": 7.6208513561667184e-06,
+      "loss": 0.3272,
+      "step": 2528
+    },
+    {
+      "epoch": 1.176096729189273,
+      "grad_norm": 0.4325427711009979,
+      "learning_rate": 7.618546560188968e-06,
+      "loss": 0.3653,
+      "step": 2529
+    },
+    {
+      "epoch": 1.1765617733684701,
+      "grad_norm": 0.43984878063201904,
+      "learning_rate": 7.616240997281465e-06,
+      "loss": 0.3544,
+      "step": 2530
+    },
+    {
+      "epoch": 1.177026817547667,
+      "grad_norm": 0.42820581793785095,
+      "learning_rate": 7.613934668119467e-06,
+      "loss": 0.3906,
+      "step": 2531
+    },
+    {
+      "epoch": 1.177491861726864,
+      "grad_norm": 0.44489437341690063,
+      "learning_rate": 7.611627573378466e-06,
+      "loss": 0.3505,
+      "step": 2532
+    },
+    {
+      "epoch": 1.177956905906061,
+      "grad_norm": 0.4203685522079468,
+      "learning_rate": 7.609319713734169e-06,
+      "loss": 0.3437,
+      "step": 2533
+    },
+    {
+      "epoch": 1.1784219500852582,
+      "grad_norm": 0.47429367899894714,
+      "learning_rate": 7.607011089862512e-06,
+      "loss": 0.3635,
+      "step": 2534
+    },
+    {
+      "epoch": 1.178886994264455,
+      "grad_norm": 0.4002319276332855,
+      "learning_rate": 7.604701702439652e-06,
+      "loss": 0.3221,
+      "step": 2535
+    },
+    {
+      "epoch": 1.1793520384436522,
+      "grad_norm": 0.45908692479133606,
+      "learning_rate": 7.602391552141972e-06,
+      "loss": 0.3561,
+      "step": 2536
+    },
+    {
+      "epoch": 1.1798170826228491,
+      "grad_norm": 0.4550569951534271,
+      "learning_rate": 7.600080639646077e-06,
+      "loss": 0.3201,
+      "step": 2537
+    },
+    {
+      "epoch": 1.1802821268020462,
+      "grad_norm": 0.4702633321285248,
+      "learning_rate": 7.597768965628798e-06,
+      "loss": 0.36,
+      "step": 2538
+    },
+    {
+      "epoch": 1.1807471709812432,
+      "grad_norm": 0.5423824787139893,
+      "learning_rate": 7.595456530767185e-06,
+      "loss": 0.3632,
+      "step": 2539
+    },
+    {
+      "epoch": 1.1812122151604403,
+      "grad_norm": 0.44125810265541077,
+      "learning_rate": 7.593143335738511e-06,
+      "loss": 0.3491,
+      "step": 2540
+    },
+    {
+      "epoch": 1.1816772593396372,
+      "grad_norm": 0.4145257771015167,
+      "learning_rate": 7.590829381220275e-06,
+      "loss": 0.3175,
+      "step": 2541
+    },
+    {
+      "epoch": 1.1821423035188343,
+      "grad_norm": 0.5215808153152466,
+      "learning_rate": 7.5885146678901954e-06,
+      "loss": 0.3891,
+      "step": 2542
+    },
+    {
+      "epoch": 1.1826073476980312,
+      "grad_norm": 0.4379088878631592,
+      "learning_rate": 7.586199196426216e-06,
+      "loss": 0.3551,
+      "step": 2543
+    },
+    {
+      "epoch": 1.1830723918772283,
+      "grad_norm": 0.42644643783569336,
+      "learning_rate": 7.583882967506502e-06,
+      "loss": 0.3282,
+      "step": 2544
+    },
+    {
+      "epoch": 1.1835374360564255,
+      "grad_norm": 0.4398006200790405,
+      "learning_rate": 7.581565981809435e-06,
+      "loss": 0.3172,
+      "step": 2545
+    },
+    {
+      "epoch": 1.1840024802356224,
+      "grad_norm": 0.46293768286705017,
+      "learning_rate": 7.579248240013626e-06,
+      "loss": 0.3876,
+      "step": 2546
+    },
+    {
+      "epoch": 1.1844675244148193,
+      "grad_norm": 0.46180588006973267,
+      "learning_rate": 7.576929742797902e-06,
+      "loss": 0.3936,
+      "step": 2547
+    },
+    {
+      "epoch": 1.1849325685940164,
+      "grad_norm": 0.4307175874710083,
+      "learning_rate": 7.574610490841312e-06,
+      "loss": 0.3386,
+      "step": 2548
+    },
+    {
+      "epoch": 1.1853976127732135,
+      "grad_norm": 0.3494255542755127,
+      "learning_rate": 7.5722904848231315e-06,
+      "loss": 0.3124,
+      "step": 2549
+    },
+    {
+      "epoch": 1.1858626569524104,
+      "grad_norm": 0.4179413914680481,
+      "learning_rate": 7.5699697254228496e-06,
+      "loss": 0.3271,
+      "step": 2550
+    },
+    {
+      "epoch": 1.1863277011316076,
+      "grad_norm": 0.47171148657798767,
+      "learning_rate": 7.56764821332018e-06,
+      "loss": 0.3899,
+      "step": 2551
+    },
+    {
+      "epoch": 1.1867927453108045,
+      "grad_norm": 0.4213588535785675,
+      "learning_rate": 7.565325949195055e-06,
+      "loss": 0.3373,
+      "step": 2552
+    },
+    {
+      "epoch": 1.1872577894900016,
+      "grad_norm": 0.48638585209846497,
+      "learning_rate": 7.563002933727628e-06,
+      "loss": 0.3578,
+      "step": 2553
+    },
+    {
+      "epoch": 1.1877228336691985,
+      "grad_norm": 0.4090135097503662,
+      "learning_rate": 7.560679167598273e-06,
+      "loss": 0.3496,
+      "step": 2554
+    },
+    {
+      "epoch": 1.1881878778483956,
+      "grad_norm": 0.44248971343040466,
+      "learning_rate": 7.558354651487583e-06,
+      "loss": 0.3446,
+      "step": 2555
+    },
+    {
+      "epoch": 1.1886529220275925,
+      "grad_norm": 0.40973594784736633,
+      "learning_rate": 7.556029386076371e-06,
+      "loss": 0.3623,
+      "step": 2556
+    },
+    {
+      "epoch": 1.1891179662067897,
+      "grad_norm": 0.4441074728965759,
+      "learning_rate": 7.553703372045671e-06,
+      "loss": 0.3421,
+      "step": 2557
+    },
+    {
+      "epoch": 1.1895830103859866,
+      "grad_norm": 0.48949727416038513,
+      "learning_rate": 7.5513766100767334e-06,
+      "loss": 0.3372,
+      "step": 2558
+    },
+    {
+      "epoch": 1.1900480545651837,
+      "grad_norm": 0.4103635251522064,
+      "learning_rate": 7.549049100851029e-06,
+      "loss": 0.3492,
+      "step": 2559
+    },
+    {
+      "epoch": 1.1905130987443808,
+      "grad_norm": 0.395436555147171,
+      "learning_rate": 7.546720845050247e-06,
+      "loss": 0.3475,
+      "step": 2560
+    },
+    {
+      "epoch": 1.1909781429235777,
+      "grad_norm": 0.4113844335079193,
+      "learning_rate": 7.544391843356298e-06,
+      "loss": 0.3744,
+      "step": 2561
+    },
+    {
+      "epoch": 1.1914431871027749,
+      "grad_norm": 0.397686243057251,
+      "learning_rate": 7.542062096451306e-06,
+      "loss": 0.333,
+      "step": 2562
+    },
+    {
+      "epoch": 1.1919082312819718,
+      "grad_norm": 0.4004164934158325,
+      "learning_rate": 7.539731605017616e-06,
+      "loss": 0.3354,
+      "step": 2563
+    },
+    {
+      "epoch": 1.192373275461169,
+      "grad_norm": 0.43825533986091614,
+      "learning_rate": 7.537400369737793e-06,
+      "loss": 0.3887,
+      "step": 2564
+    },
+    {
+      "epoch": 1.1928383196403658,
+      "grad_norm": 0.433907151222229,
+      "learning_rate": 7.535068391294618e-06,
+      "loss": 0.3413,
+      "step": 2565
+    },
+    {
+      "epoch": 1.193303363819563,
+      "grad_norm": 0.3985256254673004,
+      "learning_rate": 7.532735670371088e-06,
+      "loss": 0.3353,
+      "step": 2566
+    },
+    {
+      "epoch": 1.1937684079987598,
+      "grad_norm": 0.4107392728328705,
+      "learning_rate": 7.530402207650418e-06,
+      "loss": 0.3753,
+      "step": 2567
+    },
+    {
+      "epoch": 1.194233452177957,
+      "grad_norm": 0.40422889590263367,
+      "learning_rate": 7.528068003816045e-06,
+      "loss": 0.3373,
+      "step": 2568
+    },
+    {
+      "epoch": 1.1946984963571539,
+      "grad_norm": 0.44744202494621277,
+      "learning_rate": 7.525733059551618e-06,
+      "loss": 0.3985,
+      "step": 2569
+    },
+    {
+      "epoch": 1.195163540536351,
+      "grad_norm": 0.4296942353248596,
+      "learning_rate": 7.523397375541003e-06,
+      "loss": 0.3075,
+      "step": 2570
+    },
+    {
+      "epoch": 1.195628584715548,
+      "grad_norm": 0.47275644540786743,
+      "learning_rate": 7.521060952468284e-06,
+      "loss": 0.3916,
+      "step": 2571
+    },
+    {
+      "epoch": 1.196093628894745,
+      "grad_norm": 0.4506266117095947,
+      "learning_rate": 7.518723791017762e-06,
+      "loss": 0.3457,
+      "step": 2572
+    },
+    {
+      "epoch": 1.196558673073942,
+      "grad_norm": 0.5241588354110718,
+      "learning_rate": 7.516385891873954e-06,
+      "loss": 0.398,
+      "step": 2573
+    },
+    {
+      "epoch": 1.197023717253139,
+      "grad_norm": 0.38779518008232117,
+      "learning_rate": 7.5140472557215945e-06,
+      "loss": 0.3554,
+      "step": 2574
+    },
+    {
+      "epoch": 1.1974887614323362,
+      "grad_norm": 0.418839693069458,
+      "learning_rate": 7.511707883245627e-06,
+      "loss": 0.3467,
+      "step": 2575
+    },
+    {
+      "epoch": 1.197953805611533,
+      "grad_norm": 0.5039485692977905,
+      "learning_rate": 7.50936777513122e-06,
+      "loss": 0.4058,
+      "step": 2576
+    },
+    {
+      "epoch": 1.1984188497907302,
+      "grad_norm": 0.4700038433074951,
+      "learning_rate": 7.50702693206375e-06,
+      "loss": 0.3266,
+      "step": 2577
+    },
+    {
+      "epoch": 1.1988838939699271,
+      "grad_norm": 0.43291565775871277,
+      "learning_rate": 7.5046853547288155e-06,
+      "loss": 0.3645,
+      "step": 2578
+    },
+    {
+      "epoch": 1.1993489381491242,
+      "grad_norm": 0.45220211148262024,
+      "learning_rate": 7.502343043812224e-06,
+      "loss": 0.3327,
+      "step": 2579
+    },
+    {
+      "epoch": 1.1998139823283211,
+      "grad_norm": 0.4677691161632538,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.3708,
+      "step": 2580
+    },
+    {
+      "epoch": 1.2002790265075183,
+      "grad_norm": 0.43373778462409973,
+      "learning_rate": 7.497656223978385e-06,
+      "loss": 0.339,
+      "step": 2581
+    },
+    {
+      "epoch": 1.2007440706867152,
+      "grad_norm": 0.44764167070388794,
+      "learning_rate": 7.495311716433833e-06,
+      "loss": 0.3613,
+      "step": 2582
+    },
+    {
+      "epoch": 1.2012091148659123,
+      "grad_norm": 0.4182588756084442,
+      "learning_rate": 7.492966478053009e-06,
+      "loss": 0.3325,
+      "step": 2583
+    },
+    {
+      "epoch": 1.2016741590451092,
+      "grad_norm": 0.35634884238243103,
+      "learning_rate": 7.490620509522797e-06,
+      "loss": 0.3005,
+      "step": 2584
+    },
+    {
+      "epoch": 1.2021392032243063,
+      "grad_norm": 0.48205626010894775,
+      "learning_rate": 7.488273811530294e-06,
+      "loss": 0.4052,
+      "step": 2585
+    },
+    {
+      "epoch": 1.2026042474035032,
+      "grad_norm": 0.41472601890563965,
+      "learning_rate": 7.48592638476281e-06,
+      "loss": 0.3258,
+      "step": 2586
+    },
+    {
+      "epoch": 1.2030692915827004,
+      "grad_norm": 0.4383096992969513,
+      "learning_rate": 7.483578229907866e-06,
+      "loss": 0.3553,
+      "step": 2587
+    },
+    {
+      "epoch": 1.2035343357618973,
+      "grad_norm": 0.45658400654792786,
+      "learning_rate": 7.481229347653201e-06,
+      "loss": 0.3551,
+      "step": 2588
+    },
+    {
+      "epoch": 1.2039993799410944,
+      "grad_norm": 0.4237629771232605,
+      "learning_rate": 7.4788797386867596e-06,
+      "loss": 0.3481,
+      "step": 2589
+    },
+    {
+      "epoch": 1.2044644241202915,
+      "grad_norm": 0.4323580861091614,
+      "learning_rate": 7.47652940369671e-06,
+      "loss": 0.3566,
+      "step": 2590
+    },
+    {
+      "epoch": 1.2049294682994884,
+      "grad_norm": 0.4833803176879883,
+      "learning_rate": 7.474178343371425e-06,
+      "loss": 0.3718,
+      "step": 2591
+    },
+    {
+      "epoch": 1.2053945124786856,
+      "grad_norm": 0.4594928026199341,
+      "learning_rate": 7.471826558399492e-06,
+      "loss": 0.3428,
+      "step": 2592
+    },
+    {
+      "epoch": 1.2058595566578825,
+      "grad_norm": 0.4109436571598053,
+      "learning_rate": 7.469474049469709e-06,
+      "loss": 0.3287,
+      "step": 2593
+    },
+    {
+      "epoch": 1.2063246008370796,
+      "grad_norm": 0.43107515573501587,
+      "learning_rate": 7.467120817271091e-06,
+      "loss": 0.3694,
+      "step": 2594
+    },
+    {
+      "epoch": 1.2067896450162765,
+      "grad_norm": 0.4271430969238281,
+      "learning_rate": 7.464766862492856e-06,
+      "loss": 0.3546,
+      "step": 2595
+    },
+    {
+      "epoch": 1.2072546891954736,
+      "grad_norm": 0.4466875195503235,
+      "learning_rate": 7.4624121858244455e-06,
+      "loss": 0.357,
+      "step": 2596
+    },
+    {
+      "epoch": 1.2077197333746705,
+      "grad_norm": 0.3861998915672302,
+      "learning_rate": 7.460056787955502e-06,
+      "loss": 0.3172,
+      "step": 2597
+    },
+    {
+      "epoch": 1.2081847775538677,
+      "grad_norm": 0.4079117774963379,
+      "learning_rate": 7.4577006695758845e-06,
+      "loss": 0.3402,
+      "step": 2598
+    },
+    {
+      "epoch": 1.2086498217330646,
+      "grad_norm": 0.521819531917572,
+      "learning_rate": 7.455343831375662e-06,
+      "loss": 0.4022,
+      "step": 2599
+    },
+    {
+      "epoch": 1.2091148659122617,
+      "grad_norm": 0.45120519399642944,
+      "learning_rate": 7.452986274045114e-06,
+      "loss": 0.3237,
+      "step": 2600
+    },
+    {
+      "epoch": 1.2095799100914586,
+      "grad_norm": 0.42835864424705505,
+      "learning_rate": 7.45062799827473e-06,
+      "loss": 0.3277,
+      "step": 2601
+    },
+    {
+      "epoch": 1.2100449542706557,
+      "grad_norm": 0.5196753144264221,
+      "learning_rate": 7.4482690047552105e-06,
+      "loss": 0.3841,
+      "step": 2602
+    },
+    {
+      "epoch": 1.2105099984498526,
+      "grad_norm": 0.616445004940033,
+      "learning_rate": 7.445909294177469e-06,
+      "loss": 0.4043,
+      "step": 2603
+    },
+    {
+      "epoch": 1.2109750426290498,
+      "grad_norm": 0.35232779383659363,
+      "learning_rate": 7.4435488672326235e-06,
+      "loss": 0.293,
+      "step": 2604
+    },
+    {
+      "epoch": 1.2114400868082469,
+      "grad_norm": 0.5146499276161194,
+      "learning_rate": 7.441187724612007e-06,
+      "loss": 0.3402,
+      "step": 2605
+    },
+    {
+      "epoch": 1.2119051309874438,
+      "grad_norm": 0.5325912833213806,
+      "learning_rate": 7.438825867007156e-06,
+      "loss": 0.3614,
+      "step": 2606
+    },
+    {
+      "epoch": 1.212370175166641,
+      "grad_norm": 0.4090646803379059,
+      "learning_rate": 7.436463295109824e-06,
+      "loss": 0.3651,
+      "step": 2607
+    },
+    {
+      "epoch": 1.2128352193458378,
+      "grad_norm": 0.4489936828613281,
+      "learning_rate": 7.434100009611969e-06,
+      "loss": 0.3855,
+      "step": 2608
+    },
+    {
+      "epoch": 1.213300263525035,
+      "grad_norm": 0.410040020942688,
+      "learning_rate": 7.431736011205759e-06,
+      "loss": 0.3056,
+      "step": 2609
+    },
+    {
+      "epoch": 1.2137653077042319,
+      "grad_norm": 0.4164040982723236,
+      "learning_rate": 7.42937130058357e-06,
+      "loss": 0.3439,
+      "step": 2610
+    },
+    {
+      "epoch": 1.214230351883429,
+      "grad_norm": 0.4618469774723053,
+      "learning_rate": 7.427005878437989e-06,
+      "loss": 0.3526,
+      "step": 2611
+    },
+    {
+      "epoch": 1.2146953960626259,
+      "grad_norm": 0.4020163416862488,
+      "learning_rate": 7.4246397454618054e-06,
+      "loss": 0.3325,
+      "step": 2612
+    },
+    {
+      "epoch": 1.215160440241823,
+      "grad_norm": 0.442120224237442,
+      "learning_rate": 7.422272902348027e-06,
+      "loss": 0.3265,
+      "step": 2613
+    },
+    {
+      "epoch": 1.21562548442102,
+      "grad_norm": 0.527863085269928,
+      "learning_rate": 7.41990534978986e-06,
+      "loss": 0.3838,
+      "step": 2614
+    },
+    {
+      "epoch": 1.216090528600217,
+      "grad_norm": 0.4838730990886688,
+      "learning_rate": 7.417537088480722e-06,
+      "loss": 0.3671,
+      "step": 2615
+    },
+    {
+      "epoch": 1.216555572779414,
+      "grad_norm": 0.3663950264453888,
+      "learning_rate": 7.41516811911424e-06,
+      "loss": 0.3008,
+      "step": 2616
+    },
+    {
+      "epoch": 1.217020616958611,
+      "grad_norm": 0.4481393098831177,
+      "learning_rate": 7.412798442384246e-06,
+      "loss": 0.375,
+      "step": 2617
+    },
+    {
+      "epoch": 1.217485661137808,
+      "grad_norm": 0.500136137008667,
+      "learning_rate": 7.4104280589847785e-06,
+      "loss": 0.3493,
+      "step": 2618
+    },
+    {
+      "epoch": 1.2179507053170051,
+      "grad_norm": 0.4056425988674164,
+      "learning_rate": 7.408056969610087e-06,
+      "loss": 0.3305,
+      "step": 2619
+    },
+    {
+      "epoch": 1.2184157494962022,
+      "grad_norm": 0.3582363724708557,
+      "learning_rate": 7.405685174954623e-06,
+      "loss": 0.3193,
+      "step": 2620
+    },
+    {
+      "epoch": 1.2188807936753991,
+      "grad_norm": 0.5270162224769592,
+      "learning_rate": 7.403312675713047e-06,
+      "loss": 0.3598,
+      "step": 2621
+    },
+    {
+      "epoch": 1.2193458378545963,
+      "grad_norm": 0.5136641263961792,
+      "learning_rate": 7.400939472580227e-06,
+      "loss": 0.3283,
+      "step": 2622
+    },
+    {
+      "epoch": 1.2198108820337932,
+      "grad_norm": 0.46304944157600403,
+      "learning_rate": 7.398565566251232e-06,
+      "loss": 0.3431,
+      "step": 2623
+    },
+    {
+      "epoch": 1.2202759262129903,
+      "grad_norm": 0.45918190479278564,
+      "learning_rate": 7.396190957421343e-06,
+      "loss": 0.3276,
+      "step": 2624
+    },
+    {
+      "epoch": 1.2207409703921872,
+      "grad_norm": 0.5070422291755676,
+      "learning_rate": 7.393815646786047e-06,
+      "loss": 0.3519,
+      "step": 2625
+    },
+    {
+      "epoch": 1.2212060145713843,
+      "grad_norm": 0.45554548501968384,
+      "learning_rate": 7.3914396350410285e-06,
+      "loss": 0.3358,
+      "step": 2626
+    },
+    {
+      "epoch": 1.2216710587505812,
+      "grad_norm": 0.43549612164497375,
+      "learning_rate": 7.389062922882187e-06,
+      "loss": 0.3329,
+      "step": 2627
+    },
+    {
+      "epoch": 1.2221361029297784,
+      "grad_norm": 0.4475855231285095,
+      "learning_rate": 7.3866855110056205e-06,
+      "loss": 0.3651,
+      "step": 2628
+    },
+    {
+      "epoch": 1.2226011471089753,
+      "grad_norm": 0.47854486107826233,
+      "learning_rate": 7.384307400107635e-06,
+      "loss": 0.3577,
+      "step": 2629
+    },
+    {
+      "epoch": 1.2230661912881724,
+      "grad_norm": 0.4717871844768524,
+      "learning_rate": 7.381928590884741e-06,
+      "loss": 0.4082,
+      "step": 2630
+    },
+    {
+      "epoch": 1.2235312354673693,
+      "grad_norm": 0.42780524492263794,
+      "learning_rate": 7.379549084033653e-06,
+      "loss": 0.3446,
+      "step": 2631
+    },
+    {
+      "epoch": 1.2239962796465664,
+      "grad_norm": 0.41821184754371643,
+      "learning_rate": 7.37716888025129e-06,
+      "loss": 0.3719,
+      "step": 2632
+    },
+    {
+      "epoch": 1.2244613238257633,
+      "grad_norm": 0.43905121088027954,
+      "learning_rate": 7.374787980234775e-06,
+      "loss": 0.3395,
+      "step": 2633
+    },
+    {
+      "epoch": 1.2249263680049605,
+      "grad_norm": 0.44092613458633423,
+      "learning_rate": 7.372406384681433e-06,
+      "loss": 0.3066,
+      "step": 2634
+    },
+    {
+      "epoch": 1.2253914121841576,
+      "grad_norm": 0.5179281234741211,
+      "learning_rate": 7.370024094288797e-06,
+      "loss": 0.3801,
+      "step": 2635
+    },
+    {
+      "epoch": 1.2258564563633545,
+      "grad_norm": 0.4214264452457428,
+      "learning_rate": 7.3676411097546e-06,
+      "loss": 0.3758,
+      "step": 2636
+    },
+    {
+      "epoch": 1.2263215005425516,
+      "grad_norm": 0.43435075879096985,
+      "learning_rate": 7.36525743177678e-06,
+      "loss": 0.3487,
+      "step": 2637
+    },
+    {
+      "epoch": 1.2267865447217485,
+      "grad_norm": 0.4535754323005676,
+      "learning_rate": 7.362873061053479e-06,
+      "loss": 0.3527,
+      "step": 2638
+    },
+    {
+      "epoch": 1.2272515889009457,
+      "grad_norm": 0.4143884778022766,
+      "learning_rate": 7.360487998283038e-06,
+      "loss": 0.3025,
+      "step": 2639
+    },
+    {
+      "epoch": 1.2277166330801426,
+      "grad_norm": 0.4507713317871094,
+      "learning_rate": 7.358102244164003e-06,
+      "loss": 0.3842,
+      "step": 2640
+    },
+    {
+      "epoch": 1.2281816772593397,
+      "grad_norm": 0.4342232942581177,
+      "learning_rate": 7.355715799395126e-06,
+      "loss": 0.37,
+      "step": 2641
+    },
+    {
+      "epoch": 1.2286467214385366,
+      "grad_norm": 0.38228335976600647,
+      "learning_rate": 7.353328664675353e-06,
+      "loss": 0.3279,
+      "step": 2642
+    },
+    {
+      "epoch": 1.2291117656177337,
+      "grad_norm": 0.42254742980003357,
+      "learning_rate": 7.350940840703842e-06,
+      "loss": 0.3522,
+      "step": 2643
+    },
+    {
+      "epoch": 1.2295768097969306,
+      "grad_norm": 0.48193877935409546,
+      "learning_rate": 7.348552328179945e-06,
+      "loss": 0.3823,
+      "step": 2644
+    },
+    {
+      "epoch": 1.2300418539761278,
+      "grad_norm": 0.4392937123775482,
+      "learning_rate": 7.3461631278032175e-06,
+      "loss": 0.3688,
+      "step": 2645
+    },
+    {
+      "epoch": 1.2305068981553249,
+      "grad_norm": 0.545441210269928,
+      "learning_rate": 7.34377324027342e-06,
+      "loss": 0.3537,
+      "step": 2646
+    },
+    {
+      "epoch": 1.2309719423345218,
+      "grad_norm": 0.43187499046325684,
+      "learning_rate": 7.3413826662905104e-06,
+      "loss": 0.314,
+      "step": 2647
+    },
+    {
+      "epoch": 1.2314369865137187,
+      "grad_norm": 0.4856451451778412,
+      "learning_rate": 7.33899140655465e-06,
+      "loss": 0.386,
+      "step": 2648
+    },
+    {
+      "epoch": 1.2319020306929158,
+      "grad_norm": 0.4500555396080017,
+      "learning_rate": 7.336599461766199e-06,
+      "loss": 0.33,
+      "step": 2649
+    },
+    {
+      "epoch": 1.232367074872113,
+      "grad_norm": 0.41082900762557983,
+      "learning_rate": 7.334206832625719e-06,
+      "loss": 0.3631,
+      "step": 2650
+    },
+    {
+      "epoch": 1.2328321190513098,
+      "grad_norm": 0.43610817193984985,
+      "learning_rate": 7.331813519833972e-06,
+      "loss": 0.3594,
+      "step": 2651
+    },
+    {
+      "epoch": 1.233297163230507,
+      "grad_norm": 0.3768427073955536,
+      "learning_rate": 7.329419524091923e-06,
+      "loss": 0.3397,
+      "step": 2652
+    },
+    {
+      "epoch": 1.2337622074097039,
+      "grad_norm": 0.4704180657863617,
+      "learning_rate": 7.32702484610073e-06,
+      "loss": 0.339,
+      "step": 2653
+    },
+    {
+      "epoch": 1.234227251588901,
+      "grad_norm": 0.5134731531143188,
+      "learning_rate": 7.32462948656176e-06,
+      "loss": 0.389,
+      "step": 2654
+    },
+    {
+      "epoch": 1.234692295768098,
+      "grad_norm": 0.454752117395401,
+      "learning_rate": 7.322233446176571e-06,
+      "loss": 0.3489,
+      "step": 2655
+    },
+    {
+      "epoch": 1.235157339947295,
+      "grad_norm": 0.4670364260673523,
+      "learning_rate": 7.319836725646927e-06,
+      "loss": 0.3759,
+      "step": 2656
+    },
+    {
+      "epoch": 1.235622384126492,
+      "grad_norm": 0.423016220331192,
+      "learning_rate": 7.317439325674788e-06,
+      "loss": 0.3212,
+      "step": 2657
+    },
+    {
+      "epoch": 1.236087428305689,
+      "grad_norm": 0.5560788512229919,
+      "learning_rate": 7.315041246962313e-06,
+      "loss": 0.3528,
+      "step": 2658
+    },
+    {
+      "epoch": 1.236552472484886,
+      "grad_norm": 0.3992140591144562,
+      "learning_rate": 7.312642490211857e-06,
+      "loss": 0.3611,
+      "step": 2659
+    },
+    {
+      "epoch": 1.237017516664083,
+      "grad_norm": 0.49364173412323,
+      "learning_rate": 7.310243056125984e-06,
+      "loss": 0.3515,
+      "step": 2660
+    },
+    {
+      "epoch": 1.2374825608432802,
+      "grad_norm": 0.5286840796470642,
+      "learning_rate": 7.307842945407444e-06,
+      "loss": 0.3108,
+      "step": 2661
+    },
+    {
+      "epoch": 1.2379476050224771,
+      "grad_norm": 0.4787932336330414,
+      "learning_rate": 7.3054421587591925e-06,
+      "loss": 0.3592,
+      "step": 2662
+    },
+    {
+      "epoch": 1.238412649201674,
+      "grad_norm": 0.4455261528491974,
+      "learning_rate": 7.30304069688438e-06,
+      "loss": 0.3157,
+      "step": 2663
+    },
+    {
+      "epoch": 1.2388776933808712,
+      "grad_norm": 0.4244392216205597,
+      "learning_rate": 7.300638560486357e-06,
+      "loss": 0.3612,
+      "step": 2664
+    },
+    {
+      "epoch": 1.2393427375600683,
+      "grad_norm": 0.4475732147693634,
+      "learning_rate": 7.298235750268667e-06,
+      "loss": 0.3858,
+      "step": 2665
+    },
+    {
+      "epoch": 1.2398077817392652,
+      "grad_norm": 0.4742658734321594,
+      "learning_rate": 7.295832266935059e-06,
+      "loss": 0.3213,
+      "step": 2666
+    },
+    {
+      "epoch": 1.2402728259184623,
+      "grad_norm": 0.4781046509742737,
+      "learning_rate": 7.29342811118947e-06,
+      "loss": 0.3711,
+      "step": 2667
+    },
+    {
+      "epoch": 1.2407378700976592,
+      "grad_norm": 0.4173920750617981,
+      "learning_rate": 7.29102328373604e-06,
+      "loss": 0.3644,
+      "step": 2668
+    },
+    {
+      "epoch": 1.2412029142768564,
+      "grad_norm": 0.4399764835834503,
+      "learning_rate": 7.288617785279104e-06,
+      "loss": 0.3529,
+      "step": 2669
+    },
+    {
+      "epoch": 1.2416679584560533,
+      "grad_norm": 0.41541317105293274,
+      "learning_rate": 7.286211616523193e-06,
+      "loss": 0.3791,
+      "step": 2670
+    },
+    {
+      "epoch": 1.2421330026352504,
+      "grad_norm": 0.4661194682121277,
+      "learning_rate": 7.283804778173035e-06,
+      "loss": 0.3629,
+      "step": 2671
+    },
+    {
+      "epoch": 1.2425980468144473,
+      "grad_norm": 0.45775488018989563,
+      "learning_rate": 7.281397270933553e-06,
+      "loss": 0.3493,
+      "step": 2672
+    },
+    {
+      "epoch": 1.2430630909936444,
+      "grad_norm": 0.45113879442214966,
+      "learning_rate": 7.278989095509867e-06,
+      "loss": 0.3641,
+      "step": 2673
+    },
+    {
+      "epoch": 1.2435281351728413,
+      "grad_norm": 0.4759120047092438,
+      "learning_rate": 7.2765802526072945e-06,
+      "loss": 0.3871,
+      "step": 2674
+    },
+    {
+      "epoch": 1.2439931793520385,
+      "grad_norm": 0.35728970170021057,
+      "learning_rate": 7.274170742931345e-06,
+      "loss": 0.3399,
+      "step": 2675
+    },
+    {
+      "epoch": 1.2444582235312356,
+      "grad_norm": 0.41665512323379517,
+      "learning_rate": 7.271760567187723e-06,
+      "loss": 0.3501,
+      "step": 2676
+    },
+    {
+      "epoch": 1.2449232677104325,
+      "grad_norm": 0.5279736518859863,
+      "learning_rate": 7.269349726082332e-06,
+      "loss": 0.3652,
+      "step": 2677
+    },
+    {
+      "epoch": 1.2453883118896294,
+      "grad_norm": 0.3690641522407532,
+      "learning_rate": 7.266938220321268e-06,
+      "loss": 0.3099,
+      "step": 2678
+    },
+    {
+      "epoch": 1.2458533560688265,
+      "grad_norm": 0.49263355135917664,
+      "learning_rate": 7.264526050610821e-06,
+      "loss": 0.3927,
+      "step": 2679
+    },
+    {
+      "epoch": 1.2463184002480237,
+      "grad_norm": 0.3883993327617645,
+      "learning_rate": 7.2621132176574774e-06,
+      "loss": 0.3367,
+      "step": 2680
+    },
+    {
+      "epoch": 1.2467834444272206,
+      "grad_norm": 0.4446150064468384,
+      "learning_rate": 7.2596997221679156e-06,
+      "loss": 0.3588,
+      "step": 2681
+    },
+    {
+      "epoch": 1.2472484886064177,
+      "grad_norm": 0.5633254051208496,
+      "learning_rate": 7.2572855648490105e-06,
+      "loss": 0.3633,
+      "step": 2682
+    },
+    {
+      "epoch": 1.2477135327856146,
+      "grad_norm": 0.4423445165157318,
+      "learning_rate": 7.25487074640783e-06,
+      "loss": 0.3665,
+      "step": 2683
+    },
+    {
+      "epoch": 1.2481785769648117,
+      "grad_norm": 0.48402294516563416,
+      "learning_rate": 7.252455267551631e-06,
+      "loss": 0.3508,
+      "step": 2684
+    },
+    {
+      "epoch": 1.2486436211440086,
+      "grad_norm": 0.53091961145401,
+      "learning_rate": 7.250039128987874e-06,
+      "loss": 0.3603,
+      "step": 2685
+    },
+    {
+      "epoch": 1.2491086653232057,
+      "grad_norm": 0.47006434202194214,
+      "learning_rate": 7.247622331424202e-06,
+      "loss": 0.3463,
+      "step": 2686
+    },
+    {
+      "epoch": 1.2495737095024027,
+      "grad_norm": 0.5498688817024231,
+      "learning_rate": 7.245204875568459e-06,
+      "loss": 0.3642,
+      "step": 2687
+    },
+    {
+      "epoch": 1.2500387536815998,
+      "grad_norm": 0.4160943627357483,
+      "learning_rate": 7.242786762128675e-06,
+      "loss": 0.3097,
+      "step": 2688
+    },
+    {
+      "epoch": 1.2505037978607967,
+      "grad_norm": 0.38064897060394287,
+      "learning_rate": 7.240367991813079e-06,
+      "loss": 0.3235,
+      "step": 2689
+    },
+    {
+      "epoch": 1.2509688420399938,
+      "grad_norm": 0.3919866383075714,
+      "learning_rate": 7.237948565330091e-06,
+      "loss": 0.3477,
+      "step": 2690
+    },
+    {
+      "epoch": 1.251433886219191,
+      "grad_norm": 0.4545346200466156,
+      "learning_rate": 7.235528483388318e-06,
+      "loss": 0.3706,
+      "step": 2691
+    },
+    {
+      "epoch": 1.2518989303983878,
+      "grad_norm": 0.43774187564849854,
+      "learning_rate": 7.233107746696564e-06,
+      "loss": 0.3314,
+      "step": 2692
+    },
+    {
+      "epoch": 1.2523639745775847,
+      "grad_norm": 0.44677144289016724,
+      "learning_rate": 7.230686355963824e-06,
+      "loss": 0.3369,
+      "step": 2693
+    },
+    {
+      "epoch": 1.2528290187567819,
+      "grad_norm": 0.4839126467704773,
+      "learning_rate": 7.228264311899284e-06,
+      "loss": 0.3949,
+      "step": 2694
+    },
+    {
+      "epoch": 1.253294062935979,
+      "grad_norm": 0.5056577920913696,
+      "learning_rate": 7.22584161521232e-06,
+      "loss": 0.3997,
+      "step": 2695
+    },
+    {
+      "epoch": 1.253759107115176,
+      "grad_norm": 0.35565152764320374,
+      "learning_rate": 7.223418266612503e-06,
+      "loss": 0.3188,
+      "step": 2696
+    },
+    {
+      "epoch": 1.254224151294373,
+      "grad_norm": 0.3856726884841919,
+      "learning_rate": 7.220994266809591e-06,
+      "loss": 0.3519,
+      "step": 2697
+    },
+    {
+      "epoch": 1.25468919547357,
+      "grad_norm": 0.46455973386764526,
+      "learning_rate": 7.218569616513533e-06,
+      "loss": 0.357,
+      "step": 2698
+    },
+    {
+      "epoch": 1.255154239652767,
+      "grad_norm": 0.37406978011131287,
+      "learning_rate": 7.2161443164344725e-06,
+      "loss": 0.3416,
+      "step": 2699
+    },
+    {
+      "epoch": 1.255619283831964,
+      "grad_norm": 0.43167954683303833,
+      "learning_rate": 7.213718367282737e-06,
+      "loss": 0.3567,
+      "step": 2700
+    },
+    {
+      "epoch": 1.256084328011161,
+      "grad_norm": 0.41151732206344604,
+      "learning_rate": 7.21129176976885e-06,
+      "loss": 0.3307,
+      "step": 2701
+    },
+    {
+      "epoch": 1.256549372190358,
+      "grad_norm": 0.4612469971179962,
+      "learning_rate": 7.208864524603523e-06,
+      "loss": 0.3684,
+      "step": 2702
+    },
+    {
+      "epoch": 1.2570144163695551,
+      "grad_norm": 0.4303331673145294,
+      "learning_rate": 7.206436632497656e-06,
+      "loss": 0.3358,
+      "step": 2703
+    },
+    {
+      "epoch": 1.257479460548752,
+      "grad_norm": 0.4344772696495056,
+      "learning_rate": 7.204008094162338e-06,
+      "loss": 0.3417,
+      "step": 2704
+    },
+    {
+      "epoch": 1.2579445047279492,
+      "grad_norm": 0.4836406409740448,
+      "learning_rate": 7.201578910308848e-06,
+      "loss": 0.3793,
+      "step": 2705
+    },
+    {
+      "epoch": 1.2584095489071463,
+      "grad_norm": 0.41068270802497864,
+      "learning_rate": 7.199149081648658e-06,
+      "loss": 0.3643,
+      "step": 2706
+    },
+    {
+      "epoch": 1.2588745930863432,
+      "grad_norm": 0.4286365211009979,
+      "learning_rate": 7.196718608893421e-06,
+      "loss": 0.3287,
+      "step": 2707
+    },
+    {
+      "epoch": 1.25933963726554,
+      "grad_norm": 0.4405948519706726,
+      "learning_rate": 7.1942874927549865e-06,
+      "loss": 0.3697,
+      "step": 2708
+    },
+    {
+      "epoch": 1.2598046814447372,
+      "grad_norm": 0.36979782581329346,
+      "learning_rate": 7.191855733945388e-06,
+      "loss": 0.3286,
+      "step": 2709
+    },
+    {
+      "epoch": 1.2602697256239344,
+      "grad_norm": 0.40628305077552795,
+      "learning_rate": 7.189423333176847e-06,
+      "loss": 0.3815,
+      "step": 2710
+    },
+    {
+      "epoch": 1.2607347698031313,
+      "grad_norm": 0.42872005701065063,
+      "learning_rate": 7.186990291161775e-06,
+      "loss": 0.3473,
+      "step": 2711
+    },
+    {
+      "epoch": 1.2611998139823284,
+      "grad_norm": 0.4314822852611542,
+      "learning_rate": 7.184556608612769e-06,
+      "loss": 0.3791,
+      "step": 2712
+    },
+    {
+      "epoch": 1.2616648581615253,
+      "grad_norm": 0.42616841197013855,
+      "learning_rate": 7.182122286242617e-06,
+      "loss": 0.3513,
+      "step": 2713
+    },
+    {
+      "epoch": 1.2621299023407224,
+      "grad_norm": 0.4406971037387848,
+      "learning_rate": 7.1796873247642925e-06,
+      "loss": 0.3609,
+      "step": 2714
+    },
+    {
+      "epoch": 1.2625949465199193,
+      "grad_norm": 0.38528046011924744,
+      "learning_rate": 7.177251724890957e-06,
+      "loss": 0.3372,
+      "step": 2715
+    },
+    {
+      "epoch": 1.2630599906991165,
+      "grad_norm": 0.4626089334487915,
+      "learning_rate": 7.1748154873359556e-06,
+      "loss": 0.3702,
+      "step": 2716
+    },
+    {
+      "epoch": 1.2635250348783134,
+      "grad_norm": 0.3866967260837555,
+      "learning_rate": 7.172378612812824e-06,
+      "loss": 0.3385,
+      "step": 2717
+    },
+    {
+      "epoch": 1.2639900790575105,
+      "grad_norm": 0.43559587001800537,
+      "learning_rate": 7.169941102035281e-06,
+      "loss": 0.3711,
+      "step": 2718
+    },
+    {
+      "epoch": 1.2644551232367074,
+      "grad_norm": 0.4723300635814667,
+      "learning_rate": 7.167502955717238e-06,
+      "loss": 0.3782,
+      "step": 2719
+    },
+    {
+      "epoch": 1.2649201674159045,
+      "grad_norm": 0.4137257933616638,
+      "learning_rate": 7.165064174572787e-06,
+      "loss": 0.3603,
+      "step": 2720
+    },
+    {
+      "epoch": 1.2653852115951016,
+      "grad_norm": 0.5321688652038574,
+      "learning_rate": 7.162624759316205e-06,
+      "loss": 0.3839,
+      "step": 2721
+    },
+    {
+      "epoch": 1.2658502557742985,
+      "grad_norm": 0.47028544545173645,
+      "learning_rate": 7.160184710661959e-06,
+      "loss": 0.3513,
+      "step": 2722
+    },
+    {
+      "epoch": 1.2663152999534955,
+      "grad_norm": 0.5355919599533081,
+      "learning_rate": 7.1577440293247e-06,
+      "loss": 0.3701,
+      "step": 2723
+    },
+    {
+      "epoch": 1.2667803441326926,
+      "grad_norm": 0.4389120936393738,
+      "learning_rate": 7.155302716019263e-06,
+      "loss": 0.3273,
+      "step": 2724
+    },
+    {
+      "epoch": 1.2672453883118897,
+      "grad_norm": 0.4850804805755615,
+      "learning_rate": 7.1528607714606706e-06,
+      "loss": 0.3639,
+      "step": 2725
+    },
+    {
+      "epoch": 1.2677104324910866,
+      "grad_norm": 0.4661966562271118,
+      "learning_rate": 7.1504181963641265e-06,
+      "loss": 0.3097,
+      "step": 2726
+    },
+    {
+      "epoch": 1.2681754766702837,
+      "grad_norm": 0.4354253113269806,
+      "learning_rate": 7.147974991445021e-06,
+      "loss": 0.3555,
+      "step": 2727
+    },
+    {
+      "epoch": 1.2686405208494806,
+      "grad_norm": 0.46909964084625244,
+      "learning_rate": 7.145531157418932e-06,
+      "loss": 0.3747,
+      "step": 2728
+    },
+    {
+      "epoch": 1.2691055650286778,
+      "grad_norm": 0.5229499340057373,
+      "learning_rate": 7.143086695001616e-06,
+      "loss": 0.3623,
+      "step": 2729
+    },
+    {
+      "epoch": 1.2695706092078747,
+      "grad_norm": 0.42865434288978577,
+      "learning_rate": 7.14064160490902e-06,
+      "loss": 0.3392,
+      "step": 2730
+    },
+    {
+      "epoch": 1.2700356533870718,
+      "grad_norm": 0.46530023217201233,
+      "learning_rate": 7.1381958878572665e-06,
+      "loss": 0.3611,
+      "step": 2731
+    },
+    {
+      "epoch": 1.270500697566269,
+      "grad_norm": 0.5126709938049316,
+      "learning_rate": 7.135749544562669e-06,
+      "loss": 0.3519,
+      "step": 2732
+    },
+    {
+      "epoch": 1.2709657417454658,
+      "grad_norm": 0.4154670536518097,
+      "learning_rate": 7.133302575741722e-06,
+      "loss": 0.3326,
+      "step": 2733
+    },
+    {
+      "epoch": 1.2714307859246627,
+      "grad_norm": 0.4609792232513428,
+      "learning_rate": 7.130854982111102e-06,
+      "loss": 0.375,
+      "step": 2734
+    },
+    {
+      "epoch": 1.2718958301038599,
+      "grad_norm": 0.4117323160171509,
+      "learning_rate": 7.128406764387667e-06,
+      "loss": 0.3381,
+      "step": 2735
+    },
+    {
+      "epoch": 1.272360874283057,
+      "grad_norm": 0.47705698013305664,
+      "learning_rate": 7.1259579232884655e-06,
+      "loss": 0.3487,
+      "step": 2736
+    },
+    {
+      "epoch": 1.272825918462254,
+      "grad_norm": 0.464181512594223,
+      "learning_rate": 7.12350845953072e-06,
+      "loss": 0.333,
+      "step": 2737
+    },
+    {
+      "epoch": 1.2732909626414508,
+      "grad_norm": 0.41154077649116516,
+      "learning_rate": 7.12105837383184e-06,
+      "loss": 0.3457,
+      "step": 2738
+    },
+    {
+      "epoch": 1.273756006820648,
+      "grad_norm": 0.4273618459701538,
+      "learning_rate": 7.118607666909413e-06,
+      "loss": 0.338,
+      "step": 2739
+    },
+    {
+      "epoch": 1.274221050999845,
+      "grad_norm": 0.4530821740627289,
+      "learning_rate": 7.116156339481215e-06,
+      "loss": 0.3423,
+      "step": 2740
+    },
+    {
+      "epoch": 1.274686095179042,
+      "grad_norm": 0.434441477060318,
+      "learning_rate": 7.113704392265196e-06,
+      "loss": 0.3672,
+      "step": 2741
+    },
+    {
+      "epoch": 1.275151139358239,
+      "grad_norm": 0.44967541098594666,
+      "learning_rate": 7.1112518259794946e-06,
+      "loss": 0.3257,
+      "step": 2742
+    },
+    {
+      "epoch": 1.275616183537436,
+      "grad_norm": 0.44372931122779846,
+      "learning_rate": 7.108798641342428e-06,
+      "loss": 0.313,
+      "step": 2743
+    },
+    {
+      "epoch": 1.2760812277166331,
+      "grad_norm": 0.4219914674758911,
+      "learning_rate": 7.106344839072492e-06,
+      "loss": 0.361,
+      "step": 2744
+    },
+    {
+      "epoch": 1.27654627189583,
+      "grad_norm": 0.4370742738246918,
+      "learning_rate": 7.103890419888367e-06,
+      "loss": 0.3304,
+      "step": 2745
+    },
+    {
+      "epoch": 1.2770113160750272,
+      "grad_norm": 0.41166791319847107,
+      "learning_rate": 7.1014353845089126e-06,
+      "loss": 0.3618,
+      "step": 2746
+    },
+    {
+      "epoch": 1.2774763602542243,
+      "grad_norm": 0.3665507137775421,
+      "learning_rate": 7.098979733653165e-06,
+      "loss": 0.3125,
+      "step": 2747
+    },
+    {
+      "epoch": 1.2779414044334212,
+      "grad_norm": 0.4135091006755829,
+      "learning_rate": 7.096523468040349e-06,
+      "loss": 0.3494,
+      "step": 2748
+    },
+    {
+      "epoch": 1.278406448612618,
+      "grad_norm": 0.5211425423622131,
+      "learning_rate": 7.094066588389864e-06,
+      "loss": 0.3392,
+      "step": 2749
+    },
+    {
+      "epoch": 1.2788714927918152,
+      "grad_norm": 0.4214104413986206,
+      "learning_rate": 7.09160909542129e-06,
+      "loss": 0.3606,
+      "step": 2750
+    },
+    {
+      "epoch": 1.2793365369710124,
+      "grad_norm": 0.4723619818687439,
+      "learning_rate": 7.089150989854385e-06,
+      "loss": 0.3315,
+      "step": 2751
+    },
+    {
+      "epoch": 1.2798015811502093,
+      "grad_norm": 0.5669287443161011,
+      "learning_rate": 7.08669227240909e-06,
+      "loss": 0.3577,
+      "step": 2752
+    },
+    {
+      "epoch": 1.2802666253294062,
+      "grad_norm": 0.43773379921913147,
+      "learning_rate": 7.084232943805522e-06,
+      "loss": 0.3266,
+      "step": 2753
+    },
+    {
+      "epoch": 1.2807316695086033,
+      "grad_norm": 0.4219588041305542,
+      "learning_rate": 7.081773004763981e-06,
+      "loss": 0.3465,
+      "step": 2754
+    },
+    {
+      "epoch": 1.2811967136878004,
+      "grad_norm": 0.45649266242980957,
+      "learning_rate": 7.079312456004941e-06,
+      "loss": 0.337,
+      "step": 2755
+    },
+    {
+      "epoch": 1.2816617578669973,
+      "grad_norm": 0.49470004439353943,
+      "learning_rate": 7.076851298249057e-06,
+      "loss": 0.3538,
+      "step": 2756
+    },
+    {
+      "epoch": 1.2821268020461944,
+      "grad_norm": 0.4066573679447174,
+      "learning_rate": 7.074389532217163e-06,
+      "loss": 0.3736,
+      "step": 2757
+    },
+    {
+      "epoch": 1.2825918462253914,
+      "grad_norm": 0.41550639271736145,
+      "learning_rate": 7.0719271586302675e-06,
+      "loss": 0.3284,
+      "step": 2758
+    },
+    {
+      "epoch": 1.2830568904045885,
+      "grad_norm": 0.45798173546791077,
+      "learning_rate": 7.069464178209563e-06,
+      "loss": 0.3366,
+      "step": 2759
+    },
+    {
+      "epoch": 1.2835219345837854,
+      "grad_norm": 0.44275960326194763,
+      "learning_rate": 7.067000591676416e-06,
+      "loss": 0.3668,
+      "step": 2760
+    },
+    {
+      "epoch": 1.2839869787629825,
+      "grad_norm": 0.40415266156196594,
+      "learning_rate": 7.06453639975237e-06,
+      "loss": 0.3625,
+      "step": 2761
+    },
+    {
+      "epoch": 1.2844520229421796,
+      "grad_norm": 0.4864114224910736,
+      "learning_rate": 7.062071603159147e-06,
+      "loss": 0.3692,
+      "step": 2762
+    },
+    {
+      "epoch": 1.2849170671213765,
+      "grad_norm": 0.4446645975112915,
+      "learning_rate": 7.059606202618646e-06,
+      "loss": 0.3724,
+      "step": 2763
+    },
+    {
+      "epoch": 1.2853821113005734,
+      "grad_norm": 0.406984806060791,
+      "learning_rate": 7.057140198852942e-06,
+      "loss": 0.3384,
+      "step": 2764
+    },
+    {
+      "epoch": 1.2858471554797706,
+      "grad_norm": 0.44974416494369507,
+      "learning_rate": 7.054673592584289e-06,
+      "loss": 0.3329,
+      "step": 2765
+    },
+    {
+      "epoch": 1.2863121996589677,
+      "grad_norm": 0.4763135313987732,
+      "learning_rate": 7.052206384535115e-06,
+      "loss": 0.3411,
+      "step": 2766
+    },
+    {
+      "epoch": 1.2867772438381646,
+      "grad_norm": 0.44069865345954895,
+      "learning_rate": 7.0497385754280255e-06,
+      "loss": 0.351,
+      "step": 2767
+    },
+    {
+      "epoch": 1.2872422880173615,
+      "grad_norm": 0.40506044030189514,
+      "learning_rate": 7.047270165985801e-06,
+      "loss": 0.3856,
+      "step": 2768
+    },
+    {
+      "epoch": 1.2877073321965586,
+      "grad_norm": 0.4121856689453125,
+      "learning_rate": 7.0448011569314e-06,
+      "loss": 0.3631,
+      "step": 2769
+    },
+    {
+      "epoch": 1.2881723763757558,
+      "grad_norm": 0.4210520088672638,
+      "learning_rate": 7.042331548987953e-06,
+      "loss": 0.3079,
+      "step": 2770
+    },
+    {
+      "epoch": 1.2886374205549527,
+      "grad_norm": 0.46315163373947144,
+      "learning_rate": 7.039861342878769e-06,
+      "loss": 0.3727,
+      "step": 2771
+    },
+    {
+      "epoch": 1.2891024647341498,
+      "grad_norm": 0.43790408968925476,
+      "learning_rate": 7.037390539327332e-06,
+      "loss": 0.3682,
+      "step": 2772
+    },
+    {
+      "epoch": 1.2895675089133467,
+      "grad_norm": 0.4229748845100403,
+      "learning_rate": 7.0349191390573e-06,
+      "loss": 0.3808,
+      "step": 2773
+    },
+    {
+      "epoch": 1.2900325530925438,
+      "grad_norm": 0.44016340374946594,
+      "learning_rate": 7.032447142792506e-06,
+      "loss": 0.3573,
+      "step": 2774
+    },
+    {
+      "epoch": 1.2904975972717407,
+      "grad_norm": 0.4395344853401184,
+      "learning_rate": 7.029974551256957e-06,
+      "loss": 0.3609,
+      "step": 2775
+    },
+    {
+      "epoch": 1.2909626414509379,
+      "grad_norm": 0.4566013514995575,
+      "learning_rate": 7.027501365174833e-06,
+      "loss": 0.3355,
+      "step": 2776
+    },
+    {
+      "epoch": 1.291427685630135,
+      "grad_norm": 0.44307535886764526,
+      "learning_rate": 7.025027585270495e-06,
+      "loss": 0.3421,
+      "step": 2777
+    },
+    {
+      "epoch": 1.291892729809332,
+      "grad_norm": 0.4520743489265442,
+      "learning_rate": 7.022553212268469e-06,
+      "loss": 0.3466,
+      "step": 2778
+    },
+    {
+      "epoch": 1.2923577739885288,
+      "grad_norm": 0.4748930335044861,
+      "learning_rate": 7.0200782468934605e-06,
+      "loss": 0.359,
+      "step": 2779
+    },
+    {
+      "epoch": 1.292822818167726,
+      "grad_norm": 0.4355354905128479,
+      "learning_rate": 7.017602689870345e-06,
+      "loss": 0.3346,
+      "step": 2780
+    },
+    {
+      "epoch": 1.293287862346923,
+      "grad_norm": 0.4658071994781494,
+      "learning_rate": 7.015126541924174e-06,
+      "loss": 0.3986,
+      "step": 2781
+    },
+    {
+      "epoch": 1.29375290652612,
+      "grad_norm": 0.45585206151008606,
+      "learning_rate": 7.012649803780171e-06,
+      "loss": 0.3595,
+      "step": 2782
+    },
+    {
+      "epoch": 1.294217950705317,
+      "grad_norm": 0.4219159781932831,
+      "learning_rate": 7.010172476163732e-06,
+      "loss": 0.3379,
+      "step": 2783
+    },
+    {
+      "epoch": 1.294682994884514,
+      "grad_norm": 0.40543264150619507,
+      "learning_rate": 7.007694559800427e-06,
+      "loss": 0.3518,
+      "step": 2784
+    },
+    {
+      "epoch": 1.2951480390637111,
+      "grad_norm": 0.4673357605934143,
+      "learning_rate": 7.005216055415996e-06,
+      "loss": 0.3548,
+      "step": 2785
+    },
+    {
+      "epoch": 1.295613083242908,
+      "grad_norm": 0.43243858218193054,
+      "learning_rate": 7.002736963736354e-06,
+      "loss": 0.3667,
+      "step": 2786
+    },
+    {
+      "epoch": 1.2960781274221052,
+      "grad_norm": 0.4363313913345337,
+      "learning_rate": 7.000257285487586e-06,
+      "loss": 0.3485,
+      "step": 2787
+    },
+    {
+      "epoch": 1.296543171601302,
+      "grad_norm": 0.44257044792175293,
+      "learning_rate": 6.997777021395949e-06,
+      "loss": 0.3625,
+      "step": 2788
+    },
+    {
+      "epoch": 1.2970082157804992,
+      "grad_norm": 0.448543906211853,
+      "learning_rate": 6.9952961721878734e-06,
+      "loss": 0.373,
+      "step": 2789
+    },
+    {
+      "epoch": 1.297473259959696,
+      "grad_norm": 0.41273754835128784,
+      "learning_rate": 6.992814738589958e-06,
+      "loss": 0.3761,
+      "step": 2790
+    },
+    {
+      "epoch": 1.2979383041388932,
+      "grad_norm": 0.540025532245636,
+      "learning_rate": 6.990332721328978e-06,
+      "loss": 0.3636,
+      "step": 2791
+    },
+    {
+      "epoch": 1.2984033483180903,
+      "grad_norm": 0.42229142785072327,
+      "learning_rate": 6.987850121131871e-06,
+      "loss": 0.3319,
+      "step": 2792
+    },
+    {
+      "epoch": 1.2988683924972872,
+      "grad_norm": 0.4574054777622223,
+      "learning_rate": 6.985366938725751e-06,
+      "loss": 0.3782,
+      "step": 2793
+    },
+    {
+      "epoch": 1.2993334366764842,
+      "grad_norm": 0.4920805096626282,
+      "learning_rate": 6.982883174837904e-06,
+      "loss": 0.376,
+      "step": 2794
+    },
+    {
+      "epoch": 1.2997984808556813,
+      "grad_norm": 0.40307578444480896,
+      "learning_rate": 6.980398830195785e-06,
+      "loss": 0.3285,
+      "step": 2795
+    },
+    {
+      "epoch": 1.3002635250348784,
+      "grad_norm": 0.41613009572029114,
+      "learning_rate": 6.977913905527016e-06,
+      "loss": 0.3433,
+      "step": 2796
+    },
+    {
+      "epoch": 1.3007285692140753,
+      "grad_norm": 0.44297316670417786,
+      "learning_rate": 6.975428401559392e-06,
+      "loss": 0.3558,
+      "step": 2797
+    },
+    {
+      "epoch": 1.3011936133932724,
+      "grad_norm": 0.3465144634246826,
+      "learning_rate": 6.972942319020876e-06,
+      "loss": 0.3213,
+      "step": 2798
+    },
+    {
+      "epoch": 1.3016586575724693,
+      "grad_norm": 0.40050673484802246,
+      "learning_rate": 6.9704556586396e-06,
+      "loss": 0.3431,
+      "step": 2799
+    },
+    {
+      "epoch": 1.3021237017516665,
+      "grad_norm": 0.45212817192077637,
+      "learning_rate": 6.967968421143869e-06,
+      "loss": 0.3546,
+      "step": 2800
+    },
+    {
+      "epoch": 1.3025887459308634,
+      "grad_norm": 0.4688425064086914,
+      "learning_rate": 6.965480607262154e-06,
+      "loss": 0.3559,
+      "step": 2801
+    },
+    {
+      "epoch": 1.3030537901100605,
+      "grad_norm": 0.4439482092857361,
+      "learning_rate": 6.962992217723094e-06,
+      "loss": 0.3234,
+      "step": 2802
+    },
+    {
+      "epoch": 1.3035188342892574,
+      "grad_norm": 0.43428176641464233,
+      "learning_rate": 6.960503253255499e-06,
+      "loss": 0.3435,
+      "step": 2803
+    },
+    {
+      "epoch": 1.3039838784684545,
+      "grad_norm": 0.3863378167152405,
+      "learning_rate": 6.958013714588348e-06,
+      "loss": 0.3557,
+      "step": 2804
+    },
+    {
+      "epoch": 1.3044489226476514,
+      "grad_norm": 0.40679073333740234,
+      "learning_rate": 6.95552360245078e-06,
+      "loss": 0.3292,
+      "step": 2805
+    },
+    {
+      "epoch": 1.3049139668268486,
+      "grad_norm": 0.45496124029159546,
+      "learning_rate": 6.953032917572116e-06,
+      "loss": 0.3961,
+      "step": 2806
+    },
+    {
+      "epoch": 1.3053790110060457,
+      "grad_norm": 0.3845018446445465,
+      "learning_rate": 6.950541660681833e-06,
+      "loss": 0.3522,
+      "step": 2807
+    },
+    {
+      "epoch": 1.3058440551852426,
+      "grad_norm": 0.4297010004520416,
+      "learning_rate": 6.94804983250958e-06,
+      "loss": 0.3595,
+      "step": 2808
+    },
+    {
+      "epoch": 1.3063090993644395,
+      "grad_norm": 0.3819620609283447,
+      "learning_rate": 6.945557433785175e-06,
+      "loss": 0.3066,
+      "step": 2809
+    },
+    {
+      "epoch": 1.3067741435436366,
+      "grad_norm": 0.42445626854896545,
+      "learning_rate": 6.943064465238598e-06,
+      "loss": 0.3727,
+      "step": 2810
+    },
+    {
+      "epoch": 1.3072391877228338,
+      "grad_norm": 0.4176945686340332,
+      "learning_rate": 6.940570927600001e-06,
+      "loss": 0.3371,
+      "step": 2811
+    },
+    {
+      "epoch": 1.3077042319020307,
+      "grad_norm": 0.4216483235359192,
+      "learning_rate": 6.9380768215996995e-06,
+      "loss": 0.337,
+      "step": 2812
+    },
+    {
+      "epoch": 1.3081692760812278,
+      "grad_norm": 0.4133038818836212,
+      "learning_rate": 6.93558214796818e-06,
+      "loss": 0.3371,
+      "step": 2813
+    },
+    {
+      "epoch": 1.3086343202604247,
+      "grad_norm": 0.47481316328048706,
+      "learning_rate": 6.933086907436087e-06,
+      "loss": 0.4007,
+      "step": 2814
+    },
+    {
+      "epoch": 1.3090993644396218,
+      "grad_norm": 0.41611889004707336,
+      "learning_rate": 6.93059110073424e-06,
+      "loss": 0.3471,
+      "step": 2815
+    },
+    {
+      "epoch": 1.3095644086188187,
+      "grad_norm": 0.45148274302482605,
+      "learning_rate": 6.928094728593617e-06,
+      "loss": 0.3308,
+      "step": 2816
+    },
+    {
+      "epoch": 1.3100294527980159,
+      "grad_norm": 0.4194914400577545,
+      "learning_rate": 6.9255977917453665e-06,
+      "loss": 0.3708,
+      "step": 2817
+    },
+    {
+      "epoch": 1.3104944969772128,
+      "grad_norm": 0.472514808177948,
+      "learning_rate": 6.923100290920801e-06,
+      "loss": 0.3499,
+      "step": 2818
+    },
+    {
+      "epoch": 1.31095954115641,
+      "grad_norm": 0.4619661569595337,
+      "learning_rate": 6.920602226851397e-06,
+      "loss": 0.3644,
+      "step": 2819
+    },
+    {
+      "epoch": 1.3114245853356068,
+      "grad_norm": 0.3975506126880646,
+      "learning_rate": 6.9181036002687985e-06,
+      "loss": 0.3624,
+      "step": 2820
+    },
+    {
+      "epoch": 1.311889629514804,
+      "grad_norm": 0.48552754521369934,
+      "learning_rate": 6.91560441190481e-06,
+      "loss": 0.3961,
+      "step": 2821
+    },
+    {
+      "epoch": 1.312354673694001,
+      "grad_norm": 0.4577177166938782,
+      "learning_rate": 6.913104662491406e-06,
+      "loss": 0.3915,
+      "step": 2822
+    },
+    {
+      "epoch": 1.312819717873198,
+      "grad_norm": 0.47331976890563965,
+      "learning_rate": 6.910604352760721e-06,
+      "loss": 0.3035,
+      "step": 2823
+    },
+    {
+      "epoch": 1.3132847620523949,
+      "grad_norm": 0.4498820900917053,
+      "learning_rate": 6.908103483445056e-06,
+      "loss": 0.3242,
+      "step": 2824
+    },
+    {
+      "epoch": 1.313749806231592,
+      "grad_norm": 0.5082352757453918,
+      "learning_rate": 6.905602055276872e-06,
+      "loss": 0.3974,
+      "step": 2825
+    },
+    {
+      "epoch": 1.3142148504107891,
+      "grad_norm": 0.4770672023296356,
+      "learning_rate": 6.9031000689888015e-06,
+      "loss": 0.3467,
+      "step": 2826
+    },
+    {
+      "epoch": 1.314679894589986,
+      "grad_norm": 0.5072925090789795,
+      "learning_rate": 6.9005975253136324e-06,
+      "loss": 0.3451,
+      "step": 2827
+    },
+    {
+      "epoch": 1.3151449387691831,
+      "grad_norm": 0.4119420051574707,
+      "learning_rate": 6.898094424984319e-06,
+      "loss": 0.3581,
+      "step": 2828
+    },
+    {
+      "epoch": 1.31560998294838,
+      "grad_norm": 0.453228235244751,
+      "learning_rate": 6.89559076873398e-06,
+      "loss": 0.3201,
+      "step": 2829
+    },
+    {
+      "epoch": 1.3160750271275772,
+      "grad_norm": 0.5295499563217163,
+      "learning_rate": 6.893086557295896e-06,
+      "loss": 0.387,
+      "step": 2830
+    },
+    {
+      "epoch": 1.316540071306774,
+      "grad_norm": 0.428733229637146,
+      "learning_rate": 6.890581791403509e-06,
+      "loss": 0.3681,
+      "step": 2831
+    },
+    {
+      "epoch": 1.3170051154859712,
+      "grad_norm": 0.4751463234424591,
+      "learning_rate": 6.888076471790423e-06,
+      "loss": 0.345,
+      "step": 2832
+    },
+    {
+      "epoch": 1.3174701596651681,
+      "grad_norm": 0.44530734419822693,
+      "learning_rate": 6.885570599190408e-06,
+      "loss": 0.3325,
+      "step": 2833
+    },
+    {
+      "epoch": 1.3179352038443652,
+      "grad_norm": 0.5325456857681274,
+      "learning_rate": 6.88306417433739e-06,
+      "loss": 0.3644,
+      "step": 2834
+    },
+    {
+      "epoch": 1.3184002480235621,
+      "grad_norm": 0.3941551446914673,
+      "learning_rate": 6.880557197965465e-06,
+      "loss": 0.3159,
+      "step": 2835
+    },
+    {
+      "epoch": 1.3188652922027593,
+      "grad_norm": 0.40330174565315247,
+      "learning_rate": 6.878049670808882e-06,
+      "loss": 0.3403,
+      "step": 2836
+    },
+    {
+      "epoch": 1.3193303363819564,
+      "grad_norm": 0.5025704503059387,
+      "learning_rate": 6.875541593602055e-06,
+      "loss": 0.3683,
+      "step": 2837
+    },
+    {
+      "epoch": 1.3197953805611533,
+      "grad_norm": 0.48347654938697815,
+      "learning_rate": 6.873032967079562e-06,
+      "loss": 0.3385,
+      "step": 2838
+    },
+    {
+      "epoch": 1.3202604247403502,
+      "grad_norm": 0.45224472880363464,
+      "learning_rate": 6.8705237919761344e-06,
+      "loss": 0.4031,
+      "step": 2839
+    },
+    {
+      "epoch": 1.3207254689195473,
+      "grad_norm": 0.4854915142059326,
+      "learning_rate": 6.868014069026672e-06,
+      "loss": 0.3297,
+      "step": 2840
+    },
+    {
+      "epoch": 1.3211905130987445,
+      "grad_norm": 0.5176591873168945,
+      "learning_rate": 6.865503798966232e-06,
+      "loss": 0.3377,
+      "step": 2841
+    },
+    {
+      "epoch": 1.3216555572779414,
+      "grad_norm": 0.43474873900413513,
+      "learning_rate": 6.86299298253003e-06,
+      "loss": 0.3378,
+      "step": 2842
+    },
+    {
+      "epoch": 1.3221206014571385,
+      "grad_norm": 0.5134937167167664,
+      "learning_rate": 6.860481620453445e-06,
+      "loss": 0.3775,
+      "step": 2843
+    },
+    {
+      "epoch": 1.3225856456363354,
+      "grad_norm": 0.47907009720802307,
+      "learning_rate": 6.857969713472015e-06,
+      "loss": 0.3326,
+      "step": 2844
+    },
+    {
+      "epoch": 1.3230506898155325,
+      "grad_norm": 0.4050777554512024,
+      "learning_rate": 6.855457262321433e-06,
+      "loss": 0.3613,
+      "step": 2845
+    },
+    {
+      "epoch": 1.3235157339947294,
+      "grad_norm": 0.4659248888492584,
+      "learning_rate": 6.852944267737557e-06,
+      "loss": 0.3669,
+      "step": 2846
+    },
+    {
+      "epoch": 1.3239807781739266,
+      "grad_norm": 0.42381158471107483,
+      "learning_rate": 6.850430730456403e-06,
+      "loss": 0.37,
+      "step": 2847
+    },
+    {
+      "epoch": 1.3244458223531235,
+      "grad_norm": 0.39930933713912964,
+      "learning_rate": 6.847916651214146e-06,
+      "loss": 0.3577,
+      "step": 2848
+    },
+    {
+      "epoch": 1.3249108665323206,
+      "grad_norm": 0.3815561532974243,
+      "learning_rate": 6.845402030747118e-06,
+      "loss": 0.3432,
+      "step": 2849
+    },
+    {
+      "epoch": 1.3253759107115175,
+      "grad_norm": 0.41214823722839355,
+      "learning_rate": 6.84288686979181e-06,
+      "loss": 0.3228,
+      "step": 2850
+    },
+    {
+      "epoch": 1.3258409548907146,
+      "grad_norm": 0.41365644335746765,
+      "learning_rate": 6.8403711690848715e-06,
+      "loss": 0.3858,
+      "step": 2851
+    },
+    {
+      "epoch": 1.3263059990699118,
+      "grad_norm": 0.4793543219566345,
+      "learning_rate": 6.83785492936311e-06,
+      "loss": 0.3466,
+      "step": 2852
+    },
+    {
+      "epoch": 1.3267710432491087,
+      "grad_norm": 0.39055493474006653,
+      "learning_rate": 6.8353381513634945e-06,
+      "loss": 0.3124,
+      "step": 2853
+    },
+    {
+      "epoch": 1.3272360874283056,
+      "grad_norm": 0.37652844190597534,
+      "learning_rate": 6.832820835823145e-06,
+      "loss": 0.3251,
+      "step": 2854
+    },
+    {
+      "epoch": 1.3277011316075027,
+      "grad_norm": 0.4325464069843292,
+      "learning_rate": 6.830302983479344e-06,
+      "loss": 0.3776,
+      "step": 2855
+    },
+    {
+      "epoch": 1.3281661757866998,
+      "grad_norm": 0.3975974917411804,
+      "learning_rate": 6.827784595069529e-06,
+      "loss": 0.3446,
+      "step": 2856
+    },
+    {
+      "epoch": 1.3286312199658967,
+      "grad_norm": 0.3837142288684845,
+      "learning_rate": 6.825265671331295e-06,
+      "loss": 0.3613,
+      "step": 2857
+    },
+    {
+      "epoch": 1.3290962641450939,
+      "grad_norm": 0.44566547870635986,
+      "learning_rate": 6.822746213002393e-06,
+      "loss": 0.3458,
+      "step": 2858
+    },
+    {
+      "epoch": 1.3295613083242908,
+      "grad_norm": 0.38540929555892944,
+      "learning_rate": 6.820226220820733e-06,
+      "loss": 0.3582,
+      "step": 2859
+    },
+    {
+      "epoch": 1.3300263525034879,
+      "grad_norm": 0.4004163444042206,
+      "learning_rate": 6.8177056955243794e-06,
+      "loss": 0.3327,
+      "step": 2860
+    },
+    {
+      "epoch": 1.3304913966826848,
+      "grad_norm": 0.3982474207878113,
+      "learning_rate": 6.815184637851553e-06,
+      "loss": 0.3696,
+      "step": 2861
+    },
+    {
+      "epoch": 1.330956440861882,
+      "grad_norm": 0.36453261971473694,
+      "learning_rate": 6.812663048540631e-06,
+      "loss": 0.3276,
+      "step": 2862
+    },
+    {
+      "epoch": 1.3314214850410788,
+      "grad_norm": 0.41965192556381226,
+      "learning_rate": 6.810140928330144e-06,
+      "loss": 0.3681,
+      "step": 2863
+    },
+    {
+      "epoch": 1.331886529220276,
+      "grad_norm": 0.4125366806983948,
+      "learning_rate": 6.807618277958783e-06,
+      "loss": 0.3285,
+      "step": 2864
+    },
+    {
+      "epoch": 1.3323515733994729,
+      "grad_norm": 0.3705972731113434,
+      "learning_rate": 6.805095098165388e-06,
+      "loss": 0.3291,
+      "step": 2865
+    },
+    {
+      "epoch": 1.33281661757867,
+      "grad_norm": 0.48080259561538696,
+      "learning_rate": 6.8025713896889615e-06,
+      "loss": 0.3379,
+      "step": 2866
+    },
+    {
+      "epoch": 1.333281661757867,
+      "grad_norm": 0.39710700511932373,
+      "learning_rate": 6.800047153268653e-06,
+      "loss": 0.3732,
+      "step": 2867
+    },
+    {
+      "epoch": 1.333746705937064,
+      "grad_norm": 0.3951427638530731,
+      "learning_rate": 6.797522389643772e-06,
+      "loss": 0.3569,
+      "step": 2868
+    },
+    {
+      "epoch": 1.334211750116261,
+      "grad_norm": 0.33918821811676025,
+      "learning_rate": 6.79499709955378e-06,
+      "loss": 0.3214,
+      "step": 2869
+    },
+    {
+      "epoch": 1.334676794295458,
+      "grad_norm": 0.39031147956848145,
+      "learning_rate": 6.792471283738293e-06,
+      "loss": 0.3722,
+      "step": 2870
+    },
+    {
+      "epoch": 1.3351418384746552,
+      "grad_norm": 0.3802085518836975,
+      "learning_rate": 6.789944942937084e-06,
+      "loss": 0.3364,
+      "step": 2871
+    },
+    {
+      "epoch": 1.335606882653852,
+      "grad_norm": 0.36190685629844666,
+      "learning_rate": 6.787418077890076e-06,
+      "loss": 0.3239,
+      "step": 2872
+    },
+    {
+      "epoch": 1.3360719268330492,
+      "grad_norm": 0.4449096620082855,
+      "learning_rate": 6.784890689337346e-06,
+      "loss": 0.3719,
+      "step": 2873
+    },
+    {
+      "epoch": 1.3365369710122461,
+      "grad_norm": 0.4153344929218292,
+      "learning_rate": 6.782362778019125e-06,
+      "loss": 0.3713,
+      "step": 2874
+    },
+    {
+      "epoch": 1.3370020151914432,
+      "grad_norm": 0.3987843692302704,
+      "learning_rate": 6.779834344675797e-06,
+      "loss": 0.3889,
+      "step": 2875
+    },
+    {
+      "epoch": 1.3374670593706401,
+      "grad_norm": 0.39704734086990356,
+      "learning_rate": 6.7773053900478995e-06,
+      "loss": 0.2937,
+      "step": 2876
+    },
+    {
+      "epoch": 1.3379321035498373,
+      "grad_norm": 0.42171093821525574,
+      "learning_rate": 6.774775914876123e-06,
+      "loss": 0.3604,
+      "step": 2877
+    },
+    {
+      "epoch": 1.3383971477290342,
+      "grad_norm": 0.42374035716056824,
+      "learning_rate": 6.7722459199013095e-06,
+      "loss": 0.3642,
+      "step": 2878
+    },
+    {
+      "epoch": 1.3388621919082313,
+      "grad_norm": 0.43570584058761597,
+      "learning_rate": 6.769715405864452e-06,
+      "loss": 0.391,
+      "step": 2879
+    },
+    {
+      "epoch": 1.3393272360874282,
+      "grad_norm": 0.361253947019577,
+      "learning_rate": 6.767184373506698e-06,
+      "loss": 0.3158,
+      "step": 2880
+    },
+    {
+      "epoch": 1.3397922802666253,
+      "grad_norm": 0.46323809027671814,
+      "learning_rate": 6.7646528235693445e-06,
+      "loss": 0.337,
+      "step": 2881
+    },
+    {
+      "epoch": 1.3402573244458225,
+      "grad_norm": 0.5015746355056763,
+      "learning_rate": 6.762120756793844e-06,
+      "loss": 0.3892,
+      "step": 2882
+    },
+    {
+      "epoch": 1.3407223686250194,
+      "grad_norm": 0.43376925587654114,
+      "learning_rate": 6.759588173921796e-06,
+      "loss": 0.3749,
+      "step": 2883
+    },
+    {
+      "epoch": 1.3411874128042163,
+      "grad_norm": 0.4221652150154114,
+      "learning_rate": 6.757055075694954e-06,
+      "loss": 0.3531,
+      "step": 2884
+    },
+    {
+      "epoch": 1.3416524569834134,
+      "grad_norm": 0.49050408601760864,
+      "learning_rate": 6.754521462855219e-06,
+      "loss": 0.3557,
+      "step": 2885
+    },
+    {
+      "epoch": 1.3421175011626105,
+      "grad_norm": 0.4941301941871643,
+      "learning_rate": 6.7519873361446475e-06,
+      "loss": 0.3587,
+      "step": 2886
+    },
+    {
+      "epoch": 1.3425825453418074,
+      "grad_norm": 0.378880113363266,
+      "learning_rate": 6.749452696305442e-06,
+      "loss": 0.3779,
+      "step": 2887
+    },
+    {
+      "epoch": 1.3430475895210046,
+      "grad_norm": 0.41799113154411316,
+      "learning_rate": 6.74691754407996e-06,
+      "loss": 0.339,
+      "step": 2888
+    },
+    {
+      "epoch": 1.3435126337002015,
+      "grad_norm": 0.39678072929382324,
+      "learning_rate": 6.744381880210703e-06,
+      "loss": 0.3487,
+      "step": 2889
+    },
+    {
+      "epoch": 1.3439776778793986,
+      "grad_norm": 0.41609761118888855,
+      "learning_rate": 6.741845705440329e-06,
+      "loss": 0.3937,
+      "step": 2890
+    },
+    {
+      "epoch": 1.3444427220585955,
+      "grad_norm": 0.4008464217185974,
+      "learning_rate": 6.739309020511641e-06,
+      "loss": 0.3155,
+      "step": 2891
+    },
+    {
+      "epoch": 1.3449077662377926,
+      "grad_norm": 0.45266151428222656,
+      "learning_rate": 6.736771826167592e-06,
+      "loss": 0.3634,
+      "step": 2892
+    },
+    {
+      "epoch": 1.3453728104169898,
+      "grad_norm": 0.423709511756897,
+      "learning_rate": 6.734234123151284e-06,
+      "loss": 0.3839,
+      "step": 2893
+    },
+    {
+      "epoch": 1.3458378545961867,
+      "grad_norm": 0.4130350649356842,
+      "learning_rate": 6.731695912205974e-06,
+      "loss": 0.3357,
+      "step": 2894
+    },
+    {
+      "epoch": 1.3463028987753836,
+      "grad_norm": 0.38522496819496155,
+      "learning_rate": 6.7291571940750575e-06,
+      "loss": 0.316,
+      "step": 2895
+    },
+    {
+      "epoch": 1.3467679429545807,
+      "grad_norm": 0.47391700744628906,
+      "learning_rate": 6.726617969502088e-06,
+      "loss": 0.3453,
+      "step": 2896
+    },
+    {
+      "epoch": 1.3472329871337778,
+      "grad_norm": 0.3721844255924225,
+      "learning_rate": 6.724078239230758e-06,
+      "loss": 0.3274,
+      "step": 2897
+    },
+    {
+      "epoch": 1.3476980313129747,
+      "grad_norm": 0.45813339948654175,
+      "learning_rate": 6.721538004004918e-06,
+      "loss": 0.3747,
+      "step": 2898
+    },
+    {
+      "epoch": 1.3481630754921716,
+      "grad_norm": 0.44626984000205994,
+      "learning_rate": 6.71899726456856e-06,
+      "loss": 0.3592,
+      "step": 2899
+    },
+    {
+      "epoch": 1.3486281196713688,
+      "grad_norm": 0.4165315330028534,
+      "learning_rate": 6.716456021665825e-06,
+      "loss": 0.3405,
+      "step": 2900
+    },
+    {
+      "epoch": 1.3490931638505659,
+      "grad_norm": 0.40565526485443115,
+      "learning_rate": 6.713914276041001e-06,
+      "loss": 0.351,
+      "step": 2901
+    },
+    {
+      "epoch": 1.3495582080297628,
+      "grad_norm": 0.4150947332382202,
+      "learning_rate": 6.7113720284385255e-06,
+      "loss": 0.3609,
+      "step": 2902
+    },
+    {
+      "epoch": 1.35002325220896,
+      "grad_norm": 0.35937878489494324,
+      "learning_rate": 6.708829279602982e-06,
+      "loss": 0.3121,
+      "step": 2903
+    },
+    {
+      "epoch": 1.3504882963881568,
+      "grad_norm": 0.4622684121131897,
+      "learning_rate": 6.7062860302790965e-06,
+      "loss": 0.3757,
+      "step": 2904
+    },
+    {
+      "epoch": 1.350953340567354,
+      "grad_norm": 0.3929741680622101,
+      "learning_rate": 6.70374228121175e-06,
+      "loss": 0.3468,
+      "step": 2905
+    },
+    {
+      "epoch": 1.3514183847465508,
+      "grad_norm": 0.3833123743534088,
+      "learning_rate": 6.7011980331459635e-06,
+      "loss": 0.3254,
+      "step": 2906
+    },
+    {
+      "epoch": 1.351883428925748,
+      "grad_norm": 0.4471578598022461,
+      "learning_rate": 6.698653286826906e-06,
+      "loss": 0.3819,
+      "step": 2907
+    },
+    {
+      "epoch": 1.352348473104945,
+      "grad_norm": 0.42805638909339905,
+      "learning_rate": 6.696108042999892e-06,
+      "loss": 0.3425,
+      "step": 2908
+    },
+    {
+      "epoch": 1.352813517284142,
+      "grad_norm": 0.40058985352516174,
+      "learning_rate": 6.693562302410384e-06,
+      "loss": 0.3844,
+      "step": 2909
+    },
+    {
+      "epoch": 1.353278561463339,
+      "grad_norm": 0.417976051568985,
+      "learning_rate": 6.6910160658039835e-06,
+      "loss": 0.3821,
+      "step": 2910
+    },
+    {
+      "epoch": 1.353743605642536,
+      "grad_norm": 0.36969244480133057,
+      "learning_rate": 6.6884693339264466e-06,
+      "loss": 0.3038,
+      "step": 2911
+    },
+    {
+      "epoch": 1.3542086498217332,
+      "grad_norm": 0.36144527792930603,
+      "learning_rate": 6.685922107523667e-06,
+      "loss": 0.3405,
+      "step": 2912
+    },
+    {
+      "epoch": 1.35467369400093,
+      "grad_norm": 0.4299263060092926,
+      "learning_rate": 6.683374387341688e-06,
+      "loss": 0.3538,
+      "step": 2913
+    },
+    {
+      "epoch": 1.355138738180127,
+      "grad_norm": 0.41073358058929443,
+      "learning_rate": 6.680826174126693e-06,
+      "loss": 0.3451,
+      "step": 2914
+    },
+    {
+      "epoch": 1.355603782359324,
+      "grad_norm": 0.42276236414909363,
+      "learning_rate": 6.678277468625014e-06,
+      "loss": 0.3641,
+      "step": 2915
+    },
+    {
+      "epoch": 1.3560688265385212,
+      "grad_norm": 0.39464181661605835,
+      "learning_rate": 6.675728271583124e-06,
+      "loss": 0.3313,
+      "step": 2916
+    },
+    {
+      "epoch": 1.3565338707177181,
+      "grad_norm": 0.3857058882713318,
+      "learning_rate": 6.673178583747644e-06,
+      "loss": 0.3363,
+      "step": 2917
+    },
+    {
+      "epoch": 1.3569989148969153,
+      "grad_norm": 0.40449684858322144,
+      "learning_rate": 6.670628405865334e-06,
+      "loss": 0.3499,
+      "step": 2918
+    },
+    {
+      "epoch": 1.3574639590761122,
+      "grad_norm": 0.43879038095474243,
+      "learning_rate": 6.6680777386831e-06,
+      "loss": 0.3781,
+      "step": 2919
+    },
+    {
+      "epoch": 1.3579290032553093,
+      "grad_norm": 0.4013797640800476,
+      "learning_rate": 6.665526582947991e-06,
+      "loss": 0.3759,
+      "step": 2920
+    },
+    {
+      "epoch": 1.3583940474345062,
+      "grad_norm": 0.5068005323410034,
+      "learning_rate": 6.6629749394071995e-06,
+      "loss": 0.3319,
+      "step": 2921
+    },
+    {
+      "epoch": 1.3588590916137033,
+      "grad_norm": 0.43432286381721497,
+      "learning_rate": 6.6604228088080605e-06,
+      "loss": 0.3545,
+      "step": 2922
+    },
+    {
+      "epoch": 1.3593241357929005,
+      "grad_norm": 0.5053911805152893,
+      "learning_rate": 6.657870191898051e-06,
+      "loss": 0.3545,
+      "step": 2923
+    },
+    {
+      "epoch": 1.3597891799720974,
+      "grad_norm": 0.37629908323287964,
+      "learning_rate": 6.655317089424791e-06,
+      "loss": 0.3084,
+      "step": 2924
+    },
+    {
+      "epoch": 1.3602542241512943,
+      "grad_norm": 0.40685826539993286,
+      "learning_rate": 6.652763502136044e-06,
+      "loss": 0.3445,
+      "step": 2925
+    },
+    {
+      "epoch": 1.3607192683304914,
+      "grad_norm": 0.46807217597961426,
+      "learning_rate": 6.6502094307797124e-06,
+      "loss": 0.3464,
+      "step": 2926
+    },
+    {
+      "epoch": 1.3611843125096885,
+      "grad_norm": 0.43712517619132996,
+      "learning_rate": 6.647654876103844e-06,
+      "loss": 0.3325,
+      "step": 2927
+    },
+    {
+      "epoch": 1.3616493566888854,
+      "grad_norm": 0.5149297118186951,
+      "learning_rate": 6.645099838856624e-06,
+      "loss": 0.405,
+      "step": 2928
+    },
+    {
+      "epoch": 1.3621144008680823,
+      "grad_norm": 0.41195258498191833,
+      "learning_rate": 6.6425443197863836e-06,
+      "loss": 0.3373,
+      "step": 2929
+    },
+    {
+      "epoch": 1.3625794450472795,
+      "grad_norm": 0.5708654522895813,
+      "learning_rate": 6.639988319641592e-06,
+      "loss": 0.3448,
+      "step": 2930
+    },
+    {
+      "epoch": 1.3630444892264766,
+      "grad_norm": 0.45768019556999207,
+      "learning_rate": 6.637431839170861e-06,
+      "loss": 0.3502,
+      "step": 2931
+    },
+    {
+      "epoch": 1.3635095334056735,
+      "grad_norm": 0.42099401354789734,
+      "learning_rate": 6.6348748791229416e-06,
+      "loss": 0.3665,
+      "step": 2932
+    },
+    {
+      "epoch": 1.3639745775848706,
+      "grad_norm": 0.45441335439682007,
+      "learning_rate": 6.632317440246725e-06,
+      "loss": 0.3322,
+      "step": 2933
+    },
+    {
+      "epoch": 1.3644396217640675,
+      "grad_norm": 0.5002641677856445,
+      "learning_rate": 6.629759523291242e-06,
+      "loss": 0.3342,
+      "step": 2934
+    },
+    {
+      "epoch": 1.3649046659432646,
+      "grad_norm": 0.4336529076099396,
+      "learning_rate": 6.627201129005672e-06,
+      "loss": 0.3839,
+      "step": 2935
+    },
+    {
+      "epoch": 1.3653697101224616,
+      "grad_norm": 0.47555792331695557,
+      "learning_rate": 6.624642258139318e-06,
+      "loss": 0.3358,
+      "step": 2936
+    },
+    {
+      "epoch": 1.3658347543016587,
+      "grad_norm": 0.47870880365371704,
+      "learning_rate": 6.622082911441637e-06,
+      "loss": 0.3573,
+      "step": 2937
+    },
+    {
+      "epoch": 1.3662997984808558,
+      "grad_norm": 0.44583725929260254,
+      "learning_rate": 6.619523089662219e-06,
+      "loss": 0.3536,
+      "step": 2938
+    },
+    {
+      "epoch": 1.3667648426600527,
+      "grad_norm": 0.44511839747428894,
+      "learning_rate": 6.616962793550794e-06,
+      "loss": 0.3496,
+      "step": 2939
+    },
+    {
+      "epoch": 1.3672298868392496,
+      "grad_norm": 0.4455788731575012,
+      "learning_rate": 6.614402023857231e-06,
+      "loss": 0.3345,
+      "step": 2940
+    },
+    {
+      "epoch": 1.3676949310184467,
+      "grad_norm": 0.42460909485816956,
+      "learning_rate": 6.61184078133154e-06,
+      "loss": 0.3473,
+      "step": 2941
+    },
+    {
+      "epoch": 1.3681599751976439,
+      "grad_norm": 0.44328975677490234,
+      "learning_rate": 6.609279066723865e-06,
+      "loss": 0.3733,
+      "step": 2942
+    },
+    {
+      "epoch": 1.3686250193768408,
+      "grad_norm": 0.4914466142654419,
+      "learning_rate": 6.606716880784491e-06,
+      "loss": 0.373,
+      "step": 2943
+    },
+    {
+      "epoch": 1.369090063556038,
+      "grad_norm": 0.4081343710422516,
+      "learning_rate": 6.604154224263839e-06,
+      "loss": 0.3409,
+      "step": 2944
+    },
+    {
+      "epoch": 1.3695551077352348,
+      "grad_norm": 0.44254809617996216,
+      "learning_rate": 6.601591097912472e-06,
+      "loss": 0.3659,
+      "step": 2945
+    },
+    {
+      "epoch": 1.370020151914432,
+      "grad_norm": 0.4107268452644348,
+      "learning_rate": 6.599027502481088e-06,
+      "loss": 0.3359,
+      "step": 2946
+    },
+    {
+      "epoch": 1.3704851960936288,
+      "grad_norm": 0.4272344410419464,
+      "learning_rate": 6.596463438720522e-06,
+      "loss": 0.3476,
+      "step": 2947
+    },
+    {
+      "epoch": 1.370950240272826,
+      "grad_norm": 0.47288623452186584,
+      "learning_rate": 6.593898907381746e-06,
+      "loss": 0.3282,
+      "step": 2948
+    },
+    {
+      "epoch": 1.3714152844520229,
+      "grad_norm": 0.4827193319797516,
+      "learning_rate": 6.59133390921587e-06,
+      "loss": 0.3738,
+      "step": 2949
+    },
+    {
+      "epoch": 1.37188032863122,
+      "grad_norm": 0.4004787504673004,
+      "learning_rate": 6.588768444974141e-06,
+      "loss": 0.3441,
+      "step": 2950
+    },
+    {
+      "epoch": 1.372345372810417,
+      "grad_norm": 0.44598275423049927,
+      "learning_rate": 6.58620251540794e-06,
+      "loss": 0.3349,
+      "step": 2951
+    },
+    {
+      "epoch": 1.372810416989614,
+      "grad_norm": 0.49700048565864563,
+      "learning_rate": 6.583636121268787e-06,
+      "loss": 0.3879,
+      "step": 2952
+    },
+    {
+      "epoch": 1.3732754611688112,
+      "grad_norm": 0.46100863814353943,
+      "learning_rate": 6.581069263308338e-06,
+      "loss": 0.3708,
+      "step": 2953
+    },
+    {
+      "epoch": 1.373740505348008,
+      "grad_norm": 0.47054529190063477,
+      "learning_rate": 6.5785019422783836e-06,
+      "loss": 0.352,
+      "step": 2954
+    },
+    {
+      "epoch": 1.374205549527205,
+      "grad_norm": 0.4136134386062622,
+      "learning_rate": 6.57593415893085e-06,
+      "loss": 0.3491,
+      "step": 2955
+    },
+    {
+      "epoch": 1.374670593706402,
+      "grad_norm": 0.49520325660705566,
+      "learning_rate": 6.5733659140178e-06,
+      "loss": 0.3456,
+      "step": 2956
+    },
+    {
+      "epoch": 1.3751356378855992,
+      "grad_norm": 0.3895581066608429,
+      "learning_rate": 6.5707972082914275e-06,
+      "loss": 0.3026,
+      "step": 2957
+    },
+    {
+      "epoch": 1.3756006820647961,
+      "grad_norm": 0.4688425660133362,
+      "learning_rate": 6.56822804250407e-06,
+      "loss": 0.3848,
+      "step": 2958
+    },
+    {
+      "epoch": 1.3760657262439933,
+      "grad_norm": 0.45133891701698303,
+      "learning_rate": 6.56565841740819e-06,
+      "loss": 0.3607,
+      "step": 2959
+    },
+    {
+      "epoch": 1.3765307704231902,
+      "grad_norm": 0.49955958127975464,
+      "learning_rate": 6.5630883337563935e-06,
+      "loss": 0.3493,
+      "step": 2960
+    },
+    {
+      "epoch": 1.3769958146023873,
+      "grad_norm": 0.43193817138671875,
+      "learning_rate": 6.560517792301412e-06,
+      "loss": 0.3393,
+      "step": 2961
+    },
+    {
+      "epoch": 1.3774608587815842,
+      "grad_norm": 0.4623286724090576,
+      "learning_rate": 6.557946793796116e-06,
+      "loss": 0.367,
+      "step": 2962
+    },
+    {
+      "epoch": 1.3779259029607813,
+      "grad_norm": 0.44610491394996643,
+      "learning_rate": 6.55537533899351e-06,
+      "loss": 0.3393,
+      "step": 2963
+    },
+    {
+      "epoch": 1.3783909471399782,
+      "grad_norm": 0.4373582601547241,
+      "learning_rate": 6.552803428646732e-06,
+      "loss": 0.3219,
+      "step": 2964
+    },
+    {
+      "epoch": 1.3788559913191754,
+      "grad_norm": 0.4308757185935974,
+      "learning_rate": 6.550231063509054e-06,
+      "loss": 0.3711,
+      "step": 2965
+    },
+    {
+      "epoch": 1.3793210354983723,
+      "grad_norm": 0.42997416853904724,
+      "learning_rate": 6.547658244333876e-06,
+      "loss": 0.3208,
+      "step": 2966
+    },
+    {
+      "epoch": 1.3797860796775694,
+      "grad_norm": 0.4055032432079315,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.3543,
+      "step": 2967
+    },
+    {
+      "epoch": 1.3802511238567665,
+      "grad_norm": 0.39875516295433044,
+      "learning_rate": 6.542511246885308e-06,
+      "loss": 0.3392,
+      "step": 2968
+    },
+    {
+      "epoch": 1.3807161680359634,
+      "grad_norm": 0.4401739239692688,
+      "learning_rate": 6.539937070119389e-06,
+      "loss": 0.3887,
+      "step": 2969
+    },
+    {
+      "epoch": 1.3811812122151603,
+      "grad_norm": 0.38288089632987976,
+      "learning_rate": 6.5373624423309165e-06,
+      "loss": 0.3448,
+      "step": 2970
+    },
+    {
+      "epoch": 1.3816462563943575,
+      "grad_norm": 0.44091928005218506,
+      "learning_rate": 6.534787364273957e-06,
+      "loss": 0.3297,
+      "step": 2971
+    },
+    {
+      "epoch": 1.3821113005735546,
+      "grad_norm": 0.5266174674034119,
+      "learning_rate": 6.532211836702708e-06,
+      "loss": 0.3979,
+      "step": 2972
+    },
+    {
+      "epoch": 1.3825763447527515,
+      "grad_norm": 0.40957939624786377,
+      "learning_rate": 6.529635860371497e-06,
+      "loss": 0.3353,
+      "step": 2973
+    },
+    {
+      "epoch": 1.3830413889319486,
+      "grad_norm": 0.4384903311729431,
+      "learning_rate": 6.527059436034791e-06,
+      "loss": 0.3386,
+      "step": 2974
+    },
+    {
+      "epoch": 1.3835064331111455,
+      "grad_norm": 0.3974456787109375,
+      "learning_rate": 6.524482564447181e-06,
+      "loss": 0.3046,
+      "step": 2975
+    },
+    {
+      "epoch": 1.3839714772903426,
+      "grad_norm": 0.43884211778640747,
+      "learning_rate": 6.521905246363389e-06,
+      "loss": 0.412,
+      "step": 2976
+    },
+    {
+      "epoch": 1.3844365214695395,
+      "grad_norm": 0.4392377436161041,
+      "learning_rate": 6.5193274825382724e-06,
+      "loss": 0.3413,
+      "step": 2977
+    },
+    {
+      "epoch": 1.3849015656487367,
+      "grad_norm": 0.4340946674346924,
+      "learning_rate": 6.516749273726814e-06,
+      "loss": 0.3686,
+      "step": 2978
+    },
+    {
+      "epoch": 1.3853666098279336,
+      "grad_norm": 0.3607385456562042,
+      "learning_rate": 6.514170620684128e-06,
+      "loss": 0.3132,
+      "step": 2979
+    },
+    {
+      "epoch": 1.3858316540071307,
+      "grad_norm": 0.5139884352684021,
+      "learning_rate": 6.511591524165465e-06,
+      "loss": 0.3757,
+      "step": 2980
+    },
+    {
+      "epoch": 1.3862966981863276,
+      "grad_norm": 0.43492573499679565,
+      "learning_rate": 6.509011984926197e-06,
+      "loss": 0.3282,
+      "step": 2981
+    },
+    {
+      "epoch": 1.3867617423655247,
+      "grad_norm": 0.3985266089439392,
+      "learning_rate": 6.50643200372183e-06,
+      "loss": 0.3391,
+      "step": 2982
+    },
+    {
+      "epoch": 1.3872267865447219,
+      "grad_norm": 0.5525508522987366,
+      "learning_rate": 6.503851581307997e-06,
+      "loss": 0.4244,
+      "step": 2983
+    },
+    {
+      "epoch": 1.3876918307239188,
+      "grad_norm": 0.3948424756526947,
+      "learning_rate": 6.5012707184404645e-06,
+      "loss": 0.3381,
+      "step": 2984
+    },
+    {
+      "epoch": 1.3881568749031157,
+      "grad_norm": 0.4646177589893341,
+      "learning_rate": 6.498689415875121e-06,
+      "loss": 0.3424,
+      "step": 2985
+    },
+    {
+      "epoch": 1.3886219190823128,
+      "grad_norm": 0.4283245801925659,
+      "learning_rate": 6.496107674367994e-06,
+      "loss": 0.3178,
+      "step": 2986
+    },
+    {
+      "epoch": 1.38908696326151,
+      "grad_norm": 0.45199066400527954,
+      "learning_rate": 6.49352549467523e-06,
+      "loss": 0.3601,
+      "step": 2987
+    },
+    {
+      "epoch": 1.3895520074407068,
+      "grad_norm": 0.3850160837173462,
+      "learning_rate": 6.4909428775531095e-06,
+      "loss": 0.3472,
+      "step": 2988
+    },
+    {
+      "epoch": 1.390017051619904,
+      "grad_norm": 0.4370267391204834,
+      "learning_rate": 6.488359823758036e-06,
+      "loss": 0.3827,
+      "step": 2989
+    },
+    {
+      "epoch": 1.3904820957991009,
+      "grad_norm": 0.4112876355648041,
+      "learning_rate": 6.485776334046546e-06,
+      "loss": 0.3564,
+      "step": 2990
+    },
+    {
+      "epoch": 1.390947139978298,
+      "grad_norm": 0.4564457833766937,
+      "learning_rate": 6.483192409175301e-06,
+      "loss": 0.3507,
+      "step": 2991
+    },
+    {
+      "epoch": 1.391412184157495,
+      "grad_norm": 0.4476487636566162,
+      "learning_rate": 6.4806080499010916e-06,
+      "loss": 0.3607,
+      "step": 2992
+    },
+    {
+      "epoch": 1.391877228336692,
+      "grad_norm": 0.4968775808811188,
+      "learning_rate": 6.478023256980835e-06,
+      "loss": 0.3816,
+      "step": 2993
+    },
+    {
+      "epoch": 1.392342272515889,
+      "grad_norm": 0.38123828172683716,
+      "learning_rate": 6.475438031171574e-06,
+      "loss": 0.3263,
+      "step": 2994
+    },
+    {
+      "epoch": 1.392807316695086,
+      "grad_norm": 0.46535784006118774,
+      "learning_rate": 6.472852373230478e-06,
+      "loss": 0.3571,
+      "step": 2995
+    },
+    {
+      "epoch": 1.393272360874283,
+      "grad_norm": 0.6109573841094971,
+      "learning_rate": 6.4702662839148476e-06,
+      "loss": 0.3365,
+      "step": 2996
+    },
+    {
+      "epoch": 1.39373740505348,
+      "grad_norm": 0.41744059324264526,
+      "learning_rate": 6.467679763982103e-06,
+      "loss": 0.3295,
+      "step": 2997
+    },
+    {
+      "epoch": 1.3942024492326772,
+      "grad_norm": 0.3972165286540985,
+      "learning_rate": 6.465092814189795e-06,
+      "loss": 0.3273,
+      "step": 2998
+    },
+    {
+      "epoch": 1.3946674934118741,
+      "grad_norm": 0.5437119007110596,
+      "learning_rate": 6.462505435295601e-06,
+      "loss": 0.3738,
+      "step": 2999
+    },
+    {
+      "epoch": 1.395132537591071,
+      "grad_norm": 0.4269973337650299,
+      "learning_rate": 6.459917628057319e-06,
+      "loss": 0.321,
+      "step": 3000
+    },
+    {
+      "epoch": 1.3955975817702682,
+      "grad_norm": 0.46754154562950134,
+      "learning_rate": 6.457329393232878e-06,
+      "loss": 0.3345,
+      "step": 3001
+    },
+    {
+      "epoch": 1.3960626259494653,
+      "grad_norm": 0.5135738253593445,
+      "learning_rate": 6.454740731580331e-06,
+      "loss": 0.3541,
+      "step": 3002
+    },
+    {
+      "epoch": 1.3965276701286622,
+      "grad_norm": 0.5451841950416565,
+      "learning_rate": 6.452151643857851e-06,
+      "loss": 0.3808,
+      "step": 3003
+    },
+    {
+      "epoch": 1.3969927143078593,
+      "grad_norm": 0.5181593298912048,
+      "learning_rate": 6.4495621308237435e-06,
+      "loss": 0.348,
+      "step": 3004
+    },
+    {
+      "epoch": 1.3974577584870562,
+      "grad_norm": 0.5118134617805481,
+      "learning_rate": 6.446972193236433e-06,
+      "loss": 0.3291,
+      "step": 3005
+    },
+    {
+      "epoch": 1.3979228026662534,
+      "grad_norm": 0.48063331842422485,
+      "learning_rate": 6.444381831854469e-06,
+      "loss": 0.3796,
+      "step": 3006
+    },
+    {
+      "epoch": 1.3983878468454503,
+      "grad_norm": 0.4797370433807373,
+      "learning_rate": 6.44179104743653e-06,
+      "loss": 0.3617,
+      "step": 3007
+    },
+    {
+      "epoch": 1.3988528910246474,
+      "grad_norm": 0.4799365699291229,
+      "learning_rate": 6.439199840741412e-06,
+      "loss": 0.3273,
+      "step": 3008
+    },
+    {
+      "epoch": 1.3993179352038443,
+      "grad_norm": 0.4450487494468689,
+      "learning_rate": 6.436608212528037e-06,
+      "loss": 0.3871,
+      "step": 3009
+    },
+    {
+      "epoch": 1.3997829793830414,
+      "grad_norm": 0.4422202706336975,
+      "learning_rate": 6.434016163555452e-06,
+      "loss": 0.3508,
+      "step": 3010
+    },
+    {
+      "epoch": 1.4002480235622383,
+      "grad_norm": 0.4393402636051178,
+      "learning_rate": 6.431423694582825e-06,
+      "loss": 0.3253,
+      "step": 3011
+    },
+    {
+      "epoch": 1.4007130677414354,
+      "grad_norm": 0.4574717879295349,
+      "learning_rate": 6.428830806369451e-06,
+      "loss": 0.3499,
+      "step": 3012
+    },
+    {
+      "epoch": 1.4011781119206326,
+      "grad_norm": 0.4212042987346649,
+      "learning_rate": 6.42623749967474e-06,
+      "loss": 0.353,
+      "step": 3013
+    },
+    {
+      "epoch": 1.4016431560998295,
+      "grad_norm": 0.3749128580093384,
+      "learning_rate": 6.423643775258232e-06,
+      "loss": 0.3304,
+      "step": 3014
+    },
+    {
+      "epoch": 1.4021082002790264,
+      "grad_norm": 0.4546009302139282,
+      "learning_rate": 6.421049633879588e-06,
+      "loss": 0.3276,
+      "step": 3015
+    },
+    {
+      "epoch": 1.4025732444582235,
+      "grad_norm": 0.4350774586200714,
+      "learning_rate": 6.418455076298587e-06,
+      "loss": 0.3725,
+      "step": 3016
+    },
+    {
+      "epoch": 1.4030382886374206,
+      "grad_norm": 0.4265797436237335,
+      "learning_rate": 6.415860103275136e-06,
+      "loss": 0.3663,
+      "step": 3017
+    },
+    {
+      "epoch": 1.4035033328166175,
+      "grad_norm": 0.3795028030872345,
+      "learning_rate": 6.413264715569259e-06,
+      "loss": 0.3185,
+      "step": 3018
+    },
+    {
+      "epoch": 1.4039683769958147,
+      "grad_norm": 0.39534875750541687,
+      "learning_rate": 6.4106689139411015e-06,
+      "loss": 0.3468,
+      "step": 3019
+    },
+    {
+      "epoch": 1.4044334211750116,
+      "grad_norm": 0.3997578024864197,
+      "learning_rate": 6.408072699150933e-06,
+      "loss": 0.371,
+      "step": 3020
+    },
+    {
+      "epoch": 1.4048984653542087,
+      "grad_norm": 0.3619006872177124,
+      "learning_rate": 6.405476071959142e-06,
+      "loss": 0.3284,
+      "step": 3021
+    },
+    {
+      "epoch": 1.4053635095334056,
+      "grad_norm": 0.41694512963294983,
+      "learning_rate": 6.402879033126239e-06,
+      "loss": 0.3559,
+      "step": 3022
+    },
+    {
+      "epoch": 1.4058285537126027,
+      "grad_norm": 0.41265085339546204,
+      "learning_rate": 6.400281583412855e-06,
+      "loss": 0.3171,
+      "step": 3023
+    },
+    {
+      "epoch": 1.4062935978917996,
+      "grad_norm": 0.5016796588897705,
+      "learning_rate": 6.397683723579741e-06,
+      "loss": 0.3731,
+      "step": 3024
+    },
+    {
+      "epoch": 1.4067586420709968,
+      "grad_norm": 0.39977139234542847,
+      "learning_rate": 6.395085454387766e-06,
+      "loss": 0.3065,
+      "step": 3025
+    },
+    {
+      "epoch": 1.4072236862501937,
+      "grad_norm": 0.41781142354011536,
+      "learning_rate": 6.392486776597921e-06,
+      "loss": 0.3491,
+      "step": 3026
+    },
+    {
+      "epoch": 1.4076887304293908,
+      "grad_norm": 0.4563649594783783,
+      "learning_rate": 6.389887690971319e-06,
+      "loss": 0.3464,
+      "step": 3027
+    },
+    {
+      "epoch": 1.408153774608588,
+      "grad_norm": 0.4177623987197876,
+      "learning_rate": 6.387288198269189e-06,
+      "loss": 0.3782,
+      "step": 3028
+    },
+    {
+      "epoch": 1.4086188187877848,
+      "grad_norm": 0.3898913264274597,
+      "learning_rate": 6.384688299252879e-06,
+      "loss": 0.3564,
+      "step": 3029
+    },
+    {
+      "epoch": 1.4090838629669817,
+      "grad_norm": 0.3614499568939209,
+      "learning_rate": 6.3820879946838585e-06,
+      "loss": 0.3215,
+      "step": 3030
+    },
+    {
+      "epoch": 1.4095489071461789,
+      "grad_norm": 0.45202839374542236,
+      "learning_rate": 6.379487285323713e-06,
+      "loss": 0.3829,
+      "step": 3031
+    },
+    {
+      "epoch": 1.410013951325376,
+      "grad_norm": 0.39168378710746765,
+      "learning_rate": 6.3768861719341475e-06,
+      "loss": 0.3619,
+      "step": 3032
+    },
+    {
+      "epoch": 1.410478995504573,
+      "grad_norm": 0.3808189630508423,
+      "learning_rate": 6.374284655276989e-06,
+      "loss": 0.3376,
+      "step": 3033
+    },
+    {
+      "epoch": 1.41094403968377,
+      "grad_norm": 0.43050652742385864,
+      "learning_rate": 6.371682736114178e-06,
+      "loss": 0.3206,
+      "step": 3034
+    },
+    {
+      "epoch": 1.411409083862967,
+      "grad_norm": 0.4489372968673706,
+      "learning_rate": 6.369080415207773e-06,
+      "loss": 0.3917,
+      "step": 3035
+    },
+    {
+      "epoch": 1.411874128042164,
+      "grad_norm": 0.37065914273262024,
+      "learning_rate": 6.366477693319953e-06,
+      "loss": 0.3446,
+      "step": 3036
+    },
+    {
+      "epoch": 1.412339172221361,
+      "grad_norm": 0.3754143714904785,
+      "learning_rate": 6.363874571213013e-06,
+      "loss": 0.356,
+      "step": 3037
+    },
+    {
+      "epoch": 1.412804216400558,
+      "grad_norm": 0.3956548273563385,
+      "learning_rate": 6.361271049649363e-06,
+      "loss": 0.3558,
+      "step": 3038
+    },
+    {
+      "epoch": 1.4132692605797552,
+      "grad_norm": 0.4007726013660431,
+      "learning_rate": 6.358667129391536e-06,
+      "loss": 0.3204,
+      "step": 3039
+    },
+    {
+      "epoch": 1.4137343047589521,
+      "grad_norm": 0.42144933342933655,
+      "learning_rate": 6.356062811202175e-06,
+      "loss": 0.3674,
+      "step": 3040
+    },
+    {
+      "epoch": 1.414199348938149,
+      "grad_norm": 0.4103334844112396,
+      "learning_rate": 6.3534580958440425e-06,
+      "loss": 0.3627,
+      "step": 3041
+    },
+    {
+      "epoch": 1.4146643931173462,
+      "grad_norm": 0.34175369143486023,
+      "learning_rate": 6.35085298408002e-06,
+      "loss": 0.3123,
+      "step": 3042
+    },
+    {
+      "epoch": 1.4151294372965433,
+      "grad_norm": 0.38748347759246826,
+      "learning_rate": 6.348247476673099e-06,
+      "loss": 0.3575,
+      "step": 3043
+    },
+    {
+      "epoch": 1.4155944814757402,
+      "grad_norm": 0.413284569978714,
+      "learning_rate": 6.345641574386393e-06,
+      "loss": 0.3341,
+      "step": 3044
+    },
+    {
+      "epoch": 1.416059525654937,
+      "grad_norm": 0.37702399492263794,
+      "learning_rate": 6.3430352779831275e-06,
+      "loss": 0.3339,
+      "step": 3045
+    },
+    {
+      "epoch": 1.4165245698341342,
+      "grad_norm": 0.4544568955898285,
+      "learning_rate": 6.340428588226643e-06,
+      "loss": 0.3564,
+      "step": 3046
+    },
+    {
+      "epoch": 1.4169896140133313,
+      "grad_norm": 0.3882623016834259,
+      "learning_rate": 6.337821505880399e-06,
+      "loss": 0.3699,
+      "step": 3047
+    },
+    {
+      "epoch": 1.4174546581925282,
+      "grad_norm": 0.3817591369152069,
+      "learning_rate": 6.335214031707966e-06,
+      "loss": 0.3165,
+      "step": 3048
+    },
+    {
+      "epoch": 1.4179197023717254,
+      "grad_norm": 0.39398717880249023,
+      "learning_rate": 6.33260616647303e-06,
+      "loss": 0.3486,
+      "step": 3049
+    },
+    {
+      "epoch": 1.4183847465509223,
+      "grad_norm": 0.4238983988761902,
+      "learning_rate": 6.329997910939394e-06,
+      "loss": 0.3892,
+      "step": 3050
+    },
+    {
+      "epoch": 1.4188497907301194,
+      "grad_norm": 0.43442922830581665,
+      "learning_rate": 6.327389265870974e-06,
+      "loss": 0.3235,
+      "step": 3051
+    },
+    {
+      "epoch": 1.4193148349093163,
+      "grad_norm": 0.4579649865627289,
+      "learning_rate": 6.324780232031799e-06,
+      "loss": 0.3464,
+      "step": 3052
+    },
+    {
+      "epoch": 1.4197798790885134,
+      "grad_norm": 0.3303820490837097,
+      "learning_rate": 6.322170810186013e-06,
+      "loss": 0.285,
+      "step": 3053
+    },
+    {
+      "epoch": 1.4202449232677106,
+      "grad_norm": 0.4255482256412506,
+      "learning_rate": 6.319561001097871e-06,
+      "loss": 0.3688,
+      "step": 3054
+    },
+    {
+      "epoch": 1.4207099674469075,
+      "grad_norm": 0.46313193440437317,
+      "learning_rate": 6.316950805531746e-06,
+      "loss": 0.3576,
+      "step": 3055
+    },
+    {
+      "epoch": 1.4211750116261044,
+      "grad_norm": 0.412056565284729,
+      "learning_rate": 6.314340224252124e-06,
+      "loss": 0.3525,
+      "step": 3056
+    },
+    {
+      "epoch": 1.4216400558053015,
+      "grad_norm": 0.4473661184310913,
+      "learning_rate": 6.311729258023597e-06,
+      "loss": 0.3644,
+      "step": 3057
+    },
+    {
+      "epoch": 1.4221050999844986,
+      "grad_norm": 0.3638128638267517,
+      "learning_rate": 6.309117907610878e-06,
+      "loss": 0.3013,
+      "step": 3058
+    },
+    {
+      "epoch": 1.4225701441636955,
+      "grad_norm": 0.471097856760025,
+      "learning_rate": 6.306506173778788e-06,
+      "loss": 0.3941,
+      "step": 3059
+    },
+    {
+      "epoch": 1.4230351883428924,
+      "grad_norm": 0.42378881573677063,
+      "learning_rate": 6.303894057292261e-06,
+      "loss": 0.3553,
+      "step": 3060
+    },
+    {
+      "epoch": 1.4235002325220896,
+      "grad_norm": 0.44823458790779114,
+      "learning_rate": 6.3012815589163435e-06,
+      "loss": 0.3624,
+      "step": 3061
+    },
+    {
+      "epoch": 1.4239652767012867,
+      "grad_norm": 0.39349788427352905,
+      "learning_rate": 6.2986686794161955e-06,
+      "loss": 0.335,
+      "step": 3062
+    },
+    {
+      "epoch": 1.4244303208804836,
+      "grad_norm": 0.39323508739471436,
+      "learning_rate": 6.296055419557086e-06,
+      "loss": 0.3689,
+      "step": 3063
+    },
+    {
+      "epoch": 1.4248953650596807,
+      "grad_norm": 0.39461150765419006,
+      "learning_rate": 6.293441780104394e-06,
+      "loss": 0.351,
+      "step": 3064
+    },
+    {
+      "epoch": 1.4253604092388776,
+      "grad_norm": 0.3934915065765381,
+      "learning_rate": 6.290827761823617e-06,
+      "loss": 0.3515,
+      "step": 3065
+    },
+    {
+      "epoch": 1.4258254534180748,
+      "grad_norm": 0.407262921333313,
+      "learning_rate": 6.2882133654803535e-06,
+      "loss": 0.3113,
+      "step": 3066
+    },
+    {
+      "epoch": 1.4262904975972717,
+      "grad_norm": 0.4266803562641144,
+      "learning_rate": 6.28559859184032e-06,
+      "loss": 0.3751,
+      "step": 3067
+    },
+    {
+      "epoch": 1.4267555417764688,
+      "grad_norm": 0.39150920510292053,
+      "learning_rate": 6.282983441669343e-06,
+      "loss": 0.3722,
+      "step": 3068
+    },
+    {
+      "epoch": 1.427220585955666,
+      "grad_norm": 0.42089784145355225,
+      "learning_rate": 6.280367915733354e-06,
+      "loss": 0.3574,
+      "step": 3069
+    },
+    {
+      "epoch": 1.4276856301348628,
+      "grad_norm": 0.3548668324947357,
+      "learning_rate": 6.277752014798401e-06,
+      "loss": 0.3241,
+      "step": 3070
+    },
+    {
+      "epoch": 1.4281506743140597,
+      "grad_norm": 0.4519556164741516,
+      "learning_rate": 6.275135739630636e-06,
+      "loss": 0.3671,
+      "step": 3071
+    },
+    {
+      "epoch": 1.4286157184932569,
+      "grad_norm": 0.4052821695804596,
+      "learning_rate": 6.272519090996326e-06,
+      "loss": 0.3497,
+      "step": 3072
+    },
+    {
+      "epoch": 1.429080762672454,
+      "grad_norm": 0.4167952835559845,
+      "learning_rate": 6.269902069661843e-06,
+      "loss": 0.3615,
+      "step": 3073
+    },
+    {
+      "epoch": 1.429545806851651,
+      "grad_norm": 0.4397052526473999,
+      "learning_rate": 6.267284676393672e-06,
+      "loss": 0.3694,
+      "step": 3074
+    },
+    {
+      "epoch": 1.4300108510308478,
+      "grad_norm": 0.42949673533439636,
+      "learning_rate": 6.264666911958404e-06,
+      "loss": 0.329,
+      "step": 3075
+    },
+    {
+      "epoch": 1.430475895210045,
+      "grad_norm": 0.39663806557655334,
+      "learning_rate": 6.26204877712274e-06,
+      "loss": 0.3553,
+      "step": 3076
+    },
+    {
+      "epoch": 1.430940939389242,
+      "grad_norm": 0.3915885388851166,
+      "learning_rate": 6.259430272653489e-06,
+      "loss": 0.349,
+      "step": 3077
+    },
+    {
+      "epoch": 1.431405983568439,
+      "grad_norm": 0.4367777705192566,
+      "learning_rate": 6.256811399317567e-06,
+      "loss": 0.343,
+      "step": 3078
+    },
+    {
+      "epoch": 1.431871027747636,
+      "grad_norm": 0.4142322242259979,
+      "learning_rate": 6.254192157882002e-06,
+      "loss": 0.3547,
+      "step": 3079
+    },
+    {
+      "epoch": 1.432336071926833,
+      "grad_norm": 0.41435766220092773,
+      "learning_rate": 6.251572549113925e-06,
+      "loss": 0.3363,
+      "step": 3080
+    },
+    {
+      "epoch": 1.4328011161060301,
+      "grad_norm": 0.3753543496131897,
+      "learning_rate": 6.248952573780578e-06,
+      "loss": 0.3099,
+      "step": 3081
+    },
+    {
+      "epoch": 1.433266160285227,
+      "grad_norm": 0.4042842388153076,
+      "learning_rate": 6.246332232649309e-06,
+      "loss": 0.3347,
+      "step": 3082
+    },
+    {
+      "epoch": 1.4337312044644241,
+      "grad_norm": 0.48114094138145447,
+      "learning_rate": 6.243711526487575e-06,
+      "loss": 0.4166,
+      "step": 3083
+    },
+    {
+      "epoch": 1.4341962486436213,
+      "grad_norm": 0.36960569024086,
+      "learning_rate": 6.241090456062934e-06,
+      "loss": 0.3309,
+      "step": 3084
+    },
+    {
+      "epoch": 1.4346612928228182,
+      "grad_norm": 0.4094170928001404,
+      "learning_rate": 6.238469022143059e-06,
+      "loss": 0.3325,
+      "step": 3085
+    },
+    {
+      "epoch": 1.435126337002015,
+      "grad_norm": 0.418476402759552,
+      "learning_rate": 6.235847225495724e-06,
+      "loss": 0.3791,
+      "step": 3086
+    },
+    {
+      "epoch": 1.4355913811812122,
+      "grad_norm": 0.3743997812271118,
+      "learning_rate": 6.23322506688881e-06,
+      "loss": 0.3432,
+      "step": 3087
+    },
+    {
+      "epoch": 1.4360564253604093,
+      "grad_norm": 0.44286203384399414,
+      "learning_rate": 6.230602547090307e-06,
+      "loss": 0.3745,
+      "step": 3088
+    },
+    {
+      "epoch": 1.4365214695396062,
+      "grad_norm": 0.40265190601348877,
+      "learning_rate": 6.227979666868307e-06,
+      "loss": 0.3441,
+      "step": 3089
+    },
+    {
+      "epoch": 1.4369865137188031,
+      "grad_norm": 0.4036969244480133,
+      "learning_rate": 6.225356426991007e-06,
+      "loss": 0.3609,
+      "step": 3090
+    },
+    {
+      "epoch": 1.4374515578980003,
+      "grad_norm": 0.3730718493461609,
+      "learning_rate": 6.222732828226714e-06,
+      "loss": 0.3337,
+      "step": 3091
+    },
+    {
+      "epoch": 1.4379166020771974,
+      "grad_norm": 0.4534870684146881,
+      "learning_rate": 6.2201088713438366e-06,
+      "loss": 0.369,
+      "step": 3092
+    },
+    {
+      "epoch": 1.4383816462563943,
+      "grad_norm": 0.4100678563117981,
+      "learning_rate": 6.2174845571108884e-06,
+      "loss": 0.3617,
+      "step": 3093
+    },
+    {
+      "epoch": 1.4388466904355914,
+      "grad_norm": 0.458183616399765,
+      "learning_rate": 6.214859886296491e-06,
+      "loss": 0.3556,
+      "step": 3094
+    },
+    {
+      "epoch": 1.4393117346147883,
+      "grad_norm": 0.4203883111476898,
+      "learning_rate": 6.212234859669366e-06,
+      "loss": 0.3503,
+      "step": 3095
+    },
+    {
+      "epoch": 1.4397767787939855,
+      "grad_norm": 0.43506595492362976,
+      "learning_rate": 6.209609477998339e-06,
+      "loss": 0.3604,
+      "step": 3096
+    },
+    {
+      "epoch": 1.4402418229731824,
+      "grad_norm": 0.37604156136512756,
+      "learning_rate": 6.206983742052345e-06,
+      "loss": 0.317,
+      "step": 3097
+    },
+    {
+      "epoch": 1.4407068671523795,
+      "grad_norm": 0.3998353183269501,
+      "learning_rate": 6.204357652600419e-06,
+      "loss": 0.3476,
+      "step": 3098
+    },
+    {
+      "epoch": 1.4411719113315766,
+      "grad_norm": 0.4210034906864166,
+      "learning_rate": 6.201731210411698e-06,
+      "loss": 0.368,
+      "step": 3099
+    },
+    {
+      "epoch": 1.4416369555107735,
+      "grad_norm": 0.44328659772872925,
+      "learning_rate": 6.199104416255426e-06,
+      "loss": 0.3929,
+      "step": 3100
+    },
+    {
+      "epoch": 1.4421019996899704,
+      "grad_norm": 0.44677838683128357,
+      "learning_rate": 6.196477270900947e-06,
+      "loss": 0.3404,
+      "step": 3101
+    },
+    {
+      "epoch": 1.4425670438691676,
+      "grad_norm": 0.42385387420654297,
+      "learning_rate": 6.193849775117709e-06,
+      "loss": 0.3779,
+      "step": 3102
+    },
+    {
+      "epoch": 1.4430320880483647,
+      "grad_norm": 0.4066379964351654,
+      "learning_rate": 6.191221929675266e-06,
+      "loss": 0.3221,
+      "step": 3103
+    },
+    {
+      "epoch": 1.4434971322275616,
+      "grad_norm": 0.40358468890190125,
+      "learning_rate": 6.188593735343269e-06,
+      "loss": 0.332,
+      "step": 3104
+    },
+    {
+      "epoch": 1.4439621764067587,
+      "grad_norm": 0.5439780950546265,
+      "learning_rate": 6.185965192891472e-06,
+      "loss": 0.3834,
+      "step": 3105
+    },
+    {
+      "epoch": 1.4444272205859556,
+      "grad_norm": 0.4213503897190094,
+      "learning_rate": 6.183336303089735e-06,
+      "loss": 0.3605,
+      "step": 3106
+    },
+    {
+      "epoch": 1.4448922647651528,
+      "grad_norm": 0.39889928698539734,
+      "learning_rate": 6.1807070667080145e-06,
+      "loss": 0.3439,
+      "step": 3107
+    },
+    {
+      "epoch": 1.4453573089443497,
+      "grad_norm": 0.5045921802520752,
+      "learning_rate": 6.1780774845163736e-06,
+      "loss": 0.3687,
+      "step": 3108
+    },
+    {
+      "epoch": 1.4458223531235468,
+      "grad_norm": 0.4484003186225891,
+      "learning_rate": 6.175447557284972e-06,
+      "loss": 0.349,
+      "step": 3109
+    },
+    {
+      "epoch": 1.4462873973027437,
+      "grad_norm": 0.4632101058959961,
+      "learning_rate": 6.172817285784076e-06,
+      "loss": 0.3897,
+      "step": 3110
+    },
+    {
+      "epoch": 1.4467524414819408,
+      "grad_norm": 0.43179023265838623,
+      "learning_rate": 6.170186670784047e-06,
+      "loss": 0.3275,
+      "step": 3111
+    },
+    {
+      "epoch": 1.4472174856611377,
+      "grad_norm": 0.48543015122413635,
+      "learning_rate": 6.1675557130553475e-06,
+      "loss": 0.3395,
+      "step": 3112
+    },
+    {
+      "epoch": 1.4476825298403349,
+      "grad_norm": 0.4830179214477539,
+      "learning_rate": 6.164924413368546e-06,
+      "loss": 0.3724,
+      "step": 3113
+    },
+    {
+      "epoch": 1.448147574019532,
+      "grad_norm": 0.4735562801361084,
+      "learning_rate": 6.162292772494305e-06,
+      "loss": 0.362,
+      "step": 3114
+    },
+    {
+      "epoch": 1.4486126181987289,
+      "grad_norm": 0.43904268741607666,
+      "learning_rate": 6.159660791203392e-06,
+      "loss": 0.3283,
+      "step": 3115
+    },
+    {
+      "epoch": 1.4490776623779258,
+      "grad_norm": 0.4425245523452759,
+      "learning_rate": 6.157028470266669e-06,
+      "loss": 0.3637,
+      "step": 3116
+    },
+    {
+      "epoch": 1.449542706557123,
+      "grad_norm": 0.40749219059944153,
+      "learning_rate": 6.1543958104551e-06,
+      "loss": 0.3578,
+      "step": 3117
+    },
+    {
+      "epoch": 1.45000775073632,
+      "grad_norm": 0.4425008296966553,
+      "learning_rate": 6.15176281253975e-06,
+      "loss": 0.3405,
+      "step": 3118
+    },
+    {
+      "epoch": 1.450472794915517,
+      "grad_norm": 0.38328611850738525,
+      "learning_rate": 6.1491294772917785e-06,
+      "loss": 0.3225,
+      "step": 3119
+    },
+    {
+      "epoch": 1.450937839094714,
+      "grad_norm": 0.47361764311790466,
+      "learning_rate": 6.146495805482451e-06,
+      "loss": 0.3796,
+      "step": 3120
+    },
+    {
+      "epoch": 1.451402883273911,
+      "grad_norm": 0.41950133442878723,
+      "learning_rate": 6.143861797883124e-06,
+      "loss": 0.355,
+      "step": 3121
+    },
+    {
+      "epoch": 1.451867927453108,
+      "grad_norm": 0.46674826741218567,
+      "learning_rate": 6.141227455265256e-06,
+      "loss": 0.37,
+      "step": 3122
+    },
+    {
+      "epoch": 1.452332971632305,
+      "grad_norm": 0.46177905797958374,
+      "learning_rate": 6.138592778400404e-06,
+      "loss": 0.3742,
+      "step": 3123
+    },
+    {
+      "epoch": 1.4527980158115021,
+      "grad_norm": 0.4034256637096405,
+      "learning_rate": 6.135957768060221e-06,
+      "loss": 0.3395,
+      "step": 3124
+    },
+    {
+      "epoch": 1.453263059990699,
+      "grad_norm": 0.47683438658714294,
+      "learning_rate": 6.133322425016459e-06,
+      "loss": 0.3755,
+      "step": 3125
+    },
+    {
+      "epoch": 1.4537281041698962,
+      "grad_norm": 0.46114102005958557,
+      "learning_rate": 6.1306867500409685e-06,
+      "loss": 0.3385,
+      "step": 3126
+    },
+    {
+      "epoch": 1.454193148349093,
+      "grad_norm": 0.40554434061050415,
+      "learning_rate": 6.128050743905695e-06,
+      "loss": 0.3636,
+      "step": 3127
+    },
+    {
+      "epoch": 1.4546581925282902,
+      "grad_norm": 0.4275938868522644,
+      "learning_rate": 6.12541440738268e-06,
+      "loss": 0.3578,
+      "step": 3128
+    },
+    {
+      "epoch": 1.4551232367074873,
+      "grad_norm": 0.42898625135421753,
+      "learning_rate": 6.122777741244067e-06,
+      "loss": 0.3635,
+      "step": 3129
+    },
+    {
+      "epoch": 1.4555882808866842,
+      "grad_norm": 0.4090706706047058,
+      "learning_rate": 6.120140746262091e-06,
+      "loss": 0.329,
+      "step": 3130
+    },
+    {
+      "epoch": 1.4560533250658811,
+      "grad_norm": 0.45951342582702637,
+      "learning_rate": 6.117503423209084e-06,
+      "loss": 0.3656,
+      "step": 3131
+    },
+    {
+      "epoch": 1.4565183692450783,
+      "grad_norm": 0.3707917332649231,
+      "learning_rate": 6.1148657728574765e-06,
+      "loss": 0.3696,
+      "step": 3132
+    },
+    {
+      "epoch": 1.4569834134242754,
+      "grad_norm": 0.4463897943496704,
+      "learning_rate": 6.1122277959797925e-06,
+      "loss": 0.335,
+      "step": 3133
+    },
+    {
+      "epoch": 1.4574484576034723,
+      "grad_norm": 0.48531046509742737,
+      "learning_rate": 6.109589493348655e-06,
+      "loss": 0.3419,
+      "step": 3134
+    },
+    {
+      "epoch": 1.4579135017826694,
+      "grad_norm": 0.3899511396884918,
+      "learning_rate": 6.106950865736777e-06,
+      "loss": 0.3556,
+      "step": 3135
+    },
+    {
+      "epoch": 1.4583785459618663,
+      "grad_norm": 0.3700244426727295,
+      "learning_rate": 6.10431191391697e-06,
+      "loss": 0.3633,
+      "step": 3136
+    },
+    {
+      "epoch": 1.4588435901410635,
+      "grad_norm": 0.43946826457977295,
+      "learning_rate": 6.101672638662141e-06,
+      "loss": 0.3294,
+      "step": 3137
+    },
+    {
+      "epoch": 1.4593086343202604,
+      "grad_norm": 0.407183438539505,
+      "learning_rate": 6.099033040745292e-06,
+      "loss": 0.3605,
+      "step": 3138
+    },
+    {
+      "epoch": 1.4597736784994575,
+      "grad_norm": 0.39797452092170715,
+      "learning_rate": 6.0963931209395165e-06,
+      "loss": 0.3581,
+      "step": 3139
+    },
+    {
+      "epoch": 1.4602387226786544,
+      "grad_norm": 0.429902583360672,
+      "learning_rate": 6.0937528800180056e-06,
+      "loss": 0.3423,
+      "step": 3140
+    },
+    {
+      "epoch": 1.4607037668578515,
+      "grad_norm": 0.43174222111701965,
+      "learning_rate": 6.0911123187540414e-06,
+      "loss": 0.3604,
+      "step": 3141
+    },
+    {
+      "epoch": 1.4611688110370484,
+      "grad_norm": 0.4118611514568329,
+      "learning_rate": 6.088471437921002e-06,
+      "loss": 0.3742,
+      "step": 3142
+    },
+    {
+      "epoch": 1.4616338552162456,
+      "grad_norm": 0.37793856859207153,
+      "learning_rate": 6.0858302382923585e-06,
+      "loss": 0.3462,
+      "step": 3143
+    },
+    {
+      "epoch": 1.4620988993954427,
+      "grad_norm": 0.4251405894756317,
+      "learning_rate": 6.083188720641676e-06,
+      "loss": 0.3283,
+      "step": 3144
+    },
+    {
+      "epoch": 1.4625639435746396,
+      "grad_norm": 0.3901570439338684,
+      "learning_rate": 6.080546885742611e-06,
+      "loss": 0.3649,
+      "step": 3145
+    },
+    {
+      "epoch": 1.4630289877538365,
+      "grad_norm": 0.39368346333503723,
+      "learning_rate": 6.077904734368915e-06,
+      "loss": 0.3652,
+      "step": 3146
+    },
+    {
+      "epoch": 1.4634940319330336,
+      "grad_norm": 0.4399242103099823,
+      "learning_rate": 6.075262267294432e-06,
+      "loss": 0.3585,
+      "step": 3147
+    },
+    {
+      "epoch": 1.4639590761122308,
+      "grad_norm": 0.37305113673210144,
+      "learning_rate": 6.072619485293095e-06,
+      "loss": 0.339,
+      "step": 3148
+    },
+    {
+      "epoch": 1.4644241202914277,
+      "grad_norm": 0.4052191376686096,
+      "learning_rate": 6.069976389138934e-06,
+      "loss": 0.3698,
+      "step": 3149
+    },
+    {
+      "epoch": 1.4648891644706248,
+      "grad_norm": 0.4240652620792389,
+      "learning_rate": 6.067332979606069e-06,
+      "loss": 0.3417,
+      "step": 3150
+    },
+    {
+      "epoch": 1.4653542086498217,
+      "grad_norm": 0.4290774464607239,
+      "learning_rate": 6.064689257468711e-06,
+      "loss": 0.3381,
+      "step": 3151
+    },
+    {
+      "epoch": 1.4658192528290188,
+      "grad_norm": 0.3692978024482727,
+      "learning_rate": 6.062045223501163e-06,
+      "loss": 0.3469,
+      "step": 3152
+    },
+    {
+      "epoch": 1.4662842970082157,
+      "grad_norm": 0.37645047903060913,
+      "learning_rate": 6.0594008784778206e-06,
+      "loss": 0.3889,
+      "step": 3153
+    },
+    {
+      "epoch": 1.4667493411874128,
+      "grad_norm": 0.4383949637413025,
+      "learning_rate": 6.056756223173167e-06,
+      "loss": 0.3384,
+      "step": 3154
+    },
+    {
+      "epoch": 1.4672143853666098,
+      "grad_norm": 0.4462020695209503,
+      "learning_rate": 6.054111258361782e-06,
+      "loss": 0.3641,
+      "step": 3155
+    },
+    {
+      "epoch": 1.4676794295458069,
+      "grad_norm": 0.38466379046440125,
+      "learning_rate": 6.051465984818332e-06,
+      "loss": 0.3538,
+      "step": 3156
+    },
+    {
+      "epoch": 1.4681444737250038,
+      "grad_norm": 0.42687302827835083,
+      "learning_rate": 6.048820403317575e-06,
+      "loss": 0.3286,
+      "step": 3157
+    },
+    {
+      "epoch": 1.468609517904201,
+      "grad_norm": 0.4158511161804199,
+      "learning_rate": 6.046174514634355e-06,
+      "loss": 0.3368,
+      "step": 3158
+    },
+    {
+      "epoch": 1.469074562083398,
+      "grad_norm": 0.409820020198822,
+      "learning_rate": 6.043528319543615e-06,
+      "loss": 0.3446,
+      "step": 3159
+    },
+    {
+      "epoch": 1.469539606262595,
+      "grad_norm": 0.4203595817089081,
+      "learning_rate": 6.04088181882038e-06,
+      "loss": 0.3366,
+      "step": 3160
+    },
+    {
+      "epoch": 1.4700046504417918,
+      "grad_norm": 0.4121120274066925,
+      "learning_rate": 6.038235013239767e-06,
+      "loss": 0.3271,
+      "step": 3161
+    },
+    {
+      "epoch": 1.470469694620989,
+      "grad_norm": 0.40760576725006104,
+      "learning_rate": 6.035587903576984e-06,
+      "loss": 0.3536,
+      "step": 3162
+    },
+    {
+      "epoch": 1.470934738800186,
+      "grad_norm": 0.4346151649951935,
+      "learning_rate": 6.032940490607324e-06,
+      "loss": 0.3356,
+      "step": 3163
+    },
+    {
+      "epoch": 1.471399782979383,
+      "grad_norm": 0.459246426820755,
+      "learning_rate": 6.030292775106173e-06,
+      "loss": 0.3679,
+      "step": 3164
+    },
+    {
+      "epoch": 1.4718648271585801,
+      "grad_norm": 0.42355459928512573,
+      "learning_rate": 6.027644757849004e-06,
+      "loss": 0.386,
+      "step": 3165
+    },
+    {
+      "epoch": 1.472329871337777,
+      "grad_norm": 0.4206094741821289,
+      "learning_rate": 6.024996439611376e-06,
+      "loss": 0.3344,
+      "step": 3166
+    },
+    {
+      "epoch": 1.4727949155169742,
+      "grad_norm": 0.36260485649108887,
+      "learning_rate": 6.022347821168941e-06,
+      "loss": 0.3362,
+      "step": 3167
+    },
+    {
+      "epoch": 1.473259959696171,
+      "grad_norm": 0.4596819281578064,
+      "learning_rate": 6.0196989032974366e-06,
+      "loss": 0.4029,
+      "step": 3168
+    },
+    {
+      "epoch": 1.4737250038753682,
+      "grad_norm": 0.4887857139110565,
+      "learning_rate": 6.017049686772685e-06,
+      "loss": 0.335,
+      "step": 3169
+    },
+    {
+      "epoch": 1.474190048054565,
+      "grad_norm": 0.4173000156879425,
+      "learning_rate": 6.0144001723706e-06,
+      "loss": 0.329,
+      "step": 3170
+    },
+    {
+      "epoch": 1.4746550922337622,
+      "grad_norm": 0.419015496969223,
+      "learning_rate": 6.011750360867183e-06,
+      "loss": 0.3812,
+      "step": 3171
+    },
+    {
+      "epoch": 1.4751201364129591,
+      "grad_norm": 0.5123944282531738,
+      "learning_rate": 6.009100253038518e-06,
+      "loss": 0.3391,
+      "step": 3172
+    },
+    {
+      "epoch": 1.4755851805921563,
+      "grad_norm": 0.44715067744255066,
+      "learning_rate": 6.00644984966078e-06,
+      "loss": 0.3153,
+      "step": 3173
+    },
+    {
+      "epoch": 1.4760502247713534,
+      "grad_norm": 0.4377215802669525,
+      "learning_rate": 6.003799151510229e-06,
+      "loss": 0.3815,
+      "step": 3174
+    },
+    {
+      "epoch": 1.4765152689505503,
+      "grad_norm": 0.5287955403327942,
+      "learning_rate": 6.001148159363213e-06,
+      "loss": 0.362,
+      "step": 3175
+    },
+    {
+      "epoch": 1.4769803131297472,
+      "grad_norm": 0.3864191770553589,
+      "learning_rate": 5.998496873996161e-06,
+      "loss": 0.3175,
+      "step": 3176
+    },
+    {
+      "epoch": 1.4774453573089443,
+      "grad_norm": 0.45076581835746765,
+      "learning_rate": 5.995845296185594e-06,
+      "loss": 0.3572,
+      "step": 3177
+    },
+    {
+      "epoch": 1.4779104014881415,
+      "grad_norm": 0.4349561929702759,
+      "learning_rate": 5.993193426708115e-06,
+      "loss": 0.3452,
+      "step": 3178
+    },
+    {
+      "epoch": 1.4783754456673384,
+      "grad_norm": 0.45883357524871826,
+      "learning_rate": 5.990541266340414e-06,
+      "loss": 0.3639,
+      "step": 3179
+    },
+    {
+      "epoch": 1.4788404898465355,
+      "grad_norm": 0.4192543923854828,
+      "learning_rate": 5.987888815859266e-06,
+      "loss": 0.3199,
+      "step": 3180
+    },
+    {
+      "epoch": 1.4793055340257324,
+      "grad_norm": 0.4741514325141907,
+      "learning_rate": 5.985236076041531e-06,
+      "loss": 0.3624,
+      "step": 3181
+    },
+    {
+      "epoch": 1.4797705782049295,
+      "grad_norm": 0.4844418168067932,
+      "learning_rate": 5.982583047664151e-06,
+      "loss": 0.3577,
+      "step": 3182
+    },
+    {
+      "epoch": 1.4802356223841264,
+      "grad_norm": 0.4832378923892975,
+      "learning_rate": 5.979929731504158e-06,
+      "loss": 0.3504,
+      "step": 3183
+    },
+    {
+      "epoch": 1.4807006665633236,
+      "grad_norm": 0.42179495096206665,
+      "learning_rate": 5.9772761283386626e-06,
+      "loss": 0.3384,
+      "step": 3184
+    },
+    {
+      "epoch": 1.4811657107425205,
+      "grad_norm": 0.4342862665653229,
+      "learning_rate": 5.9746222389448635e-06,
+      "loss": 0.3624,
+      "step": 3185
+    },
+    {
+      "epoch": 1.4816307549217176,
+      "grad_norm": 0.47698426246643066,
+      "learning_rate": 5.971968064100042e-06,
+      "loss": 0.3483,
+      "step": 3186
+    },
+    {
+      "epoch": 1.4820957991009145,
+      "grad_norm": 0.4859686493873596,
+      "learning_rate": 5.969313604581564e-06,
+      "loss": 0.3691,
+      "step": 3187
+    },
+    {
+      "epoch": 1.4825608432801116,
+      "grad_norm": 0.4531085789203644,
+      "learning_rate": 5.966658861166874e-06,
+      "loss": 0.3741,
+      "step": 3188
+    },
+    {
+      "epoch": 1.4830258874593087,
+      "grad_norm": 0.46530213952064514,
+      "learning_rate": 5.9640038346335045e-06,
+      "loss": 0.3501,
+      "step": 3189
+    },
+    {
+      "epoch": 1.4834909316385056,
+      "grad_norm": 0.4475176930427551,
+      "learning_rate": 5.961348525759072e-06,
+      "loss": 0.3515,
+      "step": 3190
+    },
+    {
+      "epoch": 1.4839559758177026,
+      "grad_norm": 0.4891159236431122,
+      "learning_rate": 5.958692935321271e-06,
+      "loss": 0.3485,
+      "step": 3191
+    },
+    {
+      "epoch": 1.4844210199968997,
+      "grad_norm": 0.4750014841556549,
+      "learning_rate": 5.956037064097881e-06,
+      "loss": 0.3773,
+      "step": 3192
+    },
+    {
+      "epoch": 1.4848860641760968,
+      "grad_norm": 0.37893229722976685,
+      "learning_rate": 5.953380912866764e-06,
+      "loss": 0.372,
+      "step": 3193
+    },
+    {
+      "epoch": 1.4853511083552937,
+      "grad_norm": 0.41365838050842285,
+      "learning_rate": 5.9507244824058644e-06,
+      "loss": 0.2967,
+      "step": 3194
+    },
+    {
+      "epoch": 1.4858161525344908,
+      "grad_norm": 0.5796143412590027,
+      "learning_rate": 5.948067773493205e-06,
+      "loss": 0.4209,
+      "step": 3195
+    },
+    {
+      "epoch": 1.4862811967136877,
+      "grad_norm": 0.388439804315567,
+      "learning_rate": 5.945410786906896e-06,
+      "loss": 0.3745,
+      "step": 3196
+    },
+    {
+      "epoch": 1.4867462408928849,
+      "grad_norm": 0.3897601068019867,
+      "learning_rate": 5.9427535234251235e-06,
+      "loss": 0.3073,
+      "step": 3197
+    },
+    {
+      "epoch": 1.4872112850720818,
+      "grad_norm": 0.5110495090484619,
+      "learning_rate": 5.940095983826157e-06,
+      "loss": 0.3641,
+      "step": 3198
+    },
+    {
+      "epoch": 1.487676329251279,
+      "grad_norm": 0.46713900566101074,
+      "learning_rate": 5.9374381688883475e-06,
+      "loss": 0.3414,
+      "step": 3199
+    },
+    {
+      "epoch": 1.488141373430476,
+      "grad_norm": 0.4882988929748535,
+      "learning_rate": 5.9347800793901245e-06,
+      "loss": 0.3823,
+      "step": 3200
+    },
+    {
+      "epoch": 1.488606417609673,
+      "grad_norm": 0.45495909452438354,
+      "learning_rate": 5.93212171611e-06,
+      "loss": 0.3497,
+      "step": 3201
+    },
+    {
+      "epoch": 1.4890714617888698,
+      "grad_norm": 0.38749927282333374,
+      "learning_rate": 5.929463079826565e-06,
+      "loss": 0.3585,
+      "step": 3202
+    },
+    {
+      "epoch": 1.489536505968067,
+      "grad_norm": 0.4609941244125366,
+      "learning_rate": 5.9268041713184934e-06,
+      "loss": 0.3501,
+      "step": 3203
+    },
+    {
+      "epoch": 1.490001550147264,
+      "grad_norm": 0.38551583886146545,
+      "learning_rate": 5.924144991364533e-06,
+      "loss": 0.3611,
+      "step": 3204
+    },
+    {
+      "epoch": 1.490466594326461,
+      "grad_norm": 0.37728172540664673,
+      "learning_rate": 5.921485540743516e-06,
+      "loss": 0.3233,
+      "step": 3205
+    },
+    {
+      "epoch": 1.490931638505658,
+      "grad_norm": 0.4780239760875702,
+      "learning_rate": 5.918825820234352e-06,
+      "loss": 0.3523,
+      "step": 3206
+    },
+    {
+      "epoch": 1.491396682684855,
+      "grad_norm": 0.43843451142311096,
+      "learning_rate": 5.9161658306160286e-06,
+      "loss": 0.3479,
+      "step": 3207
+    },
+    {
+      "epoch": 1.4918617268640522,
+      "grad_norm": 0.43270936608314514,
+      "learning_rate": 5.913505572667615e-06,
+      "loss": 0.3498,
+      "step": 3208
+    },
+    {
+      "epoch": 1.492326771043249,
+      "grad_norm": 0.4177435338497162,
+      "learning_rate": 5.910845047168259e-06,
+      "loss": 0.3481,
+      "step": 3209
+    },
+    {
+      "epoch": 1.4927918152224462,
+      "grad_norm": 0.4250778257846832,
+      "learning_rate": 5.908184254897183e-06,
+      "loss": 0.338,
+      "step": 3210
+    },
+    {
+      "epoch": 1.493256859401643,
+      "grad_norm": 0.40649059414863586,
+      "learning_rate": 5.90552319663369e-06,
+      "loss": 0.3486,
+      "step": 3211
+    },
+    {
+      "epoch": 1.4937219035808402,
+      "grad_norm": 0.40856921672821045,
+      "learning_rate": 5.902861873157162e-06,
+      "loss": 0.3573,
+      "step": 3212
+    },
+    {
+      "epoch": 1.4941869477600371,
+      "grad_norm": 0.5227980613708496,
+      "learning_rate": 5.900200285247055e-06,
+      "loss": 0.3883,
+      "step": 3213
+    },
+    {
+      "epoch": 1.4946519919392343,
+      "grad_norm": 0.3745517134666443,
+      "learning_rate": 5.897538433682909e-06,
+      "loss": 0.3327,
+      "step": 3214
+    },
+    {
+      "epoch": 1.4951170361184314,
+      "grad_norm": 0.38121265172958374,
+      "learning_rate": 5.894876319244334e-06,
+      "loss": 0.3519,
+      "step": 3215
+    },
+    {
+      "epoch": 1.4955820802976283,
+      "grad_norm": 0.45746102929115295,
+      "learning_rate": 5.892213942711019e-06,
+      "loss": 0.3917,
+      "step": 3216
+    },
+    {
+      "epoch": 1.4960471244768252,
+      "grad_norm": 0.47537440061569214,
+      "learning_rate": 5.889551304862735e-06,
+      "loss": 0.3296,
+      "step": 3217
+    },
+    {
+      "epoch": 1.4965121686560223,
+      "grad_norm": 0.427947461605072,
+      "learning_rate": 5.8868884064793215e-06,
+      "loss": 0.3772,
+      "step": 3218
+    },
+    {
+      "epoch": 1.4969772128352195,
+      "grad_norm": 0.40912094712257385,
+      "learning_rate": 5.884225248340699e-06,
+      "loss": 0.3814,
+      "step": 3219
+    },
+    {
+      "epoch": 1.4974422570144164,
+      "grad_norm": 0.40474238991737366,
+      "learning_rate": 5.881561831226865e-06,
+      "loss": 0.346,
+      "step": 3220
+    },
+    {
+      "epoch": 1.4979073011936133,
+      "grad_norm": 0.4801974296569824,
+      "learning_rate": 5.878898155917889e-06,
+      "loss": 0.3668,
+      "step": 3221
+    },
+    {
+      "epoch": 1.4983723453728104,
+      "grad_norm": 0.4175057113170624,
+      "learning_rate": 5.8762342231939205e-06,
+      "loss": 0.2951,
+      "step": 3222
+    },
+    {
+      "epoch": 1.4988373895520075,
+      "grad_norm": 0.41870221495628357,
+      "learning_rate": 5.873570033835181e-06,
+      "loss": 0.3757,
+      "step": 3223
+    },
+    {
+      "epoch": 1.4993024337312044,
+      "grad_norm": 0.3709333539009094,
+      "learning_rate": 5.8709055886219665e-06,
+      "loss": 0.3465,
+      "step": 3224
+    },
+    {
+      "epoch": 1.4997674779104015,
+      "grad_norm": 0.4033331871032715,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.3314,
+      "step": 3225
+    },
+    {
+      "epoch": 1.5002325220895987,
+      "grad_norm": 0.45277902483940125,
+      "learning_rate": 5.865575933753686e-06,
+      "loss": 0.3526,
+      "step": 3226
+    },
+    {
+      "epoch": 1.5006975662687956,
+      "grad_norm": 0.476517915725708,
+      "learning_rate": 5.862910725659586e-06,
+      "loss": 0.3651,
+      "step": 3227
+    },
+    {
+      "epoch": 1.5011626104479925,
+      "grad_norm": 0.372768372297287,
+      "learning_rate": 5.860245264832952e-06,
+      "loss": 0.3189,
+      "step": 3228
+    },
+    {
+      "epoch": 1.5016276546271896,
+      "grad_norm": 0.4148904085159302,
+      "learning_rate": 5.857579552054454e-06,
+      "loss": 0.3241,
+      "step": 3229
+    },
+    {
+      "epoch": 1.5020926988063867,
+      "grad_norm": 0.38212111592292786,
+      "learning_rate": 5.854913588104832e-06,
+      "loss": 0.3015,
+      "step": 3230
+    },
+    {
+      "epoch": 1.5025577429855836,
+      "grad_norm": 0.5229579210281372,
+      "learning_rate": 5.85224737376491e-06,
+      "loss": 0.3802,
+      "step": 3231
+    },
+    {
+      "epoch": 1.5030227871647805,
+      "grad_norm": 0.37114113569259644,
+      "learning_rate": 5.849580909815573e-06,
+      "loss": 0.3221,
+      "step": 3232
+    },
+    {
+      "epoch": 1.5034878313439777,
+      "grad_norm": 0.4346405863761902,
+      "learning_rate": 5.84691419703779e-06,
+      "loss": 0.3551,
+      "step": 3233
+    },
+    {
+      "epoch": 1.5039528755231748,
+      "grad_norm": 0.463509202003479,
+      "learning_rate": 5.844247236212593e-06,
+      "loss": 0.3559,
+      "step": 3234
+    },
+    {
+      "epoch": 1.5044179197023717,
+      "grad_norm": 0.5213866829872131,
+      "learning_rate": 5.8415800281210945e-06,
+      "loss": 0.3356,
+      "step": 3235
+    },
+    {
+      "epoch": 1.5048829638815686,
+      "grad_norm": 0.4189200699329376,
+      "learning_rate": 5.838912573544475e-06,
+      "loss": 0.349,
+      "step": 3236
+    },
+    {
+      "epoch": 1.5053480080607657,
+      "grad_norm": 0.41886717081069946,
+      "learning_rate": 5.836244873263989e-06,
+      "loss": 0.3107,
+      "step": 3237
+    },
+    {
+      "epoch": 1.5058130522399629,
+      "grad_norm": 0.4343456029891968,
+      "learning_rate": 5.833576928060964e-06,
+      "loss": 0.3686,
+      "step": 3238
+    },
+    {
+      "epoch": 1.5062780964191598,
+      "grad_norm": 0.41297784447669983,
+      "learning_rate": 5.830908738716797e-06,
+      "loss": 0.342,
+      "step": 3239
+    },
+    {
+      "epoch": 1.5067431405983567,
+      "grad_norm": 0.44648993015289307,
+      "learning_rate": 5.828240306012957e-06,
+      "loss": 0.3716,
+      "step": 3240
+    },
+    {
+      "epoch": 1.507208184777554,
+      "grad_norm": 0.40866819024086,
+      "learning_rate": 5.825571630730984e-06,
+      "loss": 0.3111,
+      "step": 3241
+    },
+    {
+      "epoch": 1.507673228956751,
+      "grad_norm": 0.45804688334465027,
+      "learning_rate": 5.8229027136524896e-06,
+      "loss": 0.3408,
+      "step": 3242
+    },
+    {
+      "epoch": 1.5081382731359478,
+      "grad_norm": 0.4614756107330322,
+      "learning_rate": 5.820233555559157e-06,
+      "loss": 0.3852,
+      "step": 3243
+    },
+    {
+      "epoch": 1.508603317315145,
+      "grad_norm": 0.39953604340553284,
+      "learning_rate": 5.81756415723274e-06,
+      "loss": 0.298,
+      "step": 3244
+    },
+    {
+      "epoch": 1.509068361494342,
+      "grad_norm": 0.4138416349887848,
+      "learning_rate": 5.814894519455061e-06,
+      "loss": 0.3585,
+      "step": 3245
+    },
+    {
+      "epoch": 1.509533405673539,
+      "grad_norm": 0.4480072557926178,
+      "learning_rate": 5.812224643008014e-06,
+      "loss": 0.3583,
+      "step": 3246
+    },
+    {
+      "epoch": 1.509998449852736,
+      "grad_norm": 0.3700372576713562,
+      "learning_rate": 5.809554528673562e-06,
+      "loss": 0.2974,
+      "step": 3247
+    },
+    {
+      "epoch": 1.510463494031933,
+      "grad_norm": 0.4125301241874695,
+      "learning_rate": 5.806884177233737e-06,
+      "loss": 0.3842,
+      "step": 3248
+    },
+    {
+      "epoch": 1.5109285382111302,
+      "grad_norm": 0.4698508679866791,
+      "learning_rate": 5.804213589470644e-06,
+      "loss": 0.3631,
+      "step": 3249
+    },
+    {
+      "epoch": 1.511393582390327,
+      "grad_norm": 0.37476885318756104,
+      "learning_rate": 5.801542766166454e-06,
+      "loss": 0.3072,
+      "step": 3250
+    },
+    {
+      "epoch": 1.511858626569524,
+      "grad_norm": 0.43928253650665283,
+      "learning_rate": 5.7988717081034066e-06,
+      "loss": 0.3819,
+      "step": 3251
+    },
+    {
+      "epoch": 1.512323670748721,
+      "grad_norm": 0.4448581635951996,
+      "learning_rate": 5.796200416063813e-06,
+      "loss": 0.325,
+      "step": 3252
+    },
+    {
+      "epoch": 1.5127887149279182,
+      "grad_norm": 0.5041275024414062,
+      "learning_rate": 5.793528890830049e-06,
+      "loss": 0.3903,
+      "step": 3253
+    },
+    {
+      "epoch": 1.5132537591071151,
+      "grad_norm": 0.3757099211215973,
+      "learning_rate": 5.790857133184563e-06,
+      "loss": 0.3212,
+      "step": 3254
+    },
+    {
+      "epoch": 1.5137188032863123,
+      "grad_norm": 0.5224563479423523,
+      "learning_rate": 5.788185143909868e-06,
+      "loss": 0.3684,
+      "step": 3255
+    },
+    {
+      "epoch": 1.5141838474655094,
+      "grad_norm": 0.46224769949913025,
+      "learning_rate": 5.785512923788549e-06,
+      "loss": 0.354,
+      "step": 3256
+    },
+    {
+      "epoch": 1.5146488916447063,
+      "grad_norm": 0.4003010392189026,
+      "learning_rate": 5.7828404736032515e-06,
+      "loss": 0.3085,
+      "step": 3257
+    },
+    {
+      "epoch": 1.5151139358239032,
+      "grad_norm": 0.4108293950557709,
+      "learning_rate": 5.780167794136696e-06,
+      "loss": 0.3633,
+      "step": 3258
+    },
+    {
+      "epoch": 1.5155789800031003,
+      "grad_norm": 0.4604003131389618,
+      "learning_rate": 5.777494886171667e-06,
+      "loss": 0.3237,
+      "step": 3259
+    },
+    {
+      "epoch": 1.5160440241822974,
+      "grad_norm": 0.49842220544815063,
+      "learning_rate": 5.7748217504910145e-06,
+      "loss": 0.3905,
+      "step": 3260
+    },
+    {
+      "epoch": 1.5165090683614944,
+      "grad_norm": 0.4144563376903534,
+      "learning_rate": 5.772148387877656e-06,
+      "loss": 0.3754,
+      "step": 3261
+    },
+    {
+      "epoch": 1.5169741125406913,
+      "grad_norm": 0.4629983901977539,
+      "learning_rate": 5.7694747991145775e-06,
+      "loss": 0.3352,
+      "step": 3262
+    },
+    {
+      "epoch": 1.5174391567198884,
+      "grad_norm": 0.3761589825153351,
+      "learning_rate": 5.766800984984828e-06,
+      "loss": 0.3488,
+      "step": 3263
+    },
+    {
+      "epoch": 1.5179042008990855,
+      "grad_norm": 0.3842097222805023,
+      "learning_rate": 5.764126946271526e-06,
+      "loss": 0.3479,
+      "step": 3264
+    },
+    {
+      "epoch": 1.5183692450782824,
+      "grad_norm": 0.39669203758239746,
+      "learning_rate": 5.7614526837578525e-06,
+      "loss": 0.3421,
+      "step": 3265
+    },
+    {
+      "epoch": 1.5188342892574793,
+      "grad_norm": 0.42259714007377625,
+      "learning_rate": 5.758778198227057e-06,
+      "loss": 0.36,
+      "step": 3266
+    },
+    {
+      "epoch": 1.5192993334366764,
+      "grad_norm": 0.39835333824157715,
+      "learning_rate": 5.7561034904624525e-06,
+      "loss": 0.3514,
+      "step": 3267
+    },
+    {
+      "epoch": 1.5197643776158736,
+      "grad_norm": 0.383436918258667,
+      "learning_rate": 5.753428561247416e-06,
+      "loss": 0.3601,
+      "step": 3268
+    },
+    {
+      "epoch": 1.5202294217950705,
+      "grad_norm": 0.39197006821632385,
+      "learning_rate": 5.750753411365394e-06,
+      "loss": 0.3377,
+      "step": 3269
+    },
+    {
+      "epoch": 1.5206944659742676,
+      "grad_norm": 0.4436843991279602,
+      "learning_rate": 5.74807804159989e-06,
+      "loss": 0.3428,
+      "step": 3270
+    },
+    {
+      "epoch": 1.5211595101534647,
+      "grad_norm": 0.3785765469074249,
+      "learning_rate": 5.74540245273448e-06,
+      "loss": 0.3329,
+      "step": 3271
+    },
+    {
+      "epoch": 1.5216245543326616,
+      "grad_norm": 0.3833819329738617,
+      "learning_rate": 5.7427266455528e-06,
+      "loss": 0.3533,
+      "step": 3272
+    },
+    {
+      "epoch": 1.5220895985118585,
+      "grad_norm": 0.4789281189441681,
+      "learning_rate": 5.7400506208385486e-06,
+      "loss": 0.3841,
+      "step": 3273
+    },
+    {
+      "epoch": 1.5225546426910557,
+      "grad_norm": 0.4086497724056244,
+      "learning_rate": 5.737374379375491e-06,
+      "loss": 0.3316,
+      "step": 3274
+    },
+    {
+      "epoch": 1.5230196868702528,
+      "grad_norm": 0.39715808629989624,
+      "learning_rate": 5.734697921947456e-06,
+      "loss": 0.3326,
+      "step": 3275
+    },
+    {
+      "epoch": 1.5234847310494497,
+      "grad_norm": 0.45392605662345886,
+      "learning_rate": 5.732021249338333e-06,
+      "loss": 0.3432,
+      "step": 3276
+    },
+    {
+      "epoch": 1.5239497752286466,
+      "grad_norm": 0.4217411279678345,
+      "learning_rate": 5.729344362332075e-06,
+      "loss": 0.3766,
+      "step": 3277
+    },
+    {
+      "epoch": 1.5244148194078437,
+      "grad_norm": 0.4188506603240967,
+      "learning_rate": 5.7266672617127014e-06,
+      "loss": 0.3234,
+      "step": 3278
+    },
+    {
+      "epoch": 1.5248798635870409,
+      "grad_norm": 0.40585535764694214,
+      "learning_rate": 5.723989948264291e-06,
+      "loss": 0.3499,
+      "step": 3279
+    },
+    {
+      "epoch": 1.5253449077662378,
+      "grad_norm": 0.4009935259819031,
+      "learning_rate": 5.721312422770984e-06,
+      "loss": 0.359,
+      "step": 3280
+    },
+    {
+      "epoch": 1.5258099519454347,
+      "grad_norm": 0.4260950982570648,
+      "learning_rate": 5.718634686016985e-06,
+      "loss": 0.3418,
+      "step": 3281
+    },
+    {
+      "epoch": 1.5262749961246318,
+      "grad_norm": 0.40163952112197876,
+      "learning_rate": 5.715956738786559e-06,
+      "loss": 0.3451,
+      "step": 3282
+    },
+    {
+      "epoch": 1.526740040303829,
+      "grad_norm": 0.38455137610435486,
+      "learning_rate": 5.713278581864032e-06,
+      "loss": 0.3522,
+      "step": 3283
+    },
+    {
+      "epoch": 1.5272050844830258,
+      "grad_norm": 0.3814459443092346,
+      "learning_rate": 5.710600216033797e-06,
+      "loss": 0.3334,
+      "step": 3284
+    },
+    {
+      "epoch": 1.527670128662223,
+      "grad_norm": 0.4370444416999817,
+      "learning_rate": 5.7079216420803e-06,
+      "loss": 0.3619,
+      "step": 3285
+    },
+    {
+      "epoch": 1.52813517284142,
+      "grad_norm": 0.40159928798675537,
+      "learning_rate": 5.705242860788052e-06,
+      "loss": 0.3379,
+      "step": 3286
+    },
+    {
+      "epoch": 1.528600217020617,
+      "grad_norm": 0.4162231385707855,
+      "learning_rate": 5.7025638729416275e-06,
+      "loss": 0.3562,
+      "step": 3287
+    },
+    {
+      "epoch": 1.529065261199814,
+      "grad_norm": 0.36917373538017273,
+      "learning_rate": 5.699884679325656e-06,
+      "loss": 0.3496,
+      "step": 3288
+    },
+    {
+      "epoch": 1.529530305379011,
+      "grad_norm": 0.36086469888687134,
+      "learning_rate": 5.697205280724828e-06,
+      "loss": 0.33,
+      "step": 3289
+    },
+    {
+      "epoch": 1.5299953495582082,
+      "grad_norm": 0.434421181678772,
+      "learning_rate": 5.6945256779239e-06,
+      "loss": 0.373,
+      "step": 3290
+    },
+    {
+      "epoch": 1.530460393737405,
+      "grad_norm": 0.39509040117263794,
+      "learning_rate": 5.691845871707682e-06,
+      "loss": 0.3426,
+      "step": 3291
+    },
+    {
+      "epoch": 1.530925437916602,
+      "grad_norm": 0.38857322931289673,
+      "learning_rate": 5.689165862861046e-06,
+      "loss": 0.3389,
+      "step": 3292
+    },
+    {
+      "epoch": 1.531390482095799,
+      "grad_norm": 0.38726887106895447,
+      "learning_rate": 5.686485652168923e-06,
+      "loss": 0.3262,
+      "step": 3293
+    },
+    {
+      "epoch": 1.5318555262749962,
+      "grad_norm": 0.4127531051635742,
+      "learning_rate": 5.683805240416302e-06,
+      "loss": 0.3861,
+      "step": 3294
+    },
+    {
+      "epoch": 1.5323205704541931,
+      "grad_norm": 0.3922295868396759,
+      "learning_rate": 5.681124628388235e-06,
+      "loss": 0.3506,
+      "step": 3295
+    },
+    {
+      "epoch": 1.53278561463339,
+      "grad_norm": 0.38945940136909485,
+      "learning_rate": 5.678443816869828e-06,
+      "loss": 0.34,
+      "step": 3296
+    },
+    {
+      "epoch": 1.5332506588125872,
+      "grad_norm": 0.38314536213874817,
+      "learning_rate": 5.675762806646247e-06,
+      "loss": 0.3136,
+      "step": 3297
+    },
+    {
+      "epoch": 1.5337157029917843,
+      "grad_norm": 0.4257749021053314,
+      "learning_rate": 5.673081598502715e-06,
+      "loss": 0.3641,
+      "step": 3298
+    },
+    {
+      "epoch": 1.5341807471709812,
+      "grad_norm": 0.4407237470149994,
+      "learning_rate": 5.670400193224516e-06,
+      "loss": 0.3381,
+      "step": 3299
+    },
+    {
+      "epoch": 1.5346457913501783,
+      "grad_norm": 0.4940028190612793,
+      "learning_rate": 5.66771859159699e-06,
+      "loss": 0.336,
+      "step": 3300
+    },
+    {
+      "epoch": 1.5351108355293754,
+      "grad_norm": 0.40870407223701477,
+      "learning_rate": 5.665036794405535e-06,
+      "loss": 0.3687,
+      "step": 3301
+    },
+    {
+      "epoch": 1.5355758797085723,
+      "grad_norm": 0.4065590500831604,
+      "learning_rate": 5.662354802435606e-06,
+      "loss": 0.3937,
+      "step": 3302
+    },
+    {
+      "epoch": 1.5360409238877692,
+      "grad_norm": 0.39256197214126587,
+      "learning_rate": 5.659672616472712e-06,
+      "loss": 0.3364,
+      "step": 3303
+    },
+    {
+      "epoch": 1.5365059680669664,
+      "grad_norm": 0.4631505608558655,
+      "learning_rate": 5.656990237302426e-06,
+      "loss": 0.3388,
+      "step": 3304
+    },
+    {
+      "epoch": 1.5369710122461635,
+      "grad_norm": 0.3832097053527832,
+      "learning_rate": 5.6543076657103705e-06,
+      "loss": 0.3108,
+      "step": 3305
+    },
+    {
+      "epoch": 1.5374360564253604,
+      "grad_norm": 0.44715628027915955,
+      "learning_rate": 5.651624902482225e-06,
+      "loss": 0.3716,
+      "step": 3306
+    },
+    {
+      "epoch": 1.5379011006045573,
+      "grad_norm": 0.5109707713127136,
+      "learning_rate": 5.648941948403732e-06,
+      "loss": 0.3541,
+      "step": 3307
+    },
+    {
+      "epoch": 1.5383661447837544,
+      "grad_norm": 0.35396090149879456,
+      "learning_rate": 5.646258804260685e-06,
+      "loss": 0.3068,
+      "step": 3308
+    },
+    {
+      "epoch": 1.5388311889629516,
+      "grad_norm": 0.5051494836807251,
+      "learning_rate": 5.643575470838929e-06,
+      "loss": 0.4256,
+      "step": 3309
+    },
+    {
+      "epoch": 1.5392962331421485,
+      "grad_norm": 0.4141186475753784,
+      "learning_rate": 5.640891948924373e-06,
+      "loss": 0.3335,
+      "step": 3310
+    },
+    {
+      "epoch": 1.5397612773213454,
+      "grad_norm": 0.3828221559524536,
+      "learning_rate": 5.638208239302975e-06,
+      "loss": 0.3464,
+      "step": 3311
+    },
+    {
+      "epoch": 1.5402263215005425,
+      "grad_norm": 0.39551404118537903,
+      "learning_rate": 5.6355243427607475e-06,
+      "loss": 0.387,
+      "step": 3312
+    },
+    {
+      "epoch": 1.5406913656797396,
+      "grad_norm": 0.3690919280052185,
+      "learning_rate": 5.632840260083766e-06,
+      "loss": 0.2866,
+      "step": 3313
+    },
+    {
+      "epoch": 1.5411564098589365,
+      "grad_norm": 0.458450585603714,
+      "learning_rate": 5.630155992058151e-06,
+      "loss": 0.4064,
+      "step": 3314
+    },
+    {
+      "epoch": 1.5416214540381337,
+      "grad_norm": 0.4220149517059326,
+      "learning_rate": 5.6274715394700805e-06,
+      "loss": 0.377,
+      "step": 3315
+    },
+    {
+      "epoch": 1.5420864982173308,
+      "grad_norm": 0.41717949509620667,
+      "learning_rate": 5.62478690310579e-06,
+      "loss": 0.3553,
+      "step": 3316
+    },
+    {
+      "epoch": 1.5425515423965277,
+      "grad_norm": 0.42877861857414246,
+      "learning_rate": 5.622102083751563e-06,
+      "loss": 0.3538,
+      "step": 3317
+    },
+    {
+      "epoch": 1.5430165865757246,
+      "grad_norm": 0.44071346521377563,
+      "learning_rate": 5.61941708219374e-06,
+      "loss": 0.3461,
+      "step": 3318
+    },
+    {
+      "epoch": 1.5434816307549217,
+      "grad_norm": 0.47426027059555054,
+      "learning_rate": 5.6167318992187155e-06,
+      "loss": 0.3504,
+      "step": 3319
+    },
+    {
+      "epoch": 1.5439466749341189,
+      "grad_norm": 0.4373508393764496,
+      "learning_rate": 5.614046535612936e-06,
+      "loss": 0.3585,
+      "step": 3320
+    },
+    {
+      "epoch": 1.5444117191133158,
+      "grad_norm": 0.48637011647224426,
+      "learning_rate": 5.6113609921629e-06,
+      "loss": 0.3465,
+      "step": 3321
+    },
+    {
+      "epoch": 1.5448767632925127,
+      "grad_norm": 0.3798205852508545,
+      "learning_rate": 5.60867526965516e-06,
+      "loss": 0.3345,
+      "step": 3322
+    },
+    {
+      "epoch": 1.5453418074717098,
+      "grad_norm": 0.3805210590362549,
+      "learning_rate": 5.60598936887632e-06,
+      "loss": 0.3352,
+      "step": 3323
+    },
+    {
+      "epoch": 1.545806851650907,
+      "grad_norm": 0.44706982374191284,
+      "learning_rate": 5.603303290613036e-06,
+      "loss": 0.3934,
+      "step": 3324
+    },
+    {
+      "epoch": 1.5462718958301038,
+      "grad_norm": 0.38400477170944214,
+      "learning_rate": 5.600617035652019e-06,
+      "loss": 0.3168,
+      "step": 3325
+    },
+    {
+      "epoch": 1.5467369400093007,
+      "grad_norm": 0.3989894688129425,
+      "learning_rate": 5.597930604780028e-06,
+      "loss": 0.3661,
+      "step": 3326
+    },
+    {
+      "epoch": 1.5472019841884979,
+      "grad_norm": 0.40619131922721863,
+      "learning_rate": 5.595243998783876e-06,
+      "loss": 0.355,
+      "step": 3327
+    },
+    {
+      "epoch": 1.547667028367695,
+      "grad_norm": 0.4321429431438446,
+      "learning_rate": 5.592557218450424e-06,
+      "loss": 0.3404,
+      "step": 3328
+    },
+    {
+      "epoch": 1.548132072546892,
+      "grad_norm": 0.3868047893047333,
+      "learning_rate": 5.589870264566588e-06,
+      "loss": 0.3609,
+      "step": 3329
+    },
+    {
+      "epoch": 1.548597116726089,
+      "grad_norm": 0.4493039548397064,
+      "learning_rate": 5.587183137919332e-06,
+      "loss": 0.3628,
+      "step": 3330
+    },
+    {
+      "epoch": 1.5490621609052861,
+      "grad_norm": 0.4076448082923889,
+      "learning_rate": 5.584495839295674e-06,
+      "loss": 0.3429,
+      "step": 3331
+    },
+    {
+      "epoch": 1.549527205084483,
+      "grad_norm": 0.4267314374446869,
+      "learning_rate": 5.58180836948268e-06,
+      "loss": 0.3768,
+      "step": 3332
+    },
+    {
+      "epoch": 1.54999224926368,
+      "grad_norm": 0.4145706295967102,
+      "learning_rate": 5.579120729267463e-06,
+      "loss": 0.3438,
+      "step": 3333
+    },
+    {
+      "epoch": 1.550457293442877,
+      "grad_norm": 0.42286407947540283,
+      "learning_rate": 5.576432919437193e-06,
+      "loss": 0.3253,
+      "step": 3334
+    },
+    {
+      "epoch": 1.5509223376220742,
+      "grad_norm": 0.3949490785598755,
+      "learning_rate": 5.57374494077908e-06,
+      "loss": 0.346,
+      "step": 3335
+    },
+    {
+      "epoch": 1.5513873818012711,
+      "grad_norm": 0.44627898931503296,
+      "learning_rate": 5.571056794080396e-06,
+      "loss": 0.3721,
+      "step": 3336
+    },
+    {
+      "epoch": 1.551852425980468,
+      "grad_norm": 0.3899642527103424,
+      "learning_rate": 5.568368480128453e-06,
+      "loss": 0.3405,
+      "step": 3337
+    },
+    {
+      "epoch": 1.5523174701596651,
+      "grad_norm": 0.4098135530948639,
+      "learning_rate": 5.565679999710614e-06,
+      "loss": 0.3319,
+      "step": 3338
+    },
+    {
+      "epoch": 1.5527825143388623,
+      "grad_norm": 0.39709505438804626,
+      "learning_rate": 5.562991353614292e-06,
+      "loss": 0.3294,
+      "step": 3339
+    },
+    {
+      "epoch": 1.5532475585180592,
+      "grad_norm": 0.3686107099056244,
+      "learning_rate": 5.560302542626947e-06,
+      "loss": 0.359,
+      "step": 3340
+    },
+    {
+      "epoch": 1.553712602697256,
+      "grad_norm": 0.3666870892047882,
+      "learning_rate": 5.557613567536087e-06,
+      "loss": 0.3197,
+      "step": 3341
+    },
+    {
+      "epoch": 1.5541776468764532,
+      "grad_norm": 0.4867759048938751,
+      "learning_rate": 5.554924429129271e-06,
+      "loss": 0.3515,
+      "step": 3342
+    },
+    {
+      "epoch": 1.5546426910556503,
+      "grad_norm": 0.3786954879760742,
+      "learning_rate": 5.552235128194105e-06,
+      "loss": 0.3297,
+      "step": 3343
+    },
+    {
+      "epoch": 1.5551077352348472,
+      "grad_norm": 0.3850143849849701,
+      "learning_rate": 5.5495456655182376e-06,
+      "loss": 0.3379,
+      "step": 3344
+    },
+    {
+      "epoch": 1.5555727794140444,
+      "grad_norm": 0.4032987952232361,
+      "learning_rate": 5.546856041889374e-06,
+      "loss": 0.3273,
+      "step": 3345
+    },
+    {
+      "epoch": 1.5560378235932415,
+      "grad_norm": 0.4159895181655884,
+      "learning_rate": 5.544166258095256e-06,
+      "loss": 0.3851,
+      "step": 3346
+    },
+    {
+      "epoch": 1.5565028677724384,
+      "grad_norm": 0.41379231214523315,
+      "learning_rate": 5.54147631492368e-06,
+      "loss": 0.3465,
+      "step": 3347
+    },
+    {
+      "epoch": 1.5569679119516353,
+      "grad_norm": 0.36884400248527527,
+      "learning_rate": 5.538786213162487e-06,
+      "loss": 0.3264,
+      "step": 3348
+    },
+    {
+      "epoch": 1.5574329561308324,
+      "grad_norm": 0.3928481638431549,
+      "learning_rate": 5.536095953599565e-06,
+      "loss": 0.3351,
+      "step": 3349
+    },
+    {
+      "epoch": 1.5578980003100296,
+      "grad_norm": 0.4361253082752228,
+      "learning_rate": 5.533405537022846e-06,
+      "loss": 0.3428,
+      "step": 3350
+    },
+    {
+      "epoch": 1.5583630444892265,
+      "grad_norm": 0.36980322003364563,
+      "learning_rate": 5.530714964220308e-06,
+      "loss": 0.3544,
+      "step": 3351
+    },
+    {
+      "epoch": 1.5588280886684234,
+      "grad_norm": 0.3828973174095154,
+      "learning_rate": 5.528024235979978e-06,
+      "loss": 0.3581,
+      "step": 3352
+    },
+    {
+      "epoch": 1.5592931328476205,
+      "grad_norm": 0.3844209313392639,
+      "learning_rate": 5.525333353089926e-06,
+      "loss": 0.3084,
+      "step": 3353
+    },
+    {
+      "epoch": 1.5597581770268176,
+      "grad_norm": 0.4435608386993408,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.3475,
+      "step": 3354
+    },
+    {
+      "epoch": 1.5602232212060145,
+      "grad_norm": 0.36281177401542664,
+      "learning_rate": 5.519951126513164e-06,
+      "loss": 0.3257,
+      "step": 3355
+    },
+    {
+      "epoch": 1.5606882653852114,
+      "grad_norm": 0.4244144558906555,
+      "learning_rate": 5.517259784402823e-06,
+      "loss": 0.3716,
+      "step": 3356
+    },
+    {
+      "epoch": 1.5611533095644086,
+      "grad_norm": 0.39875200390815735,
+      "learning_rate": 5.514568290795492e-06,
+      "loss": 0.3374,
+      "step": 3357
+    },
+    {
+      "epoch": 1.5616183537436057,
+      "grad_norm": 0.42064738273620605,
+      "learning_rate": 5.511876646479466e-06,
+      "loss": 0.3643,
+      "step": 3358
+    },
+    {
+      "epoch": 1.5620833979228026,
+      "grad_norm": 0.4080541133880615,
+      "learning_rate": 5.509184852243084e-06,
+      "loss": 0.3452,
+      "step": 3359
+    },
+    {
+      "epoch": 1.5625484421019997,
+      "grad_norm": 0.3764135539531708,
+      "learning_rate": 5.5064929088747324e-06,
+      "loss": 0.3365,
+      "step": 3360
+    },
+    {
+      "epoch": 1.5630134862811969,
+      "grad_norm": 0.39057546854019165,
+      "learning_rate": 5.503800817162833e-06,
+      "loss": 0.3422,
+      "step": 3361
+    },
+    {
+      "epoch": 1.5634785304603938,
+      "grad_norm": 0.3925474286079407,
+      "learning_rate": 5.501108577895858e-06,
+      "loss": 0.3133,
+      "step": 3362
+    },
+    {
+      "epoch": 1.5639435746395907,
+      "grad_norm": 0.4536259174346924,
+      "learning_rate": 5.49841619186232e-06,
+      "loss": 0.3756,
+      "step": 3363
+    },
+    {
+      "epoch": 1.5644086188187878,
+      "grad_norm": 0.4168597161769867,
+      "learning_rate": 5.495723659850776e-06,
+      "loss": 0.3313,
+      "step": 3364
+    },
+    {
+      "epoch": 1.564873662997985,
+      "grad_norm": 0.35290852189064026,
+      "learning_rate": 5.493030982649823e-06,
+      "loss": 0.3392,
+      "step": 3365
+    },
+    {
+      "epoch": 1.5653387071771818,
+      "grad_norm": 0.36755627393722534,
+      "learning_rate": 5.4903381610481034e-06,
+      "loss": 0.3271,
+      "step": 3366
+    },
+    {
+      "epoch": 1.5658037513563787,
+      "grad_norm": 0.42110899090766907,
+      "learning_rate": 5.487645195834302e-06,
+      "loss": 0.3726,
+      "step": 3367
+    },
+    {
+      "epoch": 1.5662687955355759,
+      "grad_norm": 0.3488036096096039,
+      "learning_rate": 5.484952087797144e-06,
+      "loss": 0.3097,
+      "step": 3368
+    },
+    {
+      "epoch": 1.566733839714773,
+      "grad_norm": 0.4092177152633667,
+      "learning_rate": 5.482258837725397e-06,
+      "loss": 0.3727,
+      "step": 3369
+    },
+    {
+      "epoch": 1.5671988838939699,
+      "grad_norm": 0.3953334093093872,
+      "learning_rate": 5.479565446407867e-06,
+      "loss": 0.3146,
+      "step": 3370
+    },
+    {
+      "epoch": 1.5676639280731668,
+      "grad_norm": 0.3620423376560211,
+      "learning_rate": 5.47687191463341e-06,
+      "loss": 0.3622,
+      "step": 3371
+    },
+    {
+      "epoch": 1.5681289722523641,
+      "grad_norm": 0.3988654613494873,
+      "learning_rate": 5.4741782431909144e-06,
+      "loss": 0.345,
+      "step": 3372
+    },
+    {
+      "epoch": 1.568594016431561,
+      "grad_norm": 0.3772416114807129,
+      "learning_rate": 5.471484432869314e-06,
+      "loss": 0.3463,
+      "step": 3373
+    },
+    {
+      "epoch": 1.569059060610758,
+      "grad_norm": 0.4692382216453552,
+      "learning_rate": 5.4687904844575814e-06,
+      "loss": 0.3523,
+      "step": 3374
+    },
+    {
+      "epoch": 1.569524104789955,
+      "grad_norm": 0.38080427050590515,
+      "learning_rate": 5.46609639874473e-06,
+      "loss": 0.3665,
+      "step": 3375
+    },
+    {
+      "epoch": 1.5699891489691522,
+      "grad_norm": 0.3956538438796997,
+      "learning_rate": 5.4634021765198135e-06,
+      "loss": 0.3536,
+      "step": 3376
+    },
+    {
+      "epoch": 1.570454193148349,
+      "grad_norm": 0.46239396929740906,
+      "learning_rate": 5.460707818571928e-06,
+      "loss": 0.3549,
+      "step": 3377
+    },
+    {
+      "epoch": 1.570919237327546,
+      "grad_norm": 0.42876169085502625,
+      "learning_rate": 5.458013325690205e-06,
+      "loss": 0.3599,
+      "step": 3378
+    },
+    {
+      "epoch": 1.5713842815067431,
+      "grad_norm": 0.3865092992782593,
+      "learning_rate": 5.455318698663819e-06,
+      "loss": 0.3271,
+      "step": 3379
+    },
+    {
+      "epoch": 1.5718493256859403,
+      "grad_norm": 0.40359625220298767,
+      "learning_rate": 5.452623938281983e-06,
+      "loss": 0.3451,
+      "step": 3380
+    },
+    {
+      "epoch": 1.5723143698651372,
+      "grad_norm": 0.4457720220088959,
+      "learning_rate": 5.449929045333946e-06,
+      "loss": 0.3897,
+      "step": 3381
+    },
+    {
+      "epoch": 1.572779414044334,
+      "grad_norm": 0.4244859516620636,
+      "learning_rate": 5.447234020608999e-06,
+      "loss": 0.3683,
+      "step": 3382
+    },
+    {
+      "epoch": 1.5732444582235312,
+      "grad_norm": 0.37125489115715027,
+      "learning_rate": 5.444538864896472e-06,
+      "loss": 0.3132,
+      "step": 3383
+    },
+    {
+      "epoch": 1.5737095024027283,
+      "grad_norm": 0.3960449993610382,
+      "learning_rate": 5.441843578985735e-06,
+      "loss": 0.3133,
+      "step": 3384
+    },
+    {
+      "epoch": 1.5741745465819252,
+      "grad_norm": 0.4028528034687042,
+      "learning_rate": 5.439148163666188e-06,
+      "loss": 0.367,
+      "step": 3385
+    },
+    {
+      "epoch": 1.5746395907611221,
+      "grad_norm": 0.38724517822265625,
+      "learning_rate": 5.436452619727278e-06,
+      "loss": 0.3709,
+      "step": 3386
+    },
+    {
+      "epoch": 1.5751046349403195,
+      "grad_norm": 0.41816893219947815,
+      "learning_rate": 5.4337569479584866e-06,
+      "loss": 0.3545,
+      "step": 3387
+    },
+    {
+      "epoch": 1.5755696791195164,
+      "grad_norm": 0.3713048994541168,
+      "learning_rate": 5.431061149149327e-06,
+      "loss": 0.3023,
+      "step": 3388
+    },
+    {
+      "epoch": 1.5760347232987133,
+      "grad_norm": 0.37138456106185913,
+      "learning_rate": 5.428365224089362e-06,
+      "loss": 0.3382,
+      "step": 3389
+    },
+    {
+      "epoch": 1.5764997674779104,
+      "grad_norm": 0.39625900983810425,
+      "learning_rate": 5.425669173568179e-06,
+      "loss": 0.3549,
+      "step": 3390
+    },
+    {
+      "epoch": 1.5769648116571076,
+      "grad_norm": 0.4074704349040985,
+      "learning_rate": 5.42297299837541e-06,
+      "loss": 0.3267,
+      "step": 3391
+    },
+    {
+      "epoch": 1.5774298558363045,
+      "grad_norm": 0.41984236240386963,
+      "learning_rate": 5.42027669930072e-06,
+      "loss": 0.3526,
+      "step": 3392
+    },
+    {
+      "epoch": 1.5778949000155014,
+      "grad_norm": 0.38742825388908386,
+      "learning_rate": 5.417580277133812e-06,
+      "loss": 0.3161,
+      "step": 3393
+    },
+    {
+      "epoch": 1.5783599441946985,
+      "grad_norm": 0.41518786549568176,
+      "learning_rate": 5.414883732664422e-06,
+      "loss": 0.3263,
+      "step": 3394
+    },
+    {
+      "epoch": 1.5788249883738956,
+      "grad_norm": 0.41794711351394653,
+      "learning_rate": 5.412187066682327e-06,
+      "loss": 0.3718,
+      "step": 3395
+    },
+    {
+      "epoch": 1.5792900325530925,
+      "grad_norm": 0.42802178859710693,
+      "learning_rate": 5.409490279977335e-06,
+      "loss": 0.3883,
+      "step": 3396
+    },
+    {
+      "epoch": 1.5797550767322894,
+      "grad_norm": 0.3825167715549469,
+      "learning_rate": 5.406793373339292e-06,
+      "loss": 0.3426,
+      "step": 3397
+    },
+    {
+      "epoch": 1.5802201209114866,
+      "grad_norm": 0.3995634913444519,
+      "learning_rate": 5.404096347558078e-06,
+      "loss": 0.3749,
+      "step": 3398
+    },
+    {
+      "epoch": 1.5806851650906837,
+      "grad_norm": 0.35901325941085815,
+      "learning_rate": 5.4013992034236065e-06,
+      "loss": 0.3438,
+      "step": 3399
+    },
+    {
+      "epoch": 1.5811502092698806,
+      "grad_norm": 0.3686612546443939,
+      "learning_rate": 5.398701941725827e-06,
+      "loss": 0.3198,
+      "step": 3400
+    },
+    {
+      "epoch": 1.5816152534490775,
+      "grad_norm": 0.3947772979736328,
+      "learning_rate": 5.396004563254728e-06,
+      "loss": 0.3274,
+      "step": 3401
+    },
+    {
+      "epoch": 1.5820802976282748,
+      "grad_norm": 0.45323094725608826,
+      "learning_rate": 5.393307068800322e-06,
+      "loss": 0.4098,
+      "step": 3402
+    },
+    {
+      "epoch": 1.5825453418074718,
+      "grad_norm": 0.346132755279541,
+      "learning_rate": 5.390609459152666e-06,
+      "loss": 0.3316,
+      "step": 3403
+    },
+    {
+      "epoch": 1.5830103859866687,
+      "grad_norm": 0.39992961287498474,
+      "learning_rate": 5.387911735101845e-06,
+      "loss": 0.3548,
+      "step": 3404
+    },
+    {
+      "epoch": 1.5834754301658658,
+      "grad_norm": 0.40256255865097046,
+      "learning_rate": 5.385213897437975e-06,
+      "loss": 0.3207,
+      "step": 3405
+    },
+    {
+      "epoch": 1.583940474345063,
+      "grad_norm": 0.428801029920578,
+      "learning_rate": 5.3825159469512135e-06,
+      "loss": 0.3389,
+      "step": 3406
+    },
+    {
+      "epoch": 1.5844055185242598,
+      "grad_norm": 0.40780290961265564,
+      "learning_rate": 5.3798178844317435e-06,
+      "loss": 0.3221,
+      "step": 3407
+    },
+    {
+      "epoch": 1.5848705627034567,
+      "grad_norm": 0.38632360100746155,
+      "learning_rate": 5.377119710669785e-06,
+      "loss": 0.3096,
+      "step": 3408
+    },
+    {
+      "epoch": 1.5853356068826538,
+      "grad_norm": 0.45682859420776367,
+      "learning_rate": 5.374421426455589e-06,
+      "loss": 0.3737,
+      "step": 3409
+    },
+    {
+      "epoch": 1.585800651061851,
+      "grad_norm": 0.3837336599826813,
+      "learning_rate": 5.371723032579439e-06,
+      "loss": 0.3283,
+      "step": 3410
+    },
+    {
+      "epoch": 1.5862656952410479,
+      "grad_norm": 0.4327690303325653,
+      "learning_rate": 5.369024529831649e-06,
+      "loss": 0.3497,
+      "step": 3411
+    },
+    {
+      "epoch": 1.5867307394202448,
+      "grad_norm": 0.4153764843940735,
+      "learning_rate": 5.366325919002569e-06,
+      "loss": 0.3354,
+      "step": 3412
+    },
+    {
+      "epoch": 1.587195783599442,
+      "grad_norm": 0.399143785238266,
+      "learning_rate": 5.36362720088258e-06,
+      "loss": 0.3495,
+      "step": 3413
+    },
+    {
+      "epoch": 1.587660827778639,
+      "grad_norm": 0.40303295850753784,
+      "learning_rate": 5.3609283762620875e-06,
+      "loss": 0.3219,
+      "step": 3414
+    },
+    {
+      "epoch": 1.588125871957836,
+      "grad_norm": 0.40428194403648376,
+      "learning_rate": 5.358229445931538e-06,
+      "loss": 0.3199,
+      "step": 3415
+    },
+    {
+      "epoch": 1.588590916137033,
+      "grad_norm": 0.41889360547065735,
+      "learning_rate": 5.355530410681402e-06,
+      "loss": 0.3695,
+      "step": 3416
+    },
+    {
+      "epoch": 1.5890559603162302,
+      "grad_norm": 0.43096888065338135,
+      "learning_rate": 5.352831271302183e-06,
+      "loss": 0.3836,
+      "step": 3417
+    },
+    {
+      "epoch": 1.589521004495427,
+      "grad_norm": 0.40473613142967224,
+      "learning_rate": 5.350132028584416e-06,
+      "loss": 0.3579,
+      "step": 3418
+    },
+    {
+      "epoch": 1.589986048674624,
+      "grad_norm": 0.39574551582336426,
+      "learning_rate": 5.3474326833186656e-06,
+      "loss": 0.3357,
+      "step": 3419
+    },
+    {
+      "epoch": 1.5904510928538211,
+      "grad_norm": 0.38140177726745605,
+      "learning_rate": 5.344733236295525e-06,
+      "loss": 0.3498,
+      "step": 3420
+    },
+    {
+      "epoch": 1.5909161370330183,
+      "grad_norm": 0.4418228566646576,
+      "learning_rate": 5.34203368830562e-06,
+      "loss": 0.3608,
+      "step": 3421
+    },
+    {
+      "epoch": 1.5913811812122152,
+      "grad_norm": 0.34866416454315186,
+      "learning_rate": 5.339334040139603e-06,
+      "loss": 0.3159,
+      "step": 3422
+    },
+    {
+      "epoch": 1.591846225391412,
+      "grad_norm": 0.4378582239151001,
+      "learning_rate": 5.336634292588156e-06,
+      "loss": 0.4073,
+      "step": 3423
+    },
+    {
+      "epoch": 1.5923112695706092,
+      "grad_norm": 0.3770564794540405,
+      "learning_rate": 5.333934446441994e-06,
+      "loss": 0.3165,
+      "step": 3424
+    },
+    {
+      "epoch": 1.5927763137498063,
+      "grad_norm": 0.4754953682422638,
+      "learning_rate": 5.3312345024918575e-06,
+      "loss": 0.4054,
+      "step": 3425
+    },
+    {
+      "epoch": 1.5932413579290032,
+      "grad_norm": 0.3792072534561157,
+      "learning_rate": 5.328534461528515e-06,
+      "loss": 0.2908,
+      "step": 3426
+    },
+    {
+      "epoch": 1.5937064021082001,
+      "grad_norm": 0.4275760054588318,
+      "learning_rate": 5.325834324342765e-06,
+      "loss": 0.3807,
+      "step": 3427
+    },
+    {
+      "epoch": 1.5941714462873973,
+      "grad_norm": 0.4156563878059387,
+      "learning_rate": 5.323134091725434e-06,
+      "loss": 0.3607,
+      "step": 3428
+    },
+    {
+      "epoch": 1.5946364904665944,
+      "grad_norm": 0.3854392170906067,
+      "learning_rate": 5.320433764467375e-06,
+      "loss": 0.3535,
+      "step": 3429
+    },
+    {
+      "epoch": 1.5951015346457913,
+      "grad_norm": 0.40652233362197876,
+      "learning_rate": 5.3177333433594734e-06,
+      "loss": 0.338,
+      "step": 3430
+    },
+    {
+      "epoch": 1.5955665788249884,
+      "grad_norm": 0.4638773202896118,
+      "learning_rate": 5.315032829192636e-06,
+      "loss": 0.3495,
+      "step": 3431
+    },
+    {
+      "epoch": 1.5960316230041856,
+      "grad_norm": 0.40089505910873413,
+      "learning_rate": 5.312332222757799e-06,
+      "loss": 0.309,
+      "step": 3432
+    },
+    {
+      "epoch": 1.5964966671833825,
+      "grad_norm": 0.41283005475997925,
+      "learning_rate": 5.309631524845929e-06,
+      "loss": 0.3703,
+      "step": 3433
+    },
+    {
+      "epoch": 1.5969617113625794,
+      "grad_norm": 0.4164558947086334,
+      "learning_rate": 5.306930736248013e-06,
+      "loss": 0.3475,
+      "step": 3434
+    },
+    {
+      "epoch": 1.5974267555417765,
+      "grad_norm": 0.38702747225761414,
+      "learning_rate": 5.30422985775507e-06,
+      "loss": 0.3319,
+      "step": 3435
+    },
+    {
+      "epoch": 1.5978917997209736,
+      "grad_norm": 0.3754745423793793,
+      "learning_rate": 5.301528890158143e-06,
+      "loss": 0.3057,
+      "step": 3436
+    },
+    {
+      "epoch": 1.5983568439001705,
+      "grad_norm": 0.41013574600219727,
+      "learning_rate": 5.298827834248303e-06,
+      "loss": 0.3675,
+      "step": 3437
+    },
+    {
+      "epoch": 1.5988218880793674,
+      "grad_norm": 0.43470141291618347,
+      "learning_rate": 5.296126690816644e-06,
+      "loss": 0.3695,
+      "step": 3438
+    },
+    {
+      "epoch": 1.5992869322585646,
+      "grad_norm": 0.42870068550109863,
+      "learning_rate": 5.293425460654288e-06,
+      "loss": 0.3576,
+      "step": 3439
+    },
+    {
+      "epoch": 1.5997519764377617,
+      "grad_norm": 0.3700660765171051,
+      "learning_rate": 5.290724144552379e-06,
+      "loss": 0.3345,
+      "step": 3440
+    },
+    {
+      "epoch": 1.6002170206169586,
+      "grad_norm": 0.37624475359916687,
+      "learning_rate": 5.288022743302093e-06,
+      "loss": 0.3434,
+      "step": 3441
+    },
+    {
+      "epoch": 1.6006820647961555,
+      "grad_norm": 0.4745047092437744,
+      "learning_rate": 5.2853212576946225e-06,
+      "loss": 0.367,
+      "step": 3442
+    },
+    {
+      "epoch": 1.6011471089753526,
+      "grad_norm": 0.3658214807510376,
+      "learning_rate": 5.282619688521189e-06,
+      "loss": 0.3596,
+      "step": 3443
+    },
+    {
+      "epoch": 1.6016121531545497,
+      "grad_norm": 0.44426223635673523,
+      "learning_rate": 5.2799180365730405e-06,
+      "loss": 0.3377,
+      "step": 3444
+    },
+    {
+      "epoch": 1.6020771973337466,
+      "grad_norm": 0.3807421326637268,
+      "learning_rate": 5.2772163026414455e-06,
+      "loss": 0.3378,
+      "step": 3445
+    },
+    {
+      "epoch": 1.6025422415129438,
+      "grad_norm": 0.4405421316623688,
+      "learning_rate": 5.274514487517698e-06,
+      "loss": 0.3689,
+      "step": 3446
+    },
+    {
+      "epoch": 1.603007285692141,
+      "grad_norm": 0.36217546463012695,
+      "learning_rate": 5.271812591993116e-06,
+      "loss": 0.3242,
+      "step": 3447
+    },
+    {
+      "epoch": 1.6034723298713378,
+      "grad_norm": 0.4049311876296997,
+      "learning_rate": 5.269110616859041e-06,
+      "loss": 0.3661,
+      "step": 3448
+    },
+    {
+      "epoch": 1.6039373740505347,
+      "grad_norm": 0.4050389230251312,
+      "learning_rate": 5.266408562906838e-06,
+      "loss": 0.3447,
+      "step": 3449
+    },
+    {
+      "epoch": 1.6044024182297318,
+      "grad_norm": 0.40756580233573914,
+      "learning_rate": 5.263706430927895e-06,
+      "loss": 0.3566,
+      "step": 3450
+    },
+    {
+      "epoch": 1.604867462408929,
+      "grad_norm": 0.3964058458805084,
+      "learning_rate": 5.261004221713621e-06,
+      "loss": 0.3193,
+      "step": 3451
+    },
+    {
+      "epoch": 1.6053325065881259,
+      "grad_norm": 0.439263254404068,
+      "learning_rate": 5.258301936055449e-06,
+      "loss": 0.359,
+      "step": 3452
+    },
+    {
+      "epoch": 1.6057975507673228,
+      "grad_norm": 0.4526593089103699,
+      "learning_rate": 5.255599574744836e-06,
+      "loss": 0.3515,
+      "step": 3453
+    },
+    {
+      "epoch": 1.60626259494652,
+      "grad_norm": 0.43033599853515625,
+      "learning_rate": 5.252897138573261e-06,
+      "loss": 0.3492,
+      "step": 3454
+    },
+    {
+      "epoch": 1.606727639125717,
+      "grad_norm": 0.43363288044929504,
+      "learning_rate": 5.2501946283322204e-06,
+      "loss": 0.3523,
+      "step": 3455
+    },
+    {
+      "epoch": 1.607192683304914,
+      "grad_norm": 0.3905112147331238,
+      "learning_rate": 5.247492044813237e-06,
+      "loss": 0.3388,
+      "step": 3456
+    },
+    {
+      "epoch": 1.6076577274841108,
+      "grad_norm": 0.45133721828460693,
+      "learning_rate": 5.244789388807855e-06,
+      "loss": 0.3748,
+      "step": 3457
+    },
+    {
+      "epoch": 1.608122771663308,
+      "grad_norm": 0.5364686250686646,
+      "learning_rate": 5.242086661107635e-06,
+      "loss": 0.3814,
+      "step": 3458
+    },
+    {
+      "epoch": 1.608587815842505,
+      "grad_norm": 0.4719644784927368,
+      "learning_rate": 5.239383862504166e-06,
+      "loss": 0.3346,
+      "step": 3459
+    },
+    {
+      "epoch": 1.609052860021702,
+      "grad_norm": 0.4059002697467804,
+      "learning_rate": 5.236680993789052e-06,
+      "loss": 0.3241,
+      "step": 3460
+    },
+    {
+      "epoch": 1.6095179042008991,
+      "grad_norm": 0.4945448040962219,
+      "learning_rate": 5.2339780557539185e-06,
+      "loss": 0.3588,
+      "step": 3461
+    },
+    {
+      "epoch": 1.6099829483800963,
+      "grad_norm": 0.49811092019081116,
+      "learning_rate": 5.231275049190414e-06,
+      "loss": 0.3167,
+      "step": 3462
+    },
+    {
+      "epoch": 1.6104479925592932,
+      "grad_norm": 0.47977715730667114,
+      "learning_rate": 5.228571974890204e-06,
+      "loss": 0.3564,
+      "step": 3463
+    },
+    {
+      "epoch": 1.61091303673849,
+      "grad_norm": 0.39883333444595337,
+      "learning_rate": 5.225868833644973e-06,
+      "loss": 0.3644,
+      "step": 3464
+    },
+    {
+      "epoch": 1.6113780809176872,
+      "grad_norm": 0.38860994577407837,
+      "learning_rate": 5.223165626246432e-06,
+      "loss": 0.347,
+      "step": 3465
+    },
+    {
+      "epoch": 1.6118431250968843,
+      "grad_norm": 0.4335678815841675,
+      "learning_rate": 5.220462353486304e-06,
+      "loss": 0.3534,
+      "step": 3466
+    },
+    {
+      "epoch": 1.6123081692760812,
+      "grad_norm": 0.4631887376308441,
+      "learning_rate": 5.217759016156333e-06,
+      "loss": 0.3494,
+      "step": 3467
+    },
+    {
+      "epoch": 1.6127732134552781,
+      "grad_norm": 0.4094836711883545,
+      "learning_rate": 5.215055615048283e-06,
+      "loss": 0.3841,
+      "step": 3468
+    },
+    {
+      "epoch": 1.6132382576344753,
+      "grad_norm": 0.36185774207115173,
+      "learning_rate": 5.2123521509539375e-06,
+      "loss": 0.3246,
+      "step": 3469
+    },
+    {
+      "epoch": 1.6137033018136724,
+      "grad_norm": 0.46863463521003723,
+      "learning_rate": 5.209648624665095e-06,
+      "loss": 0.3801,
+      "step": 3470
+    },
+    {
+      "epoch": 1.6141683459928693,
+      "grad_norm": 0.4251573085784912,
+      "learning_rate": 5.206945036973577e-06,
+      "loss": 0.3449,
+      "step": 3471
+    },
+    {
+      "epoch": 1.6146333901720662,
+      "grad_norm": 0.4120125472545624,
+      "learning_rate": 5.2042413886712176e-06,
+      "loss": 0.3312,
+      "step": 3472
+    },
+    {
+      "epoch": 1.6150984343512633,
+      "grad_norm": 0.41511979699134827,
+      "learning_rate": 5.201537680549874e-06,
+      "loss": 0.3495,
+      "step": 3473
+    },
+    {
+      "epoch": 1.6155634785304605,
+      "grad_norm": 0.43771082162857056,
+      "learning_rate": 5.198833913401418e-06,
+      "loss": 0.3843,
+      "step": 3474
+    },
+    {
+      "epoch": 1.6160285227096574,
+      "grad_norm": 0.39838284254074097,
+      "learning_rate": 5.196130088017737e-06,
+      "loss": 0.308,
+      "step": 3475
+    },
+    {
+      "epoch": 1.6164935668888545,
+      "grad_norm": 0.4643276631832123,
+      "learning_rate": 5.19342620519074e-06,
+      "loss": 0.351,
+      "step": 3476
+    },
+    {
+      "epoch": 1.6169586110680516,
+      "grad_norm": 0.34968698024749756,
+      "learning_rate": 5.190722265712349e-06,
+      "loss": 0.3099,
+      "step": 3477
+    },
+    {
+      "epoch": 1.6174236552472485,
+      "grad_norm": 0.40132972598075867,
+      "learning_rate": 5.188018270374504e-06,
+      "loss": 0.3358,
+      "step": 3478
+    },
+    {
+      "epoch": 1.6178886994264454,
+      "grad_norm": 0.3743666410446167,
+      "learning_rate": 5.185314219969163e-06,
+      "loss": 0.3669,
+      "step": 3479
+    },
+    {
+      "epoch": 1.6183537436056425,
+      "grad_norm": 0.3797807991504669,
+      "learning_rate": 5.182610115288296e-06,
+      "loss": 0.3471,
+      "step": 3480
+    },
+    {
+      "epoch": 1.6188187877848397,
+      "grad_norm": 0.43480437994003296,
+      "learning_rate": 5.179905957123891e-06,
+      "loss": 0.3913,
+      "step": 3481
+    },
+    {
+      "epoch": 1.6192838319640366,
+      "grad_norm": 0.3873792886734009,
+      "learning_rate": 5.177201746267955e-06,
+      "loss": 0.354,
+      "step": 3482
+    },
+    {
+      "epoch": 1.6197488761432335,
+      "grad_norm": 0.38425004482269287,
+      "learning_rate": 5.174497483512506e-06,
+      "loss": 0.3294,
+      "step": 3483
+    },
+    {
+      "epoch": 1.6202139203224306,
+      "grad_norm": 0.4359165132045746,
+      "learning_rate": 5.171793169649578e-06,
+      "loss": 0.3585,
+      "step": 3484
+    },
+    {
+      "epoch": 1.6206789645016277,
+      "grad_norm": 0.3433784246444702,
+      "learning_rate": 5.1690888054712215e-06,
+      "loss": 0.3329,
+      "step": 3485
+    },
+    {
+      "epoch": 1.6211440086808246,
+      "grad_norm": 0.4633955657482147,
+      "learning_rate": 5.1663843917695e-06,
+      "loss": 0.3745,
+      "step": 3486
+    },
+    {
+      "epoch": 1.6216090528600215,
+      "grad_norm": 0.44534632563591003,
+      "learning_rate": 5.163679929336491e-06,
+      "loss": 0.3692,
+      "step": 3487
+    },
+    {
+      "epoch": 1.6220740970392187,
+      "grad_norm": 0.3649267852306366,
+      "learning_rate": 5.160975418964293e-06,
+      "loss": 0.362,
+      "step": 3488
+    },
+    {
+      "epoch": 1.6225391412184158,
+      "grad_norm": 0.3883490264415741,
+      "learning_rate": 5.158270861445007e-06,
+      "loss": 0.3561,
+      "step": 3489
+    },
+    {
+      "epoch": 1.6230041853976127,
+      "grad_norm": 0.3937935531139374,
+      "learning_rate": 5.155566257570758e-06,
+      "loss": 0.32,
+      "step": 3490
+    },
+    {
+      "epoch": 1.6234692295768098,
+      "grad_norm": 0.36953264474868774,
+      "learning_rate": 5.152861608133678e-06,
+      "loss": 0.362,
+      "step": 3491
+    },
+    {
+      "epoch": 1.623934273756007,
+      "grad_norm": 0.49353107810020447,
+      "learning_rate": 5.150156913925916e-06,
+      "loss": 0.3646,
+      "step": 3492
+    },
+    {
+      "epoch": 1.6243993179352039,
+      "grad_norm": 0.393684446811676,
+      "learning_rate": 5.147452175739633e-06,
+      "loss": 0.3571,
+      "step": 3493
+    },
+    {
+      "epoch": 1.6248643621144008,
+      "grad_norm": 0.369238942861557,
+      "learning_rate": 5.144747394367002e-06,
+      "loss": 0.3446,
+      "step": 3494
+    },
+    {
+      "epoch": 1.625329406293598,
+      "grad_norm": 0.3968108296394348,
+      "learning_rate": 5.142042570600212e-06,
+      "loss": 0.3458,
+      "step": 3495
+    },
+    {
+      "epoch": 1.625794450472795,
+      "grad_norm": 0.44049766659736633,
+      "learning_rate": 5.139337705231459e-06,
+      "loss": 0.3774,
+      "step": 3496
+    },
+    {
+      "epoch": 1.626259494651992,
+      "grad_norm": 0.37539348006248474,
+      "learning_rate": 5.136632799052957e-06,
+      "loss": 0.3373,
+      "step": 3497
+    },
+    {
+      "epoch": 1.6267245388311888,
+      "grad_norm": 0.4646461009979248,
+      "learning_rate": 5.133927852856927e-06,
+      "loss": 0.352,
+      "step": 3498
+    },
+    {
+      "epoch": 1.627189583010386,
+      "grad_norm": 0.4138670861721039,
+      "learning_rate": 5.131222867435602e-06,
+      "loss": 0.3393,
+      "step": 3499
+    },
+    {
+      "epoch": 1.627654627189583,
+      "grad_norm": 0.40055084228515625,
+      "learning_rate": 5.128517843581233e-06,
+      "loss": 0.3605,
+      "step": 3500
+    },
+    {
+      "epoch": 1.62811967136878,
+      "grad_norm": 0.4990515112876892,
+      "learning_rate": 5.125812782086075e-06,
+      "loss": 0.338,
+      "step": 3501
+    },
+    {
+      "epoch": 1.628584715547977,
+      "grad_norm": 0.49589017033576965,
+      "learning_rate": 5.123107683742397e-06,
+      "loss": 0.3181,
+      "step": 3502
+    },
+    {
+      "epoch": 1.629049759727174,
+      "grad_norm": 0.4138791263103485,
+      "learning_rate": 5.1204025493424766e-06,
+      "loss": 0.3751,
+      "step": 3503
+    },
+    {
+      "epoch": 1.6295148039063712,
+      "grad_norm": 0.38624855875968933,
+      "learning_rate": 5.117697379678606e-06,
+      "loss": 0.3404,
+      "step": 3504
+    },
+    {
+      "epoch": 1.629979848085568,
+      "grad_norm": 0.5305654406547546,
+      "learning_rate": 5.114992175543084e-06,
+      "loss": 0.3786,
+      "step": 3505
+    },
+    {
+      "epoch": 1.6304448922647652,
+      "grad_norm": 0.4678351581096649,
+      "learning_rate": 5.112286937728223e-06,
+      "loss": 0.3552,
+      "step": 3506
+    },
+    {
+      "epoch": 1.6309099364439623,
+      "grad_norm": 0.4201079308986664,
+      "learning_rate": 5.109581667026341e-06,
+      "loss": 0.388,
+      "step": 3507
+    },
+    {
+      "epoch": 1.6313749806231592,
+      "grad_norm": 0.39116141200065613,
+      "learning_rate": 5.106876364229768e-06,
+      "loss": 0.338,
+      "step": 3508
+    },
+    {
+      "epoch": 1.6318400248023561,
+      "grad_norm": 0.3895026743412018,
+      "learning_rate": 5.104171030130846e-06,
+      "loss": 0.3374,
+      "step": 3509
+    },
+    {
+      "epoch": 1.6323050689815533,
+      "grad_norm": 0.3727634847164154,
+      "learning_rate": 5.101465665521919e-06,
+      "loss": 0.372,
+      "step": 3510
+    },
+    {
+      "epoch": 1.6327701131607504,
+      "grad_norm": 0.3340446352958679,
+      "learning_rate": 5.098760271195348e-06,
+      "loss": 0.3047,
+      "step": 3511
+    },
+    {
+      "epoch": 1.6332351573399473,
+      "grad_norm": 0.42071983218193054,
+      "learning_rate": 5.096054847943498e-06,
+      "loss": 0.3951,
+      "step": 3512
+    },
+    {
+      "epoch": 1.6337002015191442,
+      "grad_norm": 0.3533824682235718,
+      "learning_rate": 5.093349396558744e-06,
+      "loss": 0.3344,
+      "step": 3513
+    },
+    {
+      "epoch": 1.6341652456983413,
+      "grad_norm": 0.44178301095962524,
+      "learning_rate": 5.090643917833465e-06,
+      "loss": 0.3444,
+      "step": 3514
+    },
+    {
+      "epoch": 1.6346302898775384,
+      "grad_norm": 0.42436400055885315,
+      "learning_rate": 5.0879384125600565e-06,
+      "loss": 0.3877,
+      "step": 3515
+    },
+    {
+      "epoch": 1.6350953340567354,
+      "grad_norm": 0.4048612415790558,
+      "learning_rate": 5.085232881530916e-06,
+      "loss": 0.3495,
+      "step": 3516
+    },
+    {
+      "epoch": 1.6355603782359323,
+      "grad_norm": 0.3634321689605713,
+      "learning_rate": 5.082527325538449e-06,
+      "loss": 0.3845,
+      "step": 3517
+    },
+    {
+      "epoch": 1.6360254224151294,
+      "grad_norm": 0.3633408546447754,
+      "learning_rate": 5.0798217453750675e-06,
+      "loss": 0.3352,
+      "step": 3518
+    },
+    {
+      "epoch": 1.6364904665943265,
+      "grad_norm": 0.4096944034099579,
+      "learning_rate": 5.077116141833195e-06,
+      "loss": 0.3372,
+      "step": 3519
+    },
+    {
+      "epoch": 1.6369555107735234,
+      "grad_norm": 0.37343043088912964,
+      "learning_rate": 5.074410515705256e-06,
+      "loss": 0.345,
+      "step": 3520
+    },
+    {
+      "epoch": 1.6374205549527205,
+      "grad_norm": 0.4189586639404297,
+      "learning_rate": 5.071704867783684e-06,
+      "loss": 0.3634,
+      "step": 3521
+    },
+    {
+      "epoch": 1.6378855991319177,
+      "grad_norm": 0.358604371547699,
+      "learning_rate": 5.068999198860924e-06,
+      "loss": 0.3367,
+      "step": 3522
+    },
+    {
+      "epoch": 1.6383506433111146,
+      "grad_norm": 0.3955291509628296,
+      "learning_rate": 5.066293509729418e-06,
+      "loss": 0.3635,
+      "step": 3523
+    },
+    {
+      "epoch": 1.6388156874903115,
+      "grad_norm": 0.3491485118865967,
+      "learning_rate": 5.063587801181621e-06,
+      "loss": 0.3198,
+      "step": 3524
+    },
+    {
+      "epoch": 1.6392807316695086,
+      "grad_norm": 0.39579418301582336,
+      "learning_rate": 5.060882074009988e-06,
+      "loss": 0.3553,
+      "step": 3525
+    },
+    {
+      "epoch": 1.6397457758487057,
+      "grad_norm": 0.404255747795105,
+      "learning_rate": 5.0581763290069865e-06,
+      "loss": 0.3354,
+      "step": 3526
+    },
+    {
+      "epoch": 1.6402108200279026,
+      "grad_norm": 0.38574931025505066,
+      "learning_rate": 5.055470566965082e-06,
+      "loss": 0.3344,
+      "step": 3527
+    },
+    {
+      "epoch": 1.6406758642070995,
+      "grad_norm": 0.3492719829082489,
+      "learning_rate": 5.052764788676749e-06,
+      "loss": 0.3031,
+      "step": 3528
+    },
+    {
+      "epoch": 1.6411409083862967,
+      "grad_norm": 0.4036017954349518,
+      "learning_rate": 5.050058994934467e-06,
+      "loss": 0.3784,
+      "step": 3529
+    },
+    {
+      "epoch": 1.6416059525654938,
+      "grad_norm": 0.4062303900718689,
+      "learning_rate": 5.047353186530718e-06,
+      "loss": 0.38,
+      "step": 3530
+    },
+    {
+      "epoch": 1.6420709967446907,
+      "grad_norm": 0.3702806532382965,
+      "learning_rate": 5.04464736425799e-06,
+      "loss": 0.3046,
+      "step": 3531
+    },
+    {
+      "epoch": 1.6425360409238876,
+      "grad_norm": 0.40208354592323303,
+      "learning_rate": 5.0419415289087755e-06,
+      "loss": 0.3609,
+      "step": 3532
+    },
+    {
+      "epoch": 1.643001085103085,
+      "grad_norm": 0.35900288820266724,
+      "learning_rate": 5.039235681275568e-06,
+      "loss": 0.305,
+      "step": 3533
+    },
+    {
+      "epoch": 1.6434661292822819,
+      "grad_norm": 0.4107772707939148,
+      "learning_rate": 5.036529822150865e-06,
+      "loss": 0.367,
+      "step": 3534
+    },
+    {
+      "epoch": 1.6439311734614788,
+      "grad_norm": 0.36534854769706726,
+      "learning_rate": 5.033823952327173e-06,
+      "loss": 0.3382,
+      "step": 3535
+    },
+    {
+      "epoch": 1.644396217640676,
+      "grad_norm": 0.3700648844242096,
+      "learning_rate": 5.031118072596993e-06,
+      "loss": 0.3338,
+      "step": 3536
+    },
+    {
+      "epoch": 1.644861261819873,
+      "grad_norm": 0.3885965943336487,
+      "learning_rate": 5.028412183752835e-06,
+      "loss": 0.3292,
+      "step": 3537
+    },
+    {
+      "epoch": 1.64532630599907,
+      "grad_norm": 0.49532386660575867,
+      "learning_rate": 5.025706286587211e-06,
+      "loss": 0.3665,
+      "step": 3538
+    },
+    {
+      "epoch": 1.6457913501782668,
+      "grad_norm": 0.42976298928260803,
+      "learning_rate": 5.023000381892633e-06,
+      "loss": 0.3094,
+      "step": 3539
+    },
+    {
+      "epoch": 1.646256394357464,
+      "grad_norm": 0.4339199662208557,
+      "learning_rate": 5.020294470461615e-06,
+      "loss": 0.3615,
+      "step": 3540
+    },
+    {
+      "epoch": 1.646721438536661,
+      "grad_norm": 0.4408279061317444,
+      "learning_rate": 5.017588553086677e-06,
+      "loss": 0.3422,
+      "step": 3541
+    },
+    {
+      "epoch": 1.647186482715858,
+      "grad_norm": 0.4677359163761139,
+      "learning_rate": 5.014882630560339e-06,
+      "loss": 0.3658,
+      "step": 3542
+    },
+    {
+      "epoch": 1.647651526895055,
+      "grad_norm": 0.416978657245636,
+      "learning_rate": 5.01217670367512e-06,
+      "loss": 0.3575,
+      "step": 3543
+    },
+    {
+      "epoch": 1.648116571074252,
+      "grad_norm": 0.42392852902412415,
+      "learning_rate": 5.009470773223541e-06,
+      "loss": 0.3341,
+      "step": 3544
+    },
+    {
+      "epoch": 1.6485816152534492,
+      "grad_norm": 0.48418161273002625,
+      "learning_rate": 5.006764839998128e-06,
+      "loss": 0.3392,
+      "step": 3545
+    },
+    {
+      "epoch": 1.649046659432646,
+      "grad_norm": 0.4479508697986603,
+      "learning_rate": 5.004058904791402e-06,
+      "loss": 0.3764,
+      "step": 3546
+    },
+    {
+      "epoch": 1.649511703611843,
+      "grad_norm": 0.43179401755332947,
+      "learning_rate": 5.0013529683958885e-06,
+      "loss": 0.3507,
+      "step": 3547
+    },
+    {
+      "epoch": 1.6499767477910403,
+      "grad_norm": 0.3624599874019623,
+      "learning_rate": 4.998647031604114e-06,
+      "loss": 0.3289,
+      "step": 3548
+    },
+    {
+      "epoch": 1.6504417919702372,
+      "grad_norm": 0.46299293637275696,
+      "learning_rate": 4.995941095208599e-06,
+      "loss": 0.3391,
+      "step": 3549
+    },
+    {
+      "epoch": 1.6509068361494341,
+      "grad_norm": 0.41883838176727295,
+      "learning_rate": 4.993235160001874e-06,
+      "loss": 0.3454,
+      "step": 3550
+    },
+    {
+      "epoch": 1.6513718803286312,
+      "grad_norm": 0.40758463740348816,
+      "learning_rate": 4.99052922677646e-06,
+      "loss": 0.3604,
+      "step": 3551
+    },
+    {
+      "epoch": 1.6518369245078284,
+      "grad_norm": 0.4133349359035492,
+      "learning_rate": 4.987823296324882e-06,
+      "loss": 0.3388,
+      "step": 3552
+    },
+    {
+      "epoch": 1.6523019686870253,
+      "grad_norm": 0.4033692479133606,
+      "learning_rate": 4.985117369439661e-06,
+      "loss": 0.3454,
+      "step": 3553
+    },
+    {
+      "epoch": 1.6527670128662222,
+      "grad_norm": 0.38803398609161377,
+      "learning_rate": 4.982411446913324e-06,
+      "loss": 0.3475,
+      "step": 3554
+    },
+    {
+      "epoch": 1.6532320570454193,
+      "grad_norm": 0.4113628566265106,
+      "learning_rate": 4.979705529538385e-06,
+      "loss": 0.337,
+      "step": 3555
+    },
+    {
+      "epoch": 1.6536971012246164,
+      "grad_norm": 0.4451198875904083,
+      "learning_rate": 4.976999618107369e-06,
+      "loss": 0.3973,
+      "step": 3556
+    },
+    {
+      "epoch": 1.6541621454038133,
+      "grad_norm": 0.36026668548583984,
+      "learning_rate": 4.974293713412791e-06,
+      "loss": 0.3212,
+      "step": 3557
+    },
+    {
+      "epoch": 1.6546271895830102,
+      "grad_norm": 0.38637468218803406,
+      "learning_rate": 4.971587816247166e-06,
+      "loss": 0.3623,
+      "step": 3558
+    },
+    {
+      "epoch": 1.6550922337622074,
+      "grad_norm": 0.3913023769855499,
+      "learning_rate": 4.9688819274030074e-06,
+      "loss": 0.3304,
+      "step": 3559
+    },
+    {
+      "epoch": 1.6555572779414045,
+      "grad_norm": 0.3982096016407013,
+      "learning_rate": 4.96617604767283e-06,
+      "loss": 0.3727,
+      "step": 3560
+    },
+    {
+      "epoch": 1.6560223221206014,
+      "grad_norm": 0.3954141139984131,
+      "learning_rate": 4.963470177849135e-06,
+      "loss": 0.328,
+      "step": 3561
+    },
+    {
+      "epoch": 1.6564873662997985,
+      "grad_norm": 0.40723100304603577,
+      "learning_rate": 4.960764318724434e-06,
+      "loss": 0.344,
+      "step": 3562
+    },
+    {
+      "epoch": 1.6569524104789957,
+      "grad_norm": 0.35724228620529175,
+      "learning_rate": 4.958058471091225e-06,
+      "loss": 0.3379,
+      "step": 3563
+    },
+    {
+      "epoch": 1.6574174546581926,
+      "grad_norm": 0.3636278808116913,
+      "learning_rate": 4.9553526357420104e-06,
+      "loss": 0.3522,
+      "step": 3564
+    },
+    {
+      "epoch": 1.6578824988373895,
+      "grad_norm": 0.5159657597541809,
+      "learning_rate": 4.952646813469282e-06,
+      "loss": 0.3819,
+      "step": 3565
+    },
+    {
+      "epoch": 1.6583475430165866,
+      "grad_norm": 0.40651264786720276,
+      "learning_rate": 4.949941005065534e-06,
+      "loss": 0.3147,
+      "step": 3566
+    },
+    {
+      "epoch": 1.6588125871957837,
+      "grad_norm": 0.4187847673892975,
+      "learning_rate": 4.947235211323253e-06,
+      "loss": 0.3342,
+      "step": 3567
+    },
+    {
+      "epoch": 1.6592776313749806,
+      "grad_norm": 0.43306782841682434,
+      "learning_rate": 4.944529433034919e-06,
+      "loss": 0.3691,
+      "step": 3568
+    },
+    {
+      "epoch": 1.6597426755541775,
+      "grad_norm": 0.4201146960258484,
+      "learning_rate": 4.941823670993016e-06,
+      "loss": 0.3509,
+      "step": 3569
+    },
+    {
+      "epoch": 1.6602077197333747,
+      "grad_norm": 0.4412671625614166,
+      "learning_rate": 4.939117925990013e-06,
+      "loss": 0.3418,
+      "step": 3570
+    },
+    {
+      "epoch": 1.6606727639125718,
+      "grad_norm": 0.3967730700969696,
+      "learning_rate": 4.936412198818382e-06,
+      "loss": 0.3747,
+      "step": 3571
+    },
+    {
+      "epoch": 1.6611378080917687,
+      "grad_norm": 0.3809349834918976,
+      "learning_rate": 4.933706490270583e-06,
+      "loss": 0.3397,
+      "step": 3572
+    },
+    {
+      "epoch": 1.6616028522709656,
+      "grad_norm": 0.3735617697238922,
+      "learning_rate": 4.9310008011390774e-06,
+      "loss": 0.3398,
+      "step": 3573
+    },
+    {
+      "epoch": 1.6620678964501627,
+      "grad_norm": 0.44322773814201355,
+      "learning_rate": 4.9282951322163166e-06,
+      "loss": 0.3258,
+      "step": 3574
+    },
+    {
+      "epoch": 1.6625329406293599,
+      "grad_norm": 0.438362181186676,
+      "learning_rate": 4.925589484294747e-06,
+      "loss": 0.3926,
+      "step": 3575
+    },
+    {
+      "epoch": 1.6629979848085568,
+      "grad_norm": 0.36551767587661743,
+      "learning_rate": 4.922883858166807e-06,
+      "loss": 0.3317,
+      "step": 3576
+    },
+    {
+      "epoch": 1.6634630289877539,
+      "grad_norm": 0.39981698989868164,
+      "learning_rate": 4.920178254624935e-06,
+      "loss": 0.3266,
+      "step": 3577
+    },
+    {
+      "epoch": 1.663928073166951,
+      "grad_norm": 0.4227448105812073,
+      "learning_rate": 4.917472674461553e-06,
+      "loss": 0.3333,
+      "step": 3578
+    },
+    {
+      "epoch": 1.664393117346148,
+      "grad_norm": 0.4146755337715149,
+      "learning_rate": 4.9147671184690855e-06,
+      "loss": 0.3662,
+      "step": 3579
+    },
+    {
+      "epoch": 1.6648581615253448,
+      "grad_norm": 0.4139578342437744,
+      "learning_rate": 4.912061587439944e-06,
+      "loss": 0.3087,
+      "step": 3580
+    },
+    {
+      "epoch": 1.665323205704542,
+      "grad_norm": 0.4043808579444885,
+      "learning_rate": 4.9093560821665365e-06,
+      "loss": 0.3842,
+      "step": 3581
+    },
+    {
+      "epoch": 1.665788249883739,
+      "grad_norm": 0.3665468990802765,
+      "learning_rate": 4.906650603441259e-06,
+      "loss": 0.3199,
+      "step": 3582
+    },
+    {
+      "epoch": 1.666253294062936,
+      "grad_norm": 0.46738430857658386,
+      "learning_rate": 4.903945152056505e-06,
+      "loss": 0.374,
+      "step": 3583
+    },
+    {
+      "epoch": 1.666718338242133,
+      "grad_norm": 0.4571022689342499,
+      "learning_rate": 4.901239728804653e-06,
+      "loss": 0.3221,
+      "step": 3584
+    },
+    {
+      "epoch": 1.66718338242133,
+      "grad_norm": 0.41317233443260193,
+      "learning_rate": 4.8985343344780815e-06,
+      "loss": 0.3583,
+      "step": 3585
+    },
+    {
+      "epoch": 1.6676484266005271,
+      "grad_norm": 0.36751988530158997,
+      "learning_rate": 4.895828969869157e-06,
+      "loss": 0.3409,
+      "step": 3586
+    },
+    {
+      "epoch": 1.668113470779724,
+      "grad_norm": 0.44614648818969727,
+      "learning_rate": 4.8931236357702326e-06,
+      "loss": 0.3393,
+      "step": 3587
+    },
+    {
+      "epoch": 1.668578514958921,
+      "grad_norm": 0.3438377380371094,
+      "learning_rate": 4.89041833297366e-06,
+      "loss": 0.3349,
+      "step": 3588
+    },
+    {
+      "epoch": 1.669043559138118,
+      "grad_norm": 0.393669992685318,
+      "learning_rate": 4.88771306227178e-06,
+      "loss": 0.3534,
+      "step": 3589
+    },
+    {
+      "epoch": 1.6695086033173152,
+      "grad_norm": 0.40612417459487915,
+      "learning_rate": 4.885007824456917e-06,
+      "loss": 0.3435,
+      "step": 3590
+    },
+    {
+      "epoch": 1.6699736474965121,
+      "grad_norm": 0.3694112300872803,
+      "learning_rate": 4.882302620321395e-06,
+      "loss": 0.3275,
+      "step": 3591
+    },
+    {
+      "epoch": 1.6704386916757092,
+      "grad_norm": 0.3994767665863037,
+      "learning_rate": 4.879597450657525e-06,
+      "loss": 0.3593,
+      "step": 3592
+    },
+    {
+      "epoch": 1.6709037358549064,
+      "grad_norm": 0.3501642346382141,
+      "learning_rate": 4.876892316257605e-06,
+      "loss": 0.3165,
+      "step": 3593
+    },
+    {
+      "epoch": 1.6713687800341033,
+      "grad_norm": 0.3949240744113922,
+      "learning_rate": 4.874187217913926e-06,
+      "loss": 0.3407,
+      "step": 3594
+    },
+    {
+      "epoch": 1.6718338242133002,
+      "grad_norm": 0.4457712471485138,
+      "learning_rate": 4.871482156418769e-06,
+      "loss": 0.3615,
+      "step": 3595
+    },
+    {
+      "epoch": 1.6722988683924973,
+      "grad_norm": 0.36873969435691833,
+      "learning_rate": 4.868777132564398e-06,
+      "loss": 0.3671,
+      "step": 3596
+    },
+    {
+      "epoch": 1.6727639125716944,
+      "grad_norm": 0.35475531220436096,
+      "learning_rate": 4.866072147143075e-06,
+      "loss": 0.3593,
+      "step": 3597
+    },
+    {
+      "epoch": 1.6732289567508913,
+      "grad_norm": 0.3828467130661011,
+      "learning_rate": 4.863367200947044e-06,
+      "loss": 0.344,
+      "step": 3598
+    },
+    {
+      "epoch": 1.6736940009300882,
+      "grad_norm": 0.44382408261299133,
+      "learning_rate": 4.8606622947685415e-06,
+      "loss": 0.3121,
+      "step": 3599
+    },
+    {
+      "epoch": 1.6741590451092854,
+      "grad_norm": 0.4533607065677643,
+      "learning_rate": 4.857957429399788e-06,
+      "loss": 0.3396,
+      "step": 3600
+    },
+    {
+      "epoch": 1.6746240892884825,
+      "grad_norm": 0.4441719651222229,
+      "learning_rate": 4.855252605632999e-06,
+      "loss": 0.3897,
+      "step": 3601
+    },
+    {
+      "epoch": 1.6750891334676794,
+      "grad_norm": 0.36047327518463135,
+      "learning_rate": 4.852547824260369e-06,
+      "loss": 0.3307,
+      "step": 3602
+    },
+    {
+      "epoch": 1.6755541776468763,
+      "grad_norm": 0.39020687341690063,
+      "learning_rate": 4.849843086074085e-06,
+      "loss": 0.3056,
+      "step": 3603
+    },
+    {
+      "epoch": 1.6760192218260734,
+      "grad_norm": 0.5388873219490051,
+      "learning_rate": 4.847138391866325e-06,
+      "loss": 0.3605,
+      "step": 3604
+    },
+    {
+      "epoch": 1.6764842660052706,
+      "grad_norm": 0.4391537010669708,
+      "learning_rate": 4.8444337424292445e-06,
+      "loss": 0.3831,
+      "step": 3605
+    },
+    {
+      "epoch": 1.6769493101844675,
+      "grad_norm": 0.3435923457145691,
+      "learning_rate": 4.841729138554996e-06,
+      "loss": 0.3155,
+      "step": 3606
+    },
+    {
+      "epoch": 1.6774143543636646,
+      "grad_norm": 0.3751065135002136,
+      "learning_rate": 4.839024581035709e-06,
+      "loss": 0.3387,
+      "step": 3607
+    },
+    {
+      "epoch": 1.6778793985428617,
+      "grad_norm": 0.47324469685554504,
+      "learning_rate": 4.83632007066351e-06,
+      "loss": 0.3561,
+      "step": 3608
+    },
+    {
+      "epoch": 1.6783444427220586,
+      "grad_norm": 0.4418434202671051,
+      "learning_rate": 4.833615608230501e-06,
+      "loss": 0.3445,
+      "step": 3609
+    },
+    {
+      "epoch": 1.6788094869012555,
+      "grad_norm": 0.4009036719799042,
+      "learning_rate": 4.830911194528781e-06,
+      "loss": 0.3336,
+      "step": 3610
+    },
+    {
+      "epoch": 1.6792745310804527,
+      "grad_norm": 0.42753738164901733,
+      "learning_rate": 4.828206830350423e-06,
+      "loss": 0.3613,
+      "step": 3611
+    },
+    {
+      "epoch": 1.6797395752596498,
+      "grad_norm": 0.3864332139492035,
+      "learning_rate": 4.825502516487497e-06,
+      "loss": 0.3251,
+      "step": 3612
+    },
+    {
+      "epoch": 1.6802046194388467,
+      "grad_norm": 0.38312116265296936,
+      "learning_rate": 4.822798253732046e-06,
+      "loss": 0.3146,
+      "step": 3613
+    },
+    {
+      "epoch": 1.6806696636180436,
+      "grad_norm": 0.4273233413696289,
+      "learning_rate": 4.82009404287611e-06,
+      "loss": 0.34,
+      "step": 3614
+    },
+    {
+      "epoch": 1.6811347077972407,
+      "grad_norm": 0.4202836751937866,
+      "learning_rate": 4.817389884711706e-06,
+      "loss": 0.3703,
+      "step": 3615
+    },
+    {
+      "epoch": 1.6815997519764379,
+      "grad_norm": 0.4047132730484009,
+      "learning_rate": 4.81468578003084e-06,
+      "loss": 0.3389,
+      "step": 3616
+    },
+    {
+      "epoch": 1.6820647961556348,
+      "grad_norm": 0.4253717362880707,
+      "learning_rate": 4.8119817296254965e-06,
+      "loss": 0.3647,
+      "step": 3617
+    },
+    {
+      "epoch": 1.6825298403348317,
+      "grad_norm": 0.4105294644832611,
+      "learning_rate": 4.809277734287654e-06,
+      "loss": 0.3298,
+      "step": 3618
+    },
+    {
+      "epoch": 1.6829948845140288,
+      "grad_norm": 0.367784321308136,
+      "learning_rate": 4.8065737948092615e-06,
+      "loss": 0.3189,
+      "step": 3619
+    },
+    {
+      "epoch": 1.683459928693226,
+      "grad_norm": 0.42748555541038513,
+      "learning_rate": 4.803869911982264e-06,
+      "loss": 0.3723,
+      "step": 3620
+    },
+    {
+      "epoch": 1.6839249728724228,
+      "grad_norm": 0.44452059268951416,
+      "learning_rate": 4.801166086598584e-06,
+      "loss": 0.324,
+      "step": 3621
+    },
+    {
+      "epoch": 1.68439001705162,
+      "grad_norm": 0.4104439914226532,
+      "learning_rate": 4.798462319450127e-06,
+      "loss": 0.316,
+      "step": 3622
+    },
+    {
+      "epoch": 1.684855061230817,
+      "grad_norm": 0.37423205375671387,
+      "learning_rate": 4.795758611328782e-06,
+      "loss": 0.3399,
+      "step": 3623
+    },
+    {
+      "epoch": 1.685320105410014,
+      "grad_norm": 0.39451467990875244,
+      "learning_rate": 4.793054963026425e-06,
+      "loss": 0.3372,
+      "step": 3624
+    },
+    {
+      "epoch": 1.6857851495892109,
+      "grad_norm": 0.42838355898857117,
+      "learning_rate": 4.790351375334906e-06,
+      "loss": 0.358,
+      "step": 3625
+    },
+    {
+      "epoch": 1.686250193768408,
+      "grad_norm": 0.4108628034591675,
+      "learning_rate": 4.787647849046064e-06,
+      "loss": 0.31,
+      "step": 3626
+    },
+    {
+      "epoch": 1.6867152379476051,
+      "grad_norm": 0.481794536113739,
+      "learning_rate": 4.784944384951718e-06,
+      "loss": 0.3697,
+      "step": 3627
+    },
+    {
+      "epoch": 1.687180282126802,
+      "grad_norm": 0.4808673858642578,
+      "learning_rate": 4.782240983843668e-06,
+      "loss": 0.3575,
+      "step": 3628
+    },
+    {
+      "epoch": 1.687645326305999,
+      "grad_norm": 0.40240907669067383,
+      "learning_rate": 4.779537646513697e-06,
+      "loss": 0.3278,
+      "step": 3629
+    },
+    {
+      "epoch": 1.688110370485196,
+      "grad_norm": 0.3663456439971924,
+      "learning_rate": 4.7768343737535694e-06,
+      "loss": 0.299,
+      "step": 3630
+    },
+    {
+      "epoch": 1.6885754146643932,
+      "grad_norm": 0.470187246799469,
+      "learning_rate": 4.774131166355027e-06,
+      "loss": 0.4006,
+      "step": 3631
+    },
+    {
+      "epoch": 1.68904045884359,
+      "grad_norm": 0.4300827383995056,
+      "learning_rate": 4.771428025109798e-06,
+      "loss": 0.3551,
+      "step": 3632
+    },
+    {
+      "epoch": 1.689505503022787,
+      "grad_norm": 0.45297130942344666,
+      "learning_rate": 4.768724950809587e-06,
+      "loss": 0.3158,
+      "step": 3633
+    },
+    {
+      "epoch": 1.6899705472019841,
+      "grad_norm": 0.39991265535354614,
+      "learning_rate": 4.766021944246082e-06,
+      "loss": 0.3333,
+      "step": 3634
+    },
+    {
+      "epoch": 1.6904355913811813,
+      "grad_norm": 0.42256665229797363,
+      "learning_rate": 4.763319006210949e-06,
+      "loss": 0.3593,
+      "step": 3635
+    },
+    {
+      "epoch": 1.6909006355603782,
+      "grad_norm": 0.4344233274459839,
+      "learning_rate": 4.7606161374958355e-06,
+      "loss": 0.402,
+      "step": 3636
+    },
+    {
+      "epoch": 1.6913656797395753,
+      "grad_norm": 0.39570462703704834,
+      "learning_rate": 4.757913338892365e-06,
+      "loss": 0.3526,
+      "step": 3637
+    },
+    {
+      "epoch": 1.6918307239187724,
+      "grad_norm": 0.38186073303222656,
+      "learning_rate": 4.755210611192146e-06,
+      "loss": 0.328,
+      "step": 3638
+    },
+    {
+      "epoch": 1.6922957680979693,
+      "grad_norm": 0.3946187198162079,
+      "learning_rate": 4.752507955186765e-06,
+      "loss": 0.3226,
+      "step": 3639
+    },
+    {
+      "epoch": 1.6927608122771662,
+      "grad_norm": 0.4827534258365631,
+      "learning_rate": 4.749805371667781e-06,
+      "loss": 0.3992,
+      "step": 3640
+    },
+    {
+      "epoch": 1.6932258564563634,
+      "grad_norm": 0.3777780830860138,
+      "learning_rate": 4.747102861426742e-06,
+      "loss": 0.3196,
+      "step": 3641
+    },
+    {
+      "epoch": 1.6936909006355605,
+      "grad_norm": 0.4654094874858856,
+      "learning_rate": 4.744400425255165e-06,
+      "loss": 0.3779,
+      "step": 3642
+    },
+    {
+      "epoch": 1.6941559448147574,
+      "grad_norm": 0.3912859559059143,
+      "learning_rate": 4.741698063944553e-06,
+      "loss": 0.3197,
+      "step": 3643
+    },
+    {
+      "epoch": 1.6946209889939543,
+      "grad_norm": 0.40966254472732544,
+      "learning_rate": 4.7389957782863806e-06,
+      "loss": 0.3348,
+      "step": 3644
+    },
+    {
+      "epoch": 1.6950860331731514,
+      "grad_norm": 0.41277340054512024,
+      "learning_rate": 4.736293569072108e-06,
+      "loss": 0.3528,
+      "step": 3645
+    },
+    {
+      "epoch": 1.6955510773523486,
+      "grad_norm": 0.42663225531578064,
+      "learning_rate": 4.733591437093163e-06,
+      "loss": 0.3359,
+      "step": 3646
+    },
+    {
+      "epoch": 1.6960161215315455,
+      "grad_norm": 0.36344727873802185,
+      "learning_rate": 4.730889383140961e-06,
+      "loss": 0.35,
+      "step": 3647
+    },
+    {
+      "epoch": 1.6964811657107424,
+      "grad_norm": 0.41764384508132935,
+      "learning_rate": 4.7281874080068855e-06,
+      "loss": 0.375,
+      "step": 3648
+    },
+    {
+      "epoch": 1.6969462098899395,
+      "grad_norm": 0.4017370641231537,
+      "learning_rate": 4.725485512482304e-06,
+      "loss": 0.315,
+      "step": 3649
+    },
+    {
+      "epoch": 1.6974112540691366,
+      "grad_norm": 0.4225184917449951,
+      "learning_rate": 4.722783697358555e-06,
+      "loss": 0.3786,
+      "step": 3650
+    },
+    {
+      "epoch": 1.6978762982483335,
+      "grad_norm": 0.39845162630081177,
+      "learning_rate": 4.720081963426962e-06,
+      "loss": 0.3394,
+      "step": 3651
+    },
+    {
+      "epoch": 1.6983413424275307,
+      "grad_norm": 0.3981609642505646,
+      "learning_rate": 4.717380311478813e-06,
+      "loss": 0.3866,
+      "step": 3652
+    },
+    {
+      "epoch": 1.6988063866067278,
+      "grad_norm": 0.3535573184490204,
+      "learning_rate": 4.714678742305381e-06,
+      "loss": 0.3103,
+      "step": 3653
+    },
+    {
+      "epoch": 1.6992714307859247,
+      "grad_norm": 0.43783578276634216,
+      "learning_rate": 4.711977256697909e-06,
+      "loss": 0.3763,
+      "step": 3654
+    },
+    {
+      "epoch": 1.6997364749651216,
+      "grad_norm": 0.39808762073516846,
+      "learning_rate": 4.7092758554476215e-06,
+      "loss": 0.3485,
+      "step": 3655
+    },
+    {
+      "epoch": 1.7002015191443187,
+      "grad_norm": 0.3975062668323517,
+      "learning_rate": 4.706574539345712e-06,
+      "loss": 0.3592,
+      "step": 3656
+    },
+    {
+      "epoch": 1.7006665633235158,
+      "grad_norm": 0.3854736089706421,
+      "learning_rate": 4.703873309183357e-06,
+      "loss": 0.3236,
+      "step": 3657
+    },
+    {
+      "epoch": 1.7011316075027128,
+      "grad_norm": 0.3952430486679077,
+      "learning_rate": 4.7011721657516966e-06,
+      "loss": 0.363,
+      "step": 3658
+    },
+    {
+      "epoch": 1.7015966516819097,
+      "grad_norm": 0.47998958826065063,
+      "learning_rate": 4.698471109841858e-06,
+      "loss": 0.4107,
+      "step": 3659
+    },
+    {
+      "epoch": 1.7020616958611068,
+      "grad_norm": 0.37458357214927673,
+      "learning_rate": 4.695770142244931e-06,
+      "loss": 0.3453,
+      "step": 3660
+    },
+    {
+      "epoch": 1.702526740040304,
+      "grad_norm": 0.3739306330680847,
+      "learning_rate": 4.693069263751989e-06,
+      "loss": 0.3587,
+      "step": 3661
+    },
+    {
+      "epoch": 1.7029917842195008,
+      "grad_norm": 0.38612034916877747,
+      "learning_rate": 4.690368475154072e-06,
+      "loss": 0.3531,
+      "step": 3662
+    },
+    {
+      "epoch": 1.7034568283986977,
+      "grad_norm": 0.41020193696022034,
+      "learning_rate": 4.687667777242203e-06,
+      "loss": 0.3569,
+      "step": 3663
+    },
+    {
+      "epoch": 1.7039218725778948,
+      "grad_norm": 0.3760784864425659,
+      "learning_rate": 4.684967170807365e-06,
+      "loss": 0.3417,
+      "step": 3664
+    },
+    {
+      "epoch": 1.704386916757092,
+      "grad_norm": 0.3778047263622284,
+      "learning_rate": 4.682266656640528e-06,
+      "loss": 0.3226,
+      "step": 3665
+    },
+    {
+      "epoch": 1.7048519609362889,
+      "grad_norm": 0.4011727273464203,
+      "learning_rate": 4.679566235532625e-06,
+      "loss": 0.3394,
+      "step": 3666
+    },
+    {
+      "epoch": 1.705317005115486,
+      "grad_norm": 0.38582319021224976,
+      "learning_rate": 4.676865908274567e-06,
+      "loss": 0.3972,
+      "step": 3667
+    },
+    {
+      "epoch": 1.7057820492946831,
+      "grad_norm": 0.3527704179286957,
+      "learning_rate": 4.674165675657236e-06,
+      "loss": 0.3193,
+      "step": 3668
+    },
+    {
+      "epoch": 1.70624709347388,
+      "grad_norm": 0.4227658808231354,
+      "learning_rate": 4.671465538471487e-06,
+      "loss": 0.3571,
+      "step": 3669
+    },
+    {
+      "epoch": 1.706712137653077,
+      "grad_norm": 0.36377352476119995,
+      "learning_rate": 4.668765497508143e-06,
+      "loss": 0.3433,
+      "step": 3670
+    },
+    {
+      "epoch": 1.707177181832274,
+      "grad_norm": 0.3703608512878418,
+      "learning_rate": 4.666065553558007e-06,
+      "loss": 0.3152,
+      "step": 3671
+    },
+    {
+      "epoch": 1.7076422260114712,
+      "grad_norm": 0.39327993988990784,
+      "learning_rate": 4.663365707411845e-06,
+      "loss": 0.3686,
+      "step": 3672
+    },
+    {
+      "epoch": 1.708107270190668,
+      "grad_norm": 0.41823717951774597,
+      "learning_rate": 4.660665959860399e-06,
+      "loss": 0.3777,
+      "step": 3673
+    },
+    {
+      "epoch": 1.708572314369865,
+      "grad_norm": 0.3737585246562958,
+      "learning_rate": 4.657966311694383e-06,
+      "loss": 0.3365,
+      "step": 3674
+    },
+    {
+      "epoch": 1.7090373585490621,
+      "grad_norm": 0.43027764558792114,
+      "learning_rate": 4.655266763704476e-06,
+      "loss": 0.3414,
+      "step": 3675
+    },
+    {
+      "epoch": 1.7095024027282593,
+      "grad_norm": 0.376740962266922,
+      "learning_rate": 4.652567316681337e-06,
+      "loss": 0.3267,
+      "step": 3676
+    },
+    {
+      "epoch": 1.7099674469074562,
+      "grad_norm": 0.40848153829574585,
+      "learning_rate": 4.649867971415585e-06,
+      "loss": 0.361,
+      "step": 3677
+    },
+    {
+      "epoch": 1.710432491086653,
+      "grad_norm": 0.40334048867225647,
+      "learning_rate": 4.647168728697819e-06,
+      "loss": 0.3619,
+      "step": 3678
+    },
+    {
+      "epoch": 1.7108975352658504,
+      "grad_norm": 0.38001689314842224,
+      "learning_rate": 4.6444695893185994e-06,
+      "loss": 0.3197,
+      "step": 3679
+    },
+    {
+      "epoch": 1.7113625794450473,
+      "grad_norm": 0.4063551425933838,
+      "learning_rate": 4.641770554068465e-06,
+      "loss": 0.3519,
+      "step": 3680
+    },
+    {
+      "epoch": 1.7118276236242442,
+      "grad_norm": 0.46857863664627075,
+      "learning_rate": 4.639071623737913e-06,
+      "loss": 0.3636,
+      "step": 3681
+    },
+    {
+      "epoch": 1.7122926678034414,
+      "grad_norm": 0.4151492416858673,
+      "learning_rate": 4.636372799117424e-06,
+      "loss": 0.3372,
+      "step": 3682
+    },
+    {
+      "epoch": 1.7127577119826385,
+      "grad_norm": 0.43509939312934875,
+      "learning_rate": 4.6336740809974315e-06,
+      "loss": 0.3731,
+      "step": 3683
+    },
+    {
+      "epoch": 1.7132227561618354,
+      "grad_norm": 0.4010016918182373,
+      "learning_rate": 4.630975470168352e-06,
+      "loss": 0.3233,
+      "step": 3684
+    },
+    {
+      "epoch": 1.7136878003410323,
+      "grad_norm": 0.39253363013267517,
+      "learning_rate": 4.628276967420563e-06,
+      "loss": 0.3614,
+      "step": 3685
+    },
+    {
+      "epoch": 1.7141528445202294,
+      "grad_norm": 0.36530962586402893,
+      "learning_rate": 4.625578573544414e-06,
+      "loss": 0.3289,
+      "step": 3686
+    },
+    {
+      "epoch": 1.7146178886994266,
+      "grad_norm": 0.4295608103275299,
+      "learning_rate": 4.622880289330217e-06,
+      "loss": 0.3404,
+      "step": 3687
+    },
+    {
+      "epoch": 1.7150829328786235,
+      "grad_norm": 0.39288264513015747,
+      "learning_rate": 4.620182115568259e-06,
+      "loss": 0.3217,
+      "step": 3688
+    },
+    {
+      "epoch": 1.7155479770578204,
+      "grad_norm": 0.43459996581077576,
+      "learning_rate": 4.617484053048788e-06,
+      "loss": 0.3493,
+      "step": 3689
+    },
+    {
+      "epoch": 1.7160130212370175,
+      "grad_norm": 0.39573171734809875,
+      "learning_rate": 4.614786102562026e-06,
+      "loss": 0.3547,
+      "step": 3690
+    },
+    {
+      "epoch": 1.7164780654162146,
+      "grad_norm": 0.39884433150291443,
+      "learning_rate": 4.6120882648981565e-06,
+      "loss": 0.35,
+      "step": 3691
+    },
+    {
+      "epoch": 1.7169431095954115,
+      "grad_norm": 0.4326457679271698,
+      "learning_rate": 4.609390540847336e-06,
+      "loss": 0.3687,
+      "step": 3692
+    },
+    {
+      "epoch": 1.7174081537746084,
+      "grad_norm": 0.3984185755252838,
+      "learning_rate": 4.606692931199678e-06,
+      "loss": 0.3789,
+      "step": 3693
+    },
+    {
+      "epoch": 1.7178731979538058,
+      "grad_norm": 0.3940320312976837,
+      "learning_rate": 4.603995436745274e-06,
+      "loss": 0.3263,
+      "step": 3694
+    },
+    {
+      "epoch": 1.7183382421330027,
+      "grad_norm": 0.4455493688583374,
+      "learning_rate": 4.6012980582741725e-06,
+      "loss": 0.3492,
+      "step": 3695
+    },
+    {
+      "epoch": 1.7188032863121996,
+      "grad_norm": 0.4288008511066437,
+      "learning_rate": 4.598600796576395e-06,
+      "loss": 0.3087,
+      "step": 3696
+    },
+    {
+      "epoch": 1.7192683304913967,
+      "grad_norm": 0.3839088976383209,
+      "learning_rate": 4.595903652441923e-06,
+      "loss": 0.3869,
+      "step": 3697
+    },
+    {
+      "epoch": 1.7197333746705938,
+      "grad_norm": 0.3795816898345947,
+      "learning_rate": 4.59320662666071e-06,
+      "loss": 0.3492,
+      "step": 3698
+    },
+    {
+      "epoch": 1.7201984188497907,
+      "grad_norm": 0.39537104964256287,
+      "learning_rate": 4.590509720022665e-06,
+      "loss": 0.3488,
+      "step": 3699
+    },
+    {
+      "epoch": 1.7206634630289876,
+      "grad_norm": 0.3996933102607727,
+      "learning_rate": 4.587812933317674e-06,
+      "loss": 0.3653,
+      "step": 3700
+    },
+    {
+      "epoch": 1.7211285072081848,
+      "grad_norm": 0.3842652142047882,
+      "learning_rate": 4.5851162673355785e-06,
+      "loss": 0.3643,
+      "step": 3701
+    },
+    {
+      "epoch": 1.721593551387382,
+      "grad_norm": 0.3390163779258728,
+      "learning_rate": 4.58241972286619e-06,
+      "loss": 0.3105,
+      "step": 3702
+    },
+    {
+      "epoch": 1.7220585955665788,
+      "grad_norm": 0.3907347619533539,
+      "learning_rate": 4.5797233006992805e-06,
+      "loss": 0.3597,
+      "step": 3703
+    },
+    {
+      "epoch": 1.7225236397457757,
+      "grad_norm": 0.38712698221206665,
+      "learning_rate": 4.5770270016245915e-06,
+      "loss": 0.3285,
+      "step": 3704
+    },
+    {
+      "epoch": 1.7229886839249728,
+      "grad_norm": 0.40449246764183044,
+      "learning_rate": 4.574330826431822e-06,
+      "loss": 0.3639,
+      "step": 3705
+    },
+    {
+      "epoch": 1.72345372810417,
+      "grad_norm": 0.39589884877204895,
+      "learning_rate": 4.571634775910641e-06,
+      "loss": 0.3156,
+      "step": 3706
+    },
+    {
+      "epoch": 1.7239187722833669,
+      "grad_norm": 0.4132598042488098,
+      "learning_rate": 4.568938850850673e-06,
+      "loss": 0.3397,
+      "step": 3707
+    },
+    {
+      "epoch": 1.7243838164625638,
+      "grad_norm": 0.3807767927646637,
+      "learning_rate": 4.566243052041516e-06,
+      "loss": 0.3297,
+      "step": 3708
+    },
+    {
+      "epoch": 1.7248488606417611,
+      "grad_norm": 0.3866475820541382,
+      "learning_rate": 4.5635473802727225e-06,
+      "loss": 0.3258,
+      "step": 3709
+    },
+    {
+      "epoch": 1.725313904820958,
+      "grad_norm": 0.38793426752090454,
+      "learning_rate": 4.560851836333813e-06,
+      "loss": 0.3257,
+      "step": 3710
+    },
+    {
+      "epoch": 1.725778949000155,
+      "grad_norm": 0.440531849861145,
+      "learning_rate": 4.558156421014268e-06,
+      "loss": 0.3547,
+      "step": 3711
+    },
+    {
+      "epoch": 1.726243993179352,
+      "grad_norm": 0.4038770794868469,
+      "learning_rate": 4.555461135103529e-06,
+      "loss": 0.35,
+      "step": 3712
+    },
+    {
+      "epoch": 1.7267090373585492,
+      "grad_norm": 0.32992175221443176,
+      "learning_rate": 4.5527659793910025e-06,
+      "loss": 0.3144,
+      "step": 3713
+    },
+    {
+      "epoch": 1.727174081537746,
+      "grad_norm": 0.4566981792449951,
+      "learning_rate": 4.550070954666056e-06,
+      "loss": 0.4192,
+      "step": 3714
+    },
+    {
+      "epoch": 1.727639125716943,
+      "grad_norm": 0.419821172952652,
+      "learning_rate": 4.547376061718021e-06,
+      "loss": 0.3432,
+      "step": 3715
+    },
+    {
+      "epoch": 1.7281041698961401,
+      "grad_norm": 0.4241842031478882,
+      "learning_rate": 4.544681301336182e-06,
+      "loss": 0.3418,
+      "step": 3716
+    },
+    {
+      "epoch": 1.7285692140753373,
+      "grad_norm": 0.3986935019493103,
+      "learning_rate": 4.541986674309798e-06,
+      "loss": 0.3934,
+      "step": 3717
+    },
+    {
+      "epoch": 1.7290342582545342,
+      "grad_norm": 0.360019713640213,
+      "learning_rate": 4.539292181428074e-06,
+      "loss": 0.3393,
+      "step": 3718
+    },
+    {
+      "epoch": 1.729499302433731,
+      "grad_norm": 0.39975255727767944,
+      "learning_rate": 4.536597823480188e-06,
+      "loss": 0.3076,
+      "step": 3719
+    },
+    {
+      "epoch": 1.7299643466129282,
+      "grad_norm": 0.42592859268188477,
+      "learning_rate": 4.533903601255272e-06,
+      "loss": 0.3379,
+      "step": 3720
+    },
+    {
+      "epoch": 1.7304293907921253,
+      "grad_norm": 0.3946399986743927,
+      "learning_rate": 4.531209515542422e-06,
+      "loss": 0.377,
+      "step": 3721
+    },
+    {
+      "epoch": 1.7308944349713222,
+      "grad_norm": 0.39706122875213623,
+      "learning_rate": 4.528515567130688e-06,
+      "loss": 0.3682,
+      "step": 3722
+    },
+    {
+      "epoch": 1.7313594791505194,
+      "grad_norm": 0.3509725332260132,
+      "learning_rate": 4.525821756809088e-06,
+      "loss": 0.28,
+      "step": 3723
+    },
+    {
+      "epoch": 1.7318245233297165,
+      "grad_norm": 0.4142434895038605,
+      "learning_rate": 4.523128085366592e-06,
+      "loss": 0.3259,
+      "step": 3724
+    },
+    {
+      "epoch": 1.7322895675089134,
+      "grad_norm": 0.39951425790786743,
+      "learning_rate": 4.520434553592134e-06,
+      "loss": 0.3618,
+      "step": 3725
+    },
+    {
+      "epoch": 1.7327546116881103,
+      "grad_norm": 0.3927379250526428,
+      "learning_rate": 4.517741162274605e-06,
+      "loss": 0.3323,
+      "step": 3726
+    },
+    {
+      "epoch": 1.7332196558673074,
+      "grad_norm": 0.3651019036769867,
+      "learning_rate": 4.515047912202858e-06,
+      "loss": 0.3219,
+      "step": 3727
+    },
+    {
+      "epoch": 1.7336847000465045,
+      "grad_norm": 0.40421250462532043,
+      "learning_rate": 4.5123548041656984e-06,
+      "loss": 0.4112,
+      "step": 3728
+    },
+    {
+      "epoch": 1.7341497442257015,
+      "grad_norm": 0.39693304896354675,
+      "learning_rate": 4.509661838951897e-06,
+      "loss": 0.3038,
+      "step": 3729
+    },
+    {
+      "epoch": 1.7346147884048984,
+      "grad_norm": 0.3815270960330963,
+      "learning_rate": 4.506969017350178e-06,
+      "loss": 0.3384,
+      "step": 3730
+    },
+    {
+      "epoch": 1.7350798325840955,
+      "grad_norm": 0.3915429711341858,
+      "learning_rate": 4.5042763401492256e-06,
+      "loss": 0.3403,
+      "step": 3731
+    },
+    {
+      "epoch": 1.7355448767632926,
+      "grad_norm": 0.40549951791763306,
+      "learning_rate": 4.50158380813768e-06,
+      "loss": 0.3575,
+      "step": 3732
+    },
+    {
+      "epoch": 1.7360099209424895,
+      "grad_norm": 0.40941810607910156,
+      "learning_rate": 4.498891422104143e-06,
+      "loss": 0.3286,
+      "step": 3733
+    },
+    {
+      "epoch": 1.7364749651216864,
+      "grad_norm": 0.38618361949920654,
+      "learning_rate": 4.496199182837167e-06,
+      "loss": 0.3338,
+      "step": 3734
+    },
+    {
+      "epoch": 1.7369400093008835,
+      "grad_norm": 0.3790382146835327,
+      "learning_rate": 4.493507091125269e-06,
+      "loss": 0.3541,
+      "step": 3735
+    },
+    {
+      "epoch": 1.7374050534800807,
+      "grad_norm": 0.39466914534568787,
+      "learning_rate": 4.490815147756915e-06,
+      "loss": 0.3831,
+      "step": 3736
+    },
+    {
+      "epoch": 1.7378700976592776,
+      "grad_norm": 0.4218112528324127,
+      "learning_rate": 4.4881233535205345e-06,
+      "loss": 0.3405,
+      "step": 3737
+    },
+    {
+      "epoch": 1.7383351418384747,
+      "grad_norm": 0.41723906993865967,
+      "learning_rate": 4.4854317092045085e-06,
+      "loss": 0.37,
+      "step": 3738
+    },
+    {
+      "epoch": 1.7388001860176718,
+      "grad_norm": 0.3990018367767334,
+      "learning_rate": 4.482740215597179e-06,
+      "loss": 0.3546,
+      "step": 3739
+    },
+    {
+      "epoch": 1.7392652301968687,
+      "grad_norm": 0.4663386046886444,
+      "learning_rate": 4.480048873486836e-06,
+      "loss": 0.3627,
+      "step": 3740
+    },
+    {
+      "epoch": 1.7397302743760656,
+      "grad_norm": 0.36172690987586975,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 0.2987,
+      "step": 3741
+    },
+    {
+      "epoch": 1.7401953185552628,
+      "grad_norm": 0.3599375784397125,
+      "learning_rate": 4.474666646910074e-06,
+      "loss": 0.3212,
+      "step": 3742
+    },
+    {
+      "epoch": 1.74066036273446,
+      "grad_norm": 0.43623727560043335,
+      "learning_rate": 4.471975764020023e-06,
+      "loss": 0.3696,
+      "step": 3743
+    },
+    {
+      "epoch": 1.7411254069136568,
+      "grad_norm": 0.40363821387290955,
+      "learning_rate": 4.469285035779693e-06,
+      "loss": 0.3287,
+      "step": 3744
+    },
+    {
+      "epoch": 1.7415904510928537,
+      "grad_norm": 0.40188169479370117,
+      "learning_rate": 4.466594462977156e-06,
+      "loss": 0.3471,
+      "step": 3745
+    },
+    {
+      "epoch": 1.7420554952720508,
+      "grad_norm": 0.4021664261817932,
+      "learning_rate": 4.463904046400438e-06,
+      "loss": 0.3482,
+      "step": 3746
+    },
+    {
+      "epoch": 1.742520539451248,
+      "grad_norm": 0.3717334270477295,
+      "learning_rate": 4.4612137868375136e-06,
+      "loss": 0.3174,
+      "step": 3747
+    },
+    {
+      "epoch": 1.7429855836304449,
+      "grad_norm": 0.4289903938770294,
+      "learning_rate": 4.458523685076321e-06,
+      "loss": 0.3474,
+      "step": 3748
+    },
+    {
+      "epoch": 1.7434506278096418,
+      "grad_norm": 0.44443613290786743,
+      "learning_rate": 4.455833741904746e-06,
+      "loss": 0.3668,
+      "step": 3749
+    },
+    {
+      "epoch": 1.743915671988839,
+      "grad_norm": 0.36988234519958496,
+      "learning_rate": 4.4531439581106295e-06,
+      "loss": 0.3533,
+      "step": 3750
+    },
+    {
+      "epoch": 1.744380716168036,
+      "grad_norm": 0.4682080149650574,
+      "learning_rate": 4.450454334481763e-06,
+      "loss": 0.3772,
+      "step": 3751
+    },
+    {
+      "epoch": 1.744845760347233,
+      "grad_norm": 0.4056577682495117,
+      "learning_rate": 4.447764871805899e-06,
+      "loss": 0.3237,
+      "step": 3752
+    },
+    {
+      "epoch": 1.74531080452643,
+      "grad_norm": 0.4110541045665741,
+      "learning_rate": 4.4450755708707305e-06,
+      "loss": 0.3519,
+      "step": 3753
+    },
+    {
+      "epoch": 1.7457758487056272,
+      "grad_norm": 0.39790642261505127,
+      "learning_rate": 4.442386432463915e-06,
+      "loss": 0.3489,
+      "step": 3754
+    },
+    {
+      "epoch": 1.746240892884824,
+      "grad_norm": 0.3980329632759094,
+      "learning_rate": 4.439697457373055e-06,
+      "loss": 0.3612,
+      "step": 3755
+    },
+    {
+      "epoch": 1.746705937064021,
+      "grad_norm": 0.3943917751312256,
+      "learning_rate": 4.437008646385711e-06,
+      "loss": 0.3486,
+      "step": 3756
+    },
+    {
+      "epoch": 1.7471709812432181,
+      "grad_norm": 0.34319841861724854,
+      "learning_rate": 4.434320000289387e-06,
+      "loss": 0.3213,
+      "step": 3757
+    },
+    {
+      "epoch": 1.7476360254224153,
+      "grad_norm": 0.4046364724636078,
+      "learning_rate": 4.431631519871549e-06,
+      "loss": 0.343,
+      "step": 3758
+    },
+    {
+      "epoch": 1.7481010696016122,
+      "grad_norm": 0.4628947377204895,
+      "learning_rate": 4.428943205919605e-06,
+      "loss": 0.3734,
+      "step": 3759
+    },
+    {
+      "epoch": 1.748566113780809,
+      "grad_norm": 0.3862752914428711,
+      "learning_rate": 4.426255059220921e-06,
+      "loss": 0.3305,
+      "step": 3760
+    },
+    {
+      "epoch": 1.7490311579600062,
+      "grad_norm": 0.37330740690231323,
+      "learning_rate": 4.42356708056281e-06,
+      "loss": 0.3254,
+      "step": 3761
+    },
+    {
+      "epoch": 1.7494962021392033,
+      "grad_norm": 0.4342025816440582,
+      "learning_rate": 4.420879270732539e-06,
+      "loss": 0.3663,
+      "step": 3762
+    },
+    {
+      "epoch": 1.7499612463184002,
+      "grad_norm": 0.42904359102249146,
+      "learning_rate": 4.418191630517322e-06,
+      "loss": 0.3377,
+      "step": 3763
+    },
+    {
+      "epoch": 1.7504262904975971,
+      "grad_norm": 0.40578049421310425,
+      "learning_rate": 4.415504160704327e-06,
+      "loss": 0.3785,
+      "step": 3764
+    },
+    {
+      "epoch": 1.7508913346767943,
+      "grad_norm": 0.34141576290130615,
+      "learning_rate": 4.412816862080668e-06,
+      "loss": 0.2916,
+      "step": 3765
+    },
+    {
+      "epoch": 1.7513563788559914,
+      "grad_norm": 0.4625285565853119,
+      "learning_rate": 4.4101297354334135e-06,
+      "loss": 0.3609,
+      "step": 3766
+    },
+    {
+      "epoch": 1.7518214230351883,
+      "grad_norm": 0.40238505601882935,
+      "learning_rate": 4.407442781549577e-06,
+      "loss": 0.3996,
+      "step": 3767
+    },
+    {
+      "epoch": 1.7522864672143854,
+      "grad_norm": 0.40571123361587524,
+      "learning_rate": 4.404756001216126e-06,
+      "loss": 0.3533,
+      "step": 3768
+    },
+    {
+      "epoch": 1.7527515113935825,
+      "grad_norm": 0.37566980719566345,
+      "learning_rate": 4.4020693952199726e-06,
+      "loss": 0.3394,
+      "step": 3769
+    },
+    {
+      "epoch": 1.7532165555727794,
+      "grad_norm": 0.4016037583351135,
+      "learning_rate": 4.3993829643479825e-06,
+      "loss": 0.324,
+      "step": 3770
+    },
+    {
+      "epoch": 1.7536815997519763,
+      "grad_norm": 0.41683268547058105,
+      "learning_rate": 4.396696709386964e-06,
+      "loss": 0.3341,
+      "step": 3771
+    },
+    {
+      "epoch": 1.7541466439311735,
+      "grad_norm": 0.3792789876461029,
+      "learning_rate": 4.394010631123681e-06,
+      "loss": 0.3193,
+      "step": 3772
+    },
+    {
+      "epoch": 1.7546116881103706,
+      "grad_norm": 0.38343918323516846,
+      "learning_rate": 4.39132473034484e-06,
+      "loss": 0.3722,
+      "step": 3773
+    },
+    {
+      "epoch": 1.7550767322895675,
+      "grad_norm": 0.3963862359523773,
+      "learning_rate": 4.388639007837101e-06,
+      "loss": 0.3609,
+      "step": 3774
+    },
+    {
+      "epoch": 1.7555417764687644,
+      "grad_norm": 0.3763202428817749,
+      "learning_rate": 4.385953464387064e-06,
+      "loss": 0.3183,
+      "step": 3775
+    },
+    {
+      "epoch": 1.7560068206479615,
+      "grad_norm": 0.3931805491447449,
+      "learning_rate": 4.383268100781285e-06,
+      "loss": 0.3821,
+      "step": 3776
+    },
+    {
+      "epoch": 1.7564718648271587,
+      "grad_norm": 0.3617267906665802,
+      "learning_rate": 4.38058291780626e-06,
+      "loss": 0.3264,
+      "step": 3777
+    },
+    {
+      "epoch": 1.7569369090063556,
+      "grad_norm": 0.38972967863082886,
+      "learning_rate": 4.377897916248438e-06,
+      "loss": 0.3538,
+      "step": 3778
+    },
+    {
+      "epoch": 1.7574019531855525,
+      "grad_norm": 0.4271766245365143,
+      "learning_rate": 4.37521309689421e-06,
+      "loss": 0.3718,
+      "step": 3779
+    },
+    {
+      "epoch": 1.7578669973647496,
+      "grad_norm": 0.3271983861923218,
+      "learning_rate": 4.37252846052992e-06,
+      "loss": 0.3391,
+      "step": 3780
+    },
+    {
+      "epoch": 1.7583320415439467,
+      "grad_norm": 0.41087400913238525,
+      "learning_rate": 4.36984400794185e-06,
+      "loss": 0.3266,
+      "step": 3781
+    },
+    {
+      "epoch": 1.7587970857231436,
+      "grad_norm": 0.44612687826156616,
+      "learning_rate": 4.367159739916236e-06,
+      "loss": 0.334,
+      "step": 3782
+    },
+    {
+      "epoch": 1.7592621299023408,
+      "grad_norm": 0.3791424632072449,
+      "learning_rate": 4.364475657239253e-06,
+      "loss": 0.3684,
+      "step": 3783
+    },
+    {
+      "epoch": 1.759727174081538,
+      "grad_norm": 0.4565580189228058,
+      "learning_rate": 4.361791760697027e-06,
+      "loss": 0.3547,
+      "step": 3784
+    },
+    {
+      "epoch": 1.7601922182607348,
+      "grad_norm": 0.48362359404563904,
+      "learning_rate": 4.35910805107563e-06,
+      "loss": 0.3483,
+      "step": 3785
+    },
+    {
+      "epoch": 1.7606572624399317,
+      "grad_norm": 0.37739720940589905,
+      "learning_rate": 4.356424529161072e-06,
+      "loss": 0.351,
+      "step": 3786
+    },
+    {
+      "epoch": 1.7611223066191288,
+      "grad_norm": 0.36957481503486633,
+      "learning_rate": 4.353741195739318e-06,
+      "loss": 0.362,
+      "step": 3787
+    },
+    {
+      "epoch": 1.761587350798326,
+      "grad_norm": 0.39713501930236816,
+      "learning_rate": 4.351058051596269e-06,
+      "loss": 0.3421,
+      "step": 3788
+    },
+    {
+      "epoch": 1.7620523949775229,
+      "grad_norm": 0.37897831201553345,
+      "learning_rate": 4.348375097517776e-06,
+      "loss": 0.3196,
+      "step": 3789
+    },
+    {
+      "epoch": 1.7625174391567198,
+      "grad_norm": 0.38132160902023315,
+      "learning_rate": 4.345692334289632e-06,
+      "loss": 0.3764,
+      "step": 3790
+    },
+    {
+      "epoch": 1.762982483335917,
+      "grad_norm": 0.39471349120140076,
+      "learning_rate": 4.343009762697577e-06,
+      "loss": 0.3472,
+      "step": 3791
+    },
+    {
+      "epoch": 1.763447527515114,
+      "grad_norm": 0.35921382904052734,
+      "learning_rate": 4.340327383527289e-06,
+      "loss": 0.3489,
+      "step": 3792
+    },
+    {
+      "epoch": 1.763912571694311,
+      "grad_norm": 0.3691664934158325,
+      "learning_rate": 4.337645197564398e-06,
+      "loss": 0.3357,
+      "step": 3793
+    },
+    {
+      "epoch": 1.7643776158735078,
+      "grad_norm": 0.38761186599731445,
+      "learning_rate": 4.334963205594467e-06,
+      "loss": 0.3535,
+      "step": 3794
+    },
+    {
+      "epoch": 1.764842660052705,
+      "grad_norm": 0.4036206305027008,
+      "learning_rate": 4.332281408403011e-06,
+      "loss": 0.3786,
+      "step": 3795
+    },
+    {
+      "epoch": 1.765307704231902,
+      "grad_norm": 0.37231314182281494,
+      "learning_rate": 4.3295998067754844e-06,
+      "loss": 0.3068,
+      "step": 3796
+    },
+    {
+      "epoch": 1.765772748411099,
+      "grad_norm": 0.38186076283454895,
+      "learning_rate": 4.326918401497287e-06,
+      "loss": 0.334,
+      "step": 3797
+    },
+    {
+      "epoch": 1.7662377925902961,
+      "grad_norm": 0.4085099697113037,
+      "learning_rate": 4.3242371933537554e-06,
+      "loss": 0.3273,
+      "step": 3798
+    },
+    {
+      "epoch": 1.7667028367694932,
+      "grad_norm": 0.4823872447013855,
+      "learning_rate": 4.321556183130175e-06,
+      "loss": 0.3616,
+      "step": 3799
+    },
+    {
+      "epoch": 1.7671678809486902,
+      "grad_norm": 0.38693398237228394,
+      "learning_rate": 4.318875371611766e-06,
+      "loss": 0.3751,
+      "step": 3800
+    },
+    {
+      "epoch": 1.767632925127887,
+      "grad_norm": 0.4294361472129822,
+      "learning_rate": 4.3161947595836985e-06,
+      "loss": 0.3404,
+      "step": 3801
+    },
+    {
+      "epoch": 1.7680979693070842,
+      "grad_norm": 0.46202561259269714,
+      "learning_rate": 4.313514347831077e-06,
+      "loss": 0.3245,
+      "step": 3802
+    },
+    {
+      "epoch": 1.7685630134862813,
+      "grad_norm": 0.42043882608413696,
+      "learning_rate": 4.310834137138956e-06,
+      "loss": 0.3401,
+      "step": 3803
+    },
+    {
+      "epoch": 1.7690280576654782,
+      "grad_norm": 0.4185798168182373,
+      "learning_rate": 4.308154128292318e-06,
+      "loss": 0.359,
+      "step": 3804
+    },
+    {
+      "epoch": 1.7694931018446751,
+      "grad_norm": 0.44950559735298157,
+      "learning_rate": 4.305474322076102e-06,
+      "loss": 0.3354,
+      "step": 3805
+    },
+    {
+      "epoch": 1.7699581460238722,
+      "grad_norm": 0.4571090638637543,
+      "learning_rate": 4.302794719275173e-06,
+      "loss": 0.3484,
+      "step": 3806
+    },
+    {
+      "epoch": 1.7704231902030694,
+      "grad_norm": 0.42826327681541443,
+      "learning_rate": 4.300115320674346e-06,
+      "loss": 0.3595,
+      "step": 3807
+    },
+    {
+      "epoch": 1.7708882343822663,
+      "grad_norm": 0.42987918853759766,
+      "learning_rate": 4.297436127058373e-06,
+      "loss": 0.3028,
+      "step": 3808
+    },
+    {
+      "epoch": 1.7713532785614632,
+      "grad_norm": 0.4052066206932068,
+      "learning_rate": 4.294757139211948e-06,
+      "loss": 0.3719,
+      "step": 3809
+    },
+    {
+      "epoch": 1.7718183227406603,
+      "grad_norm": 0.4018250107765198,
+      "learning_rate": 4.292078357919701e-06,
+      "loss": 0.3937,
+      "step": 3810
+    },
+    {
+      "epoch": 1.7722833669198574,
+      "grad_norm": 0.40564820170402527,
+      "learning_rate": 4.289399783966205e-06,
+      "loss": 0.344,
+      "step": 3811
+    },
+    {
+      "epoch": 1.7727484110990543,
+      "grad_norm": 0.4192334711551666,
+      "learning_rate": 4.286721418135968e-06,
+      "loss": 0.3571,
+      "step": 3812
+    },
+    {
+      "epoch": 1.7732134552782515,
+      "grad_norm": 0.36322298645973206,
+      "learning_rate": 4.284043261213442e-06,
+      "loss": 0.3576,
+      "step": 3813
+    },
+    {
+      "epoch": 1.7736784994574486,
+      "grad_norm": 0.35039764642715454,
+      "learning_rate": 4.281365313983016e-06,
+      "loss": 0.3473,
+      "step": 3814
+    },
+    {
+      "epoch": 1.7741435436366455,
+      "grad_norm": 0.3839752972126007,
+      "learning_rate": 4.278687577229018e-06,
+      "loss": 0.3502,
+      "step": 3815
+    },
+    {
+      "epoch": 1.7746085878158424,
+      "grad_norm": 0.4036378264427185,
+      "learning_rate": 4.2760100517357095e-06,
+      "loss": 0.3351,
+      "step": 3816
+    },
+    {
+      "epoch": 1.7750736319950395,
+      "grad_norm": 0.3815009891986847,
+      "learning_rate": 4.273332738287299e-06,
+      "loss": 0.3496,
+      "step": 3817
+    },
+    {
+      "epoch": 1.7755386761742367,
+      "grad_norm": 0.38633623719215393,
+      "learning_rate": 4.270655637667926e-06,
+      "loss": 0.3215,
+      "step": 3818
+    },
+    {
+      "epoch": 1.7760037203534336,
+      "grad_norm": 0.39606183767318726,
+      "learning_rate": 4.267978750661669e-06,
+      "loss": 0.3445,
+      "step": 3819
+    },
+    {
+      "epoch": 1.7764687645326305,
+      "grad_norm": 0.39145374298095703,
+      "learning_rate": 4.265302078052546e-06,
+      "loss": 0.3274,
+      "step": 3820
+    },
+    {
+      "epoch": 1.7769338087118276,
+      "grad_norm": 0.37993159890174866,
+      "learning_rate": 4.26262562062451e-06,
+      "loss": 0.3546,
+      "step": 3821
+    },
+    {
+      "epoch": 1.7773988528910247,
+      "grad_norm": 0.41166287660598755,
+      "learning_rate": 4.259949379161454e-06,
+      "loss": 0.3559,
+      "step": 3822
+    },
+    {
+      "epoch": 1.7778638970702216,
+      "grad_norm": 0.4474835991859436,
+      "learning_rate": 4.2572733544472025e-06,
+      "loss": 0.3525,
+      "step": 3823
+    },
+    {
+      "epoch": 1.7783289412494185,
+      "grad_norm": 0.38438984751701355,
+      "learning_rate": 4.2545975472655215e-06,
+      "loss": 0.3461,
+      "step": 3824
+    },
+    {
+      "epoch": 1.7787939854286157,
+      "grad_norm": 0.37471044063568115,
+      "learning_rate": 4.2519219584001106e-06,
+      "loss": 0.3681,
+      "step": 3825
+    },
+    {
+      "epoch": 1.7792590296078128,
+      "grad_norm": 0.3626500368118286,
+      "learning_rate": 4.249246588634609e-06,
+      "loss": 0.334,
+      "step": 3826
+    },
+    {
+      "epoch": 1.7797240737870097,
+      "grad_norm": 0.34971731901168823,
+      "learning_rate": 4.246571438752585e-06,
+      "loss": 0.3683,
+      "step": 3827
+    },
+    {
+      "epoch": 1.7801891179662068,
+      "grad_norm": 0.37381622195243835,
+      "learning_rate": 4.243896509537551e-06,
+      "loss": 0.3602,
+      "step": 3828
+    },
+    {
+      "epoch": 1.780654162145404,
+      "grad_norm": 0.4181731343269348,
+      "learning_rate": 4.241221801772945e-06,
+      "loss": 0.3684,
+      "step": 3829
+    },
+    {
+      "epoch": 1.7811192063246009,
+      "grad_norm": 0.3543993830680847,
+      "learning_rate": 4.238547316242149e-06,
+      "loss": 0.3383,
+      "step": 3830
+    },
+    {
+      "epoch": 1.7815842505037978,
+      "grad_norm": 0.4083957076072693,
+      "learning_rate": 4.235873053728475e-06,
+      "loss": 0.372,
+      "step": 3831
+    },
+    {
+      "epoch": 1.7820492946829949,
+      "grad_norm": 0.3555138409137726,
+      "learning_rate": 4.2331990150151745e-06,
+      "loss": 0.351,
+      "step": 3832
+    },
+    {
+      "epoch": 1.782514338862192,
+      "grad_norm": 0.40743115544319153,
+      "learning_rate": 4.230525200885425e-06,
+      "loss": 0.3407,
+      "step": 3833
+    },
+    {
+      "epoch": 1.782979383041389,
+      "grad_norm": 0.435773104429245,
+      "learning_rate": 4.227851612122347e-06,
+      "loss": 0.3665,
+      "step": 3834
+    },
+    {
+      "epoch": 1.7834444272205858,
+      "grad_norm": 0.3714430332183838,
+      "learning_rate": 4.225178249508988e-06,
+      "loss": 0.3557,
+      "step": 3835
+    },
+    {
+      "epoch": 1.783909471399783,
+      "grad_norm": 0.35675156116485596,
+      "learning_rate": 4.222505113828335e-06,
+      "loss": 0.3125,
+      "step": 3836
+    },
+    {
+      "epoch": 1.78437451557898,
+      "grad_norm": 0.3813643753528595,
+      "learning_rate": 4.219832205863303e-06,
+      "loss": 0.3653,
+      "step": 3837
+    },
+    {
+      "epoch": 1.784839559758177,
+      "grad_norm": 0.4000234603881836,
+      "learning_rate": 4.217159526396749e-06,
+      "loss": 0.3811,
+      "step": 3838
+    },
+    {
+      "epoch": 1.785304603937374,
+      "grad_norm": 0.3764249086380005,
+      "learning_rate": 4.214487076211452e-06,
+      "loss": 0.2882,
+      "step": 3839
+    },
+    {
+      "epoch": 1.7857696481165712,
+      "grad_norm": 0.47971102595329285,
+      "learning_rate": 4.2118148560901325e-06,
+      "loss": 0.3693,
+      "step": 3840
+    },
+    {
+      "epoch": 1.7862346922957681,
+      "grad_norm": 0.3611883521080017,
+      "learning_rate": 4.209142866815438e-06,
+      "loss": 0.3483,
+      "step": 3841
+    },
+    {
+      "epoch": 1.786699736474965,
+      "grad_norm": 0.3263775110244751,
+      "learning_rate": 4.206471109169952e-06,
+      "loss": 0.3338,
+      "step": 3842
+    },
+    {
+      "epoch": 1.7871647806541622,
+      "grad_norm": 0.4258721172809601,
+      "learning_rate": 4.2037995839361876e-06,
+      "loss": 0.3556,
+      "step": 3843
+    },
+    {
+      "epoch": 1.7876298248333593,
+      "grad_norm": 0.43471091985702515,
+      "learning_rate": 4.201128291896594e-06,
+      "loss": 0.349,
+      "step": 3844
+    },
+    {
+      "epoch": 1.7880948690125562,
+      "grad_norm": 0.4108128249645233,
+      "learning_rate": 4.198457233833546e-06,
+      "loss": 0.3643,
+      "step": 3845
+    },
+    {
+      "epoch": 1.7885599131917531,
+      "grad_norm": 0.3940763771533966,
+      "learning_rate": 4.195786410529357e-06,
+      "loss": 0.3659,
+      "step": 3846
+    },
+    {
+      "epoch": 1.7890249573709502,
+      "grad_norm": 0.4194324314594269,
+      "learning_rate": 4.193115822766263e-06,
+      "loss": 0.3359,
+      "step": 3847
+    },
+    {
+      "epoch": 1.7894900015501474,
+      "grad_norm": 0.37528660893440247,
+      "learning_rate": 4.19044547132644e-06,
+      "loss": 0.3255,
+      "step": 3848
+    },
+    {
+      "epoch": 1.7899550457293443,
+      "grad_norm": 0.4335568845272064,
+      "learning_rate": 4.1877753569919865e-06,
+      "loss": 0.3884,
+      "step": 3849
+    },
+    {
+      "epoch": 1.7904200899085412,
+      "grad_norm": 0.41178300976753235,
+      "learning_rate": 4.18510548054494e-06,
+      "loss": 0.3179,
+      "step": 3850
+    },
+    {
+      "epoch": 1.7908851340877383,
+      "grad_norm": 0.41776037216186523,
+      "learning_rate": 4.18243584276726e-06,
+      "loss": 0.3448,
+      "step": 3851
+    },
+    {
+      "epoch": 1.7913501782669354,
+      "grad_norm": 0.40628287196159363,
+      "learning_rate": 4.179766444440844e-06,
+      "loss": 0.3959,
+      "step": 3852
+    },
+    {
+      "epoch": 1.7918152224461323,
+      "grad_norm": 0.37497735023498535,
+      "learning_rate": 4.177097286347511e-06,
+      "loss": 0.3118,
+      "step": 3853
+    },
+    {
+      "epoch": 1.7922802666253292,
+      "grad_norm": 0.4827350378036499,
+      "learning_rate": 4.174428369269018e-06,
+      "loss": 0.3839,
+      "step": 3854
+    },
+    {
+      "epoch": 1.7927453108045266,
+      "grad_norm": 0.3956352472305298,
+      "learning_rate": 4.171759693987046e-06,
+      "loss": 0.3378,
+      "step": 3855
+    },
+    {
+      "epoch": 1.7932103549837235,
+      "grad_norm": 0.3712174892425537,
+      "learning_rate": 4.169091261283205e-06,
+      "loss": 0.3207,
+      "step": 3856
+    },
+    {
+      "epoch": 1.7936753991629204,
+      "grad_norm": 0.48411279916763306,
+      "learning_rate": 4.166423071939038e-06,
+      "loss": 0.3747,
+      "step": 3857
+    },
+    {
+      "epoch": 1.7941404433421175,
+      "grad_norm": 0.38150522112846375,
+      "learning_rate": 4.163755126736011e-06,
+      "loss": 0.3312,
+      "step": 3858
+    },
+    {
+      "epoch": 1.7946054875213147,
+      "grad_norm": 0.3953656554222107,
+      "learning_rate": 4.1610874264555265e-06,
+      "loss": 0.3349,
+      "step": 3859
+    },
+    {
+      "epoch": 1.7950705317005116,
+      "grad_norm": 0.4521973431110382,
+      "learning_rate": 4.158419971878907e-06,
+      "loss": 0.3255,
+      "step": 3860
+    },
+    {
+      "epoch": 1.7955355758797085,
+      "grad_norm": 0.3754480481147766,
+      "learning_rate": 4.155752763787409e-06,
+      "loss": 0.3399,
+      "step": 3861
+    },
+    {
+      "epoch": 1.7960006200589056,
+      "grad_norm": 0.36341235041618347,
+      "learning_rate": 4.1530858029622125e-06,
+      "loss": 0.3306,
+      "step": 3862
+    },
+    {
+      "epoch": 1.7964656642381027,
+      "grad_norm": 0.5212793946266174,
+      "learning_rate": 4.150419090184428e-06,
+      "loss": 0.4108,
+      "step": 3863
+    },
+    {
+      "epoch": 1.7969307084172996,
+      "grad_norm": 0.37436044216156006,
+      "learning_rate": 4.147752626235092e-06,
+      "loss": 0.3122,
+      "step": 3864
+    },
+    {
+      "epoch": 1.7973957525964965,
+      "grad_norm": 0.3866221606731415,
+      "learning_rate": 4.145086411895168e-06,
+      "loss": 0.3834,
+      "step": 3865
+    },
+    {
+      "epoch": 1.7978607967756937,
+      "grad_norm": 0.42558223009109497,
+      "learning_rate": 4.142420447945548e-06,
+      "loss": 0.4011,
+      "step": 3866
+    },
+    {
+      "epoch": 1.7983258409548908,
+      "grad_norm": 0.3522246479988098,
+      "learning_rate": 4.13975473516705e-06,
+      "loss": 0.2951,
+      "step": 3867
+    },
+    {
+      "epoch": 1.7987908851340877,
+      "grad_norm": 0.37685632705688477,
+      "learning_rate": 4.137089274340415e-06,
+      "loss": 0.3199,
+      "step": 3868
+    },
+    {
+      "epoch": 1.7992559293132846,
+      "grad_norm": 0.3827194571495056,
+      "learning_rate": 4.134424066246318e-06,
+      "loss": 0.3436,
+      "step": 3869
+    },
+    {
+      "epoch": 1.799720973492482,
+      "grad_norm": 0.4337718188762665,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.3462,
+      "step": 3870
+    },
+    {
+      "epoch": 1.8001860176716789,
+      "grad_norm": 0.40320268273353577,
+      "learning_rate": 4.129094411378034e-06,
+      "loss": 0.3414,
+      "step": 3871
+    },
+    {
+      "epoch": 1.8006510618508758,
+      "grad_norm": 0.37748897075653076,
+      "learning_rate": 4.1264299661648195e-06,
+      "loss": 0.3365,
+      "step": 3872
+    },
+    {
+      "epoch": 1.8011161060300729,
+      "grad_norm": 0.3657347857952118,
+      "learning_rate": 4.123765776806081e-06,
+      "loss": 0.339,
+      "step": 3873
+    },
+    {
+      "epoch": 1.80158115020927,
+      "grad_norm": 0.4097522497177124,
+      "learning_rate": 4.121101844082111e-06,
+      "loss": 0.3451,
+      "step": 3874
+    },
+    {
+      "epoch": 1.802046194388467,
+      "grad_norm": 0.3985246419906616,
+      "learning_rate": 4.118438168773137e-06,
+      "loss": 0.3597,
+      "step": 3875
+    },
+    {
+      "epoch": 1.8025112385676638,
+      "grad_norm": 0.41197559237480164,
+      "learning_rate": 4.115774751659302e-06,
+      "loss": 0.3268,
+      "step": 3876
+    },
+    {
+      "epoch": 1.802976282746861,
+      "grad_norm": 0.3859005272388458,
+      "learning_rate": 4.11311159352068e-06,
+      "loss": 0.3354,
+      "step": 3877
+    },
+    {
+      "epoch": 1.803441326926058,
+      "grad_norm": 0.3585662841796875,
+      "learning_rate": 4.110448695137266e-06,
+      "loss": 0.3303,
+      "step": 3878
+    },
+    {
+      "epoch": 1.803906371105255,
+      "grad_norm": 0.38057127594947815,
+      "learning_rate": 4.107786057288982e-06,
+      "loss": 0.3306,
+      "step": 3879
+    },
+    {
+      "epoch": 1.8043714152844519,
+      "grad_norm": 0.4793699085712433,
+      "learning_rate": 4.105123680755667e-06,
+      "loss": 0.3414,
+      "step": 3880
+    },
+    {
+      "epoch": 1.804836459463649,
+      "grad_norm": 0.4067707061767578,
+      "learning_rate": 4.102461566317093e-06,
+      "loss": 0.322,
+      "step": 3881
+    },
+    {
+      "epoch": 1.8053015036428461,
+      "grad_norm": 0.4255411922931671,
+      "learning_rate": 4.099799714752944e-06,
+      "loss": 0.296,
+      "step": 3882
+    },
+    {
+      "epoch": 1.805766547822043,
+      "grad_norm": 0.40467318892478943,
+      "learning_rate": 4.097138126842839e-06,
+      "loss": 0.3672,
+      "step": 3883
+    },
+    {
+      "epoch": 1.8062315920012402,
+      "grad_norm": 0.41983935236930847,
+      "learning_rate": 4.09447680336631e-06,
+      "loss": 0.3322,
+      "step": 3884
+    },
+    {
+      "epoch": 1.8066966361804373,
+      "grad_norm": 0.37576842308044434,
+      "learning_rate": 4.091815745102818e-06,
+      "loss": 0.3064,
+      "step": 3885
+    },
+    {
+      "epoch": 1.8071616803596342,
+      "grad_norm": 0.44206252694129944,
+      "learning_rate": 4.089154952831741e-06,
+      "loss": 0.3848,
+      "step": 3886
+    },
+    {
+      "epoch": 1.807626724538831,
+      "grad_norm": 0.3728598356246948,
+      "learning_rate": 4.086494427332386e-06,
+      "loss": 0.3392,
+      "step": 3887
+    },
+    {
+      "epoch": 1.8080917687180282,
+      "grad_norm": 0.41656002402305603,
+      "learning_rate": 4.083834169383972e-06,
+      "loss": 0.3432,
+      "step": 3888
+    },
+    {
+      "epoch": 1.8085568128972254,
+      "grad_norm": 0.40540826320648193,
+      "learning_rate": 4.0811741797656505e-06,
+      "loss": 0.3381,
+      "step": 3889
+    },
+    {
+      "epoch": 1.8090218570764223,
+      "grad_norm": 0.40355539321899414,
+      "learning_rate": 4.078514459256485e-06,
+      "loss": 0.3464,
+      "step": 3890
+    },
+    {
+      "epoch": 1.8094869012556192,
+      "grad_norm": 0.37891075015068054,
+      "learning_rate": 4.075855008635468e-06,
+      "loss": 0.3557,
+      "step": 3891
+    },
+    {
+      "epoch": 1.8099519454348163,
+      "grad_norm": 0.35543182492256165,
+      "learning_rate": 4.073195828681509e-06,
+      "loss": 0.3177,
+      "step": 3892
+    },
+    {
+      "epoch": 1.8104169896140134,
+      "grad_norm": 0.400921106338501,
+      "learning_rate": 4.070536920173435e-06,
+      "loss": 0.3601,
+      "step": 3893
+    },
+    {
+      "epoch": 1.8108820337932103,
+      "grad_norm": 0.35732823610305786,
+      "learning_rate": 4.067878283890002e-06,
+      "loss": 0.345,
+      "step": 3894
+    },
+    {
+      "epoch": 1.8113470779724072,
+      "grad_norm": 0.39798301458358765,
+      "learning_rate": 4.065219920609877e-06,
+      "loss": 0.3464,
+      "step": 3895
+    },
+    {
+      "epoch": 1.8118121221516044,
+      "grad_norm": 0.3960334062576294,
+      "learning_rate": 4.062561831111656e-06,
+      "loss": 0.3632,
+      "step": 3896
+    },
+    {
+      "epoch": 1.8122771663308015,
+      "grad_norm": 0.3809148669242859,
+      "learning_rate": 4.059904016173844e-06,
+      "loss": 0.3482,
+      "step": 3897
+    },
+    {
+      "epoch": 1.8127422105099984,
+      "grad_norm": 0.397305965423584,
+      "learning_rate": 4.05724647657488e-06,
+      "loss": 0.3559,
+      "step": 3898
+    },
+    {
+      "epoch": 1.8132072546891955,
+      "grad_norm": 0.3920312225818634,
+      "learning_rate": 4.0545892130931065e-06,
+      "loss": 0.3402,
+      "step": 3899
+    },
+    {
+      "epoch": 1.8136722988683927,
+      "grad_norm": 0.39747723937034607,
+      "learning_rate": 4.051932226506797e-06,
+      "loss": 0.3551,
+      "step": 3900
+    },
+    {
+      "epoch": 1.8141373430475896,
+      "grad_norm": 0.4466649889945984,
+      "learning_rate": 4.049275517594137e-06,
+      "loss": 0.3301,
+      "step": 3901
+    },
+    {
+      "epoch": 1.8146023872267865,
+      "grad_norm": 0.413919597864151,
+      "learning_rate": 4.046619087133238e-06,
+      "loss": 0.3609,
+      "step": 3902
+    },
+    {
+      "epoch": 1.8150674314059836,
+      "grad_norm": 0.34546610713005066,
+      "learning_rate": 4.04396293590212e-06,
+      "loss": 0.3098,
+      "step": 3903
+    },
+    {
+      "epoch": 1.8155324755851807,
+      "grad_norm": 0.40819764137268066,
+      "learning_rate": 4.0413070646787325e-06,
+      "loss": 0.3451,
+      "step": 3904
+    },
+    {
+      "epoch": 1.8159975197643776,
+      "grad_norm": 0.3749530017375946,
+      "learning_rate": 4.03865147424093e-06,
+      "loss": 0.3456,
+      "step": 3905
+    },
+    {
+      "epoch": 1.8164625639435745,
+      "grad_norm": 0.3659387528896332,
+      "learning_rate": 4.035996165366497e-06,
+      "loss": 0.3042,
+      "step": 3906
+    },
+    {
+      "epoch": 1.8169276081227717,
+      "grad_norm": 0.41476431488990784,
+      "learning_rate": 4.033341138833127e-06,
+      "loss": 0.3314,
+      "step": 3907
+    },
+    {
+      "epoch": 1.8173926523019688,
+      "grad_norm": 0.39919087290763855,
+      "learning_rate": 4.030686395418439e-06,
+      "loss": 0.3424,
+      "step": 3908
+    },
+    {
+      "epoch": 1.8178576964811657,
+      "grad_norm": 0.40032151341438293,
+      "learning_rate": 4.028031935899958e-06,
+      "loss": 0.3332,
+      "step": 3909
+    },
+    {
+      "epoch": 1.8183227406603626,
+      "grad_norm": 0.41616636514663696,
+      "learning_rate": 4.025377761055138e-06,
+      "loss": 0.3357,
+      "step": 3910
+    },
+    {
+      "epoch": 1.8187877848395597,
+      "grad_norm": 0.40295374393463135,
+      "learning_rate": 4.022723871661338e-06,
+      "loss": 0.3281,
+      "step": 3911
+    },
+    {
+      "epoch": 1.8192528290187568,
+      "grad_norm": 0.49246686697006226,
+      "learning_rate": 4.020070268495844e-06,
+      "loss": 0.3867,
+      "step": 3912
+    },
+    {
+      "epoch": 1.8197178731979538,
+      "grad_norm": 0.38608917593955994,
+      "learning_rate": 4.017416952335849e-06,
+      "loss": 0.3213,
+      "step": 3913
+    },
+    {
+      "epoch": 1.8201829173771509,
+      "grad_norm": 0.3529578745365143,
+      "learning_rate": 4.014763923958471e-06,
+      "loss": 0.3246,
+      "step": 3914
+    },
+    {
+      "epoch": 1.820647961556348,
+      "grad_norm": 0.42414984107017517,
+      "learning_rate": 4.0121111841407345e-06,
+      "loss": 0.3668,
+      "step": 3915
+    },
+    {
+      "epoch": 1.821113005735545,
+      "grad_norm": 0.406495064496994,
+      "learning_rate": 4.0094587336595875e-06,
+      "loss": 0.3176,
+      "step": 3916
+    },
+    {
+      "epoch": 1.8215780499147418,
+      "grad_norm": 0.39914000034332275,
+      "learning_rate": 4.006806573291886e-06,
+      "loss": 0.343,
+      "step": 3917
+    },
+    {
+      "epoch": 1.822043094093939,
+      "grad_norm": 0.3730376958847046,
+      "learning_rate": 4.004154703814407e-06,
+      "loss": 0.2956,
+      "step": 3918
+    },
+    {
+      "epoch": 1.822508138273136,
+      "grad_norm": 0.43556198477745056,
+      "learning_rate": 4.00150312600384e-06,
+      "loss": 0.3393,
+      "step": 3919
+    },
+    {
+      "epoch": 1.822973182452333,
+      "grad_norm": 0.3719125986099243,
+      "learning_rate": 3.998851840636789e-06,
+      "loss": 0.3421,
+      "step": 3920
+    },
+    {
+      "epoch": 1.8234382266315299,
+      "grad_norm": 0.40865078568458557,
+      "learning_rate": 3.996200848489771e-06,
+      "loss": 0.3597,
+      "step": 3921
+    },
+    {
+      "epoch": 1.823903270810727,
+      "grad_norm": 0.38333582878112793,
+      "learning_rate": 3.9935501503392214e-06,
+      "loss": 0.3604,
+      "step": 3922
+    },
+    {
+      "epoch": 1.8243683149899241,
+      "grad_norm": 0.40074682235717773,
+      "learning_rate": 3.990899746961483e-06,
+      "loss": 0.3406,
+      "step": 3923
+    },
+    {
+      "epoch": 1.824833359169121,
+      "grad_norm": 0.4388389587402344,
+      "learning_rate": 3.9882496391328185e-06,
+      "loss": 0.3327,
+      "step": 3924
+    },
+    {
+      "epoch": 1.825298403348318,
+      "grad_norm": 0.40420886874198914,
+      "learning_rate": 3.9855998276294e-06,
+      "loss": 0.3187,
+      "step": 3925
+    },
+    {
+      "epoch": 1.825763447527515,
+      "grad_norm": 0.4143049418926239,
+      "learning_rate": 3.982950313227317e-06,
+      "loss": 0.3311,
+      "step": 3926
+    },
+    {
+      "epoch": 1.8262284917067122,
+      "grad_norm": 0.4336578845977783,
+      "learning_rate": 3.980301096702567e-06,
+      "loss": 0.3923,
+      "step": 3927
+    },
+    {
+      "epoch": 1.826693535885909,
+      "grad_norm": 0.395292729139328,
+      "learning_rate": 3.9776521788310605e-06,
+      "loss": 0.3403,
+      "step": 3928
+    },
+    {
+      "epoch": 1.8271585800651062,
+      "grad_norm": 0.38078877329826355,
+      "learning_rate": 3.975003560388625e-06,
+      "loss": 0.3363,
+      "step": 3929
+    },
+    {
+      "epoch": 1.8276236242443034,
+      "grad_norm": 0.375010222196579,
+      "learning_rate": 3.9723552421509975e-06,
+      "loss": 0.3441,
+      "step": 3930
+    },
+    {
+      "epoch": 1.8280886684235003,
+      "grad_norm": 0.3816002607345581,
+      "learning_rate": 3.969707224893829e-06,
+      "loss": 0.3388,
+      "step": 3931
+    },
+    {
+      "epoch": 1.8285537126026972,
+      "grad_norm": 0.3783058524131775,
+      "learning_rate": 3.967059509392677e-06,
+      "loss": 0.3584,
+      "step": 3932
+    },
+    {
+      "epoch": 1.8290187567818943,
+      "grad_norm": 0.4086306095123291,
+      "learning_rate": 3.964412096423019e-06,
+      "loss": 0.3437,
+      "step": 3933
+    },
+    {
+      "epoch": 1.8294838009610914,
+      "grad_norm": 0.3575378954410553,
+      "learning_rate": 3.961764986760234e-06,
+      "loss": 0.3245,
+      "step": 3934
+    },
+    {
+      "epoch": 1.8299488451402883,
+      "grad_norm": 0.4207441806793213,
+      "learning_rate": 3.959118181179622e-06,
+      "loss": 0.3375,
+      "step": 3935
+    },
+    {
+      "epoch": 1.8304138893194852,
+      "grad_norm": 0.39432263374328613,
+      "learning_rate": 3.9564716804563855e-06,
+      "loss": 0.3562,
+      "step": 3936
+    },
+    {
+      "epoch": 1.8308789334986824,
+      "grad_norm": 0.3997967541217804,
+      "learning_rate": 3.9538254853656465e-06,
+      "loss": 0.3544,
+      "step": 3937
+    },
+    {
+      "epoch": 1.8313439776778795,
+      "grad_norm": 0.369190514087677,
+      "learning_rate": 3.951179596682427e-06,
+      "loss": 0.3524,
+      "step": 3938
+    },
+    {
+      "epoch": 1.8318090218570764,
+      "grad_norm": 0.37524935603141785,
+      "learning_rate": 3.948534015181671e-06,
+      "loss": 0.3549,
+      "step": 3939
+    },
+    {
+      "epoch": 1.8322740660362733,
+      "grad_norm": 0.36414414644241333,
+      "learning_rate": 3.945888741638219e-06,
+      "loss": 0.3104,
+      "step": 3940
+    },
+    {
+      "epoch": 1.8327391102154704,
+      "grad_norm": 0.3788766860961914,
+      "learning_rate": 3.943243776826834e-06,
+      "loss": 0.3434,
+      "step": 3941
+    },
+    {
+      "epoch": 1.8332041543946676,
+      "grad_norm": 0.3915799558162689,
+      "learning_rate": 3.94059912152218e-06,
+      "loss": 0.3332,
+      "step": 3942
+    },
+    {
+      "epoch": 1.8336691985738645,
+      "grad_norm": 0.4084991216659546,
+      "learning_rate": 3.937954776498839e-06,
+      "loss": 0.356,
+      "step": 3943
+    },
+    {
+      "epoch": 1.8341342427530616,
+      "grad_norm": 0.38015294075012207,
+      "learning_rate": 3.93531074253129e-06,
+      "loss": 0.3525,
+      "step": 3944
+    },
+    {
+      "epoch": 1.8345992869322587,
+      "grad_norm": 0.3744085133075714,
+      "learning_rate": 3.932667020393933e-06,
+      "loss": 0.3628,
+      "step": 3945
+    },
+    {
+      "epoch": 1.8350643311114556,
+      "grad_norm": 0.32951560616493225,
+      "learning_rate": 3.930023610861067e-06,
+      "loss": 0.3227,
+      "step": 3946
+    },
+    {
+      "epoch": 1.8355293752906525,
+      "grad_norm": 0.3684207499027252,
+      "learning_rate": 3.927380514706906e-06,
+      "loss": 0.3635,
+      "step": 3947
+    },
+    {
+      "epoch": 1.8359944194698496,
+      "grad_norm": 0.34596481919288635,
+      "learning_rate": 3.924737732705568e-06,
+      "loss": 0.3528,
+      "step": 3948
+    },
+    {
+      "epoch": 1.8364594636490468,
+      "grad_norm": 0.3639676868915558,
+      "learning_rate": 3.9220952656310855e-06,
+      "loss": 0.3836,
+      "step": 3949
+    },
+    {
+      "epoch": 1.8369245078282437,
+      "grad_norm": 0.34727898240089417,
+      "learning_rate": 3.919453114257389e-06,
+      "loss": 0.31,
+      "step": 3950
+    },
+    {
+      "epoch": 1.8373895520074406,
+      "grad_norm": 0.3681538701057434,
+      "learning_rate": 3.916811279358326e-06,
+      "loss": 0.3589,
+      "step": 3951
+    },
+    {
+      "epoch": 1.8378545961866377,
+      "grad_norm": 0.41221883893013,
+      "learning_rate": 3.9141697617076414e-06,
+      "loss": 0.3592,
+      "step": 3952
+    },
+    {
+      "epoch": 1.8383196403658348,
+      "grad_norm": 0.3793887794017792,
+      "learning_rate": 3.911528562078999e-06,
+      "loss": 0.3131,
+      "step": 3953
+    },
+    {
+      "epoch": 1.8387846845450317,
+      "grad_norm": 0.3928253650665283,
+      "learning_rate": 3.9088876812459585e-06,
+      "loss": 0.3305,
+      "step": 3954
+    },
+    {
+      "epoch": 1.8392497287242286,
+      "grad_norm": 0.4711274802684784,
+      "learning_rate": 3.906247119981995e-06,
+      "loss": 0.4163,
+      "step": 3955
+    },
+    {
+      "epoch": 1.8397147729034258,
+      "grad_norm": 0.3746742606163025,
+      "learning_rate": 3.903606879060483e-06,
+      "loss": 0.3223,
+      "step": 3956
+    },
+    {
+      "epoch": 1.840179817082623,
+      "grad_norm": 0.4172133207321167,
+      "learning_rate": 3.900966959254709e-06,
+      "loss": 0.3512,
+      "step": 3957
+    },
+    {
+      "epoch": 1.8406448612618198,
+      "grad_norm": 0.3593921959400177,
+      "learning_rate": 3.898327361337859e-06,
+      "loss": 0.3413,
+      "step": 3958
+    },
+    {
+      "epoch": 1.841109905441017,
+      "grad_norm": 0.40812554955482483,
+      "learning_rate": 3.89568808608303e-06,
+      "loss": 0.3318,
+      "step": 3959
+    },
+    {
+      "epoch": 1.841574949620214,
+      "grad_norm": 0.42885205149650574,
+      "learning_rate": 3.8930491342632235e-06,
+      "loss": 0.3515,
+      "step": 3960
+    },
+    {
+      "epoch": 1.842039993799411,
+      "grad_norm": 0.3699992597103119,
+      "learning_rate": 3.890410506651346e-06,
+      "loss": 0.3183,
+      "step": 3961
+    },
+    {
+      "epoch": 1.8425050379786079,
+      "grad_norm": 0.44218167662620544,
+      "learning_rate": 3.887772204020207e-06,
+      "loss": 0.3596,
+      "step": 3962
+    },
+    {
+      "epoch": 1.842970082157805,
+      "grad_norm": 0.38194793462753296,
+      "learning_rate": 3.885134227142525e-06,
+      "loss": 0.3587,
+      "step": 3963
+    },
+    {
+      "epoch": 1.8434351263370021,
+      "grad_norm": 0.3831635117530823,
+      "learning_rate": 3.882496576790918e-06,
+      "loss": 0.3255,
+      "step": 3964
+    },
+    {
+      "epoch": 1.843900170516199,
+      "grad_norm": 0.4331647455692291,
+      "learning_rate": 3.879859253737911e-06,
+      "loss": 0.3348,
+      "step": 3965
+    },
+    {
+      "epoch": 1.844365214695396,
+      "grad_norm": 0.41714081168174744,
+      "learning_rate": 3.8772222587559345e-06,
+      "loss": 0.317,
+      "step": 3966
+    },
+    {
+      "epoch": 1.844830258874593,
+      "grad_norm": 0.3952346444129944,
+      "learning_rate": 3.8745855926173205e-06,
+      "loss": 0.3647,
+      "step": 3967
+    },
+    {
+      "epoch": 1.8452953030537902,
+      "grad_norm": 0.38905858993530273,
+      "learning_rate": 3.871949256094308e-06,
+      "loss": 0.3791,
+      "step": 3968
+    },
+    {
+      "epoch": 1.845760347232987,
+      "grad_norm": 0.4492240250110626,
+      "learning_rate": 3.869313249959033e-06,
+      "loss": 0.3301,
+      "step": 3969
+    },
+    {
+      "epoch": 1.846225391412184,
+      "grad_norm": 0.4257234036922455,
+      "learning_rate": 3.866677574983542e-06,
+      "loss": 0.3827,
+      "step": 3970
+    },
+    {
+      "epoch": 1.8466904355913811,
+      "grad_norm": 0.3498407304286957,
+      "learning_rate": 3.86404223193978e-06,
+      "loss": 0.356,
+      "step": 3971
+    },
+    {
+      "epoch": 1.8471554797705783,
+      "grad_norm": 0.35863253474235535,
+      "learning_rate": 3.861407221599598e-06,
+      "loss": 0.3364,
+      "step": 3972
+    },
+    {
+      "epoch": 1.8476205239497752,
+      "grad_norm": 0.3656211495399475,
+      "learning_rate": 3.858772544734745e-06,
+      "loss": 0.3302,
+      "step": 3973
+    },
+    {
+      "epoch": 1.8480855681289723,
+      "grad_norm": 0.46760740876197815,
+      "learning_rate": 3.856138202116878e-06,
+      "loss": 0.3454,
+      "step": 3974
+    },
+    {
+      "epoch": 1.8485506123081694,
+      "grad_norm": 0.4070281684398651,
+      "learning_rate": 3.853504194517551e-06,
+      "loss": 0.3385,
+      "step": 3975
+    },
+    {
+      "epoch": 1.8490156564873663,
+      "grad_norm": 0.3659380376338959,
+      "learning_rate": 3.850870522708222e-06,
+      "loss": 0.3412,
+      "step": 3976
+    },
+    {
+      "epoch": 1.8494807006665632,
+      "grad_norm": 0.3773347735404968,
+      "learning_rate": 3.848237187460252e-06,
+      "loss": 0.3553,
+      "step": 3977
+    },
+    {
+      "epoch": 1.8499457448457604,
+      "grad_norm": 0.4317549765110016,
+      "learning_rate": 3.845604189544902e-06,
+      "loss": 0.3417,
+      "step": 3978
+    },
+    {
+      "epoch": 1.8504107890249575,
+      "grad_norm": 0.4554439187049866,
+      "learning_rate": 3.842971529733333e-06,
+      "loss": 0.3545,
+      "step": 3979
+    },
+    {
+      "epoch": 1.8508758332041544,
+      "grad_norm": 0.4062388241291046,
+      "learning_rate": 3.840339208796611e-06,
+      "loss": 0.3467,
+      "step": 3980
+    },
+    {
+      "epoch": 1.8513408773833513,
+      "grad_norm": 0.36868301033973694,
+      "learning_rate": 3.837707227505696e-06,
+      "loss": 0.3109,
+      "step": 3981
+    },
+    {
+      "epoch": 1.8518059215625484,
+      "grad_norm": 0.4157978594303131,
+      "learning_rate": 3.8350755866314555e-06,
+      "loss": 0.3758,
+      "step": 3982
+    },
+    {
+      "epoch": 1.8522709657417455,
+      "grad_norm": 0.4424905776977539,
+      "learning_rate": 3.8324442869446525e-06,
+      "loss": 0.3405,
+      "step": 3983
+    },
+    {
+      "epoch": 1.8527360099209425,
+      "grad_norm": 0.3770042657852173,
+      "learning_rate": 3.829813329215956e-06,
+      "loss": 0.3587,
+      "step": 3984
+    },
+    {
+      "epoch": 1.8532010541001394,
+      "grad_norm": 0.39475545287132263,
+      "learning_rate": 3.827182714215925e-06,
+      "loss": 0.3133,
+      "step": 3985
+    },
+    {
+      "epoch": 1.8536660982793365,
+      "grad_norm": 0.48087528347969055,
+      "learning_rate": 3.824552442715029e-06,
+      "loss": 0.3636,
+      "step": 3986
+    },
+    {
+      "epoch": 1.8541311424585336,
+      "grad_norm": 0.41496339440345764,
+      "learning_rate": 3.821922515483627e-06,
+      "loss": 0.3245,
+      "step": 3987
+    },
+    {
+      "epoch": 1.8545961866377305,
+      "grad_norm": 0.38482680916786194,
+      "learning_rate": 3.819292933291986e-06,
+      "loss": 0.3443,
+      "step": 3988
+    },
+    {
+      "epoch": 1.8550612308169276,
+      "grad_norm": 0.40696123242378235,
+      "learning_rate": 3.8166636969102655e-06,
+      "loss": 0.3456,
+      "step": 3989
+    },
+    {
+      "epoch": 1.8555262749961248,
+      "grad_norm": 0.37776339054107666,
+      "learning_rate": 3.814034807108529e-06,
+      "loss": 0.3619,
+      "step": 3990
+    },
+    {
+      "epoch": 1.8559913191753217,
+      "grad_norm": 0.40916046500205994,
+      "learning_rate": 3.8114062646567317e-06,
+      "loss": 0.3444,
+      "step": 3991
+    },
+    {
+      "epoch": 1.8564563633545186,
+      "grad_norm": 0.3734774589538574,
+      "learning_rate": 3.808778070324735e-06,
+      "loss": 0.341,
+      "step": 3992
+    },
+    {
+      "epoch": 1.8569214075337157,
+      "grad_norm": 0.38891860842704773,
+      "learning_rate": 3.80615022488229e-06,
+      "loss": 0.3358,
+      "step": 3993
+    },
+    {
+      "epoch": 1.8573864517129128,
+      "grad_norm": 0.40861591696739197,
+      "learning_rate": 3.803522729099054e-06,
+      "loss": 0.3432,
+      "step": 3994
+    },
+    {
+      "epoch": 1.8578514958921097,
+      "grad_norm": 0.45857205986976624,
+      "learning_rate": 3.8008955837445742e-06,
+      "loss": 0.3707,
+      "step": 3995
+    },
+    {
+      "epoch": 1.8583165400713066,
+      "grad_norm": 0.38723862171173096,
+      "learning_rate": 3.7982687895883036e-06,
+      "loss": 0.3699,
+      "step": 3996
+    },
+    {
+      "epoch": 1.8587815842505038,
+      "grad_norm": 0.37383538484573364,
+      "learning_rate": 3.795642347399582e-06,
+      "loss": 0.3368,
+      "step": 3997
+    },
+    {
+      "epoch": 1.859246628429701,
+      "grad_norm": 0.38865286111831665,
+      "learning_rate": 3.7930162579476566e-06,
+      "loss": 0.3678,
+      "step": 3998
+    },
+    {
+      "epoch": 1.8597116726088978,
+      "grad_norm": 0.42665407061576843,
+      "learning_rate": 3.790390522001662e-06,
+      "loss": 0.3673,
+      "step": 3999
+    },
+    {
+      "epoch": 1.8601767167880947,
+      "grad_norm": 0.42146944999694824,
+      "learning_rate": 3.787765140330636e-06,
+      "loss": 0.3535,
+      "step": 4000
+    },
+    {
+      "epoch": 1.860641760967292,
+      "grad_norm": 0.36168861389160156,
+      "learning_rate": 3.7851401137035114e-06,
+      "loss": 0.3267,
+      "step": 4001
+    },
+    {
+      "epoch": 1.861106805146489,
+      "grad_norm": 0.4150577485561371,
+      "learning_rate": 3.782515442889112e-06,
+      "loss": 0.3572,
+      "step": 4002
+    },
+    {
+      "epoch": 1.8615718493256859,
+      "grad_norm": 0.4027031362056732,
+      "learning_rate": 3.7798911286561655e-06,
+      "loss": 0.3416,
+      "step": 4003
+    },
+    {
+      "epoch": 1.862036893504883,
+      "grad_norm": 0.4184225797653198,
+      "learning_rate": 3.777267171773288e-06,
+      "loss": 0.38,
+      "step": 4004
+    },
+    {
+      "epoch": 1.8625019376840801,
+      "grad_norm": 0.3823773264884949,
+      "learning_rate": 3.774643573008995e-06,
+      "loss": 0.2838,
+      "step": 4005
+    },
+    {
+      "epoch": 1.862966981863277,
+      "grad_norm": 0.4209512770175934,
+      "learning_rate": 3.7720203331316946e-06,
+      "loss": 0.3812,
+      "step": 4006
+    },
+    {
+      "epoch": 1.863432026042474,
+      "grad_norm": 0.37146735191345215,
+      "learning_rate": 3.769397452909695e-06,
+      "loss": 0.3212,
+      "step": 4007
+    },
+    {
+      "epoch": 1.863897070221671,
+      "grad_norm": 0.334895521402359,
+      "learning_rate": 3.76677493311119e-06,
+      "loss": 0.3148,
+      "step": 4008
+    },
+    {
+      "epoch": 1.8643621144008682,
+      "grad_norm": 0.372598797082901,
+      "learning_rate": 3.7641527745042784e-06,
+      "loss": 0.3257,
+      "step": 4009
+    },
+    {
+      "epoch": 1.864827158580065,
+      "grad_norm": 0.4964183270931244,
+      "learning_rate": 3.7615309778569427e-06,
+      "loss": 0.3659,
+      "step": 4010
+    },
+    {
+      "epoch": 1.865292202759262,
+      "grad_norm": 0.37151283025741577,
+      "learning_rate": 3.7589095439370676e-06,
+      "loss": 0.3561,
+      "step": 4011
+    },
+    {
+      "epoch": 1.8657572469384591,
+      "grad_norm": 0.43840986490249634,
+      "learning_rate": 3.7562884735124273e-06,
+      "loss": 0.3004,
+      "step": 4012
+    },
+    {
+      "epoch": 1.8662222911176563,
+      "grad_norm": 0.4607786536216736,
+      "learning_rate": 3.7536677673506926e-06,
+      "loss": 0.4157,
+      "step": 4013
+    },
+    {
+      "epoch": 1.8666873352968532,
+      "grad_norm": 0.3740638792514801,
+      "learning_rate": 3.751047426219423e-06,
+      "loss": 0.3024,
+      "step": 4014
+    },
+    {
+      "epoch": 1.86715237947605,
+      "grad_norm": 0.41392412781715393,
+      "learning_rate": 3.7484274508860776e-06,
+      "loss": 0.3458,
+      "step": 4015
+    },
+    {
+      "epoch": 1.8676174236552474,
+      "grad_norm": 0.46794068813323975,
+      "learning_rate": 3.745807842118e-06,
+      "loss": 0.3345,
+      "step": 4016
+    },
+    {
+      "epoch": 1.8680824678344443,
+      "grad_norm": 0.400758296251297,
+      "learning_rate": 3.7431886006824347e-06,
+      "loss": 0.339,
+      "step": 4017
+    },
+    {
+      "epoch": 1.8685475120136412,
+      "grad_norm": 0.3659023642539978,
+      "learning_rate": 3.7405697273465125e-06,
+      "loss": 0.3756,
+      "step": 4018
+    },
+    {
+      "epoch": 1.8690125561928383,
+      "grad_norm": 0.4087202548980713,
+      "learning_rate": 3.7379512228772618e-06,
+      "loss": 0.3622,
+      "step": 4019
+    },
+    {
+      "epoch": 1.8694776003720355,
+      "grad_norm": 0.3660109341144562,
+      "learning_rate": 3.7353330880415963e-06,
+      "loss": 0.3189,
+      "step": 4020
+    },
+    {
+      "epoch": 1.8699426445512324,
+      "grad_norm": 0.3783299922943115,
+      "learning_rate": 3.7327153236063295e-06,
+      "loss": 0.3218,
+      "step": 4021
+    },
+    {
+      "epoch": 1.8704076887304293,
+      "grad_norm": 0.3811236023902893,
+      "learning_rate": 3.7300979303381576e-06,
+      "loss": 0.3672,
+      "step": 4022
+    },
+    {
+      "epoch": 1.8708727329096264,
+      "grad_norm": 0.3874422013759613,
+      "learning_rate": 3.7274809090036757e-06,
+      "loss": 0.377,
+      "step": 4023
+    },
+    {
+      "epoch": 1.8713377770888235,
+      "grad_norm": 0.40507352352142334,
+      "learning_rate": 3.724864260369364e-06,
+      "loss": 0.3598,
+      "step": 4024
+    },
+    {
+      "epoch": 1.8718028212680204,
+      "grad_norm": 0.3792964220046997,
+      "learning_rate": 3.7222479852016015e-06,
+      "loss": 0.3304,
+      "step": 4025
+    },
+    {
+      "epoch": 1.8722678654472173,
+      "grad_norm": 0.3633670210838318,
+      "learning_rate": 3.7196320842666467e-06,
+      "loss": 0.2997,
+      "step": 4026
+    },
+    {
+      "epoch": 1.8727329096264145,
+      "grad_norm": 0.3763364851474762,
+      "learning_rate": 3.7170165583306595e-06,
+      "loss": 0.3596,
+      "step": 4027
+    },
+    {
+      "epoch": 1.8731979538056116,
+      "grad_norm": 0.3550674617290497,
+      "learning_rate": 3.71440140815968e-06,
+      "loss": 0.3433,
+      "step": 4028
+    },
+    {
+      "epoch": 1.8736629979848085,
+      "grad_norm": 0.3978072702884674,
+      "learning_rate": 3.7117866345196473e-06,
+      "loss": 0.3898,
+      "step": 4029
+    },
+    {
+      "epoch": 1.8741280421640054,
+      "grad_norm": 0.417102575302124,
+      "learning_rate": 3.709172238176384e-06,
+      "loss": 0.337,
+      "step": 4030
+    },
+    {
+      "epoch": 1.8745930863432028,
+      "grad_norm": 0.36194002628326416,
+      "learning_rate": 3.706558219895607e-06,
+      "loss": 0.3347,
+      "step": 4031
+    },
+    {
+      "epoch": 1.8750581305223997,
+      "grad_norm": 0.39445623755455017,
+      "learning_rate": 3.7039445804429154e-06,
+      "loss": 0.3578,
+      "step": 4032
+    },
+    {
+      "epoch": 1.8755231747015966,
+      "grad_norm": 0.3856761157512665,
+      "learning_rate": 3.7013313205838066e-06,
+      "loss": 0.3493,
+      "step": 4033
+    },
+    {
+      "epoch": 1.8759882188807937,
+      "grad_norm": 0.3411782383918762,
+      "learning_rate": 3.698718441083657e-06,
+      "loss": 0.3297,
+      "step": 4034
+    },
+    {
+      "epoch": 1.8764532630599908,
+      "grad_norm": 0.421271950006485,
+      "learning_rate": 3.6961059427077407e-06,
+      "loss": 0.3634,
+      "step": 4035
+    },
+    {
+      "epoch": 1.8769183072391877,
+      "grad_norm": 0.4099300801753998,
+      "learning_rate": 3.693493826221215e-06,
+      "loss": 0.3815,
+      "step": 4036
+    },
+    {
+      "epoch": 1.8773833514183846,
+      "grad_norm": 0.37041401863098145,
+      "learning_rate": 3.6908820923891235e-06,
+      "loss": 0.3621,
+      "step": 4037
+    },
+    {
+      "epoch": 1.8778483955975818,
+      "grad_norm": 0.36186453700065613,
+      "learning_rate": 3.6882707419764053e-06,
+      "loss": 0.3328,
+      "step": 4038
+    },
+    {
+      "epoch": 1.878313439776779,
+      "grad_norm": 0.38483038544654846,
+      "learning_rate": 3.6856597757478784e-06,
+      "loss": 0.3676,
+      "step": 4039
+    },
+    {
+      "epoch": 1.8787784839559758,
+      "grad_norm": 0.3237282335758209,
+      "learning_rate": 3.6830491944682543e-06,
+      "loss": 0.3129,
+      "step": 4040
+    },
+    {
+      "epoch": 1.8792435281351727,
+      "grad_norm": 0.3765997588634491,
+      "learning_rate": 3.6804389989021292e-06,
+      "loss": 0.3639,
+      "step": 4041
+    },
+    {
+      "epoch": 1.8797085723143698,
+      "grad_norm": 0.3259957432746887,
+      "learning_rate": 3.6778291898139907e-06,
+      "loss": 0.3183,
+      "step": 4042
+    },
+    {
+      "epoch": 1.880173616493567,
+      "grad_norm": 0.37962499260902405,
+      "learning_rate": 3.675219767968203e-06,
+      "loss": 0.3432,
+      "step": 4043
+    },
+    {
+      "epoch": 1.8806386606727639,
+      "grad_norm": 0.34864094853401184,
+      "learning_rate": 3.6726107341290285e-06,
+      "loss": 0.3045,
+      "step": 4044
+    },
+    {
+      "epoch": 1.881103704851961,
+      "grad_norm": 0.3533942997455597,
+      "learning_rate": 3.6700020890606068e-06,
+      "loss": 0.3542,
+      "step": 4045
+    },
+    {
+      "epoch": 1.8815687490311581,
+      "grad_norm": 0.351633220911026,
+      "learning_rate": 3.667393833526972e-06,
+      "loss": 0.3474,
+      "step": 4046
+    },
+    {
+      "epoch": 1.882033793210355,
+      "grad_norm": 0.3609406650066376,
+      "learning_rate": 3.664785968292036e-06,
+      "loss": 0.3546,
+      "step": 4047
+    },
+    {
+      "epoch": 1.882498837389552,
+      "grad_norm": 0.3869522511959076,
+      "learning_rate": 3.6621784941196036e-06,
+      "loss": 0.3715,
+      "step": 4048
+    },
+    {
+      "epoch": 1.882963881568749,
+      "grad_norm": 0.36395883560180664,
+      "learning_rate": 3.6595714117733583e-06,
+      "loss": 0.3602,
+      "step": 4049
+    },
+    {
+      "epoch": 1.8834289257479462,
+      "grad_norm": 0.36276188492774963,
+      "learning_rate": 3.656964722016875e-06,
+      "loss": 0.3046,
+      "step": 4050
+    },
+    {
+      "epoch": 1.883893969927143,
+      "grad_norm": 0.3588871955871582,
+      "learning_rate": 3.6543584256136076e-06,
+      "loss": 0.3384,
+      "step": 4051
+    },
+    {
+      "epoch": 1.88435901410634,
+      "grad_norm": 0.39765575528144836,
+      "learning_rate": 3.6517525233269015e-06,
+      "loss": 0.3879,
+      "step": 4052
+    },
+    {
+      "epoch": 1.8848240582855371,
+      "grad_norm": 0.34334567189216614,
+      "learning_rate": 3.6491470159199806e-06,
+      "loss": 0.316,
+      "step": 4053
+    },
+    {
+      "epoch": 1.8852891024647342,
+      "grad_norm": 0.3806763291358948,
+      "learning_rate": 3.646541904155958e-06,
+      "loss": 0.3474,
+      "step": 4054
+    },
+    {
+      "epoch": 1.8857541466439312,
+      "grad_norm": 0.3665696978569031,
+      "learning_rate": 3.643937188797826e-06,
+      "loss": 0.3653,
+      "step": 4055
+    },
+    {
+      "epoch": 1.886219190823128,
+      "grad_norm": 0.3573841154575348,
+      "learning_rate": 3.641332870608466e-06,
+      "loss": 0.3426,
+      "step": 4056
+    },
+    {
+      "epoch": 1.8866842350023252,
+      "grad_norm": 0.3744082450866699,
+      "learning_rate": 3.6387289503506375e-06,
+      "loss": 0.3397,
+      "step": 4057
+    },
+    {
+      "epoch": 1.8871492791815223,
+      "grad_norm": 0.37690484523773193,
+      "learning_rate": 3.6361254287869886e-06,
+      "loss": 0.3628,
+      "step": 4058
+    },
+    {
+      "epoch": 1.8876143233607192,
+      "grad_norm": 0.34452709555625916,
+      "learning_rate": 3.6335223066800466e-06,
+      "loss": 0.3244,
+      "step": 4059
+    },
+    {
+      "epoch": 1.8880793675399163,
+      "grad_norm": 0.4054635167121887,
+      "learning_rate": 3.6309195847922284e-06,
+      "loss": 0.3616,
+      "step": 4060
+    },
+    {
+      "epoch": 1.8885444117191135,
+      "grad_norm": 0.40721654891967773,
+      "learning_rate": 3.628317263885823e-06,
+      "loss": 0.3951,
+      "step": 4061
+    },
+    {
+      "epoch": 1.8890094558983104,
+      "grad_norm": 0.3469128906726837,
+      "learning_rate": 3.625715344723012e-06,
+      "loss": 0.2801,
+      "step": 4062
+    },
+    {
+      "epoch": 1.8894745000775073,
+      "grad_norm": 0.35594481229782104,
+      "learning_rate": 3.623113828065853e-06,
+      "loss": 0.3517,
+      "step": 4063
+    },
+    {
+      "epoch": 1.8899395442567044,
+      "grad_norm": 0.3937917947769165,
+      "learning_rate": 3.6205127146762885e-06,
+      "loss": 0.3679,
+      "step": 4064
+    },
+    {
+      "epoch": 1.8904045884359015,
+      "grad_norm": 0.3829907476902008,
+      "learning_rate": 3.617912005316142e-06,
+      "loss": 0.3391,
+      "step": 4065
+    },
+    {
+      "epoch": 1.8908696326150984,
+      "grad_norm": 0.36475563049316406,
+      "learning_rate": 3.615311700747122e-06,
+      "loss": 0.335,
+      "step": 4066
+    },
+    {
+      "epoch": 1.8913346767942953,
+      "grad_norm": 0.3677665591239929,
+      "learning_rate": 3.6127118017308116e-06,
+      "loss": 0.3373,
+      "step": 4067
+    },
+    {
+      "epoch": 1.8917997209734925,
+      "grad_norm": 0.3832724094390869,
+      "learning_rate": 3.6101123090286814e-06,
+      "loss": 0.3258,
+      "step": 4068
+    },
+    {
+      "epoch": 1.8922647651526896,
+      "grad_norm": 0.3629043698310852,
+      "learning_rate": 3.607513223402078e-06,
+      "loss": 0.3766,
+      "step": 4069
+    },
+    {
+      "epoch": 1.8927298093318865,
+      "grad_norm": 0.3398493826389313,
+      "learning_rate": 3.6049145456122347e-06,
+      "loss": 0.3425,
+      "step": 4070
+    },
+    {
+      "epoch": 1.8931948535110834,
+      "grad_norm": 0.36889976263046265,
+      "learning_rate": 3.6023162764202613e-06,
+      "loss": 0.3559,
+      "step": 4071
+    },
+    {
+      "epoch": 1.8936598976902805,
+      "grad_norm": 0.393019437789917,
+      "learning_rate": 3.599718416587146e-06,
+      "loss": 0.3255,
+      "step": 4072
+    },
+    {
+      "epoch": 1.8941249418694777,
+      "grad_norm": 0.401734322309494,
+      "learning_rate": 3.5971209668737626e-06,
+      "loss": 0.3756,
+      "step": 4073
+    },
+    {
+      "epoch": 1.8945899860486746,
+      "grad_norm": 0.34484103322029114,
+      "learning_rate": 3.5945239280408596e-06,
+      "loss": 0.321,
+      "step": 4074
+    },
+    {
+      "epoch": 1.8950550302278717,
+      "grad_norm": 0.41480597853660583,
+      "learning_rate": 3.591927300849069e-06,
+      "loss": 0.3551,
+      "step": 4075
+    },
+    {
+      "epoch": 1.8955200744070688,
+      "grad_norm": 0.3733346462249756,
+      "learning_rate": 3.5893310860588997e-06,
+      "loss": 0.3334,
+      "step": 4076
+    },
+    {
+      "epoch": 1.8959851185862657,
+      "grad_norm": 0.3196105360984802,
+      "learning_rate": 3.5867352844307433e-06,
+      "loss": 0.3302,
+      "step": 4077
+    },
+    {
+      "epoch": 1.8964501627654626,
+      "grad_norm": 0.3849073052406311,
+      "learning_rate": 3.5841398967248654e-06,
+      "loss": 0.3548,
+      "step": 4078
+    },
+    {
+      "epoch": 1.8969152069446598,
+      "grad_norm": 0.38783520460128784,
+      "learning_rate": 3.5815449237014144e-06,
+      "loss": 0.3313,
+      "step": 4079
+    },
+    {
+      "epoch": 1.8973802511238569,
+      "grad_norm": 0.3971535563468933,
+      "learning_rate": 3.578950366120414e-06,
+      "loss": 0.3581,
+      "step": 4080
+    },
+    {
+      "epoch": 1.8978452953030538,
+      "grad_norm": 0.33792009949684143,
+      "learning_rate": 3.5763562247417694e-06,
+      "loss": 0.3269,
+      "step": 4081
+    },
+    {
+      "epoch": 1.8983103394822507,
+      "grad_norm": 0.36532509326934814,
+      "learning_rate": 3.5737625003252606e-06,
+      "loss": 0.3342,
+      "step": 4082
+    },
+    {
+      "epoch": 1.8987753836614478,
+      "grad_norm": 0.36312729120254517,
+      "learning_rate": 3.5711691936305522e-06,
+      "loss": 0.3322,
+      "step": 4083
+    },
+    {
+      "epoch": 1.899240427840645,
+      "grad_norm": 0.36655622720718384,
+      "learning_rate": 3.568576305417175e-06,
+      "loss": 0.3621,
+      "step": 4084
+    },
+    {
+      "epoch": 1.8997054720198419,
+      "grad_norm": 0.38222381472587585,
+      "learning_rate": 3.5659838364445505e-06,
+      "loss": 0.3598,
+      "step": 4085
+    },
+    {
+      "epoch": 1.9001705161990388,
+      "grad_norm": 0.356871098279953,
+      "learning_rate": 3.5633917874719642e-06,
+      "loss": 0.3531,
+      "step": 4086
+    },
+    {
+      "epoch": 1.9006355603782359,
+      "grad_norm": 0.34935128688812256,
+      "learning_rate": 3.5608001592585895e-06,
+      "loss": 0.3085,
+      "step": 4087
+    },
+    {
+      "epoch": 1.901100604557433,
+      "grad_norm": 0.3709997236728668,
+      "learning_rate": 3.55820895256347e-06,
+      "loss": 0.3215,
+      "step": 4088
+    },
+    {
+      "epoch": 1.90156564873663,
+      "grad_norm": 0.39799949526786804,
+      "learning_rate": 3.5556181681455314e-06,
+      "loss": 0.3538,
+      "step": 4089
+    },
+    {
+      "epoch": 1.902030692915827,
+      "grad_norm": 0.35450243949890137,
+      "learning_rate": 3.553027806763568e-06,
+      "loss": 0.3455,
+      "step": 4090
+    },
+    {
+      "epoch": 1.9024957370950242,
+      "grad_norm": 0.43028944730758667,
+      "learning_rate": 3.5504378691762586e-06,
+      "loss": 0.3607,
+      "step": 4091
+    },
+    {
+      "epoch": 1.902960781274221,
+      "grad_norm": 0.35662081837654114,
+      "learning_rate": 3.5478483561421497e-06,
+      "loss": 0.312,
+      "step": 4092
+    },
+    {
+      "epoch": 1.903425825453418,
+      "grad_norm": 0.38295355439186096,
+      "learning_rate": 3.5452592684196707e-06,
+      "loss": 0.3484,
+      "step": 4093
+    },
+    {
+      "epoch": 1.9038908696326151,
+      "grad_norm": 0.38135382533073425,
+      "learning_rate": 3.542670606767121e-06,
+      "loss": 0.3646,
+      "step": 4094
+    },
+    {
+      "epoch": 1.9043559138118122,
+      "grad_norm": 0.3967789113521576,
+      "learning_rate": 3.540082371942682e-06,
+      "loss": 0.3255,
+      "step": 4095
+    },
+    {
+      "epoch": 1.9048209579910091,
+      "grad_norm": 0.4134833514690399,
+      "learning_rate": 3.5374945647044e-06,
+      "loss": 0.3487,
+      "step": 4096
+    },
+    {
+      "epoch": 1.905286002170206,
+      "grad_norm": 0.36041533946990967,
+      "learning_rate": 3.5349071858102056e-06,
+      "loss": 0.3114,
+      "step": 4097
+    },
+    {
+      "epoch": 1.9057510463494032,
+      "grad_norm": 0.40801483392715454,
+      "learning_rate": 3.5323202360178976e-06,
+      "loss": 0.3595,
+      "step": 4098
+    },
+    {
+      "epoch": 1.9062160905286003,
+      "grad_norm": 0.39261582493782043,
+      "learning_rate": 3.529733716085154e-06,
+      "loss": 0.3617,
+      "step": 4099
+    },
+    {
+      "epoch": 1.9066811347077972,
+      "grad_norm": 0.3464079797267914,
+      "learning_rate": 3.5271476267695216e-06,
+      "loss": 0.3262,
+      "step": 4100
+    },
+    {
+      "epoch": 1.9071461788869941,
+      "grad_norm": 0.461972177028656,
+      "learning_rate": 3.5245619688284277e-06,
+      "loss": 0.3115,
+      "step": 4101
+    },
+    {
+      "epoch": 1.9076112230661912,
+      "grad_norm": 0.35122260451316833,
+      "learning_rate": 3.5219767430191653e-06,
+      "loss": 0.3377,
+      "step": 4102
+    },
+    {
+      "epoch": 1.9080762672453884,
+      "grad_norm": 0.4606049060821533,
+      "learning_rate": 3.5193919500989093e-06,
+      "loss": 0.3413,
+      "step": 4103
+    },
+    {
+      "epoch": 1.9085413114245853,
+      "grad_norm": 0.40791839361190796,
+      "learning_rate": 3.516807590824699e-06,
+      "loss": 0.3605,
+      "step": 4104
+    },
+    {
+      "epoch": 1.9090063556037824,
+      "grad_norm": 0.3770095407962799,
+      "learning_rate": 3.514223665953455e-06,
+      "loss": 0.3335,
+      "step": 4105
+    },
+    {
+      "epoch": 1.9094713997829795,
+      "grad_norm": 0.3994191884994507,
+      "learning_rate": 3.5116401762419643e-06,
+      "loss": 0.2999,
+      "step": 4106
+    },
+    {
+      "epoch": 1.9099364439621764,
+      "grad_norm": 0.4128780663013458,
+      "learning_rate": 3.509057122446893e-06,
+      "loss": 0.3623,
+      "step": 4107
+    },
+    {
+      "epoch": 1.9104014881413733,
+      "grad_norm": 0.39665135741233826,
+      "learning_rate": 3.506474505324772e-06,
+      "loss": 0.3586,
+      "step": 4108
+    },
+    {
+      "epoch": 1.9108665323205705,
+      "grad_norm": 0.3825541138648987,
+      "learning_rate": 3.503892325632007e-06,
+      "loss": 0.3502,
+      "step": 4109
+    },
+    {
+      "epoch": 1.9113315764997676,
+      "grad_norm": 0.3846040368080139,
+      "learning_rate": 3.5013105841248794e-06,
+      "loss": 0.3397,
+      "step": 4110
+    },
+    {
+      "epoch": 1.9117966206789645,
+      "grad_norm": 0.4117841422557831,
+      "learning_rate": 3.4987292815595376e-06,
+      "loss": 0.3676,
+      "step": 4111
+    },
+    {
+      "epoch": 1.9122616648581614,
+      "grad_norm": 0.38701215386390686,
+      "learning_rate": 3.496148418692006e-06,
+      "loss": 0.3433,
+      "step": 4112
+    },
+    {
+      "epoch": 1.9127267090373585,
+      "grad_norm": 0.3598323464393616,
+      "learning_rate": 3.4935679962781722e-06,
+      "loss": 0.346,
+      "step": 4113
+    },
+    {
+      "epoch": 1.9131917532165557,
+      "grad_norm": 0.40813809633255005,
+      "learning_rate": 3.4909880150738057e-06,
+      "loss": 0.3356,
+      "step": 4114
+    },
+    {
+      "epoch": 1.9136567973957526,
+      "grad_norm": 0.3636961579322815,
+      "learning_rate": 3.4884084758345365e-06,
+      "loss": 0.3225,
+      "step": 4115
+    },
+    {
+      "epoch": 1.9141218415749495,
+      "grad_norm": 0.3554372489452362,
+      "learning_rate": 3.4858293793158727e-06,
+      "loss": 0.3349,
+      "step": 4116
+    },
+    {
+      "epoch": 1.9145868857541466,
+      "grad_norm": 0.4285871684551239,
+      "learning_rate": 3.4832507262731876e-06,
+      "loss": 0.3442,
+      "step": 4117
+    },
+    {
+      "epoch": 1.9150519299333437,
+      "grad_norm": 0.38151922821998596,
+      "learning_rate": 3.4806725174617305e-06,
+      "loss": 0.3202,
+      "step": 4118
+    },
+    {
+      "epoch": 1.9155169741125406,
+      "grad_norm": 0.4228741228580475,
+      "learning_rate": 3.4780947536366115e-06,
+      "loss": 0.3516,
+      "step": 4119
+    },
+    {
+      "epoch": 1.9159820182917378,
+      "grad_norm": 0.3742198944091797,
+      "learning_rate": 3.4755174355528214e-06,
+      "loss": 0.3673,
+      "step": 4120
+    },
+    {
+      "epoch": 1.9164470624709349,
+      "grad_norm": 0.3714161217212677,
+      "learning_rate": 3.4729405639652102e-06,
+      "loss": 0.3152,
+      "step": 4121
+    },
+    {
+      "epoch": 1.9169121066501318,
+      "grad_norm": 0.3599150478839874,
+      "learning_rate": 3.470364139628504e-06,
+      "loss": 0.3102,
+      "step": 4122
+    },
+    {
+      "epoch": 1.9173771508293287,
+      "grad_norm": 0.4052508771419525,
+      "learning_rate": 3.467788163297294e-06,
+      "loss": 0.3409,
+      "step": 4123
+    },
+    {
+      "epoch": 1.9178421950085258,
+      "grad_norm": 0.39652350544929504,
+      "learning_rate": 3.465212635726045e-06,
+      "loss": 0.4028,
+      "step": 4124
+    },
+    {
+      "epoch": 1.918307239187723,
+      "grad_norm": 0.3475923240184784,
+      "learning_rate": 3.462637557669084e-06,
+      "loss": 0.3194,
+      "step": 4125
+    },
+    {
+      "epoch": 1.9187722833669199,
+      "grad_norm": 0.35219481587409973,
+      "learning_rate": 3.460062929880612e-06,
+      "loss": 0.3094,
+      "step": 4126
+    },
+    {
+      "epoch": 1.9192373275461168,
+      "grad_norm": 0.3954284191131592,
+      "learning_rate": 3.4574887531146926e-06,
+      "loss": 0.3428,
+      "step": 4127
+    },
+    {
+      "epoch": 1.9197023717253139,
+      "grad_norm": 0.4456411600112915,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.3701,
+      "step": 4128
+    },
+    {
+      "epoch": 1.920167415904511,
+      "grad_norm": 0.3940032720565796,
+      "learning_rate": 3.4523417556661244e-06,
+      "loss": 0.3499,
+      "step": 4129
+    },
+    {
+      "epoch": 1.920632460083708,
+      "grad_norm": 0.36356043815612793,
+      "learning_rate": 3.4497689364909483e-06,
+      "loss": 0.362,
+      "step": 4130
+    },
+    {
+      "epoch": 1.9210975042629048,
+      "grad_norm": 0.34411749243736267,
+      "learning_rate": 3.4471965713532675e-06,
+      "loss": 0.3274,
+      "step": 4131
+    },
+    {
+      "epoch": 1.921562548442102,
+      "grad_norm": 0.36001211404800415,
+      "learning_rate": 3.444624661006491e-06,
+      "loss": 0.3625,
+      "step": 4132
+    },
+    {
+      "epoch": 1.922027592621299,
+      "grad_norm": 0.3642420172691345,
+      "learning_rate": 3.4420532062038846e-06,
+      "loss": 0.3257,
+      "step": 4133
+    },
+    {
+      "epoch": 1.922492636800496,
+      "grad_norm": 0.3693174719810486,
+      "learning_rate": 3.43948220769859e-06,
+      "loss": 0.3414,
+      "step": 4134
+    },
+    {
+      "epoch": 1.922957680979693,
+      "grad_norm": 0.4484702944755554,
+      "learning_rate": 3.4369116662436074e-06,
+      "loss": 0.3864,
+      "step": 4135
+    },
+    {
+      "epoch": 1.9234227251588902,
+      "grad_norm": 0.3540983200073242,
+      "learning_rate": 3.4343415825918102e-06,
+      "loss": 0.3052,
+      "step": 4136
+    },
+    {
+      "epoch": 1.9238877693380871,
+      "grad_norm": 0.35332098603248596,
+      "learning_rate": 3.4317719574959307e-06,
+      "loss": 0.3524,
+      "step": 4137
+    },
+    {
+      "epoch": 1.924352813517284,
+      "grad_norm": 0.41718167066574097,
+      "learning_rate": 3.4292027917085733e-06,
+      "loss": 0.3519,
+      "step": 4138
+    },
+    {
+      "epoch": 1.9248178576964812,
+      "grad_norm": 0.42364948987960815,
+      "learning_rate": 3.4266340859822023e-06,
+      "loss": 0.3231,
+      "step": 4139
+    },
+    {
+      "epoch": 1.9252829018756783,
+      "grad_norm": 0.38962802290916443,
+      "learning_rate": 3.424065841069152e-06,
+      "loss": 0.3384,
+      "step": 4140
+    },
+    {
+      "epoch": 1.9257479460548752,
+      "grad_norm": 0.37895727157592773,
+      "learning_rate": 3.421498057721617e-06,
+      "loss": 0.3477,
+      "step": 4141
+    },
+    {
+      "epoch": 1.926212990234072,
+      "grad_norm": 0.40839117765426636,
+      "learning_rate": 3.4189307366916635e-06,
+      "loss": 0.3364,
+      "step": 4142
+    },
+    {
+      "epoch": 1.9266780344132692,
+      "grad_norm": 0.39300408959388733,
+      "learning_rate": 3.4163638787312152e-06,
+      "loss": 0.3477,
+      "step": 4143
+    },
+    {
+      "epoch": 1.9271430785924664,
+      "grad_norm": 0.39907246828079224,
+      "learning_rate": 3.4137974845920616e-06,
+      "loss": 0.3376,
+      "step": 4144
+    },
+    {
+      "epoch": 1.9276081227716633,
+      "grad_norm": 0.38821595907211304,
+      "learning_rate": 3.411231555025861e-06,
+      "loss": 0.3703,
+      "step": 4145
+    },
+    {
+      "epoch": 1.9280731669508602,
+      "grad_norm": 0.3863977789878845,
+      "learning_rate": 3.4086660907841307e-06,
+      "loss": 0.3541,
+      "step": 4146
+    },
+    {
+      "epoch": 1.9285382111300573,
+      "grad_norm": 0.36651626229286194,
+      "learning_rate": 3.4061010926182557e-06,
+      "loss": 0.3067,
+      "step": 4147
+    },
+    {
+      "epoch": 1.9290032553092544,
+      "grad_norm": 0.4503706693649292,
+      "learning_rate": 3.403536561279479e-06,
+      "loss": 0.3646,
+      "step": 4148
+    },
+    {
+      "epoch": 1.9294682994884513,
+      "grad_norm": 0.417574942111969,
+      "learning_rate": 3.400972497518914e-06,
+      "loss": 0.3605,
+      "step": 4149
+    },
+    {
+      "epoch": 1.9299333436676485,
+      "grad_norm": 0.3759341239929199,
+      "learning_rate": 3.398408902087529e-06,
+      "loss": 0.3452,
+      "step": 4150
+    },
+    {
+      "epoch": 1.9303983878468456,
+      "grad_norm": 0.36154696345329285,
+      "learning_rate": 3.395845775736163e-06,
+      "loss": 0.3387,
+      "step": 4151
+    },
+    {
+      "epoch": 1.9308634320260425,
+      "grad_norm": 0.35659950971603394,
+      "learning_rate": 3.3932831192155115e-06,
+      "loss": 0.331,
+      "step": 4152
+    },
+    {
+      "epoch": 1.9313284762052394,
+      "grad_norm": 0.36559152603149414,
+      "learning_rate": 3.3907209332761383e-06,
+      "loss": 0.3622,
+      "step": 4153
+    },
+    {
+      "epoch": 1.9317935203844365,
+      "grad_norm": 0.365607887506485,
+      "learning_rate": 3.3881592186684616e-06,
+      "loss": 0.3181,
+      "step": 4154
+    },
+    {
+      "epoch": 1.9322585645636337,
+      "grad_norm": 0.3874776065349579,
+      "learning_rate": 3.3855979761427705e-06,
+      "loss": 0.3239,
+      "step": 4155
+    },
+    {
+      "epoch": 1.9327236087428306,
+      "grad_norm": 0.3716128170490265,
+      "learning_rate": 3.383037206449207e-06,
+      "loss": 0.3572,
+      "step": 4156
+    },
+    {
+      "epoch": 1.9331886529220275,
+      "grad_norm": 0.38972359895706177,
+      "learning_rate": 3.3804769103377827e-06,
+      "loss": 0.3569,
+      "step": 4157
+    },
+    {
+      "epoch": 1.9336536971012246,
+      "grad_norm": 0.3612194061279297,
+      "learning_rate": 3.377917088558364e-06,
+      "loss": 0.3391,
+      "step": 4158
+    },
+    {
+      "epoch": 1.9341187412804217,
+      "grad_norm": 0.3686378002166748,
+      "learning_rate": 3.3753577418606844e-06,
+      "loss": 0.3359,
+      "step": 4159
+    },
+    {
+      "epoch": 1.9345837854596186,
+      "grad_norm": 0.35202595591545105,
+      "learning_rate": 3.3727988709943303e-06,
+      "loss": 0.3094,
+      "step": 4160
+    },
+    {
+      "epoch": 1.9350488296388155,
+      "grad_norm": 0.3890860676765442,
+      "learning_rate": 3.370240476708759e-06,
+      "loss": 0.346,
+      "step": 4161
+    },
+    {
+      "epoch": 1.9355138738180129,
+      "grad_norm": 0.3682956397533417,
+      "learning_rate": 3.367682559753277e-06,
+      "loss": 0.3363,
+      "step": 4162
+    },
+    {
+      "epoch": 1.9359789179972098,
+      "grad_norm": 0.37430447340011597,
+      "learning_rate": 3.36512512087706e-06,
+      "loss": 0.335,
+      "step": 4163
+    },
+    {
+      "epoch": 1.9364439621764067,
+      "grad_norm": 0.371165007352829,
+      "learning_rate": 3.3625681608291393e-06,
+      "loss": 0.3534,
+      "step": 4164
+    },
+    {
+      "epoch": 1.9369090063556038,
+      "grad_norm": 0.3415764570236206,
+      "learning_rate": 3.360011680358409e-06,
+      "loss": 0.3183,
+      "step": 4165
+    },
+    {
+      "epoch": 1.937374050534801,
+      "grad_norm": 0.41029369831085205,
+      "learning_rate": 3.3574556802136164e-06,
+      "loss": 0.3807,
+      "step": 4166
+    },
+    {
+      "epoch": 1.9378390947139978,
+      "grad_norm": 0.41946086287498474,
+      "learning_rate": 3.354900161143377e-06,
+      "loss": 0.3535,
+      "step": 4167
+    },
+    {
+      "epoch": 1.9383041388931948,
+      "grad_norm": 0.36154383420944214,
+      "learning_rate": 3.352345123896158e-06,
+      "loss": 0.3153,
+      "step": 4168
+    },
+    {
+      "epoch": 1.9387691830723919,
+      "grad_norm": 0.3600989878177643,
+      "learning_rate": 3.3497905692202892e-06,
+      "loss": 0.3591,
+      "step": 4169
+    },
+    {
+      "epoch": 1.939234227251589,
+      "grad_norm": 0.37105801701545715,
+      "learning_rate": 3.347236497863957e-06,
+      "loss": 0.3842,
+      "step": 4170
+    },
+    {
+      "epoch": 1.939699271430786,
+      "grad_norm": 0.32860368490219116,
+      "learning_rate": 3.3446829105752103e-06,
+      "loss": 0.289,
+      "step": 4171
+    },
+    {
+      "epoch": 1.9401643156099828,
+      "grad_norm": 0.3648563325405121,
+      "learning_rate": 3.34212980810195e-06,
+      "loss": 0.3297,
+      "step": 4172
+    },
+    {
+      "epoch": 1.94062935978918,
+      "grad_norm": 0.3792143166065216,
+      "learning_rate": 3.3395771911919416e-06,
+      "loss": 0.3211,
+      "step": 4173
+    },
+    {
+      "epoch": 1.941094403968377,
+      "grad_norm": 0.4156646430492401,
+      "learning_rate": 3.3370250605928013e-06,
+      "loss": 0.3774,
+      "step": 4174
+    },
+    {
+      "epoch": 1.941559448147574,
+      "grad_norm": 0.3592356741428375,
+      "learning_rate": 3.33447341705201e-06,
+      "loss": 0.3604,
+      "step": 4175
+    },
+    {
+      "epoch": 1.9420244923267709,
+      "grad_norm": 0.4310251474380493,
+      "learning_rate": 3.3319222613169e-06,
+      "loss": 0.3351,
+      "step": 4176
+    },
+    {
+      "epoch": 1.9424895365059682,
+      "grad_norm": 0.40586546063423157,
+      "learning_rate": 3.3293715941346676e-06,
+      "loss": 0.312,
+      "step": 4177
+    },
+    {
+      "epoch": 1.9429545806851651,
+      "grad_norm": 0.374381422996521,
+      "learning_rate": 3.3268214162523563e-06,
+      "loss": 0.3451,
+      "step": 4178
+    },
+    {
+      "epoch": 1.943419624864362,
+      "grad_norm": 0.3669213056564331,
+      "learning_rate": 3.324271728416877e-06,
+      "loss": 0.3387,
+      "step": 4179
+    },
+    {
+      "epoch": 1.9438846690435592,
+      "grad_norm": 0.39169323444366455,
+      "learning_rate": 3.321722531374988e-06,
+      "loss": 0.3568,
+      "step": 4180
+    },
+    {
+      "epoch": 1.9443497132227563,
+      "grad_norm": 0.3700978457927704,
+      "learning_rate": 3.3191738258733085e-06,
+      "loss": 0.3265,
+      "step": 4181
+    },
+    {
+      "epoch": 1.9448147574019532,
+      "grad_norm": 0.46921029686927795,
+      "learning_rate": 3.316625612658315e-06,
+      "loss": 0.3883,
+      "step": 4182
+    },
+    {
+      "epoch": 1.94527980158115,
+      "grad_norm": 0.38272857666015625,
+      "learning_rate": 3.314077892476334e-06,
+      "loss": 0.3328,
+      "step": 4183
+    },
+    {
+      "epoch": 1.9457448457603472,
+      "grad_norm": 0.40746691823005676,
+      "learning_rate": 3.3115306660735564e-06,
+      "loss": 0.3558,
+      "step": 4184
+    },
+    {
+      "epoch": 1.9462098899395444,
+      "grad_norm": 0.33452197909355164,
+      "learning_rate": 3.308983934196018e-06,
+      "loss": 0.3091,
+      "step": 4185
+    },
+    {
+      "epoch": 1.9466749341187413,
+      "grad_norm": 0.4103114902973175,
+      "learning_rate": 3.3064376975896197e-06,
+      "loss": 0.3777,
+      "step": 4186
+    },
+    {
+      "epoch": 1.9471399782979382,
+      "grad_norm": 0.3644181489944458,
+      "learning_rate": 3.3038919570001086e-06,
+      "loss": 0.3238,
+      "step": 4187
+    },
+    {
+      "epoch": 1.9476050224771353,
+      "grad_norm": 0.42216435074806213,
+      "learning_rate": 3.301346713173096e-06,
+      "loss": 0.3551,
+      "step": 4188
+    },
+    {
+      "epoch": 1.9480700666563324,
+      "grad_norm": 0.39056453108787537,
+      "learning_rate": 3.2988019668540373e-06,
+      "loss": 0.3612,
+      "step": 4189
+    },
+    {
+      "epoch": 1.9485351108355293,
+      "grad_norm": 0.37854042649269104,
+      "learning_rate": 3.2962577187882517e-06,
+      "loss": 0.3339,
+      "step": 4190
+    },
+    {
+      "epoch": 1.9490001550147265,
+      "grad_norm": 0.34841084480285645,
+      "learning_rate": 3.2937139697209043e-06,
+      "loss": 0.305,
+      "step": 4191
+    },
+    {
+      "epoch": 1.9494651991939236,
+      "grad_norm": 0.4083377718925476,
+      "learning_rate": 3.2911707203970213e-06,
+      "loss": 0.3892,
+      "step": 4192
+    },
+    {
+      "epoch": 1.9499302433731205,
+      "grad_norm": 0.3734639585018158,
+      "learning_rate": 3.2886279715614754e-06,
+      "loss": 0.327,
+      "step": 4193
+    },
+    {
+      "epoch": 1.9503952875523174,
+      "grad_norm": 0.34090444445610046,
+      "learning_rate": 3.286085723959001e-06,
+      "loss": 0.3013,
+      "step": 4194
+    },
+    {
+      "epoch": 1.9508603317315145,
+      "grad_norm": 0.42237943410873413,
+      "learning_rate": 3.283543978334177e-06,
+      "loss": 0.3566,
+      "step": 4195
+    },
+    {
+      "epoch": 1.9513253759107116,
+      "grad_norm": 0.36327943205833435,
+      "learning_rate": 3.281002735431442e-06,
+      "loss": 0.3294,
+      "step": 4196
+    },
+    {
+      "epoch": 1.9517904200899086,
+      "grad_norm": 0.3818821310997009,
+      "learning_rate": 3.2784619959950832e-06,
+      "loss": 0.3715,
+      "step": 4197
+    },
+    {
+      "epoch": 1.9522554642691055,
+      "grad_norm": 0.3964533805847168,
+      "learning_rate": 3.2759217607692427e-06,
+      "loss": 0.3286,
+      "step": 4198
+    },
+    {
+      "epoch": 1.9527205084483026,
+      "grad_norm": 0.404161661863327,
+      "learning_rate": 3.2733820304979136e-06,
+      "loss": 0.3412,
+      "step": 4199
+    },
+    {
+      "epoch": 1.9531855526274997,
+      "grad_norm": 0.397344708442688,
+      "learning_rate": 3.2708428059249437e-06,
+      "loss": 0.3277,
+      "step": 4200
+    },
+    {
+      "epoch": 1.9536505968066966,
+      "grad_norm": 0.3819953501224518,
+      "learning_rate": 3.268304087794027e-06,
+      "loss": 0.325,
+      "step": 4201
+    },
+    {
+      "epoch": 1.9541156409858935,
+      "grad_norm": 0.40046122670173645,
+      "learning_rate": 3.2657658768487164e-06,
+      "loss": 0.3437,
+      "step": 4202
+    },
+    {
+      "epoch": 1.9545806851650906,
+      "grad_norm": 0.3873307704925537,
+      "learning_rate": 3.26322817383241e-06,
+      "loss": 0.3454,
+      "step": 4203
+    },
+    {
+      "epoch": 1.9550457293442878,
+      "grad_norm": 0.3931639790534973,
+      "learning_rate": 3.260690979488361e-06,
+      "loss": 0.3232,
+      "step": 4204
+    },
+    {
+      "epoch": 1.9555107735234847,
+      "grad_norm": 0.48349693417549133,
+      "learning_rate": 3.258154294559671e-06,
+      "loss": 0.4106,
+      "step": 4205
+    },
+    {
+      "epoch": 1.9559758177026818,
+      "grad_norm": 0.3327430486679077,
+      "learning_rate": 3.255618119789298e-06,
+      "loss": 0.305,
+      "step": 4206
+    },
+    {
+      "epoch": 1.956440861881879,
+      "grad_norm": 0.4101545512676239,
+      "learning_rate": 3.2530824559200415e-06,
+      "loss": 0.3245,
+      "step": 4207
+    },
+    {
+      "epoch": 1.9569059060610758,
+      "grad_norm": 0.4183330833911896,
+      "learning_rate": 3.2505473036945588e-06,
+      "loss": 0.3395,
+      "step": 4208
+    },
+    {
+      "epoch": 1.9573709502402727,
+      "grad_norm": 0.463460236787796,
+      "learning_rate": 3.2480126638553533e-06,
+      "loss": 0.3707,
+      "step": 4209
+    },
+    {
+      "epoch": 1.9578359944194699,
+      "grad_norm": 0.3990997076034546,
+      "learning_rate": 3.2454785371447817e-06,
+      "loss": 0.3211,
+      "step": 4210
+    },
+    {
+      "epoch": 1.958301038598667,
+      "grad_norm": 0.38769689202308655,
+      "learning_rate": 3.2429449243050464e-06,
+      "loss": 0.3219,
+      "step": 4211
+    },
+    {
+      "epoch": 1.958766082777864,
+      "grad_norm": 0.3971726596355438,
+      "learning_rate": 3.2404118260782047e-06,
+      "loss": 0.3451,
+      "step": 4212
+    },
+    {
+      "epoch": 1.9592311269570608,
+      "grad_norm": 0.3939190208911896,
+      "learning_rate": 3.2378792432061557e-06,
+      "loss": 0.3454,
+      "step": 4213
+    },
+    {
+      "epoch": 1.959696171136258,
+      "grad_norm": 0.3693125247955322,
+      "learning_rate": 3.2353471764306567e-06,
+      "loss": 0.3299,
+      "step": 4214
+    },
+    {
+      "epoch": 1.960161215315455,
+      "grad_norm": 0.42611974477767944,
+      "learning_rate": 3.2328156264933043e-06,
+      "loss": 0.3396,
+      "step": 4215
+    },
+    {
+      "epoch": 1.960626259494652,
+      "grad_norm": 0.4172080159187317,
+      "learning_rate": 3.23028459413555e-06,
+      "loss": 0.3579,
+      "step": 4216
+    },
+    {
+      "epoch": 1.9610913036738489,
+      "grad_norm": 0.4222067594528198,
+      "learning_rate": 3.227754080098694e-06,
+      "loss": 0.3634,
+      "step": 4217
+    },
+    {
+      "epoch": 1.961556347853046,
+      "grad_norm": 0.3900541365146637,
+      "learning_rate": 3.2252240851238786e-06,
+      "loss": 0.3096,
+      "step": 4218
+    },
+    {
+      "epoch": 1.9620213920322431,
+      "grad_norm": 0.3957304060459137,
+      "learning_rate": 3.2226946099521026e-06,
+      "loss": 0.3681,
+      "step": 4219
+    },
+    {
+      "epoch": 1.96248643621144,
+      "grad_norm": 0.44212228059768677,
+      "learning_rate": 3.2201656553242054e-06,
+      "loss": 0.3601,
+      "step": 4220
+    },
+    {
+      "epoch": 1.9629514803906372,
+      "grad_norm": 0.38398414850234985,
+      "learning_rate": 3.217637221980878e-06,
+      "loss": 0.308,
+      "step": 4221
+    },
+    {
+      "epoch": 1.9634165245698343,
+      "grad_norm": 0.4060637056827545,
+      "learning_rate": 3.215109310662656e-06,
+      "loss": 0.342,
+      "step": 4222
+    },
+    {
+      "epoch": 1.9638815687490312,
+      "grad_norm": 0.4322217106819153,
+      "learning_rate": 3.2125819221099265e-06,
+      "loss": 0.3304,
+      "step": 4223
+    },
+    {
+      "epoch": 1.964346612928228,
+      "grad_norm": 0.37583181262016296,
+      "learning_rate": 3.210055057062917e-06,
+      "loss": 0.2955,
+      "step": 4224
+    },
+    {
+      "epoch": 1.9648116571074252,
+      "grad_norm": 0.37987688183784485,
+      "learning_rate": 3.2075287162617084e-06,
+      "loss": 0.3506,
+      "step": 4225
+    },
+    {
+      "epoch": 1.9652767012866224,
+      "grad_norm": 0.42965471744537354,
+      "learning_rate": 3.2050029004462226e-06,
+      "loss": 0.3463,
+      "step": 4226
+    },
+    {
+      "epoch": 1.9657417454658193,
+      "grad_norm": 0.3723207712173462,
+      "learning_rate": 3.2024776103562304e-06,
+      "loss": 0.3088,
+      "step": 4227
+    },
+    {
+      "epoch": 1.9662067896450162,
+      "grad_norm": 0.3601841628551483,
+      "learning_rate": 3.199952846731349e-06,
+      "loss": 0.3396,
+      "step": 4228
+    },
+    {
+      "epoch": 1.9666718338242133,
+      "grad_norm": 0.45041221380233765,
+      "learning_rate": 3.197428610311042e-06,
+      "loss": 0.3399,
+      "step": 4229
+    },
+    {
+      "epoch": 1.9671368780034104,
+      "grad_norm": 0.3874083459377289,
+      "learning_rate": 3.194904901834613e-06,
+      "loss": 0.3283,
+      "step": 4230
+    },
+    {
+      "epoch": 1.9676019221826073,
+      "grad_norm": 0.3825784921646118,
+      "learning_rate": 3.19238172204122e-06,
+      "loss": 0.3291,
+      "step": 4231
+    },
+    {
+      "epoch": 1.9680669663618042,
+      "grad_norm": 0.3952145278453827,
+      "learning_rate": 3.1898590716698574e-06,
+      "loss": 0.3482,
+      "step": 4232
+    },
+    {
+      "epoch": 1.9685320105410014,
+      "grad_norm": 0.4280111789703369,
+      "learning_rate": 3.1873369514593712e-06,
+      "loss": 0.3488,
+      "step": 4233
+    },
+    {
+      "epoch": 1.9689970547201985,
+      "grad_norm": 0.44719386100769043,
+      "learning_rate": 3.184815362148448e-06,
+      "loss": 0.3692,
+      "step": 4234
+    },
+    {
+      "epoch": 1.9694620988993954,
+      "grad_norm": 0.4005783796310425,
+      "learning_rate": 3.1822943044756222e-06,
+      "loss": 0.3451,
+      "step": 4235
+    },
+    {
+      "epoch": 1.9699271430785925,
+      "grad_norm": 0.3623179495334625,
+      "learning_rate": 3.1797737791792672e-06,
+      "loss": 0.3514,
+      "step": 4236
+    },
+    {
+      "epoch": 1.9703921872577896,
+      "grad_norm": 0.4718107581138611,
+      "learning_rate": 3.177253786997609e-06,
+      "loss": 0.3607,
+      "step": 4237
+    },
+    {
+      "epoch": 1.9708572314369865,
+      "grad_norm": 0.44240835309028625,
+      "learning_rate": 3.1747343286687065e-06,
+      "loss": 0.3568,
+      "step": 4238
+    },
+    {
+      "epoch": 1.9713222756161835,
+      "grad_norm": 0.3883517384529114,
+      "learning_rate": 3.1722154049304728e-06,
+      "loss": 0.3407,
+      "step": 4239
+    },
+    {
+      "epoch": 1.9717873197953806,
+      "grad_norm": 0.4095773994922638,
+      "learning_rate": 3.1696970165206564e-06,
+      "loss": 0.3302,
+      "step": 4240
+    },
+    {
+      "epoch": 1.9722523639745777,
+      "grad_norm": 0.4610760509967804,
+      "learning_rate": 3.167179164176857e-06,
+      "loss": 0.3149,
+      "step": 4241
+    },
+    {
+      "epoch": 1.9727174081537746,
+      "grad_norm": 0.4434893727302551,
+      "learning_rate": 3.1646618486365068e-06,
+      "loss": 0.3795,
+      "step": 4242
+    },
+    {
+      "epoch": 1.9731824523329715,
+      "grad_norm": 0.3962785601615906,
+      "learning_rate": 3.1621450706368904e-06,
+      "loss": 0.3624,
+      "step": 4243
+    },
+    {
+      "epoch": 1.9736474965121686,
+      "grad_norm": 0.3942055404186249,
+      "learning_rate": 3.15962883091513e-06,
+      "loss": 0.3408,
+      "step": 4244
+    },
+    {
+      "epoch": 1.9741125406913658,
+      "grad_norm": 0.4068949520587921,
+      "learning_rate": 3.1571131302081916e-06,
+      "loss": 0.3692,
+      "step": 4245
+    },
+    {
+      "epoch": 1.9745775848705627,
+      "grad_norm": 0.46068617701530457,
+      "learning_rate": 3.154597969252883e-06,
+      "loss": 0.3489,
+      "step": 4246
+    },
+    {
+      "epoch": 1.9750426290497596,
+      "grad_norm": 0.47146105766296387,
+      "learning_rate": 3.1520833487858547e-06,
+      "loss": 0.3245,
+      "step": 4247
+    },
+    {
+      "epoch": 1.9755076732289567,
+      "grad_norm": 0.3786758482456207,
+      "learning_rate": 3.1495692695435966e-06,
+      "loss": 0.3297,
+      "step": 4248
+    },
+    {
+      "epoch": 1.9759727174081538,
+      "grad_norm": 0.4309166967868805,
+      "learning_rate": 3.147055732262444e-06,
+      "loss": 0.3595,
+      "step": 4249
+    },
+    {
+      "epoch": 1.9764377615873507,
+      "grad_norm": 0.4204605519771576,
+      "learning_rate": 3.1445427376785687e-06,
+      "loss": 0.3285,
+      "step": 4250
+    },
+    {
+      "epoch": 1.9769028057665479,
+      "grad_norm": 0.4231455624103546,
+      "learning_rate": 3.142030286527987e-06,
+      "loss": 0.3434,
+      "step": 4251
+    },
+    {
+      "epoch": 1.977367849945745,
+      "grad_norm": 0.39683791995048523,
+      "learning_rate": 3.1395183795465565e-06,
+      "loss": 0.3404,
+      "step": 4252
+    },
+    {
+      "epoch": 1.977832894124942,
+      "grad_norm": 0.43231719732284546,
+      "learning_rate": 3.137007017469971e-06,
+      "loss": 0.3534,
+      "step": 4253
+    },
+    {
+      "epoch": 1.9782979383041388,
+      "grad_norm": 0.45599350333213806,
+      "learning_rate": 3.1344962010337703e-06,
+      "loss": 0.3811,
+      "step": 4254
+    },
+    {
+      "epoch": 1.978762982483336,
+      "grad_norm": 0.4510321021080017,
+      "learning_rate": 3.131985930973329e-06,
+      "loss": 0.374,
+      "step": 4255
+    },
+    {
+      "epoch": 1.979228026662533,
+      "grad_norm": 0.44332581758499146,
+      "learning_rate": 3.1294762080238672e-06,
+      "loss": 0.3416,
+      "step": 4256
+    },
+    {
+      "epoch": 1.97969307084173,
+      "grad_norm": 0.34618374705314636,
+      "learning_rate": 3.12696703292044e-06,
+      "loss": 0.2877,
+      "step": 4257
+    },
+    {
+      "epoch": 1.9801581150209269,
+      "grad_norm": 0.455598384141922,
+      "learning_rate": 3.1244584063979467e-06,
+      "loss": 0.3718,
+      "step": 4258
+    },
+    {
+      "epoch": 1.980623159200124,
+      "grad_norm": 0.4085701107978821,
+      "learning_rate": 3.121950329191119e-06,
+      "loss": 0.3747,
+      "step": 4259
+    },
+    {
+      "epoch": 1.9810882033793211,
+      "grad_norm": 0.3253307044506073,
+      "learning_rate": 3.1194428020345375e-06,
+      "loss": 0.3125,
+      "step": 4260
+    },
+    {
+      "epoch": 1.981553247558518,
+      "grad_norm": 0.36619746685028076,
+      "learning_rate": 3.11693582566261e-06,
+      "loss": 0.343,
+      "step": 4261
+    },
+    {
+      "epoch": 1.982018291737715,
+      "grad_norm": 0.47953325510025024,
+      "learning_rate": 3.1144294008095942e-06,
+      "loss": 0.3779,
+      "step": 4262
+    },
+    {
+      "epoch": 1.982483335916912,
+      "grad_norm": 0.4037295877933502,
+      "learning_rate": 3.111923528209577e-06,
+      "loss": 0.3665,
+      "step": 4263
+    },
+    {
+      "epoch": 1.9829483800961092,
+      "grad_norm": 0.376261442899704,
+      "learning_rate": 3.1094182085964935e-06,
+      "loss": 0.3332,
+      "step": 4264
+    },
+    {
+      "epoch": 1.983413424275306,
+      "grad_norm": 0.4503709673881531,
+      "learning_rate": 3.1069134427041047e-06,
+      "loss": 0.3675,
+      "step": 4265
+    },
+    {
+      "epoch": 1.9838784684545032,
+      "grad_norm": 0.4101594388484955,
+      "learning_rate": 3.1044092312660213e-06,
+      "loss": 0.3787,
+      "step": 4266
+    },
+    {
+      "epoch": 1.9843435126337003,
+      "grad_norm": 0.45172303915023804,
+      "learning_rate": 3.101905575015682e-06,
+      "loss": 0.358,
+      "step": 4267
+    },
+    {
+      "epoch": 1.9848085568128973,
+      "grad_norm": 0.3164076805114746,
+      "learning_rate": 3.0994024746863692e-06,
+      "loss": 0.3028,
+      "step": 4268
+    },
+    {
+      "epoch": 1.9852736009920942,
+      "grad_norm": 0.3954414427280426,
+      "learning_rate": 3.0968999310111993e-06,
+      "loss": 0.3998,
+      "step": 4269
+    },
+    {
+      "epoch": 1.9857386451712913,
+      "grad_norm": 0.40084972977638245,
+      "learning_rate": 3.0943979447231287e-06,
+      "loss": 0.3176,
+      "step": 4270
+    },
+    {
+      "epoch": 1.9862036893504884,
+      "grad_norm": 0.4292057752609253,
+      "learning_rate": 3.091896516554945e-06,
+      "loss": 0.3524,
+      "step": 4271
+    },
+    {
+      "epoch": 1.9866687335296853,
+      "grad_norm": 0.4974363148212433,
+      "learning_rate": 3.0893956472392805e-06,
+      "loss": 0.3649,
+      "step": 4272
+    },
+    {
+      "epoch": 1.9871337777088822,
+      "grad_norm": 0.37316155433654785,
+      "learning_rate": 3.086895337508594e-06,
+      "loss": 0.3477,
+      "step": 4273
+    },
+    {
+      "epoch": 1.9875988218880793,
+      "grad_norm": 0.3752647638320923,
+      "learning_rate": 3.0843955880951906e-06,
+      "loss": 0.3585,
+      "step": 4274
+    },
+    {
+      "epoch": 1.9880638660672765,
+      "grad_norm": 0.3850919008255005,
+      "learning_rate": 3.081896399731202e-06,
+      "loss": 0.3594,
+      "step": 4275
+    },
+    {
+      "epoch": 1.9885289102464734,
+      "grad_norm": 0.40935975313186646,
+      "learning_rate": 3.0793977731486034e-06,
+      "loss": 0.3427,
+      "step": 4276
+    },
+    {
+      "epoch": 1.9889939544256703,
+      "grad_norm": 0.4816490411758423,
+      "learning_rate": 3.0768997090791995e-06,
+      "loss": 0.4007,
+      "step": 4277
+    },
+    {
+      "epoch": 1.9894589986048674,
+      "grad_norm": 0.3938618302345276,
+      "learning_rate": 3.0744022082546356e-06,
+      "loss": 0.3432,
+      "step": 4278
+    },
+    {
+      "epoch": 1.9899240427840645,
+      "grad_norm": 0.34527865052223206,
+      "learning_rate": 3.071905271406384e-06,
+      "loss": 0.3668,
+      "step": 4279
+    },
+    {
+      "epoch": 1.9903890869632614,
+      "grad_norm": 0.3559629023075104,
+      "learning_rate": 3.0694088992657617e-06,
+      "loss": 0.3488,
+      "step": 4280
+    },
+    {
+      "epoch": 1.9908541311424586,
+      "grad_norm": 0.4233744144439697,
+      "learning_rate": 3.066913092563913e-06,
+      "loss": 0.3593,
+      "step": 4281
+    },
+    {
+      "epoch": 1.9913191753216557,
+      "grad_norm": 0.36039772629737854,
+      "learning_rate": 3.064417852031822e-06,
+      "loss": 0.3275,
+      "step": 4282
+    },
+    {
+      "epoch": 1.9917842195008526,
+      "grad_norm": 0.41235482692718506,
+      "learning_rate": 3.0619231784003e-06,
+      "loss": 0.3885,
+      "step": 4283
+    },
+    {
+      "epoch": 1.9922492636800495,
+      "grad_norm": 0.3795316219329834,
+      "learning_rate": 3.059429072400001e-06,
+      "loss": 0.3176,
+      "step": 4284
+    },
+    {
+      "epoch": 1.9927143078592466,
+      "grad_norm": 0.4195452332496643,
+      "learning_rate": 3.0569355347614033e-06,
+      "loss": 0.3609,
+      "step": 4285
+    },
+    {
+      "epoch": 1.9931793520384438,
+      "grad_norm": 0.32910481095314026,
+      "learning_rate": 3.054442566214827e-06,
+      "loss": 0.3151,
+      "step": 4286
+    },
+    {
+      "epoch": 1.9936443962176407,
+      "grad_norm": 0.3915290832519531,
+      "learning_rate": 3.051950167490422e-06,
+      "loss": 0.3448,
+      "step": 4287
+    },
+    {
+      "epoch": 1.9941094403968376,
+      "grad_norm": 0.39752107858657837,
+      "learning_rate": 3.049458339318169e-06,
+      "loss": 0.3286,
+      "step": 4288
+    },
+    {
+      "epoch": 1.9945744845760347,
+      "grad_norm": 0.45507895946502686,
+      "learning_rate": 3.0469670824278863e-06,
+      "loss": 0.3607,
+      "step": 4289
+    },
+    {
+      "epoch": 1.9950395287552318,
+      "grad_norm": 0.3670455515384674,
+      "learning_rate": 3.044476397549221e-06,
+      "loss": 0.332,
+      "step": 4290
+    },
+    {
+      "epoch": 1.9955045729344287,
+      "grad_norm": 0.4452548623085022,
+      "learning_rate": 3.0419862854116554e-06,
+      "loss": 0.3882,
+      "step": 4291
+    },
+    {
+      "epoch": 1.9959696171136256,
+      "grad_norm": 0.3838987946510315,
+      "learning_rate": 3.0394967467445014e-06,
+      "loss": 0.3492,
+      "step": 4292
+    },
+    {
+      "epoch": 1.9964346612928228,
+      "grad_norm": 0.43697041273117065,
+      "learning_rate": 3.0370077822769073e-06,
+      "loss": 0.3688,
+      "step": 4293
+    },
+    {
+      "epoch": 1.99689970547202,
+      "grad_norm": 0.38790035247802734,
+      "learning_rate": 3.034519392737847e-06,
+      "loss": 0.2995,
+      "step": 4294
+    },
+    {
+      "epoch": 1.9973647496512168,
+      "grad_norm": 0.41673675179481506,
+      "learning_rate": 3.0320315788561334e-06,
+      "loss": 0.3068,
+      "step": 4295
+    },
+    {
+      "epoch": 1.997829793830414,
+      "grad_norm": 0.5042394995689392,
+      "learning_rate": 3.029544341360402e-06,
+      "loss": 0.3846,
+      "step": 4296
+    },
+    {
+      "epoch": 1.998294838009611,
+      "grad_norm": 0.3903370201587677,
+      "learning_rate": 3.0270576809791273e-06,
+      "loss": 0.3873,
+      "step": 4297
+    },
+    {
+      "epoch": 1.998759882188808,
+      "grad_norm": 0.4303033947944641,
+      "learning_rate": 3.02457159844061e-06,
+      "loss": 0.3656,
+      "step": 4298
+    },
+    {
+      "epoch": 1.9992249263680049,
+      "grad_norm": 0.44149455428123474,
+      "learning_rate": 3.022086094472986e-06,
+      "loss": 0.3233,
+      "step": 4299
+    },
+    {
+      "epoch": 1.999689970547202,
+      "grad_norm": 0.38690122961997986,
+      "learning_rate": 3.019601169804216e-06,
+      "loss": 0.3404,
+      "step": 4300
+    },
+    {
+      "epoch": 2.000155014726399,
+      "grad_norm": 0.7339352369308472,
+      "learning_rate": 3.0171168251620974e-06,
+      "loss": 0.5187,
+      "step": 4301
+    },
+    {
+      "epoch": 2.000620058905596,
+      "grad_norm": 0.42907798290252686,
+      "learning_rate": 3.01463306127425e-06,
+      "loss": 0.3366,
+      "step": 4302
+    },
+    {
+      "epoch": 2.001085103084793,
+      "grad_norm": 0.4182564616203308,
+      "learning_rate": 3.012149878868132e-06,
+      "loss": 0.2779,
+      "step": 4303
+    },
+    {
+      "epoch": 2.0015501472639903,
+      "grad_norm": 0.37000566720962524,
+      "learning_rate": 3.009667278671024e-06,
+      "loss": 0.3469,
+      "step": 4304
+    },
+    {
+      "epoch": 2.002015191443187,
+      "grad_norm": 0.36970627307891846,
+      "learning_rate": 3.0071852614100427e-06,
+      "loss": 0.2855,
+      "step": 4305
+    },
+    {
+      "epoch": 2.002480235622384,
+      "grad_norm": 0.39976659417152405,
+      "learning_rate": 3.004703827812128e-06,
+      "loss": 0.3605,
+      "step": 4306
+    },
+    {
+      "epoch": 2.002945279801581,
+      "grad_norm": 0.3608565032482147,
+      "learning_rate": 3.0022229786040526e-06,
+      "loss": 0.2864,
+      "step": 4307
+    },
+    {
+      "epoch": 2.0034103239807783,
+      "grad_norm": 0.37464091181755066,
+      "learning_rate": 2.999742714512415e-06,
+      "loss": 0.3287,
+      "step": 4308
+    },
+    {
+      "epoch": 2.0038753681599752,
+      "grad_norm": 0.37484025955200195,
+      "learning_rate": 2.997263036263647e-06,
+      "loss": 0.3256,
+      "step": 4309
+    },
+    {
+      "epoch": 2.004340412339172,
+      "grad_norm": 0.34249240159988403,
+      "learning_rate": 2.9947839445840045e-06,
+      "loss": 0.2974,
+      "step": 4310
+    },
+    {
+      "epoch": 2.004805456518369,
+      "grad_norm": 0.38859909772872925,
+      "learning_rate": 2.9923054401995745e-06,
+      "loss": 0.3534,
+      "step": 4311
+    },
+    {
+      "epoch": 2.0052705006975664,
+      "grad_norm": 0.3693307936191559,
+      "learning_rate": 2.9898275238362686e-06,
+      "loss": 0.3355,
+      "step": 4312
+    },
+    {
+      "epoch": 2.0057355448767633,
+      "grad_norm": 0.3429759442806244,
+      "learning_rate": 2.98735019621983e-06,
+      "loss": 0.2918,
+      "step": 4313
+    },
+    {
+      "epoch": 2.00620058905596,
+      "grad_norm": 0.3832457959651947,
+      "learning_rate": 2.984873458075827e-06,
+      "loss": 0.3288,
+      "step": 4314
+    },
+    {
+      "epoch": 2.006665633235157,
+      "grad_norm": 0.34353703260421753,
+      "learning_rate": 2.9823973101296564e-06,
+      "loss": 0.3088,
+      "step": 4315
+    },
+    {
+      "epoch": 2.0071306774143545,
+      "grad_norm": 0.37068885564804077,
+      "learning_rate": 2.9799217531065407e-06,
+      "loss": 0.3203,
+      "step": 4316
+    },
+    {
+      "epoch": 2.0075957215935514,
+      "grad_norm": 0.3725161850452423,
+      "learning_rate": 2.977446787731532e-06,
+      "loss": 0.3242,
+      "step": 4317
+    },
+    {
+      "epoch": 2.0080607657727483,
+      "grad_norm": 0.36433523893356323,
+      "learning_rate": 2.9749724147295054e-06,
+      "loss": 0.3368,
+      "step": 4318
+    },
+    {
+      "epoch": 2.0085258099519456,
+      "grad_norm": 0.4169025421142578,
+      "learning_rate": 2.972498634825168e-06,
+      "loss": 0.315,
+      "step": 4319
+    },
+    {
+      "epoch": 2.0089908541311425,
+      "grad_norm": 0.41936197876930237,
+      "learning_rate": 2.9700254487430448e-06,
+      "loss": 0.3359,
+      "step": 4320
+    },
+    {
+      "epoch": 2.0094558983103394,
+      "grad_norm": 0.3963291347026825,
+      "learning_rate": 2.9675528572074953e-06,
+      "loss": 0.3093,
+      "step": 4321
+    },
+    {
+      "epoch": 2.0099209424895363,
+      "grad_norm": 0.38462764024734497,
+      "learning_rate": 2.9650808609427e-06,
+      "loss": 0.3439,
+      "step": 4322
+    },
+    {
+      "epoch": 2.0103859866687337,
+      "grad_norm": 0.36186519265174866,
+      "learning_rate": 2.962609460672669e-06,
+      "loss": 0.3149,
+      "step": 4323
+    },
+    {
+      "epoch": 2.0108510308479306,
+      "grad_norm": 0.3426899015903473,
+      "learning_rate": 2.960138657121233e-06,
+      "loss": 0.2876,
+      "step": 4324
+    },
+    {
+      "epoch": 2.0113160750271275,
+      "grad_norm": 0.3653092682361603,
+      "learning_rate": 2.957668451012049e-06,
+      "loss": 0.347,
+      "step": 4325
+    },
+    {
+      "epoch": 2.0117811192063244,
+      "grad_norm": 0.3865642547607422,
+      "learning_rate": 2.955198843068603e-06,
+      "loss": 0.3204,
+      "step": 4326
+    },
+    {
+      "epoch": 2.0122461633855218,
+      "grad_norm": 0.4121185839176178,
+      "learning_rate": 2.9527298340142e-06,
+      "loss": 0.3238,
+      "step": 4327
+    },
+    {
+      "epoch": 2.0127112075647187,
+      "grad_norm": 0.32982850074768066,
+      "learning_rate": 2.950261424571977e-06,
+      "loss": 0.289,
+      "step": 4328
+    },
+    {
+      "epoch": 2.0131762517439156,
+      "grad_norm": 0.36676689982414246,
+      "learning_rate": 2.9477936154648866e-06,
+      "loss": 0.3467,
+      "step": 4329
+    },
+    {
+      "epoch": 2.0136412959231125,
+      "grad_norm": 0.368436336517334,
+      "learning_rate": 2.9453264074157134e-06,
+      "loss": 0.352,
+      "step": 4330
+    },
+    {
+      "epoch": 2.01410634010231,
+      "grad_norm": 0.4215182363986969,
+      "learning_rate": 2.9428598011470597e-06,
+      "loss": 0.3065,
+      "step": 4331
+    },
+    {
+      "epoch": 2.0145713842815067,
+      "grad_norm": 0.3354952037334442,
+      "learning_rate": 2.9403937973813564e-06,
+      "loss": 0.3313,
+      "step": 4332
+    },
+    {
+      "epoch": 2.0150364284607036,
+      "grad_norm": 0.3583640456199646,
+      "learning_rate": 2.9379283968408546e-06,
+      "loss": 0.2962,
+      "step": 4333
+    },
+    {
+      "epoch": 2.015501472639901,
+      "grad_norm": 0.33064424991607666,
+      "learning_rate": 2.9354636002476324e-06,
+      "loss": 0.3023,
+      "step": 4334
+    },
+    {
+      "epoch": 2.015966516819098,
+      "grad_norm": 0.33953142166137695,
+      "learning_rate": 2.9329994083235857e-06,
+      "loss": 0.3078,
+      "step": 4335
+    },
+    {
+      "epoch": 2.016431560998295,
+      "grad_norm": 0.3762052059173584,
+      "learning_rate": 2.930535821790439e-06,
+      "loss": 0.3592,
+      "step": 4336
+    },
+    {
+      "epoch": 2.0168966051774917,
+      "grad_norm": 0.35094988346099854,
+      "learning_rate": 2.928072841369734e-06,
+      "loss": 0.3177,
+      "step": 4337
+    },
+    {
+      "epoch": 2.017361649356689,
+      "grad_norm": 0.3473728597164154,
+      "learning_rate": 2.92561046778284e-06,
+      "loss": 0.35,
+      "step": 4338
+    },
+    {
+      "epoch": 2.017826693535886,
+      "grad_norm": 0.3321429491043091,
+      "learning_rate": 2.9231487017509442e-06,
+      "loss": 0.3325,
+      "step": 4339
+    },
+    {
+      "epoch": 2.018291737715083,
+      "grad_norm": 0.3240537941455841,
+      "learning_rate": 2.920687543995061e-06,
+      "loss": 0.2927,
+      "step": 4340
+    },
+    {
+      "epoch": 2.0187567818942798,
+      "grad_norm": 0.36483100056648254,
+      "learning_rate": 2.91822699523602e-06,
+      "loss": 0.3395,
+      "step": 4341
+    },
+    {
+      "epoch": 2.019221826073477,
+      "grad_norm": 0.3937772512435913,
+      "learning_rate": 2.915767056194479e-06,
+      "loss": 0.3633,
+      "step": 4342
+    },
+    {
+      "epoch": 2.019686870252674,
+      "grad_norm": 0.33440476655960083,
+      "learning_rate": 2.9133077275909112e-06,
+      "loss": 0.3051,
+      "step": 4343
+    },
+    {
+      "epoch": 2.020151914431871,
+      "grad_norm": 0.3541061282157898,
+      "learning_rate": 2.910849010145617e-06,
+      "loss": 0.3395,
+      "step": 4344
+    },
+    {
+      "epoch": 2.0206169586110683,
+      "grad_norm": 0.34767183661460876,
+      "learning_rate": 2.9083909045787116e-06,
+      "loss": 0.3267,
+      "step": 4345
+    },
+    {
+      "epoch": 2.021082002790265,
+      "grad_norm": 0.3523595333099365,
+      "learning_rate": 2.905933411610136e-06,
+      "loss": 0.3337,
+      "step": 4346
+    },
+    {
+      "epoch": 2.021547046969462,
+      "grad_norm": 0.35873571038246155,
+      "learning_rate": 2.9034765319596497e-06,
+      "loss": 0.3286,
+      "step": 4347
+    },
+    {
+      "epoch": 2.022012091148659,
+      "grad_norm": 0.38059335947036743,
+      "learning_rate": 2.9010202663468353e-06,
+      "loss": 0.3044,
+      "step": 4348
+    },
+    {
+      "epoch": 2.0224771353278563,
+      "grad_norm": 0.3723052144050598,
+      "learning_rate": 2.8985646154910887e-06,
+      "loss": 0.3459,
+      "step": 4349
+    },
+    {
+      "epoch": 2.0229421795070532,
+      "grad_norm": 0.3536323308944702,
+      "learning_rate": 2.896109580111634e-06,
+      "loss": 0.3042,
+      "step": 4350
+    },
+    {
+      "epoch": 2.02340722368625,
+      "grad_norm": 0.3543092608451843,
+      "learning_rate": 2.8936551609275078e-06,
+      "loss": 0.3361,
+      "step": 4351
+    },
+    {
+      "epoch": 2.023872267865447,
+      "grad_norm": 0.3618990182876587,
+      "learning_rate": 2.8912013586575733e-06,
+      "loss": 0.3393,
+      "step": 4352
+    },
+    {
+      "epoch": 2.0243373120446444,
+      "grad_norm": 0.35705238580703735,
+      "learning_rate": 2.8887481740205046e-06,
+      "loss": 0.3271,
+      "step": 4353
+    },
+    {
+      "epoch": 2.0248023562238413,
+      "grad_norm": 0.36953264474868774,
+      "learning_rate": 2.8862956077348054e-06,
+      "loss": 0.2905,
+      "step": 4354
+    },
+    {
+      "epoch": 2.025267400403038,
+      "grad_norm": 0.4037397801876068,
+      "learning_rate": 2.883843660518787e-06,
+      "loss": 0.3424,
+      "step": 4355
+    },
+    {
+      "epoch": 2.025732444582235,
+      "grad_norm": 0.3392297327518463,
+      "learning_rate": 2.881392333090589e-06,
+      "loss": 0.3267,
+      "step": 4356
+    },
+    {
+      "epoch": 2.0261974887614325,
+      "grad_norm": 0.36336854100227356,
+      "learning_rate": 2.8789416261681624e-06,
+      "loss": 0.3284,
+      "step": 4357
+    },
+    {
+      "epoch": 2.0266625329406294,
+      "grad_norm": 0.35528960824012756,
+      "learning_rate": 2.8764915404692805e-06,
+      "loss": 0.3275,
+      "step": 4358
+    },
+    {
+      "epoch": 2.0271275771198263,
+      "grad_norm": 0.36412903666496277,
+      "learning_rate": 2.874042076711536e-06,
+      "loss": 0.3135,
+      "step": 4359
+    },
+    {
+      "epoch": 2.0275926212990236,
+      "grad_norm": 0.3423042595386505,
+      "learning_rate": 2.871593235612333e-06,
+      "loss": 0.2972,
+      "step": 4360
+    },
+    {
+      "epoch": 2.0280576654782205,
+      "grad_norm": 0.36381906270980835,
+      "learning_rate": 2.8691450178889013e-06,
+      "loss": 0.3177,
+      "step": 4361
+    },
+    {
+      "epoch": 2.0285227096574174,
+      "grad_norm": 0.36212357878685,
+      "learning_rate": 2.8666974242582794e-06,
+      "loss": 0.337,
+      "step": 4362
+    },
+    {
+      "epoch": 2.0289877538366143,
+      "grad_norm": 0.38655251264572144,
+      "learning_rate": 2.864250455437333e-06,
+      "loss": 0.3042,
+      "step": 4363
+    },
+    {
+      "epoch": 2.0294527980158117,
+      "grad_norm": 0.32187148928642273,
+      "learning_rate": 2.8618041121427347e-06,
+      "loss": 0.3092,
+      "step": 4364
+    },
+    {
+      "epoch": 2.0299178421950086,
+      "grad_norm": 0.3990195691585541,
+      "learning_rate": 2.8593583950909833e-06,
+      "loss": 0.3392,
+      "step": 4365
+    },
+    {
+      "epoch": 2.0303828863742055,
+      "grad_norm": 0.35381731390953064,
+      "learning_rate": 2.8569133049983843e-06,
+      "loss": 0.325,
+      "step": 4366
+    },
+    {
+      "epoch": 2.0308479305534024,
+      "grad_norm": 0.3144361674785614,
+      "learning_rate": 2.8544688425810707e-06,
+      "loss": 0.2756,
+      "step": 4367
+    },
+    {
+      "epoch": 2.0313129747325998,
+      "grad_norm": 0.4486398994922638,
+      "learning_rate": 2.8520250085549807e-06,
+      "loss": 0.337,
+      "step": 4368
+    },
+    {
+      "epoch": 2.0317780189117967,
+      "grad_norm": 0.38106465339660645,
+      "learning_rate": 2.8495818036358756e-06,
+      "loss": 0.3487,
+      "step": 4369
+    },
+    {
+      "epoch": 2.0322430630909936,
+      "grad_norm": 0.34248411655426025,
+      "learning_rate": 2.8471392285393307e-06,
+      "loss": 0.3025,
+      "step": 4370
+    },
+    {
+      "epoch": 2.0327081072701905,
+      "grad_norm": 0.3708108067512512,
+      "learning_rate": 2.8446972839807384e-06,
+      "loss": 0.3159,
+      "step": 4371
+    },
+    {
+      "epoch": 2.033173151449388,
+      "grad_norm": 0.3808947503566742,
+      "learning_rate": 2.8422559706753004e-06,
+      "loss": 0.2909,
+      "step": 4372
+    },
+    {
+      "epoch": 2.0336381956285847,
+      "grad_norm": 0.36368271708488464,
+      "learning_rate": 2.8398152893380426e-06,
+      "loss": 0.3181,
+      "step": 4373
+    },
+    {
+      "epoch": 2.0341032398077816,
+      "grad_norm": 0.38620734214782715,
+      "learning_rate": 2.8373752406837963e-06,
+      "loss": 0.3181,
+      "step": 4374
+    },
+    {
+      "epoch": 2.034568283986979,
+      "grad_norm": 0.3720152974128723,
+      "learning_rate": 2.834935825427216e-06,
+      "loss": 0.335,
+      "step": 4375
+    },
+    {
+      "epoch": 2.035033328166176,
+      "grad_norm": 0.36220601201057434,
+      "learning_rate": 2.8324970442827627e-06,
+      "loss": 0.3028,
+      "step": 4376
+    },
+    {
+      "epoch": 2.035498372345373,
+      "grad_norm": 0.36199939250946045,
+      "learning_rate": 2.8300588979647202e-06,
+      "loss": 0.3198,
+      "step": 4377
+    },
+    {
+      "epoch": 2.0359634165245697,
+      "grad_norm": 0.3544287085533142,
+      "learning_rate": 2.827621387187178e-06,
+      "loss": 0.3024,
+      "step": 4378
+    },
+    {
+      "epoch": 2.036428460703767,
+      "grad_norm": 0.37387457489967346,
+      "learning_rate": 2.825184512664048e-06,
+      "loss": 0.3213,
+      "step": 4379
+    },
+    {
+      "epoch": 2.036893504882964,
+      "grad_norm": 0.37123391032218933,
+      "learning_rate": 2.8227482751090445e-06,
+      "loss": 0.2892,
+      "step": 4380
+    },
+    {
+      "epoch": 2.037358549062161,
+      "grad_norm": 0.34186607599258423,
+      "learning_rate": 2.8203126752357067e-06,
+      "loss": 0.2994,
+      "step": 4381
+    },
+    {
+      "epoch": 2.0378235932413578,
+      "grad_norm": 0.34526383876800537,
+      "learning_rate": 2.8178777137573814e-06,
+      "loss": 0.3343,
+      "step": 4382
+    },
+    {
+      "epoch": 2.038288637420555,
+      "grad_norm": 0.3409002721309662,
+      "learning_rate": 2.8154433913872314e-06,
+      "loss": 0.3204,
+      "step": 4383
+    },
+    {
+      "epoch": 2.038753681599752,
+      "grad_norm": 0.3608821630477905,
+      "learning_rate": 2.8130097088382256e-06,
+      "loss": 0.3083,
+      "step": 4384
+    },
+    {
+      "epoch": 2.039218725778949,
+      "grad_norm": 0.39243847131729126,
+      "learning_rate": 2.8105766668231548e-06,
+      "loss": 0.2993,
+      "step": 4385
+    },
+    {
+      "epoch": 2.039683769958146,
+      "grad_norm": 0.34924769401550293,
+      "learning_rate": 2.8081442660546126e-06,
+      "loss": 0.3602,
+      "step": 4386
+    },
+    {
+      "epoch": 2.040148814137343,
+      "grad_norm": 0.3145473003387451,
+      "learning_rate": 2.8057125072450143e-06,
+      "loss": 0.3078,
+      "step": 4387
+    },
+    {
+      "epoch": 2.04061385831654,
+      "grad_norm": 0.35963907837867737,
+      "learning_rate": 2.8032813911065795e-06,
+      "loss": 0.3269,
+      "step": 4388
+    },
+    {
+      "epoch": 2.041078902495737,
+      "grad_norm": 0.34630414843559265,
+      "learning_rate": 2.8008509183513444e-06,
+      "loss": 0.2996,
+      "step": 4389
+    },
+    {
+      "epoch": 2.0415439466749343,
+      "grad_norm": 0.32224032282829285,
+      "learning_rate": 2.7984210896911525e-06,
+      "loss": 0.3007,
+      "step": 4390
+    },
+    {
+      "epoch": 2.0420089908541312,
+      "grad_norm": 0.36333921551704407,
+      "learning_rate": 2.795991905837665e-06,
+      "loss": 0.3216,
+      "step": 4391
+    },
+    {
+      "epoch": 2.042474035033328,
+      "grad_norm": 0.3574282228946686,
+      "learning_rate": 2.793563367502346e-06,
+      "loss": 0.282,
+      "step": 4392
+    },
+    {
+      "epoch": 2.042939079212525,
+      "grad_norm": 0.3617522418498993,
+      "learning_rate": 2.791135475396477e-06,
+      "loss": 0.3471,
+      "step": 4393
+    },
+    {
+      "epoch": 2.0434041233917224,
+      "grad_norm": 0.3639947474002838,
+      "learning_rate": 2.7887082302311486e-06,
+      "loss": 0.3791,
+      "step": 4394
+    },
+    {
+      "epoch": 2.0438691675709193,
+      "grad_norm": 0.3631550967693329,
+      "learning_rate": 2.786281632717264e-06,
+      "loss": 0.292,
+      "step": 4395
+    },
+    {
+      "epoch": 2.044334211750116,
+      "grad_norm": 0.3522656261920929,
+      "learning_rate": 2.7838556835655304e-06,
+      "loss": 0.3327,
+      "step": 4396
+    },
+    {
+      "epoch": 2.044799255929313,
+      "grad_norm": 0.3434469699859619,
+      "learning_rate": 2.781430383486468e-06,
+      "loss": 0.3274,
+      "step": 4397
+    },
+    {
+      "epoch": 2.0452643001085105,
+      "grad_norm": 0.36457687616348267,
+      "learning_rate": 2.779005733190412e-06,
+      "loss": 0.3083,
+      "step": 4398
+    },
+    {
+      "epoch": 2.0457293442877074,
+      "grad_norm": 0.3624141812324524,
+      "learning_rate": 2.7765817333874984e-06,
+      "loss": 0.2975,
+      "step": 4399
+    },
+    {
+      "epoch": 2.0461943884669043,
+      "grad_norm": 0.3684295415878296,
+      "learning_rate": 2.7741583847876816e-06,
+      "loss": 0.3205,
+      "step": 4400
+    },
+    {
+      "epoch": 2.046659432646101,
+      "grad_norm": 0.3599317967891693,
+      "learning_rate": 2.7717356881007185e-06,
+      "loss": 0.3371,
+      "step": 4401
+    },
+    {
+      "epoch": 2.0471244768252985,
+      "grad_norm": 0.3688468039035797,
+      "learning_rate": 2.769313644036179e-06,
+      "loss": 0.3187,
+      "step": 4402
+    },
+    {
+      "epoch": 2.0475895210044954,
+      "grad_norm": 0.43112462759017944,
+      "learning_rate": 2.766892253303438e-06,
+      "loss": 0.3259,
+      "step": 4403
+    },
+    {
+      "epoch": 2.0480545651836923,
+      "grad_norm": 0.38703638315200806,
+      "learning_rate": 2.7644715166116835e-06,
+      "loss": 0.2869,
+      "step": 4404
+    },
+    {
+      "epoch": 2.0485196093628897,
+      "grad_norm": 0.39289000630378723,
+      "learning_rate": 2.7620514346699103e-06,
+      "loss": 0.3227,
+      "step": 4405
+    },
+    {
+      "epoch": 2.0489846535420866,
+      "grad_norm": 0.403656542301178,
+      "learning_rate": 2.7596320081869214e-06,
+      "loss": 0.3242,
+      "step": 4406
+    },
+    {
+      "epoch": 2.0494496977212835,
+      "grad_norm": 0.42674005031585693,
+      "learning_rate": 2.7572132378713255e-06,
+      "loss": 0.3151,
+      "step": 4407
+    },
+    {
+      "epoch": 2.0499147419004804,
+      "grad_norm": 0.35708075761795044,
+      "learning_rate": 2.754795124431544e-06,
+      "loss": 0.3025,
+      "step": 4408
+    },
+    {
+      "epoch": 2.0503797860796777,
+      "grad_norm": 0.41090068221092224,
+      "learning_rate": 2.752377668575799e-06,
+      "loss": 0.3488,
+      "step": 4409
+    },
+    {
+      "epoch": 2.0508448302588747,
+      "grad_norm": 0.3659050166606903,
+      "learning_rate": 2.749960871012129e-06,
+      "loss": 0.3236,
+      "step": 4410
+    },
+    {
+      "epoch": 2.0513098744380716,
+      "grad_norm": 0.4140544533729553,
+      "learning_rate": 2.7475447324483697e-06,
+      "loss": 0.317,
+      "step": 4411
+    },
+    {
+      "epoch": 2.0517749186172685,
+      "grad_norm": 0.3611926734447479,
+      "learning_rate": 2.7451292535921738e-06,
+      "loss": 0.2912,
+      "step": 4412
+    },
+    {
+      "epoch": 2.052239962796466,
+      "grad_norm": 0.3759238123893738,
+      "learning_rate": 2.7427144351509904e-06,
+      "loss": 0.3198,
+      "step": 4413
+    },
+    {
+      "epoch": 2.0527050069756627,
+      "grad_norm": 0.37930646538734436,
+      "learning_rate": 2.7403002778320865e-06,
+      "loss": 0.3247,
+      "step": 4414
+    },
+    {
+      "epoch": 2.0531700511548596,
+      "grad_norm": 0.3774726390838623,
+      "learning_rate": 2.737886782342524e-06,
+      "loss": 0.3085,
+      "step": 4415
+    },
+    {
+      "epoch": 2.0536350953340565,
+      "grad_norm": 0.4020675718784332,
+      "learning_rate": 2.735473949389179e-06,
+      "loss": 0.2979,
+      "step": 4416
+    },
+    {
+      "epoch": 2.054100139513254,
+      "grad_norm": 0.4103391766548157,
+      "learning_rate": 2.733061779678732e-06,
+      "loss": 0.3147,
+      "step": 4417
+    },
+    {
+      "epoch": 2.054565183692451,
+      "grad_norm": 0.41063860058784485,
+      "learning_rate": 2.7306502739176686e-06,
+      "loss": 0.3333,
+      "step": 4418
+    },
+    {
+      "epoch": 2.0550302278716477,
+      "grad_norm": 0.36508336663246155,
+      "learning_rate": 2.728239432812277e-06,
+      "loss": 0.3225,
+      "step": 4419
+    },
+    {
+      "epoch": 2.055495272050845,
+      "grad_norm": 0.34332776069641113,
+      "learning_rate": 2.7258292570686566e-06,
+      "loss": 0.2756,
+      "step": 4420
+    },
+    {
+      "epoch": 2.055960316230042,
+      "grad_norm": 0.400511771440506,
+      "learning_rate": 2.7234197473927054e-06,
+      "loss": 0.3704,
+      "step": 4421
+    },
+    {
+      "epoch": 2.056425360409239,
+      "grad_norm": 0.32971513271331787,
+      "learning_rate": 2.7210109044901335e-06,
+      "loss": 0.2891,
+      "step": 4422
+    },
+    {
+      "epoch": 2.0568904045884358,
+      "grad_norm": 0.357746422290802,
+      "learning_rate": 2.7186027290664474e-06,
+      "loss": 0.3439,
+      "step": 4423
+    },
+    {
+      "epoch": 2.057355448767633,
+      "grad_norm": 0.355873703956604,
+      "learning_rate": 2.716195221826967e-06,
+      "loss": 0.329,
+      "step": 4424
+    },
+    {
+      "epoch": 2.05782049294683,
+      "grad_norm": 0.3346231281757355,
+      "learning_rate": 2.7137883834768076e-06,
+      "loss": 0.3268,
+      "step": 4425
+    },
+    {
+      "epoch": 2.058285537126027,
+      "grad_norm": 0.35958102345466614,
+      "learning_rate": 2.711382214720898e-06,
+      "loss": 0.3305,
+      "step": 4426
+    },
+    {
+      "epoch": 2.058750581305224,
+      "grad_norm": 0.3547190725803375,
+      "learning_rate": 2.708976716263961e-06,
+      "loss": 0.3142,
+      "step": 4427
+    },
+    {
+      "epoch": 2.059215625484421,
+      "grad_norm": 0.3430922329425812,
+      "learning_rate": 2.7065718888105298e-06,
+      "loss": 0.3278,
+      "step": 4428
+    },
+    {
+      "epoch": 2.059680669663618,
+      "grad_norm": 0.36960989236831665,
+      "learning_rate": 2.7041677330649408e-06,
+      "loss": 0.3641,
+      "step": 4429
+    },
+    {
+      "epoch": 2.060145713842815,
+      "grad_norm": 0.3565279543399811,
+      "learning_rate": 2.7017642497313324e-06,
+      "loss": 0.3027,
+      "step": 4430
+    },
+    {
+      "epoch": 2.060610758022012,
+      "grad_norm": 0.3367528021335602,
+      "learning_rate": 2.6993614395136454e-06,
+      "loss": 0.3133,
+      "step": 4431
+    },
+    {
+      "epoch": 2.0610758022012092,
+      "grad_norm": 0.35852697491645813,
+      "learning_rate": 2.6969593031156205e-06,
+      "loss": 0.342,
+      "step": 4432
+    },
+    {
+      "epoch": 2.061540846380406,
+      "grad_norm": 0.34888017177581787,
+      "learning_rate": 2.694557841240809e-06,
+      "loss": 0.3282,
+      "step": 4433
+    },
+    {
+      "epoch": 2.062005890559603,
+      "grad_norm": 0.35499727725982666,
+      "learning_rate": 2.692157054592557e-06,
+      "loss": 0.2948,
+      "step": 4434
+    },
+    {
+      "epoch": 2.0624709347388004,
+      "grad_norm": 0.33716249465942383,
+      "learning_rate": 2.689756943874019e-06,
+      "loss": 0.2994,
+      "step": 4435
+    },
+    {
+      "epoch": 2.0629359789179973,
+      "grad_norm": 0.33978933095932007,
+      "learning_rate": 2.687357509788143e-06,
+      "loss": 0.3069,
+      "step": 4436
+    },
+    {
+      "epoch": 2.063401023097194,
+      "grad_norm": 0.3515429198741913,
+      "learning_rate": 2.684958753037691e-06,
+      "loss": 0.3488,
+      "step": 4437
+    },
+    {
+      "epoch": 2.063866067276391,
+      "grad_norm": 0.3632570207118988,
+      "learning_rate": 2.682560674325215e-06,
+      "loss": 0.326,
+      "step": 4438
+    },
+    {
+      "epoch": 2.0643311114555885,
+      "grad_norm": 0.36300498247146606,
+      "learning_rate": 2.680163274353075e-06,
+      "loss": 0.2862,
+      "step": 4439
+    },
+    {
+      "epoch": 2.0647961556347854,
+      "grad_norm": 0.3675801753997803,
+      "learning_rate": 2.6777665538234292e-06,
+      "loss": 0.3299,
+      "step": 4440
+    },
+    {
+      "epoch": 2.0652611998139823,
+      "grad_norm": 0.3985215127468109,
+      "learning_rate": 2.6753705134382425e-06,
+      "loss": 0.3205,
+      "step": 4441
+    },
+    {
+      "epoch": 2.065726243993179,
+      "grad_norm": 0.35572221875190735,
+      "learning_rate": 2.6729751538992704e-06,
+      "loss": 0.3302,
+      "step": 4442
+    },
+    {
+      "epoch": 2.0661912881723765,
+      "grad_norm": 0.35556313395500183,
+      "learning_rate": 2.67058047590808e-06,
+      "loss": 0.2882,
+      "step": 4443
+    },
+    {
+      "epoch": 2.0666563323515734,
+      "grad_norm": 0.4020705819129944,
+      "learning_rate": 2.6681864801660284e-06,
+      "loss": 0.3273,
+      "step": 4444
+    },
+    {
+      "epoch": 2.0671213765307703,
+      "grad_norm": 0.40115731954574585,
+      "learning_rate": 2.6657931673742834e-06,
+      "loss": 0.3591,
+      "step": 4445
+    },
+    {
+      "epoch": 2.0675864207099672,
+      "grad_norm": 0.320512592792511,
+      "learning_rate": 2.6634005382338025e-06,
+      "loss": 0.2962,
+      "step": 4446
+    },
+    {
+      "epoch": 2.0680514648891646,
+      "grad_norm": 0.35881075263023376,
+      "learning_rate": 2.6610085934453523e-06,
+      "loss": 0.3679,
+      "step": 4447
+    },
+    {
+      "epoch": 2.0685165090683615,
+      "grad_norm": 0.35448718070983887,
+      "learning_rate": 2.6586173337094904e-06,
+      "loss": 0.2736,
+      "step": 4448
+    },
+    {
+      "epoch": 2.0689815532475584,
+      "grad_norm": 0.378061980009079,
+      "learning_rate": 2.656226759726582e-06,
+      "loss": 0.3526,
+      "step": 4449
+    },
+    {
+      "epoch": 2.0694465974267557,
+      "grad_norm": 0.3106938600540161,
+      "learning_rate": 2.6538368721967838e-06,
+      "loss": 0.2833,
+      "step": 4450
+    },
+    {
+      "epoch": 2.0699116416059526,
+      "grad_norm": 0.3516807556152344,
+      "learning_rate": 2.6514476718200566e-06,
+      "loss": 0.3392,
+      "step": 4451
+    },
+    {
+      "epoch": 2.0703766857851496,
+      "grad_norm": 0.3462466299533844,
+      "learning_rate": 2.649059159296158e-06,
+      "loss": 0.3042,
+      "step": 4452
+    },
+    {
+      "epoch": 2.0708417299643465,
+      "grad_norm": 0.36509615182876587,
+      "learning_rate": 2.646671335324647e-06,
+      "loss": 0.3551,
+      "step": 4453
+    },
+    {
+      "epoch": 2.071306774143544,
+      "grad_norm": 0.347945898771286,
+      "learning_rate": 2.644284200604874e-06,
+      "loss": 0.3431,
+      "step": 4454
+    },
+    {
+      "epoch": 2.0717718183227407,
+      "grad_norm": 0.3239026963710785,
+      "learning_rate": 2.641897755835997e-06,
+      "loss": 0.2844,
+      "step": 4455
+    },
+    {
+      "epoch": 2.0722368625019376,
+      "grad_norm": 0.35373586416244507,
+      "learning_rate": 2.6395120017169627e-06,
+      "loss": 0.3261,
+      "step": 4456
+    },
+    {
+      "epoch": 2.0727019066811345,
+      "grad_norm": 0.3564927875995636,
+      "learning_rate": 2.6371269389465227e-06,
+      "loss": 0.3213,
+      "step": 4457
+    },
+    {
+      "epoch": 2.073166950860332,
+      "grad_norm": 0.3221034109592438,
+      "learning_rate": 2.6347425682232196e-06,
+      "loss": 0.2954,
+      "step": 4458
+    },
+    {
+      "epoch": 2.0736319950395288,
+      "grad_norm": 0.36184707283973694,
+      "learning_rate": 2.6323588902454013e-06,
+      "loss": 0.3331,
+      "step": 4459
+    },
+    {
+      "epoch": 2.0740970392187257,
+      "grad_norm": 0.3728765845298767,
+      "learning_rate": 2.629975905711204e-06,
+      "loss": 0.3231,
+      "step": 4460
+    },
+    {
+      "epoch": 2.0745620833979226,
+      "grad_norm": 0.35533544421195984,
+      "learning_rate": 2.6275936153185694e-06,
+      "loss": 0.3132,
+      "step": 4461
+    },
+    {
+      "epoch": 2.07502712757712,
+      "grad_norm": 0.337354451417923,
+      "learning_rate": 2.6252120197652277e-06,
+      "loss": 0.2933,
+      "step": 4462
+    },
+    {
+      "epoch": 2.075492171756317,
+      "grad_norm": 0.33192774653434753,
+      "learning_rate": 2.622831119748711e-06,
+      "loss": 0.3354,
+      "step": 4463
+    },
+    {
+      "epoch": 2.0759572159355137,
+      "grad_norm": 0.36651748418807983,
+      "learning_rate": 2.620450915966346e-06,
+      "loss": 0.3294,
+      "step": 4464
+    },
+    {
+      "epoch": 2.076422260114711,
+      "grad_norm": 0.3576674461364746,
+      "learning_rate": 2.618071409115259e-06,
+      "loss": 0.3174,
+      "step": 4465
+    },
+    {
+      "epoch": 2.076887304293908,
+      "grad_norm": 0.33194971084594727,
+      "learning_rate": 2.615692599892364e-06,
+      "loss": 0.3175,
+      "step": 4466
+    },
+    {
+      "epoch": 2.077352348473105,
+      "grad_norm": 0.40041840076446533,
+      "learning_rate": 2.6133144889943808e-06,
+      "loss": 0.317,
+      "step": 4467
+    },
+    {
+      "epoch": 2.077817392652302,
+      "grad_norm": 0.39717933535575867,
+      "learning_rate": 2.6109370771178155e-06,
+      "loss": 0.3476,
+      "step": 4468
+    },
+    {
+      "epoch": 2.078282436831499,
+      "grad_norm": 0.35431498289108276,
+      "learning_rate": 2.6085603649589723e-06,
+      "loss": 0.3019,
+      "step": 4469
+    },
+    {
+      "epoch": 2.078747481010696,
+      "grad_norm": 0.3609691560268402,
+      "learning_rate": 2.6061843532139563e-06,
+      "loss": 0.3173,
+      "step": 4470
+    },
+    {
+      "epoch": 2.079212525189893,
+      "grad_norm": 0.37636587023735046,
+      "learning_rate": 2.6038090425786577e-06,
+      "loss": 0.3253,
+      "step": 4471
+    },
+    {
+      "epoch": 2.07967756936909,
+      "grad_norm": 0.34753209352493286,
+      "learning_rate": 2.601434433748771e-06,
+      "loss": 0.2982,
+      "step": 4472
+    },
+    {
+      "epoch": 2.0801426135482872,
+      "grad_norm": 0.3801688849925995,
+      "learning_rate": 2.5990605274197763e-06,
+      "loss": 0.336,
+      "step": 4473
+    },
+    {
+      "epoch": 2.080607657727484,
+      "grad_norm": 0.37925639748573303,
+      "learning_rate": 2.596687324286954e-06,
+      "loss": 0.3277,
+      "step": 4474
+    },
+    {
+      "epoch": 2.081072701906681,
+      "grad_norm": 0.3775731027126312,
+      "learning_rate": 2.5943148250453774e-06,
+      "loss": 0.3043,
+      "step": 4475
+    },
+    {
+      "epoch": 2.0815377460858784,
+      "grad_norm": 0.37141674757003784,
+      "learning_rate": 2.5919430303899144e-06,
+      "loss": 0.3141,
+      "step": 4476
+    },
+    {
+      "epoch": 2.0820027902650753,
+      "grad_norm": 0.38826245069503784,
+      "learning_rate": 2.589571941015222e-06,
+      "loss": 0.312,
+      "step": 4477
+    },
+    {
+      "epoch": 2.082467834444272,
+      "grad_norm": 0.3552106022834778,
+      "learning_rate": 2.587201557615756e-06,
+      "loss": 0.2905,
+      "step": 4478
+    },
+    {
+      "epoch": 2.082932878623469,
+      "grad_norm": 0.40252459049224854,
+      "learning_rate": 2.584831880885761e-06,
+      "loss": 0.3681,
+      "step": 4479
+    },
+    {
+      "epoch": 2.0833979228026664,
+      "grad_norm": 0.35997259616851807,
+      "learning_rate": 2.58246291151928e-06,
+      "loss": 0.3051,
+      "step": 4480
+    },
+    {
+      "epoch": 2.0838629669818634,
+      "grad_norm": 0.3523218333721161,
+      "learning_rate": 2.580094650210142e-06,
+      "loss": 0.2918,
+      "step": 4481
+    },
+    {
+      "epoch": 2.0843280111610603,
+      "grad_norm": 0.4533151090145111,
+      "learning_rate": 2.577727097651976e-06,
+      "loss": 0.3553,
+      "step": 4482
+    },
+    {
+      "epoch": 2.084793055340257,
+      "grad_norm": 0.35789787769317627,
+      "learning_rate": 2.575360254538195e-06,
+      "loss": 0.2959,
+      "step": 4483
+    },
+    {
+      "epoch": 2.0852580995194545,
+      "grad_norm": 0.39699408411979675,
+      "learning_rate": 2.5729941215620148e-06,
+      "loss": 0.3356,
+      "step": 4484
+    },
+    {
+      "epoch": 2.0857231436986514,
+      "grad_norm": 0.3448418974876404,
+      "learning_rate": 2.5706286994164315e-06,
+      "loss": 0.2892,
+      "step": 4485
+    },
+    {
+      "epoch": 2.0861881878778483,
+      "grad_norm": 0.3884351849555969,
+      "learning_rate": 2.568263988794242e-06,
+      "loss": 0.3183,
+      "step": 4486
+    },
+    {
+      "epoch": 2.0866532320570452,
+      "grad_norm": 0.4005531072616577,
+      "learning_rate": 2.56589999038803e-06,
+      "loss": 0.3394,
+      "step": 4487
+    },
+    {
+      "epoch": 2.0871182762362426,
+      "grad_norm": 0.3274531960487366,
+      "learning_rate": 2.563536704890176e-06,
+      "loss": 0.2915,
+      "step": 4488
+    },
+    {
+      "epoch": 2.0875833204154395,
+      "grad_norm": 0.418525367975235,
+      "learning_rate": 2.5611741329928436e-06,
+      "loss": 0.347,
+      "step": 4489
+    },
+    {
+      "epoch": 2.0880483645946364,
+      "grad_norm": 0.3738149106502533,
+      "learning_rate": 2.558812275387995e-06,
+      "loss": 0.3231,
+      "step": 4490
+    },
+    {
+      "epoch": 2.0885134087738333,
+      "grad_norm": 0.3588385581970215,
+      "learning_rate": 2.556451132767377e-06,
+      "loss": 0.3281,
+      "step": 4491
+    },
+    {
+      "epoch": 2.0889784529530306,
+      "grad_norm": 0.3635399639606476,
+      "learning_rate": 2.554090705822533e-06,
+      "loss": 0.3367,
+      "step": 4492
+    },
+    {
+      "epoch": 2.0894434971322275,
+      "grad_norm": 0.342986136674881,
+      "learning_rate": 2.5517309952447887e-06,
+      "loss": 0.3051,
+      "step": 4493
+    },
+    {
+      "epoch": 2.0899085413114245,
+      "grad_norm": 0.3733537495136261,
+      "learning_rate": 2.549372001725272e-06,
+      "loss": 0.3154,
+      "step": 4494
+    },
+    {
+      "epoch": 2.090373585490622,
+      "grad_norm": 0.404184490442276,
+      "learning_rate": 2.547013725954887e-06,
+      "loss": 0.3149,
+      "step": 4495
+    },
+    {
+      "epoch": 2.0908386296698187,
+      "grad_norm": 0.39761337637901306,
+      "learning_rate": 2.5446561686243397e-06,
+      "loss": 0.3353,
+      "step": 4496
+    },
+    {
+      "epoch": 2.0913036738490156,
+      "grad_norm": 0.36512845754623413,
+      "learning_rate": 2.5422993304241163e-06,
+      "loss": 0.3217,
+      "step": 4497
+    },
+    {
+      "epoch": 2.0917687180282125,
+      "grad_norm": 0.35249972343444824,
+      "learning_rate": 2.5399432120444985e-06,
+      "loss": 0.2946,
+      "step": 4498
+    },
+    {
+      "epoch": 2.09223376220741,
+      "grad_norm": 0.3695544898509979,
+      "learning_rate": 2.537587814175554e-06,
+      "loss": 0.3332,
+      "step": 4499
+    },
+    {
+      "epoch": 2.0926988063866068,
+      "grad_norm": 0.3467404246330261,
+      "learning_rate": 2.5352331375071437e-06,
+      "loss": 0.3147,
+      "step": 4500
+    },
+    {
+      "epoch": 2.0931638505658037,
+      "grad_norm": 0.3637906610965729,
+      "learning_rate": 2.53287918272891e-06,
+      "loss": 0.3054,
+      "step": 4501
+    },
+    {
+      "epoch": 2.0936288947450006,
+      "grad_norm": 0.3547385036945343,
+      "learning_rate": 2.5305259505302914e-06,
+      "loss": 0.3231,
+      "step": 4502
+    },
+    {
+      "epoch": 2.094093938924198,
+      "grad_norm": 0.36574193835258484,
+      "learning_rate": 2.5281734416005107e-06,
+      "loss": 0.3125,
+      "step": 4503
+    },
+    {
+      "epoch": 2.094558983103395,
+      "grad_norm": 0.3319716155529022,
+      "learning_rate": 2.5258216566285758e-06,
+      "loss": 0.3083,
+      "step": 4504
+    },
+    {
+      "epoch": 2.0950240272825917,
+      "grad_norm": 0.3524357080459595,
+      "learning_rate": 2.5234705963032917e-06,
+      "loss": 0.3457,
+      "step": 4505
+    },
+    {
+      "epoch": 2.095489071461789,
+      "grad_norm": 0.34479501843452454,
+      "learning_rate": 2.5211202613132413e-06,
+      "loss": 0.3415,
+      "step": 4506
+    },
+    {
+      "epoch": 2.095954115640986,
+      "grad_norm": 0.3542589247226715,
+      "learning_rate": 2.5187706523468034e-06,
+      "loss": 0.3051,
+      "step": 4507
+    },
+    {
+      "epoch": 2.096419159820183,
+      "grad_norm": 0.3570210337638855,
+      "learning_rate": 2.516421770092136e-06,
+      "loss": 0.3496,
+      "step": 4508
+    },
+    {
+      "epoch": 2.09688420399938,
+      "grad_norm": 0.3366619050502777,
+      "learning_rate": 2.5140736152371916e-06,
+      "loss": 0.3166,
+      "step": 4509
+    },
+    {
+      "epoch": 2.097349248178577,
+      "grad_norm": 0.3258604407310486,
+      "learning_rate": 2.5117261884697066e-06,
+      "loss": 0.2912,
+      "step": 4510
+    },
+    {
+      "epoch": 2.097814292357774,
+      "grad_norm": 0.3651737868785858,
+      "learning_rate": 2.509379490477204e-06,
+      "loss": 0.3397,
+      "step": 4511
+    },
+    {
+      "epoch": 2.098279336536971,
+      "grad_norm": 0.335183322429657,
+      "learning_rate": 2.507033521946992e-06,
+      "loss": 0.2965,
+      "step": 4512
+    },
+    {
+      "epoch": 2.098744380716168,
+      "grad_norm": 0.37854668498039246,
+      "learning_rate": 2.5046882835661694e-06,
+      "loss": 0.3288,
+      "step": 4513
+    },
+    {
+      "epoch": 2.099209424895365,
+      "grad_norm": 0.39023512601852417,
+      "learning_rate": 2.502343776021615e-06,
+      "loss": 0.3475,
+      "step": 4514
+    },
+    {
+      "epoch": 2.099674469074562,
+      "grad_norm": 0.37429752945899963,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.325,
+      "step": 4515
+    },
+    {
+      "epoch": 2.100139513253759,
+      "grad_norm": 0.3756406605243683,
+      "learning_rate": 2.4976569561877774e-06,
+      "loss": 0.327,
+      "step": 4516
+    },
+    {
+      "epoch": 2.100604557432956,
+      "grad_norm": 0.36603933572769165,
+      "learning_rate": 2.4953146452711866e-06,
+      "loss": 0.341,
+      "step": 4517
+    },
+    {
+      "epoch": 2.1010696016121533,
+      "grad_norm": 0.34370267391204834,
+      "learning_rate": 2.492973067936251e-06,
+      "loss": 0.317,
+      "step": 4518
+    },
+    {
+      "epoch": 2.10153464579135,
+      "grad_norm": 0.40021899342536926,
+      "learning_rate": 2.490632224868783e-06,
+      "loss": 0.3222,
+      "step": 4519
+    },
+    {
+      "epoch": 2.101999689970547,
+      "grad_norm": 0.39490172266960144,
+      "learning_rate": 2.4882921167543745e-06,
+      "loss": 0.3088,
+      "step": 4520
+    },
+    {
+      "epoch": 2.102464734149744,
+      "grad_norm": 0.37795066833496094,
+      "learning_rate": 2.485952744278407e-06,
+      "loss": 0.3156,
+      "step": 4521
+    },
+    {
+      "epoch": 2.1029297783289413,
+      "grad_norm": 0.3622766435146332,
+      "learning_rate": 2.483614108126045e-06,
+      "loss": 0.3002,
+      "step": 4522
+    },
+    {
+      "epoch": 2.1033948225081383,
+      "grad_norm": 0.33893483877182007,
+      "learning_rate": 2.4812762089822384e-06,
+      "loss": 0.3183,
+      "step": 4523
+    },
+    {
+      "epoch": 2.103859866687335,
+      "grad_norm": 0.3871879279613495,
+      "learning_rate": 2.478939047531716e-06,
+      "loss": 0.3391,
+      "step": 4524
+    },
+    {
+      "epoch": 2.1043249108665325,
+      "grad_norm": 0.34787362813949585,
+      "learning_rate": 2.4766026244589986e-06,
+      "loss": 0.2944,
+      "step": 4525
+    },
+    {
+      "epoch": 2.1047899550457294,
+      "grad_norm": 0.36640048027038574,
+      "learning_rate": 2.4742669404483825e-06,
+      "loss": 0.3368,
+      "step": 4526
+    },
+    {
+      "epoch": 2.1052549992249263,
+      "grad_norm": 0.36439046263694763,
+      "learning_rate": 2.471931996183956e-06,
+      "loss": 0.3047,
+      "step": 4527
+    },
+    {
+      "epoch": 2.105720043404123,
+      "grad_norm": 0.33959466218948364,
+      "learning_rate": 2.4695977923495816e-06,
+      "loss": 0.2989,
+      "step": 4528
+    },
+    {
+      "epoch": 2.1061850875833206,
+      "grad_norm": 0.3902963697910309,
+      "learning_rate": 2.4672643296289145e-06,
+      "loss": 0.3183,
+      "step": 4529
+    },
+    {
+      "epoch": 2.1066501317625175,
+      "grad_norm": 0.4295981228351593,
+      "learning_rate": 2.464931608705384e-06,
+      "loss": 0.2918,
+      "step": 4530
+    },
+    {
+      "epoch": 2.1071151759417144,
+      "grad_norm": 0.36755356192588806,
+      "learning_rate": 2.462599630262209e-06,
+      "loss": 0.324,
+      "step": 4531
+    },
+    {
+      "epoch": 2.1075802201209113,
+      "grad_norm": 0.3295392394065857,
+      "learning_rate": 2.4602683949823853e-06,
+      "loss": 0.3026,
+      "step": 4532
+    },
+    {
+      "epoch": 2.1080452643001086,
+      "grad_norm": 0.34905338287353516,
+      "learning_rate": 2.457937903548695e-06,
+      "loss": 0.3265,
+      "step": 4533
+    },
+    {
+      "epoch": 2.1085103084793055,
+      "grad_norm": 0.37717199325561523,
+      "learning_rate": 2.4556081566437025e-06,
+      "loss": 0.3326,
+      "step": 4534
+    },
+    {
+      "epoch": 2.1089753526585024,
+      "grad_norm": 0.34889674186706543,
+      "learning_rate": 2.453279154949753e-06,
+      "loss": 0.3113,
+      "step": 4535
+    },
+    {
+      "epoch": 2.1094403968377,
+      "grad_norm": 0.35179656744003296,
+      "learning_rate": 2.4509508991489704e-06,
+      "loss": 0.3188,
+      "step": 4536
+    },
+    {
+      "epoch": 2.1099054410168967,
+      "grad_norm": 0.35203817486763,
+      "learning_rate": 2.4486233899232674e-06,
+      "loss": 0.3106,
+      "step": 4537
+    },
+    {
+      "epoch": 2.1103704851960936,
+      "grad_norm": 0.3592386841773987,
+      "learning_rate": 2.4462966279543287e-06,
+      "loss": 0.3241,
+      "step": 4538
+    },
+    {
+      "epoch": 2.1108355293752905,
+      "grad_norm": 0.37422919273376465,
+      "learning_rate": 2.4439706139236295e-06,
+      "loss": 0.2891,
+      "step": 4539
+    },
+    {
+      "epoch": 2.111300573554488,
+      "grad_norm": 0.37109753489494324,
+      "learning_rate": 2.4416453485124196e-06,
+      "loss": 0.3353,
+      "step": 4540
+    },
+    {
+      "epoch": 2.1117656177336848,
+      "grad_norm": 0.35821226239204407,
+      "learning_rate": 2.4393208324017294e-06,
+      "loss": 0.3409,
+      "step": 4541
+    },
+    {
+      "epoch": 2.1122306619128817,
+      "grad_norm": 0.3365572690963745,
+      "learning_rate": 2.4369970662723756e-06,
+      "loss": 0.2886,
+      "step": 4542
+    },
+    {
+      "epoch": 2.1126957060920786,
+      "grad_norm": 0.3579605221748352,
+      "learning_rate": 2.4346740508049484e-06,
+      "loss": 0.3297,
+      "step": 4543
+    },
+    {
+      "epoch": 2.113160750271276,
+      "grad_norm": 0.32147571444511414,
+      "learning_rate": 2.432351786679822e-06,
+      "loss": 0.2968,
+      "step": 4544
+    },
+    {
+      "epoch": 2.113625794450473,
+      "grad_norm": 0.34730449318885803,
+      "learning_rate": 2.430030274577151e-06,
+      "loss": 0.3432,
+      "step": 4545
+    },
+    {
+      "epoch": 2.1140908386296697,
+      "grad_norm": 0.3238038718700409,
+      "learning_rate": 2.4277095151768698e-06,
+      "loss": 0.2922,
+      "step": 4546
+    },
+    {
+      "epoch": 2.1145558828088666,
+      "grad_norm": 0.3770063817501068,
+      "learning_rate": 2.4253895091586883e-06,
+      "loss": 0.2981,
+      "step": 4547
+    },
+    {
+      "epoch": 2.115020926988064,
+      "grad_norm": 0.39007771015167236,
+      "learning_rate": 2.423070257202101e-06,
+      "loss": 0.3372,
+      "step": 4548
+    },
+    {
+      "epoch": 2.115485971167261,
+      "grad_norm": 0.3754151463508606,
+      "learning_rate": 2.420751759986376e-06,
+      "loss": 0.318,
+      "step": 4549
+    },
+    {
+      "epoch": 2.115951015346458,
+      "grad_norm": 0.3623104989528656,
+      "learning_rate": 2.4184340181905675e-06,
+      "loss": 0.3214,
+      "step": 4550
+    },
+    {
+      "epoch": 2.116416059525655,
+      "grad_norm": 0.34606826305389404,
+      "learning_rate": 2.4161170324935e-06,
+      "loss": 0.3351,
+      "step": 4551
+    },
+    {
+      "epoch": 2.116881103704852,
+      "grad_norm": 0.36132997274398804,
+      "learning_rate": 2.4138008035737858e-06,
+      "loss": 0.3475,
+      "step": 4552
+    },
+    {
+      "epoch": 2.117346147884049,
+      "grad_norm": 0.36312153935432434,
+      "learning_rate": 2.411485332109806e-06,
+      "loss": 0.3111,
+      "step": 4553
+    },
+    {
+      "epoch": 2.117811192063246,
+      "grad_norm": 0.38179275393486023,
+      "learning_rate": 2.4091706187797286e-06,
+      "loss": 0.321,
+      "step": 4554
+    },
+    {
+      "epoch": 2.118276236242443,
+      "grad_norm": 0.37554097175598145,
+      "learning_rate": 2.4068566642614923e-06,
+      "loss": 0.2791,
+      "step": 4555
+    },
+    {
+      "epoch": 2.11874128042164,
+      "grad_norm": 0.3833089768886566,
+      "learning_rate": 2.4045434692328172e-06,
+      "loss": 0.3357,
+      "step": 4556
+    },
+    {
+      "epoch": 2.119206324600837,
+      "grad_norm": 0.35031166672706604,
+      "learning_rate": 2.4022310343712022e-06,
+      "loss": 0.2867,
+      "step": 4557
+    },
+    {
+      "epoch": 2.119671368780034,
+      "grad_norm": 0.37646663188934326,
+      "learning_rate": 2.3999193603539234e-06,
+      "loss": 0.3037,
+      "step": 4558
+    },
+    {
+      "epoch": 2.1201364129592313,
+      "grad_norm": 0.34302130341529846,
+      "learning_rate": 2.3976084478580282e-06,
+      "loss": 0.3407,
+      "step": 4559
+    },
+    {
+      "epoch": 2.120601457138428,
+      "grad_norm": 0.336687296628952,
+      "learning_rate": 2.3952982975603494e-06,
+      "loss": 0.2947,
+      "step": 4560
+    },
+    {
+      "epoch": 2.121066501317625,
+      "grad_norm": 0.3723877966403961,
+      "learning_rate": 2.3929889101374887e-06,
+      "loss": 0.3331,
+      "step": 4561
+    },
+    {
+      "epoch": 2.121531545496822,
+      "grad_norm": 0.369495153427124,
+      "learning_rate": 2.3906802862658325e-06,
+      "loss": 0.3428,
+      "step": 4562
+    },
+    {
+      "epoch": 2.1219965896760193,
+      "grad_norm": 0.3618394136428833,
+      "learning_rate": 2.388372426621534e-06,
+      "loss": 0.3158,
+      "step": 4563
+    },
+    {
+      "epoch": 2.1224616338552162,
+      "grad_norm": 0.34468409419059753,
+      "learning_rate": 2.386065331880534e-06,
+      "loss": 0.3068,
+      "step": 4564
+    },
+    {
+      "epoch": 2.122926678034413,
+      "grad_norm": 0.3602186441421509,
+      "learning_rate": 2.3837590027185364e-06,
+      "loss": 0.326,
+      "step": 4565
+    },
+    {
+      "epoch": 2.1233917222136105,
+      "grad_norm": 0.339193195104599,
+      "learning_rate": 2.381453439811034e-06,
+      "loss": 0.3237,
+      "step": 4566
+    },
+    {
+      "epoch": 2.1238567663928074,
+      "grad_norm": 0.3470028340816498,
+      "learning_rate": 2.379148643833283e-06,
+      "loss": 0.3042,
+      "step": 4567
+    },
+    {
+      "epoch": 2.1243218105720043,
+      "grad_norm": 0.35298952460289,
+      "learning_rate": 2.3768446154603234e-06,
+      "loss": 0.3114,
+      "step": 4568
+    },
+    {
+      "epoch": 2.124786854751201,
+      "grad_norm": 0.39458978176116943,
+      "learning_rate": 2.3745413553669678e-06,
+      "loss": 0.331,
+      "step": 4569
+    },
+    {
+      "epoch": 2.1252518989303986,
+      "grad_norm": 0.36957892775535583,
+      "learning_rate": 2.3722388642278047e-06,
+      "loss": 0.2894,
+      "step": 4570
+    },
+    {
+      "epoch": 2.1257169431095955,
+      "grad_norm": 0.38413190841674805,
+      "learning_rate": 2.369937142717194e-06,
+      "loss": 0.3403,
+      "step": 4571
+    },
+    {
+      "epoch": 2.1261819872887924,
+      "grad_norm": 0.32337597012519836,
+      "learning_rate": 2.3676361915092757e-06,
+      "loss": 0.2965,
+      "step": 4572
+    },
+    {
+      "epoch": 2.1266470314679893,
+      "grad_norm": 0.3725772500038147,
+      "learning_rate": 2.3653360112779567e-06,
+      "loss": 0.3541,
+      "step": 4573
+    },
+    {
+      "epoch": 2.1271120756471866,
+      "grad_norm": 0.3611546456813812,
+      "learning_rate": 2.363036602696927e-06,
+      "loss": 0.2821,
+      "step": 4574
+    },
+    {
+      "epoch": 2.1275771198263835,
+      "grad_norm": 0.362352192401886,
+      "learning_rate": 2.3607379664396414e-06,
+      "loss": 0.3477,
+      "step": 4575
+    },
+    {
+      "epoch": 2.1280421640055804,
+      "grad_norm": 0.3450590968132019,
+      "learning_rate": 2.3584401031793377e-06,
+      "loss": 0.2978,
+      "step": 4576
+    },
+    {
+      "epoch": 2.128507208184778,
+      "grad_norm": 0.4075535535812378,
+      "learning_rate": 2.35614301358902e-06,
+      "loss": 0.358,
+      "step": 4577
+    },
+    {
+      "epoch": 2.1289722523639747,
+      "grad_norm": 0.3503515422344208,
+      "learning_rate": 2.353846698341468e-06,
+      "loss": 0.3289,
+      "step": 4578
+    },
+    {
+      "epoch": 2.1294372965431716,
+      "grad_norm": 0.38733020424842834,
+      "learning_rate": 2.351551158109235e-06,
+      "loss": 0.2964,
+      "step": 4579
+    },
+    {
+      "epoch": 2.1299023407223685,
+      "grad_norm": 0.35857686400413513,
+      "learning_rate": 2.3492563935646493e-06,
+      "loss": 0.3276,
+      "step": 4580
+    },
+    {
+      "epoch": 2.130367384901566,
+      "grad_norm": 0.34984397888183594,
+      "learning_rate": 2.3469624053798117e-06,
+      "loss": 0.3162,
+      "step": 4581
+    },
+    {
+      "epoch": 2.1308324290807628,
+      "grad_norm": 0.34331047534942627,
+      "learning_rate": 2.34466919422659e-06,
+      "loss": 0.309,
+      "step": 4582
+    },
+    {
+      "epoch": 2.1312974732599597,
+      "grad_norm": 0.32851484417915344,
+      "learning_rate": 2.3423767607766316e-06,
+      "loss": 0.3186,
+      "step": 4583
+    },
+    {
+      "epoch": 2.1317625174391566,
+      "grad_norm": 0.360724538564682,
+      "learning_rate": 2.34008510570135e-06,
+      "loss": 0.2974,
+      "step": 4584
+    },
+    {
+      "epoch": 2.132227561618354,
+      "grad_norm": 0.3493599593639374,
+      "learning_rate": 2.337794229671938e-06,
+      "loss": 0.3563,
+      "step": 4585
+    },
+    {
+      "epoch": 2.132692605797551,
+      "grad_norm": 0.3586118519306183,
+      "learning_rate": 2.3355041333593507e-06,
+      "loss": 0.3515,
+      "step": 4586
+    },
+    {
+      "epoch": 2.1331576499767477,
+      "grad_norm": 0.3299708664417267,
+      "learning_rate": 2.3332148174343257e-06,
+      "loss": 0.3117,
+      "step": 4587
+    },
+    {
+      "epoch": 2.1336226941559446,
+      "grad_norm": 0.31566065549850464,
+      "learning_rate": 2.3309262825673616e-06,
+      "loss": 0.3165,
+      "step": 4588
+    },
+    {
+      "epoch": 2.134087738335142,
+      "grad_norm": 0.35169747471809387,
+      "learning_rate": 2.3286385294287367e-06,
+      "loss": 0.3247,
+      "step": 4589
+    },
+    {
+      "epoch": 2.134552782514339,
+      "grad_norm": 0.3681233525276184,
+      "learning_rate": 2.3263515586884935e-06,
+      "loss": 0.3373,
+      "step": 4590
+    },
+    {
+      "epoch": 2.135017826693536,
+      "grad_norm": 0.3795618414878845,
+      "learning_rate": 2.32406537101645e-06,
+      "loss": 0.3029,
+      "step": 4591
+    },
+    {
+      "epoch": 2.1354828708727327,
+      "grad_norm": 0.3609673082828522,
+      "learning_rate": 2.3217799670821938e-06,
+      "loss": 0.3362,
+      "step": 4592
+    },
+    {
+      "epoch": 2.13594791505193,
+      "grad_norm": 0.3580998480319977,
+      "learning_rate": 2.3194953475550846e-06,
+      "loss": 0.3337,
+      "step": 4593
+    },
+    {
+      "epoch": 2.136412959231127,
+      "grad_norm": 0.33573347330093384,
+      "learning_rate": 2.3172115131042466e-06,
+      "loss": 0.2821,
+      "step": 4594
+    },
+    {
+      "epoch": 2.136878003410324,
+      "grad_norm": 0.3740091919898987,
+      "learning_rate": 2.314928464398581e-06,
+      "loss": 0.3662,
+      "step": 4595
+    },
+    {
+      "epoch": 2.137343047589521,
+      "grad_norm": 0.385172963142395,
+      "learning_rate": 2.3126462021067518e-06,
+      "loss": 0.3709,
+      "step": 4596
+    },
+    {
+      "epoch": 2.137808091768718,
+      "grad_norm": 0.3355659246444702,
+      "learning_rate": 2.310364726897202e-06,
+      "loss": 0.3105,
+      "step": 4597
+    },
+    {
+      "epoch": 2.138273135947915,
+      "grad_norm": 0.36844390630722046,
+      "learning_rate": 2.3080840394381327e-06,
+      "loss": 0.3155,
+      "step": 4598
+    },
+    {
+      "epoch": 2.138738180127112,
+      "grad_norm": 0.3672247529029846,
+      "learning_rate": 2.305804140397525e-06,
+      "loss": 0.3089,
+      "step": 4599
+    },
+    {
+      "epoch": 2.1392032243063093,
+      "grad_norm": 0.364468514919281,
+      "learning_rate": 2.3035250304431206e-06,
+      "loss": 0.3194,
+      "step": 4600
+    },
+    {
+      "epoch": 2.139668268485506,
+      "grad_norm": 0.37167108058929443,
+      "learning_rate": 2.3012467102424373e-06,
+      "loss": 0.3588,
+      "step": 4601
+    },
+    {
+      "epoch": 2.140133312664703,
+      "grad_norm": 0.3726482689380646,
+      "learning_rate": 2.2989691804627544e-06,
+      "loss": 0.3193,
+      "step": 4602
+    },
+    {
+      "epoch": 2.1405983568439,
+      "grad_norm": 0.3460210859775543,
+      "learning_rate": 2.296692441771125e-06,
+      "loss": 0.3113,
+      "step": 4603
+    },
+    {
+      "epoch": 2.1410634010230973,
+      "grad_norm": 0.3563862442970276,
+      "learning_rate": 2.29441649483437e-06,
+      "loss": 0.3405,
+      "step": 4604
+    },
+    {
+      "epoch": 2.1415284452022942,
+      "grad_norm": 0.3387638032436371,
+      "learning_rate": 2.2921413403190774e-06,
+      "loss": 0.3125,
+      "step": 4605
+    },
+    {
+      "epoch": 2.141993489381491,
+      "grad_norm": 0.38365113735198975,
+      "learning_rate": 2.2898669788916006e-06,
+      "loss": 0.3201,
+      "step": 4606
+    },
+    {
+      "epoch": 2.1424585335606885,
+      "grad_norm": 0.369579553604126,
+      "learning_rate": 2.2875934112180664e-06,
+      "loss": 0.3501,
+      "step": 4607
+    },
+    {
+      "epoch": 2.1429235777398854,
+      "grad_norm": 0.36825814843177795,
+      "learning_rate": 2.285320637964362e-06,
+      "loss": 0.3147,
+      "step": 4608
+    },
+    {
+      "epoch": 2.1433886219190823,
+      "grad_norm": 0.3360195457935333,
+      "learning_rate": 2.2830486597961504e-06,
+      "loss": 0.2841,
+      "step": 4609
+    },
+    {
+      "epoch": 2.143853666098279,
+      "grad_norm": 0.3493092358112335,
+      "learning_rate": 2.2807774773788518e-06,
+      "loss": 0.3183,
+      "step": 4610
+    },
+    {
+      "epoch": 2.1443187102774766,
+      "grad_norm": 0.3550946116447449,
+      "learning_rate": 2.2785070913776635e-06,
+      "loss": 0.3226,
+      "step": 4611
+    },
+    {
+      "epoch": 2.1447837544566735,
+      "grad_norm": 0.3619578778743744,
+      "learning_rate": 2.2762375024575424e-06,
+      "loss": 0.3189,
+      "step": 4612
+    },
+    {
+      "epoch": 2.1452487986358704,
+      "grad_norm": 0.388798326253891,
+      "learning_rate": 2.2739687112832125e-06,
+      "loss": 0.3142,
+      "step": 4613
+    },
+    {
+      "epoch": 2.1457138428150673,
+      "grad_norm": 0.3689102530479431,
+      "learning_rate": 2.2717007185191673e-06,
+      "loss": 0.3336,
+      "step": 4614
+    },
+    {
+      "epoch": 2.1461788869942646,
+      "grad_norm": 0.3550708591938019,
+      "learning_rate": 2.269433524829666e-06,
+      "loss": 0.3183,
+      "step": 4615
+    },
+    {
+      "epoch": 2.1466439311734615,
+      "grad_norm": 0.3319764733314514,
+      "learning_rate": 2.267167130878734e-06,
+      "loss": 0.3024,
+      "step": 4616
+    },
+    {
+      "epoch": 2.1471089753526584,
+      "grad_norm": 0.35746893286705017,
+      "learning_rate": 2.2649015373301574e-06,
+      "loss": 0.2984,
+      "step": 4617
+    },
+    {
+      "epoch": 2.1475740195318553,
+      "grad_norm": 0.3471658527851105,
+      "learning_rate": 2.2626367448474963e-06,
+      "loss": 0.3148,
+      "step": 4618
+    },
+    {
+      "epoch": 2.1480390637110527,
+      "grad_norm": 0.3576483130455017,
+      "learning_rate": 2.2603727540940673e-06,
+      "loss": 0.3567,
+      "step": 4619
+    },
+    {
+      "epoch": 2.1485041078902496,
+      "grad_norm": 0.31263265013694763,
+      "learning_rate": 2.25810956573296e-06,
+      "loss": 0.294,
+      "step": 4620
+    },
+    {
+      "epoch": 2.1489691520694465,
+      "grad_norm": 0.3338179588317871,
+      "learning_rate": 2.255847180427022e-06,
+      "loss": 0.319,
+      "step": 4621
+    },
+    {
+      "epoch": 2.1494341962486434,
+      "grad_norm": 0.33670803904533386,
+      "learning_rate": 2.2535855988388734e-06,
+      "loss": 0.3338,
+      "step": 4622
+    },
+    {
+      "epoch": 2.1498992404278408,
+      "grad_norm": 0.3188488185405731,
+      "learning_rate": 2.2513248216308897e-06,
+      "loss": 0.2784,
+      "step": 4623
+    },
+    {
+      "epoch": 2.1503642846070377,
+      "grad_norm": 0.40295830368995667,
+      "learning_rate": 2.249064849465221e-06,
+      "loss": 0.3436,
+      "step": 4624
+    },
+    {
+      "epoch": 2.1508293287862346,
+      "grad_norm": 0.35394805669784546,
+      "learning_rate": 2.2468056830037725e-06,
+      "loss": 0.3132,
+      "step": 4625
+    },
+    {
+      "epoch": 2.151294372965432,
+      "grad_norm": 0.3387957215309143,
+      "learning_rate": 2.2445473229082186e-06,
+      "loss": 0.3198,
+      "step": 4626
+    },
+    {
+      "epoch": 2.151759417144629,
+      "grad_norm": 0.34971579909324646,
+      "learning_rate": 2.2422897698399964e-06,
+      "loss": 0.2797,
+      "step": 4627
+    },
+    {
+      "epoch": 2.1522244613238257,
+      "grad_norm": 0.4021191895008087,
+      "learning_rate": 2.240033024460309e-06,
+      "loss": 0.3572,
+      "step": 4628
+    },
+    {
+      "epoch": 2.1526895055030226,
+      "grad_norm": 0.34752318263053894,
+      "learning_rate": 2.2377770874301157e-06,
+      "loss": 0.3026,
+      "step": 4629
+    },
+    {
+      "epoch": 2.15315454968222,
+      "grad_norm": 0.338008850812912,
+      "learning_rate": 2.2355219594101483e-06,
+      "loss": 0.3024,
+      "step": 4630
+    },
+    {
+      "epoch": 2.153619593861417,
+      "grad_norm": 0.331898033618927,
+      "learning_rate": 2.2332676410608937e-06,
+      "loss": 0.3002,
+      "step": 4631
+    },
+    {
+      "epoch": 2.154084638040614,
+      "grad_norm": 0.36906805634498596,
+      "learning_rate": 2.231014133042608e-06,
+      "loss": 0.3276,
+      "step": 4632
+    },
+    {
+      "epoch": 2.1545496822198107,
+      "grad_norm": 0.35603275895118713,
+      "learning_rate": 2.2287614360153042e-06,
+      "loss": 0.3409,
+      "step": 4633
+    },
+    {
+      "epoch": 2.155014726399008,
+      "grad_norm": 0.3487357199192047,
+      "learning_rate": 2.226509550638764e-06,
+      "loss": 0.326,
+      "step": 4634
+    },
+    {
+      "epoch": 2.155479770578205,
+      "grad_norm": 0.3237025737762451,
+      "learning_rate": 2.224258477572524e-06,
+      "loss": 0.2981,
+      "step": 4635
+    },
+    {
+      "epoch": 2.155944814757402,
+      "grad_norm": 0.35809311270713806,
+      "learning_rate": 2.222008217475891e-06,
+      "loss": 0.307,
+      "step": 4636
+    },
+    {
+      "epoch": 2.156409858936599,
+      "grad_norm": 0.3373444080352783,
+      "learning_rate": 2.219758771007926e-06,
+      "loss": 0.3359,
+      "step": 4637
+    },
+    {
+      "epoch": 2.156874903115796,
+      "grad_norm": 0.36277028918266296,
+      "learning_rate": 2.217510138827457e-06,
+      "loss": 0.3766,
+      "step": 4638
+    },
+    {
+      "epoch": 2.157339947294993,
+      "grad_norm": 0.32372814416885376,
+      "learning_rate": 2.215262321593072e-06,
+      "loss": 0.2788,
+      "step": 4639
+    },
+    {
+      "epoch": 2.15780499147419,
+      "grad_norm": 0.37135806679725647,
+      "learning_rate": 2.2130153199631214e-06,
+      "loss": 0.3146,
+      "step": 4640
+    },
+    {
+      "epoch": 2.1582700356533873,
+      "grad_norm": 0.3312724828720093,
+      "learning_rate": 2.2107691345957133e-06,
+      "loss": 0.3261,
+      "step": 4641
+    },
+    {
+      "epoch": 2.158735079832584,
+      "grad_norm": 0.3527110517024994,
+      "learning_rate": 2.208523766148721e-06,
+      "loss": 0.3189,
+      "step": 4642
+    },
+    {
+      "epoch": 2.159200124011781,
+      "grad_norm": 0.3383033275604248,
+      "learning_rate": 2.2062792152797733e-06,
+      "loss": 0.3293,
+      "step": 4643
+    },
+    {
+      "epoch": 2.159665168190978,
+      "grad_norm": 0.4051493704319,
+      "learning_rate": 2.204035482646267e-06,
+      "loss": 0.3627,
+      "step": 4644
+    },
+    {
+      "epoch": 2.1601302123701753,
+      "grad_norm": 0.3591381311416626,
+      "learning_rate": 2.20179256890535e-06,
+      "loss": 0.2986,
+      "step": 4645
+    },
+    {
+      "epoch": 2.1605952565493722,
+      "grad_norm": 0.38014209270477295,
+      "learning_rate": 2.1995504747139397e-06,
+      "loss": 0.306,
+      "step": 4646
+    },
+    {
+      "epoch": 2.161060300728569,
+      "grad_norm": 0.33583784103393555,
+      "learning_rate": 2.1973092007287054e-06,
+      "loss": 0.3432,
+      "step": 4647
+    },
+    {
+      "epoch": 2.161525344907766,
+      "grad_norm": 0.3436342775821686,
+      "learning_rate": 2.195068747606084e-06,
+      "loss": 0.3368,
+      "step": 4648
+    },
+    {
+      "epoch": 2.1619903890869634,
+      "grad_norm": 0.33585768938064575,
+      "learning_rate": 2.1928291160022634e-06,
+      "loss": 0.3136,
+      "step": 4649
+    },
+    {
+      "epoch": 2.1624554332661603,
+      "grad_norm": 0.35160210728645325,
+      "learning_rate": 2.190590306573198e-06,
+      "loss": 0.3302,
+      "step": 4650
+    },
+    {
+      "epoch": 2.162920477445357,
+      "grad_norm": 0.35300135612487793,
+      "learning_rate": 2.1883523199745987e-06,
+      "loss": 0.2987,
+      "step": 4651
+    },
+    {
+      "epoch": 2.163385521624554,
+      "grad_norm": 0.4004720151424408,
+      "learning_rate": 2.1861151568619336e-06,
+      "loss": 0.3809,
+      "step": 4652
+    },
+    {
+      "epoch": 2.1638505658037515,
+      "grad_norm": 0.3017584979534149,
+      "learning_rate": 2.1838788178904346e-06,
+      "loss": 0.2655,
+      "step": 4653
+    },
+    {
+      "epoch": 2.1643156099829484,
+      "grad_norm": 0.3657880127429962,
+      "learning_rate": 2.1816433037150856e-06,
+      "loss": 0.3563,
+      "step": 4654
+    },
+    {
+      "epoch": 2.1647806541621453,
+      "grad_norm": 0.362338125705719,
+      "learning_rate": 2.179408614990635e-06,
+      "loss": 0.3145,
+      "step": 4655
+    },
+    {
+      "epoch": 2.1652456983413426,
+      "grad_norm": 0.3580748736858368,
+      "learning_rate": 2.177174752371585e-06,
+      "loss": 0.3113,
+      "step": 4656
+    },
+    {
+      "epoch": 2.1657107425205395,
+      "grad_norm": 0.37395989894866943,
+      "learning_rate": 2.1749417165121994e-06,
+      "loss": 0.3297,
+      "step": 4657
+    },
+    {
+      "epoch": 2.1661757866997364,
+      "grad_norm": 0.35120391845703125,
+      "learning_rate": 2.1727095080664956e-06,
+      "loss": 0.3417,
+      "step": 4658
+    },
+    {
+      "epoch": 2.1666408308789333,
+      "grad_norm": 0.3511483371257782,
+      "learning_rate": 2.1704781276882547e-06,
+      "loss": 0.3232,
+      "step": 4659
+    },
+    {
+      "epoch": 2.1671058750581307,
+      "grad_norm": 0.3221030533313751,
+      "learning_rate": 2.168247576031008e-06,
+      "loss": 0.3062,
+      "step": 4660
+    },
+    {
+      "epoch": 2.1675709192373276,
+      "grad_norm": 0.3379102349281311,
+      "learning_rate": 2.16601785374805e-06,
+      "loss": 0.3103,
+      "step": 4661
+    },
+    {
+      "epoch": 2.1680359634165245,
+      "grad_norm": 0.35525548458099365,
+      "learning_rate": 2.163788961492429e-06,
+      "loss": 0.3502,
+      "step": 4662
+    },
+    {
+      "epoch": 2.1685010075957214,
+      "grad_norm": 0.33914363384246826,
+      "learning_rate": 2.161560899916954e-06,
+      "loss": 0.3336,
+      "step": 4663
+    },
+    {
+      "epoch": 2.1689660517749187,
+      "grad_norm": 0.36163097620010376,
+      "learning_rate": 2.159333669674185e-06,
+      "loss": 0.3073,
+      "step": 4664
+    },
+    {
+      "epoch": 2.1694310959541157,
+      "grad_norm": 0.34654346108436584,
+      "learning_rate": 2.1571072714164445e-06,
+      "loss": 0.3466,
+      "step": 4665
+    },
+    {
+      "epoch": 2.1698961401333126,
+      "grad_norm": 0.33517155051231384,
+      "learning_rate": 2.1548817057958043e-06,
+      "loss": 0.3351,
+      "step": 4666
+    },
+    {
+      "epoch": 2.17036118431251,
+      "grad_norm": 0.3178362250328064,
+      "learning_rate": 2.152656973464101e-06,
+      "loss": 0.2925,
+      "step": 4667
+    },
+    {
+      "epoch": 2.170826228491707,
+      "grad_norm": 0.4054083824157715,
+      "learning_rate": 2.1504330750729185e-06,
+      "loss": 0.3176,
+      "step": 4668
+    },
+    {
+      "epoch": 2.1712912726709037,
+      "grad_norm": 0.31070810556411743,
+      "learning_rate": 2.1482100112736044e-06,
+      "loss": 0.2867,
+      "step": 4669
+    },
+    {
+      "epoch": 2.1717563168501006,
+      "grad_norm": 0.3718927800655365,
+      "learning_rate": 2.1459877827172538e-06,
+      "loss": 0.3633,
+      "step": 4670
+    },
+    {
+      "epoch": 2.172221361029298,
+      "grad_norm": 0.36688584089279175,
+      "learning_rate": 2.1437663900547255e-06,
+      "loss": 0.3169,
+      "step": 4671
+    },
+    {
+      "epoch": 2.172686405208495,
+      "grad_norm": 0.35190173983573914,
+      "learning_rate": 2.141545833936625e-06,
+      "loss": 0.3077,
+      "step": 4672
+    },
+    {
+      "epoch": 2.173151449387692,
+      "grad_norm": 0.3758264482021332,
+      "learning_rate": 2.13932611501332e-06,
+      "loss": 0.36,
+      "step": 4673
+    },
+    {
+      "epoch": 2.1736164935668887,
+      "grad_norm": 0.33575019240379333,
+      "learning_rate": 2.1371072339349293e-06,
+      "loss": 0.3308,
+      "step": 4674
+    },
+    {
+      "epoch": 2.174081537746086,
+      "grad_norm": 0.33818864822387695,
+      "learning_rate": 2.1348891913513293e-06,
+      "loss": 0.3019,
+      "step": 4675
+    },
+    {
+      "epoch": 2.174546581925283,
+      "grad_norm": 0.3424922227859497,
+      "learning_rate": 2.132671987912145e-06,
+      "loss": 0.2985,
+      "step": 4676
+    },
+    {
+      "epoch": 2.17501162610448,
+      "grad_norm": 0.3274764120578766,
+      "learning_rate": 2.130455624266762e-06,
+      "loss": 0.3055,
+      "step": 4677
+    },
+    {
+      "epoch": 2.1754766702836768,
+      "grad_norm": 0.3366675078868866,
+      "learning_rate": 2.128240101064315e-06,
+      "loss": 0.3189,
+      "step": 4678
+    },
+    {
+      "epoch": 2.175941714462874,
+      "grad_norm": 0.38803109526634216,
+      "learning_rate": 2.126025418953698e-06,
+      "loss": 0.3544,
+      "step": 4679
+    },
+    {
+      "epoch": 2.176406758642071,
+      "grad_norm": 0.32740333676338196,
+      "learning_rate": 2.1238115785835512e-06,
+      "loss": 0.3153,
+      "step": 4680
+    },
+    {
+      "epoch": 2.176871802821268,
+      "grad_norm": 0.30742892622947693,
+      "learning_rate": 2.1215985806022765e-06,
+      "loss": 0.2642,
+      "step": 4681
+    },
+    {
+      "epoch": 2.177336847000465,
+      "grad_norm": 0.3487866520881653,
+      "learning_rate": 2.1193864256580215e-06,
+      "loss": 0.3637,
+      "step": 4682
+    },
+    {
+      "epoch": 2.177801891179662,
+      "grad_norm": 0.35300448536872864,
+      "learning_rate": 2.117175114398694e-06,
+      "loss": 0.3174,
+      "step": 4683
+    },
+    {
+      "epoch": 2.178266935358859,
+      "grad_norm": 0.3436736762523651,
+      "learning_rate": 2.1149646474719475e-06,
+      "loss": 0.3213,
+      "step": 4684
+    },
+    {
+      "epoch": 2.178731979538056,
+      "grad_norm": 0.32160818576812744,
+      "learning_rate": 2.112755025525193e-06,
+      "loss": 0.3061,
+      "step": 4685
+    },
+    {
+      "epoch": 2.1791970237172533,
+      "grad_norm": 0.3675907552242279,
+      "learning_rate": 2.110546249205597e-06,
+      "loss": 0.3207,
+      "step": 4686
+    },
+    {
+      "epoch": 2.1796620678964502,
+      "grad_norm": 0.3436359465122223,
+      "learning_rate": 2.1083383191600676e-06,
+      "loss": 0.3222,
+      "step": 4687
+    },
+    {
+      "epoch": 2.180127112075647,
+      "grad_norm": 0.32878395915031433,
+      "learning_rate": 2.106131236035277e-06,
+      "loss": 0.3153,
+      "step": 4688
+    },
+    {
+      "epoch": 2.180592156254844,
+      "grad_norm": 0.3562372624874115,
+      "learning_rate": 2.1039250004776397e-06,
+      "loss": 0.349,
+      "step": 4689
+    },
+    {
+      "epoch": 2.1810572004340414,
+      "grad_norm": 0.33452579379081726,
+      "learning_rate": 2.1017196131333306e-06,
+      "loss": 0.3016,
+      "step": 4690
+    },
+    {
+      "epoch": 2.1815222446132383,
+      "grad_norm": 0.37932857871055603,
+      "learning_rate": 2.099515074648267e-06,
+      "loss": 0.3182,
+      "step": 4691
+    },
+    {
+      "epoch": 2.181987288792435,
+      "grad_norm": 0.35700809955596924,
+      "learning_rate": 2.0973113856681277e-06,
+      "loss": 0.3217,
+      "step": 4692
+    },
+    {
+      "epoch": 2.182452332971632,
+      "grad_norm": 0.37079155445098877,
+      "learning_rate": 2.0951085468383326e-06,
+      "loss": 0.3399,
+      "step": 4693
+    },
+    {
+      "epoch": 2.1829173771508295,
+      "grad_norm": 0.3523573577404022,
+      "learning_rate": 2.0929065588040615e-06,
+      "loss": 0.3152,
+      "step": 4694
+    },
+    {
+      "epoch": 2.1833824213300264,
+      "grad_norm": 0.33730053901672363,
+      "learning_rate": 2.0907054222102367e-06,
+      "loss": 0.3066,
+      "step": 4695
+    },
+    {
+      "epoch": 2.1838474655092233,
+      "grad_norm": 0.34836792945861816,
+      "learning_rate": 2.088505137701538e-06,
+      "loss": 0.3236,
+      "step": 4696
+    },
+    {
+      "epoch": 2.1843125096884206,
+      "grad_norm": 0.3443335294723511,
+      "learning_rate": 2.0863057059223923e-06,
+      "loss": 0.2979,
+      "step": 4697
+    },
+    {
+      "epoch": 2.1847775538676175,
+      "grad_norm": 0.32043859362602234,
+      "learning_rate": 2.08410712751698e-06,
+      "loss": 0.3132,
+      "step": 4698
+    },
+    {
+      "epoch": 2.1852425980468144,
+      "grad_norm": 0.3443070948123932,
+      "learning_rate": 2.081909403129225e-06,
+      "loss": 0.297,
+      "step": 4699
+    },
+    {
+      "epoch": 2.1857076422260113,
+      "grad_norm": 0.3612792193889618,
+      "learning_rate": 2.079712533402808e-06,
+      "loss": 0.3369,
+      "step": 4700
+    },
+    {
+      "epoch": 2.1861726864052087,
+      "grad_norm": 0.3530242443084717,
+      "learning_rate": 2.0775165189811534e-06,
+      "loss": 0.2989,
+      "step": 4701
+    },
+    {
+      "epoch": 2.1866377305844056,
+      "grad_norm": 0.33956217765808105,
+      "learning_rate": 2.0753213605074424e-06,
+      "loss": 0.3195,
+      "step": 4702
+    },
+    {
+      "epoch": 2.1871027747636025,
+      "grad_norm": 0.3230421245098114,
+      "learning_rate": 2.0731270586245972e-06,
+      "loss": 0.3039,
+      "step": 4703
+    },
+    {
+      "epoch": 2.1875678189427994,
+      "grad_norm": 0.41725921630859375,
+      "learning_rate": 2.070933613975296e-06,
+      "loss": 0.3244,
+      "step": 4704
+    },
+    {
+      "epoch": 2.1880328631219967,
+      "grad_norm": 0.33200743794441223,
+      "learning_rate": 2.068741027201961e-06,
+      "loss": 0.3153,
+      "step": 4705
+    },
+    {
+      "epoch": 2.1884979073011936,
+      "grad_norm": 0.3534869849681854,
+      "learning_rate": 2.066549298946767e-06,
+      "loss": 0.344,
+      "step": 4706
+    },
+    {
+      "epoch": 2.1889629514803906,
+      "grad_norm": 0.308760404586792,
+      "learning_rate": 2.064358429851634e-06,
+      "loss": 0.2962,
+      "step": 4707
+    },
+    {
+      "epoch": 2.1894279956595875,
+      "grad_norm": 0.36021149158477783,
+      "learning_rate": 2.062168420558232e-06,
+      "loss": 0.3497,
+      "step": 4708
+    },
+    {
+      "epoch": 2.189893039838785,
+      "grad_norm": 0.3680166006088257,
+      "learning_rate": 2.0599792717079807e-06,
+      "loss": 0.3084,
+      "step": 4709
+    },
+    {
+      "epoch": 2.1903580840179817,
+      "grad_norm": 0.3480318486690521,
+      "learning_rate": 2.0577909839420468e-06,
+      "loss": 0.3213,
+      "step": 4710
+    },
+    {
+      "epoch": 2.1908231281971786,
+      "grad_norm": 0.33287766575813293,
+      "learning_rate": 2.0556035579013417e-06,
+      "loss": 0.2899,
+      "step": 4711
+    },
+    {
+      "epoch": 2.1912881723763755,
+      "grad_norm": 0.36140042543411255,
+      "learning_rate": 2.0534169942265298e-06,
+      "loss": 0.3306,
+      "step": 4712
+    },
+    {
+      "epoch": 2.191753216555573,
+      "grad_norm": 0.3293544352054596,
+      "learning_rate": 2.0512312935580163e-06,
+      "loss": 0.2999,
+      "step": 4713
+    },
+    {
+      "epoch": 2.1922182607347698,
+      "grad_norm": 0.3596755266189575,
+      "learning_rate": 2.0490464565359615e-06,
+      "loss": 0.3341,
+      "step": 4714
+    },
+    {
+      "epoch": 2.1926833049139667,
+      "grad_norm": 0.3495473563671112,
+      "learning_rate": 2.0468624838002647e-06,
+      "loss": 0.3237,
+      "step": 4715
+    },
+    {
+      "epoch": 2.193148349093164,
+      "grad_norm": 0.3244265615940094,
+      "learning_rate": 2.044679375990581e-06,
+      "loss": 0.3232,
+      "step": 4716
+    },
+    {
+      "epoch": 2.193613393272361,
+      "grad_norm": 0.34927719831466675,
+      "learning_rate": 2.0424971337463017e-06,
+      "loss": 0.3305,
+      "step": 4717
+    },
+    {
+      "epoch": 2.194078437451558,
+      "grad_norm": 0.3691467046737671,
+      "learning_rate": 2.0403157577065746e-06,
+      "loss": 0.3485,
+      "step": 4718
+    },
+    {
+      "epoch": 2.1945434816307547,
+      "grad_norm": 0.3577490448951721,
+      "learning_rate": 2.0381352485102857e-06,
+      "loss": 0.3291,
+      "step": 4719
+    },
+    {
+      "epoch": 2.195008525809952,
+      "grad_norm": 0.344149649143219,
+      "learning_rate": 2.0359556067960727e-06,
+      "loss": 0.3146,
+      "step": 4720
+    },
+    {
+      "epoch": 2.195473569989149,
+      "grad_norm": 0.33729931712150574,
+      "learning_rate": 2.0337768332023185e-06,
+      "loss": 0.2903,
+      "step": 4721
+    },
+    {
+      "epoch": 2.195938614168346,
+      "grad_norm": 0.36155104637145996,
+      "learning_rate": 2.0315989283671474e-06,
+      "loss": 0.3305,
+      "step": 4722
+    },
+    {
+      "epoch": 2.196403658347543,
+      "grad_norm": 0.3909613788127899,
+      "learning_rate": 2.029421892928436e-06,
+      "loss": 0.3343,
+      "step": 4723
+    },
+    {
+      "epoch": 2.19686870252674,
+      "grad_norm": 0.3571639358997345,
+      "learning_rate": 2.027245727523798e-06,
+      "loss": 0.2831,
+      "step": 4724
+    },
+    {
+      "epoch": 2.197333746705937,
+      "grad_norm": 0.3750656247138977,
+      "learning_rate": 2.0250704327906025e-06,
+      "loss": 0.3303,
+      "step": 4725
+    },
+    {
+      "epoch": 2.197798790885134,
+      "grad_norm": 0.35987409949302673,
+      "learning_rate": 2.022896009365952e-06,
+      "loss": 0.3574,
+      "step": 4726
+    },
+    {
+      "epoch": 2.1982638350643313,
+      "grad_norm": 0.3430827558040619,
+      "learning_rate": 2.020722457886705e-06,
+      "loss": 0.3216,
+      "step": 4727
+    },
+    {
+      "epoch": 2.1987288792435282,
+      "grad_norm": 0.3294491171836853,
+      "learning_rate": 2.018549778989456e-06,
+      "loss": 0.3249,
+      "step": 4728
+    },
+    {
+      "epoch": 2.199193923422725,
+      "grad_norm": 0.39135822653770447,
+      "learning_rate": 2.0163779733105497e-06,
+      "loss": 0.3184,
+      "step": 4729
+    },
+    {
+      "epoch": 2.199658967601922,
+      "grad_norm": 0.3750271201133728,
+      "learning_rate": 2.0142070414860704e-06,
+      "loss": 0.3288,
+      "step": 4730
+    },
+    {
+      "epoch": 2.2001240117811194,
+      "grad_norm": 0.35246542096138,
+      "learning_rate": 2.0120369841518496e-06,
+      "loss": 0.3001,
+      "step": 4731
+    },
+    {
+      "epoch": 2.2005890559603163,
+      "grad_norm": 0.3286935091018677,
+      "learning_rate": 2.009867801943462e-06,
+      "loss": 0.331,
+      "step": 4732
+    },
+    {
+      "epoch": 2.201054100139513,
+      "grad_norm": 0.37689733505249023,
+      "learning_rate": 2.007699495496228e-06,
+      "loss": 0.3225,
+      "step": 4733
+    },
+    {
+      "epoch": 2.20151914431871,
+      "grad_norm": 0.358782559633255,
+      "learning_rate": 2.0055320654452055e-06,
+      "loss": 0.3297,
+      "step": 4734
+    },
+    {
+      "epoch": 2.2019841884979074,
+      "grad_norm": 0.33447474241256714,
+      "learning_rate": 2.0033655124252033e-06,
+      "loss": 0.3196,
+      "step": 4735
+    },
+    {
+      "epoch": 2.2024492326771044,
+      "grad_norm": 0.36790183186531067,
+      "learning_rate": 2.001199837070766e-06,
+      "loss": 0.2924,
+      "step": 4736
+    },
+    {
+      "epoch": 2.2029142768563013,
+      "grad_norm": 0.384657084941864,
+      "learning_rate": 1.999035040016188e-06,
+      "loss": 0.355,
+      "step": 4737
+    },
+    {
+      "epoch": 2.2033793210354986,
+      "grad_norm": 0.32145190238952637,
+      "learning_rate": 1.9968711218954994e-06,
+      "loss": 0.3055,
+      "step": 4738
+    },
+    {
+      "epoch": 2.2038443652146955,
+      "grad_norm": 0.34638088941574097,
+      "learning_rate": 1.9947080833424816e-06,
+      "loss": 0.3214,
+      "step": 4739
+    },
+    {
+      "epoch": 2.2043094093938924,
+      "grad_norm": 0.33720049262046814,
+      "learning_rate": 1.9925459249906488e-06,
+      "loss": 0.3181,
+      "step": 4740
+    },
+    {
+      "epoch": 2.2047744535730893,
+      "grad_norm": 0.3274615705013275,
+      "learning_rate": 1.990384647473265e-06,
+      "loss": 0.3015,
+      "step": 4741
+    },
+    {
+      "epoch": 2.2052394977522867,
+      "grad_norm": 0.3554820716381073,
+      "learning_rate": 1.9882242514233313e-06,
+      "loss": 0.3141,
+      "step": 4742
+    },
+    {
+      "epoch": 2.2057045419314836,
+      "grad_norm": 0.35325977206230164,
+      "learning_rate": 1.9860647374735937e-06,
+      "loss": 0.309,
+      "step": 4743
+    },
+    {
+      "epoch": 2.2061695861106805,
+      "grad_norm": 0.3346039652824402,
+      "learning_rate": 1.9839061062565384e-06,
+      "loss": 0.2945,
+      "step": 4744
+    },
+    {
+      "epoch": 2.2066346302898774,
+      "grad_norm": 0.3343028426170349,
+      "learning_rate": 1.9817483584043954e-06,
+      "loss": 0.3328,
+      "step": 4745
+    },
+    {
+      "epoch": 2.2070996744690747,
+      "grad_norm": 0.33999499678611755,
+      "learning_rate": 1.9795914945491305e-06,
+      "loss": 0.323,
+      "step": 4746
+    },
+    {
+      "epoch": 2.2075647186482716,
+      "grad_norm": 0.32321181893348694,
+      "learning_rate": 1.977435515322458e-06,
+      "loss": 0.2882,
+      "step": 4747
+    },
+    {
+      "epoch": 2.2080297628274685,
+      "grad_norm": 0.38171252608299255,
+      "learning_rate": 1.9752804213558254e-06,
+      "loss": 0.3073,
+      "step": 4748
+    },
+    {
+      "epoch": 2.2084948070066655,
+      "grad_norm": 0.37734076380729675,
+      "learning_rate": 1.9731262132804275e-06,
+      "loss": 0.3522,
+      "step": 4749
+    },
+    {
+      "epoch": 2.208959851185863,
+      "grad_norm": 0.34129461646080017,
+      "learning_rate": 1.970972891727194e-06,
+      "loss": 0.3072,
+      "step": 4750
+    },
+    {
+      "epoch": 2.2094248953650597,
+      "grad_norm": 0.36008572578430176,
+      "learning_rate": 1.9688204573268015e-06,
+      "loss": 0.3233,
+      "step": 4751
+    },
+    {
+      "epoch": 2.2098899395442566,
+      "grad_norm": 0.3995450735092163,
+      "learning_rate": 1.9666689107096597e-06,
+      "loss": 0.3445,
+      "step": 4752
+    },
+    {
+      "epoch": 2.2103549837234535,
+      "grad_norm": 0.3434513211250305,
+      "learning_rate": 1.964518252505925e-06,
+      "loss": 0.3023,
+      "step": 4753
+    },
+    {
+      "epoch": 2.210820027902651,
+      "grad_norm": 0.31967639923095703,
+      "learning_rate": 1.962368483345486e-06,
+      "loss": 0.3327,
+      "step": 4754
+    },
+    {
+      "epoch": 2.2112850720818478,
+      "grad_norm": 0.30786556005477905,
+      "learning_rate": 1.9602196038579774e-06,
+      "loss": 0.3022,
+      "step": 4755
+    },
+    {
+      "epoch": 2.2117501162610447,
+      "grad_norm": 0.3452177345752716,
+      "learning_rate": 1.9580716146727734e-06,
+      "loss": 0.3256,
+      "step": 4756
+    },
+    {
+      "epoch": 2.212215160440242,
+      "grad_norm": 0.33406537771224976,
+      "learning_rate": 1.9559245164189812e-06,
+      "loss": 0.3443,
+      "step": 4757
+    },
+    {
+      "epoch": 2.212680204619439,
+      "grad_norm": 0.31842756271362305,
+      "learning_rate": 1.9537783097254543e-06,
+      "loss": 0.3247,
+      "step": 4758
+    },
+    {
+      "epoch": 2.213145248798636,
+      "grad_norm": 0.3189336955547333,
+      "learning_rate": 1.9516329952207787e-06,
+      "loss": 0.3472,
+      "step": 4759
+    },
+    {
+      "epoch": 2.2136102929778327,
+      "grad_norm": 0.3681834638118744,
+      "learning_rate": 1.949488573533285e-06,
+      "loss": 0.3135,
+      "step": 4760
+    },
+    {
+      "epoch": 2.21407533715703,
+      "grad_norm": 0.3542393445968628,
+      "learning_rate": 1.9473450452910365e-06,
+      "loss": 0.3136,
+      "step": 4761
+    },
+    {
+      "epoch": 2.214540381336227,
+      "grad_norm": 0.3548336327075958,
+      "learning_rate": 1.9452024111218414e-06,
+      "loss": 0.3259,
+      "step": 4762
+    },
+    {
+      "epoch": 2.215005425515424,
+      "grad_norm": 0.34949004650115967,
+      "learning_rate": 1.9430606716532393e-06,
+      "loss": 0.331,
+      "step": 4763
+    },
+    {
+      "epoch": 2.215470469694621,
+      "grad_norm": 0.3351438641548157,
+      "learning_rate": 1.940919827512513e-06,
+      "loss": 0.2809,
+      "step": 4764
+    },
+    {
+      "epoch": 2.215935513873818,
+      "grad_norm": 0.3751305639743805,
+      "learning_rate": 1.938779879326679e-06,
+      "loss": 0.3407,
+      "step": 4765
+    },
+    {
+      "epoch": 2.216400558053015,
+      "grad_norm": 0.3458346724510193,
+      "learning_rate": 1.936640827722494e-06,
+      "loss": 0.3044,
+      "step": 4766
+    },
+    {
+      "epoch": 2.216865602232212,
+      "grad_norm": 0.31323379278182983,
+      "learning_rate": 1.934502673326452e-06,
+      "loss": 0.2983,
+      "step": 4767
+    },
+    {
+      "epoch": 2.2173306464114093,
+      "grad_norm": 0.35735267400741577,
+      "learning_rate": 1.9323654167647854e-06,
+      "loss": 0.3316,
+      "step": 4768
+    },
+    {
+      "epoch": 2.217795690590606,
+      "grad_norm": 0.3447880744934082,
+      "learning_rate": 1.930229058663459e-06,
+      "loss": 0.2798,
+      "step": 4769
+    },
+    {
+      "epoch": 2.218260734769803,
+      "grad_norm": 0.3700970709323883,
+      "learning_rate": 1.9280935996481792e-06,
+      "loss": 0.319,
+      "step": 4770
+    },
+    {
+      "epoch": 2.218725778949,
+      "grad_norm": 0.34711953997612,
+      "learning_rate": 1.9259590403443857e-06,
+      "loss": 0.3449,
+      "step": 4771
+    },
+    {
+      "epoch": 2.2191908231281974,
+      "grad_norm": 0.3344910144805908,
+      "learning_rate": 1.923825381377259e-06,
+      "loss": 0.2688,
+      "step": 4772
+    },
+    {
+      "epoch": 2.2196558673073943,
+      "grad_norm": 0.38596683740615845,
+      "learning_rate": 1.9216926233717087e-06,
+      "loss": 0.3415,
+      "step": 4773
+    },
+    {
+      "epoch": 2.220120911486591,
+      "grad_norm": 0.36054715514183044,
+      "learning_rate": 1.9195607669523903e-06,
+      "loss": 0.2957,
+      "step": 4774
+    },
+    {
+      "epoch": 2.220585955665788,
+      "grad_norm": 0.33113333582878113,
+      "learning_rate": 1.9174298127436845e-06,
+      "loss": 0.3162,
+      "step": 4775
+    },
+    {
+      "epoch": 2.2210509998449854,
+      "grad_norm": 0.3493165373802185,
+      "learning_rate": 1.9152997613697184e-06,
+      "loss": 0.3168,
+      "step": 4776
+    },
+    {
+      "epoch": 2.2215160440241823,
+      "grad_norm": 0.37465736269950867,
+      "learning_rate": 1.913170613454345e-06,
+      "loss": 0.351,
+      "step": 4777
+    },
+    {
+      "epoch": 2.2219810882033793,
+      "grad_norm": 0.3411155045032501,
+      "learning_rate": 1.9110423696211588e-06,
+      "loss": 0.3113,
+      "step": 4778
+    },
+    {
+      "epoch": 2.222446132382576,
+      "grad_norm": 0.35005953907966614,
+      "learning_rate": 1.9089150304934883e-06,
+      "loss": 0.3244,
+      "step": 4779
+    },
+    {
+      "epoch": 2.2229111765617735,
+      "grad_norm": 0.3763487935066223,
+      "learning_rate": 1.9067885966943983e-06,
+      "loss": 0.3509,
+      "step": 4780
+    },
+    {
+      "epoch": 2.2233762207409704,
+      "grad_norm": 0.3517736792564392,
+      "learning_rate": 1.9046630688466827e-06,
+      "loss": 0.3222,
+      "step": 4781
+    },
+    {
+      "epoch": 2.2238412649201673,
+      "grad_norm": 0.3561932146549225,
+      "learning_rate": 1.9025384475728787e-06,
+      "loss": 0.3047,
+      "step": 4782
+    },
+    {
+      "epoch": 2.224306309099364,
+      "grad_norm": 0.3545885980129242,
+      "learning_rate": 1.9004147334952483e-06,
+      "loss": 0.3192,
+      "step": 4783
+    },
+    {
+      "epoch": 2.2247713532785616,
+      "grad_norm": 0.3312489092350006,
+      "learning_rate": 1.8982919272357974e-06,
+      "loss": 0.3082,
+      "step": 4784
+    },
+    {
+      "epoch": 2.2252363974577585,
+      "grad_norm": 0.36054548621177673,
+      "learning_rate": 1.8961700294162578e-06,
+      "loss": 0.3275,
+      "step": 4785
+    },
+    {
+      "epoch": 2.2257014416369554,
+      "grad_norm": 0.350093811750412,
+      "learning_rate": 1.8940490406581018e-06,
+      "loss": 0.2902,
+      "step": 4786
+    },
+    {
+      "epoch": 2.2261664858161527,
+      "grad_norm": 0.34745752811431885,
+      "learning_rate": 1.8919289615825304e-06,
+      "loss": 0.3468,
+      "step": 4787
+    },
+    {
+      "epoch": 2.2266315299953496,
+      "grad_norm": 0.3414476811885834,
+      "learning_rate": 1.8898097928104825e-06,
+      "loss": 0.2941,
+      "step": 4788
+    },
+    {
+      "epoch": 2.2270965741745465,
+      "grad_norm": 0.3686240315437317,
+      "learning_rate": 1.8876915349626258e-06,
+      "loss": 0.3379,
+      "step": 4789
+    },
+    {
+      "epoch": 2.2275616183537434,
+      "grad_norm": 0.3241952657699585,
+      "learning_rate": 1.8855741886593643e-06,
+      "loss": 0.3107,
+      "step": 4790
+    },
+    {
+      "epoch": 2.228026662532941,
+      "grad_norm": 0.3654305934906006,
+      "learning_rate": 1.883457754520835e-06,
+      "loss": 0.3199,
+      "step": 4791
+    },
+    {
+      "epoch": 2.2284917067121377,
+      "grad_norm": 0.3426763415336609,
+      "learning_rate": 1.8813422331669084e-06,
+      "loss": 0.3018,
+      "step": 4792
+    },
+    {
+      "epoch": 2.2289567508913346,
+      "grad_norm": 0.39639508724212646,
+      "learning_rate": 1.8792276252171855e-06,
+      "loss": 0.3529,
+      "step": 4793
+    },
+    {
+      "epoch": 2.2294217950705315,
+      "grad_norm": 0.3361477553844452,
+      "learning_rate": 1.8771139312909976e-06,
+      "loss": 0.2977,
+      "step": 4794
+    },
+    {
+      "epoch": 2.229886839249729,
+      "grad_norm": 0.3340165913105011,
+      "learning_rate": 1.8750011520074158e-06,
+      "loss": 0.2923,
+      "step": 4795
+    },
+    {
+      "epoch": 2.2303518834289258,
+      "grad_norm": 0.3512633442878723,
+      "learning_rate": 1.8728892879852345e-06,
+      "loss": 0.3543,
+      "step": 4796
+    },
+    {
+      "epoch": 2.2308169276081227,
+      "grad_norm": 0.35980337858200073,
+      "learning_rate": 1.870778339842989e-06,
+      "loss": 0.3258,
+      "step": 4797
+    },
+    {
+      "epoch": 2.23128197178732,
+      "grad_norm": 0.3445799946784973,
+      "learning_rate": 1.8686683081989371e-06,
+      "loss": 0.3035,
+      "step": 4798
+    },
+    {
+      "epoch": 2.231747015966517,
+      "grad_norm": 0.3666563630104065,
+      "learning_rate": 1.866559193671077e-06,
+      "loss": 0.2764,
+      "step": 4799
+    },
+    {
+      "epoch": 2.232212060145714,
+      "grad_norm": 0.36018139123916626,
+      "learning_rate": 1.8644509968771302e-06,
+      "loss": 0.3334,
+      "step": 4800
+    },
+    {
+      "epoch": 2.2326771043249107,
+      "grad_norm": 0.3475804626941681,
+      "learning_rate": 1.8623437184345556e-06,
+      "loss": 0.34,
+      "step": 4801
+    },
+    {
+      "epoch": 2.233142148504108,
+      "grad_norm": 0.33812299370765686,
+      "learning_rate": 1.86023735896054e-06,
+      "loss": 0.3183,
+      "step": 4802
+    },
+    {
+      "epoch": 2.233607192683305,
+      "grad_norm": 0.3337913453578949,
+      "learning_rate": 1.8581319190720038e-06,
+      "loss": 0.3004,
+      "step": 4803
+    },
+    {
+      "epoch": 2.234072236862502,
+      "grad_norm": 0.3546447157859802,
+      "learning_rate": 1.8560273993855938e-06,
+      "loss": 0.3009,
+      "step": 4804
+    },
+    {
+      "epoch": 2.234537281041699,
+      "grad_norm": 0.37179654836654663,
+      "learning_rate": 1.8539238005176912e-06,
+      "loss": 0.3277,
+      "step": 4805
+    },
+    {
+      "epoch": 2.235002325220896,
+      "grad_norm": 0.36469766497612,
+      "learning_rate": 1.8518211230844042e-06,
+      "loss": 0.3292,
+      "step": 4806
+    },
+    {
+      "epoch": 2.235467369400093,
+      "grad_norm": 0.3361158072948456,
+      "learning_rate": 1.849719367701575e-06,
+      "loss": 0.3354,
+      "step": 4807
+    },
+    {
+      "epoch": 2.23593241357929,
+      "grad_norm": 0.3247022330760956,
+      "learning_rate": 1.8476185349847713e-06,
+      "loss": 0.3196,
+      "step": 4808
+    },
+    {
+      "epoch": 2.236397457758487,
+      "grad_norm": 0.36382511258125305,
+      "learning_rate": 1.8455186255492956e-06,
+      "loss": 0.342,
+      "step": 4809
+    },
+    {
+      "epoch": 2.236862501937684,
+      "grad_norm": 0.33687886595726013,
+      "learning_rate": 1.8434196400101744e-06,
+      "loss": 0.2892,
+      "step": 4810
+    },
+    {
+      "epoch": 2.237327546116881,
+      "grad_norm": 0.3983268737792969,
+      "learning_rate": 1.8413215789821692e-06,
+      "loss": 0.3366,
+      "step": 4811
+    },
+    {
+      "epoch": 2.237792590296078,
+      "grad_norm": 0.37651053071022034,
+      "learning_rate": 1.839224443079765e-06,
+      "loss": 0.35,
+      "step": 4812
+    },
+    {
+      "epoch": 2.238257634475275,
+      "grad_norm": 0.33488836884498596,
+      "learning_rate": 1.8371282329171803e-06,
+      "loss": 0.2801,
+      "step": 4813
+    },
+    {
+      "epoch": 2.2387226786544723,
+      "grad_norm": 0.341766357421875,
+      "learning_rate": 1.8350329491083613e-06,
+      "loss": 0.3194,
+      "step": 4814
+    },
+    {
+      "epoch": 2.239187722833669,
+      "grad_norm": 0.3747437596321106,
+      "learning_rate": 1.832938592266984e-06,
+      "loss": 0.3123,
+      "step": 4815
+    },
+    {
+      "epoch": 2.239652767012866,
+      "grad_norm": 0.351366251707077,
+      "learning_rate": 1.8308451630064484e-06,
+      "loss": 0.309,
+      "step": 4816
+    },
+    {
+      "epoch": 2.2401178111920634,
+      "grad_norm": 0.40906503796577454,
+      "learning_rate": 1.8287526619398888e-06,
+      "loss": 0.3532,
+      "step": 4817
+    },
+    {
+      "epoch": 2.2405828553712603,
+      "grad_norm": 0.3506213426589966,
+      "learning_rate": 1.8266610896801624e-06,
+      "loss": 0.3236,
+      "step": 4818
+    },
+    {
+      "epoch": 2.2410478995504572,
+      "grad_norm": 0.3513505756855011,
+      "learning_rate": 1.824570446839859e-06,
+      "loss": 0.3113,
+      "step": 4819
+    },
+    {
+      "epoch": 2.241512943729654,
+      "grad_norm": 0.33928313851356506,
+      "learning_rate": 1.8224807340312912e-06,
+      "loss": 0.3216,
+      "step": 4820
+    },
+    {
+      "epoch": 2.2419779879088515,
+      "grad_norm": 0.3162320852279663,
+      "learning_rate": 1.8203919518665049e-06,
+      "loss": 0.3097,
+      "step": 4821
+    },
+    {
+      "epoch": 2.2424430320880484,
+      "grad_norm": 0.36044222116470337,
+      "learning_rate": 1.8183041009572678e-06,
+      "loss": 0.33,
+      "step": 4822
+    },
+    {
+      "epoch": 2.2429080762672453,
+      "grad_norm": 0.4068707525730133,
+      "learning_rate": 1.8162171819150798e-06,
+      "loss": 0.3338,
+      "step": 4823
+    },
+    {
+      "epoch": 2.243373120446442,
+      "grad_norm": 0.3441452383995056,
+      "learning_rate": 1.8141311953511637e-06,
+      "loss": 0.3387,
+      "step": 4824
+    },
+    {
+      "epoch": 2.2438381646256396,
+      "grad_norm": 0.33748528361320496,
+      "learning_rate": 1.8120461418764711e-06,
+      "loss": 0.3243,
+      "step": 4825
+    },
+    {
+      "epoch": 2.2443032088048365,
+      "grad_norm": 0.3908786475658417,
+      "learning_rate": 1.8099620221016818e-06,
+      "loss": 0.3282,
+      "step": 4826
+    },
+    {
+      "epoch": 2.2447682529840334,
+      "grad_norm": 0.3668133616447449,
+      "learning_rate": 1.8078788366372008e-06,
+      "loss": 0.2975,
+      "step": 4827
+    },
+    {
+      "epoch": 2.2452332971632307,
+      "grad_norm": 0.3833257257938385,
+      "learning_rate": 1.8057965860931593e-06,
+      "loss": 0.312,
+      "step": 4828
+    },
+    {
+      "epoch": 2.2456983413424276,
+      "grad_norm": 0.3340941369533539,
+      "learning_rate": 1.8037152710794115e-06,
+      "loss": 0.3094,
+      "step": 4829
+    },
+    {
+      "epoch": 2.2461633855216245,
+      "grad_norm": 0.3409848213195801,
+      "learning_rate": 1.8016348922055448e-06,
+      "loss": 0.2966,
+      "step": 4830
+    },
+    {
+      "epoch": 2.2466284297008214,
+      "grad_norm": 0.3662347197532654,
+      "learning_rate": 1.7995554500808655e-06,
+      "loss": 0.3617,
+      "step": 4831
+    },
+    {
+      "epoch": 2.247093473880019,
+      "grad_norm": 0.4032919406890869,
+      "learning_rate": 1.7974769453144102e-06,
+      "loss": 0.3042,
+      "step": 4832
+    },
+    {
+      "epoch": 2.2475585180592157,
+      "grad_norm": 0.3917955160140991,
+      "learning_rate": 1.7953993785149377e-06,
+      "loss": 0.3188,
+      "step": 4833
+    },
+    {
+      "epoch": 2.2480235622384126,
+      "grad_norm": 0.36748558282852173,
+      "learning_rate": 1.7933227502909361e-06,
+      "loss": 0.3355,
+      "step": 4834
+    },
+    {
+      "epoch": 2.2484886064176095,
+      "grad_norm": 0.36093080043792725,
+      "learning_rate": 1.7912470612506123e-06,
+      "loss": 0.3389,
+      "step": 4835
+    },
+    {
+      "epoch": 2.248953650596807,
+      "grad_norm": 0.34391847252845764,
+      "learning_rate": 1.7891723120019038e-06,
+      "loss": 0.3463,
+      "step": 4836
+    },
+    {
+      "epoch": 2.2494186947760038,
+      "grad_norm": 0.3272242844104767,
+      "learning_rate": 1.7870985031524718e-06,
+      "loss": 0.3093,
+      "step": 4837
+    },
+    {
+      "epoch": 2.2498837389552007,
+      "grad_norm": 0.3449787199497223,
+      "learning_rate": 1.7850256353097017e-06,
+      "loss": 0.3272,
+      "step": 4838
+    },
+    {
+      "epoch": 2.250348783134398,
+      "grad_norm": 0.36595356464385986,
+      "learning_rate": 1.7829537090807002e-06,
+      "loss": 0.325,
+      "step": 4839
+    },
+    {
+      "epoch": 2.250813827313595,
+      "grad_norm": 0.35330867767333984,
+      "learning_rate": 1.7808827250723043e-06,
+      "loss": 0.3626,
+      "step": 4840
+    },
+    {
+      "epoch": 2.251278871492792,
+      "grad_norm": 0.3324193060398102,
+      "learning_rate": 1.7788126838910674e-06,
+      "loss": 0.3043,
+      "step": 4841
+    },
+    {
+      "epoch": 2.2517439156719887,
+      "grad_norm": 0.36338382959365845,
+      "learning_rate": 1.7767435861432752e-06,
+      "loss": 0.3228,
+      "step": 4842
+    },
+    {
+      "epoch": 2.2522089598511856,
+      "grad_norm": 0.3679875135421753,
+      "learning_rate": 1.7746754324349291e-06,
+      "loss": 0.366,
+      "step": 4843
+    },
+    {
+      "epoch": 2.252674004030383,
+      "grad_norm": 0.3294106423854828,
+      "learning_rate": 1.7726082233717607e-06,
+      "loss": 0.3195,
+      "step": 4844
+    },
+    {
+      "epoch": 2.25313904820958,
+      "grad_norm": 0.3595585227012634,
+      "learning_rate": 1.7705419595592193e-06,
+      "loss": 0.3553,
+      "step": 4845
+    },
+    {
+      "epoch": 2.253604092388777,
+      "grad_norm": 0.34528473019599915,
+      "learning_rate": 1.7684766416024828e-06,
+      "loss": 0.302,
+      "step": 4846
+    },
+    {
+      "epoch": 2.254069136567974,
+      "grad_norm": 0.38927531242370605,
+      "learning_rate": 1.7664122701064462e-06,
+      "loss": 0.3594,
+      "step": 4847
+    },
+    {
+      "epoch": 2.254534180747171,
+      "grad_norm": 0.30416175723075867,
+      "learning_rate": 1.7643488456757324e-06,
+      "loss": 0.3064,
+      "step": 4848
+    },
+    {
+      "epoch": 2.254999224926368,
+      "grad_norm": 0.37094974517822266,
+      "learning_rate": 1.7622863689146841e-06,
+      "loss": 0.3446,
+      "step": 4849
+    },
+    {
+      "epoch": 2.255464269105565,
+      "grad_norm": 0.31221383810043335,
+      "learning_rate": 1.7602248404273692e-06,
+      "loss": 0.2903,
+      "step": 4850
+    },
+    {
+      "epoch": 2.255929313284762,
+      "grad_norm": 0.33237534761428833,
+      "learning_rate": 1.758164260817573e-06,
+      "loss": 0.3272,
+      "step": 4851
+    },
+    {
+      "epoch": 2.256394357463959,
+      "grad_norm": 0.3221777677536011,
+      "learning_rate": 1.7561046306888092e-06,
+      "loss": 0.3238,
+      "step": 4852
+    },
+    {
+      "epoch": 2.256859401643156,
+      "grad_norm": 0.3504292964935303,
+      "learning_rate": 1.7540459506443052e-06,
+      "loss": 0.3285,
+      "step": 4853
+    },
+    {
+      "epoch": 2.257324445822353,
+      "grad_norm": 0.3311966359615326,
+      "learning_rate": 1.7519882212870204e-06,
+      "loss": 0.3255,
+      "step": 4854
+    },
+    {
+      "epoch": 2.2577894900015503,
+      "grad_norm": 0.36982831358909607,
+      "learning_rate": 1.7499314432196257e-06,
+      "loss": 0.3495,
+      "step": 4855
+    },
+    {
+      "epoch": 2.258254534180747,
+      "grad_norm": 0.35521289706230164,
+      "learning_rate": 1.747875617044521e-06,
+      "loss": 0.3106,
+      "step": 4856
+    },
+    {
+      "epoch": 2.258719578359944,
+      "grad_norm": 0.3456581234931946,
+      "learning_rate": 1.7458207433638225e-06,
+      "loss": 0.3315,
+      "step": 4857
+    },
+    {
+      "epoch": 2.2591846225391414,
+      "grad_norm": 0.3502543270587921,
+      "learning_rate": 1.7437668227793714e-06,
+      "loss": 0.308,
+      "step": 4858
+    },
+    {
+      "epoch": 2.2596496667183383,
+      "grad_norm": 0.3269558846950531,
+      "learning_rate": 1.7417138558927244e-06,
+      "loss": 0.2988,
+      "step": 4859
+    },
+    {
+      "epoch": 2.2601147108975352,
+      "grad_norm": 0.3678964376449585,
+      "learning_rate": 1.7396618433051648e-06,
+      "loss": 0.336,
+      "step": 4860
+    },
+    {
+      "epoch": 2.260579755076732,
+      "grad_norm": 0.3400316834449768,
+      "learning_rate": 1.7376107856176928e-06,
+      "loss": 0.3004,
+      "step": 4861
+    },
+    {
+      "epoch": 2.2610447992559295,
+      "grad_norm": 0.35273489356040955,
+      "learning_rate": 1.7355606834310317e-06,
+      "loss": 0.3352,
+      "step": 4862
+    },
+    {
+      "epoch": 2.2615098434351264,
+      "grad_norm": 0.35708242654800415,
+      "learning_rate": 1.7335115373456202e-06,
+      "loss": 0.3263,
+      "step": 4863
+    },
+    {
+      "epoch": 2.2619748876143233,
+      "grad_norm": 0.31572067737579346,
+      "learning_rate": 1.7314633479616227e-06,
+      "loss": 0.2921,
+      "step": 4864
+    },
+    {
+      "epoch": 2.26243993179352,
+      "grad_norm": 0.3223276436328888,
+      "learning_rate": 1.7294161158789197e-06,
+      "loss": 0.3239,
+      "step": 4865
+    },
+    {
+      "epoch": 2.2629049759727176,
+      "grad_norm": 0.32633039355278015,
+      "learning_rate": 1.7273698416971095e-06,
+      "loss": 0.3272,
+      "step": 4866
+    },
+    {
+      "epoch": 2.2633700201519145,
+      "grad_norm": 0.3426065444946289,
+      "learning_rate": 1.725324526015517e-06,
+      "loss": 0.3702,
+      "step": 4867
+    },
+    {
+      "epoch": 2.2638350643311114,
+      "grad_norm": 0.33616894483566284,
+      "learning_rate": 1.723280169433178e-06,
+      "loss": 0.3157,
+      "step": 4868
+    },
+    {
+      "epoch": 2.2643001085103087,
+      "grad_norm": 0.3157622516155243,
+      "learning_rate": 1.7212367725488544e-06,
+      "loss": 0.282,
+      "step": 4869
+    },
+    {
+      "epoch": 2.2647651526895056,
+      "grad_norm": 0.3689837157726288,
+      "learning_rate": 1.7191943359610214e-06,
+      "loss": 0.3646,
+      "step": 4870
+    },
+    {
+      "epoch": 2.2652301968687025,
+      "grad_norm": 0.337637722492218,
+      "learning_rate": 1.7171528602678767e-06,
+      "loss": 0.3203,
+      "step": 4871
+    },
+    {
+      "epoch": 2.2656952410478994,
+      "grad_norm": 0.34560054540634155,
+      "learning_rate": 1.7151123460673353e-06,
+      "loss": 0.3025,
+      "step": 4872
+    },
+    {
+      "epoch": 2.2661602852270963,
+      "grad_norm": 0.3452926576137543,
+      "learning_rate": 1.7130727939570325e-06,
+      "loss": 0.3539,
+      "step": 4873
+    },
+    {
+      "epoch": 2.2666253294062937,
+      "grad_norm": 0.3485986888408661,
+      "learning_rate": 1.7110342045343164e-06,
+      "loss": 0.3004,
+      "step": 4874
+    },
+    {
+      "epoch": 2.2670903735854906,
+      "grad_norm": 0.3492816984653473,
+      "learning_rate": 1.7089965783962608e-06,
+      "loss": 0.2975,
+      "step": 4875
+    },
+    {
+      "epoch": 2.2675554177646875,
+      "grad_norm": 0.3906685709953308,
+      "learning_rate": 1.7069599161396488e-06,
+      "loss": 0.3477,
+      "step": 4876
+    },
+    {
+      "epoch": 2.268020461943885,
+      "grad_norm": 0.35646331310272217,
+      "learning_rate": 1.70492421836099e-06,
+      "loss": 0.3069,
+      "step": 4877
+    },
+    {
+      "epoch": 2.2684855061230818,
+      "grad_norm": 0.35431551933288574,
+      "learning_rate": 1.7028894856565036e-06,
+      "loss": 0.305,
+      "step": 4878
+    },
+    {
+      "epoch": 2.2689505503022787,
+      "grad_norm": 0.3560280501842499,
+      "learning_rate": 1.700855718622133e-06,
+      "loss": 0.3439,
+      "step": 4879
+    },
+    {
+      "epoch": 2.2694155944814756,
+      "grad_norm": 0.3366261124610901,
+      "learning_rate": 1.698822917853532e-06,
+      "loss": 0.3093,
+      "step": 4880
+    },
+    {
+      "epoch": 2.269880638660673,
+      "grad_norm": 0.3454442322254181,
+      "learning_rate": 1.6967910839460788e-06,
+      "loss": 0.3045,
+      "step": 4881
+    },
+    {
+      "epoch": 2.27034568283987,
+      "grad_norm": 0.360445499420166,
+      "learning_rate": 1.6947602174948609e-06,
+      "loss": 0.3238,
+      "step": 4882
+    },
+    {
+      "epoch": 2.2708107270190667,
+      "grad_norm": 0.342626690864563,
+      "learning_rate": 1.6927303190946876e-06,
+      "loss": 0.3257,
+      "step": 4883
+    },
+    {
+      "epoch": 2.2712757711982636,
+      "grad_norm": 0.3575427532196045,
+      "learning_rate": 1.6907013893400838e-06,
+      "loss": 0.334,
+      "step": 4884
+    },
+    {
+      "epoch": 2.271740815377461,
+      "grad_norm": 0.3537481427192688,
+      "learning_rate": 1.6886734288252904e-06,
+      "loss": 0.3557,
+      "step": 4885
+    },
+    {
+      "epoch": 2.272205859556658,
+      "grad_norm": 0.32836195826530457,
+      "learning_rate": 1.6866464381442622e-06,
+      "loss": 0.3094,
+      "step": 4886
+    },
+    {
+      "epoch": 2.272670903735855,
+      "grad_norm": 0.3523426651954651,
+      "learning_rate": 1.6846204178906744e-06,
+      "loss": 0.3144,
+      "step": 4887
+    },
+    {
+      "epoch": 2.273135947915052,
+      "grad_norm": 0.35294294357299805,
+      "learning_rate": 1.6825953686579126e-06,
+      "loss": 0.3225,
+      "step": 4888
+    },
+    {
+      "epoch": 2.273600992094249,
+      "grad_norm": 0.33541440963745117,
+      "learning_rate": 1.6805712910390836e-06,
+      "loss": 0.3106,
+      "step": 4889
+    },
+    {
+      "epoch": 2.274066036273446,
+      "grad_norm": 0.3566812574863434,
+      "learning_rate": 1.6785481856270042e-06,
+      "loss": 0.3216,
+      "step": 4890
+    },
+    {
+      "epoch": 2.274531080452643,
+      "grad_norm": 0.3586536645889282,
+      "learning_rate": 1.6765260530142114e-06,
+      "loss": 0.3015,
+      "step": 4891
+    },
+    {
+      "epoch": 2.27499612463184,
+      "grad_norm": 0.31585457921028137,
+      "learning_rate": 1.6745048937929525e-06,
+      "loss": 0.295,
+      "step": 4892
+    },
+    {
+      "epoch": 2.275461168811037,
+      "grad_norm": 0.36500096321105957,
+      "learning_rate": 1.6724847085551955e-06,
+      "loss": 0.3317,
+      "step": 4893
+    },
+    {
+      "epoch": 2.275926212990234,
+      "grad_norm": 0.3432101309299469,
+      "learning_rate": 1.6704654978926167e-06,
+      "loss": 0.3005,
+      "step": 4894
+    },
+    {
+      "epoch": 2.276391257169431,
+      "grad_norm": 0.34062010049819946,
+      "learning_rate": 1.6684472623966108e-06,
+      "loss": 0.2974,
+      "step": 4895
+    },
+    {
+      "epoch": 2.2768563013486283,
+      "grad_norm": 0.33995094895362854,
+      "learning_rate": 1.666430002658287e-06,
+      "loss": 0.3219,
+      "step": 4896
+    },
+    {
+      "epoch": 2.277321345527825,
+      "grad_norm": 0.32602110505104065,
+      "learning_rate": 1.6644137192684694e-06,
+      "loss": 0.3365,
+      "step": 4897
+    },
+    {
+      "epoch": 2.277786389707022,
+      "grad_norm": 0.3418422341346741,
+      "learning_rate": 1.6623984128176912e-06,
+      "loss": 0.3219,
+      "step": 4898
+    },
+    {
+      "epoch": 2.2782514338862194,
+      "grad_norm": 0.3250005841255188,
+      "learning_rate": 1.6603840838962066e-06,
+      "loss": 0.3283,
+      "step": 4899
+    },
+    {
+      "epoch": 2.2787164780654163,
+      "grad_norm": 0.366692453622818,
+      "learning_rate": 1.6583707330939774e-06,
+      "loss": 0.3365,
+      "step": 4900
+    },
+    {
+      "epoch": 2.2791815222446132,
+      "grad_norm": 0.29823222756385803,
+      "learning_rate": 1.6563583610006806e-06,
+      "loss": 0.3037,
+      "step": 4901
+    },
+    {
+      "epoch": 2.27964656642381,
+      "grad_norm": 0.36057764291763306,
+      "learning_rate": 1.6543469682057105e-06,
+      "loss": 0.3333,
+      "step": 4902
+    },
+    {
+      "epoch": 2.280111610603007,
+      "grad_norm": 0.34210824966430664,
+      "learning_rate": 1.6523365552981674e-06,
+      "loss": 0.3378,
+      "step": 4903
+    },
+    {
+      "epoch": 2.2805766547822044,
+      "grad_norm": 0.3876771926879883,
+      "learning_rate": 1.6503271228668726e-06,
+      "loss": 0.3132,
+      "step": 4904
+    },
+    {
+      "epoch": 2.2810416989614013,
+      "grad_norm": 0.36838680505752563,
+      "learning_rate": 1.6483186715003523e-06,
+      "loss": 0.3283,
+      "step": 4905
+    },
+    {
+      "epoch": 2.281506743140598,
+      "grad_norm": 0.32702258229255676,
+      "learning_rate": 1.6463112017868516e-06,
+      "loss": 0.3193,
+      "step": 4906
+    },
+    {
+      "epoch": 2.2819717873197956,
+      "grad_norm": 0.33795708417892456,
+      "learning_rate": 1.6443047143143248e-06,
+      "loss": 0.3234,
+      "step": 4907
+    },
+    {
+      "epoch": 2.2824368314989925,
+      "grad_norm": 0.3263553977012634,
+      "learning_rate": 1.6422992096704422e-06,
+      "loss": 0.2925,
+      "step": 4908
+    },
+    {
+      "epoch": 2.2829018756781894,
+      "grad_norm": 0.38260510563850403,
+      "learning_rate": 1.6402946884425796e-06,
+      "loss": 0.3497,
+      "step": 4909
+    },
+    {
+      "epoch": 2.2833669198573863,
+      "grad_norm": 0.38565513491630554,
+      "learning_rate": 1.6382911512178323e-06,
+      "loss": 0.32,
+      "step": 4910
+    },
+    {
+      "epoch": 2.2838319640365836,
+      "grad_norm": 0.3692467212677002,
+      "learning_rate": 1.6362885985830001e-06,
+      "loss": 0.3332,
+      "step": 4911
+    },
+    {
+      "epoch": 2.2842970082157805,
+      "grad_norm": 0.3257414996623993,
+      "learning_rate": 1.6342870311246024e-06,
+      "loss": 0.3331,
+      "step": 4912
+    },
+    {
+      "epoch": 2.2847620523949774,
+      "grad_norm": 0.3339729607105255,
+      "learning_rate": 1.6322864494288616e-06,
+      "loss": 0.3091,
+      "step": 4913
+    },
+    {
+      "epoch": 2.2852270965741743,
+      "grad_norm": 0.32545122504234314,
+      "learning_rate": 1.6302868540817184e-06,
+      "loss": 0.3183,
+      "step": 4914
+    },
+    {
+      "epoch": 2.2856921407533717,
+      "grad_norm": 0.33002758026123047,
+      "learning_rate": 1.6282882456688197e-06,
+      "loss": 0.3126,
+      "step": 4915
+    },
+    {
+      "epoch": 2.2861571849325686,
+      "grad_norm": 0.3634626865386963,
+      "learning_rate": 1.6262906247755284e-06,
+      "loss": 0.3097,
+      "step": 4916
+    },
+    {
+      "epoch": 2.2866222291117655,
+      "grad_norm": 0.33301109075546265,
+      "learning_rate": 1.6242939919869117e-06,
+      "loss": 0.3269,
+      "step": 4917
+    },
+    {
+      "epoch": 2.287087273290963,
+      "grad_norm": 0.3408260941505432,
+      "learning_rate": 1.6222983478877525e-06,
+      "loss": 0.3362,
+      "step": 4918
+    },
+    {
+      "epoch": 2.2875523174701597,
+      "grad_norm": 0.35351526737213135,
+      "learning_rate": 1.6203036930625427e-06,
+      "loss": 0.3216,
+      "step": 4919
+    },
+    {
+      "epoch": 2.2880173616493567,
+      "grad_norm": 0.3414493799209595,
+      "learning_rate": 1.618310028095486e-06,
+      "loss": 0.356,
+      "step": 4920
+    },
+    {
+      "epoch": 2.2884824058285536,
+      "grad_norm": 0.3114349842071533,
+      "learning_rate": 1.6163173535704913e-06,
+      "loss": 0.3032,
+      "step": 4921
+    },
+    {
+      "epoch": 2.288947450007751,
+      "grad_norm": 0.3438819944858551,
+      "learning_rate": 1.6143256700711835e-06,
+      "loss": 0.3043,
+      "step": 4922
+    },
+    {
+      "epoch": 2.289412494186948,
+      "grad_norm": 0.35198479890823364,
+      "learning_rate": 1.6123349781808911e-06,
+      "loss": 0.3026,
+      "step": 4923
+    },
+    {
+      "epoch": 2.2898775383661447,
+      "grad_norm": 0.339100182056427,
+      "learning_rate": 1.610345278482659e-06,
+      "loss": 0.3188,
+      "step": 4924
+    },
+    {
+      "epoch": 2.2903425825453416,
+      "grad_norm": 0.3438659608364105,
+      "learning_rate": 1.6083565715592343e-06,
+      "loss": 0.311,
+      "step": 4925
+    },
+    {
+      "epoch": 2.290807626724539,
+      "grad_norm": 0.3378322720527649,
+      "learning_rate": 1.606368857993081e-06,
+      "loss": 0.3261,
+      "step": 4926
+    },
+    {
+      "epoch": 2.291272670903736,
+      "grad_norm": 0.3330194652080536,
+      "learning_rate": 1.6043821383663638e-06,
+      "loss": 0.3251,
+      "step": 4927
+    },
+    {
+      "epoch": 2.291737715082933,
+      "grad_norm": 0.3478861451148987,
+      "learning_rate": 1.6023964132609642e-06,
+      "loss": 0.3453,
+      "step": 4928
+    },
+    {
+      "epoch": 2.29220275926213,
+      "grad_norm": 0.33348485827445984,
+      "learning_rate": 1.600411683258466e-06,
+      "loss": 0.3251,
+      "step": 4929
+    },
+    {
+      "epoch": 2.292667803441327,
+      "grad_norm": 0.33961689472198486,
+      "learning_rate": 1.5984279489401655e-06,
+      "loss": 0.3144,
+      "step": 4930
+    },
+    {
+      "epoch": 2.293132847620524,
+      "grad_norm": 0.33800020813941956,
+      "learning_rate": 1.596445210887067e-06,
+      "loss": 0.3234,
+      "step": 4931
+    },
+    {
+      "epoch": 2.293597891799721,
+      "grad_norm": 0.3581719398498535,
+      "learning_rate": 1.5944634696798827e-06,
+      "loss": 0.3088,
+      "step": 4932
+    },
+    {
+      "epoch": 2.2940629359789177,
+      "grad_norm": 0.32927682995796204,
+      "learning_rate": 1.5924827258990305e-06,
+      "loss": 0.2917,
+      "step": 4933
+    },
+    {
+      "epoch": 2.294527980158115,
+      "grad_norm": 0.3471710681915283,
+      "learning_rate": 1.5905029801246401e-06,
+      "loss": 0.3399,
+      "step": 4934
+    },
+    {
+      "epoch": 2.294993024337312,
+      "grad_norm": 0.36960649490356445,
+      "learning_rate": 1.5885242329365448e-06,
+      "loss": 0.319,
+      "step": 4935
+    },
+    {
+      "epoch": 2.295458068516509,
+      "grad_norm": 0.3120764493942261,
+      "learning_rate": 1.5865464849142897e-06,
+      "loss": 0.309,
+      "step": 4936
+    },
+    {
+      "epoch": 2.2959231126957063,
+      "grad_norm": 0.32464638352394104,
+      "learning_rate": 1.5845697366371237e-06,
+      "loss": 0.3333,
+      "step": 4937
+    },
+    {
+      "epoch": 2.296388156874903,
+      "grad_norm": 0.35078492760658264,
+      "learning_rate": 1.5825939886840036e-06,
+      "loss": 0.3327,
+      "step": 4938
+    },
+    {
+      "epoch": 2.2968532010541,
+      "grad_norm": 0.3693057596683502,
+      "learning_rate": 1.5806192416335959e-06,
+      "loss": 0.2898,
+      "step": 4939
+    },
+    {
+      "epoch": 2.297318245233297,
+      "grad_norm": 0.39447295665740967,
+      "learning_rate": 1.5786454960642694e-06,
+      "loss": 0.3496,
+      "step": 4940
+    },
+    {
+      "epoch": 2.2977832894124943,
+      "grad_norm": 0.30697473883628845,
+      "learning_rate": 1.576672752554103e-06,
+      "loss": 0.2768,
+      "step": 4941
+    },
+    {
+      "epoch": 2.2982483335916912,
+      "grad_norm": 0.33453670144081116,
+      "learning_rate": 1.574701011680882e-06,
+      "loss": 0.3336,
+      "step": 4942
+    },
+    {
+      "epoch": 2.298713377770888,
+      "grad_norm": 0.3246103525161743,
+      "learning_rate": 1.572730274022099e-06,
+      "loss": 0.318,
+      "step": 4943
+    },
+    {
+      "epoch": 2.299178421950085,
+      "grad_norm": 0.38411572575569153,
+      "learning_rate": 1.570760540154947e-06,
+      "loss": 0.3616,
+      "step": 4944
+    },
+    {
+      "epoch": 2.2996434661292824,
+      "grad_norm": 0.35833507776260376,
+      "learning_rate": 1.5687918106563326e-06,
+      "loss": 0.323,
+      "step": 4945
+    },
+    {
+      "epoch": 2.3001085103084793,
+      "grad_norm": 0.3165144920349121,
+      "learning_rate": 1.566824086102862e-06,
+      "loss": 0.2817,
+      "step": 4946
+    },
+    {
+      "epoch": 2.300573554487676,
+      "grad_norm": 0.3616514801979065,
+      "learning_rate": 1.5648573670708527e-06,
+      "loss": 0.3341,
+      "step": 4947
+    },
+    {
+      "epoch": 2.3010385986668735,
+      "grad_norm": 0.35235756635665894,
+      "learning_rate": 1.562891654136321e-06,
+      "loss": 0.3283,
+      "step": 4948
+    },
+    {
+      "epoch": 2.3015036428460705,
+      "grad_norm": 0.3527206480503082,
+      "learning_rate": 1.560926947874996e-06,
+      "loss": 0.3329,
+      "step": 4949
+    },
+    {
+      "epoch": 2.3019686870252674,
+      "grad_norm": 0.35835447907447815,
+      "learning_rate": 1.5589632488623053e-06,
+      "loss": 0.3169,
+      "step": 4950
+    },
+    {
+      "epoch": 2.3024337312044643,
+      "grad_norm": 0.33192187547683716,
+      "learning_rate": 1.557000557673387e-06,
+      "loss": 0.2644,
+      "step": 4951
+    },
+    {
+      "epoch": 2.3028987753836616,
+      "grad_norm": 0.3572143614292145,
+      "learning_rate": 1.5550388748830786e-06,
+      "loss": 0.3176,
+      "step": 4952
+    },
+    {
+      "epoch": 2.3033638195628585,
+      "grad_norm": 0.34204208850860596,
+      "learning_rate": 1.5530782010659267e-06,
+      "loss": 0.3265,
+      "step": 4953
+    },
+    {
+      "epoch": 2.3038288637420554,
+      "grad_norm": 0.3122238516807556,
+      "learning_rate": 1.5511185367961813e-06,
+      "loss": 0.2969,
+      "step": 4954
+    },
+    {
+      "epoch": 2.3042939079212523,
+      "grad_norm": 0.34026968479156494,
+      "learning_rate": 1.5491598826477967e-06,
+      "loss": 0.3238,
+      "step": 4955
+    },
+    {
+      "epoch": 2.3047589521004497,
+      "grad_norm": 0.31696975231170654,
+      "learning_rate": 1.5472022391944285e-06,
+      "loss": 0.2969,
+      "step": 4956
+    },
+    {
+      "epoch": 2.3052239962796466,
+      "grad_norm": 0.38175731897354126,
+      "learning_rate": 1.5452456070094419e-06,
+      "loss": 0.3834,
+      "step": 4957
+    },
+    {
+      "epoch": 2.3056890404588435,
+      "grad_norm": 0.31309717893600464,
+      "learning_rate": 1.543289986665899e-06,
+      "loss": 0.2931,
+      "step": 4958
+    },
+    {
+      "epoch": 2.306154084638041,
+      "grad_norm": 0.3553571403026581,
+      "learning_rate": 1.5413353787365726e-06,
+      "loss": 0.3586,
+      "step": 4959
+    },
+    {
+      "epoch": 2.3066191288172377,
+      "grad_norm": 0.3609411418437958,
+      "learning_rate": 1.5393817837939328e-06,
+      "loss": 0.2954,
+      "step": 4960
+    },
+    {
+      "epoch": 2.3070841729964346,
+      "grad_norm": 0.3583955466747284,
+      "learning_rate": 1.537429202410159e-06,
+      "loss": 0.3251,
+      "step": 4961
+    },
+    {
+      "epoch": 2.3075492171756316,
+      "grad_norm": 0.3274742662906647,
+      "learning_rate": 1.5354776351571266e-06,
+      "loss": 0.2977,
+      "step": 4962
+    },
+    {
+      "epoch": 2.308014261354829,
+      "grad_norm": 0.3816562294960022,
+      "learning_rate": 1.533527082606422e-06,
+      "loss": 0.3322,
+      "step": 4963
+    },
+    {
+      "epoch": 2.308479305534026,
+      "grad_norm": 0.33534786105155945,
+      "learning_rate": 1.5315775453293269e-06,
+      "loss": 0.3119,
+      "step": 4964
+    },
+    {
+      "epoch": 2.3089443497132227,
+      "grad_norm": 0.33446696400642395,
+      "learning_rate": 1.5296290238968303e-06,
+      "loss": 0.3481,
+      "step": 4965
+    },
+    {
+      "epoch": 2.3094093938924196,
+      "grad_norm": 0.32276517152786255,
+      "learning_rate": 1.5276815188796235e-06,
+      "loss": 0.3235,
+      "step": 4966
+    },
+    {
+      "epoch": 2.309874438071617,
+      "grad_norm": 0.3487582504749298,
+      "learning_rate": 1.5257350308480994e-06,
+      "loss": 0.3368,
+      "step": 4967
+    },
+    {
+      "epoch": 2.310339482250814,
+      "grad_norm": 0.32539722323417664,
+      "learning_rate": 1.5237895603723501e-06,
+      "loss": 0.3009,
+      "step": 4968
+    },
+    {
+      "epoch": 2.3108045264300108,
+      "grad_norm": 0.3673970699310303,
+      "learning_rate": 1.5218451080221763e-06,
+      "loss": 0.3305,
+      "step": 4969
+    },
+    {
+      "epoch": 2.311269570609208,
+      "grad_norm": 0.31702741980552673,
+      "learning_rate": 1.5199016743670719e-06,
+      "loss": 0.2929,
+      "step": 4970
+    },
+    {
+      "epoch": 2.311734614788405,
+      "grad_norm": 0.3536950647830963,
+      "learning_rate": 1.517959259976241e-06,
+      "loss": 0.3543,
+      "step": 4971
+    },
+    {
+      "epoch": 2.312199658967602,
+      "grad_norm": 0.3390069603919983,
+      "learning_rate": 1.5160178654185836e-06,
+      "loss": 0.3179,
+      "step": 4972
+    },
+    {
+      "epoch": 2.312664703146799,
+      "grad_norm": 0.3432963192462921,
+      "learning_rate": 1.5140774912627005e-06,
+      "loss": 0.3357,
+      "step": 4973
+    },
+    {
+      "epoch": 2.3131297473259957,
+      "grad_norm": 0.35164210200309753,
+      "learning_rate": 1.5121381380769002e-06,
+      "loss": 0.3311,
+      "step": 4974
+    },
+    {
+      "epoch": 2.313594791505193,
+      "grad_norm": 0.33762413263320923,
+      "learning_rate": 1.5101998064291828e-06,
+      "loss": 0.3111,
+      "step": 4975
+    },
+    {
+      "epoch": 2.31405983568439,
+      "grad_norm": 0.3147108852863312,
+      "learning_rate": 1.5082624968872578e-06,
+      "loss": 0.2938,
+      "step": 4976
+    },
+    {
+      "epoch": 2.314524879863587,
+      "grad_norm": 0.3715774714946747,
+      "learning_rate": 1.50632621001853e-06,
+      "loss": 0.3709,
+      "step": 4977
+    },
+    {
+      "epoch": 2.3149899240427843,
+      "grad_norm": 0.32812413573265076,
+      "learning_rate": 1.5043909463901086e-06,
+      "loss": 0.3058,
+      "step": 4978
+    },
+    {
+      "epoch": 2.315454968221981,
+      "grad_norm": 0.3472438156604767,
+      "learning_rate": 1.5024567065687977e-06,
+      "loss": 0.3163,
+      "step": 4979
+    },
+    {
+      "epoch": 2.315920012401178,
+      "grad_norm": 0.3634364604949951,
+      "learning_rate": 1.500523491121108e-06,
+      "loss": 0.3306,
+      "step": 4980
+    },
+    {
+      "epoch": 2.316385056580375,
+      "grad_norm": 0.3792670667171478,
+      "learning_rate": 1.4985913006132435e-06,
+      "loss": 0.336,
+      "step": 4981
+    },
+    {
+      "epoch": 2.3168501007595723,
+      "grad_norm": 0.32765889167785645,
+      "learning_rate": 1.496660135611115e-06,
+      "loss": 0.3053,
+      "step": 4982
+    },
+    {
+      "epoch": 2.3173151449387692,
+      "grad_norm": 0.35471829771995544,
+      "learning_rate": 1.4947299966803259e-06,
+      "loss": 0.3403,
+      "step": 4983
+    },
+    {
+      "epoch": 2.317780189117966,
+      "grad_norm": 0.3708203136920929,
+      "learning_rate": 1.4928008843861851e-06,
+      "loss": 0.2946,
+      "step": 4984
+    },
+    {
+      "epoch": 2.318245233297163,
+      "grad_norm": 0.3640722930431366,
+      "learning_rate": 1.490872799293696e-06,
+      "loss": 0.3321,
+      "step": 4985
+    },
+    {
+      "epoch": 2.3187102774763604,
+      "grad_norm": 0.33545586466789246,
+      "learning_rate": 1.4889457419675669e-06,
+      "loss": 0.2907,
+      "step": 4986
+    },
+    {
+      "epoch": 2.3191753216555573,
+      "grad_norm": 0.3201104402542114,
+      "learning_rate": 1.487019712972197e-06,
+      "loss": 0.3066,
+      "step": 4987
+    },
+    {
+      "epoch": 2.319640365834754,
+      "grad_norm": 0.32187438011169434,
+      "learning_rate": 1.4850947128716914e-06,
+      "loss": 0.3197,
+      "step": 4988
+    },
+    {
+      "epoch": 2.3201054100139515,
+      "grad_norm": 0.3507417142391205,
+      "learning_rate": 1.4831707422298513e-06,
+      "loss": 0.323,
+      "step": 4989
+    },
+    {
+      "epoch": 2.3205704541931484,
+      "grad_norm": 0.3247746229171753,
+      "learning_rate": 1.4812478016101784e-06,
+      "loss": 0.3163,
+      "step": 4990
+    },
+    {
+      "epoch": 2.3210354983723454,
+      "grad_norm": 0.34104475378990173,
+      "learning_rate": 1.4793258915758668e-06,
+      "loss": 0.3392,
+      "step": 4991
+    },
+    {
+      "epoch": 2.3215005425515423,
+      "grad_norm": 0.32340723276138306,
+      "learning_rate": 1.4774050126898164e-06,
+      "loss": 0.3241,
+      "step": 4992
+    },
+    {
+      "epoch": 2.3219655867307396,
+      "grad_norm": 0.348816454410553,
+      "learning_rate": 1.475485165514618e-06,
+      "loss": 0.3026,
+      "step": 4993
+    },
+    {
+      "epoch": 2.3224306309099365,
+      "grad_norm": 0.372328519821167,
+      "learning_rate": 1.473566350612567e-06,
+      "loss": 0.3019,
+      "step": 4994
+    },
+    {
+      "epoch": 2.3228956750891334,
+      "grad_norm": 0.3437679409980774,
+      "learning_rate": 1.47164856854565e-06,
+      "loss": 0.3165,
+      "step": 4995
+    },
+    {
+      "epoch": 2.3233607192683303,
+      "grad_norm": 0.36129605770111084,
+      "learning_rate": 1.4697318198755572e-06,
+      "loss": 0.3472,
+      "step": 4996
+    },
+    {
+      "epoch": 2.3238257634475277,
+      "grad_norm": 0.32315579056739807,
+      "learning_rate": 1.4678161051636703e-06,
+      "loss": 0.3008,
+      "step": 4997
+    },
+    {
+      "epoch": 2.3242908076267246,
+      "grad_norm": 0.35113513469696045,
+      "learning_rate": 1.4659014249710734e-06,
+      "loss": 0.3222,
+      "step": 4998
+    },
+    {
+      "epoch": 2.3247558518059215,
+      "grad_norm": 0.3178529143333435,
+      "learning_rate": 1.4639877798585434e-06,
+      "loss": 0.3028,
+      "step": 4999
+    },
+    {
+      "epoch": 2.325220895985119,
+      "grad_norm": 0.31862160563468933,
+      "learning_rate": 1.462075170386556e-06,
+      "loss": 0.306,
+      "step": 5000
+    },
+    {
+      "epoch": 2.3256859401643157,
+      "grad_norm": 0.3178994953632355,
+      "learning_rate": 1.4601635971152844e-06,
+      "loss": 0.3442,
+      "step": 5001
+    },
+    {
+      "epoch": 2.3261509843435126,
+      "grad_norm": 0.33765512704849243,
+      "learning_rate": 1.4582530606045986e-06,
+      "loss": 0.3313,
+      "step": 5002
+    },
+    {
+      "epoch": 2.3266160285227095,
+      "grad_norm": 0.3193725347518921,
+      "learning_rate": 1.456343561414061e-06,
+      "loss": 0.2883,
+      "step": 5003
+    },
+    {
+      "epoch": 2.3270810727019065,
+      "grad_norm": 0.3627553880214691,
+      "learning_rate": 1.4544351001029349e-06,
+      "loss": 0.3271,
+      "step": 5004
+    },
+    {
+      "epoch": 2.327546116881104,
+      "grad_norm": 0.3708834648132324,
+      "learning_rate": 1.4525276772301761e-06,
+      "loss": 0.3387,
+      "step": 5005
+    },
+    {
+      "epoch": 2.3280111610603007,
+      "grad_norm": 0.33413180708885193,
+      "learning_rate": 1.4506212933544394e-06,
+      "loss": 0.3144,
+      "step": 5006
+    },
+    {
+      "epoch": 2.3284762052394976,
+      "grad_norm": 0.3867631256580353,
+      "learning_rate": 1.4487159490340714e-06,
+      "loss": 0.3447,
+      "step": 5007
+    },
+    {
+      "epoch": 2.328941249418695,
+      "grad_norm": 0.3161846697330475,
+      "learning_rate": 1.4468116448271196e-06,
+      "loss": 0.3167,
+      "step": 5008
+    },
+    {
+      "epoch": 2.329406293597892,
+      "grad_norm": 0.3061203956604004,
+      "learning_rate": 1.4449083812913217e-06,
+      "loss": 0.3078,
+      "step": 5009
+    },
+    {
+      "epoch": 2.3298713377770888,
+      "grad_norm": 0.33758896589279175,
+      "learning_rate": 1.4430061589841122e-06,
+      "loss": 0.3081,
+      "step": 5010
+    },
+    {
+      "epoch": 2.3303363819562857,
+      "grad_norm": 0.37586334347724915,
+      "learning_rate": 1.4411049784626213e-06,
+      "loss": 0.324,
+      "step": 5011
+    },
+    {
+      "epoch": 2.330801426135483,
+      "grad_norm": 0.36526045203208923,
+      "learning_rate": 1.4392048402836744e-06,
+      "loss": 0.3264,
+      "step": 5012
+    },
+    {
+      "epoch": 2.33126647031468,
+      "grad_norm": 0.3223170042037964,
+      "learning_rate": 1.437305745003793e-06,
+      "loss": 0.2935,
+      "step": 5013
+    },
+    {
+      "epoch": 2.331731514493877,
+      "grad_norm": 0.3323318660259247,
+      "learning_rate": 1.4354076931791876e-06,
+      "loss": 0.3261,
+      "step": 5014
+    },
+    {
+      "epoch": 2.3321965586730737,
+      "grad_norm": 0.3371385931968689,
+      "learning_rate": 1.433510685365771e-06,
+      "loss": 0.3157,
+      "step": 5015
+    },
+    {
+      "epoch": 2.332661602852271,
+      "grad_norm": 0.31770241260528564,
+      "learning_rate": 1.4316147221191411e-06,
+      "loss": 0.3195,
+      "step": 5016
+    },
+    {
+      "epoch": 2.333126647031468,
+      "grad_norm": 0.35454097390174866,
+      "learning_rate": 1.4297198039945998e-06,
+      "loss": 0.3435,
+      "step": 5017
+    },
+    {
+      "epoch": 2.333591691210665,
+      "grad_norm": 0.34481361508369446,
+      "learning_rate": 1.4278259315471332e-06,
+      "loss": 0.312,
+      "step": 5018
+    },
+    {
+      "epoch": 2.3340567353898622,
+      "grad_norm": 0.37690550088882446,
+      "learning_rate": 1.425933105331429e-06,
+      "loss": 0.3404,
+      "step": 5019
+    },
+    {
+      "epoch": 2.334521779569059,
+      "grad_norm": 0.32657355070114136,
+      "learning_rate": 1.4240413259018632e-06,
+      "loss": 0.3181,
+      "step": 5020
+    },
+    {
+      "epoch": 2.334986823748256,
+      "grad_norm": 0.33882707357406616,
+      "learning_rate": 1.4221505938125097e-06,
+      "loss": 0.3137,
+      "step": 5021
+    },
+    {
+      "epoch": 2.335451867927453,
+      "grad_norm": 0.35311123728752136,
+      "learning_rate": 1.42026090961713e-06,
+      "loss": 0.3127,
+      "step": 5022
+    },
+    {
+      "epoch": 2.3359169121066503,
+      "grad_norm": 0.3354308307170868,
+      "learning_rate": 1.4183722738691834e-06,
+      "loss": 0.3445,
+      "step": 5023
+    },
+    {
+      "epoch": 2.336381956285847,
+      "grad_norm": 0.3541325330734253,
+      "learning_rate": 1.4164846871218213e-06,
+      "loss": 0.3136,
+      "step": 5024
+    },
+    {
+      "epoch": 2.336847000465044,
+      "grad_norm": 0.32369163632392883,
+      "learning_rate": 1.4145981499278877e-06,
+      "loss": 0.3183,
+      "step": 5025
+    },
+    {
+      "epoch": 2.337312044644241,
+      "grad_norm": 0.33266210556030273,
+      "learning_rate": 1.4127126628399168e-06,
+      "loss": 0.3208,
+      "step": 5026
+    },
+    {
+      "epoch": 2.3377770888234384,
+      "grad_norm": 0.3531791567802429,
+      "learning_rate": 1.410828226410139e-06,
+      "loss": 0.3399,
+      "step": 5027
+    },
+    {
+      "epoch": 2.3382421330026353,
+      "grad_norm": 0.3102116882801056,
+      "learning_rate": 1.4089448411904733e-06,
+      "loss": 0.3203,
+      "step": 5028
+    },
+    {
+      "epoch": 2.338707177181832,
+      "grad_norm": 0.39151903986930847,
+      "learning_rate": 1.4070625077325345e-06,
+      "loss": 0.3452,
+      "step": 5029
+    },
+    {
+      "epoch": 2.3391722213610295,
+      "grad_norm": 0.3667720854282379,
+      "learning_rate": 1.4051812265876257e-06,
+      "loss": 0.2859,
+      "step": 5030
+    },
+    {
+      "epoch": 2.3396372655402264,
+      "grad_norm": 0.30760088562965393,
+      "learning_rate": 1.4033009983067454e-06,
+      "loss": 0.2963,
+      "step": 5031
+    },
+    {
+      "epoch": 2.3401023097194233,
+      "grad_norm": 0.3295292854309082,
+      "learning_rate": 1.4014218234405796e-06,
+      "loss": 0.3199,
+      "step": 5032
+    },
+    {
+      "epoch": 2.3405673538986203,
+      "grad_norm": 0.321175754070282,
+      "learning_rate": 1.3995437025395109e-06,
+      "loss": 0.3068,
+      "step": 5033
+    },
+    {
+      "epoch": 2.341032398077817,
+      "grad_norm": 0.35323259234428406,
+      "learning_rate": 1.3976666361536074e-06,
+      "loss": 0.2936,
+      "step": 5034
+    },
+    {
+      "epoch": 2.3414974422570145,
+      "grad_norm": 0.3676642179489136,
+      "learning_rate": 1.395790624832633e-06,
+      "loss": 0.3327,
+      "step": 5035
+    },
+    {
+      "epoch": 2.3419624864362114,
+      "grad_norm": 0.31863468885421753,
+      "learning_rate": 1.3939156691260407e-06,
+      "loss": 0.2977,
+      "step": 5036
+    },
+    {
+      "epoch": 2.3424275306154083,
+      "grad_norm": 0.3256819248199463,
+      "learning_rate": 1.392041769582977e-06,
+      "loss": 0.3326,
+      "step": 5037
+    },
+    {
+      "epoch": 2.3428925747946057,
+      "grad_norm": 0.3244876265525818,
+      "learning_rate": 1.3901689267522718e-06,
+      "loss": 0.3405,
+      "step": 5038
+    },
+    {
+      "epoch": 2.3433576189738026,
+      "grad_norm": 0.31455737352371216,
+      "learning_rate": 1.3882971411824547e-06,
+      "loss": 0.3042,
+      "step": 5039
+    },
+    {
+      "epoch": 2.3438226631529995,
+      "grad_norm": 0.32401779294013977,
+      "learning_rate": 1.386426413421738e-06,
+      "loss": 0.3078,
+      "step": 5040
+    },
+    {
+      "epoch": 2.3442877073321964,
+      "grad_norm": 0.33598318696022034,
+      "learning_rate": 1.3845567440180308e-06,
+      "loss": 0.297,
+      "step": 5041
+    },
+    {
+      "epoch": 2.3447527515113937,
+      "grad_norm": 0.35334211587905884,
+      "learning_rate": 1.3826881335189258e-06,
+      "loss": 0.3692,
+      "step": 5042
+    },
+    {
+      "epoch": 2.3452177956905906,
+      "grad_norm": 0.35467445850372314,
+      "learning_rate": 1.3808205824717108e-06,
+      "loss": 0.3316,
+      "step": 5043
+    },
+    {
+      "epoch": 2.3456828398697875,
+      "grad_norm": 0.34783780574798584,
+      "learning_rate": 1.3789540914233607e-06,
+      "loss": 0.316,
+      "step": 5044
+    },
+    {
+      "epoch": 2.3461478840489844,
+      "grad_norm": 0.36762094497680664,
+      "learning_rate": 1.3770886609205381e-06,
+      "loss": 0.3563,
+      "step": 5045
+    },
+    {
+      "epoch": 2.346612928228182,
+      "grad_norm": 0.35472989082336426,
+      "learning_rate": 1.3752242915095993e-06,
+      "loss": 0.337,
+      "step": 5046
+    },
+    {
+      "epoch": 2.3470779724073787,
+      "grad_norm": 0.33739933371543884,
+      "learning_rate": 1.373360983736588e-06,
+      "loss": 0.2881,
+      "step": 5047
+    },
+    {
+      "epoch": 2.3475430165865756,
+      "grad_norm": 0.374533474445343,
+      "learning_rate": 1.3714987381472378e-06,
+      "loss": 0.3296,
+      "step": 5048
+    },
+    {
+      "epoch": 2.348008060765773,
+      "grad_norm": 0.32988038659095764,
+      "learning_rate": 1.3696375552869673e-06,
+      "loss": 0.316,
+      "step": 5049
+    },
+    {
+      "epoch": 2.34847310494497,
+      "grad_norm": 0.33644863963127136,
+      "learning_rate": 1.36777743570089e-06,
+      "loss": 0.3117,
+      "step": 5050
+    },
+    {
+      "epoch": 2.3489381491241668,
+      "grad_norm": 0.3263695538043976,
+      "learning_rate": 1.365918379933801e-06,
+      "loss": 0.299,
+      "step": 5051
+    },
+    {
+      "epoch": 2.3494031933033637,
+      "grad_norm": 0.3548851013183594,
+      "learning_rate": 1.3640603885301917e-06,
+      "loss": 0.3732,
+      "step": 5052
+    },
+    {
+      "epoch": 2.349868237482561,
+      "grad_norm": 0.32212895154953003,
+      "learning_rate": 1.362203462034234e-06,
+      "loss": 0.3201,
+      "step": 5053
+    },
+    {
+      "epoch": 2.350333281661758,
+      "grad_norm": 0.3988026976585388,
+      "learning_rate": 1.3603476009897942e-06,
+      "loss": 0.3429,
+      "step": 5054
+    },
+    {
+      "epoch": 2.350798325840955,
+      "grad_norm": 0.3584456145763397,
+      "learning_rate": 1.3584928059404207e-06,
+      "loss": 0.3324,
+      "step": 5055
+    },
+    {
+      "epoch": 2.3512633700201517,
+      "grad_norm": 0.315178781747818,
+      "learning_rate": 1.356639077429357e-06,
+      "loss": 0.3269,
+      "step": 5056
+    },
+    {
+      "epoch": 2.351728414199349,
+      "grad_norm": 0.3512877821922302,
+      "learning_rate": 1.354786415999526e-06,
+      "loss": 0.3097,
+      "step": 5057
+    },
+    {
+      "epoch": 2.352193458378546,
+      "grad_norm": 0.35129326581954956,
+      "learning_rate": 1.352934822193544e-06,
+      "loss": 0.3193,
+      "step": 5058
+    },
+    {
+      "epoch": 2.352658502557743,
+      "grad_norm": 0.3396090269088745,
+      "learning_rate": 1.351084296553713e-06,
+      "loss": 0.2999,
+      "step": 5059
+    },
+    {
+      "epoch": 2.3531235467369402,
+      "grad_norm": 0.3327389359474182,
+      "learning_rate": 1.3492348396220229e-06,
+      "loss": 0.3217,
+      "step": 5060
+    },
+    {
+      "epoch": 2.353588590916137,
+      "grad_norm": 0.31320449709892273,
+      "learning_rate": 1.3473864519401463e-06,
+      "loss": 0.2914,
+      "step": 5061
+    },
+    {
+      "epoch": 2.354053635095334,
+      "grad_norm": 0.343872994184494,
+      "learning_rate": 1.34553913404945e-06,
+      "loss": 0.3318,
+      "step": 5062
+    },
+    {
+      "epoch": 2.354518679274531,
+      "grad_norm": 0.31648504734039307,
+      "learning_rate": 1.3436928864909799e-06,
+      "loss": 0.2752,
+      "step": 5063
+    },
+    {
+      "epoch": 2.354983723453728,
+      "grad_norm": 0.38725370168685913,
+      "learning_rate": 1.341847709805475e-06,
+      "loss": 0.3429,
+      "step": 5064
+    },
+    {
+      "epoch": 2.355448767632925,
+      "grad_norm": 0.3474979102611542,
+      "learning_rate": 1.3400036045333542e-06,
+      "loss": 0.3161,
+      "step": 5065
+    },
+    {
+      "epoch": 2.355913811812122,
+      "grad_norm": 0.3323841392993927,
+      "learning_rate": 1.3381605712147294e-06,
+      "loss": 0.2989,
+      "step": 5066
+    },
+    {
+      "epoch": 2.356378855991319,
+      "grad_norm": 0.3497825562953949,
+      "learning_rate": 1.3363186103893916e-06,
+      "loss": 0.356,
+      "step": 5067
+    },
+    {
+      "epoch": 2.3568439001705164,
+      "grad_norm": 0.35843122005462646,
+      "learning_rate": 1.3344777225968247e-06,
+      "loss": 0.3005,
+      "step": 5068
+    },
+    {
+      "epoch": 2.3573089443497133,
+      "grad_norm": 0.3550424873828888,
+      "learning_rate": 1.332637908376192e-06,
+      "loss": 0.3315,
+      "step": 5069
+    },
+    {
+      "epoch": 2.35777398852891,
+      "grad_norm": 0.32174983620643616,
+      "learning_rate": 1.3307991682663463e-06,
+      "loss": 0.2728,
+      "step": 5070
+    },
+    {
+      "epoch": 2.358239032708107,
+      "grad_norm": 0.37295761704444885,
+      "learning_rate": 1.328961502805825e-06,
+      "loss": 0.3422,
+      "step": 5071
+    },
+    {
+      "epoch": 2.3587040768873044,
+      "grad_norm": 0.34429284930229187,
+      "learning_rate": 1.3271249125328512e-06,
+      "loss": 0.3188,
+      "step": 5072
+    },
+    {
+      "epoch": 2.3591691210665013,
+      "grad_norm": 0.32858747243881226,
+      "learning_rate": 1.3252893979853304e-06,
+      "loss": 0.2941,
+      "step": 5073
+    },
+    {
+      "epoch": 2.3596341652456982,
+      "grad_norm": 0.3379555344581604,
+      "learning_rate": 1.3234549597008572e-06,
+      "loss": 0.3626,
+      "step": 5074
+    },
+    {
+      "epoch": 2.360099209424895,
+      "grad_norm": 0.33183690905570984,
+      "learning_rate": 1.3216215982167064e-06,
+      "loss": 0.3313,
+      "step": 5075
+    },
+    {
+      "epoch": 2.3605642536040925,
+      "grad_norm": 0.386362761259079,
+      "learning_rate": 1.3197893140698426e-06,
+      "loss": 0.3743,
+      "step": 5076
+    },
+    {
+      "epoch": 2.3610292977832894,
+      "grad_norm": 0.3396303653717041,
+      "learning_rate": 1.3179581077969084e-06,
+      "loss": 0.2914,
+      "step": 5077
+    },
+    {
+      "epoch": 2.3614943419624863,
+      "grad_norm": 0.3387748599052429,
+      "learning_rate": 1.3161279799342385e-06,
+      "loss": 0.3185,
+      "step": 5078
+    },
+    {
+      "epoch": 2.3619593861416837,
+      "grad_norm": 0.32866546511650085,
+      "learning_rate": 1.314298931017844e-06,
+      "loss": 0.2987,
+      "step": 5079
+    },
+    {
+      "epoch": 2.3624244303208806,
+      "grad_norm": 0.3564012348651886,
+      "learning_rate": 1.3124709615834263e-06,
+      "loss": 0.3356,
+      "step": 5080
+    },
+    {
+      "epoch": 2.3628894745000775,
+      "grad_norm": 0.34756961464881897,
+      "learning_rate": 1.3106440721663655e-06,
+      "loss": 0.358,
+      "step": 5081
+    },
+    {
+      "epoch": 2.3633545186792744,
+      "grad_norm": 0.3577297627925873,
+      "learning_rate": 1.3088182633017294e-06,
+      "loss": 0.3194,
+      "step": 5082
+    },
+    {
+      "epoch": 2.3638195628584717,
+      "grad_norm": 0.3133348822593689,
+      "learning_rate": 1.306993535524269e-06,
+      "loss": 0.3099,
+      "step": 5083
+    },
+    {
+      "epoch": 2.3642846070376686,
+      "grad_norm": 0.3245254456996918,
+      "learning_rate": 1.3051698893684144e-06,
+      "loss": 0.3172,
+      "step": 5084
+    },
+    {
+      "epoch": 2.3647496512168655,
+      "grad_norm": 0.3281231224536896,
+      "learning_rate": 1.303347325368285e-06,
+      "loss": 0.3242,
+      "step": 5085
+    },
+    {
+      "epoch": 2.3652146953960624,
+      "grad_norm": 0.3328626751899719,
+      "learning_rate": 1.3015258440576767e-06,
+      "loss": 0.3048,
+      "step": 5086
+    },
+    {
+      "epoch": 2.36567973957526,
+      "grad_norm": 0.34870782494544983,
+      "learning_rate": 1.299705445970076e-06,
+      "loss": 0.2954,
+      "step": 5087
+    },
+    {
+      "epoch": 2.3661447837544567,
+      "grad_norm": 0.31751129031181335,
+      "learning_rate": 1.2978861316386437e-06,
+      "loss": 0.3195,
+      "step": 5088
+    },
+    {
+      "epoch": 2.3666098279336536,
+      "grad_norm": 0.3291054368019104,
+      "learning_rate": 1.2960679015962313e-06,
+      "loss": 0.3173,
+      "step": 5089
+    },
+    {
+      "epoch": 2.367074872112851,
+      "grad_norm": 0.34506654739379883,
+      "learning_rate": 1.2942507563753653e-06,
+      "loss": 0.3536,
+      "step": 5090
+    },
+    {
+      "epoch": 2.367539916292048,
+      "grad_norm": 0.3213541507720947,
+      "learning_rate": 1.2924346965082612e-06,
+      "loss": 0.3214,
+      "step": 5091
+    },
+    {
+      "epoch": 2.3680049604712448,
+      "grad_norm": 0.3615787625312805,
+      "learning_rate": 1.2906197225268108e-06,
+      "loss": 0.3476,
+      "step": 5092
+    },
+    {
+      "epoch": 2.3684700046504417,
+      "grad_norm": 0.32223793864250183,
+      "learning_rate": 1.288805834962591e-06,
+      "loss": 0.2795,
+      "step": 5093
+    },
+    {
+      "epoch": 2.3689350488296386,
+      "grad_norm": 0.357710599899292,
+      "learning_rate": 1.2869930343468611e-06,
+      "loss": 0.3424,
+      "step": 5094
+    },
+    {
+      "epoch": 2.369400093008836,
+      "grad_norm": 0.3196834921836853,
+      "learning_rate": 1.285181321210562e-06,
+      "loss": 0.2891,
+      "step": 5095
+    },
+    {
+      "epoch": 2.369865137188033,
+      "grad_norm": 0.33670881390571594,
+      "learning_rate": 1.2833706960843118e-06,
+      "loss": 0.3391,
+      "step": 5096
+    },
+    {
+      "epoch": 2.3703301813672297,
+      "grad_norm": 0.30696120858192444,
+      "learning_rate": 1.2815611594984162e-06,
+      "loss": 0.275,
+      "step": 5097
+    },
+    {
+      "epoch": 2.370795225546427,
+      "grad_norm": 0.35666152834892273,
+      "learning_rate": 1.2797527119828567e-06,
+      "loss": 0.342,
+      "step": 5098
+    },
+    {
+      "epoch": 2.371260269725624,
+      "grad_norm": 0.31616926193237305,
+      "learning_rate": 1.2779453540673009e-06,
+      "loss": 0.3172,
+      "step": 5099
+    },
+    {
+      "epoch": 2.371725313904821,
+      "grad_norm": 0.3312261700630188,
+      "learning_rate": 1.2761390862810907e-06,
+      "loss": 0.3278,
+      "step": 5100
+    },
+    {
+      "epoch": 2.372190358084018,
+      "grad_norm": 0.33016443252563477,
+      "learning_rate": 1.274333909153257e-06,
+      "loss": 0.3146,
+      "step": 5101
+    },
+    {
+      "epoch": 2.372655402263215,
+      "grad_norm": 0.34783631563186646,
+      "learning_rate": 1.2725298232125034e-06,
+      "loss": 0.3116,
+      "step": 5102
+    },
+    {
+      "epoch": 2.373120446442412,
+      "grad_norm": 0.34138286113739014,
+      "learning_rate": 1.27072682898722e-06,
+      "loss": 0.3093,
+      "step": 5103
+    },
+    {
+      "epoch": 2.373585490621609,
+      "grad_norm": 0.34735816717147827,
+      "learning_rate": 1.2689249270054716e-06,
+      "loss": 0.3073,
+      "step": 5104
+    },
+    {
+      "epoch": 2.374050534800806,
+      "grad_norm": 0.3784792721271515,
+      "learning_rate": 1.2671241177950078e-06,
+      "loss": 0.3441,
+      "step": 5105
+    },
+    {
+      "epoch": 2.374515578980003,
+      "grad_norm": 0.3438231945037842,
+      "learning_rate": 1.2653244018832562e-06,
+      "loss": 0.2984,
+      "step": 5106
+    },
+    {
+      "epoch": 2.3749806231592,
+      "grad_norm": 0.3264318108558655,
+      "learning_rate": 1.2635257797973255e-06,
+      "loss": 0.3278,
+      "step": 5107
+    },
+    {
+      "epoch": 2.375445667338397,
+      "grad_norm": 0.3333817422389984,
+      "learning_rate": 1.2617282520640007e-06,
+      "loss": 0.2979,
+      "step": 5108
+    },
+    {
+      "epoch": 2.3759107115175944,
+      "grad_norm": 0.37426552176475525,
+      "learning_rate": 1.2599318192097509e-06,
+      "loss": 0.3836,
+      "step": 5109
+    },
+    {
+      "epoch": 2.3763757556967913,
+      "grad_norm": 0.3461456596851349,
+      "learning_rate": 1.2581364817607194e-06,
+      "loss": 0.3214,
+      "step": 5110
+    },
+    {
+      "epoch": 2.376840799875988,
+      "grad_norm": 0.35083499550819397,
+      "learning_rate": 1.2563422402427339e-06,
+      "loss": 0.3083,
+      "step": 5111
+    },
+    {
+      "epoch": 2.377305844055185,
+      "grad_norm": 0.3530008792877197,
+      "learning_rate": 1.254549095181296e-06,
+      "loss": 0.2944,
+      "step": 5112
+    },
+    {
+      "epoch": 2.3777708882343824,
+      "grad_norm": 0.3381586968898773,
+      "learning_rate": 1.2527570471015915e-06,
+      "loss": 0.3435,
+      "step": 5113
+    },
+    {
+      "epoch": 2.3782359324135793,
+      "grad_norm": 0.3386650085449219,
+      "learning_rate": 1.2509660965284797e-06,
+      "loss": 0.3303,
+      "step": 5114
+    },
+    {
+      "epoch": 2.3787009765927762,
+      "grad_norm": 0.333477646112442,
+      "learning_rate": 1.2491762439865034e-06,
+      "loss": 0.3108,
+      "step": 5115
+    },
+    {
+      "epoch": 2.379166020771973,
+      "grad_norm": 0.32221055030822754,
+      "learning_rate": 1.247387489999879e-06,
+      "loss": 0.307,
+      "step": 5116
+    },
+    {
+      "epoch": 2.3796310649511705,
+      "grad_norm": 0.35751134157180786,
+      "learning_rate": 1.2455998350925042e-06,
+      "loss": 0.3388,
+      "step": 5117
+    },
+    {
+      "epoch": 2.3800961091303674,
+      "grad_norm": 0.3639261722564697,
+      "learning_rate": 1.2438132797879554e-06,
+      "loss": 0.348,
+      "step": 5118
+    },
+    {
+      "epoch": 2.3805611533095643,
+      "grad_norm": 0.31775328516960144,
+      "learning_rate": 1.2420278246094835e-06,
+      "loss": 0.2986,
+      "step": 5119
+    },
+    {
+      "epoch": 2.3810261974887617,
+      "grad_norm": 0.36067667603492737,
+      "learning_rate": 1.240243470080022e-06,
+      "loss": 0.3553,
+      "step": 5120
+    },
+    {
+      "epoch": 2.3814912416679586,
+      "grad_norm": 0.31480494141578674,
+      "learning_rate": 1.2384602167221765e-06,
+      "loss": 0.2935,
+      "step": 5121
+    },
+    {
+      "epoch": 2.3819562858471555,
+      "grad_norm": 0.30289679765701294,
+      "learning_rate": 1.2366780650582355e-06,
+      "loss": 0.2969,
+      "step": 5122
+    },
+    {
+      "epoch": 2.3824213300263524,
+      "grad_norm": 0.3624131977558136,
+      "learning_rate": 1.2348970156101592e-06,
+      "loss": 0.3583,
+      "step": 5123
+    },
+    {
+      "epoch": 2.3828863742055497,
+      "grad_norm": 0.3010319769382477,
+      "learning_rate": 1.233117068899592e-06,
+      "loss": 0.2917,
+      "step": 5124
+    },
+    {
+      "epoch": 2.3833514183847466,
+      "grad_norm": 0.3690085709095001,
+      "learning_rate": 1.2313382254478473e-06,
+      "loss": 0.3563,
+      "step": 5125
+    },
+    {
+      "epoch": 2.3838164625639435,
+      "grad_norm": 0.33377185463905334,
+      "learning_rate": 1.229560485775923e-06,
+      "loss": 0.3258,
+      "step": 5126
+    },
+    {
+      "epoch": 2.3842815067431404,
+      "grad_norm": 0.31576281785964966,
+      "learning_rate": 1.227783850404487e-06,
+      "loss": 0.3138,
+      "step": 5127
+    },
+    {
+      "epoch": 2.384746550922338,
+      "grad_norm": 0.3261266350746155,
+      "learning_rate": 1.2260083198538886e-06,
+      "loss": 0.3253,
+      "step": 5128
+    },
+    {
+      "epoch": 2.3852115951015347,
+      "grad_norm": 0.34286534786224365,
+      "learning_rate": 1.2242338946441518e-06,
+      "loss": 0.357,
+      "step": 5129
+    },
+    {
+      "epoch": 2.3856766392807316,
+      "grad_norm": 0.3160213530063629,
+      "learning_rate": 1.2224605752949786e-06,
+      "loss": 0.3051,
+      "step": 5130
+    },
+    {
+      "epoch": 2.386141683459929,
+      "grad_norm": 0.34241053462028503,
+      "learning_rate": 1.2206883623257421e-06,
+      "loss": 0.3066,
+      "step": 5131
+    },
+    {
+      "epoch": 2.386606727639126,
+      "grad_norm": 0.34416183829307556,
+      "learning_rate": 1.2189172562554973e-06,
+      "loss": 0.3326,
+      "step": 5132
+    },
+    {
+      "epoch": 2.3870717718183228,
+      "grad_norm": 0.3707297742366791,
+      "learning_rate": 1.2171472576029707e-06,
+      "loss": 0.3504,
+      "step": 5133
+    },
+    {
+      "epoch": 2.3875368159975197,
+      "grad_norm": 0.31945255398750305,
+      "learning_rate": 1.2153783668865681e-06,
+      "loss": 0.3035,
+      "step": 5134
+    },
+    {
+      "epoch": 2.3880018601767166,
+      "grad_norm": 0.33111390471458435,
+      "learning_rate": 1.2136105846243662e-06,
+      "loss": 0.3256,
+      "step": 5135
+    },
+    {
+      "epoch": 2.388466904355914,
+      "grad_norm": 0.3572760224342346,
+      "learning_rate": 1.2118439113341224e-06,
+      "loss": 0.3382,
+      "step": 5136
+    },
+    {
+      "epoch": 2.388931948535111,
+      "grad_norm": 0.32829439640045166,
+      "learning_rate": 1.210078347533264e-06,
+      "loss": 0.3172,
+      "step": 5137
+    },
+    {
+      "epoch": 2.3893969927143077,
+      "grad_norm": 0.3324892818927765,
+      "learning_rate": 1.2083138937388989e-06,
+      "loss": 0.3209,
+      "step": 5138
+    },
+    {
+      "epoch": 2.389862036893505,
+      "grad_norm": 0.31397631764411926,
+      "learning_rate": 1.2065505504678038e-06,
+      "loss": 0.3149,
+      "step": 5139
+    },
+    {
+      "epoch": 2.390327081072702,
+      "grad_norm": 0.3297927677631378,
+      "learning_rate": 1.2047883182364351e-06,
+      "loss": 0.2948,
+      "step": 5140
+    },
+    {
+      "epoch": 2.390792125251899,
+      "grad_norm": 0.33224743604660034,
+      "learning_rate": 1.2030271975609214e-06,
+      "loss": 0.3089,
+      "step": 5141
+    },
+    {
+      "epoch": 2.391257169431096,
+      "grad_norm": 0.37292471528053284,
+      "learning_rate": 1.2012671889570683e-06,
+      "loss": 0.3241,
+      "step": 5142
+    },
+    {
+      "epoch": 2.391722213610293,
+      "grad_norm": 0.3111366927623749,
+      "learning_rate": 1.1995082929403507e-06,
+      "loss": 0.2858,
+      "step": 5143
+    },
+    {
+      "epoch": 2.39218725778949,
+      "grad_norm": 0.37733349204063416,
+      "learning_rate": 1.1977505100259235e-06,
+      "loss": 0.3593,
+      "step": 5144
+    },
+    {
+      "epoch": 2.392652301968687,
+      "grad_norm": 0.36650681495666504,
+      "learning_rate": 1.1959938407286099e-06,
+      "loss": 0.3479,
+      "step": 5145
+    },
+    {
+      "epoch": 2.393117346147884,
+      "grad_norm": 0.31650200486183167,
+      "learning_rate": 1.1942382855629131e-06,
+      "loss": 0.2705,
+      "step": 5146
+    },
+    {
+      "epoch": 2.393582390327081,
+      "grad_norm": 0.33961057662963867,
+      "learning_rate": 1.1924838450430032e-06,
+      "loss": 0.3244,
+      "step": 5147
+    },
+    {
+      "epoch": 2.394047434506278,
+      "grad_norm": 0.31274569034576416,
+      "learning_rate": 1.19073051968273e-06,
+      "loss": 0.2908,
+      "step": 5148
+    },
+    {
+      "epoch": 2.394512478685475,
+      "grad_norm": 0.32725125551223755,
+      "learning_rate": 1.188978309995612e-06,
+      "loss": 0.311,
+      "step": 5149
+    },
+    {
+      "epoch": 2.3949775228646724,
+      "grad_norm": 0.3675840198993683,
+      "learning_rate": 1.1872272164948456e-06,
+      "loss": 0.356,
+      "step": 5150
+    },
+    {
+      "epoch": 2.3954425670438693,
+      "grad_norm": 0.3493657410144806,
+      "learning_rate": 1.1854772396932946e-06,
+      "loss": 0.3292,
+      "step": 5151
+    },
+    {
+      "epoch": 2.395907611223066,
+      "grad_norm": 0.344338595867157,
+      "learning_rate": 1.1837283801034998e-06,
+      "loss": 0.3117,
+      "step": 5152
+    },
+    {
+      "epoch": 2.396372655402263,
+      "grad_norm": 0.344719260931015,
+      "learning_rate": 1.181980638237676e-06,
+      "loss": 0.3277,
+      "step": 5153
+    },
+    {
+      "epoch": 2.3968376995814604,
+      "grad_norm": 0.32095858454704285,
+      "learning_rate": 1.1802340146077045e-06,
+      "loss": 0.2881,
+      "step": 5154
+    },
+    {
+      "epoch": 2.3973027437606573,
+      "grad_norm": 0.3079981803894043,
+      "learning_rate": 1.1784885097251474e-06,
+      "loss": 0.3379,
+      "step": 5155
+    },
+    {
+      "epoch": 2.3977677879398542,
+      "grad_norm": 0.33286038041114807,
+      "learning_rate": 1.1767441241012307e-06,
+      "loss": 0.3374,
+      "step": 5156
+    },
+    {
+      "epoch": 2.398232832119051,
+      "grad_norm": 0.36355435848236084,
+      "learning_rate": 1.1750008582468592e-06,
+      "loss": 0.3364,
+      "step": 5157
+    },
+    {
+      "epoch": 2.3986978762982485,
+      "grad_norm": 0.3209869861602783,
+      "learning_rate": 1.1732587126726054e-06,
+      "loss": 0.3159,
+      "step": 5158
+    },
+    {
+      "epoch": 2.3991629204774454,
+      "grad_norm": 0.32692816853523254,
+      "learning_rate": 1.1715176878887174e-06,
+      "loss": 0.3143,
+      "step": 5159
+    },
+    {
+      "epoch": 2.3996279646566423,
+      "grad_norm": 0.345546692609787,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.3354,
+      "step": 5160
+    },
+    {
+      "epoch": 2.4000930088358396,
+      "grad_norm": 0.3152925670146942,
+      "learning_rate": 1.168039002731377e-06,
+      "loss": 0.3035,
+      "step": 5161
+    },
+    {
+      "epoch": 2.4005580530150366,
+      "grad_norm": 0.379638671875,
+      "learning_rate": 1.1663013433767756e-06,
+      "loss": 0.3057,
+      "step": 5162
+    },
+    {
+      "epoch": 2.4010230971942335,
+      "grad_norm": 0.31644484400749207,
+      "learning_rate": 1.164564806850239e-06,
+      "loss": 0.3047,
+      "step": 5163
+    },
+    {
+      "epoch": 2.4014881413734304,
+      "grad_norm": 0.34637710452079773,
+      "learning_rate": 1.1628293936603707e-06,
+      "loss": 0.3396,
+      "step": 5164
+    },
+    {
+      "epoch": 2.4019531855526273,
+      "grad_norm": 0.3282802700996399,
+      "learning_rate": 1.1610951043154472e-06,
+      "loss": 0.337,
+      "step": 5165
+    },
+    {
+      "epoch": 2.4024182297318246,
+      "grad_norm": 0.30934378504753113,
+      "learning_rate": 1.1593619393234096e-06,
+      "loss": 0.3007,
+      "step": 5166
+    },
+    {
+      "epoch": 2.4028832739110215,
+      "grad_norm": 0.33991992473602295,
+      "learning_rate": 1.1576298991918778e-06,
+      "loss": 0.3462,
+      "step": 5167
+    },
+    {
+      "epoch": 2.4033483180902184,
+      "grad_norm": 0.30924656987190247,
+      "learning_rate": 1.1558989844281349e-06,
+      "loss": 0.3036,
+      "step": 5168
+    },
+    {
+      "epoch": 2.4038133622694158,
+      "grad_norm": 0.341217577457428,
+      "learning_rate": 1.1541691955391403e-06,
+      "loss": 0.3183,
+      "step": 5169
+    },
+    {
+      "epoch": 2.4042784064486127,
+      "grad_norm": 0.30020660161972046,
+      "learning_rate": 1.1524405330315187e-06,
+      "loss": 0.3064,
+      "step": 5170
+    },
+    {
+      "epoch": 2.4047434506278096,
+      "grad_norm": 0.3430638611316681,
+      "learning_rate": 1.15071299741157e-06,
+      "loss": 0.3619,
+      "step": 5171
+    },
+    {
+      "epoch": 2.4052084948070065,
+      "grad_norm": 0.33357352018356323,
+      "learning_rate": 1.148986589185258e-06,
+      "loss": 0.3348,
+      "step": 5172
+    },
+    {
+      "epoch": 2.405673538986204,
+      "grad_norm": 0.34539151191711426,
+      "learning_rate": 1.147261308858223e-06,
+      "loss": 0.3226,
+      "step": 5173
+    },
+    {
+      "epoch": 2.4061385831654007,
+      "grad_norm": 0.34261536598205566,
+      "learning_rate": 1.145537156935768e-06,
+      "loss": 0.3085,
+      "step": 5174
+    },
+    {
+      "epoch": 2.4066036273445977,
+      "grad_norm": 0.3475549817085266,
+      "learning_rate": 1.143814133922872e-06,
+      "loss": 0.326,
+      "step": 5175
+    },
+    {
+      "epoch": 2.4070686715237946,
+      "grad_norm": 0.3175617754459381,
+      "learning_rate": 1.142092240324179e-06,
+      "loss": 0.3038,
+      "step": 5176
+    },
+    {
+      "epoch": 2.407533715702992,
+      "grad_norm": 0.33959630131721497,
+      "learning_rate": 1.1403714766440061e-06,
+      "loss": 0.3148,
+      "step": 5177
+    },
+    {
+      "epoch": 2.407998759882189,
+      "grad_norm": 0.34172576665878296,
+      "learning_rate": 1.1386518433863331e-06,
+      "loss": 0.3174,
+      "step": 5178
+    },
+    {
+      "epoch": 2.4084638040613857,
+      "grad_norm": 0.30820226669311523,
+      "learning_rate": 1.1369333410548166e-06,
+      "loss": 0.2975,
+      "step": 5179
+    },
+    {
+      "epoch": 2.408928848240583,
+      "grad_norm": 0.3713323473930359,
+      "learning_rate": 1.1352159701527743e-06,
+      "loss": 0.3735,
+      "step": 5180
+    },
+    {
+      "epoch": 2.40939389241978,
+      "grad_norm": 0.3337565064430237,
+      "learning_rate": 1.1334997311832003e-06,
+      "loss": 0.2932,
+      "step": 5181
+    },
+    {
+      "epoch": 2.409858936598977,
+      "grad_norm": 0.35603073239326477,
+      "learning_rate": 1.1317846246487485e-06,
+      "loss": 0.2922,
+      "step": 5182
+    },
+    {
+      "epoch": 2.410323980778174,
+      "grad_norm": 0.34648483991622925,
+      "learning_rate": 1.13007065105175e-06,
+      "loss": 0.3224,
+      "step": 5183
+    },
+    {
+      "epoch": 2.410789024957371,
+      "grad_norm": 0.3309410810470581,
+      "learning_rate": 1.128357810894196e-06,
+      "loss": 0.3081,
+      "step": 5184
+    },
+    {
+      "epoch": 2.411254069136568,
+      "grad_norm": 0.3483981788158417,
+      "learning_rate": 1.1266461046777537e-06,
+      "loss": 0.3221,
+      "step": 5185
+    },
+    {
+      "epoch": 2.411719113315765,
+      "grad_norm": 0.3361232280731201,
+      "learning_rate": 1.1249355329037498e-06,
+      "loss": 0.3369,
+      "step": 5186
+    },
+    {
+      "epoch": 2.412184157494962,
+      "grad_norm": 0.30131569504737854,
+      "learning_rate": 1.1232260960731855e-06,
+      "loss": 0.2907,
+      "step": 5187
+    },
+    {
+      "epoch": 2.412649201674159,
+      "grad_norm": 0.3845626711845398,
+      "learning_rate": 1.1215177946867262e-06,
+      "loss": 0.3715,
+      "step": 5188
+    },
+    {
+      "epoch": 2.413114245853356,
+      "grad_norm": 0.3543691337108612,
+      "learning_rate": 1.1198106292447076e-06,
+      "loss": 0.3494,
+      "step": 5189
+    },
+    {
+      "epoch": 2.413579290032553,
+      "grad_norm": 0.4014170467853546,
+      "learning_rate": 1.1181046002471292e-06,
+      "loss": 0.3547,
+      "step": 5190
+    },
+    {
+      "epoch": 2.4140443342117504,
+      "grad_norm": 0.3230542540550232,
+      "learning_rate": 1.1163997081936578e-06,
+      "loss": 0.2861,
+      "step": 5191
+    },
+    {
+      "epoch": 2.4145093783909473,
+      "grad_norm": 0.31293249130249023,
+      "learning_rate": 1.1146959535836317e-06,
+      "loss": 0.2862,
+      "step": 5192
+    },
+    {
+      "epoch": 2.414974422570144,
+      "grad_norm": 0.3535180389881134,
+      "learning_rate": 1.112993336916049e-06,
+      "loss": 0.3462,
+      "step": 5193
+    },
+    {
+      "epoch": 2.415439466749341,
+      "grad_norm": 0.32987791299819946,
+      "learning_rate": 1.1112918586895826e-06,
+      "loss": 0.3061,
+      "step": 5194
+    },
+    {
+      "epoch": 2.415904510928538,
+      "grad_norm": 0.3237574100494385,
+      "learning_rate": 1.1095915194025642e-06,
+      "loss": 0.2999,
+      "step": 5195
+    },
+    {
+      "epoch": 2.4163695551077353,
+      "grad_norm": 0.35278692841529846,
+      "learning_rate": 1.1078923195529973e-06,
+      "loss": 0.3329,
+      "step": 5196
+    },
+    {
+      "epoch": 2.4168345992869322,
+      "grad_norm": 0.34075209498405457,
+      "learning_rate": 1.1061942596385516e-06,
+      "loss": 0.3127,
+      "step": 5197
+    },
+    {
+      "epoch": 2.417299643466129,
+      "grad_norm": 0.3517972230911255,
+      "learning_rate": 1.1044973401565578e-06,
+      "loss": 0.3275,
+      "step": 5198
+    },
+    {
+      "epoch": 2.4177646876453265,
+      "grad_norm": 0.4040515720844269,
+      "learning_rate": 1.1028015616040182e-06,
+      "loss": 0.3465,
+      "step": 5199
+    },
+    {
+      "epoch": 2.4182297318245234,
+      "grad_norm": 0.3578769266605377,
+      "learning_rate": 1.1011069244775996e-06,
+      "loss": 0.3302,
+      "step": 5200
+    },
+    {
+      "epoch": 2.4186947760037203,
+      "grad_norm": 0.3215552866458893,
+      "learning_rate": 1.0994134292736307e-06,
+      "loss": 0.2755,
+      "step": 5201
+    },
+    {
+      "epoch": 2.419159820182917,
+      "grad_norm": 0.3483463227748871,
+      "learning_rate": 1.0977210764881124e-06,
+      "loss": 0.3649,
+      "step": 5202
+    },
+    {
+      "epoch": 2.4196248643621145,
+      "grad_norm": 0.32858699560165405,
+      "learning_rate": 1.096029866616704e-06,
+      "loss": 0.3099,
+      "step": 5203
+    },
+    {
+      "epoch": 2.4200899085413115,
+      "grad_norm": 0.34919896721839905,
+      "learning_rate": 1.0943398001547362e-06,
+      "loss": 0.3519,
+      "step": 5204
+    },
+    {
+      "epoch": 2.4205549527205084,
+      "grad_norm": 0.3417411148548126,
+      "learning_rate": 1.0926508775971995e-06,
+      "loss": 0.301,
+      "step": 5205
+    },
+    {
+      "epoch": 2.4210199968997053,
+      "grad_norm": 0.3448435068130493,
+      "learning_rate": 1.0909630994387538e-06,
+      "loss": 0.319,
+      "step": 5206
+    },
+    {
+      "epoch": 2.4214850410789026,
+      "grad_norm": 0.31199708580970764,
+      "learning_rate": 1.0892764661737204e-06,
+      "loss": 0.2935,
+      "step": 5207
+    },
+    {
+      "epoch": 2.4219500852580995,
+      "grad_norm": 0.33110520243644714,
+      "learning_rate": 1.0875909782960887e-06,
+      "loss": 0.3122,
+      "step": 5208
+    },
+    {
+      "epoch": 2.4224151294372964,
+      "grad_norm": 0.3433871865272522,
+      "learning_rate": 1.0859066362995085e-06,
+      "loss": 0.2962,
+      "step": 5209
+    },
+    {
+      "epoch": 2.4228801736164938,
+      "grad_norm": 0.34019917249679565,
+      "learning_rate": 1.0842234406772973e-06,
+      "loss": 0.3437,
+      "step": 5210
+    },
+    {
+      "epoch": 2.4233452177956907,
+      "grad_norm": 0.3699561655521393,
+      "learning_rate": 1.0825413919224353e-06,
+      "loss": 0.3118,
+      "step": 5211
+    },
+    {
+      "epoch": 2.4238102619748876,
+      "grad_norm": 0.364387184381485,
+      "learning_rate": 1.0808604905275693e-06,
+      "loss": 0.3258,
+      "step": 5212
+    },
+    {
+      "epoch": 2.4242753061540845,
+      "grad_norm": 0.35317301750183105,
+      "learning_rate": 1.0791807369850048e-06,
+      "loss": 0.3585,
+      "step": 5213
+    },
+    {
+      "epoch": 2.424740350333282,
+      "grad_norm": 0.341281533241272,
+      "learning_rate": 1.077502131786718e-06,
+      "loss": 0.3319,
+      "step": 5214
+    },
+    {
+      "epoch": 2.4252053945124787,
+      "grad_norm": 0.3004712164402008,
+      "learning_rate": 1.0758246754243412e-06,
+      "loss": 0.2823,
+      "step": 5215
+    },
+    {
+      "epoch": 2.4256704386916756,
+      "grad_norm": 0.34995582699775696,
+      "learning_rate": 1.0741483683891773e-06,
+      "loss": 0.2991,
+      "step": 5216
+    },
+    {
+      "epoch": 2.4261354828708726,
+      "grad_norm": 0.3658967912197113,
+      "learning_rate": 1.072473211172187e-06,
+      "loss": 0.3243,
+      "step": 5217
+    },
+    {
+      "epoch": 2.42660052705007,
+      "grad_norm": 0.3104728162288666,
+      "learning_rate": 1.0707992042639986e-06,
+      "loss": 0.3104,
+      "step": 5218
+    },
+    {
+      "epoch": 2.427065571229267,
+      "grad_norm": 0.3573016822338104,
+      "learning_rate": 1.0691263481548996e-06,
+      "loss": 0.3329,
+      "step": 5219
+    },
+    {
+      "epoch": 2.4275306154084637,
+      "grad_norm": 0.31468331813812256,
+      "learning_rate": 1.0674546433348453e-06,
+      "loss": 0.2965,
+      "step": 5220
+    },
+    {
+      "epoch": 2.427995659587661,
+      "grad_norm": 0.3958417773246765,
+      "learning_rate": 1.0657840902934469e-06,
+      "loss": 0.3601,
+      "step": 5221
+    },
+    {
+      "epoch": 2.428460703766858,
+      "grad_norm": 0.3233220875263214,
+      "learning_rate": 1.064114689519985e-06,
+      "loss": 0.3139,
+      "step": 5222
+    },
+    {
+      "epoch": 2.428925747946055,
+      "grad_norm": 0.3252237141132355,
+      "learning_rate": 1.0624464415033987e-06,
+      "loss": 0.3017,
+      "step": 5223
+    },
+    {
+      "epoch": 2.4293907921252518,
+      "grad_norm": 0.32925140857696533,
+      "learning_rate": 1.060779346732293e-06,
+      "loss": 0.3318,
+      "step": 5224
+    },
+    {
+      "epoch": 2.4298558363044487,
+      "grad_norm": 0.3464730978012085,
+      "learning_rate": 1.0591134056949314e-06,
+      "loss": 0.3449,
+      "step": 5225
+    },
+    {
+      "epoch": 2.430320880483646,
+      "grad_norm": 0.3610069155693054,
+      "learning_rate": 1.0574486188792393e-06,
+      "loss": 0.3089,
+      "step": 5226
+    },
+    {
+      "epoch": 2.430785924662843,
+      "grad_norm": 0.33958742022514343,
+      "learning_rate": 1.0557849867728088e-06,
+      "loss": 0.3386,
+      "step": 5227
+    },
+    {
+      "epoch": 2.43125096884204,
+      "grad_norm": 0.3290470540523529,
+      "learning_rate": 1.0541225098628877e-06,
+      "loss": 0.3015,
+      "step": 5228
+    },
+    {
+      "epoch": 2.431716013021237,
+      "grad_norm": 0.32628098130226135,
+      "learning_rate": 1.0524611886363912e-06,
+      "loss": 0.3187,
+      "step": 5229
+    },
+    {
+      "epoch": 2.432181057200434,
+      "grad_norm": 0.33828026056289673,
+      "learning_rate": 1.0508010235798904e-06,
+      "loss": 0.3105,
+      "step": 5230
+    },
+    {
+      "epoch": 2.432646101379631,
+      "grad_norm": 0.33995717763900757,
+      "learning_rate": 1.0491420151796227e-06,
+      "loss": 0.3409,
+      "step": 5231
+    },
+    {
+      "epoch": 2.433111145558828,
+      "grad_norm": 0.30739811062812805,
+      "learning_rate": 1.047484163921486e-06,
+      "loss": 0.3109,
+      "step": 5232
+    },
+    {
+      "epoch": 2.4335761897380253,
+      "grad_norm": 0.32465070486068726,
+      "learning_rate": 1.0458274702910347e-06,
+      "loss": 0.299,
+      "step": 5233
+    },
+    {
+      "epoch": 2.434041233917222,
+      "grad_norm": 0.36253416538238525,
+      "learning_rate": 1.044171934773489e-06,
+      "loss": 0.3325,
+      "step": 5234
+    },
+    {
+      "epoch": 2.434506278096419,
+      "grad_norm": 0.31193816661834717,
+      "learning_rate": 1.04251755785373e-06,
+      "loss": 0.3138,
+      "step": 5235
+    },
+    {
+      "epoch": 2.434971322275616,
+      "grad_norm": 0.3374314308166504,
+      "learning_rate": 1.0408643400162949e-06,
+      "loss": 0.3161,
+      "step": 5236
+    },
+    {
+      "epoch": 2.4354363664548133,
+      "grad_norm": 0.34814929962158203,
+      "learning_rate": 1.039212281745387e-06,
+      "loss": 0.3192,
+      "step": 5237
+    },
+    {
+      "epoch": 2.4359014106340102,
+      "grad_norm": 0.3480595648288727,
+      "learning_rate": 1.0375613835248648e-06,
+      "loss": 0.3186,
+      "step": 5238
+    },
+    {
+      "epoch": 2.436366454813207,
+      "grad_norm": 0.35934314131736755,
+      "learning_rate": 1.0359116458382523e-06,
+      "loss": 0.3206,
+      "step": 5239
+    },
+    {
+      "epoch": 2.4368314989924045,
+      "grad_norm": 0.34923362731933594,
+      "learning_rate": 1.0342630691687283e-06,
+      "loss": 0.3336,
+      "step": 5240
+    },
+    {
+      "epoch": 2.4372965431716014,
+      "grad_norm": 0.3344292640686035,
+      "learning_rate": 1.0326156539991361e-06,
+      "loss": 0.329,
+      "step": 5241
+    },
+    {
+      "epoch": 2.4377615873507983,
+      "grad_norm": 0.34684503078460693,
+      "learning_rate": 1.0309694008119748e-06,
+      "loss": 0.3105,
+      "step": 5242
+    },
+    {
+      "epoch": 2.438226631529995,
+      "grad_norm": 0.32323524355888367,
+      "learning_rate": 1.0293243100894068e-06,
+      "loss": 0.3101,
+      "step": 5243
+    },
+    {
+      "epoch": 2.4386916757091925,
+      "grad_norm": 0.3490697741508484,
+      "learning_rate": 1.027680382313253e-06,
+      "loss": 0.3156,
+      "step": 5244
+    },
+    {
+      "epoch": 2.4391567198883894,
+      "grad_norm": 0.33271536231040955,
+      "learning_rate": 1.0260376179649905e-06,
+      "loss": 0.3069,
+      "step": 5245
+    },
+    {
+      "epoch": 2.4396217640675864,
+      "grad_norm": 0.31948867440223694,
+      "learning_rate": 1.0243960175257605e-06,
+      "loss": 0.3263,
+      "step": 5246
+    },
+    {
+      "epoch": 2.4400868082467833,
+      "grad_norm": 0.3256017863750458,
+      "learning_rate": 1.0227555814763623e-06,
+      "loss": 0.3457,
+      "step": 5247
+    },
+    {
+      "epoch": 2.4405518524259806,
+      "grad_norm": 0.31808292865753174,
+      "learning_rate": 1.0211163102972494e-06,
+      "loss": 0.3109,
+      "step": 5248
+    },
+    {
+      "epoch": 2.4410168966051775,
+      "grad_norm": 0.33860480785369873,
+      "learning_rate": 1.0194782044685414e-06,
+      "loss": 0.3262,
+      "step": 5249
+    },
+    {
+      "epoch": 2.4414819407843744,
+      "grad_norm": 0.3586934506893158,
+      "learning_rate": 1.0178412644700093e-06,
+      "loss": 0.3337,
+      "step": 5250
+    },
+    {
+      "epoch": 2.4419469849635718,
+      "grad_norm": 0.31557610630989075,
+      "learning_rate": 1.0162054907810892e-06,
+      "loss": 0.2812,
+      "step": 5251
+    },
+    {
+      "epoch": 2.4424120291427687,
+      "grad_norm": 0.3515093922615051,
+      "learning_rate": 1.0145708838808704e-06,
+      "loss": 0.3759,
+      "step": 5252
+    },
+    {
+      "epoch": 2.4428770733219656,
+      "grad_norm": 0.3308696448802948,
+      "learning_rate": 1.012937444248105e-06,
+      "loss": 0.2839,
+      "step": 5253
+    },
+    {
+      "epoch": 2.4433421175011625,
+      "grad_norm": 0.35928985476493835,
+      "learning_rate": 1.0113051723611989e-06,
+      "loss": 0.3185,
+      "step": 5254
+    },
+    {
+      "epoch": 2.4438071616803594,
+      "grad_norm": 0.3398415744304657,
+      "learning_rate": 1.0096740686982192e-06,
+      "loss": 0.295,
+      "step": 5255
+    },
+    {
+      "epoch": 2.4442722058595567,
+      "grad_norm": 0.3303294777870178,
+      "learning_rate": 1.0080441337368884e-06,
+      "loss": 0.3409,
+      "step": 5256
+    },
+    {
+      "epoch": 2.4447372500387536,
+      "grad_norm": 0.33345457911491394,
+      "learning_rate": 1.0064153679545891e-06,
+      "loss": 0.3565,
+      "step": 5257
+    },
+    {
+      "epoch": 2.4452022942179505,
+      "grad_norm": 0.30849790573120117,
+      "learning_rate": 1.00478777182836e-06,
+      "loss": 0.3075,
+      "step": 5258
+    },
+    {
+      "epoch": 2.445667338397148,
+      "grad_norm": 0.3234153091907501,
+      "learning_rate": 1.0031613458348988e-06,
+      "loss": 0.2985,
+      "step": 5259
+    },
+    {
+      "epoch": 2.446132382576345,
+      "grad_norm": 0.3142699897289276,
+      "learning_rate": 1.0015360904505573e-06,
+      "loss": 0.2921,
+      "step": 5260
+    },
+    {
+      "epoch": 2.4465974267555417,
+      "grad_norm": 0.36145663261413574,
+      "learning_rate": 9.99912006151348e-07,
+      "loss": 0.3293,
+      "step": 5261
+    },
+    {
+      "epoch": 2.4470624709347386,
+      "grad_norm": 0.354502409696579,
+      "learning_rate": 9.98289093412938e-07,
+      "loss": 0.3246,
+      "step": 5262
+    },
+    {
+      "epoch": 2.447527515113936,
+      "grad_norm": 0.36151012778282166,
+      "learning_rate": 9.966673527106514e-07,
+      "loss": 0.3153,
+      "step": 5263
+    },
+    {
+      "epoch": 2.447992559293133,
+      "grad_norm": 0.3437816798686981,
+      "learning_rate": 9.950467845194712e-07,
+      "loss": 0.3397,
+      "step": 5264
+    },
+    {
+      "epoch": 2.4484576034723298,
+      "grad_norm": 0.335560142993927,
+      "learning_rate": 9.934273893140335e-07,
+      "loss": 0.3223,
+      "step": 5265
+    },
+    {
+      "epoch": 2.4489226476515267,
+      "grad_norm": 0.33611413836479187,
+      "learning_rate": 9.918091675686343e-07,
+      "loss": 0.3288,
+      "step": 5266
+    },
+    {
+      "epoch": 2.449387691830724,
+      "grad_norm": 0.3385131359100342,
+      "learning_rate": 9.90192119757225e-07,
+      "loss": 0.3153,
+      "step": 5267
+    },
+    {
+      "epoch": 2.449852736009921,
+      "grad_norm": 0.35654857754707336,
+      "learning_rate": 9.88576246353411e-07,
+      "loss": 0.2852,
+      "step": 5268
+    },
+    {
+      "epoch": 2.450317780189118,
+      "grad_norm": 0.3671344220638275,
+      "learning_rate": 9.869615478304567e-07,
+      "loss": 0.3212,
+      "step": 5269
+    },
+    {
+      "epoch": 2.450782824368315,
+      "grad_norm": 0.33951297402381897,
+      "learning_rate": 9.853480246612812e-07,
+      "loss": 0.3201,
+      "step": 5270
+    },
+    {
+      "epoch": 2.451247868547512,
+      "grad_norm": 0.34729960560798645,
+      "learning_rate": 9.837356773184576e-07,
+      "loss": 0.3244,
+      "step": 5271
+    },
+    {
+      "epoch": 2.451712912726709,
+      "grad_norm": 0.3578476011753082,
+      "learning_rate": 9.821245062742191e-07,
+      "loss": 0.2941,
+      "step": 5272
+    },
+    {
+      "epoch": 2.452177956905906,
+      "grad_norm": 0.3414701521396637,
+      "learning_rate": 9.805145120004478e-07,
+      "loss": 0.3043,
+      "step": 5273
+    },
+    {
+      "epoch": 2.4526430010851032,
+      "grad_norm": 0.30916067957878113,
+      "learning_rate": 9.789056949686882e-07,
+      "loss": 0.299,
+      "step": 5274
+    },
+    {
+      "epoch": 2.4531080452643,
+      "grad_norm": 0.3154885470867157,
+      "learning_rate": 9.772980556501338e-07,
+      "loss": 0.313,
+      "step": 5275
+    },
+    {
+      "epoch": 2.453573089443497,
+      "grad_norm": 0.3455050587654114,
+      "learning_rate": 9.756915945156392e-07,
+      "loss": 0.3221,
+      "step": 5276
+    },
+    {
+      "epoch": 2.454038133622694,
+      "grad_norm": 0.3348415195941925,
+      "learning_rate": 9.74086312035708e-07,
+      "loss": 0.312,
+      "step": 5277
+    },
+    {
+      "epoch": 2.4545031778018913,
+      "grad_norm": 0.34347227215766907,
+      "learning_rate": 9.724822086805019e-07,
+      "loss": 0.3398,
+      "step": 5278
+    },
+    {
+      "epoch": 2.454968221981088,
+      "grad_norm": 0.340069055557251,
+      "learning_rate": 9.70879284919839e-07,
+      "loss": 0.3009,
+      "step": 5279
+    },
+    {
+      "epoch": 2.455433266160285,
+      "grad_norm": 0.3648471236228943,
+      "learning_rate": 9.692775412231863e-07,
+      "loss": 0.3182,
+      "step": 5280
+    },
+    {
+      "epoch": 2.4558983103394825,
+      "grad_norm": 0.33610066771507263,
+      "learning_rate": 9.6767697805967e-07,
+      "loss": 0.303,
+      "step": 5281
+    },
+    {
+      "epoch": 2.4563633545186794,
+      "grad_norm": 0.3398694694042206,
+      "learning_rate": 9.660775958980712e-07,
+      "loss": 0.3386,
+      "step": 5282
+    },
+    {
+      "epoch": 2.4568283986978763,
+      "grad_norm": 0.32278192043304443,
+      "learning_rate": 9.644793952068187e-07,
+      "loss": 0.2896,
+      "step": 5283
+    },
+    {
+      "epoch": 2.457293442877073,
+      "grad_norm": 0.3328745365142822,
+      "learning_rate": 9.628823764540035e-07,
+      "loss": 0.303,
+      "step": 5284
+    },
+    {
+      "epoch": 2.4577584870562705,
+      "grad_norm": 0.3310834765434265,
+      "learning_rate": 9.612865401073634e-07,
+      "loss": 0.3166,
+      "step": 5285
+    },
+    {
+      "epoch": 2.4582235312354674,
+      "grad_norm": 0.43221089243888855,
+      "learning_rate": 9.596918866342959e-07,
+      "loss": 0.3503,
+      "step": 5286
+    },
+    {
+      "epoch": 2.4586885754146643,
+      "grad_norm": 0.36144617199897766,
+      "learning_rate": 9.580984165018458e-07,
+      "loss": 0.3356,
+      "step": 5287
+    },
+    {
+      "epoch": 2.4591536195938613,
+      "grad_norm": 0.3269920349121094,
+      "learning_rate": 9.565061301767176e-07,
+      "loss": 0.3237,
+      "step": 5288
+    },
+    {
+      "epoch": 2.4596186637730586,
+      "grad_norm": 0.32409968972206116,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.3154,
+      "step": 5289
+    },
+    {
+      "epoch": 2.4600837079522555,
+      "grad_norm": 0.3190405070781708,
+      "learning_rate": 9.533251108134922e-07,
+      "loss": 0.3,
+      "step": 5290
+    },
+    {
+      "epoch": 2.4605487521314524,
+      "grad_norm": 0.3245762884616852,
+      "learning_rate": 9.517363787070672e-07,
+      "loss": 0.3034,
+      "step": 5291
+    },
+    {
+      "epoch": 2.4610137963106498,
+      "grad_norm": 0.32275402545928955,
+      "learning_rate": 9.501488322712987e-07,
+      "loss": 0.3172,
+      "step": 5292
+    },
+    {
+      "epoch": 2.4614788404898467,
+      "grad_norm": 0.33684971928596497,
+      "learning_rate": 9.485624719711551e-07,
+      "loss": 0.3347,
+      "step": 5293
+    },
+    {
+      "epoch": 2.4619438846690436,
+      "grad_norm": 0.35810065269470215,
+      "learning_rate": 9.469772982712561e-07,
+      "loss": 0.3018,
+      "step": 5294
+    },
+    {
+      "epoch": 2.4624089288482405,
+      "grad_norm": 0.34780067205429077,
+      "learning_rate": 9.453933116358715e-07,
+      "loss": 0.3157,
+      "step": 5295
+    },
+    {
+      "epoch": 2.4628739730274374,
+      "grad_norm": 0.36027947068214417,
+      "learning_rate": 9.438105125289276e-07,
+      "loss": 0.3516,
+      "step": 5296
+    },
+    {
+      "epoch": 2.4633390172066347,
+      "grad_norm": 0.317049503326416,
+      "learning_rate": 9.422289014139996e-07,
+      "loss": 0.2994,
+      "step": 5297
+    },
+    {
+      "epoch": 2.4638040613858316,
+      "grad_norm": 0.3178118169307709,
+      "learning_rate": 9.406484787543136e-07,
+      "loss": 0.3013,
+      "step": 5298
+    },
+    {
+      "epoch": 2.4642691055650285,
+      "grad_norm": 0.3206396698951721,
+      "learning_rate": 9.390692450127531e-07,
+      "loss": 0.3182,
+      "step": 5299
+    },
+    {
+      "epoch": 2.464734149744226,
+      "grad_norm": 0.35063356161117554,
+      "learning_rate": 9.374912006518467e-07,
+      "loss": 0.3607,
+      "step": 5300
+    },
+    {
+      "epoch": 2.465199193923423,
+      "grad_norm": 0.35182520747184753,
+      "learning_rate": 9.359143461337799e-07,
+      "loss": 0.316,
+      "step": 5301
+    },
+    {
+      "epoch": 2.4656642381026197,
+      "grad_norm": 0.3166978359222412,
+      "learning_rate": 9.343386819203892e-07,
+      "loss": 0.2737,
+      "step": 5302
+    },
+    {
+      "epoch": 2.4661292822818166,
+      "grad_norm": 0.3476794362068176,
+      "learning_rate": 9.327642084731575e-07,
+      "loss": 0.3288,
+      "step": 5303
+    },
+    {
+      "epoch": 2.466594326461014,
+      "grad_norm": 0.33571144938468933,
+      "learning_rate": 9.311909262532248e-07,
+      "loss": 0.319,
+      "step": 5304
+    },
+    {
+      "epoch": 2.467059370640211,
+      "grad_norm": 0.3019503355026245,
+      "learning_rate": 9.296188357213804e-07,
+      "loss": 0.2941,
+      "step": 5305
+    },
+    {
+      "epoch": 2.4675244148194078,
+      "grad_norm": 0.37272417545318604,
+      "learning_rate": 9.280479373380624e-07,
+      "loss": 0.3377,
+      "step": 5306
+    },
+    {
+      "epoch": 2.4679894589986047,
+      "grad_norm": 0.30881267786026,
+      "learning_rate": 9.26478231563363e-07,
+      "loss": 0.3139,
+      "step": 5307
+    },
+    {
+      "epoch": 2.468454503177802,
+      "grad_norm": 0.32374900579452515,
+      "learning_rate": 9.249097188570216e-07,
+      "loss": 0.3063,
+      "step": 5308
+    },
+    {
+      "epoch": 2.468919547356999,
+      "grad_norm": 0.3490180969238281,
+      "learning_rate": 9.23342399678433e-07,
+      "loss": 0.3599,
+      "step": 5309
+    },
+    {
+      "epoch": 2.469384591536196,
+      "grad_norm": 0.34069767594337463,
+      "learning_rate": 9.21776274486636e-07,
+      "loss": 0.3225,
+      "step": 5310
+    },
+    {
+      "epoch": 2.469849635715393,
+      "grad_norm": 0.32864606380462646,
+      "learning_rate": 9.202113437403259e-07,
+      "loss": 0.3247,
+      "step": 5311
+    },
+    {
+      "epoch": 2.47031467989459,
+      "grad_norm": 0.3595259487628937,
+      "learning_rate": 9.18647607897844e-07,
+      "loss": 0.3265,
+      "step": 5312
+    },
+    {
+      "epoch": 2.470779724073787,
+      "grad_norm": 0.3525889217853546,
+      "learning_rate": 9.170850674171833e-07,
+      "loss": 0.3167,
+      "step": 5313
+    },
+    {
+      "epoch": 2.471244768252984,
+      "grad_norm": 0.3449327051639557,
+      "learning_rate": 9.155237227559883e-07,
+      "loss": 0.3431,
+      "step": 5314
+    },
+    {
+      "epoch": 2.4717098124321812,
+      "grad_norm": 0.3054923117160797,
+      "learning_rate": 9.139635743715486e-07,
+      "loss": 0.2732,
+      "step": 5315
+    },
+    {
+      "epoch": 2.472174856611378,
+      "grad_norm": 0.3427397310733795,
+      "learning_rate": 9.124046227208083e-07,
+      "loss": 0.3412,
+      "step": 5316
+    },
+    {
+      "epoch": 2.472639900790575,
+      "grad_norm": 0.34003254771232605,
+      "learning_rate": 9.108468682603594e-07,
+      "loss": 0.3276,
+      "step": 5317
+    },
+    {
+      "epoch": 2.473104944969772,
+      "grad_norm": 0.31388726830482483,
+      "learning_rate": 9.092903114464407e-07,
+      "loss": 0.3003,
+      "step": 5318
+    },
+    {
+      "epoch": 2.4735699891489693,
+      "grad_norm": 0.31117507815361023,
+      "learning_rate": 9.077349527349455e-07,
+      "loss": 0.3139,
+      "step": 5319
+    },
+    {
+      "epoch": 2.474035033328166,
+      "grad_norm": 0.32882410287857056,
+      "learning_rate": 9.061807925814098e-07,
+      "loss": 0.3111,
+      "step": 5320
+    },
+    {
+      "epoch": 2.474500077507363,
+      "grad_norm": 0.33620962500572205,
+      "learning_rate": 9.046278314410245e-07,
+      "loss": 0.3183,
+      "step": 5321
+    },
+    {
+      "epoch": 2.4749651216865605,
+      "grad_norm": 0.3354683518409729,
+      "learning_rate": 9.030760697686247e-07,
+      "loss": 0.3094,
+      "step": 5322
+    },
+    {
+      "epoch": 2.4754301658657574,
+      "grad_norm": 0.3122057616710663,
+      "learning_rate": 9.01525508018698e-07,
+      "loss": 0.3064,
+      "step": 5323
+    },
+    {
+      "epoch": 2.4758952100449543,
+      "grad_norm": 0.35190722346305847,
+      "learning_rate": 8.999761466453771e-07,
+      "loss": 0.351,
+      "step": 5324
+    },
+    {
+      "epoch": 2.476360254224151,
+      "grad_norm": 0.3170509338378906,
+      "learning_rate": 8.984279861024453e-07,
+      "loss": 0.3036,
+      "step": 5325
+    },
+    {
+      "epoch": 2.476825298403348,
+      "grad_norm": 0.33579573035240173,
+      "learning_rate": 8.968810268433347e-07,
+      "loss": 0.3135,
+      "step": 5326
+    },
+    {
+      "epoch": 2.4772903425825454,
+      "grad_norm": 0.35341697931289673,
+      "learning_rate": 8.953352693211232e-07,
+      "loss": 0.312,
+      "step": 5327
+    },
+    {
+      "epoch": 2.4777553867617423,
+      "grad_norm": 0.32664796710014343,
+      "learning_rate": 8.937907139885376e-07,
+      "loss": 0.2932,
+      "step": 5328
+    },
+    {
+      "epoch": 2.4782204309409392,
+      "grad_norm": 0.3640941083431244,
+      "learning_rate": 8.922473612979565e-07,
+      "loss": 0.3478,
+      "step": 5329
+    },
+    {
+      "epoch": 2.4786854751201366,
+      "grad_norm": 0.30336251854896545,
+      "learning_rate": 8.907052117013981e-07,
+      "loss": 0.2738,
+      "step": 5330
+    },
+    {
+      "epoch": 2.4791505192993335,
+      "grad_norm": 0.36704903841018677,
+      "learning_rate": 8.891642656505373e-07,
+      "loss": 0.3298,
+      "step": 5331
+    },
+    {
+      "epoch": 2.4796155634785304,
+      "grad_norm": 0.3368130624294281,
+      "learning_rate": 8.876245235966884e-07,
+      "loss": 0.295,
+      "step": 5332
+    },
+    {
+      "epoch": 2.4800806076577273,
+      "grad_norm": 0.3467327058315277,
+      "learning_rate": 8.860859859908199e-07,
+      "loss": 0.3234,
+      "step": 5333
+    },
+    {
+      "epoch": 2.4805456518369247,
+      "grad_norm": 0.3419296145439148,
+      "learning_rate": 8.845486532835435e-07,
+      "loss": 0.3423,
+      "step": 5334
+    },
+    {
+      "epoch": 2.4810106960161216,
+      "grad_norm": 0.3196740448474884,
+      "learning_rate": 8.830125259251171e-07,
+      "loss": 0.3017,
+      "step": 5335
+    },
+    {
+      "epoch": 2.4814757401953185,
+      "grad_norm": 0.3675239086151123,
+      "learning_rate": 8.814776043654494e-07,
+      "loss": 0.3654,
+      "step": 5336
+    },
+    {
+      "epoch": 2.4819407843745154,
+      "grad_norm": 0.3341238796710968,
+      "learning_rate": 8.799438890540929e-07,
+      "loss": 0.2776,
+      "step": 5337
+    },
+    {
+      "epoch": 2.4824058285537127,
+      "grad_norm": 0.3249601125717163,
+      "learning_rate": 8.784113804402506e-07,
+      "loss": 0.3288,
+      "step": 5338
+    },
+    {
+      "epoch": 2.4828708727329096,
+      "grad_norm": 0.3380821645259857,
+      "learning_rate": 8.768800789727655e-07,
+      "loss": 0.3253,
+      "step": 5339
+    },
+    {
+      "epoch": 2.4833359169121065,
+      "grad_norm": 0.3296220302581787,
+      "learning_rate": 8.753499851001341e-07,
+      "loss": 0.2792,
+      "step": 5340
+    },
+    {
+      "epoch": 2.483800961091304,
+      "grad_norm": 0.3442380428314209,
+      "learning_rate": 8.738210992704937e-07,
+      "loss": 0.336,
+      "step": 5341
+    },
+    {
+      "epoch": 2.484266005270501,
+      "grad_norm": 0.3112342059612274,
+      "learning_rate": 8.72293421931632e-07,
+      "loss": 0.3206,
+      "step": 5342
+    },
+    {
+      "epoch": 2.4847310494496977,
+      "grad_norm": 0.3362886905670166,
+      "learning_rate": 8.707669535309793e-07,
+      "loss": 0.3584,
+      "step": 5343
+    },
+    {
+      "epoch": 2.4851960936288946,
+      "grad_norm": 0.3538823127746582,
+      "learning_rate": 8.692416945156151e-07,
+      "loss": 0.3174,
+      "step": 5344
+    },
+    {
+      "epoch": 2.485661137808092,
+      "grad_norm": 0.33575835824012756,
+      "learning_rate": 8.677176453322611e-07,
+      "loss": 0.295,
+      "step": 5345
+    },
+    {
+      "epoch": 2.486126181987289,
+      "grad_norm": 0.3484673500061035,
+      "learning_rate": 8.66194806427288e-07,
+      "loss": 0.3196,
+      "step": 5346
+    },
+    {
+      "epoch": 2.4865912261664858,
+      "grad_norm": 0.36958253383636475,
+      "learning_rate": 8.646731782467094e-07,
+      "loss": 0.3701,
+      "step": 5347
+    },
+    {
+      "epoch": 2.4870562703456827,
+      "grad_norm": 0.31938305497169495,
+      "learning_rate": 8.631527612361861e-07,
+      "loss": 0.2928,
+      "step": 5348
+    },
+    {
+      "epoch": 2.48752131452488,
+      "grad_norm": 0.33608078956604004,
+      "learning_rate": 8.616335558410244e-07,
+      "loss": 0.3191,
+      "step": 5349
+    },
+    {
+      "epoch": 2.487986358704077,
+      "grad_norm": 0.31729990243911743,
+      "learning_rate": 8.601155625061736e-07,
+      "loss": 0.2873,
+      "step": 5350
+    },
+    {
+      "epoch": 2.488451402883274,
+      "grad_norm": 0.35453861951828003,
+      "learning_rate": 8.585987816762292e-07,
+      "loss": 0.3096,
+      "step": 5351
+    },
+    {
+      "epoch": 2.488916447062471,
+      "grad_norm": 0.3638750910758972,
+      "learning_rate": 8.570832137954333e-07,
+      "loss": 0.3396,
+      "step": 5352
+    },
+    {
+      "epoch": 2.489381491241668,
+      "grad_norm": 0.35378140211105347,
+      "learning_rate": 8.555688593076689e-07,
+      "loss": 0.3485,
+      "step": 5353
+    },
+    {
+      "epoch": 2.489846535420865,
+      "grad_norm": 0.34362515807151794,
+      "learning_rate": 8.540557186564685e-07,
+      "loss": 0.2998,
+      "step": 5354
+    },
+    {
+      "epoch": 2.490311579600062,
+      "grad_norm": 0.33936387300491333,
+      "learning_rate": 8.525437922850033e-07,
+      "loss": 0.3127,
+      "step": 5355
+    },
+    {
+      "epoch": 2.490776623779259,
+      "grad_norm": 0.3667736053466797,
+      "learning_rate": 8.51033080636095e-07,
+      "loss": 0.3394,
+      "step": 5356
+    },
+    {
+      "epoch": 2.491241667958456,
+      "grad_norm": 0.3014102578163147,
+      "learning_rate": 8.495235841522038e-07,
+      "loss": 0.2753,
+      "step": 5357
+    },
+    {
+      "epoch": 2.491706712137653,
+      "grad_norm": 0.31866636872291565,
+      "learning_rate": 8.480153032754396e-07,
+      "loss": 0.3203,
+      "step": 5358
+    },
+    {
+      "epoch": 2.49217175631685,
+      "grad_norm": 0.34170639514923096,
+      "learning_rate": 8.465082384475499e-07,
+      "loss": 0.3131,
+      "step": 5359
+    },
+    {
+      "epoch": 2.4926368004960473,
+      "grad_norm": 0.31605064868927,
+      "learning_rate": 8.450023901099314e-07,
+      "loss": 0.3298,
+      "step": 5360
+    },
+    {
+      "epoch": 2.493101844675244,
+      "grad_norm": 0.336970716714859,
+      "learning_rate": 8.434977587036242e-07,
+      "loss": 0.3574,
+      "step": 5361
+    },
+    {
+      "epoch": 2.493566888854441,
+      "grad_norm": 0.3367874324321747,
+      "learning_rate": 8.419943446693069e-07,
+      "loss": 0.3065,
+      "step": 5362
+    },
+    {
+      "epoch": 2.494031933033638,
+      "grad_norm": 0.34468874335289,
+      "learning_rate": 8.404921484473072e-07,
+      "loss": 0.3462,
+      "step": 5363
+    },
+    {
+      "epoch": 2.4944969772128354,
+      "grad_norm": 0.30258965492248535,
+      "learning_rate": 8.38991170477595e-07,
+      "loss": 0.2844,
+      "step": 5364
+    },
+    {
+      "epoch": 2.4949620213920323,
+      "grad_norm": 0.32411083579063416,
+      "learning_rate": 8.3749141119978e-07,
+      "loss": 0.3055,
+      "step": 5365
+    },
+    {
+      "epoch": 2.495427065571229,
+      "grad_norm": 0.33687278628349304,
+      "learning_rate": 8.359928710531195e-07,
+      "loss": 0.3058,
+      "step": 5366
+    },
+    {
+      "epoch": 2.495892109750426,
+      "grad_norm": 0.33974799513816833,
+      "learning_rate": 8.344955504765089e-07,
+      "loss": 0.3166,
+      "step": 5367
+    },
+    {
+      "epoch": 2.4963571539296234,
+      "grad_norm": 0.3623729646205902,
+      "learning_rate": 8.32999449908492e-07,
+      "loss": 0.3154,
+      "step": 5368
+    },
+    {
+      "epoch": 2.4968221981088203,
+      "grad_norm": 0.36364513635635376,
+      "learning_rate": 8.315045697872514e-07,
+      "loss": 0.3118,
+      "step": 5369
+    },
+    {
+      "epoch": 2.4972872422880172,
+      "grad_norm": 0.32118678092956543,
+      "learning_rate": 8.30010910550611e-07,
+      "loss": 0.3173,
+      "step": 5370
+    },
+    {
+      "epoch": 2.4977522864672146,
+      "grad_norm": 0.329588383436203,
+      "learning_rate": 8.285184726360412e-07,
+      "loss": 0.3409,
+      "step": 5371
+    },
+    {
+      "epoch": 2.4982173306464115,
+      "grad_norm": 0.29638898372650146,
+      "learning_rate": 8.27027256480653e-07,
+      "loss": 0.2994,
+      "step": 5372
+    },
+    {
+      "epoch": 2.4986823748256084,
+      "grad_norm": 0.3159889578819275,
+      "learning_rate": 8.255372625212005e-07,
+      "loss": 0.3076,
+      "step": 5373
+    },
+    {
+      "epoch": 2.4991474190048053,
+      "grad_norm": 0.3349873423576355,
+      "learning_rate": 8.240484911940755e-07,
+      "loss": 0.3355,
+      "step": 5374
+    },
+    {
+      "epoch": 2.4996124631840027,
+      "grad_norm": 0.34741565585136414,
+      "learning_rate": 8.225609429353187e-07,
+      "loss": 0.3413,
+      "step": 5375
+    },
+    {
+      "epoch": 2.5000775073631996,
+      "grad_norm": 0.3568498492240906,
+      "learning_rate": 8.210746181806051e-07,
+      "loss": 0.3315,
+      "step": 5376
+    },
+    {
+      "epoch": 2.5005425515423965,
+      "grad_norm": 0.3185107409954071,
+      "learning_rate": 8.195895173652585e-07,
+      "loss": 0.2859,
+      "step": 5377
+    },
+    {
+      "epoch": 2.5010075957215934,
+      "grad_norm": 0.32899004220962524,
+      "learning_rate": 8.181056409242377e-07,
+      "loss": 0.3093,
+      "step": 5378
+    },
+    {
+      "epoch": 2.5014726399007907,
+      "grad_norm": 0.30876079201698303,
+      "learning_rate": 8.16622989292149e-07,
+      "loss": 0.3068,
+      "step": 5379
+    },
+    {
+      "epoch": 2.5019376840799876,
+      "grad_norm": 0.31439492106437683,
+      "learning_rate": 8.151415629032338e-07,
+      "loss": 0.3405,
+      "step": 5380
+    },
+    {
+      "epoch": 2.5024027282591845,
+      "grad_norm": 0.3182373344898224,
+      "learning_rate": 8.136613621913813e-07,
+      "loss": 0.3143,
+      "step": 5381
+    },
+    {
+      "epoch": 2.502867772438382,
+      "grad_norm": 0.3052506744861603,
+      "learning_rate": 8.121823875901152e-07,
+      "loss": 0.32,
+      "step": 5382
+    },
+    {
+      "epoch": 2.503332816617579,
+      "grad_norm": 0.33077898621559143,
+      "learning_rate": 8.107046395326041e-07,
+      "loss": 0.3325,
+      "step": 5383
+    },
+    {
+      "epoch": 2.5037978607967757,
+      "grad_norm": 0.31627658009529114,
+      "learning_rate": 8.092281184516571e-07,
+      "loss": 0.2717,
+      "step": 5384
+    },
+    {
+      "epoch": 2.5042629049759726,
+      "grad_norm": 0.34410303831100464,
+      "learning_rate": 8.077528247797234e-07,
+      "loss": 0.3265,
+      "step": 5385
+    },
+    {
+      "epoch": 2.5047279491551695,
+      "grad_norm": 0.3441540002822876,
+      "learning_rate": 8.062787589488913e-07,
+      "loss": 0.3192,
+      "step": 5386
+    },
+    {
+      "epoch": 2.505192993334367,
+      "grad_norm": 0.3462598919868469,
+      "learning_rate": 8.048059213908927e-07,
+      "loss": 0.3397,
+      "step": 5387
+    },
+    {
+      "epoch": 2.5056580375135638,
+      "grad_norm": 0.302598774433136,
+      "learning_rate": 8.033343125370952e-07,
+      "loss": 0.2927,
+      "step": 5388
+    },
+    {
+      "epoch": 2.5061230816927607,
+      "grad_norm": 0.33706727623939514,
+      "learning_rate": 8.018639328185113e-07,
+      "loss": 0.3264,
+      "step": 5389
+    },
+    {
+      "epoch": 2.506588125871958,
+      "grad_norm": 0.3481765687465668,
+      "learning_rate": 8.003947826657898e-07,
+      "loss": 0.3415,
+      "step": 5390
+    },
+    {
+      "epoch": 2.507053170051155,
+      "grad_norm": 0.31380075216293335,
+      "learning_rate": 7.989268625092223e-07,
+      "loss": 0.2639,
+      "step": 5391
+    },
+    {
+      "epoch": 2.507518214230352,
+      "grad_norm": 0.34795790910720825,
+      "learning_rate": 7.974601727787374e-07,
+      "loss": 0.3427,
+      "step": 5392
+    },
+    {
+      "epoch": 2.507983258409549,
+      "grad_norm": 0.32209163904190063,
+      "learning_rate": 7.959947139039065e-07,
+      "loss": 0.3024,
+      "step": 5393
+    },
+    {
+      "epoch": 2.508448302588746,
+      "grad_norm": 0.3540125787258148,
+      "learning_rate": 7.945304863139358e-07,
+      "loss": 0.3565,
+      "step": 5394
+    },
+    {
+      "epoch": 2.508913346767943,
+      "grad_norm": 0.37601977586746216,
+      "learning_rate": 7.930674904376762e-07,
+      "loss": 0.3115,
+      "step": 5395
+    },
+    {
+      "epoch": 2.50937839094714,
+      "grad_norm": 0.33575406670570374,
+      "learning_rate": 7.916057267036159e-07,
+      "loss": 0.3297,
+      "step": 5396
+    },
+    {
+      "epoch": 2.509843435126337,
+      "grad_norm": 0.32133379578590393,
+      "learning_rate": 7.901451955398792e-07,
+      "loss": 0.3136,
+      "step": 5397
+    },
+    {
+      "epoch": 2.510308479305534,
+      "grad_norm": 0.3117270767688751,
+      "learning_rate": 7.886858973742334e-07,
+      "loss": 0.2933,
+      "step": 5398
+    },
+    {
+      "epoch": 2.510773523484731,
+      "grad_norm": 0.3320932984352112,
+      "learning_rate": 7.872278326340849e-07,
+      "loss": 0.3404,
+      "step": 5399
+    },
+    {
+      "epoch": 2.511238567663928,
+      "grad_norm": 0.3289964497089386,
+      "learning_rate": 7.857710017464737e-07,
+      "loss": 0.3172,
+      "step": 5400
+    },
+    {
+      "epoch": 2.5117036118431253,
+      "grad_norm": 0.32881319522857666,
+      "learning_rate": 7.843154051380852e-07,
+      "loss": 0.3287,
+      "step": 5401
+    },
+    {
+      "epoch": 2.512168656022322,
+      "grad_norm": 0.3330029547214508,
+      "learning_rate": 7.828610432352373e-07,
+      "loss": 0.2865,
+      "step": 5402
+    },
+    {
+      "epoch": 2.512633700201519,
+      "grad_norm": 0.3702026307582855,
+      "learning_rate": 7.814079164638915e-07,
+      "loss": 0.3376,
+      "step": 5403
+    },
+    {
+      "epoch": 2.513098744380716,
+      "grad_norm": 0.3574296832084656,
+      "learning_rate": 7.799560252496424e-07,
+      "loss": 0.3269,
+      "step": 5404
+    },
+    {
+      "epoch": 2.513563788559913,
+      "grad_norm": 0.3529759645462036,
+      "learning_rate": 7.785053700177275e-07,
+      "loss": 0.3205,
+      "step": 5405
+    },
+    {
+      "epoch": 2.5140288327391103,
+      "grad_norm": 0.33158957958221436,
+      "learning_rate": 7.770559511930187e-07,
+      "loss": 0.3341,
+      "step": 5406
+    },
+    {
+      "epoch": 2.514493876918307,
+      "grad_norm": 0.3311402499675751,
+      "learning_rate": 7.756077692000274e-07,
+      "loss": 0.3263,
+      "step": 5407
+    },
+    {
+      "epoch": 2.514958921097504,
+      "grad_norm": 0.377869188785553,
+      "learning_rate": 7.741608244629045e-07,
+      "loss": 0.3237,
+      "step": 5408
+    },
+    {
+      "epoch": 2.5154239652767014,
+      "grad_norm": 0.3278749883174896,
+      "learning_rate": 7.727151174054342e-07,
+      "loss": 0.3334,
+      "step": 5409
+    },
+    {
+      "epoch": 2.5158890094558983,
+      "grad_norm": 0.34262946248054504,
+      "learning_rate": 7.712706484510424e-07,
+      "loss": 0.332,
+      "step": 5410
+    },
+    {
+      "epoch": 2.5163540536350952,
+      "grad_norm": 0.32814016938209534,
+      "learning_rate": 7.698274180227888e-07,
+      "loss": 0.2826,
+      "step": 5411
+    },
+    {
+      "epoch": 2.5168190978142926,
+      "grad_norm": 0.3465859293937683,
+      "learning_rate": 7.683854265433737e-07,
+      "loss": 0.3271,
+      "step": 5412
+    },
+    {
+      "epoch": 2.5172841419934895,
+      "grad_norm": 0.33419957756996155,
+      "learning_rate": 7.669446744351317e-07,
+      "loss": 0.2999,
+      "step": 5413
+    },
+    {
+      "epoch": 2.5177491861726864,
+      "grad_norm": 0.3603866994380951,
+      "learning_rate": 7.655051621200377e-07,
+      "loss": 0.3173,
+      "step": 5414
+    },
+    {
+      "epoch": 2.5182142303518833,
+      "grad_norm": 0.33212265372276306,
+      "learning_rate": 7.640668900196985e-07,
+      "loss": 0.292,
+      "step": 5415
+    },
+    {
+      "epoch": 2.51867927453108,
+      "grad_norm": 0.35766664147377014,
+      "learning_rate": 7.626298585553637e-07,
+      "loss": 0.3385,
+      "step": 5416
+    },
+    {
+      "epoch": 2.5191443187102776,
+      "grad_norm": 0.35297131538391113,
+      "learning_rate": 7.611940681479141e-07,
+      "loss": 0.3162,
+      "step": 5417
+    },
+    {
+      "epoch": 2.5196093628894745,
+      "grad_norm": 0.3046947419643402,
+      "learning_rate": 7.597595192178702e-07,
+      "loss": 0.2764,
+      "step": 5418
+    },
+    {
+      "epoch": 2.5200744070686714,
+      "grad_norm": 0.38086947798728943,
+      "learning_rate": 7.583262121853879e-07,
+      "loss": 0.3521,
+      "step": 5419
+    },
+    {
+      "epoch": 2.5205394512478687,
+      "grad_norm": 0.357149600982666,
+      "learning_rate": 7.568941474702618e-07,
+      "loss": 0.3225,
+      "step": 5420
+    },
+    {
+      "epoch": 2.5210044954270656,
+      "grad_norm": 0.32697343826293945,
+      "learning_rate": 7.554633254919169e-07,
+      "loss": 0.3138,
+      "step": 5421
+    },
+    {
+      "epoch": 2.5214695396062625,
+      "grad_norm": 0.3249689042568207,
+      "learning_rate": 7.540337466694203e-07,
+      "loss": 0.3149,
+      "step": 5422
+    },
+    {
+      "epoch": 2.52193458378546,
+      "grad_norm": 0.3241865634918213,
+      "learning_rate": 7.526054114214704e-07,
+      "loss": 0.3102,
+      "step": 5423
+    },
+    {
+      "epoch": 2.5223996279646568,
+      "grad_norm": 0.33546963334083557,
+      "learning_rate": 7.511783201664053e-07,
+      "loss": 0.34,
+      "step": 5424
+    },
+    {
+      "epoch": 2.5228646721438537,
+      "grad_norm": 0.3344916105270386,
+      "learning_rate": 7.49752473322195e-07,
+      "loss": 0.3078,
+      "step": 5425
+    },
+    {
+      "epoch": 2.5233297163230506,
+      "grad_norm": 0.3365899622440338,
+      "learning_rate": 7.48327871306449e-07,
+      "loss": 0.3172,
+      "step": 5426
+    },
+    {
+      "epoch": 2.5237947605022475,
+      "grad_norm": 0.3597327768802643,
+      "learning_rate": 7.469045145364079e-07,
+      "loss": 0.3203,
+      "step": 5427
+    },
+    {
+      "epoch": 2.524259804681445,
+      "grad_norm": 0.3610391914844513,
+      "learning_rate": 7.454824034289515e-07,
+      "loss": 0.2926,
+      "step": 5428
+    },
+    {
+      "epoch": 2.5247248488606417,
+      "grad_norm": 0.33750978112220764,
+      "learning_rate": 7.440615384005917e-07,
+      "loss": 0.288,
+      "step": 5429
+    },
+    {
+      "epoch": 2.5251898930398387,
+      "grad_norm": 0.35543230175971985,
+      "learning_rate": 7.426419198674773e-07,
+      "loss": 0.3081,
+      "step": 5430
+    },
+    {
+      "epoch": 2.525654937219036,
+      "grad_norm": 0.3200836777687073,
+      "learning_rate": 7.412235482453911e-07,
+      "loss": 0.3226,
+      "step": 5431
+    },
+    {
+      "epoch": 2.526119981398233,
+      "grad_norm": 0.46929338574409485,
+      "learning_rate": 7.398064239497538e-07,
+      "loss": 0.3042,
+      "step": 5432
+    },
+    {
+      "epoch": 2.52658502557743,
+      "grad_norm": 0.31712985038757324,
+      "learning_rate": 7.383905473956137e-07,
+      "loss": 0.313,
+      "step": 5433
+    },
+    {
+      "epoch": 2.5270500697566267,
+      "grad_norm": 0.31012576818466187,
+      "learning_rate": 7.369759189976622e-07,
+      "loss": 0.3,
+      "step": 5434
+    },
+    {
+      "epoch": 2.527515113935824,
+      "grad_norm": 0.3287335932254791,
+      "learning_rate": 7.355625391702176e-07,
+      "loss": 0.3219,
+      "step": 5435
+    },
+    {
+      "epoch": 2.527980158115021,
+      "grad_norm": 0.3296997547149658,
+      "learning_rate": 7.341504083272388e-07,
+      "loss": 0.3344,
+      "step": 5436
+    },
+    {
+      "epoch": 2.528445202294218,
+      "grad_norm": 0.3109728693962097,
+      "learning_rate": 7.327395268823128e-07,
+      "loss": 0.323,
+      "step": 5437
+    },
+    {
+      "epoch": 2.528910246473415,
+      "grad_norm": 0.2954341471195221,
+      "learning_rate": 7.313298952486675e-07,
+      "loss": 0.2939,
+      "step": 5438
+    },
+    {
+      "epoch": 2.529375290652612,
+      "grad_norm": 0.3587803542613983,
+      "learning_rate": 7.299215138391574e-07,
+      "loss": 0.3491,
+      "step": 5439
+    },
+    {
+      "epoch": 2.529840334831809,
+      "grad_norm": 0.29528895020484924,
+      "learning_rate": 7.285143830662778e-07,
+      "loss": 0.2828,
+      "step": 5440
+    },
+    {
+      "epoch": 2.530305379011006,
+      "grad_norm": 0.33470654487609863,
+      "learning_rate": 7.271085033421516e-07,
+      "loss": 0.3367,
+      "step": 5441
+    },
+    {
+      "epoch": 2.5307704231902033,
+      "grad_norm": 0.3154504895210266,
+      "learning_rate": 7.2570387507854e-07,
+      "loss": 0.2794,
+      "step": 5442
+    },
+    {
+      "epoch": 2.5312354673694,
+      "grad_norm": 0.34760943055152893,
+      "learning_rate": 7.243004986868357e-07,
+      "loss": 0.3367,
+      "step": 5443
+    },
+    {
+      "epoch": 2.531700511548597,
+      "grad_norm": 0.33287641406059265,
+      "learning_rate": 7.228983745780643e-07,
+      "loss": 0.3303,
+      "step": 5444
+    },
+    {
+      "epoch": 2.532165555727794,
+      "grad_norm": 0.3180674612522125,
+      "learning_rate": 7.214975031628856e-07,
+      "loss": 0.2925,
+      "step": 5445
+    },
+    {
+      "epoch": 2.532630599906991,
+      "grad_norm": 0.3350416421890259,
+      "learning_rate": 7.200978848515911e-07,
+      "loss": 0.34,
+      "step": 5446
+    },
+    {
+      "epoch": 2.5330956440861883,
+      "grad_norm": 0.347086101770401,
+      "learning_rate": 7.186995200541086e-07,
+      "loss": 0.3296,
+      "step": 5447
+    },
+    {
+      "epoch": 2.533560688265385,
+      "grad_norm": 0.3760274350643158,
+      "learning_rate": 7.17302409179993e-07,
+      "loss": 0.349,
+      "step": 5448
+    },
+    {
+      "epoch": 2.534025732444582,
+      "grad_norm": 0.3314683437347412,
+      "learning_rate": 7.159065526384384e-07,
+      "loss": 0.2927,
+      "step": 5449
+    },
+    {
+      "epoch": 2.5344907766237794,
+      "grad_norm": 0.309643417596817,
+      "learning_rate": 7.145119508382664e-07,
+      "loss": 0.2969,
+      "step": 5450
+    },
+    {
+      "epoch": 2.5349558208029763,
+      "grad_norm": 0.4211340546607971,
+      "learning_rate": 7.131186041879357e-07,
+      "loss": 0.3603,
+      "step": 5451
+    },
+    {
+      "epoch": 2.5354208649821732,
+      "grad_norm": 0.3383927643299103,
+      "learning_rate": 7.117265130955314e-07,
+      "loss": 0.3306,
+      "step": 5452
+    },
+    {
+      "epoch": 2.5358859091613706,
+      "grad_norm": 0.30522429943084717,
+      "learning_rate": 7.10335677968777e-07,
+      "loss": 0.2957,
+      "step": 5453
+    },
+    {
+      "epoch": 2.5363509533405675,
+      "grad_norm": 0.31220120191574097,
+      "learning_rate": 7.089460992150243e-07,
+      "loss": 0.3003,
+      "step": 5454
+    },
+    {
+      "epoch": 2.5368159975197644,
+      "grad_norm": 0.36357051134109497,
+      "learning_rate": 7.075577772412607e-07,
+      "loss": 0.3585,
+      "step": 5455
+    },
+    {
+      "epoch": 2.5372810416989613,
+      "grad_norm": 0.3370119333267212,
+      "learning_rate": 7.061707124540995e-07,
+      "loss": 0.3196,
+      "step": 5456
+    },
+    {
+      "epoch": 2.537746085878158,
+      "grad_norm": 0.33187320828437805,
+      "learning_rate": 7.047849052597927e-07,
+      "loss": 0.3124,
+      "step": 5457
+    },
+    {
+      "epoch": 2.5382111300573555,
+      "grad_norm": 0.3284560739994049,
+      "learning_rate": 7.034003560642183e-07,
+      "loss": 0.321,
+      "step": 5458
+    },
+    {
+      "epoch": 2.5386761742365525,
+      "grad_norm": 0.31819069385528564,
+      "learning_rate": 7.020170652728903e-07,
+      "loss": 0.3049,
+      "step": 5459
+    },
+    {
+      "epoch": 2.5391412184157494,
+      "grad_norm": 0.2943916618824005,
+      "learning_rate": 7.006350332909495e-07,
+      "loss": 0.2641,
+      "step": 5460
+    },
+    {
+      "epoch": 2.5396062625949467,
+      "grad_norm": 0.3456489145755768,
+      "learning_rate": 6.992542605231739e-07,
+      "loss": 0.367,
+      "step": 5461
+    },
+    {
+      "epoch": 2.5400713067741436,
+      "grad_norm": 0.30786919593811035,
+      "learning_rate": 6.978747473739666e-07,
+      "loss": 0.3087,
+      "step": 5462
+    },
+    {
+      "epoch": 2.5405363509533405,
+      "grad_norm": 0.3164925277233124,
+      "learning_rate": 6.964964942473662e-07,
+      "loss": 0.3066,
+      "step": 5463
+    },
+    {
+      "epoch": 2.541001395132538,
+      "grad_norm": 0.32444530725479126,
+      "learning_rate": 6.951195015470396e-07,
+      "loss": 0.3156,
+      "step": 5464
+    },
+    {
+      "epoch": 2.5414664393117348,
+      "grad_norm": 0.3363860845565796,
+      "learning_rate": 6.937437696762861e-07,
+      "loss": 0.3014,
+      "step": 5465
+    },
+    {
+      "epoch": 2.5419314834909317,
+      "grad_norm": 0.34026390314102173,
+      "learning_rate": 6.923692990380349e-07,
+      "loss": 0.3131,
+      "step": 5466
+    },
+    {
+      "epoch": 2.5423965276701286,
+      "grad_norm": 0.31564757227897644,
+      "learning_rate": 6.909960900348483e-07,
+      "loss": 0.3066,
+      "step": 5467
+    },
+    {
+      "epoch": 2.5428615718493255,
+      "grad_norm": 0.3490791618824005,
+      "learning_rate": 6.896241430689133e-07,
+      "loss": 0.3221,
+      "step": 5468
+    },
+    {
+      "epoch": 2.543326616028523,
+      "grad_norm": 0.346028208732605,
+      "learning_rate": 6.882534585420542e-07,
+      "loss": 0.3227,
+      "step": 5469
+    },
+    {
+      "epoch": 2.5437916602077197,
+      "grad_norm": 0.3382766842842102,
+      "learning_rate": 6.868840368557194e-07,
+      "loss": 0.3471,
+      "step": 5470
+    },
+    {
+      "epoch": 2.5442567043869166,
+      "grad_norm": 0.30995672941207886,
+      "learning_rate": 6.855158784109927e-07,
+      "loss": 0.3273,
+      "step": 5471
+    },
+    {
+      "epoch": 2.544721748566114,
+      "grad_norm": 0.3215526342391968,
+      "learning_rate": 6.841489836085835e-07,
+      "loss": 0.3241,
+      "step": 5472
+    },
+    {
+      "epoch": 2.545186792745311,
+      "grad_norm": 0.3126165270805359,
+      "learning_rate": 6.827833528488348e-07,
+      "loss": 0.2946,
+      "step": 5473
+    },
+    {
+      "epoch": 2.545651836924508,
+      "grad_norm": 0.3339928388595581,
+      "learning_rate": 6.814189865317156e-07,
+      "loss": 0.3383,
+      "step": 5474
+    },
+    {
+      "epoch": 2.5461168811037047,
+      "grad_norm": 0.30406853556632996,
+      "learning_rate": 6.800558850568295e-07,
+      "loss": 0.3031,
+      "step": 5475
+    },
+    {
+      "epoch": 2.5465819252829016,
+      "grad_norm": 0.3174978494644165,
+      "learning_rate": 6.786940488234034e-07,
+      "loss": 0.3134,
+      "step": 5476
+    },
+    {
+      "epoch": 2.547046969462099,
+      "grad_norm": 0.32636481523513794,
+      "learning_rate": 6.773334782302993e-07,
+      "loss": 0.3233,
+      "step": 5477
+    },
+    {
+      "epoch": 2.547512013641296,
+      "grad_norm": 0.31358960270881653,
+      "learning_rate": 6.759741736760062e-07,
+      "loss": 0.2805,
+      "step": 5478
+    },
+    {
+      "epoch": 2.5479770578204928,
+      "grad_norm": 0.36361631751060486,
+      "learning_rate": 6.746161355586411e-07,
+      "loss": 0.3479,
+      "step": 5479
+    },
+    {
+      "epoch": 2.54844210199969,
+      "grad_norm": 0.3181024193763733,
+      "learning_rate": 6.732593642759533e-07,
+      "loss": 0.2933,
+      "step": 5480
+    },
+    {
+      "epoch": 2.548907146178887,
+      "grad_norm": 0.3153908848762512,
+      "learning_rate": 6.719038602253164e-07,
+      "loss": 0.3188,
+      "step": 5481
+    },
+    {
+      "epoch": 2.549372190358084,
+      "grad_norm": 0.3370981216430664,
+      "learning_rate": 6.705496238037379e-07,
+      "loss": 0.3397,
+      "step": 5482
+    },
+    {
+      "epoch": 2.5498372345372813,
+      "grad_norm": 0.32690051198005676,
+      "learning_rate": 6.691966554078494e-07,
+      "loss": 0.2883,
+      "step": 5483
+    },
+    {
+      "epoch": 2.550302278716478,
+      "grad_norm": 0.35293567180633545,
+      "learning_rate": 6.678449554339161e-07,
+      "loss": 0.3367,
+      "step": 5484
+    },
+    {
+      "epoch": 2.550767322895675,
+      "grad_norm": 0.2994639575481415,
+      "learning_rate": 6.664945242778264e-07,
+      "loss": 0.3071,
+      "step": 5485
+    },
+    {
+      "epoch": 2.551232367074872,
+      "grad_norm": 0.30928120017051697,
+      "learning_rate": 6.651453623351017e-07,
+      "loss": 0.3188,
+      "step": 5486
+    },
+    {
+      "epoch": 2.551697411254069,
+      "grad_norm": 0.3208847939968109,
+      "learning_rate": 6.637974700008876e-07,
+      "loss": 0.3407,
+      "step": 5487
+    },
+    {
+      "epoch": 2.5521624554332663,
+      "grad_norm": 0.3318404257297516,
+      "learning_rate": 6.624508476699609e-07,
+      "loss": 0.3058,
+      "step": 5488
+    },
+    {
+      "epoch": 2.552627499612463,
+      "grad_norm": 0.34476324915885925,
+      "learning_rate": 6.611054957367253e-07,
+      "loss": 0.2937,
+      "step": 5489
+    },
+    {
+      "epoch": 2.55309254379166,
+      "grad_norm": 0.3424232304096222,
+      "learning_rate": 6.597614145952136e-07,
+      "loss": 0.3149,
+      "step": 5490
+    },
+    {
+      "epoch": 2.5535575879708574,
+      "grad_norm": 0.336582750082016,
+      "learning_rate": 6.584186046390839e-07,
+      "loss": 0.3275,
+      "step": 5491
+    },
+    {
+      "epoch": 2.5540226321500543,
+      "grad_norm": 0.30492153763771057,
+      "learning_rate": 6.570770662616244e-07,
+      "loss": 0.3226,
+      "step": 5492
+    },
+    {
+      "epoch": 2.5544876763292512,
+      "grad_norm": 0.344941645860672,
+      "learning_rate": 6.557367998557485e-07,
+      "loss": 0.3153,
+      "step": 5493
+    },
+    {
+      "epoch": 2.5549527205084486,
+      "grad_norm": 0.3531348407268524,
+      "learning_rate": 6.543978058140005e-07,
+      "loss": 0.3301,
+      "step": 5494
+    },
+    {
+      "epoch": 2.5554177646876455,
+      "grad_norm": 0.3221784234046936,
+      "learning_rate": 6.530600845285478e-07,
+      "loss": 0.3318,
+      "step": 5495
+    },
+    {
+      "epoch": 2.5558828088668424,
+      "grad_norm": 0.30838826298713684,
+      "learning_rate": 6.517236363911894e-07,
+      "loss": 0.2999,
+      "step": 5496
+    },
+    {
+      "epoch": 2.5563478530460393,
+      "grad_norm": 0.3686372935771942,
+      "learning_rate": 6.503884617933471e-07,
+      "loss": 0.3305,
+      "step": 5497
+    },
+    {
+      "epoch": 2.556812897225236,
+      "grad_norm": 0.31591683626174927,
+      "learning_rate": 6.490545611260741e-07,
+      "loss": 0.3122,
+      "step": 5498
+    },
+    {
+      "epoch": 2.5572779414044335,
+      "grad_norm": 0.34594687819480896,
+      "learning_rate": 6.477219347800462e-07,
+      "loss": 0.3461,
+      "step": 5499
+    },
+    {
+      "epoch": 2.5577429855836304,
+      "grad_norm": 0.31076663732528687,
+      "learning_rate": 6.463905831455685e-07,
+      "loss": 0.2906,
+      "step": 5500
+    },
+    {
+      "epoch": 2.5582080297628274,
+      "grad_norm": 0.3259563446044922,
+      "learning_rate": 6.450605066125726e-07,
+      "loss": 0.3008,
+      "step": 5501
+    },
+    {
+      "epoch": 2.5586730739420247,
+      "grad_norm": 0.31875672936439514,
+      "learning_rate": 6.437317055706172e-07,
+      "loss": 0.3028,
+      "step": 5502
+    },
+    {
+      "epoch": 2.5591381181212216,
+      "grad_norm": 0.3503369390964508,
+      "learning_rate": 6.424041804088848e-07,
+      "loss": 0.3006,
+      "step": 5503
+    },
+    {
+      "epoch": 2.5596031623004185,
+      "grad_norm": 0.3309253454208374,
+      "learning_rate": 6.410779315161885e-07,
+      "loss": 0.3309,
+      "step": 5504
+    },
+    {
+      "epoch": 2.5600682064796154,
+      "grad_norm": 0.30641940236091614,
+      "learning_rate": 6.397529592809615e-07,
+      "loss": 0.3159,
+      "step": 5505
+    },
+    {
+      "epoch": 2.5605332506588123,
+      "grad_norm": 0.4112693965435028,
+      "learning_rate": 6.384292640912704e-07,
+      "loss": 0.328,
+      "step": 5506
+    },
+    {
+      "epoch": 2.5609982948380097,
+      "grad_norm": 0.3142835199832916,
+      "learning_rate": 6.371068463348006e-07,
+      "loss": 0.3059,
+      "step": 5507
+    },
+    {
+      "epoch": 2.5614633390172066,
+      "grad_norm": 0.32879573106765747,
+      "learning_rate": 6.357857063988692e-07,
+      "loss": 0.3145,
+      "step": 5508
+    },
+    {
+      "epoch": 2.5619283831964035,
+      "grad_norm": 0.3391149640083313,
+      "learning_rate": 6.344658446704155e-07,
+      "loss": 0.3231,
+      "step": 5509
+    },
+    {
+      "epoch": 2.562393427375601,
+      "grad_norm": 0.30871260166168213,
+      "learning_rate": 6.331472615360062e-07,
+      "loss": 0.3248,
+      "step": 5510
+    },
+    {
+      "epoch": 2.5628584715547977,
+      "grad_norm": 0.32863956689834595,
+      "learning_rate": 6.318299573818315e-07,
+      "loss": 0.2954,
+      "step": 5511
+    },
+    {
+      "epoch": 2.5633235157339946,
+      "grad_norm": 0.3561699688434601,
+      "learning_rate": 6.305139325937098e-07,
+      "loss": 0.3346,
+      "step": 5512
+    },
+    {
+      "epoch": 2.563788559913192,
+      "grad_norm": 0.32806986570358276,
+      "learning_rate": 6.291991875570841e-07,
+      "loss": 0.2992,
+      "step": 5513
+    },
+    {
+      "epoch": 2.564253604092389,
+      "grad_norm": 0.4038988947868347,
+      "learning_rate": 6.278857226570196e-07,
+      "loss": 0.3525,
+      "step": 5514
+    },
+    {
+      "epoch": 2.564718648271586,
+      "grad_norm": 0.35461750626564026,
+      "learning_rate": 6.265735382782106e-07,
+      "loss": 0.3138,
+      "step": 5515
+    },
+    {
+      "epoch": 2.5651836924507827,
+      "grad_norm": 0.3094637393951416,
+      "learning_rate": 6.252626348049734e-07,
+      "loss": 0.305,
+      "step": 5516
+    },
+    {
+      "epoch": 2.5656487366299796,
+      "grad_norm": 0.3154189884662628,
+      "learning_rate": 6.239530126212518e-07,
+      "loss": 0.3293,
+      "step": 5517
+    },
+    {
+      "epoch": 2.566113780809177,
+      "grad_norm": 0.3151610493659973,
+      "learning_rate": 6.226446721106111e-07,
+      "loss": 0.3055,
+      "step": 5518
+    },
+    {
+      "epoch": 2.566578824988374,
+      "grad_norm": 0.3299187421798706,
+      "learning_rate": 6.213376136562449e-07,
+      "loss": 0.3182,
+      "step": 5519
+    },
+    {
+      "epoch": 2.5670438691675708,
+      "grad_norm": 0.34355783462524414,
+      "learning_rate": 6.20031837640967e-07,
+      "loss": 0.3599,
+      "step": 5520
+    },
+    {
+      "epoch": 2.567508913346768,
+      "grad_norm": 0.3010472059249878,
+      "learning_rate": 6.187273444472202e-07,
+      "loss": 0.2715,
+      "step": 5521
+    },
+    {
+      "epoch": 2.567973957525965,
+      "grad_norm": 0.3234943151473999,
+      "learning_rate": 6.174241344570681e-07,
+      "loss": 0.3174,
+      "step": 5522
+    },
+    {
+      "epoch": 2.568439001705162,
+      "grad_norm": 0.3738628327846527,
+      "learning_rate": 6.161222080522e-07,
+      "loss": 0.3164,
+      "step": 5523
+    },
+    {
+      "epoch": 2.5689040458843593,
+      "grad_norm": 0.31000205874443054,
+      "learning_rate": 6.14821565613929e-07,
+      "loss": 0.3146,
+      "step": 5524
+    },
+    {
+      "epoch": 2.569369090063556,
+      "grad_norm": 0.34590592980384827,
+      "learning_rate": 6.135222075231933e-07,
+      "loss": 0.3303,
+      "step": 5525
+    },
+    {
+      "epoch": 2.569834134242753,
+      "grad_norm": 0.3145350217819214,
+      "learning_rate": 6.122241341605523e-07,
+      "loss": 0.3044,
+      "step": 5526
+    },
+    {
+      "epoch": 2.57029917842195,
+      "grad_norm": 0.3343406021595001,
+      "learning_rate": 6.109273459061916e-07,
+      "loss": 0.3601,
+      "step": 5527
+    },
+    {
+      "epoch": 2.570764222601147,
+      "grad_norm": 0.30140528082847595,
+      "learning_rate": 6.096318431399178e-07,
+      "loss": 0.3012,
+      "step": 5528
+    },
+    {
+      "epoch": 2.5712292667803442,
+      "grad_norm": 0.34568503499031067,
+      "learning_rate": 6.083376262411644e-07,
+      "loss": 0.3658,
+      "step": 5529
+    },
+    {
+      "epoch": 2.571694310959541,
+      "grad_norm": 0.33937206864356995,
+      "learning_rate": 6.070446955889853e-07,
+      "loss": 0.3469,
+      "step": 5530
+    },
+    {
+      "epoch": 2.572159355138738,
+      "grad_norm": 0.3581177592277527,
+      "learning_rate": 6.057530515620608e-07,
+      "loss": 0.3144,
+      "step": 5531
+    },
+    {
+      "epoch": 2.5726243993179354,
+      "grad_norm": 0.29124531149864197,
+      "learning_rate": 6.044626945386894e-07,
+      "loss": 0.3194,
+      "step": 5532
+    },
+    {
+      "epoch": 2.5730894434971323,
+      "grad_norm": 0.30630677938461304,
+      "learning_rate": 6.031736248967984e-07,
+      "loss": 0.3128,
+      "step": 5533
+    },
+    {
+      "epoch": 2.573554487676329,
+      "grad_norm": 0.3839922547340393,
+      "learning_rate": 6.018858430139335e-07,
+      "loss": 0.3398,
+      "step": 5534
+    },
+    {
+      "epoch": 2.574019531855526,
+      "grad_norm": 0.336647093296051,
+      "learning_rate": 6.005993492672657e-07,
+      "loss": 0.2957,
+      "step": 5535
+    },
+    {
+      "epoch": 2.574484576034723,
+      "grad_norm": 0.3169485330581665,
+      "learning_rate": 5.993141440335887e-07,
+      "loss": 0.3208,
+      "step": 5536
+    },
+    {
+      "epoch": 2.5749496202139204,
+      "grad_norm": 0.3146721124649048,
+      "learning_rate": 5.980302276893191e-07,
+      "loss": 0.3066,
+      "step": 5537
+    },
+    {
+      "epoch": 2.5754146643931173,
+      "grad_norm": 0.32376930117607117,
+      "learning_rate": 5.967476006104922e-07,
+      "loss": 0.3335,
+      "step": 5538
+    },
+    {
+      "epoch": 2.575879708572314,
+      "grad_norm": 0.33218416571617126,
+      "learning_rate": 5.95466263172772e-07,
+      "loss": 0.3186,
+      "step": 5539
+    },
+    {
+      "epoch": 2.5763447527515115,
+      "grad_norm": 0.31140658259391785,
+      "learning_rate": 5.941862157514383e-07,
+      "loss": 0.3257,
+      "step": 5540
+    },
+    {
+      "epoch": 2.5768097969307084,
+      "grad_norm": 0.324327677488327,
+      "learning_rate": 5.92907458721399e-07,
+      "loss": 0.3425,
+      "step": 5541
+    },
+    {
+      "epoch": 2.5772748411099053,
+      "grad_norm": 0.3449290692806244,
+      "learning_rate": 5.916299924571789e-07,
+      "loss": 0.3705,
+      "step": 5542
+    },
+    {
+      "epoch": 2.5777398852891027,
+      "grad_norm": 0.32734236121177673,
+      "learning_rate": 5.903538173329287e-07,
+      "loss": 0.2884,
+      "step": 5543
+    },
+    {
+      "epoch": 2.5782049294682996,
+      "grad_norm": 0.33377382159233093,
+      "learning_rate": 5.890789337224184e-07,
+      "loss": 0.3193,
+      "step": 5544
+    },
+    {
+      "epoch": 2.5786699736474965,
+      "grad_norm": 0.31567618250846863,
+      "learning_rate": 5.87805341999042e-07,
+      "loss": 0.3011,
+      "step": 5545
+    },
+    {
+      "epoch": 2.5791350178266934,
+      "grad_norm": 0.3374476134777069,
+      "learning_rate": 5.865330425358118e-07,
+      "loss": 0.3549,
+      "step": 5546
+    },
+    {
+      "epoch": 2.5796000620058903,
+      "grad_norm": 0.3086042106151581,
+      "learning_rate": 5.852620357053651e-07,
+      "loss": 0.3016,
+      "step": 5547
+    },
+    {
+      "epoch": 2.5800651061850877,
+      "grad_norm": 0.31613045930862427,
+      "learning_rate": 5.839923218799587e-07,
+      "loss": 0.3169,
+      "step": 5548
+    },
+    {
+      "epoch": 2.5805301503642846,
+      "grad_norm": 0.2952466309070587,
+      "learning_rate": 5.827239014314723e-07,
+      "loss": 0.3013,
+      "step": 5549
+    },
+    {
+      "epoch": 2.5809951945434815,
+      "grad_norm": 0.32223108410835266,
+      "learning_rate": 5.814567747314049e-07,
+      "loss": 0.3347,
+      "step": 5550
+    },
+    {
+      "epoch": 2.581460238722679,
+      "grad_norm": 0.3356245458126068,
+      "learning_rate": 5.801909421508756e-07,
+      "loss": 0.3239,
+      "step": 5551
+    },
+    {
+      "epoch": 2.5819252829018757,
+      "grad_norm": 0.35723865032196045,
+      "learning_rate": 5.789264040606291e-07,
+      "loss": 0.3591,
+      "step": 5552
+    },
+    {
+      "epoch": 2.5823903270810726,
+      "grad_norm": 0.3004617989063263,
+      "learning_rate": 5.776631608310257e-07,
+      "loss": 0.3046,
+      "step": 5553
+    },
+    {
+      "epoch": 2.58285537126027,
+      "grad_norm": 0.35467642545700073,
+      "learning_rate": 5.764012128320507e-07,
+      "loss": 0.3802,
+      "step": 5554
+    },
+    {
+      "epoch": 2.583320415439467,
+      "grad_norm": 0.3245300054550171,
+      "learning_rate": 5.75140560433306e-07,
+      "loss": 0.2898,
+      "step": 5555
+    },
+    {
+      "epoch": 2.583785459618664,
+      "grad_norm": 0.342220276594162,
+      "learning_rate": 5.738812040040187e-07,
+      "loss": 0.3478,
+      "step": 5556
+    },
+    {
+      "epoch": 2.5842505037978607,
+      "grad_norm": 0.3202744126319885,
+      "learning_rate": 5.726231439130314e-07,
+      "loss": 0.3162,
+      "step": 5557
+    },
+    {
+      "epoch": 2.5847155479770576,
+      "grad_norm": 0.33512887358665466,
+      "learning_rate": 5.713663805288106e-07,
+      "loss": 0.308,
+      "step": 5558
+    },
+    {
+      "epoch": 2.585180592156255,
+      "grad_norm": 0.33824869990348816,
+      "learning_rate": 5.701109142194422e-07,
+      "loss": 0.308,
+      "step": 5559
+    },
+    {
+      "epoch": 2.585645636335452,
+      "grad_norm": 0.3539738953113556,
+      "learning_rate": 5.688567453526328e-07,
+      "loss": 0.3317,
+      "step": 5560
+    },
+    {
+      "epoch": 2.5861106805146488,
+      "grad_norm": 0.32088860869407654,
+      "learning_rate": 5.676038742957057e-07,
+      "loss": 0.3272,
+      "step": 5561
+    },
+    {
+      "epoch": 2.586575724693846,
+      "grad_norm": 0.33904799818992615,
+      "learning_rate": 5.663523014156086e-07,
+      "loss": 0.2933,
+      "step": 5562
+    },
+    {
+      "epoch": 2.587040768873043,
+      "grad_norm": 0.3523750901222229,
+      "learning_rate": 5.651020270789049e-07,
+      "loss": 0.2996,
+      "step": 5563
+    },
+    {
+      "epoch": 2.58750581305224,
+      "grad_norm": 0.32999008893966675,
+      "learning_rate": 5.638530516517821e-07,
+      "loss": 0.3068,
+      "step": 5564
+    },
+    {
+      "epoch": 2.587970857231437,
+      "grad_norm": 0.3218596279621124,
+      "learning_rate": 5.626053755000421e-07,
+      "loss": 0.32,
+      "step": 5565
+    },
+    {
+      "epoch": 2.588435901410634,
+      "grad_norm": 0.32146209478378296,
+      "learning_rate": 5.613589989891116e-07,
+      "loss": 0.3225,
+      "step": 5566
+    },
+    {
+      "epoch": 2.588900945589831,
+      "grad_norm": 0.2964208424091339,
+      "learning_rate": 5.601139224840318e-07,
+      "loss": 0.2975,
+      "step": 5567
+    },
+    {
+      "epoch": 2.589365989769028,
+      "grad_norm": 0.332450270652771,
+      "learning_rate": 5.588701463494672e-07,
+      "loss": 0.3431,
+      "step": 5568
+    },
+    {
+      "epoch": 2.589831033948225,
+      "grad_norm": 0.28528085350990295,
+      "learning_rate": 5.576276709496975e-07,
+      "loss": 0.2784,
+      "step": 5569
+    },
+    {
+      "epoch": 2.5902960781274222,
+      "grad_norm": 0.3144235908985138,
+      "learning_rate": 5.563864966486254e-07,
+      "loss": 0.3402,
+      "step": 5570
+    },
+    {
+      "epoch": 2.590761122306619,
+      "grad_norm": 0.35045549273490906,
+      "learning_rate": 5.551466238097697e-07,
+      "loss": 0.3278,
+      "step": 5571
+    },
+    {
+      "epoch": 2.591226166485816,
+      "grad_norm": 0.3082601726055145,
+      "learning_rate": 5.539080527962704e-07,
+      "loss": 0.3108,
+      "step": 5572
+    },
+    {
+      "epoch": 2.5916912106650134,
+      "grad_norm": 0.30737003684043884,
+      "learning_rate": 5.526707839708834e-07,
+      "loss": 0.3004,
+      "step": 5573
+    },
+    {
+      "epoch": 2.5921562548442103,
+      "grad_norm": 0.34509941935539246,
+      "learning_rate": 5.514348176959855e-07,
+      "loss": 0.3588,
+      "step": 5574
+    },
+    {
+      "epoch": 2.592621299023407,
+      "grad_norm": 0.33185267448425293,
+      "learning_rate": 5.5020015433357e-07,
+      "loss": 0.3188,
+      "step": 5575
+    },
+    {
+      "epoch": 2.593086343202604,
+      "grad_norm": 0.32632753252983093,
+      "learning_rate": 5.489667942452515e-07,
+      "loss": 0.2991,
+      "step": 5576
+    },
+    {
+      "epoch": 2.593551387381801,
+      "grad_norm": 0.31657204031944275,
+      "learning_rate": 5.477347377922593e-07,
+      "loss": 0.3415,
+      "step": 5577
+    },
+    {
+      "epoch": 2.5940164315609984,
+      "grad_norm": 0.3294719457626343,
+      "learning_rate": 5.465039853354442e-07,
+      "loss": 0.3144,
+      "step": 5578
+    },
+    {
+      "epoch": 2.5944814757401953,
+      "grad_norm": 0.3257904052734375,
+      "learning_rate": 5.452745372352725e-07,
+      "loss": 0.2761,
+      "step": 5579
+    },
+    {
+      "epoch": 2.594946519919392,
+      "grad_norm": 0.3162056505680084,
+      "learning_rate": 5.440463938518304e-07,
+      "loss": 0.333,
+      "step": 5580
+    },
+    {
+      "epoch": 2.5954115640985895,
+      "grad_norm": 0.3013499081134796,
+      "learning_rate": 5.428195555448202e-07,
+      "loss": 0.3004,
+      "step": 5581
+    },
+    {
+      "epoch": 2.5958766082777864,
+      "grad_norm": 0.3342210352420807,
+      "learning_rate": 5.415940226735633e-07,
+      "loss": 0.3388,
+      "step": 5582
+    },
+    {
+      "epoch": 2.5963416524569833,
+      "grad_norm": 0.31008031964302063,
+      "learning_rate": 5.403697955969988e-07,
+      "loss": 0.2927,
+      "step": 5583
+    },
+    {
+      "epoch": 2.5968066966361807,
+      "grad_norm": 0.32369595766067505,
+      "learning_rate": 5.391468746736834e-07,
+      "loss": 0.335,
+      "step": 5584
+    },
+    {
+      "epoch": 2.5972717408153776,
+      "grad_norm": 0.3301704525947571,
+      "learning_rate": 5.379252602617902e-07,
+      "loss": 0.3322,
+      "step": 5585
+    },
+    {
+      "epoch": 2.5977367849945745,
+      "grad_norm": 0.2886703908443451,
+      "learning_rate": 5.367049527191093e-07,
+      "loss": 0.2747,
+      "step": 5586
+    },
+    {
+      "epoch": 2.5982018291737714,
+      "grad_norm": 0.3267196714878082,
+      "learning_rate": 5.354859524030503e-07,
+      "loss": 0.3076,
+      "step": 5587
+    },
+    {
+      "epoch": 2.5986668733529683,
+      "grad_norm": 0.34606629610061646,
+      "learning_rate": 5.342682596706372e-07,
+      "loss": 0.3378,
+      "step": 5588
+    },
+    {
+      "epoch": 2.5991319175321657,
+      "grad_norm": 0.3182050287723541,
+      "learning_rate": 5.330518748785147e-07,
+      "loss": 0.3144,
+      "step": 5589
+    },
+    {
+      "epoch": 2.5995969617113626,
+      "grad_norm": 0.3103780746459961,
+      "learning_rate": 5.318367983829393e-07,
+      "loss": 0.2717,
+      "step": 5590
+    },
+    {
+      "epoch": 2.6000620058905595,
+      "grad_norm": 0.36062633991241455,
+      "learning_rate": 5.306230305397897e-07,
+      "loss": 0.3159,
+      "step": 5591
+    },
+    {
+      "epoch": 2.600527050069757,
+      "grad_norm": 0.3258778750896454,
+      "learning_rate": 5.294105717045567e-07,
+      "loss": 0.3356,
+      "step": 5592
+    },
+    {
+      "epoch": 2.6009920942489537,
+      "grad_norm": 0.33117127418518066,
+      "learning_rate": 5.281994222323506e-07,
+      "loss": 0.329,
+      "step": 5593
+    },
+    {
+      "epoch": 2.6014571384281506,
+      "grad_norm": 0.32020604610443115,
+      "learning_rate": 5.269895824778976e-07,
+      "loss": 0.3093,
+      "step": 5594
+    },
+    {
+      "epoch": 2.6019221826073475,
+      "grad_norm": 0.33276131749153137,
+      "learning_rate": 5.25781052795541e-07,
+      "loss": 0.3143,
+      "step": 5595
+    },
+    {
+      "epoch": 2.602387226786545,
+      "grad_norm": 0.3160843551158905,
+      "learning_rate": 5.245738335392376e-07,
+      "loss": 0.2935,
+      "step": 5596
+    },
+    {
+      "epoch": 2.602852270965742,
+      "grad_norm": 0.3074491620063782,
+      "learning_rate": 5.233679250625646e-07,
+      "loss": 0.3083,
+      "step": 5597
+    },
+    {
+      "epoch": 2.6033173151449387,
+      "grad_norm": 0.3221345841884613,
+      "learning_rate": 5.221633277187104e-07,
+      "loss": 0.3324,
+      "step": 5598
+    },
+    {
+      "epoch": 2.6037823593241356,
+      "grad_norm": 0.36884576082229614,
+      "learning_rate": 5.209600418604843e-07,
+      "loss": 0.3293,
+      "step": 5599
+    },
+    {
+      "epoch": 2.604247403503333,
+      "grad_norm": 0.3398898243904114,
+      "learning_rate": 5.197580678403074e-07,
+      "loss": 0.3435,
+      "step": 5600
+    },
+    {
+      "epoch": 2.60471244768253,
+      "grad_norm": 0.2826615273952484,
+      "learning_rate": 5.185574060102206e-07,
+      "loss": 0.2942,
+      "step": 5601
+    },
+    {
+      "epoch": 2.6051774918617268,
+      "grad_norm": 0.35747769474983215,
+      "learning_rate": 5.17358056721875e-07,
+      "loss": 0.3186,
+      "step": 5602
+    },
+    {
+      "epoch": 2.605642536040924,
+      "grad_norm": 0.32651564478874207,
+      "learning_rate": 5.161600203265438e-07,
+      "loss": 0.3283,
+      "step": 5603
+    },
+    {
+      "epoch": 2.606107580220121,
+      "grad_norm": 0.305650532245636,
+      "learning_rate": 5.149632971751096e-07,
+      "loss": 0.3121,
+      "step": 5604
+    },
+    {
+      "epoch": 2.606572624399318,
+      "grad_norm": 0.30761486291885376,
+      "learning_rate": 5.137678876180746e-07,
+      "loss": 0.3106,
+      "step": 5605
+    },
+    {
+      "epoch": 2.607037668578515,
+      "grad_norm": 0.3115074038505554,
+      "learning_rate": 5.125737920055551e-07,
+      "loss": 0.3197,
+      "step": 5606
+    },
+    {
+      "epoch": 2.6075027127577117,
+      "grad_norm": 0.3347545564174652,
+      "learning_rate": 5.113810106872825e-07,
+      "loss": 0.3098,
+      "step": 5607
+    },
+    {
+      "epoch": 2.607967756936909,
+      "grad_norm": 0.3298303782939911,
+      "learning_rate": 5.10189544012602e-07,
+      "loss": 0.316,
+      "step": 5608
+    },
+    {
+      "epoch": 2.608432801116106,
+      "grad_norm": 0.33597150444984436,
+      "learning_rate": 5.089993923304759e-07,
+      "loss": 0.3258,
+      "step": 5609
+    },
+    {
+      "epoch": 2.608897845295303,
+      "grad_norm": 0.312759667634964,
+      "learning_rate": 5.078105559894791e-07,
+      "loss": 0.3181,
+      "step": 5610
+    },
+    {
+      "epoch": 2.6093628894745002,
+      "grad_norm": 0.32988691329956055,
+      "learning_rate": 5.066230353378038e-07,
+      "loss": 0.3211,
+      "step": 5611
+    },
+    {
+      "epoch": 2.609827933653697,
+      "grad_norm": 0.333040714263916,
+      "learning_rate": 5.054368307232537e-07,
+      "loss": 0.3087,
+      "step": 5612
+    },
+    {
+      "epoch": 2.610292977832894,
+      "grad_norm": 0.3228285014629364,
+      "learning_rate": 5.042519424932512e-07,
+      "loss": 0.3095,
+      "step": 5613
+    },
+    {
+      "epoch": 2.6107580220120914,
+      "grad_norm": 0.35871362686157227,
+      "learning_rate": 5.030683709948292e-07,
+      "loss": 0.3171,
+      "step": 5614
+    },
+    {
+      "epoch": 2.6112230661912883,
+      "grad_norm": 0.3096243739128113,
+      "learning_rate": 5.018861165746369e-07,
+      "loss": 0.3054,
+      "step": 5615
+    },
+    {
+      "epoch": 2.611688110370485,
+      "grad_norm": 0.2971365451812744,
+      "learning_rate": 5.007051795789375e-07,
+      "loss": 0.2977,
+      "step": 5616
+    },
+    {
+      "epoch": 2.612153154549682,
+      "grad_norm": 0.31067079305648804,
+      "learning_rate": 4.995255603536076e-07,
+      "loss": 0.3265,
+      "step": 5617
+    },
+    {
+      "epoch": 2.612618198728879,
+      "grad_norm": 0.3296055793762207,
+      "learning_rate": 4.983472592441391e-07,
+      "loss": 0.3023,
+      "step": 5618
+    },
+    {
+      "epoch": 2.6130832429080764,
+      "grad_norm": 0.33535173535346985,
+      "learning_rate": 4.971702765956388e-07,
+      "loss": 0.3282,
+      "step": 5619
+    },
+    {
+      "epoch": 2.6135482870872733,
+      "grad_norm": 0.30592218041419983,
+      "learning_rate": 4.959946127528231e-07,
+      "loss": 0.3186,
+      "step": 5620
+    },
+    {
+      "epoch": 2.61401333126647,
+      "grad_norm": 0.32471343874931335,
+      "learning_rate": 4.948202680600267e-07,
+      "loss": 0.2948,
+      "step": 5621
+    },
+    {
+      "epoch": 2.6144783754456675,
+      "grad_norm": 0.331249862909317,
+      "learning_rate": 4.936472428611961e-07,
+      "loss": 0.3168,
+      "step": 5622
+    },
+    {
+      "epoch": 2.6149434196248644,
+      "grad_norm": 0.33011171221733093,
+      "learning_rate": 4.924755374998891e-07,
+      "loss": 0.3389,
+      "step": 5623
+    },
+    {
+      "epoch": 2.6154084638040613,
+      "grad_norm": 0.3165659010410309,
+      "learning_rate": 4.913051523192819e-07,
+      "loss": 0.317,
+      "step": 5624
+    },
+    {
+      "epoch": 2.6158735079832587,
+      "grad_norm": 0.3213482201099396,
+      "learning_rate": 4.901360876621597e-07,
+      "loss": 0.3103,
+      "step": 5625
+    },
+    {
+      "epoch": 2.6163385521624556,
+      "grad_norm": 0.3293680250644684,
+      "learning_rate": 4.88968343870923e-07,
+      "loss": 0.3189,
+      "step": 5626
+    },
+    {
+      "epoch": 2.6168035963416525,
+      "grad_norm": 0.3114880323410034,
+      "learning_rate": 4.87801921287585e-07,
+      "loss": 0.317,
+      "step": 5627
+    },
+    {
+      "epoch": 2.6172686405208494,
+      "grad_norm": 0.32358518242836,
+      "learning_rate": 4.866368202537714e-07,
+      "loss": 0.3315,
+      "step": 5628
+    },
+    {
+      "epoch": 2.6177336847000463,
+      "grad_norm": 0.3022099733352661,
+      "learning_rate": 4.854730411107217e-07,
+      "loss": 0.3197,
+      "step": 5629
+    },
+    {
+      "epoch": 2.6181987288792437,
+      "grad_norm": 0.34834814071655273,
+      "learning_rate": 4.843105841992895e-07,
+      "loss": 0.3536,
+      "step": 5630
+    },
+    {
+      "epoch": 2.6186637730584406,
+      "grad_norm": 0.3669452369213104,
+      "learning_rate": 4.831494498599371e-07,
+      "loss": 0.3462,
+      "step": 5631
+    },
+    {
+      "epoch": 2.6191288172376375,
+      "grad_norm": 0.3363710343837738,
+      "learning_rate": 4.819896384327433e-07,
+      "loss": 0.3383,
+      "step": 5632
+    },
+    {
+      "epoch": 2.619593861416835,
+      "grad_norm": 0.30958956480026245,
+      "learning_rate": 4.808311502573976e-07,
+      "loss": 0.2891,
+      "step": 5633
+    },
+    {
+      "epoch": 2.6200589055960317,
+      "grad_norm": 0.3210178315639496,
+      "learning_rate": 4.796739856732024e-07,
+      "loss": 0.3327,
+      "step": 5634
+    },
+    {
+      "epoch": 2.6205239497752286,
+      "grad_norm": 0.31170886754989624,
+      "learning_rate": 4.785181450190723e-07,
+      "loss": 0.3178,
+      "step": 5635
+    },
+    {
+      "epoch": 2.6209889939544255,
+      "grad_norm": 0.3083958029747009,
+      "learning_rate": 4.773636286335348e-07,
+      "loss": 0.3016,
+      "step": 5636
+    },
+    {
+      "epoch": 2.6214540381336224,
+      "grad_norm": 0.35575956106185913,
+      "learning_rate": 4.7621043685472824e-07,
+      "loss": 0.3492,
+      "step": 5637
+    },
+    {
+      "epoch": 2.62191908231282,
+      "grad_norm": 0.3041512370109558,
+      "learning_rate": 4.750585700204047e-07,
+      "loss": 0.2905,
+      "step": 5638
+    },
+    {
+      "epoch": 2.6223841264920167,
+      "grad_norm": 0.3430851399898529,
+      "learning_rate": 4.739080284679254e-07,
+      "loss": 0.3088,
+      "step": 5639
+    },
+    {
+      "epoch": 2.6228491706712136,
+      "grad_norm": 0.3387916088104248,
+      "learning_rate": 4.727588125342669e-07,
+      "loss": 0.319,
+      "step": 5640
+    },
+    {
+      "epoch": 2.623314214850411,
+      "grad_norm": 0.3547907769680023,
+      "learning_rate": 4.716109225560156e-07,
+      "loss": 0.2953,
+      "step": 5641
+    },
+    {
+      "epoch": 2.623779259029608,
+      "grad_norm": 0.32964009046554565,
+      "learning_rate": 4.7046435886937024e-07,
+      "loss": 0.3286,
+      "step": 5642
+    },
+    {
+      "epoch": 2.6242443032088048,
+      "grad_norm": 0.38167962431907654,
+      "learning_rate": 4.6931912181014007e-07,
+      "loss": 0.3104,
+      "step": 5643
+    },
+    {
+      "epoch": 2.624709347388002,
+      "grad_norm": 0.33540862798690796,
+      "learning_rate": 4.681752117137467e-07,
+      "loss": 0.3494,
+      "step": 5644
+    },
+    {
+      "epoch": 2.625174391567199,
+      "grad_norm": 0.3289260268211365,
+      "learning_rate": 4.6703262891522214e-07,
+      "loss": 0.3212,
+      "step": 5645
+    },
+    {
+      "epoch": 2.625639435746396,
+      "grad_norm": 0.3132651150226593,
+      "learning_rate": 4.6589137374921155e-07,
+      "loss": 0.2817,
+      "step": 5646
+    },
+    {
+      "epoch": 2.626104479925593,
+      "grad_norm": 0.33676326274871826,
+      "learning_rate": 4.647514465499686e-07,
+      "loss": 0.3141,
+      "step": 5647
+    },
+    {
+      "epoch": 2.6265695241047897,
+      "grad_norm": 0.3420335054397583,
+      "learning_rate": 4.6361284765136125e-07,
+      "loss": 0.3312,
+      "step": 5648
+    },
+    {
+      "epoch": 2.627034568283987,
+      "grad_norm": 0.33695536851882935,
+      "learning_rate": 4.6247557738686445e-07,
+      "loss": 0.3208,
+      "step": 5649
+    },
+    {
+      "epoch": 2.627499612463184,
+      "grad_norm": 0.351370245218277,
+      "learning_rate": 4.613396360895683e-07,
+      "loss": 0.3156,
+      "step": 5650
+    },
+    {
+      "epoch": 2.627964656642381,
+      "grad_norm": 0.3574759364128113,
+      "learning_rate": 4.602050240921696e-07,
+      "loss": 0.305,
+      "step": 5651
+    },
+    {
+      "epoch": 2.6284297008215782,
+      "grad_norm": 0.3272240161895752,
+      "learning_rate": 4.590717417269791e-07,
+      "loss": 0.338,
+      "step": 5652
+    },
+    {
+      "epoch": 2.628894745000775,
+      "grad_norm": 0.3331619203090668,
+      "learning_rate": 4.5793978932591574e-07,
+      "loss": 0.3186,
+      "step": 5653
+    },
+    {
+      "epoch": 2.629359789179972,
+      "grad_norm": 0.3595447242259979,
+      "learning_rate": 4.568091672205122e-07,
+      "loss": 0.3238,
+      "step": 5654
+    },
+    {
+      "epoch": 2.6298248333591694,
+      "grad_norm": 0.327280730009079,
+      "learning_rate": 4.5567987574190677e-07,
+      "loss": 0.3365,
+      "step": 5655
+    },
+    {
+      "epoch": 2.6302898775383663,
+      "grad_norm": 0.28672948479652405,
+      "learning_rate": 4.5455191522085274e-07,
+      "loss": 0.2672,
+      "step": 5656
+    },
+    {
+      "epoch": 2.630754921717563,
+      "grad_norm": 0.3301229774951935,
+      "learning_rate": 4.534252859877097e-07,
+      "loss": 0.344,
+      "step": 5657
+    },
+    {
+      "epoch": 2.63121996589676,
+      "grad_norm": 0.3244812488555908,
+      "learning_rate": 4.522999883724494e-07,
+      "loss": 0.3144,
+      "step": 5658
+    },
+    {
+      "epoch": 2.631685010075957,
+      "grad_norm": 0.3279408812522888,
+      "learning_rate": 4.511760227046541e-07,
+      "loss": 0.3403,
+      "step": 5659
+    },
+    {
+      "epoch": 2.6321500542551544,
+      "grad_norm": 0.32007279992103577,
+      "learning_rate": 4.500533893135134e-07,
+      "loss": 0.3198,
+      "step": 5660
+    },
+    {
+      "epoch": 2.6326150984343513,
+      "grad_norm": 0.3092459440231323,
+      "learning_rate": 4.489320885278309e-07,
+      "loss": 0.3085,
+      "step": 5661
+    },
+    {
+      "epoch": 2.633080142613548,
+      "grad_norm": 0.32061436772346497,
+      "learning_rate": 4.4781212067601445e-07,
+      "loss": 0.3115,
+      "step": 5662
+    },
+    {
+      "epoch": 2.6335451867927455,
+      "grad_norm": 0.3219727873802185,
+      "learning_rate": 4.4669348608608664e-07,
+      "loss": 0.3278,
+      "step": 5663
+    },
+    {
+      "epoch": 2.6340102309719424,
+      "grad_norm": 0.33633241057395935,
+      "learning_rate": 4.4557618508567603e-07,
+      "loss": 0.3318,
+      "step": 5664
+    },
+    {
+      "epoch": 2.6344752751511393,
+      "grad_norm": 0.31990426778793335,
+      "learning_rate": 4.4446021800202356e-07,
+      "loss": 0.3395,
+      "step": 5665
+    },
+    {
+      "epoch": 2.6349403193303362,
+      "grad_norm": 0.33744657039642334,
+      "learning_rate": 4.4334558516197666e-07,
+      "loss": 0.3202,
+      "step": 5666
+    },
+    {
+      "epoch": 2.635405363509533,
+      "grad_norm": 0.301940381526947,
+      "learning_rate": 4.422322868919937e-07,
+      "loss": 0.2888,
+      "step": 5667
+    },
+    {
+      "epoch": 2.6358704076887305,
+      "grad_norm": 0.34101638197898865,
+      "learning_rate": 4.411203235181405e-07,
+      "loss": 0.329,
+      "step": 5668
+    },
+    {
+      "epoch": 2.6363354518679274,
+      "grad_norm": 0.3637217581272125,
+      "learning_rate": 4.400096953660948e-07,
+      "loss": 0.3665,
+      "step": 5669
+    },
+    {
+      "epoch": 2.6368004960471243,
+      "grad_norm": 0.30867865681648254,
+      "learning_rate": 4.3890040276114044e-07,
+      "loss": 0.2851,
+      "step": 5670
+    },
+    {
+      "epoch": 2.6372655402263216,
+      "grad_norm": 0.3138599693775177,
+      "learning_rate": 4.377924460281718e-07,
+      "loss": 0.3383,
+      "step": 5671
+    },
+    {
+      "epoch": 2.6377305844055186,
+      "grad_norm": 0.2975723147392273,
+      "learning_rate": 4.3668582549169005e-07,
+      "loss": 0.2762,
+      "step": 5672
+    },
+    {
+      "epoch": 2.6381956285847155,
+      "grad_norm": 0.3239637315273285,
+      "learning_rate": 4.355805414758085e-07,
+      "loss": 0.3482,
+      "step": 5673
+    },
+    {
+      "epoch": 2.638660672763913,
+      "grad_norm": 0.3464727997779846,
+      "learning_rate": 4.3447659430424507e-07,
+      "loss": 0.3351,
+      "step": 5674
+    },
+    {
+      "epoch": 2.6391257169431097,
+      "grad_norm": 0.309800386428833,
+      "learning_rate": 4.3337398430032815e-07,
+      "loss": 0.301,
+      "step": 5675
+    },
+    {
+      "epoch": 2.6395907611223066,
+      "grad_norm": 0.30942457914352417,
+      "learning_rate": 4.322727117869951e-07,
+      "loss": 0.3041,
+      "step": 5676
+    },
+    {
+      "epoch": 2.6400558053015035,
+      "grad_norm": 0.3338727355003357,
+      "learning_rate": 4.3117277708679126e-07,
+      "loss": 0.3128,
+      "step": 5677
+    },
+    {
+      "epoch": 2.6405208494807004,
+      "grad_norm": 0.33797094225883484,
+      "learning_rate": 4.3007418052186834e-07,
+      "loss": 0.3582,
+      "step": 5678
+    },
+    {
+      "epoch": 2.6409858936598978,
+      "grad_norm": 0.336549311876297,
+      "learning_rate": 4.289769224139884e-07,
+      "loss": 0.3085,
+      "step": 5679
+    },
+    {
+      "epoch": 2.6414509378390947,
+      "grad_norm": 0.3069862425327301,
+      "learning_rate": 4.278810030845193e-07,
+      "loss": 0.3163,
+      "step": 5680
+    },
+    {
+      "epoch": 2.6419159820182916,
+      "grad_norm": 0.3145103454589844,
+      "learning_rate": 4.2678642285443937e-07,
+      "loss": 0.3241,
+      "step": 5681
+    },
+    {
+      "epoch": 2.642381026197489,
+      "grad_norm": 0.30022549629211426,
+      "learning_rate": 4.2569318204433217e-07,
+      "loss": 0.2823,
+      "step": 5682
+    },
+    {
+      "epoch": 2.642846070376686,
+      "grad_norm": 0.37692004442214966,
+      "learning_rate": 4.2460128097439157e-07,
+      "loss": 0.3223,
+      "step": 5683
+    },
+    {
+      "epoch": 2.6433111145558827,
+      "grad_norm": 0.33564627170562744,
+      "learning_rate": 4.235107199644162e-07,
+      "loss": 0.318,
+      "step": 5684
+    },
+    {
+      "epoch": 2.64377615873508,
+      "grad_norm": 0.3265214264392853,
+      "learning_rate": 4.224214993338149e-07,
+      "loss": 0.3559,
+      "step": 5685
+    },
+    {
+      "epoch": 2.644241202914277,
+      "grad_norm": 0.31681397557258606,
+      "learning_rate": 4.2133361940160153e-07,
+      "loss": 0.3053,
+      "step": 5686
+    },
+    {
+      "epoch": 2.644706247093474,
+      "grad_norm": 0.317182332277298,
+      "learning_rate": 4.202470804863984e-07,
+      "loss": 0.3074,
+      "step": 5687
+    },
+    {
+      "epoch": 2.645171291272671,
+      "grad_norm": 0.3192555606365204,
+      "learning_rate": 4.1916188290643643e-07,
+      "loss": 0.3069,
+      "step": 5688
+    },
+    {
+      "epoch": 2.6456363354518677,
+      "grad_norm": 0.3164333403110504,
+      "learning_rate": 4.1807802697955256e-07,
+      "loss": 0.2939,
+      "step": 5689
+    },
+    {
+      "epoch": 2.646101379631065,
+      "grad_norm": 0.32663610577583313,
+      "learning_rate": 4.169955130231884e-07,
+      "loss": 0.3759,
+      "step": 5690
+    },
+    {
+      "epoch": 2.646566423810262,
+      "grad_norm": 0.3227081000804901,
+      "learning_rate": 4.15914341354397e-07,
+      "loss": 0.3027,
+      "step": 5691
+    },
+    {
+      "epoch": 2.647031467989459,
+      "grad_norm": 0.3126518726348877,
+      "learning_rate": 4.1483451228983453e-07,
+      "loss": 0.3067,
+      "step": 5692
+    },
+    {
+      "epoch": 2.6474965121686562,
+      "grad_norm": 0.3389665484428406,
+      "learning_rate": 4.137560261457663e-07,
+      "loss": 0.3341,
+      "step": 5693
+    },
+    {
+      "epoch": 2.647961556347853,
+      "grad_norm": 0.3213329613208771,
+      "learning_rate": 4.1267888323806294e-07,
+      "loss": 0.3122,
+      "step": 5694
+    },
+    {
+      "epoch": 2.64842660052705,
+      "grad_norm": 0.341896116733551,
+      "learning_rate": 4.1160308388220103e-07,
+      "loss": 0.3159,
+      "step": 5695
+    },
+    {
+      "epoch": 2.648891644706247,
+      "grad_norm": 0.33184269070625305,
+      "learning_rate": 4.1052862839326745e-07,
+      "loss": 0.3124,
+      "step": 5696
+    },
+    {
+      "epoch": 2.649356688885444,
+      "grad_norm": 0.32483184337615967,
+      "learning_rate": 4.0945551708594934e-07,
+      "loss": 0.3103,
+      "step": 5697
+    },
+    {
+      "epoch": 2.649821733064641,
+      "grad_norm": 0.33331936597824097,
+      "learning_rate": 4.083837502745458e-07,
+      "loss": 0.3271,
+      "step": 5698
+    },
+    {
+      "epoch": 2.650286777243838,
+      "grad_norm": 0.3043390214443207,
+      "learning_rate": 4.0731332827295966e-07,
+      "loss": 0.2737,
+      "step": 5699
+    },
+    {
+      "epoch": 2.650751821423035,
+      "grad_norm": 0.3190237581729889,
+      "learning_rate": 4.062442513947007e-07,
+      "loss": 0.3411,
+      "step": 5700
+    },
+    {
+      "epoch": 2.6512168656022324,
+      "grad_norm": 0.3191225826740265,
+      "learning_rate": 4.051765199528823e-07,
+      "loss": 0.3043,
+      "step": 5701
+    },
+    {
+      "epoch": 2.6516819097814293,
+      "grad_norm": 0.33000683784484863,
+      "learning_rate": 4.0411013426022773e-07,
+      "loss": 0.3551,
+      "step": 5702
+    },
+    {
+      "epoch": 2.652146953960626,
+      "grad_norm": 0.32607778906822205,
+      "learning_rate": 4.0304509462906203e-07,
+      "loss": 0.3328,
+      "step": 5703
+    },
+    {
+      "epoch": 2.6526119981398235,
+      "grad_norm": 0.3193208575248718,
+      "learning_rate": 4.0198140137132024e-07,
+      "loss": 0.3034,
+      "step": 5704
+    },
+    {
+      "epoch": 2.6530770423190204,
+      "grad_norm": 0.3008403778076172,
+      "learning_rate": 4.0091905479853865e-07,
+      "loss": 0.3229,
+      "step": 5705
+    },
+    {
+      "epoch": 2.6535420864982173,
+      "grad_norm": 0.3564378023147583,
+      "learning_rate": 3.9985805522186336e-07,
+      "loss": 0.3812,
+      "step": 5706
+    },
+    {
+      "epoch": 2.6540071306774142,
+      "grad_norm": 0.37547022104263306,
+      "learning_rate": 3.98798402952042e-07,
+      "loss": 0.332,
+      "step": 5707
+    },
+    {
+      "epoch": 2.654472174856611,
+      "grad_norm": 0.2705591917037964,
+      "learning_rate": 3.977400982994306e-07,
+      "loss": 0.2647,
+      "step": 5708
+    },
+    {
+      "epoch": 2.6549372190358085,
+      "grad_norm": 0.31391727924346924,
+      "learning_rate": 3.966831415739891e-07,
+      "loss": 0.326,
+      "step": 5709
+    },
+    {
+      "epoch": 2.6554022632150054,
+      "grad_norm": 0.3211490213871002,
+      "learning_rate": 3.9562753308528267e-07,
+      "loss": 0.3346,
+      "step": 5710
+    },
+    {
+      "epoch": 2.6558673073942023,
+      "grad_norm": 0.3080466687679291,
+      "learning_rate": 3.945732731424823e-07,
+      "loss": 0.3227,
+      "step": 5711
+    },
+    {
+      "epoch": 2.6563323515733996,
+      "grad_norm": 0.3370213508605957,
+      "learning_rate": 3.935203620543643e-07,
+      "loss": 0.3306,
+      "step": 5712
+    },
+    {
+      "epoch": 2.6567973957525965,
+      "grad_norm": 0.29656779766082764,
+      "learning_rate": 3.924688001293081e-07,
+      "loss": 0.2599,
+      "step": 5713
+    },
+    {
+      "epoch": 2.6572624399317935,
+      "grad_norm": 0.3250420093536377,
+      "learning_rate": 3.9141858767530014e-07,
+      "loss": 0.3182,
+      "step": 5714
+    },
+    {
+      "epoch": 2.657727484110991,
+      "grad_norm": 0.33750423789024353,
+      "learning_rate": 3.903697249999289e-07,
+      "loss": 0.3342,
+      "step": 5715
+    },
+    {
+      "epoch": 2.6581925282901877,
+      "grad_norm": 0.29559361934661865,
+      "learning_rate": 3.8932221241039125e-07,
+      "loss": 0.295,
+      "step": 5716
+    },
+    {
+      "epoch": 2.6586575724693846,
+      "grad_norm": 0.35010501742362976,
+      "learning_rate": 3.882760502134847e-07,
+      "loss": 0.328,
+      "step": 5717
+    },
+    {
+      "epoch": 2.6591226166485815,
+      "grad_norm": 0.3221285343170166,
+      "learning_rate": 3.872312387156146e-07,
+      "loss": 0.3553,
+      "step": 5718
+    },
+    {
+      "epoch": 2.6595876608277784,
+      "grad_norm": 0.3222235441207886,
+      "learning_rate": 3.8618777822278854e-07,
+      "loss": 0.3039,
+      "step": 5719
+    },
+    {
+      "epoch": 2.6600527050069758,
+      "grad_norm": 0.32205820083618164,
+      "learning_rate": 3.8514566904061967e-07,
+      "loss": 0.3102,
+      "step": 5720
+    },
+    {
+      "epoch": 2.6605177491861727,
+      "grad_norm": 0.33951982855796814,
+      "learning_rate": 3.841049114743239e-07,
+      "loss": 0.3002,
+      "step": 5721
+    },
+    {
+      "epoch": 2.6609827933653696,
+      "grad_norm": 0.33085671067237854,
+      "learning_rate": 3.8306550582872306e-07,
+      "loss": 0.3237,
+      "step": 5722
+    },
+    {
+      "epoch": 2.661447837544567,
+      "grad_norm": 0.3133414685726166,
+      "learning_rate": 3.820274524082418e-07,
+      "loss": 0.2978,
+      "step": 5723
+    },
+    {
+      "epoch": 2.661912881723764,
+      "grad_norm": 0.3263706862926483,
+      "learning_rate": 3.809907515169103e-07,
+      "loss": 0.3129,
+      "step": 5724
+    },
+    {
+      "epoch": 2.6623779259029607,
+      "grad_norm": 0.3518964350223541,
+      "learning_rate": 3.7995540345835914e-07,
+      "loss": 0.2998,
+      "step": 5725
+    },
+    {
+      "epoch": 2.6628429700821576,
+      "grad_norm": 0.32105782628059387,
+      "learning_rate": 3.7892140853582725e-07,
+      "loss": 0.3411,
+      "step": 5726
+    },
+    {
+      "epoch": 2.663308014261355,
+      "grad_norm": 0.30892109870910645,
+      "learning_rate": 3.7788876705215307e-07,
+      "loss": 0.3243,
+      "step": 5727
+    },
+    {
+      "epoch": 2.663773058440552,
+      "grad_norm": 0.31716519594192505,
+      "learning_rate": 3.7685747930978236e-07,
+      "loss": 0.341,
+      "step": 5728
+    },
+    {
+      "epoch": 2.664238102619749,
+      "grad_norm": 0.31774482131004333,
+      "learning_rate": 3.758275456107613e-07,
+      "loss": 0.2949,
+      "step": 5729
+    },
+    {
+      "epoch": 2.6647031467989457,
+      "grad_norm": 0.31866157054901123,
+      "learning_rate": 3.747989662567403e-07,
+      "loss": 0.3215,
+      "step": 5730
+    },
+    {
+      "epoch": 2.665168190978143,
+      "grad_norm": 0.33754652738571167,
+      "learning_rate": 3.73771741548975e-07,
+      "loss": 0.3054,
+      "step": 5731
+    },
+    {
+      "epoch": 2.66563323515734,
+      "grad_norm": 0.3196556270122528,
+      "learning_rate": 3.727458717883209e-07,
+      "loss": 0.3075,
+      "step": 5732
+    },
+    {
+      "epoch": 2.666098279336537,
+      "grad_norm": 0.356734037399292,
+      "learning_rate": 3.717213572752404e-07,
+      "loss": 0.3356,
+      "step": 5733
+    },
+    {
+      "epoch": 2.666563323515734,
+      "grad_norm": 0.3005916178226471,
+      "learning_rate": 3.706981983097957e-07,
+      "loss": 0.3205,
+      "step": 5734
+    },
+    {
+      "epoch": 2.667028367694931,
+      "grad_norm": 0.35259753465652466,
+      "learning_rate": 3.6967639519165546e-07,
+      "loss": 0.3188,
+      "step": 5735
+    },
+    {
+      "epoch": 2.667493411874128,
+      "grad_norm": 0.3179108500480652,
+      "learning_rate": 3.686559482200874e-07,
+      "loss": 0.3535,
+      "step": 5736
+    },
+    {
+      "epoch": 2.667958456053325,
+      "grad_norm": 0.3107699751853943,
+      "learning_rate": 3.6763685769396484e-07,
+      "loss": 0.2986,
+      "step": 5737
+    },
+    {
+      "epoch": 2.668423500232522,
+      "grad_norm": 0.29539617896080017,
+      "learning_rate": 3.6661912391176223e-07,
+      "loss": 0.3016,
+      "step": 5738
+    },
+    {
+      "epoch": 2.668888544411719,
+      "grad_norm": 0.3341010808944702,
+      "learning_rate": 3.6560274717155784e-07,
+      "loss": 0.3494,
+      "step": 5739
+    },
+    {
+      "epoch": 2.669353588590916,
+      "grad_norm": 0.3309285342693329,
+      "learning_rate": 3.645877277710308e-07,
+      "loss": 0.295,
+      "step": 5740
+    },
+    {
+      "epoch": 2.669818632770113,
+      "grad_norm": 0.3219027817249298,
+      "learning_rate": 3.635740660074655e-07,
+      "loss": 0.3116,
+      "step": 5741
+    },
+    {
+      "epoch": 2.6702836769493103,
+      "grad_norm": 0.33891165256500244,
+      "learning_rate": 3.6256176217774496e-07,
+      "loss": 0.3293,
+      "step": 5742
+    },
+    {
+      "epoch": 2.6707487211285073,
+      "grad_norm": 0.3448745608329773,
+      "learning_rate": 3.6155081657835876e-07,
+      "loss": 0.3291,
+      "step": 5743
+    },
+    {
+      "epoch": 2.671213765307704,
+      "grad_norm": 0.32368507981300354,
+      "learning_rate": 3.6054122950539447e-07,
+      "loss": 0.3126,
+      "step": 5744
+    },
+    {
+      "epoch": 2.6716788094869015,
+      "grad_norm": 0.3642202913761139,
+      "learning_rate": 3.595330012545445e-07,
+      "loss": 0.3264,
+      "step": 5745
+    },
+    {
+      "epoch": 2.6721438536660984,
+      "grad_norm": 0.3226480782032013,
+      "learning_rate": 3.5852613212110307e-07,
+      "loss": 0.3025,
+      "step": 5746
+    },
+    {
+      "epoch": 2.6726088978452953,
+      "grad_norm": 0.3831692039966583,
+      "learning_rate": 3.5752062239996554e-07,
+      "loss": 0.3591,
+      "step": 5747
+    },
+    {
+      "epoch": 2.6730739420244922,
+      "grad_norm": 0.3260175585746765,
+      "learning_rate": 3.5651647238562904e-07,
+      "loss": 0.3103,
+      "step": 5748
+    },
+    {
+      "epoch": 2.673538986203689,
+      "grad_norm": 0.3270534574985504,
+      "learning_rate": 3.555136823721933e-07,
+      "loss": 0.3214,
+      "step": 5749
+    },
+    {
+      "epoch": 2.6740040303828865,
+      "grad_norm": 0.32291585206985474,
+      "learning_rate": 3.545122526533579e-07,
+      "loss": 0.3335,
+      "step": 5750
+    },
+    {
+      "epoch": 2.6744690745620834,
+      "grad_norm": 0.2985171675682068,
+      "learning_rate": 3.5351218352242755e-07,
+      "loss": 0.3173,
+      "step": 5751
+    },
+    {
+      "epoch": 2.6749341187412803,
+      "grad_norm": 0.32506510615348816,
+      "learning_rate": 3.525134752723042e-07,
+      "loss": 0.3106,
+      "step": 5752
+    },
+    {
+      "epoch": 2.6753991629204776,
+      "grad_norm": 0.33014851808547974,
+      "learning_rate": 3.515161281954943e-07,
+      "loss": 0.3127,
+      "step": 5753
+    },
+    {
+      "epoch": 2.6758642070996745,
+      "grad_norm": 0.33311840891838074,
+      "learning_rate": 3.5052014258410426e-07,
+      "loss": 0.3143,
+      "step": 5754
+    },
+    {
+      "epoch": 2.6763292512788714,
+      "grad_norm": 0.3252905309200287,
+      "learning_rate": 3.4952551872984295e-07,
+      "loss": 0.3435,
+      "step": 5755
+    },
+    {
+      "epoch": 2.6767942954580684,
+      "grad_norm": 0.31564247608184814,
+      "learning_rate": 3.485322569240174e-07,
+      "loss": 0.3095,
+      "step": 5756
+    },
+    {
+      "epoch": 2.6772593396372657,
+      "grad_norm": 0.3435322642326355,
+      "learning_rate": 3.475403574575398e-07,
+      "loss": 0.3198,
+      "step": 5757
+    },
+    {
+      "epoch": 2.6777243838164626,
+      "grad_norm": 0.3655998706817627,
+      "learning_rate": 3.4654982062092113e-07,
+      "loss": 0.3149,
+      "step": 5758
+    },
+    {
+      "epoch": 2.6781894279956595,
+      "grad_norm": 0.32407233119010925,
+      "learning_rate": 3.455606467042738e-07,
+      "loss": 0.3186,
+      "step": 5759
+    },
+    {
+      "epoch": 2.6786544721748564,
+      "grad_norm": 0.3063153624534607,
+      "learning_rate": 3.445728359973094e-07,
+      "loss": 0.2878,
+      "step": 5760
+    },
+    {
+      "epoch": 2.6791195163540538,
+      "grad_norm": 0.35395288467407227,
+      "learning_rate": 3.435863887893431e-07,
+      "loss": 0.3123,
+      "step": 5761
+    },
+    {
+      "epoch": 2.6795845605332507,
+      "grad_norm": 0.3261089026927948,
+      "learning_rate": 3.426013053692878e-07,
+      "loss": 0.3338,
+      "step": 5762
+    },
+    {
+      "epoch": 2.6800496047124476,
+      "grad_norm": 0.3305519223213196,
+      "learning_rate": 3.4161758602566043e-07,
+      "loss": 0.3206,
+      "step": 5763
+    },
+    {
+      "epoch": 2.680514648891645,
+      "grad_norm": 0.32225289940834045,
+      "learning_rate": 3.406352310465749e-07,
+      "loss": 0.3315,
+      "step": 5764
+    },
+    {
+      "epoch": 2.680979693070842,
+      "grad_norm": 0.3169860243797302,
+      "learning_rate": 3.3965424071974727e-07,
+      "loss": 0.3072,
+      "step": 5765
+    },
+    {
+      "epoch": 2.6814447372500387,
+      "grad_norm": 0.3510488271713257,
+      "learning_rate": 3.386746153324943e-07,
+      "loss": 0.3437,
+      "step": 5766
+    },
+    {
+      "epoch": 2.6819097814292356,
+      "grad_norm": 0.3206140995025635,
+      "learning_rate": 3.3769635517173103e-07,
+      "loss": 0.2735,
+      "step": 5767
+    },
+    {
+      "epoch": 2.6823748256084325,
+      "grad_norm": 0.33280014991760254,
+      "learning_rate": 3.3671946052397486e-07,
+      "loss": 0.3409,
+      "step": 5768
+    },
+    {
+      "epoch": 2.68283986978763,
+      "grad_norm": 0.3153855502605438,
+      "learning_rate": 3.3574393167534247e-07,
+      "loss": 0.2849,
+      "step": 5769
+    },
+    {
+      "epoch": 2.683304913966827,
+      "grad_norm": 0.3374989926815033,
+      "learning_rate": 3.347697689115509e-07,
+      "loss": 0.3465,
+      "step": 5770
+    },
+    {
+      "epoch": 2.6837699581460237,
+      "grad_norm": 0.32558801770210266,
+      "learning_rate": 3.337969725179152e-07,
+      "loss": 0.2962,
+      "step": 5771
+    },
+    {
+      "epoch": 2.684235002325221,
+      "grad_norm": 0.31771788001060486,
+      "learning_rate": 3.328255427793531e-07,
+      "loss": 0.3208,
+      "step": 5772
+    },
+    {
+      "epoch": 2.684700046504418,
+      "grad_norm": 0.3207005560398102,
+      "learning_rate": 3.318554799803786e-07,
+      "loss": 0.3074,
+      "step": 5773
+    },
+    {
+      "epoch": 2.685165090683615,
+      "grad_norm": 0.3172682225704193,
+      "learning_rate": 3.3088678440511e-07,
+      "loss": 0.3266,
+      "step": 5774
+    },
+    {
+      "epoch": 2.685630134862812,
+      "grad_norm": 0.3359968960285187,
+      "learning_rate": 3.299194563372604e-07,
+      "loss": 0.2934,
+      "step": 5775
+    },
+    {
+      "epoch": 2.686095179042009,
+      "grad_norm": 0.345907598733902,
+      "learning_rate": 3.289534960601454e-07,
+      "loss": 0.3541,
+      "step": 5776
+    },
+    {
+      "epoch": 2.686560223221206,
+      "grad_norm": 0.328243225812912,
+      "learning_rate": 3.279889038566786e-07,
+      "loss": 0.3324,
+      "step": 5777
+    },
+    {
+      "epoch": 2.687025267400403,
+      "grad_norm": 0.2853672206401825,
+      "learning_rate": 3.2702568000937404e-07,
+      "loss": 0.2756,
+      "step": 5778
+    },
+    {
+      "epoch": 2.6874903115796,
+      "grad_norm": 0.32393786311149597,
+      "learning_rate": 3.260638248003434e-07,
+      "loss": 0.3193,
+      "step": 5779
+    },
+    {
+      "epoch": 2.687955355758797,
+      "grad_norm": 0.3525981903076172,
+      "learning_rate": 3.2510333851129895e-07,
+      "loss": 0.3326,
+      "step": 5780
+    },
+    {
+      "epoch": 2.688420399937994,
+      "grad_norm": 0.347851037979126,
+      "learning_rate": 3.2414422142355184e-07,
+      "loss": 0.3176,
+      "step": 5781
+    },
+    {
+      "epoch": 2.688885444117191,
+      "grad_norm": 0.3218478858470917,
+      "learning_rate": 3.2318647381801237e-07,
+      "loss": 0.311,
+      "step": 5782
+    },
+    {
+      "epoch": 2.6893504882963883,
+      "grad_norm": 0.32304203510284424,
+      "learning_rate": 3.222300959751873e-07,
+      "loss": 0.3275,
+      "step": 5783
+    },
+    {
+      "epoch": 2.6898155324755852,
+      "grad_norm": 0.33753058314323425,
+      "learning_rate": 3.2127508817518637e-07,
+      "loss": 0.3146,
+      "step": 5784
+    },
+    {
+      "epoch": 2.690280576654782,
+      "grad_norm": 0.336089164018631,
+      "learning_rate": 3.2032145069771424e-07,
+      "loss": 0.3564,
+      "step": 5785
+    },
+    {
+      "epoch": 2.6907456208339795,
+      "grad_norm": 0.3043957054615021,
+      "learning_rate": 3.1936918382207696e-07,
+      "loss": 0.3022,
+      "step": 5786
+    },
+    {
+      "epoch": 2.6912106650131764,
+      "grad_norm": 0.31595054268836975,
+      "learning_rate": 3.1841828782717685e-07,
+      "loss": 0.3224,
+      "step": 5787
+    },
+    {
+      "epoch": 2.6916757091923733,
+      "grad_norm": 0.35378995537757874,
+      "learning_rate": 3.174687629915174e-07,
+      "loss": 0.3291,
+      "step": 5788
+    },
+    {
+      "epoch": 2.69214075337157,
+      "grad_norm": 0.31758201122283936,
+      "learning_rate": 3.165206095931972e-07,
+      "loss": 0.2988,
+      "step": 5789
+    },
+    {
+      "epoch": 2.692605797550767,
+      "grad_norm": 0.3146103024482727,
+      "learning_rate": 3.1557382790991686e-07,
+      "loss": 0.332,
+      "step": 5790
+    },
+    {
+      "epoch": 2.6930708417299645,
+      "grad_norm": 0.32033929228782654,
+      "learning_rate": 3.146284182189718e-07,
+      "loss": 0.304,
+      "step": 5791
+    },
+    {
+      "epoch": 2.6935358859091614,
+      "grad_norm": 0.33226868510246277,
+      "learning_rate": 3.1368438079725784e-07,
+      "loss": 0.3309,
+      "step": 5792
+    },
+    {
+      "epoch": 2.6940009300883583,
+      "grad_norm": 0.32148411870002747,
+      "learning_rate": 3.1274171592126814e-07,
+      "loss": 0.3002,
+      "step": 5793
+    },
+    {
+      "epoch": 2.6944659742675556,
+      "grad_norm": 0.30781638622283936,
+      "learning_rate": 3.1180042386709463e-07,
+      "loss": 0.3173,
+      "step": 5794
+    },
+    {
+      "epoch": 2.6949310184467525,
+      "grad_norm": 0.31994858384132385,
+      "learning_rate": 3.108605049104246e-07,
+      "loss": 0.3421,
+      "step": 5795
+    },
+    {
+      "epoch": 2.6953960626259494,
+      "grad_norm": 0.34452393651008606,
+      "learning_rate": 3.099219593265479e-07,
+      "loss": 0.3137,
+      "step": 5796
+    },
+    {
+      "epoch": 2.6958611068051463,
+      "grad_norm": 0.30855563282966614,
+      "learning_rate": 3.089847873903462e-07,
+      "loss": 0.3054,
+      "step": 5797
+    },
+    {
+      "epoch": 2.6963261509843433,
+      "grad_norm": 0.31589534878730774,
+      "learning_rate": 3.0804898937630444e-07,
+      "loss": 0.3194,
+      "step": 5798
+    },
+    {
+      "epoch": 2.6967911951635406,
+      "grad_norm": 0.33643150329589844,
+      "learning_rate": 3.0711456555850117e-07,
+      "loss": 0.3233,
+      "step": 5799
+    },
+    {
+      "epoch": 2.6972562393427375,
+      "grad_norm": 0.34752482175827026,
+      "learning_rate": 3.0618151621061464e-07,
+      "loss": 0.3242,
+      "step": 5800
+    },
+    {
+      "epoch": 2.6977212835219344,
+      "grad_norm": 0.31614622473716736,
+      "learning_rate": 3.0524984160591963e-07,
+      "loss": 0.275,
+      "step": 5801
+    },
+    {
+      "epoch": 2.6981863277011318,
+      "grad_norm": 0.33584392070770264,
+      "learning_rate": 3.043195420172879e-07,
+      "loss": 0.3686,
+      "step": 5802
+    },
+    {
+      "epoch": 2.6986513718803287,
+      "grad_norm": 0.3025575280189514,
+      "learning_rate": 3.033906177171897e-07,
+      "loss": 0.2816,
+      "step": 5803
+    },
+    {
+      "epoch": 2.6991164160595256,
+      "grad_norm": 0.30763640999794006,
+      "learning_rate": 3.024630689776914e-07,
+      "loss": 0.294,
+      "step": 5804
+    },
+    {
+      "epoch": 2.699581460238723,
+      "grad_norm": 0.3446342647075653,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.3349,
+      "step": 5805
+    },
+    {
+      "epoch": 2.70004650441792,
+      "grad_norm": 0.3446672558784485,
+      "learning_rate": 3.006120992667499e-07,
+      "loss": 0.369,
+      "step": 5806
+    },
+    {
+      "epoch": 2.7005115485971167,
+      "grad_norm": 0.29590803384780884,
+      "learning_rate": 2.9968867883742534e-07,
+      "loss": 0.2653,
+      "step": 5807
+    },
+    {
+      "epoch": 2.7009765927763136,
+      "grad_norm": 0.33212965726852417,
+      "learning_rate": 2.9876663505293833e-07,
+      "loss": 0.3127,
+      "step": 5808
+    },
+    {
+      "epoch": 2.7014416369555105,
+      "grad_norm": 0.32688334584236145,
+      "learning_rate": 2.978459681833412e-07,
+      "loss": 0.3235,
+      "step": 5809
+    },
+    {
+      "epoch": 2.701906681134708,
+      "grad_norm": 0.3344423174858093,
+      "learning_rate": 2.969266784982822e-07,
+      "loss": 0.3131,
+      "step": 5810
+    },
+    {
+      "epoch": 2.702371725313905,
+      "grad_norm": 0.3305979371070862,
+      "learning_rate": 2.9600876626700637e-07,
+      "loss": 0.3075,
+      "step": 5811
+    },
+    {
+      "epoch": 2.7028367694931017,
+      "grad_norm": 0.36284005641937256,
+      "learning_rate": 2.9509223175835487e-07,
+      "loss": 0.3072,
+      "step": 5812
+    },
+    {
+      "epoch": 2.703301813672299,
+      "grad_norm": 0.3447756767272949,
+      "learning_rate": 2.941770752407669e-07,
+      "loss": 0.319,
+      "step": 5813
+    },
+    {
+      "epoch": 2.703766857851496,
+      "grad_norm": 0.3431076109409332,
+      "learning_rate": 2.9326329698227516e-07,
+      "loss": 0.32,
+      "step": 5814
+    },
+    {
+      "epoch": 2.704231902030693,
+      "grad_norm": 0.32309937477111816,
+      "learning_rate": 2.923508972505118e-07,
+      "loss": 0.3199,
+      "step": 5815
+    },
+    {
+      "epoch": 2.70469694620989,
+      "grad_norm": 0.3181701600551605,
+      "learning_rate": 2.9143987631270296e-07,
+      "loss": 0.311,
+      "step": 5816
+    },
+    {
+      "epoch": 2.705161990389087,
+      "grad_norm": 0.3445183038711548,
+      "learning_rate": 2.905302344356742e-07,
+      "loss": 0.3241,
+      "step": 5817
+    },
+    {
+      "epoch": 2.705627034568284,
+      "grad_norm": 0.3397235870361328,
+      "learning_rate": 2.8962197188584175e-07,
+      "loss": 0.3536,
+      "step": 5818
+    },
+    {
+      "epoch": 2.706092078747481,
+      "grad_norm": 0.3191192150115967,
+      "learning_rate": 2.8871508892922286e-07,
+      "loss": 0.3158,
+      "step": 5819
+    },
+    {
+      "epoch": 2.706557122926678,
+      "grad_norm": 0.3246058225631714,
+      "learning_rate": 2.878095858314278e-07,
+      "loss": 0.3048,
+      "step": 5820
+    },
+    {
+      "epoch": 2.707022167105875,
+      "grad_norm": 0.3111349940299988,
+      "learning_rate": 2.869054628576651e-07,
+      "loss": 0.3247,
+      "step": 5821
+    },
+    {
+      "epoch": 2.707487211285072,
+      "grad_norm": 0.3155844211578369,
+      "learning_rate": 2.860027202727361e-07,
+      "loss": 0.329,
+      "step": 5822
+    },
+    {
+      "epoch": 2.707952255464269,
+      "grad_norm": 0.2947717010974884,
+      "learning_rate": 2.851013583410406e-07,
+      "loss": 0.2873,
+      "step": 5823
+    },
+    {
+      "epoch": 2.7084172996434663,
+      "grad_norm": 0.34042876958847046,
+      "learning_rate": 2.8420137732657174e-07,
+      "loss": 0.3402,
+      "step": 5824
+    },
+    {
+      "epoch": 2.7088823438226632,
+      "grad_norm": 0.31821197271347046,
+      "learning_rate": 2.833027774929209e-07,
+      "loss": 0.3091,
+      "step": 5825
+    },
+    {
+      "epoch": 2.70934738800186,
+      "grad_norm": 0.33676040172576904,
+      "learning_rate": 2.824055591032715e-07,
+      "loss": 0.3005,
+      "step": 5826
+    },
+    {
+      "epoch": 2.709812432181057,
+      "grad_norm": 0.3211042582988739,
+      "learning_rate": 2.81509722420405e-07,
+      "loss": 0.3051,
+      "step": 5827
+    },
+    {
+      "epoch": 2.710277476360254,
+      "grad_norm": 0.3290630280971527,
+      "learning_rate": 2.8061526770669813e-07,
+      "loss": 0.3338,
+      "step": 5828
+    },
+    {
+      "epoch": 2.7107425205394513,
+      "grad_norm": 0.3298051953315735,
+      "learning_rate": 2.7972219522412194e-07,
+      "loss": 0.3061,
+      "step": 5829
+    },
+    {
+      "epoch": 2.711207564718648,
+      "grad_norm": 0.3290122449398041,
+      "learning_rate": 2.7883050523424214e-07,
+      "loss": 0.3103,
+      "step": 5830
+    },
+    {
+      "epoch": 2.711672608897845,
+      "grad_norm": 0.3085375130176544,
+      "learning_rate": 2.779401979982216e-07,
+      "loss": 0.3207,
+      "step": 5831
+    },
+    {
+      "epoch": 2.7121376530770425,
+      "grad_norm": 0.3622879385948181,
+      "learning_rate": 2.7705127377681494e-07,
+      "loss": 0.3434,
+      "step": 5832
+    },
+    {
+      "epoch": 2.7126026972562394,
+      "grad_norm": 0.30665135383605957,
+      "learning_rate": 2.7616373283037514e-07,
+      "loss": 0.2987,
+      "step": 5833
+    },
+    {
+      "epoch": 2.7130677414354363,
+      "grad_norm": 0.31220006942749023,
+      "learning_rate": 2.752775754188475e-07,
+      "loss": 0.3643,
+      "step": 5834
+    },
+    {
+      "epoch": 2.7135327856146336,
+      "grad_norm": 0.31486833095550537,
+      "learning_rate": 2.743928018017744e-07,
+      "loss": 0.3106,
+      "step": 5835
+    },
+    {
+      "epoch": 2.7139978297938305,
+      "grad_norm": 0.33274805545806885,
+      "learning_rate": 2.7350941223828975e-07,
+      "loss": 0.3206,
+      "step": 5836
+    },
+    {
+      "epoch": 2.7144628739730274,
+      "grad_norm": 0.3076656758785248,
+      "learning_rate": 2.72627406987126e-07,
+      "loss": 0.3123,
+      "step": 5837
+    },
+    {
+      "epoch": 2.7149279181522243,
+      "grad_norm": 0.3071771264076233,
+      "learning_rate": 2.71746786306607e-07,
+      "loss": 0.3229,
+      "step": 5838
+    },
+    {
+      "epoch": 2.7153929623314212,
+      "grad_norm": 0.32797494530677795,
+      "learning_rate": 2.708675504546521e-07,
+      "loss": 0.3368,
+      "step": 5839
+    },
+    {
+      "epoch": 2.7158580065106186,
+      "grad_norm": 0.30586639046669006,
+      "learning_rate": 2.699896996887763e-07,
+      "loss": 0.2811,
+      "step": 5840
+    },
+    {
+      "epoch": 2.7163230506898155,
+      "grad_norm": 0.31758421659469604,
+      "learning_rate": 2.691132342660868e-07,
+      "loss": 0.3155,
+      "step": 5841
+    },
+    {
+      "epoch": 2.7167880948690124,
+      "grad_norm": 0.34208863973617554,
+      "learning_rate": 2.682381544432866e-07,
+      "loss": 0.3223,
+      "step": 5842
+    },
+    {
+      "epoch": 2.7172531390482098,
+      "grad_norm": 0.29078277945518494,
+      "learning_rate": 2.673644604766718e-07,
+      "loss": 0.2754,
+      "step": 5843
+    },
+    {
+      "epoch": 2.7177181832274067,
+      "grad_norm": 0.33728206157684326,
+      "learning_rate": 2.664921526221348e-07,
+      "loss": 0.3452,
+      "step": 5844
+    },
+    {
+      "epoch": 2.7181832274066036,
+      "grad_norm": 0.29417553544044495,
+      "learning_rate": 2.65621231135158e-07,
+      "loss": 0.2996,
+      "step": 5845
+    },
+    {
+      "epoch": 2.718648271585801,
+      "grad_norm": 0.36140957474708557,
+      "learning_rate": 2.647516962708219e-07,
+      "loss": 0.3438,
+      "step": 5846
+    },
+    {
+      "epoch": 2.719113315764998,
+      "grad_norm": 0.30681052803993225,
+      "learning_rate": 2.6388354828379813e-07,
+      "loss": 0.3148,
+      "step": 5847
+    },
+    {
+      "epoch": 2.7195783599441947,
+      "grad_norm": 0.31248223781585693,
+      "learning_rate": 2.63016787428354e-07,
+      "loss": 0.3289,
+      "step": 5848
+    },
+    {
+      "epoch": 2.7200434041233916,
+      "grad_norm": 0.32591187953948975,
+      "learning_rate": 2.621514139583492e-07,
+      "loss": 0.3017,
+      "step": 5849
+    },
+    {
+      "epoch": 2.7205084483025885,
+      "grad_norm": 0.32240256667137146,
+      "learning_rate": 2.612874281272371e-07,
+      "loss": 0.3067,
+      "step": 5850
+    },
+    {
+      "epoch": 2.720973492481786,
+      "grad_norm": 0.3212602138519287,
+      "learning_rate": 2.6042483018806577e-07,
+      "loss": 0.2961,
+      "step": 5851
+    },
+    {
+      "epoch": 2.721438536660983,
+      "grad_norm": 0.35369014739990234,
+      "learning_rate": 2.595636203934765e-07,
+      "loss": 0.3466,
+      "step": 5852
+    },
+    {
+      "epoch": 2.7219035808401797,
+      "grad_norm": 0.3174188435077667,
+      "learning_rate": 2.587037989957031e-07,
+      "loss": 0.3181,
+      "step": 5853
+    },
+    {
+      "epoch": 2.722368625019377,
+      "grad_norm": 0.356917142868042,
+      "learning_rate": 2.5784536624657354e-07,
+      "loss": 0.3176,
+      "step": 5854
+    },
+    {
+      "epoch": 2.722833669198574,
+      "grad_norm": 0.36071768403053284,
+      "learning_rate": 2.569883223975078e-07,
+      "loss": 0.3259,
+      "step": 5855
+    },
+    {
+      "epoch": 2.723298713377771,
+      "grad_norm": 0.3190617561340332,
+      "learning_rate": 2.5613266769952183e-07,
+      "loss": 0.3215,
+      "step": 5856
+    },
+    {
+      "epoch": 2.7237637575569678,
+      "grad_norm": 0.31165647506713867,
+      "learning_rate": 2.552784024032218e-07,
+      "loss": 0.2974,
+      "step": 5857
+    },
+    {
+      "epoch": 2.7242288017361647,
+      "grad_norm": 0.32557412981987,
+      "learning_rate": 2.544255267588086e-07,
+      "loss": 0.3313,
+      "step": 5858
+    },
+    {
+      "epoch": 2.724693845915362,
+      "grad_norm": 0.329074889421463,
+      "learning_rate": 2.535740410160753e-07,
+      "loss": 0.3183,
+      "step": 5859
+    },
+    {
+      "epoch": 2.725158890094559,
+      "grad_norm": 0.3058283030986786,
+      "learning_rate": 2.5272394542440847e-07,
+      "loss": 0.3057,
+      "step": 5860
+    },
+    {
+      "epoch": 2.725623934273756,
+      "grad_norm": 0.3319613039493561,
+      "learning_rate": 2.518752402327873e-07,
+      "loss": 0.2884,
+      "step": 5861
+    },
+    {
+      "epoch": 2.726088978452953,
+      "grad_norm": 0.3359425663948059,
+      "learning_rate": 2.5102792568978354e-07,
+      "loss": 0.3352,
+      "step": 5862
+    },
+    {
+      "epoch": 2.72655402263215,
+      "grad_norm": 0.32430392503738403,
+      "learning_rate": 2.501820020435619e-07,
+      "loss": 0.341,
+      "step": 5863
+    },
+    {
+      "epoch": 2.727019066811347,
+      "grad_norm": 0.35566022992134094,
+      "learning_rate": 2.4933746954188045e-07,
+      "loss": 0.2976,
+      "step": 5864
+    },
+    {
+      "epoch": 2.7274841109905443,
+      "grad_norm": 0.3226439356803894,
+      "learning_rate": 2.4849432843208786e-07,
+      "loss": 0.3071,
+      "step": 5865
+    },
+    {
+      "epoch": 2.7279491551697412,
+      "grad_norm": 0.3440340757369995,
+      "learning_rate": 2.476525789611278e-07,
+      "loss": 0.3447,
+      "step": 5866
+    },
+    {
+      "epoch": 2.728414199348938,
+      "grad_norm": 0.32926130294799805,
+      "learning_rate": 2.4681222137553304e-07,
+      "loss": 0.3287,
+      "step": 5867
+    },
+    {
+      "epoch": 2.728879243528135,
+      "grad_norm": 0.30582261085510254,
+      "learning_rate": 2.4597325592143285e-07,
+      "loss": 0.3176,
+      "step": 5868
+    },
+    {
+      "epoch": 2.729344287707332,
+      "grad_norm": 0.3010387718677521,
+      "learning_rate": 2.4513568284454504e-07,
+      "loss": 0.3091,
+      "step": 5869
+    },
+    {
+      "epoch": 2.7298093318865293,
+      "grad_norm": 0.3204473555088043,
+      "learning_rate": 2.4429950239018285e-07,
+      "loss": 0.3201,
+      "step": 5870
+    },
+    {
+      "epoch": 2.730274376065726,
+      "grad_norm": 0.31000176072120667,
+      "learning_rate": 2.4346471480324763e-07,
+      "loss": 0.3046,
+      "step": 5871
+    },
+    {
+      "epoch": 2.730739420244923,
+      "grad_norm": 0.323900431394577,
+      "learning_rate": 2.4263132032823656e-07,
+      "loss": 0.3215,
+      "step": 5872
+    },
+    {
+      "epoch": 2.7312044644241205,
+      "grad_norm": 0.34527266025543213,
+      "learning_rate": 2.417993192092372e-07,
+      "loss": 0.3417,
+      "step": 5873
+    },
+    {
+      "epoch": 2.7316695086033174,
+      "grad_norm": 0.3277050852775574,
+      "learning_rate": 2.409687116899284e-07,
+      "loss": 0.3255,
+      "step": 5874
+    },
+    {
+      "epoch": 2.7321345527825143,
+      "grad_norm": 0.32256096601486206,
+      "learning_rate": 2.401394980135835e-07,
+      "loss": 0.301,
+      "step": 5875
+    },
+    {
+      "epoch": 2.7325995969617116,
+      "grad_norm": 0.3484511971473694,
+      "learning_rate": 2.3931167842306314e-07,
+      "loss": 0.3085,
+      "step": 5876
+    },
+    {
+      "epoch": 2.7330646411409085,
+      "grad_norm": 0.317190557718277,
+      "learning_rate": 2.3848525316082503e-07,
+      "loss": 0.3135,
+      "step": 5877
+    },
+    {
+      "epoch": 2.7335296853201054,
+      "grad_norm": 0.3190753757953644,
+      "learning_rate": 2.3766022246891284e-07,
+      "loss": 0.3283,
+      "step": 5878
+    },
+    {
+      "epoch": 2.7339947294993023,
+      "grad_norm": 0.3363284468650818,
+      "learning_rate": 2.3683658658896713e-07,
+      "loss": 0.3127,
+      "step": 5879
+    },
+    {
+      "epoch": 2.7344597736784992,
+      "grad_norm": 0.32483959197998047,
+      "learning_rate": 2.3601434576221548e-07,
+      "loss": 0.3224,
+      "step": 5880
+    },
+    {
+      "epoch": 2.7349248178576966,
+      "grad_norm": 0.29627302289009094,
+      "learning_rate": 2.3519350022948083e-07,
+      "loss": 0.3114,
+      "step": 5881
+    },
+    {
+      "epoch": 2.7353898620368935,
+      "grad_norm": 0.32418790459632874,
+      "learning_rate": 2.3437405023117366e-07,
+      "loss": 0.3299,
+      "step": 5882
+    },
+    {
+      "epoch": 2.7358549062160904,
+      "grad_norm": 0.33903124928474426,
+      "learning_rate": 2.3355599600729916e-07,
+      "loss": 0.325,
+      "step": 5883
+    },
+    {
+      "epoch": 2.7363199503952877,
+      "grad_norm": 0.2996440827846527,
+      "learning_rate": 2.3273933779745016e-07,
+      "loss": 0.2877,
+      "step": 5884
+    },
+    {
+      "epoch": 2.7367849945744847,
+      "grad_norm": 0.3519968092441559,
+      "learning_rate": 2.3192407584081423e-07,
+      "loss": 0.3545,
+      "step": 5885
+    },
+    {
+      "epoch": 2.7372500387536816,
+      "grad_norm": 0.3154807984828949,
+      "learning_rate": 2.3111021037616755e-07,
+      "loss": 0.2675,
+      "step": 5886
+    },
+    {
+      "epoch": 2.7377150829328785,
+      "grad_norm": 0.33764511346817017,
+      "learning_rate": 2.3029774164187945e-07,
+      "loss": 0.3311,
+      "step": 5887
+    },
+    {
+      "epoch": 2.738180127112076,
+      "grad_norm": 0.31895461678504944,
+      "learning_rate": 2.2948666987590683e-07,
+      "loss": 0.2989,
+      "step": 5888
+    },
+    {
+      "epoch": 2.7386451712912727,
+      "grad_norm": 0.30818265676498413,
+      "learning_rate": 2.2867699531580134e-07,
+      "loss": 0.2961,
+      "step": 5889
+    },
+    {
+      "epoch": 2.7391102154704696,
+      "grad_norm": 0.32495778799057007,
+      "learning_rate": 2.278687181987016e-07,
+      "loss": 0.3442,
+      "step": 5890
+    },
+    {
+      "epoch": 2.7395752596496665,
+      "grad_norm": 0.3362978994846344,
+      "learning_rate": 2.2706183876134047e-07,
+      "loss": 0.3452,
+      "step": 5891
+    },
+    {
+      "epoch": 2.740040303828864,
+      "grad_norm": 0.32582196593284607,
+      "learning_rate": 2.262563572400389e-07,
+      "loss": 0.3089,
+      "step": 5892
+    },
+    {
+      "epoch": 2.740505348008061,
+      "grad_norm": 0.30926793813705444,
+      "learning_rate": 2.2545227387070988e-07,
+      "loss": 0.293,
+      "step": 5893
+    },
+    {
+      "epoch": 2.7409703921872577,
+      "grad_norm": 0.3550074100494385,
+      "learning_rate": 2.2464958888885613e-07,
+      "loss": 0.3285,
+      "step": 5894
+    },
+    {
+      "epoch": 2.741435436366455,
+      "grad_norm": 0.30779311060905457,
+      "learning_rate": 2.2384830252957068e-07,
+      "loss": 0.3237,
+      "step": 5895
+    },
+    {
+      "epoch": 2.741900480545652,
+      "grad_norm": 0.32127946615219116,
+      "learning_rate": 2.2304841502753804e-07,
+      "loss": 0.3402,
+      "step": 5896
+    },
+    {
+      "epoch": 2.742365524724849,
+      "grad_norm": 0.3655702471733093,
+      "learning_rate": 2.2224992661703139e-07,
+      "loss": 0.3244,
+      "step": 5897
+    },
+    {
+      "epoch": 2.7428305689040458,
+      "grad_norm": 0.3156966269016266,
+      "learning_rate": 2.2145283753191526e-07,
+      "loss": 0.2846,
+      "step": 5898
+    },
+    {
+      "epoch": 2.7432956130832427,
+      "grad_norm": 0.32252970337867737,
+      "learning_rate": 2.206571480056452e-07,
+      "loss": 0.3368,
+      "step": 5899
+    },
+    {
+      "epoch": 2.74376065726244,
+      "grad_norm": 0.33229538798332214,
+      "learning_rate": 2.1986285827126418e-07,
+      "loss": 0.3075,
+      "step": 5900
+    },
+    {
+      "epoch": 2.744225701441637,
+      "grad_norm": 0.33600881695747375,
+      "learning_rate": 2.1906996856140783e-07,
+      "loss": 0.334,
+      "step": 5901
+    },
+    {
+      "epoch": 2.744690745620834,
+      "grad_norm": 0.4041096270084381,
+      "learning_rate": 2.1827847910830034e-07,
+      "loss": 0.3149,
+      "step": 5902
+    },
+    {
+      "epoch": 2.745155789800031,
+      "grad_norm": 0.3284103572368622,
+      "learning_rate": 2.1748839014375632e-07,
+      "loss": 0.3283,
+      "step": 5903
+    },
+    {
+      "epoch": 2.745620833979228,
+      "grad_norm": 0.2934693396091461,
+      "learning_rate": 2.16699701899179e-07,
+      "loss": 0.2902,
+      "step": 5904
+    },
+    {
+      "epoch": 2.746085878158425,
+      "grad_norm": 0.3179258108139038,
+      "learning_rate": 2.1591241460556355e-07,
+      "loss": 0.3039,
+      "step": 5905
+    },
+    {
+      "epoch": 2.7465509223376223,
+      "grad_norm": 0.32590535283088684,
+      "learning_rate": 2.151265284934928e-07,
+      "loss": 0.3358,
+      "step": 5906
+    },
+    {
+      "epoch": 2.7470159665168192,
+      "grad_norm": 0.3357870280742645,
+      "learning_rate": 2.143420437931415e-07,
+      "loss": 0.3274,
+      "step": 5907
+    },
+    {
+      "epoch": 2.747481010696016,
+      "grad_norm": 0.3348116874694824,
+      "learning_rate": 2.1355896073427028e-07,
+      "loss": 0.2949,
+      "step": 5908
+    },
+    {
+      "epoch": 2.747946054875213,
+      "grad_norm": 0.31171727180480957,
+      "learning_rate": 2.127772795462324e-07,
+      "loss": 0.3113,
+      "step": 5909
+    },
+    {
+      "epoch": 2.74841109905441,
+      "grad_norm": 0.30832639336586,
+      "learning_rate": 2.1199700045797077e-07,
+      "loss": 0.3246,
+      "step": 5910
+    },
+    {
+      "epoch": 2.7488761432336073,
+      "grad_norm": 0.3134484887123108,
+      "learning_rate": 2.112181236980143e-07,
+      "loss": 0.3223,
+      "step": 5911
+    },
+    {
+      "epoch": 2.749341187412804,
+      "grad_norm": 0.32488688826560974,
+      "learning_rate": 2.104406494944855e-07,
+      "loss": 0.3119,
+      "step": 5912
+    },
+    {
+      "epoch": 2.749806231592001,
+      "grad_norm": 0.3216099739074707,
+      "learning_rate": 2.0966457807509222e-07,
+      "loss": 0.3364,
+      "step": 5913
+    },
+    {
+      "epoch": 2.7502712757711985,
+      "grad_norm": 0.32613542675971985,
+      "learning_rate": 2.088899096671343e-07,
+      "loss": 0.3025,
+      "step": 5914
+    },
+    {
+      "epoch": 2.7507363199503954,
+      "grad_norm": 0.32675376534461975,
+      "learning_rate": 2.0811664449749857e-07,
+      "loss": 0.3239,
+      "step": 5915
+    },
+    {
+      "epoch": 2.7512013641295923,
+      "grad_norm": 0.28670793771743774,
+      "learning_rate": 2.073447827926628e-07,
+      "loss": 0.2655,
+      "step": 5916
+    },
+    {
+      "epoch": 2.751666408308789,
+      "grad_norm": 0.31409409642219543,
+      "learning_rate": 2.0657432477869165e-07,
+      "loss": 0.3476,
+      "step": 5917
+    },
+    {
+      "epoch": 2.7521314524879865,
+      "grad_norm": 0.3178742229938507,
+      "learning_rate": 2.0580527068124134e-07,
+      "loss": 0.3484,
+      "step": 5918
+    },
+    {
+      "epoch": 2.7525964966671834,
+      "grad_norm": 0.3407832980155945,
+      "learning_rate": 2.0503762072555387e-07,
+      "loss": 0.329,
+      "step": 5919
+    },
+    {
+      "epoch": 2.7530615408463803,
+      "grad_norm": 0.3183245360851288,
+      "learning_rate": 2.0427137513646167e-07,
+      "loss": 0.3063,
+      "step": 5920
+    },
+    {
+      "epoch": 2.7535265850255772,
+      "grad_norm": 0.3081268072128296,
+      "learning_rate": 2.0350653413838573e-07,
+      "loss": 0.3044,
+      "step": 5921
+    },
+    {
+      "epoch": 2.7539916292047746,
+      "grad_norm": 0.31228625774383545,
+      "learning_rate": 2.0274309795533687e-07,
+      "loss": 0.3156,
+      "step": 5922
+    },
+    {
+      "epoch": 2.7544566733839715,
+      "grad_norm": 0.3150801956653595,
+      "learning_rate": 2.0198106681091124e-07,
+      "loss": 0.3221,
+      "step": 5923
+    },
+    {
+      "epoch": 2.7549217175631684,
+      "grad_norm": 0.2927737534046173,
+      "learning_rate": 2.012204409282964e-07,
+      "loss": 0.2869,
+      "step": 5924
+    },
+    {
+      "epoch": 2.7553867617423657,
+      "grad_norm": 0.32298728823661804,
+      "learning_rate": 2.0046122053026697e-07,
+      "loss": 0.3349,
+      "step": 5925
+    },
+    {
+      "epoch": 2.7558518059215626,
+      "grad_norm": 0.34636053442955017,
+      "learning_rate": 1.9970340583918668e-07,
+      "loss": 0.33,
+      "step": 5926
+    },
+    {
+      "epoch": 2.7563168501007596,
+      "grad_norm": 0.3538902699947357,
+      "learning_rate": 1.989469970770064e-07,
+      "loss": 0.3174,
+      "step": 5927
+    },
+    {
+      "epoch": 2.7567818942799565,
+      "grad_norm": 0.32646122574806213,
+      "learning_rate": 1.9819199446526716e-07,
+      "loss": 0.2947,
+      "step": 5928
+    },
+    {
+      "epoch": 2.7572469384591534,
+      "grad_norm": 0.35514765977859497,
+      "learning_rate": 1.9743839822509547e-07,
+      "loss": 0.3108,
+      "step": 5929
+    },
+    {
+      "epoch": 2.7577119826383507,
+      "grad_norm": 0.31268438696861267,
+      "learning_rate": 1.9668620857720865e-07,
+      "loss": 0.3144,
+      "step": 5930
+    },
+    {
+      "epoch": 2.7581770268175476,
+      "grad_norm": 0.2988133728504181,
+      "learning_rate": 1.9593542574190993e-07,
+      "loss": 0.2633,
+      "step": 5931
+    },
+    {
+      "epoch": 2.7586420709967445,
+      "grad_norm": 0.3349534869194031,
+      "learning_rate": 1.9518604993909175e-07,
+      "loss": 0.3689,
+      "step": 5932
+    },
+    {
+      "epoch": 2.759107115175942,
+      "grad_norm": 0.3589261472225189,
+      "learning_rate": 1.9443808138823404e-07,
+      "loss": 0.3216,
+      "step": 5933
+    },
+    {
+      "epoch": 2.7595721593551388,
+      "grad_norm": 0.3319076895713806,
+      "learning_rate": 1.9369152030840553e-07,
+      "loss": 0.3064,
+      "step": 5934
+    },
+    {
+      "epoch": 2.7600372035343357,
+      "grad_norm": 0.30595728754997253,
+      "learning_rate": 1.9294636691826073e-07,
+      "loss": 0.3035,
+      "step": 5935
+    },
+    {
+      "epoch": 2.760502247713533,
+      "grad_norm": 0.32479432225227356,
+      "learning_rate": 1.9220262143604395e-07,
+      "loss": 0.317,
+      "step": 5936
+    },
+    {
+      "epoch": 2.76096729189273,
+      "grad_norm": 0.33985117077827454,
+      "learning_rate": 1.9146028407958483e-07,
+      "loss": 0.3118,
+      "step": 5937
+    },
+    {
+      "epoch": 2.761432336071927,
+      "grad_norm": 0.32750189304351807,
+      "learning_rate": 1.907193550663028e-07,
+      "loss": 0.3164,
+      "step": 5938
+    },
+    {
+      "epoch": 2.7618973802511237,
+      "grad_norm": 0.3345829248428345,
+      "learning_rate": 1.899798346132037e-07,
+      "loss": 0.3506,
+      "step": 5939
+    },
+    {
+      "epoch": 2.7623624244303207,
+      "grad_norm": 0.31959566473960876,
+      "learning_rate": 1.8924172293688148e-07,
+      "loss": 0.2971,
+      "step": 5940
+    },
+    {
+      "epoch": 2.762827468609518,
+      "grad_norm": 0.3132396638393402,
+      "learning_rate": 1.885050202535166e-07,
+      "loss": 0.2998,
+      "step": 5941
+    },
+    {
+      "epoch": 2.763292512788715,
+      "grad_norm": 0.3381306231021881,
+      "learning_rate": 1.877697267788775e-07,
+      "loss": 0.3455,
+      "step": 5942
+    },
+    {
+      "epoch": 2.763757556967912,
+      "grad_norm": 0.3132924735546112,
+      "learning_rate": 1.870358427283192e-07,
+      "loss": 0.3172,
+      "step": 5943
+    },
+    {
+      "epoch": 2.764222601147109,
+      "grad_norm": 0.303196519613266,
+      "learning_rate": 1.8630336831678475e-07,
+      "loss": 0.3258,
+      "step": 5944
+    },
+    {
+      "epoch": 2.764687645326306,
+      "grad_norm": 0.3348020017147064,
+      "learning_rate": 1.8557230375880364e-07,
+      "loss": 0.3243,
+      "step": 5945
+    },
+    {
+      "epoch": 2.765152689505503,
+      "grad_norm": 0.33610332012176514,
+      "learning_rate": 1.848426492684946e-07,
+      "loss": 0.362,
+      "step": 5946
+    },
+    {
+      "epoch": 2.7656177336847003,
+      "grad_norm": 0.31693512201309204,
+      "learning_rate": 1.8411440505956e-07,
+      "loss": 0.2884,
+      "step": 5947
+    },
+    {
+      "epoch": 2.7660827778638972,
+      "grad_norm": 0.3176809847354889,
+      "learning_rate": 1.833875713452904e-07,
+      "loss": 0.3094,
+      "step": 5948
+    },
+    {
+      "epoch": 2.766547822043094,
+      "grad_norm": 0.3155810534954071,
+      "learning_rate": 1.8266214833856432e-07,
+      "loss": 0.3281,
+      "step": 5949
+    },
+    {
+      "epoch": 2.767012866222291,
+      "grad_norm": 0.31794315576553345,
+      "learning_rate": 1.819381362518463e-07,
+      "loss": 0.3407,
+      "step": 5950
+    },
+    {
+      "epoch": 2.767477910401488,
+      "grad_norm": 0.2952190935611725,
+      "learning_rate": 1.8121553529718782e-07,
+      "loss": 0.296,
+      "step": 5951
+    },
+    {
+      "epoch": 2.7679429545806853,
+      "grad_norm": 0.32220354676246643,
+      "learning_rate": 1.8049434568622627e-07,
+      "loss": 0.3271,
+      "step": 5952
+    },
+    {
+      "epoch": 2.768407998759882,
+      "grad_norm": 0.3344385027885437,
+      "learning_rate": 1.7977456763018764e-07,
+      "loss": 0.321,
+      "step": 5953
+    },
+    {
+      "epoch": 2.768873042939079,
+      "grad_norm": 0.330684095621109,
+      "learning_rate": 1.7905620133988166e-07,
+      "loss": 0.3028,
+      "step": 5954
+    },
+    {
+      "epoch": 2.7693380871182764,
+      "grad_norm": 0.31999465823173523,
+      "learning_rate": 1.7833924702570725e-07,
+      "loss": 0.3465,
+      "step": 5955
+    },
+    {
+      "epoch": 2.7698031312974734,
+      "grad_norm": 0.28892767429351807,
+      "learning_rate": 1.7762370489764813e-07,
+      "loss": 0.3031,
+      "step": 5956
+    },
+    {
+      "epoch": 2.7702681754766703,
+      "grad_norm": 0.3207509219646454,
+      "learning_rate": 1.7690957516527607e-07,
+      "loss": 0.322,
+      "step": 5957
+    },
+    {
+      "epoch": 2.770733219655867,
+      "grad_norm": 0.32161515951156616,
+      "learning_rate": 1.76196858037746e-07,
+      "loss": 0.3114,
+      "step": 5958
+    },
+    {
+      "epoch": 2.771198263835064,
+      "grad_norm": 0.3222302496433258,
+      "learning_rate": 1.7548555372380372e-07,
+      "loss": 0.3151,
+      "step": 5959
+    },
+    {
+      "epoch": 2.7716633080142614,
+      "grad_norm": 0.3225840926170349,
+      "learning_rate": 1.7477566243177647e-07,
+      "loss": 0.301,
+      "step": 5960
+    },
+    {
+      "epoch": 2.7721283521934583,
+      "grad_norm": 0.33795180916786194,
+      "learning_rate": 1.740671843695818e-07,
+      "loss": 0.3138,
+      "step": 5961
+    },
+    {
+      "epoch": 2.7725933963726552,
+      "grad_norm": 0.3079066574573517,
+      "learning_rate": 1.7336011974471933e-07,
+      "loss": 0.3115,
+      "step": 5962
+    },
+    {
+      "epoch": 2.7730584405518526,
+      "grad_norm": 0.317829430103302,
+      "learning_rate": 1.7265446876427895e-07,
+      "loss": 0.3104,
+      "step": 5963
+    },
+    {
+      "epoch": 2.7735234847310495,
+      "grad_norm": 0.32310232520103455,
+      "learning_rate": 1.7195023163493253e-07,
+      "loss": 0.318,
+      "step": 5964
+    },
+    {
+      "epoch": 2.7739885289102464,
+      "grad_norm": 0.3459528386592865,
+      "learning_rate": 1.712474085629412e-07,
+      "loss": 0.3189,
+      "step": 5965
+    },
+    {
+      "epoch": 2.7744535730894437,
+      "grad_norm": 0.3017194867134094,
+      "learning_rate": 1.7054599975414866e-07,
+      "loss": 0.3033,
+      "step": 5966
+    },
+    {
+      "epoch": 2.7749186172686406,
+      "grad_norm": 0.32311463356018066,
+      "learning_rate": 1.6984600541398777e-07,
+      "loss": 0.3371,
+      "step": 5967
+    },
+    {
+      "epoch": 2.7753836614478375,
+      "grad_norm": 0.29081717133522034,
+      "learning_rate": 1.6914742574747455e-07,
+      "loss": 0.3338,
+      "step": 5968
+    },
+    {
+      "epoch": 2.7758487056270345,
+      "grad_norm": 0.3164746165275574,
+      "learning_rate": 1.6845026095921314e-07,
+      "loss": 0.3281,
+      "step": 5969
+    },
+    {
+      "epoch": 2.7763137498062314,
+      "grad_norm": 0.3146812915802002,
+      "learning_rate": 1.677545112533896e-07,
+      "loss": 0.3042,
+      "step": 5970
+    },
+    {
+      "epoch": 2.7767787939854287,
+      "grad_norm": 0.3127114474773407,
+      "learning_rate": 1.6706017683377928e-07,
+      "loss": 0.2914,
+      "step": 5971
+    },
+    {
+      "epoch": 2.7772438381646256,
+      "grad_norm": 0.3269754946231842,
+      "learning_rate": 1.663672579037412e-07,
+      "loss": 0.3242,
+      "step": 5972
+    },
+    {
+      "epoch": 2.7777088823438225,
+      "grad_norm": 0.323904424905777,
+      "learning_rate": 1.6567575466621964e-07,
+      "loss": 0.2977,
+      "step": 5973
+    },
+    {
+      "epoch": 2.77817392652302,
+      "grad_norm": 0.3124120235443115,
+      "learning_rate": 1.6498566732374433e-07,
+      "loss": 0.328,
+      "step": 5974
+    },
+    {
+      "epoch": 2.7786389707022168,
+      "grad_norm": 0.31669923663139343,
+      "learning_rate": 1.6429699607843185e-07,
+      "loss": 0.3143,
+      "step": 5975
+    },
+    {
+      "epoch": 2.7791040148814137,
+      "grad_norm": 0.3284464478492737,
+      "learning_rate": 1.6360974113198203e-07,
+      "loss": 0.3065,
+      "step": 5976
+    },
+    {
+      "epoch": 2.779569059060611,
+      "grad_norm": 0.31875741481781006,
+      "learning_rate": 1.6292390268568103e-07,
+      "loss": 0.316,
+      "step": 5977
+    },
+    {
+      "epoch": 2.780034103239808,
+      "grad_norm": 0.34978044033050537,
+      "learning_rate": 1.6223948094039876e-07,
+      "loss": 0.3321,
+      "step": 5978
+    },
+    {
+      "epoch": 2.780499147419005,
+      "grad_norm": 0.32049715518951416,
+      "learning_rate": 1.615564760965921e-07,
+      "loss": 0.3292,
+      "step": 5979
+    },
+    {
+      "epoch": 2.7809641915982017,
+      "grad_norm": 0.30878329277038574,
+      "learning_rate": 1.6087488835430208e-07,
+      "loss": 0.3102,
+      "step": 5980
+    },
+    {
+      "epoch": 2.7814292357773986,
+      "grad_norm": 0.3212791979312897,
+      "learning_rate": 1.6019471791315522e-07,
+      "loss": 0.2984,
+      "step": 5981
+    },
+    {
+      "epoch": 2.781894279956596,
+      "grad_norm": 0.3322344124317169,
+      "learning_rate": 1.5951596497236154e-07,
+      "loss": 0.3274,
+      "step": 5982
+    },
+    {
+      "epoch": 2.782359324135793,
+      "grad_norm": 0.3315823972225189,
+      "learning_rate": 1.5883862973071652e-07,
+      "loss": 0.3367,
+      "step": 5983
+    },
+    {
+      "epoch": 2.78282436831499,
+      "grad_norm": 0.33415260910987854,
+      "learning_rate": 1.5816271238660196e-07,
+      "loss": 0.3106,
+      "step": 5984
+    },
+    {
+      "epoch": 2.783289412494187,
+      "grad_norm": 0.31730806827545166,
+      "learning_rate": 1.5748821313798124e-07,
+      "loss": 0.3076,
+      "step": 5985
+    },
+    {
+      "epoch": 2.783754456673384,
+      "grad_norm": 0.3219851553440094,
+      "learning_rate": 1.5681513218240573e-07,
+      "loss": 0.3218,
+      "step": 5986
+    },
+    {
+      "epoch": 2.784219500852581,
+      "grad_norm": 0.30945420265197754,
+      "learning_rate": 1.5614346971700945e-07,
+      "loss": 0.3242,
+      "step": 5987
+    },
+    {
+      "epoch": 2.784684545031778,
+      "grad_norm": 0.3280915915966034,
+      "learning_rate": 1.554732259385111e-07,
+      "loss": 0.3007,
+      "step": 5988
+    },
+    {
+      "epoch": 2.7851495892109748,
+      "grad_norm": 0.320148229598999,
+      "learning_rate": 1.5480440104321481e-07,
+      "loss": 0.3254,
+      "step": 5989
+    },
+    {
+      "epoch": 2.785614633390172,
+      "grad_norm": 0.31728118658065796,
+      "learning_rate": 1.5413699522700775e-07,
+      "loss": 0.3202,
+      "step": 5990
+    },
+    {
+      "epoch": 2.786079677569369,
+      "grad_norm": 0.30805543065071106,
+      "learning_rate": 1.5347100868536246e-07,
+      "loss": 0.3265,
+      "step": 5991
+    },
+    {
+      "epoch": 2.786544721748566,
+      "grad_norm": 0.3176276385784149,
+      "learning_rate": 1.5280644161333625e-07,
+      "loss": 0.3275,
+      "step": 5992
+    },
+    {
+      "epoch": 2.7870097659277633,
+      "grad_norm": 0.3044149577617645,
+      "learning_rate": 1.521432942055695e-07,
+      "loss": 0.297,
+      "step": 5993
+    },
+    {
+      "epoch": 2.78747481010696,
+      "grad_norm": 0.3212355971336365,
+      "learning_rate": 1.51481566656288e-07,
+      "loss": 0.3568,
+      "step": 5994
+    },
+    {
+      "epoch": 2.787939854286157,
+      "grad_norm": 0.3032307028770447,
+      "learning_rate": 1.5082125915929946e-07,
+      "loss": 0.3037,
+      "step": 5995
+    },
+    {
+      "epoch": 2.7884048984653544,
+      "grad_norm": 0.3063505291938782,
+      "learning_rate": 1.5016237190799866e-07,
+      "loss": 0.2971,
+      "step": 5996
+    },
+    {
+      "epoch": 2.7888699426445513,
+      "grad_norm": 0.2914872169494629,
+      "learning_rate": 1.4950490509536176e-07,
+      "loss": 0.313,
+      "step": 5997
+    },
+    {
+      "epoch": 2.7893349868237483,
+      "grad_norm": 0.3118445575237274,
+      "learning_rate": 1.4884885891395196e-07,
+      "loss": 0.3192,
+      "step": 5998
+    },
+    {
+      "epoch": 2.789800031002945,
+      "grad_norm": 0.33914464712142944,
+      "learning_rate": 1.4819423355591223e-07,
+      "loss": 0.348,
+      "step": 5999
+    },
+    {
+      "epoch": 2.790265075182142,
+      "grad_norm": 0.32917240262031555,
+      "learning_rate": 1.4754102921297363e-07,
+      "loss": 0.3033,
+      "step": 6000
+    },
+    {
+      "epoch": 2.7907301193613394,
+      "grad_norm": 0.33126407861709595,
+      "learning_rate": 1.4688924607644817e-07,
+      "loss": 0.312,
+      "step": 6001
+    },
+    {
+      "epoch": 2.7911951635405363,
+      "grad_norm": 0.3238784372806549,
+      "learning_rate": 1.46238884337232e-07,
+      "loss": 0.3091,
+      "step": 6002
+    },
+    {
+      "epoch": 2.7916602077197332,
+      "grad_norm": 0.32593873143196106,
+      "learning_rate": 1.4558994418580663e-07,
+      "loss": 0.3353,
+      "step": 6003
+    },
+    {
+      "epoch": 2.7921252518989306,
+      "grad_norm": 0.2928268611431122,
+      "learning_rate": 1.4494242581223615e-07,
+      "loss": 0.2997,
+      "step": 6004
+    },
+    {
+      "epoch": 2.7925902960781275,
+      "grad_norm": 0.3363851308822632,
+      "learning_rate": 1.4429632940616721e-07,
+      "loss": 0.3243,
+      "step": 6005
+    },
+    {
+      "epoch": 2.7930553402573244,
+      "grad_norm": 0.3205671012401581,
+      "learning_rate": 1.4365165515683176e-07,
+      "loss": 0.3337,
+      "step": 6006
+    },
+    {
+      "epoch": 2.7935203844365217,
+      "grad_norm": 0.3109828233718872,
+      "learning_rate": 1.4300840325304377e-07,
+      "loss": 0.2985,
+      "step": 6007
+    },
+    {
+      "epoch": 2.7939854286157186,
+      "grad_norm": 0.29912734031677246,
+      "learning_rate": 1.4236657388320198e-07,
+      "loss": 0.3093,
+      "step": 6008
+    },
+    {
+      "epoch": 2.7944504727949155,
+      "grad_norm": 0.3295075297355652,
+      "learning_rate": 1.417261672352871e-07,
+      "loss": 0.3717,
+      "step": 6009
+    },
+    {
+      "epoch": 2.7949155169741124,
+      "grad_norm": 0.29881277680397034,
+      "learning_rate": 1.4108718349686468e-07,
+      "loss": 0.3031,
+      "step": 6010
+    },
+    {
+      "epoch": 2.7953805611533094,
+      "grad_norm": 0.32853347063064575,
+      "learning_rate": 1.4044962285508113e-07,
+      "loss": 0.3193,
+      "step": 6011
+    },
+    {
+      "epoch": 2.7958456053325067,
+      "grad_norm": 0.32658717036247253,
+      "learning_rate": 1.3981348549666928e-07,
+      "loss": 0.3161,
+      "step": 6012
+    },
+    {
+      "epoch": 2.7963106495117036,
+      "grad_norm": 0.3037392795085907,
+      "learning_rate": 1.3917877160794236e-07,
+      "loss": 0.3352,
+      "step": 6013
+    },
+    {
+      "epoch": 2.7967756936909005,
+      "grad_norm": 0.3436054289340973,
+      "learning_rate": 1.385454813747983e-07,
+      "loss": 0.328,
+      "step": 6014
+    },
+    {
+      "epoch": 2.797240737870098,
+      "grad_norm": 0.2963773012161255,
+      "learning_rate": 1.3791361498271704e-07,
+      "loss": 0.2987,
+      "step": 6015
+    },
+    {
+      "epoch": 2.7977057820492948,
+      "grad_norm": 0.3178195655345917,
+      "learning_rate": 1.3728317261676338e-07,
+      "loss": 0.3486,
+      "step": 6016
+    },
+    {
+      "epoch": 2.7981708262284917,
+      "grad_norm": 0.39669740200042725,
+      "learning_rate": 1.3665415446158182e-07,
+      "loss": 0.3288,
+      "step": 6017
+    },
+    {
+      "epoch": 2.7986358704076886,
+      "grad_norm": 0.3178330659866333,
+      "learning_rate": 1.3602656070140275e-07,
+      "loss": 0.3005,
+      "step": 6018
+    },
+    {
+      "epoch": 2.7991009145868855,
+      "grad_norm": 0.31485146284103394,
+      "learning_rate": 1.354003915200375e-07,
+      "loss": 0.3319,
+      "step": 6019
+    },
+    {
+      "epoch": 2.799565958766083,
+      "grad_norm": 0.31439441442489624,
+      "learning_rate": 1.3477564710088097e-07,
+      "loss": 0.2969,
+      "step": 6020
+    },
+    {
+      "epoch": 2.8000310029452797,
+      "grad_norm": 0.3106626570224762,
+      "learning_rate": 1.3415232762691134e-07,
+      "loss": 0.3352,
+      "step": 6021
+    },
+    {
+      "epoch": 2.8004960471244766,
+      "grad_norm": 0.3026930093765259,
+      "learning_rate": 1.335304332806875e-07,
+      "loss": 0.312,
+      "step": 6022
+    },
+    {
+      "epoch": 2.800961091303674,
+      "grad_norm": 0.31321385502815247,
+      "learning_rate": 1.3290996424435375e-07,
+      "loss": 0.2761,
+      "step": 6023
+    },
+    {
+      "epoch": 2.801426135482871,
+      "grad_norm": 0.32276248931884766,
+      "learning_rate": 1.3229092069963368e-07,
+      "loss": 0.3453,
+      "step": 6024
+    },
+    {
+      "epoch": 2.801891179662068,
+      "grad_norm": 0.3255274295806885,
+      "learning_rate": 1.3167330282783608e-07,
+      "loss": 0.3046,
+      "step": 6025
+    },
+    {
+      "epoch": 2.802356223841265,
+      "grad_norm": 0.32487818598747253,
+      "learning_rate": 1.3105711080985128e-07,
+      "loss": 0.3285,
+      "step": 6026
+    },
+    {
+      "epoch": 2.802821268020462,
+      "grad_norm": 0.43458884954452515,
+      "learning_rate": 1.3044234482615216e-07,
+      "loss": 0.3202,
+      "step": 6027
+    },
+    {
+      "epoch": 2.803286312199659,
+      "grad_norm": 0.32257169485092163,
+      "learning_rate": 1.298290050567924e-07,
+      "loss": 0.3324,
+      "step": 6028
+    },
+    {
+      "epoch": 2.803751356378856,
+      "grad_norm": 0.3104286193847656,
+      "learning_rate": 1.2921709168141116e-07,
+      "loss": 0.3087,
+      "step": 6029
+    },
+    {
+      "epoch": 2.8042164005580528,
+      "grad_norm": 0.3442256450653076,
+      "learning_rate": 1.2860660487922616e-07,
+      "loss": 0.3041,
+      "step": 6030
+    },
+    {
+      "epoch": 2.80468144473725,
+      "grad_norm": 0.33402782678604126,
+      "learning_rate": 1.2799754482903992e-07,
+      "loss": 0.3278,
+      "step": 6031
+    },
+    {
+      "epoch": 2.805146488916447,
+      "grad_norm": 0.3222576379776001,
+      "learning_rate": 1.2738991170923588e-07,
+      "loss": 0.2998,
+      "step": 6032
+    },
+    {
+      "epoch": 2.805611533095644,
+      "grad_norm": 0.3217984139919281,
+      "learning_rate": 1.2678370569778052e-07,
+      "loss": 0.3382,
+      "step": 6033
+    },
+    {
+      "epoch": 2.8060765772748413,
+      "grad_norm": 0.3216967284679413,
+      "learning_rate": 1.2617892697222135e-07,
+      "loss": 0.3328,
+      "step": 6034
+    },
+    {
+      "epoch": 2.806541621454038,
+      "grad_norm": 0.3223290741443634,
+      "learning_rate": 1.2557557570968825e-07,
+      "loss": 0.3037,
+      "step": 6035
+    },
+    {
+      "epoch": 2.807006665633235,
+      "grad_norm": 0.3248218894004822,
+      "learning_rate": 1.2497365208689272e-07,
+      "loss": 0.3154,
+      "step": 6036
+    },
+    {
+      "epoch": 2.8074717098124324,
+      "grad_norm": 0.2976897358894348,
+      "learning_rate": 1.2437315628012868e-07,
+      "loss": 0.2948,
+      "step": 6037
+    },
+    {
+      "epoch": 2.8079367539916293,
+      "grad_norm": 0.34811344742774963,
+      "learning_rate": 1.2377408846527105e-07,
+      "loss": 0.3377,
+      "step": 6038
+    },
+    {
+      "epoch": 2.8084017981708262,
+      "grad_norm": 0.3315483033657074,
+      "learning_rate": 1.231764488177789e-07,
+      "loss": 0.3402,
+      "step": 6039
+    },
+    {
+      "epoch": 2.808866842350023,
+      "grad_norm": 0.3172501027584076,
+      "learning_rate": 1.225802375126889e-07,
+      "loss": 0.3076,
+      "step": 6040
+    },
+    {
+      "epoch": 2.80933188652922,
+      "grad_norm": 0.3447789251804352,
+      "learning_rate": 1.2198545472462297e-07,
+      "loss": 0.3221,
+      "step": 6041
+    },
+    {
+      "epoch": 2.8097969307084174,
+      "grad_norm": 0.3011327385902405,
+      "learning_rate": 1.2139210062778294e-07,
+      "loss": 0.3261,
+      "step": 6042
+    },
+    {
+      "epoch": 2.8102619748876143,
+      "grad_norm": 0.2930125594139099,
+      "learning_rate": 1.2080017539595312e-07,
+      "loss": 0.294,
+      "step": 6043
+    },
+    {
+      "epoch": 2.810727019066811,
+      "grad_norm": 0.32197386026382446,
+      "learning_rate": 1.202096792024976e-07,
+      "loss": 0.3191,
+      "step": 6044
+    },
+    {
+      "epoch": 2.8111920632460086,
+      "grad_norm": 0.3363608717918396,
+      "learning_rate": 1.196206122203647e-07,
+      "loss": 0.3183,
+      "step": 6045
+    },
+    {
+      "epoch": 2.8116571074252055,
+      "grad_norm": 0.32255131006240845,
+      "learning_rate": 1.1903297462208085e-07,
+      "loss": 0.3075,
+      "step": 6046
+    },
+    {
+      "epoch": 2.8121221516044024,
+      "grad_norm": 0.32174059748649597,
+      "learning_rate": 1.1844676657975673e-07,
+      "loss": 0.3559,
+      "step": 6047
+    },
+    {
+      "epoch": 2.8125871957835993,
+      "grad_norm": 0.31685617566108704,
+      "learning_rate": 1.1786198826508277e-07,
+      "loss": 0.2865,
+      "step": 6048
+    },
+    {
+      "epoch": 2.8130522399627966,
+      "grad_norm": 0.3263781666755676,
+      "learning_rate": 1.1727863984933086e-07,
+      "loss": 0.313,
+      "step": 6049
+    },
+    {
+      "epoch": 2.8135172841419935,
+      "grad_norm": 0.3233252763748169,
+      "learning_rate": 1.1669672150335487e-07,
+      "loss": 0.3147,
+      "step": 6050
+    },
+    {
+      "epoch": 2.8139823283211904,
+      "grad_norm": 0.32005932927131653,
+      "learning_rate": 1.1611623339758904e-07,
+      "loss": 0.3125,
+      "step": 6051
+    },
+    {
+      "epoch": 2.8144473725003873,
+      "grad_norm": 0.33904585242271423,
+      "learning_rate": 1.1553717570204847e-07,
+      "loss": 0.301,
+      "step": 6052
+    },
+    {
+      "epoch": 2.8149124166795847,
+      "grad_norm": 0.33245572447776794,
+      "learning_rate": 1.149595485863303e-07,
+      "loss": 0.3283,
+      "step": 6053
+    },
+    {
+      "epoch": 2.8153774608587816,
+      "grad_norm": 0.3273317515850067,
+      "learning_rate": 1.1438335221961195e-07,
+      "loss": 0.3298,
+      "step": 6054
+    },
+    {
+      "epoch": 2.8158425050379785,
+      "grad_norm": 0.31557828187942505,
+      "learning_rate": 1.1380858677065177e-07,
+      "loss": 0.3146,
+      "step": 6055
+    },
+    {
+      "epoch": 2.816307549217176,
+      "grad_norm": 0.29916810989379883,
+      "learning_rate": 1.1323525240778954e-07,
+      "loss": 0.3233,
+      "step": 6056
+    },
+    {
+      "epoch": 2.8167725933963728,
+      "grad_norm": 0.3167668282985687,
+      "learning_rate": 1.1266334929894485e-07,
+      "loss": 0.3322,
+      "step": 6057
+    },
+    {
+      "epoch": 2.8172376375755697,
+      "grad_norm": 0.2994489371776581,
+      "learning_rate": 1.120928776116198e-07,
+      "loss": 0.3082,
+      "step": 6058
+    },
+    {
+      "epoch": 2.8177026817547666,
+      "grad_norm": 0.31627991795539856,
+      "learning_rate": 1.1152383751289575e-07,
+      "loss": 0.3136,
+      "step": 6059
+    },
+    {
+      "epoch": 2.8181677259339635,
+      "grad_norm": 0.3162107467651367,
+      "learning_rate": 1.1095622916943494e-07,
+      "loss": 0.3221,
+      "step": 6060
+    },
+    {
+      "epoch": 2.818632770113161,
+      "grad_norm": 0.31993138790130615,
+      "learning_rate": 1.103900527474816e-07,
+      "loss": 0.3206,
+      "step": 6061
+    },
+    {
+      "epoch": 2.8190978142923577,
+      "grad_norm": 0.3255464732646942,
+      "learning_rate": 1.0982530841285921e-07,
+      "loss": 0.3131,
+      "step": 6062
+    },
+    {
+      "epoch": 2.8195628584715546,
+      "grad_norm": 0.317570298910141,
+      "learning_rate": 1.0926199633097156e-07,
+      "loss": 0.2974,
+      "step": 6063
+    },
+    {
+      "epoch": 2.820027902650752,
+      "grad_norm": 0.3149189352989197,
+      "learning_rate": 1.0870011666680502e-07,
+      "loss": 0.2871,
+      "step": 6064
+    },
+    {
+      "epoch": 2.820492946829949,
+      "grad_norm": 0.3116709589958191,
+      "learning_rate": 1.081396695849235e-07,
+      "loss": 0.3331,
+      "step": 6065
+    },
+    {
+      "epoch": 2.820957991009146,
+      "grad_norm": 0.3222356140613556,
+      "learning_rate": 1.0758065524947403e-07,
+      "loss": 0.328,
+      "step": 6066
+    },
+    {
+      "epoch": 2.821423035188343,
+      "grad_norm": 0.2908288538455963,
+      "learning_rate": 1.0702307382418175e-07,
+      "loss": 0.2997,
+      "step": 6067
+    },
+    {
+      "epoch": 2.82188807936754,
+      "grad_norm": 0.3209805190563202,
+      "learning_rate": 1.0646692547235437e-07,
+      "loss": 0.3317,
+      "step": 6068
+    },
+    {
+      "epoch": 2.822353123546737,
+      "grad_norm": 0.3027243912220001,
+      "learning_rate": 1.0591221035687716e-07,
+      "loss": 0.3113,
+      "step": 6069
+    },
+    {
+      "epoch": 2.822818167725934,
+      "grad_norm": 0.30290907621383667,
+      "learning_rate": 1.0535892864021901e-07,
+      "loss": 0.3267,
+      "step": 6070
+    },
+    {
+      "epoch": 2.8232832119051308,
+      "grad_norm": 0.3006175756454468,
+      "learning_rate": 1.0480708048442589e-07,
+      "loss": 0.3119,
+      "step": 6071
+    },
+    {
+      "epoch": 2.823748256084328,
+      "grad_norm": 0.34534621238708496,
+      "learning_rate": 1.0425666605112516e-07,
+      "loss": 0.3862,
+      "step": 6072
+    },
+    {
+      "epoch": 2.824213300263525,
+      "grad_norm": 0.317735493183136,
+      "learning_rate": 1.0370768550152454e-07,
+      "loss": 0.3082,
+      "step": 6073
+    },
+    {
+      "epoch": 2.824678344442722,
+      "grad_norm": 0.36195850372314453,
+      "learning_rate": 1.0316013899641264e-07,
+      "loss": 0.3223,
+      "step": 6074
+    },
+    {
+      "epoch": 2.8251433886219193,
+      "grad_norm": 0.3154940903186798,
+      "learning_rate": 1.0261402669615505e-07,
+      "loss": 0.3041,
+      "step": 6075
+    },
+    {
+      "epoch": 2.825608432801116,
+      "grad_norm": 0.33876118063926697,
+      "learning_rate": 1.0206934876070052e-07,
+      "loss": 0.3145,
+      "step": 6076
+    },
+    {
+      "epoch": 2.826073476980313,
+      "grad_norm": 0.3257219195365906,
+      "learning_rate": 1.015261053495753e-07,
+      "loss": 0.3304,
+      "step": 6077
+    },
+    {
+      "epoch": 2.8265385211595104,
+      "grad_norm": 0.30201414227485657,
+      "learning_rate": 1.0098429662188769e-07,
+      "loss": 0.3044,
+      "step": 6078
+    },
+    {
+      "epoch": 2.8270035653387073,
+      "grad_norm": 0.31688815355300903,
+      "learning_rate": 1.0044392273632354e-07,
+      "loss": 0.3012,
+      "step": 6079
+    },
+    {
+      "epoch": 2.8274686095179042,
+      "grad_norm": 0.32895052433013916,
+      "learning_rate": 9.990498385115066e-08,
+      "loss": 0.3256,
+      "step": 6080
+    },
+    {
+      "epoch": 2.827933653697101,
+      "grad_norm": 0.32152754068374634,
+      "learning_rate": 9.936748012421504e-08,
+      "loss": 0.3223,
+      "step": 6081
+    },
+    {
+      "epoch": 2.828398697876298,
+      "grad_norm": 0.31743600964546204,
+      "learning_rate": 9.883141171294242e-08,
+      "loss": 0.2809,
+      "step": 6082
+    },
+    {
+      "epoch": 2.8288637420554954,
+      "grad_norm": 0.3381618857383728,
+      "learning_rate": 9.829677877433886e-08,
+      "loss": 0.3185,
+      "step": 6083
+    },
+    {
+      "epoch": 2.8293287862346923,
+      "grad_norm": 0.32062768936157227,
+      "learning_rate": 9.776358146498966e-08,
+      "loss": 0.3108,
+      "step": 6084
+    },
+    {
+      "epoch": 2.829793830413889,
+      "grad_norm": 0.30004435777664185,
+      "learning_rate": 9.72318199410599e-08,
+      "loss": 0.3067,
+      "step": 6085
+    },
+    {
+      "epoch": 2.8302588745930866,
+      "grad_norm": 0.3237650990486145,
+      "learning_rate": 9.670149435829334e-08,
+      "loss": 0.3609,
+      "step": 6086
+    },
+    {
+      "epoch": 2.8307239187722835,
+      "grad_norm": 0.3168402314186096,
+      "learning_rate": 9.617260487201407e-08,
+      "loss": 0.3113,
+      "step": 6087
+    },
+    {
+      "epoch": 2.8311889629514804,
+      "grad_norm": 0.3079497218132019,
+      "learning_rate": 9.564515163712595e-08,
+      "loss": 0.2955,
+      "step": 6088
+    },
+    {
+      "epoch": 2.8316540071306773,
+      "grad_norm": 0.3398304581642151,
+      "learning_rate": 9.511913480810985e-08,
+      "loss": 0.3315,
+      "step": 6089
+    },
+    {
+      "epoch": 2.832119051309874,
+      "grad_norm": 0.35483306646347046,
+      "learning_rate": 9.459455453902866e-08,
+      "loss": 0.3551,
+      "step": 6090
+    },
+    {
+      "epoch": 2.8325840954890715,
+      "grad_norm": 0.34219107031822205,
+      "learning_rate": 9.407141098352335e-08,
+      "loss": 0.3129,
+      "step": 6091
+    },
+    {
+      "epoch": 2.8330491396682684,
+      "grad_norm": 0.29722991585731506,
+      "learning_rate": 9.354970429481413e-08,
+      "loss": 0.2773,
+      "step": 6092
+    },
+    {
+      "epoch": 2.8335141838474653,
+      "grad_norm": 0.31382736563682556,
+      "learning_rate": 9.302943462569991e-08,
+      "loss": 0.3483,
+      "step": 6093
+    },
+    {
+      "epoch": 2.8339792280266627,
+      "grad_norm": 0.3087048828601837,
+      "learning_rate": 9.25106021285599e-08,
+      "loss": 0.2727,
+      "step": 6094
+    },
+    {
+      "epoch": 2.8344442722058596,
+      "grad_norm": 0.3467859923839569,
+      "learning_rate": 9.199320695535086e-08,
+      "loss": 0.3607,
+      "step": 6095
+    },
+    {
+      "epoch": 2.8349093163850565,
+      "grad_norm": 0.33929920196533203,
+      "learning_rate": 9.147724925760993e-08,
+      "loss": 0.2912,
+      "step": 6096
+    },
+    {
+      "epoch": 2.835374360564254,
+      "grad_norm": 0.31342393159866333,
+      "learning_rate": 9.096272918645343e-08,
+      "loss": 0.2958,
+      "step": 6097
+    },
+    {
+      "epoch": 2.8358394047434508,
+      "grad_norm": 0.31853240728378296,
+      "learning_rate": 9.044964689257474e-08,
+      "loss": 0.3308,
+      "step": 6098
+    },
+    {
+      "epoch": 2.8363044489226477,
+      "grad_norm": 0.32699882984161377,
+      "learning_rate": 8.993800252624863e-08,
+      "loss": 0.2928,
+      "step": 6099
+    },
+    {
+      "epoch": 2.8367694931018446,
+      "grad_norm": 0.2974814474582672,
+      "learning_rate": 8.942779623732578e-08,
+      "loss": 0.2999,
+      "step": 6100
+    },
+    {
+      "epoch": 2.8372345372810415,
+      "grad_norm": 0.32415345311164856,
+      "learning_rate": 8.89190281752389e-08,
+      "loss": 0.326,
+      "step": 6101
+    },
+    {
+      "epoch": 2.837699581460239,
+      "grad_norm": 0.31111282110214233,
+      "learning_rate": 8.841169848899711e-08,
+      "loss": 0.3196,
+      "step": 6102
+    },
+    {
+      "epoch": 2.8381646256394357,
+      "grad_norm": 0.3355112671852112,
+      "learning_rate": 8.790580732718934e-08,
+      "loss": 0.3027,
+      "step": 6103
+    },
+    {
+      "epoch": 2.8386296698186326,
+      "grad_norm": 0.33080142736434937,
+      "learning_rate": 8.740135483798207e-08,
+      "loss": 0.3272,
+      "step": 6104
+    },
+    {
+      "epoch": 2.83909471399783,
+      "grad_norm": 0.3315957486629486,
+      "learning_rate": 8.68983411691221e-08,
+      "loss": 0.3239,
+      "step": 6105
+    },
+    {
+      "epoch": 2.839559758177027,
+      "grad_norm": 0.3323264420032501,
+      "learning_rate": 8.639676646793382e-08,
+      "loss": 0.3448,
+      "step": 6106
+    },
+    {
+      "epoch": 2.840024802356224,
+      "grad_norm": 0.29447272419929504,
+      "learning_rate": 8.589663088131972e-08,
+      "loss": 0.2987,
+      "step": 6107
+    },
+    {
+      "epoch": 2.840489846535421,
+      "grad_norm": 0.3482167422771454,
+      "learning_rate": 8.539793455576207e-08,
+      "loss": 0.3381,
+      "step": 6108
+    },
+    {
+      "epoch": 2.840954890714618,
+      "grad_norm": 0.29525330662727356,
+      "learning_rate": 8.490067763732124e-08,
+      "loss": 0.3098,
+      "step": 6109
+    },
+    {
+      "epoch": 2.841419934893815,
+      "grad_norm": 0.3327022194862366,
+      "learning_rate": 8.44048602716352e-08,
+      "loss": 0.3471,
+      "step": 6110
+    },
+    {
+      "epoch": 2.841884979073012,
+      "grad_norm": 0.3110980987548828,
+      "learning_rate": 8.391048260392054e-08,
+      "loss": 0.3353,
+      "step": 6111
+    },
+    {
+      "epoch": 2.8423500232522088,
+      "grad_norm": 0.31810250878334045,
+      "learning_rate": 8.341754477897257e-08,
+      "loss": 0.3223,
+      "step": 6112
+    },
+    {
+      "epoch": 2.842815067431406,
+      "grad_norm": 0.3068527281284332,
+      "learning_rate": 8.292604694116523e-08,
+      "loss": 0.3382,
+      "step": 6113
+    },
+    {
+      "epoch": 2.843280111610603,
+      "grad_norm": 0.32268133759498596,
+      "learning_rate": 8.24359892344495e-08,
+      "loss": 0.3259,
+      "step": 6114
+    },
+    {
+      "epoch": 2.8437451557898,
+      "grad_norm": 0.31271931529045105,
+      "learning_rate": 8.194737180235668e-08,
+      "loss": 0.3327,
+      "step": 6115
+    },
+    {
+      "epoch": 2.8442101999689973,
+      "grad_norm": 0.31981274485588074,
+      "learning_rate": 8.146019478799282e-08,
+      "loss": 0.2942,
+      "step": 6116
+    },
+    {
+      "epoch": 2.844675244148194,
+      "grad_norm": 0.3350343406200409,
+      "learning_rate": 8.097445833404605e-08,
+      "loss": 0.3287,
+      "step": 6117
+    },
+    {
+      "epoch": 2.845140288327391,
+      "grad_norm": 0.316834419965744,
+      "learning_rate": 8.049016258277976e-08,
+      "loss": 0.2982,
+      "step": 6118
+    },
+    {
+      "epoch": 2.845605332506588,
+      "grad_norm": 0.3442407548427582,
+      "learning_rate": 8.000730767603604e-08,
+      "loss": 0.342,
+      "step": 6119
+    },
+    {
+      "epoch": 2.846070376685785,
+      "grad_norm": 0.3273371160030365,
+      "learning_rate": 7.952589375523567e-08,
+      "loss": 0.346,
+      "step": 6120
+    },
+    {
+      "epoch": 2.8465354208649822,
+      "grad_norm": 0.29861021041870117,
+      "learning_rate": 7.904592096137753e-08,
+      "loss": 0.3062,
+      "step": 6121
+    },
+    {
+      "epoch": 2.847000465044179,
+      "grad_norm": 0.3003145158290863,
+      "learning_rate": 7.856738943503694e-08,
+      "loss": 0.3057,
+      "step": 6122
+    },
+    {
+      "epoch": 2.847465509223376,
+      "grad_norm": 0.2993358373641968,
+      "learning_rate": 7.809029931636902e-08,
+      "loss": 0.3087,
+      "step": 6123
+    },
+    {
+      "epoch": 2.8479305534025734,
+      "grad_norm": 0.31024885177612305,
+      "learning_rate": 7.761465074510422e-08,
+      "loss": 0.3314,
+      "step": 6124
+    },
+    {
+      "epoch": 2.8483955975817703,
+      "grad_norm": 0.31834203004837036,
+      "learning_rate": 7.714044386055386e-08,
+      "loss": 0.3016,
+      "step": 6125
+    },
+    {
+      "epoch": 2.848860641760967,
+      "grad_norm": 0.3217991888523102,
+      "learning_rate": 7.666767880160464e-08,
+      "loss": 0.3448,
+      "step": 6126
+    },
+    {
+      "epoch": 2.8493256859401646,
+      "grad_norm": 0.29849451780319214,
+      "learning_rate": 7.619635570672135e-08,
+      "loss": 0.3216,
+      "step": 6127
+    },
+    {
+      "epoch": 2.8497907301193615,
+      "grad_norm": 0.31368592381477356,
+      "learning_rate": 7.5726474713948e-08,
+      "loss": 0.303,
+      "step": 6128
+    },
+    {
+      "epoch": 2.8502557742985584,
+      "grad_norm": 0.295265257358551,
+      "learning_rate": 7.525803596090397e-08,
+      "loss": 0.3073,
+      "step": 6129
+    },
+    {
+      "epoch": 2.8507208184777553,
+      "grad_norm": 0.31519418954849243,
+      "learning_rate": 7.479103958478783e-08,
+      "loss": 0.3233,
+      "step": 6130
+    },
+    {
+      "epoch": 2.851185862656952,
+      "grad_norm": 0.3676014840602875,
+      "learning_rate": 7.432548572237519e-08,
+      "loss": 0.3799,
+      "step": 6131
+    },
+    {
+      "epoch": 2.8516509068361495,
+      "grad_norm": 0.3276804983615875,
+      "learning_rate": 7.386137451001974e-08,
+      "loss": 0.3389,
+      "step": 6132
+    },
+    {
+      "epoch": 2.8521159510153464,
+      "grad_norm": 0.3015483319759369,
+      "learning_rate": 7.339870608365107e-08,
+      "loss": 0.291,
+      "step": 6133
+    },
+    {
+      "epoch": 2.8525809951945433,
+      "grad_norm": 0.3150058090686798,
+      "learning_rate": 7.293748057877859e-08,
+      "loss": 0.3113,
+      "step": 6134
+    },
+    {
+      "epoch": 2.8530460393737407,
+      "grad_norm": 0.3133685886859894,
+      "learning_rate": 7.247769813048644e-08,
+      "loss": 0.3141,
+      "step": 6135
+    },
+    {
+      "epoch": 2.8535110835529376,
+      "grad_norm": 0.3154557943344116,
+      "learning_rate": 7.201935887343858e-08,
+      "loss": 0.3067,
+      "step": 6136
+    },
+    {
+      "epoch": 2.8539761277321345,
+      "grad_norm": 0.32975584268569946,
+      "learning_rate": 7.156246294187374e-08,
+      "loss": 0.3272,
+      "step": 6137
+    },
+    {
+      "epoch": 2.854441171911332,
+      "grad_norm": 0.31270360946655273,
+      "learning_rate": 7.110701046961044e-08,
+      "loss": 0.3105,
+      "step": 6138
+    },
+    {
+      "epoch": 2.8549062160905287,
+      "grad_norm": 0.3216429650783539,
+      "learning_rate": 7.065300159004307e-08,
+      "loss": 0.299,
+      "step": 6139
+    },
+    {
+      "epoch": 2.8553712602697257,
+      "grad_norm": 0.33457085490226746,
+      "learning_rate": 7.02004364361436e-08,
+      "loss": 0.3355,
+      "step": 6140
+    },
+    {
+      "epoch": 2.8558363044489226,
+      "grad_norm": 0.31512632966041565,
+      "learning_rate": 6.974931514046046e-08,
+      "loss": 0.3328,
+      "step": 6141
+    },
+    {
+      "epoch": 2.8563013486281195,
+      "grad_norm": 0.3386141359806061,
+      "learning_rate": 6.929963783511961e-08,
+      "loss": 0.3283,
+      "step": 6142
+    },
+    {
+      "epoch": 2.856766392807317,
+      "grad_norm": 0.30703625082969666,
+      "learning_rate": 6.885140465182516e-08,
+      "loss": 0.3204,
+      "step": 6143
+    },
+    {
+      "epoch": 2.8572314369865137,
+      "grad_norm": 0.31355178356170654,
+      "learning_rate": 6.840461572185708e-08,
+      "loss": 0.2829,
+      "step": 6144
+    },
+    {
+      "epoch": 2.8576964811657106,
+      "grad_norm": 0.36136549711227417,
+      "learning_rate": 6.795927117607238e-08,
+      "loss": 0.3508,
+      "step": 6145
+    },
+    {
+      "epoch": 2.858161525344908,
+      "grad_norm": 0.30663174390792847,
+      "learning_rate": 6.751537114490503e-08,
+      "loss": 0.3015,
+      "step": 6146
+    },
+    {
+      "epoch": 2.858626569524105,
+      "grad_norm": 0.35951730608940125,
+      "learning_rate": 6.707291575836661e-08,
+      "loss": 0.3571,
+      "step": 6147
+    },
+    {
+      "epoch": 2.859091613703302,
+      "grad_norm": 0.3195739984512329,
+      "learning_rate": 6.663190514604456e-08,
+      "loss": 0.304,
+      "step": 6148
+    },
+    {
+      "epoch": 2.8595566578824987,
+      "grad_norm": 0.32330840826034546,
+      "learning_rate": 6.61923394371039e-08,
+      "loss": 0.314,
+      "step": 6149
+    },
+    {
+      "epoch": 2.8600217020616956,
+      "grad_norm": 0.30728137493133545,
+      "learning_rate": 6.575421876028721e-08,
+      "loss": 0.3342,
+      "step": 6150
+    },
+    {
+      "epoch": 2.860486746240893,
+      "grad_norm": 0.2898307740688324,
+      "learning_rate": 6.531754324391126e-08,
+      "loss": 0.3215,
+      "step": 6151
+    },
+    {
+      "epoch": 2.86095179042009,
+      "grad_norm": 0.308747798204422,
+      "learning_rate": 6.488231301587266e-08,
+      "loss": 0.3281,
+      "step": 6152
+    },
+    {
+      "epoch": 2.8614168345992868,
+      "grad_norm": 0.32874009013175964,
+      "learning_rate": 6.444852820364222e-08,
+      "loss": 0.3273,
+      "step": 6153
+    },
+    {
+      "epoch": 2.861881878778484,
+      "grad_norm": 0.3160831332206726,
+      "learning_rate": 6.401618893426886e-08,
+      "loss": 0.3156,
+      "step": 6154
+    },
+    {
+      "epoch": 2.862346922957681,
+      "grad_norm": 0.3409363031387329,
+      "learning_rate": 6.358529533437796e-08,
+      "loss": 0.3194,
+      "step": 6155
+    },
+    {
+      "epoch": 2.862811967136878,
+      "grad_norm": 0.29154202342033386,
+      "learning_rate": 6.315584753017134e-08,
+      "loss": 0.3113,
+      "step": 6156
+    },
+    {
+      "epoch": 2.8632770113160753,
+      "grad_norm": 0.3012997508049011,
+      "learning_rate": 6.272784564742673e-08,
+      "loss": 0.3508,
+      "step": 6157
+    },
+    {
+      "epoch": 2.863742055495272,
+      "grad_norm": 0.3225801885128021,
+      "learning_rate": 6.230128981149941e-08,
+      "loss": 0.3195,
+      "step": 6158
+    },
+    {
+      "epoch": 2.864207099674469,
+      "grad_norm": 0.30362215638160706,
+      "learning_rate": 6.187618014732056e-08,
+      "loss": 0.3027,
+      "step": 6159
+    },
+    {
+      "epoch": 2.864672143853666,
+      "grad_norm": 0.32935047149658203,
+      "learning_rate": 6.145251677939778e-08,
+      "loss": 0.3351,
+      "step": 6160
+    },
+    {
+      "epoch": 2.865137188032863,
+      "grad_norm": 0.3463619649410248,
+      "learning_rate": 6.103029983181519e-08,
+      "loss": 0.3409,
+      "step": 6161
+    },
+    {
+      "epoch": 2.8656022322120602,
+      "grad_norm": 0.29759082198143005,
+      "learning_rate": 6.060952942823328e-08,
+      "loss": 0.3043,
+      "step": 6162
+    },
+    {
+      "epoch": 2.866067276391257,
+      "grad_norm": 0.3281930685043335,
+      "learning_rate": 6.01902056918896e-08,
+      "loss": 0.3291,
+      "step": 6163
+    },
+    {
+      "epoch": 2.866532320570454,
+      "grad_norm": 0.3233920931816101,
+      "learning_rate": 5.977232874559535e-08,
+      "loss": 0.2848,
+      "step": 6164
+    },
+    {
+      "epoch": 2.8669973647496514,
+      "grad_norm": 0.31291288137435913,
+      "learning_rate": 5.935589871174208e-08,
+      "loss": 0.341,
+      "step": 6165
+    },
+    {
+      "epoch": 2.8674624089288483,
+      "grad_norm": 0.3225501775741577,
+      "learning_rate": 5.8940915712293875e-08,
+      "loss": 0.3351,
+      "step": 6166
+    },
+    {
+      "epoch": 2.867927453108045,
+      "grad_norm": 0.33493563532829285,
+      "learning_rate": 5.8527379868792976e-08,
+      "loss": 0.3184,
+      "step": 6167
+    },
+    {
+      "epoch": 2.8683924972872425,
+      "grad_norm": 0.32679054141044617,
+      "learning_rate": 5.811529130235749e-08,
+      "loss": 0.2963,
+      "step": 6168
+    },
+    {
+      "epoch": 2.8688575414664395,
+      "grad_norm": 0.319558709859848,
+      "learning_rate": 5.770465013368198e-08,
+      "loss": 0.33,
+      "step": 6169
+    },
+    {
+      "epoch": 2.8693225856456364,
+      "grad_norm": 0.28156691789627075,
+      "learning_rate": 5.729545648303525e-08,
+      "loss": 0.2726,
+      "step": 6170
+    },
+    {
+      "epoch": 2.8697876298248333,
+      "grad_norm": 0.3368379771709442,
+      "learning_rate": 5.688771047026476e-08,
+      "loss": 0.3513,
+      "step": 6171
+    },
+    {
+      "epoch": 2.87025267400403,
+      "grad_norm": 0.32251864671707153,
+      "learning_rate": 5.648141221479164e-08,
+      "loss": 0.3177,
+      "step": 6172
+    },
+    {
+      "epoch": 2.8707177181832275,
+      "grad_norm": 0.3322902321815491,
+      "learning_rate": 5.6076561835615164e-08,
+      "loss": 0.3423,
+      "step": 6173
+    },
+    {
+      "epoch": 2.8711827623624244,
+      "grad_norm": 0.3071969449520111,
+      "learning_rate": 5.5673159451308246e-08,
+      "loss": 0.3236,
+      "step": 6174
+    },
+    {
+      "epoch": 2.8716478065416213,
+      "grad_norm": 0.29615259170532227,
+      "learning_rate": 5.527120518002138e-08,
+      "loss": 0.2841,
+      "step": 6175
+    },
+    {
+      "epoch": 2.8721128507208187,
+      "grad_norm": 0.32115840911865234,
+      "learning_rate": 5.487069913948096e-08,
+      "loss": 0.366,
+      "step": 6176
+    },
+    {
+      "epoch": 2.8725778949000156,
+      "grad_norm": 0.31044548749923706,
+      "learning_rate": 5.447164144698758e-08,
+      "loss": 0.3026,
+      "step": 6177
+    },
+    {
+      "epoch": 2.8730429390792125,
+      "grad_norm": 0.3024675250053406,
+      "learning_rate": 5.407403221941998e-08,
+      "loss": 0.3232,
+      "step": 6178
+    },
+    {
+      "epoch": 2.8735079832584094,
+      "grad_norm": 0.3414202034473419,
+      "learning_rate": 5.367787157323057e-08,
+      "loss": 0.3284,
+      "step": 6179
+    },
+    {
+      "epoch": 2.8739730274376063,
+      "grad_norm": 0.3175242841243744,
+      "learning_rate": 5.3283159624448745e-08,
+      "loss": 0.2636,
+      "step": 6180
+    },
+    {
+      "epoch": 2.8744380716168036,
+      "grad_norm": 0.490583211183548,
+      "learning_rate": 5.2889896488679816e-08,
+      "loss": 0.3264,
+      "step": 6181
+    },
+    {
+      "epoch": 2.8749031157960006,
+      "grad_norm": 0.32635655999183655,
+      "learning_rate": 5.249808228110276e-08,
+      "loss": 0.3085,
+      "step": 6182
+    },
+    {
+      "epoch": 2.8753681599751975,
+      "grad_norm": 0.32171571254730225,
+      "learning_rate": 5.2107717116474665e-08,
+      "loss": 0.3289,
+      "step": 6183
+    },
+    {
+      "epoch": 2.875833204154395,
+      "grad_norm": 0.30658993124961853,
+      "learning_rate": 5.171880110912686e-08,
+      "loss": 0.2749,
+      "step": 6184
+    },
+    {
+      "epoch": 2.8762982483335917,
+      "grad_norm": 0.3477241098880768,
+      "learning_rate": 5.133133437296656e-08,
+      "loss": 0.3022,
+      "step": 6185
+    },
+    {
+      "epoch": 2.8767632925127886,
+      "grad_norm": 0.31438449025154114,
+      "learning_rate": 5.094531702147632e-08,
+      "loss": 0.2979,
+      "step": 6186
+    },
+    {
+      "epoch": 2.877228336691986,
+      "grad_norm": 0.32434505224227905,
+      "learning_rate": 5.056074916771458e-08,
+      "loss": 0.2972,
+      "step": 6187
+    },
+    {
+      "epoch": 2.877693380871183,
+      "grad_norm": 0.33703556656837463,
+      "learning_rate": 5.01776309243146e-08,
+      "loss": 0.3343,
+      "step": 6188
+    },
+    {
+      "epoch": 2.8781584250503798,
+      "grad_norm": 0.2929964065551758,
+      "learning_rate": 4.97959624034855e-08,
+      "loss": 0.3,
+      "step": 6189
+    },
+    {
+      "epoch": 2.8786234692295767,
+      "grad_norm": 0.3615955114364624,
+      "learning_rate": 4.9415743717012296e-08,
+      "loss": 0.3695,
+      "step": 6190
+    },
+    {
+      "epoch": 2.8790885134087736,
+      "grad_norm": 0.33232712745666504,
+      "learning_rate": 4.903697497625537e-08,
+      "loss": 0.2803,
+      "step": 6191
+    },
+    {
+      "epoch": 2.879553557587971,
+      "grad_norm": 0.32238370180130005,
+      "learning_rate": 4.865965629214819e-08,
+      "loss": 0.342,
+      "step": 6192
+    },
+    {
+      "epoch": 2.880018601767168,
+      "grad_norm": 0.29459741711616516,
+      "learning_rate": 4.828378777520293e-08,
+      "loss": 0.3223,
+      "step": 6193
+    },
+    {
+      "epoch": 2.8804836459463647,
+      "grad_norm": 0.3244211971759796,
+      "learning_rate": 4.790936953550485e-08,
+      "loss": 0.3027,
+      "step": 6194
+    },
+    {
+      "epoch": 2.880948690125562,
+      "grad_norm": 0.33812856674194336,
+      "learning_rate": 4.753640168271456e-08,
+      "loss": 0.3327,
+      "step": 6195
+    },
+    {
+      "epoch": 2.881413734304759,
+      "grad_norm": 0.3344283998012543,
+      "learning_rate": 4.7164884326068584e-08,
+      "loss": 0.3385,
+      "step": 6196
+    },
+    {
+      "epoch": 2.881878778483956,
+      "grad_norm": 0.332630455493927,
+      "learning_rate": 4.6794817574378204e-08,
+      "loss": 0.3118,
+      "step": 6197
+    },
+    {
+      "epoch": 2.8823438226631533,
+      "grad_norm": 0.33553823828697205,
+      "learning_rate": 4.6426201536030616e-08,
+      "loss": 0.3352,
+      "step": 6198
+    },
+    {
+      "epoch": 2.88280886684235,
+      "grad_norm": 0.32330450415611267,
+      "learning_rate": 4.605903631898612e-08,
+      "loss": 0.3162,
+      "step": 6199
+    },
+    {
+      "epoch": 2.883273911021547,
+      "grad_norm": 0.32343700528144836,
+      "learning_rate": 4.569332203078258e-08,
+      "loss": 0.3222,
+      "step": 6200
+    },
+    {
+      "epoch": 2.883738955200744,
+      "grad_norm": 0.310979425907135,
+      "learning_rate": 4.5329058778531e-08,
+      "loss": 0.2865,
+      "step": 6201
+    },
+    {
+      "epoch": 2.884203999379941,
+      "grad_norm": 0.3251758813858032,
+      "learning_rate": 4.4966246668919355e-08,
+      "loss": 0.3159,
+      "step": 6202
+    },
+    {
+      "epoch": 2.8846690435591382,
+      "grad_norm": 0.3213922381401062,
+      "learning_rate": 4.460488580820821e-08,
+      "loss": 0.3045,
+      "step": 6203
+    },
+    {
+      "epoch": 2.885134087738335,
+      "grad_norm": 0.3345593512058258,
+      "learning_rate": 4.424497630223512e-08,
+      "loss": 0.3395,
+      "step": 6204
+    },
+    {
+      "epoch": 2.885599131917532,
+      "grad_norm": 0.3209022879600525,
+      "learning_rate": 4.3886518256411325e-08,
+      "loss": 0.3152,
+      "step": 6205
+    },
+    {
+      "epoch": 2.8860641760967294,
+      "grad_norm": 0.31403934955596924,
+      "learning_rate": 4.35295117757234e-08,
+      "loss": 0.3292,
+      "step": 6206
+    },
+    {
+      "epoch": 2.8865292202759263,
+      "grad_norm": 0.289949893951416,
+      "learning_rate": 4.3173956964732145e-08,
+      "loss": 0.304,
+      "step": 6207
+    },
+    {
+      "epoch": 2.886994264455123,
+      "grad_norm": 0.33771443367004395,
+      "learning_rate": 4.281985392757537e-08,
+      "loss": 0.3385,
+      "step": 6208
+    },
+    {
+      "epoch": 2.88745930863432,
+      "grad_norm": 0.2989499866962433,
+      "learning_rate": 4.2467202767962346e-08,
+      "loss": 0.2799,
+      "step": 6209
+    },
+    {
+      "epoch": 2.8879243528135174,
+      "grad_norm": 0.318925142288208,
+      "learning_rate": 4.211600358917989e-08,
+      "loss": 0.3248,
+      "step": 6210
+    },
+    {
+      "epoch": 2.8883893969927144,
+      "grad_norm": 0.31016743183135986,
+      "learning_rate": 4.17662564940885e-08,
+      "loss": 0.3164,
+      "step": 6211
+    },
+    {
+      "epoch": 2.8888544411719113,
+      "grad_norm": 0.3462943732738495,
+      "learning_rate": 4.14179615851229e-08,
+      "loss": 0.3617,
+      "step": 6212
+    },
+    {
+      "epoch": 2.889319485351108,
+      "grad_norm": 0.3116661310195923,
+      "learning_rate": 4.1071118964293166e-08,
+      "loss": 0.3037,
+      "step": 6213
+    },
+    {
+      "epoch": 2.8897845295303055,
+      "grad_norm": 0.29562506079673767,
+      "learning_rate": 4.07257287331847e-08,
+      "loss": 0.3133,
+      "step": 6214
+    },
+    {
+      "epoch": 2.8902495737095024,
+      "grad_norm": 0.3385922610759735,
+      "learning_rate": 4.038179099295547e-08,
+      "loss": 0.3416,
+      "step": 6215
+    },
+    {
+      "epoch": 2.8907146178886993,
+      "grad_norm": 0.28111937642097473,
+      "learning_rate": 4.0039305844339905e-08,
+      "loss": 0.3165,
+      "step": 6216
+    },
+    {
+      "epoch": 2.8911796620678967,
+      "grad_norm": 0.3083990216255188,
+      "learning_rate": 3.969827338764665e-08,
+      "loss": 0.3474,
+      "step": 6217
+    },
+    {
+      "epoch": 2.8916447062470936,
+      "grad_norm": 0.3193405270576477,
+      "learning_rate": 3.935869372275747e-08,
+      "loss": 0.3144,
+      "step": 6218
+    },
+    {
+      "epoch": 2.8921097504262905,
+      "grad_norm": 0.3196331858634949,
+      "learning_rate": 3.9020566949131145e-08,
+      "loss": 0.3228,
+      "step": 6219
+    },
+    {
+      "epoch": 2.8925747946054874,
+      "grad_norm": 0.31240054965019226,
+      "learning_rate": 3.868389316579846e-08,
+      "loss": 0.3234,
+      "step": 6220
+    },
+    {
+      "epoch": 2.8930398387846843,
+      "grad_norm": 0.2907924950122833,
+      "learning_rate": 3.834867247136553e-08,
+      "loss": 0.3052,
+      "step": 6221
+    },
+    {
+      "epoch": 2.8935048829638816,
+      "grad_norm": 0.3368573784828186,
+      "learning_rate": 3.801490496401439e-08,
+      "loss": 0.3779,
+      "step": 6222
+    },
+    {
+      "epoch": 2.8939699271430785,
+      "grad_norm": 0.3105674088001251,
+      "learning_rate": 3.768259074149905e-08,
+      "loss": 0.3045,
+      "step": 6223
+    },
+    {
+      "epoch": 2.8944349713222755,
+      "grad_norm": 0.31339845061302185,
+      "learning_rate": 3.735172990114888e-08,
+      "loss": 0.3456,
+      "step": 6224
+    },
+    {
+      "epoch": 2.894900015501473,
+      "grad_norm": 0.32460853457450867,
+      "learning_rate": 3.702232253986804e-08,
+      "loss": 0.3115,
+      "step": 6225
+    },
+    {
+      "epoch": 2.8953650596806697,
+      "grad_norm": 0.33056336641311646,
+      "learning_rate": 3.6694368754134346e-08,
+      "loss": 0.3283,
+      "step": 6226
+    },
+    {
+      "epoch": 2.8958301038598666,
+      "grad_norm": 0.31713178753852844,
+      "learning_rate": 3.6367868640000416e-08,
+      "loss": 0.3124,
+      "step": 6227
+    },
+    {
+      "epoch": 2.896295148039064,
+      "grad_norm": 0.3436635732650757,
+      "learning_rate": 3.6042822293093083e-08,
+      "loss": 0.3683,
+      "step": 6228
+    },
+    {
+      "epoch": 2.896760192218261,
+      "grad_norm": 0.30953043699264526,
+      "learning_rate": 3.571922980861231e-08,
+      "loss": 0.3073,
+      "step": 6229
+    },
+    {
+      "epoch": 2.8972252363974578,
+      "grad_norm": 0.32656288146972656,
+      "learning_rate": 3.539709128133395e-08,
+      "loss": 0.2919,
+      "step": 6230
+    },
+    {
+      "epoch": 2.8976902805766547,
+      "grad_norm": 0.31821176409721375,
+      "learning_rate": 3.5076406805606425e-08,
+      "loss": 0.3195,
+      "step": 6231
+    },
+    {
+      "epoch": 2.8981553247558516,
+      "grad_norm": 0.31194084882736206,
+      "learning_rate": 3.475717647535348e-08,
+      "loss": 0.3455,
+      "step": 6232
+    },
+    {
+      "epoch": 2.898620368935049,
+      "grad_norm": 0.31423068046569824,
+      "learning_rate": 3.443940038407256e-08,
+      "loss": 0.3265,
+      "step": 6233
+    },
+    {
+      "epoch": 2.899085413114246,
+      "grad_norm": 0.2951379418373108,
+      "learning_rate": 3.4123078624834214e-08,
+      "loss": 0.2827,
+      "step": 6234
+    },
+    {
+      "epoch": 2.8995504572934427,
+      "grad_norm": 0.3533174395561218,
+      "learning_rate": 3.3808211290284886e-08,
+      "loss": 0.3611,
+      "step": 6235
+    },
+    {
+      "epoch": 2.90001550147264,
+      "grad_norm": 0.3347010612487793,
+      "learning_rate": 3.349479847264414e-08,
+      "loss": 0.3355,
+      "step": 6236
+    },
+    {
+      "epoch": 2.900480545651837,
+      "grad_norm": 0.3282628059387207,
+      "learning_rate": 3.318284026370522e-08,
+      "loss": 0.3093,
+      "step": 6237
+    },
+    {
+      "epoch": 2.900945589831034,
+      "grad_norm": 0.31977033615112305,
+      "learning_rate": 3.287233675483503e-08,
+      "loss": 0.328,
+      "step": 6238
+    },
+    {
+      "epoch": 2.9014106340102312,
+      "grad_norm": 0.3105235993862152,
+      "learning_rate": 3.2563288036976394e-08,
+      "loss": 0.3117,
+      "step": 6239
+    },
+    {
+      "epoch": 2.901875678189428,
+      "grad_norm": 0.3219479024410248,
+      "learning_rate": 3.2255694200643003e-08,
+      "loss": 0.3167,
+      "step": 6240
+    },
+    {
+      "epoch": 2.902340722368625,
+      "grad_norm": 0.30277541279792786,
+      "learning_rate": 3.194955533592559e-08,
+      "loss": 0.2801,
+      "step": 6241
+    },
+    {
+      "epoch": 2.902805766547822,
+      "grad_norm": 0.3230368196964264,
+      "learning_rate": 3.16448715324863e-08,
+      "loss": 0.3138,
+      "step": 6242
+    },
+    {
+      "epoch": 2.903270810727019,
+      "grad_norm": 0.3114720582962036,
+      "learning_rate": 3.13416428795621e-08,
+      "loss": 0.3266,
+      "step": 6243
+    },
+    {
+      "epoch": 2.903735854906216,
+      "grad_norm": 0.324063777923584,
+      "learning_rate": 3.103986946596415e-08,
+      "loss": 0.3465,
+      "step": 6244
+    },
+    {
+      "epoch": 2.904200899085413,
+      "grad_norm": 0.318876177072525,
+      "learning_rate": 3.073955138007734e-08,
+      "loss": 0.3217,
+      "step": 6245
+    },
+    {
+      "epoch": 2.90466594326461,
+      "grad_norm": 0.3103938698768616,
+      "learning_rate": 3.044068870985906e-08,
+      "loss": 0.3207,
+      "step": 6246
+    },
+    {
+      "epoch": 2.9051309874438074,
+      "grad_norm": 0.31359586119651794,
+      "learning_rate": 3.014328154284152e-08,
+      "loss": 0.3173,
+      "step": 6247
+    },
+    {
+      "epoch": 2.9055960316230043,
+      "grad_norm": 0.311901330947876,
+      "learning_rate": 2.98473299661306e-08,
+      "loss": 0.3194,
+      "step": 6248
+    },
+    {
+      "epoch": 2.906061075802201,
+      "grad_norm": 0.31329211592674255,
+      "learning_rate": 2.955283406640641e-08,
+      "loss": 0.2925,
+      "step": 6249
+    },
+    {
+      "epoch": 2.906526119981398,
+      "grad_norm": 0.31857290863990784,
+      "learning_rate": 2.9259793929921066e-08,
+      "loss": 0.3267,
+      "step": 6250
+    },
+    {
+      "epoch": 2.906991164160595,
+      "grad_norm": 0.3248409628868103,
+      "learning_rate": 2.8968209642501465e-08,
+      "loss": 0.323,
+      "step": 6251
+    },
+    {
+      "epoch": 2.9074562083397923,
+      "grad_norm": 0.3088364005088806,
+      "learning_rate": 2.8678081289548187e-08,
+      "loss": 0.284,
+      "step": 6252
+    },
+    {
+      "epoch": 2.9079212525189893,
+      "grad_norm": 0.3107573091983795,
+      "learning_rate": 2.8389408956034925e-08,
+      "loss": 0.3242,
+      "step": 6253
+    },
+    {
+      "epoch": 2.908386296698186,
+      "grad_norm": 0.3310908079147339,
+      "learning_rate": 2.810219272650849e-08,
+      "loss": 0.3107,
+      "step": 6254
+    },
+    {
+      "epoch": 2.9088513408773835,
+      "grad_norm": 0.30595463514328003,
+      "learning_rate": 2.7816432685091598e-08,
+      "loss": 0.3522,
+      "step": 6255
+    },
+    {
+      "epoch": 2.9093163850565804,
+      "grad_norm": 0.3248220980167389,
+      "learning_rate": 2.7532128915476742e-08,
+      "loss": 0.3116,
+      "step": 6256
+    },
+    {
+      "epoch": 2.9097814292357773,
+      "grad_norm": 0.3430537283420563,
+      "learning_rate": 2.7249281500932868e-08,
+      "loss": 0.3414,
+      "step": 6257
+    },
+    {
+      "epoch": 2.9102464734149747,
+      "grad_norm": 0.32663294672966003,
+      "learning_rate": 2.6967890524301488e-08,
+      "loss": 0.2961,
+      "step": 6258
+    },
+    {
+      "epoch": 2.9107115175941716,
+      "grad_norm": 0.32773134112358093,
+      "learning_rate": 2.6687956067997234e-08,
+      "loss": 0.2928,
+      "step": 6259
+    },
+    {
+      "epoch": 2.9111765617733685,
+      "grad_norm": 0.3033883571624756,
+      "learning_rate": 2.640947821400841e-08,
+      "loss": 0.3062,
+      "step": 6260
+    },
+    {
+      "epoch": 2.9116416059525654,
+      "grad_norm": 0.33305469155311584,
+      "learning_rate": 2.6132457043896442e-08,
+      "loss": 0.2965,
+      "step": 6261
+    },
+    {
+      "epoch": 2.9121066501317623,
+      "grad_norm": 0.33847376704216003,
+      "learning_rate": 2.585689263879643e-08,
+      "loss": 0.3387,
+      "step": 6262
+    },
+    {
+      "epoch": 2.9125716943109596,
+      "grad_norm": 0.3103272616863251,
+      "learning_rate": 2.558278507941714e-08,
+      "loss": 0.2881,
+      "step": 6263
+    },
+    {
+      "epoch": 2.9130367384901565,
+      "grad_norm": 0.33586204051971436,
+      "learning_rate": 2.5310134446039357e-08,
+      "loss": 0.3428,
+      "step": 6264
+    },
+    {
+      "epoch": 2.9135017826693534,
+      "grad_norm": 0.34026390314102173,
+      "learning_rate": 2.503894081851921e-08,
+      "loss": 0.3164,
+      "step": 6265
+    },
+    {
+      "epoch": 2.913966826848551,
+      "grad_norm": 0.3134699761867523,
+      "learning_rate": 2.476920427628371e-08,
+      "loss": 0.3022,
+      "step": 6266
+    },
+    {
+      "epoch": 2.9144318710277477,
+      "grad_norm": 0.31349417567253113,
+      "learning_rate": 2.4500924898335223e-08,
+      "loss": 0.3335,
+      "step": 6267
+    },
+    {
+      "epoch": 2.9148969152069446,
+      "grad_norm": 0.30703291296958923,
+      "learning_rate": 2.4234102763247558e-08,
+      "loss": 0.3165,
+      "step": 6268
+    },
+    {
+      "epoch": 2.915361959386142,
+      "grad_norm": 0.30995601415634155,
+      "learning_rate": 2.3968737949169318e-08,
+      "loss": 0.3276,
+      "step": 6269
+    },
+    {
+      "epoch": 2.915827003565339,
+      "grad_norm": 0.3151072561740875,
+      "learning_rate": 2.370483053382111e-08,
+      "loss": 0.3182,
+      "step": 6270
+    },
+    {
+      "epoch": 2.9162920477445358,
+      "grad_norm": 0.3360000252723694,
+      "learning_rate": 2.3442380594497215e-08,
+      "loss": 0.3073,
+      "step": 6271
+    },
+    {
+      "epoch": 2.9167570919237327,
+      "grad_norm": 0.3323414921760559,
+      "learning_rate": 2.3181388208065036e-08,
+      "loss": 0.3643,
+      "step": 6272
+    },
+    {
+      "epoch": 2.9172221361029296,
+      "grad_norm": 0.3142452538013458,
+      "learning_rate": 2.2921853450965094e-08,
+      "loss": 0.2902,
+      "step": 6273
+    },
+    {
+      "epoch": 2.917687180282127,
+      "grad_norm": 0.332619309425354,
+      "learning_rate": 2.2663776399211024e-08,
+      "loss": 0.3332,
+      "step": 6274
+    },
+    {
+      "epoch": 2.918152224461324,
+      "grad_norm": 0.30116286873817444,
+      "learning_rate": 2.2407157128389033e-08,
+      "loss": 0.2971,
+      "step": 6275
+    },
+    {
+      "epoch": 2.9186172686405207,
+      "grad_norm": 0.3235343396663666,
+      "learning_rate": 2.2151995713659e-08,
+      "loss": 0.3239,
+      "step": 6276
+    },
+    {
+      "epoch": 2.919082312819718,
+      "grad_norm": 0.3417525887489319,
+      "learning_rate": 2.189829222975337e-08,
+      "loss": 0.3347,
+      "step": 6277
+    },
+    {
+      "epoch": 2.919547356998915,
+      "grad_norm": 0.34328693151474,
+      "learning_rate": 2.1646046750978255e-08,
+      "loss": 0.3241,
+      "step": 6278
+    },
+    {
+      "epoch": 2.920012401178112,
+      "grad_norm": 0.3048785924911499,
+      "learning_rate": 2.1395259351211227e-08,
+      "loss": 0.2951,
+      "step": 6279
+    },
+    {
+      "epoch": 2.920477445357309,
+      "grad_norm": 0.31795376539230347,
+      "learning_rate": 2.1145930103904645e-08,
+      "loss": 0.3376,
+      "step": 6280
+    },
+    {
+      "epoch": 2.9209424895365057,
+      "grad_norm": 0.34835532307624817,
+      "learning_rate": 2.0898059082082868e-08,
+      "loss": 0.3394,
+      "step": 6281
+    },
+    {
+      "epoch": 2.921407533715703,
+      "grad_norm": 0.30004480481147766,
+      "learning_rate": 2.065164635834338e-08,
+      "loss": 0.2802,
+      "step": 6282
+    },
+    {
+      "epoch": 2.9218725778949,
+      "grad_norm": 0.31442791223526,
+      "learning_rate": 2.040669200485623e-08,
+      "loss": 0.3024,
+      "step": 6283
+    },
+    {
+      "epoch": 2.922337622074097,
+      "grad_norm": 0.33434808254241943,
+      "learning_rate": 2.016319609336459e-08,
+      "loss": 0.3347,
+      "step": 6284
+    },
+    {
+      "epoch": 2.922802666253294,
+      "grad_norm": 0.3095998167991638,
+      "learning_rate": 1.992115869518474e-08,
+      "loss": 0.3192,
+      "step": 6285
+    },
+    {
+      "epoch": 2.923267710432491,
+      "grad_norm": 0.2919743061065674,
+      "learning_rate": 1.968057988120553e-08,
+      "loss": 0.3011,
+      "step": 6286
+    },
+    {
+      "epoch": 2.923732754611688,
+      "grad_norm": 0.3112308084964752,
+      "learning_rate": 1.9441459721887822e-08,
+      "loss": 0.3284,
+      "step": 6287
+    },
+    {
+      "epoch": 2.9241977987908854,
+      "grad_norm": 0.32711583375930786,
+      "learning_rate": 1.920379828726726e-08,
+      "loss": 0.3329,
+      "step": 6288
+    },
+    {
+      "epoch": 2.9246628429700823,
+      "grad_norm": 0.30145832896232605,
+      "learning_rate": 1.8967595646949834e-08,
+      "loss": 0.292,
+      "step": 6289
+    },
+    {
+      "epoch": 2.925127887149279,
+      "grad_norm": 0.3420089781284332,
+      "learning_rate": 1.8732851870115755e-08,
+      "loss": 0.32,
+      "step": 6290
+    },
+    {
+      "epoch": 2.925592931328476,
+      "grad_norm": 0.34311848878860474,
+      "learning_rate": 1.849956702551836e-08,
+      "loss": 0.3236,
+      "step": 6291
+    },
+    {
+      "epoch": 2.926057975507673,
+      "grad_norm": 0.31155329942703247,
+      "learning_rate": 1.826774118148189e-08,
+      "loss": 0.3191,
+      "step": 6292
+    },
+    {
+      "epoch": 2.9265230196868703,
+      "grad_norm": 0.3236565589904785,
+      "learning_rate": 1.803737440590536e-08,
+      "loss": 0.3369,
+      "step": 6293
+    },
+    {
+      "epoch": 2.9269880638660672,
+      "grad_norm": 0.28220975399017334,
+      "learning_rate": 1.7808466766259246e-08,
+      "loss": 0.2988,
+      "step": 6294
+    },
+    {
+      "epoch": 2.927453108045264,
+      "grad_norm": 0.3211870789527893,
+      "learning_rate": 1.758101832958603e-08,
+      "loss": 0.3305,
+      "step": 6295
+    },
+    {
+      "epoch": 2.9279181522244615,
+      "grad_norm": 0.3059036135673523,
+      "learning_rate": 1.7355029162502978e-08,
+      "loss": 0.2947,
+      "step": 6296
+    },
+    {
+      "epoch": 2.9283831964036584,
+      "grad_norm": 0.31441813707351685,
+      "learning_rate": 1.7130499331197703e-08,
+      "loss": 0.2993,
+      "step": 6297
+    },
+    {
+      "epoch": 2.9288482405828553,
+      "grad_norm": 0.32152053713798523,
+      "learning_rate": 1.6907428901432045e-08,
+      "loss": 0.3135,
+      "step": 6298
+    },
+    {
+      "epoch": 2.9293132847620527,
+      "grad_norm": 0.328326940536499,
+      "learning_rate": 1.668581793853874e-08,
+      "loss": 0.3092,
+      "step": 6299
+    },
+    {
+      "epoch": 2.9297783289412496,
+      "grad_norm": 0.3213350772857666,
+      "learning_rate": 1.6465666507425314e-08,
+      "loss": 0.3505,
+      "step": 6300
+    },
+    {
+      "epoch": 2.9302433731204465,
+      "grad_norm": 0.3014281988143921,
+      "learning_rate": 1.6246974672569083e-08,
+      "loss": 0.2903,
+      "step": 6301
+    },
+    {
+      "epoch": 2.9307084172996434,
+      "grad_norm": 0.31483855843544006,
+      "learning_rate": 1.6029742498022692e-08,
+      "loss": 0.3267,
+      "step": 6302
+    },
+    {
+      "epoch": 2.9311734614788403,
+      "grad_norm": 0.32072848081588745,
+      "learning_rate": 1.5813970047409144e-08,
+      "loss": 0.2982,
+      "step": 6303
+    },
+    {
+      "epoch": 2.9316385056580376,
+      "grad_norm": 0.3045773208141327,
+      "learning_rate": 1.559965738392455e-08,
+      "loss": 0.3062,
+      "step": 6304
+    },
+    {
+      "epoch": 2.9321035498372345,
+      "grad_norm": 0.3604282736778259,
+      "learning_rate": 1.538680457033814e-08,
+      "loss": 0.3295,
+      "step": 6305
+    },
+    {
+      "epoch": 2.9325685940164314,
+      "grad_norm": 0.3138650357723236,
+      "learning_rate": 1.5175411668990613e-08,
+      "loss": 0.2937,
+      "step": 6306
+    },
+    {
+      "epoch": 2.933033638195629,
+      "grad_norm": 0.330479234457016,
+      "learning_rate": 1.496547874179577e-08,
+      "loss": 0.3091,
+      "step": 6307
+    },
+    {
+      "epoch": 2.9334986823748257,
+      "grad_norm": 0.3400230407714844,
+      "learning_rate": 1.475700585023998e-08,
+      "loss": 0.3206,
+      "step": 6308
+    },
+    {
+      "epoch": 2.9339637265540226,
+      "grad_norm": 0.33280929923057556,
+      "learning_rate": 1.4549993055380519e-08,
+      "loss": 0.3632,
+      "step": 6309
+    },
+    {
+      "epoch": 2.9344287707332195,
+      "grad_norm": 0.29419514536857605,
+      "learning_rate": 1.4344440417848882e-08,
+      "loss": 0.2756,
+      "step": 6310
+    },
+    {
+      "epoch": 2.9348938149124164,
+      "grad_norm": 0.3062974214553833,
+      "learning_rate": 1.4140347997848025e-08,
+      "loss": 0.3265,
+      "step": 6311
+    },
+    {
+      "epoch": 2.9353588590916138,
+      "grad_norm": 0.3415301740169525,
+      "learning_rate": 1.3937715855152912e-08,
+      "loss": 0.3005,
+      "step": 6312
+    },
+    {
+      "epoch": 2.9358239032708107,
+      "grad_norm": 0.3116869032382965,
+      "learning_rate": 1.3736544049111622e-08,
+      "loss": 0.2983,
+      "step": 6313
+    },
+    {
+      "epoch": 2.9362889474500076,
+      "grad_norm": 0.32688966393470764,
+      "learning_rate": 1.3536832638643693e-08,
+      "loss": 0.3142,
+      "step": 6314
+    },
+    {
+      "epoch": 2.936753991629205,
+      "grad_norm": 0.3033595681190491,
+      "learning_rate": 1.333858168224178e-08,
+      "loss": 0.3131,
+      "step": 6315
+    },
+    {
+      "epoch": 2.937219035808402,
+      "grad_norm": 0.3291476368904114,
+      "learning_rate": 1.3141791237969991e-08,
+      "loss": 0.3139,
+      "step": 6316
+    },
+    {
+      "epoch": 2.9376840799875987,
+      "grad_norm": 0.3021133542060852,
+      "learning_rate": 1.2946461363465557e-08,
+      "loss": 0.2932,
+      "step": 6317
+    },
+    {
+      "epoch": 2.938149124166796,
+      "grad_norm": 0.32439282536506653,
+      "learning_rate": 1.2752592115936601e-08,
+      "loss": 0.3559,
+      "step": 6318
+    },
+    {
+      "epoch": 2.938614168345993,
+      "grad_norm": 0.3350219428539276,
+      "learning_rate": 1.2560183552164928e-08,
+      "loss": 0.3485,
+      "step": 6319
+    },
+    {
+      "epoch": 2.93907921252519,
+      "grad_norm": 0.3032825291156769,
+      "learning_rate": 1.2369235728503792e-08,
+      "loss": 0.315,
+      "step": 6320
+    },
+    {
+      "epoch": 2.939544256704387,
+      "grad_norm": 0.311538428068161,
+      "learning_rate": 1.2179748700879013e-08,
+      "loss": 0.325,
+      "step": 6321
+    },
+    {
+      "epoch": 2.9400093008835837,
+      "grad_norm": 0.3261773884296417,
+      "learning_rate": 1.1991722524787307e-08,
+      "loss": 0.3301,
+      "step": 6322
+    },
+    {
+      "epoch": 2.940474345062781,
+      "grad_norm": 0.33156725764274597,
+      "learning_rate": 1.1805157255299626e-08,
+      "loss": 0.3229,
+      "step": 6323
+    },
+    {
+      "epoch": 2.940939389241978,
+      "grad_norm": 0.3547118008136749,
+      "learning_rate": 1.1620052947056703e-08,
+      "loss": 0.2986,
+      "step": 6324
+    },
+    {
+      "epoch": 2.941404433421175,
+      "grad_norm": 0.32039710879325867,
+      "learning_rate": 1.1436409654273506e-08,
+      "loss": 0.3285,
+      "step": 6325
+    },
+    {
+      "epoch": 2.941869477600372,
+      "grad_norm": 0.31532758474349976,
+      "learning_rate": 1.1254227430735898e-08,
+      "loss": 0.3099,
+      "step": 6326
+    },
+    {
+      "epoch": 2.942334521779569,
+      "grad_norm": 0.33241426944732666,
+      "learning_rate": 1.1073506329802309e-08,
+      "loss": 0.3177,
+      "step": 6327
+    },
+    {
+      "epoch": 2.942799565958766,
+      "grad_norm": 0.3229113221168518,
+      "learning_rate": 1.089424640440262e-08,
+      "loss": 0.3215,
+      "step": 6328
+    },
+    {
+      "epoch": 2.9432646101379634,
+      "grad_norm": 0.3332347273826599,
+      "learning_rate": 1.0716447707039279e-08,
+      "loss": 0.351,
+      "step": 6329
+    },
+    {
+      "epoch": 2.9437296543171603,
+      "grad_norm": 0.3146977126598358,
+      "learning_rate": 1.0540110289786742e-08,
+      "loss": 0.3142,
+      "step": 6330
+    },
+    {
+      "epoch": 2.944194698496357,
+      "grad_norm": 0.3125641942024231,
+      "learning_rate": 1.0365234204291475e-08,
+      "loss": 0.2814,
+      "step": 6331
+    },
+    {
+      "epoch": 2.944659742675554,
+      "grad_norm": 0.32364359498023987,
+      "learning_rate": 1.019181950177195e-08,
+      "loss": 0.3173,
+      "step": 6332
+    },
+    {
+      "epoch": 2.945124786854751,
+      "grad_norm": 0.3259218633174896,
+      "learning_rate": 1.001986623301754e-08,
+      "loss": 0.3478,
+      "step": 6333
+    },
+    {
+      "epoch": 2.9455898310339483,
+      "grad_norm": 0.3398853540420532,
+      "learning_rate": 9.849374448391846e-09,
+      "loss": 0.3149,
+      "step": 6334
+    },
+    {
+      "epoch": 2.9460548752131452,
+      "grad_norm": 0.31736600399017334,
+      "learning_rate": 9.680344197828262e-09,
+      "loss": 0.2984,
+      "step": 6335
+    },
+    {
+      "epoch": 2.946519919392342,
+      "grad_norm": 0.3133545219898224,
+      "learning_rate": 9.512775530833296e-09,
+      "loss": 0.3267,
+      "step": 6336
+    },
+    {
+      "epoch": 2.9469849635715395,
+      "grad_norm": 0.32863521575927734,
+      "learning_rate": 9.346668496485468e-09,
+      "loss": 0.3461,
+      "step": 6337
+    },
+    {
+      "epoch": 2.9474500077507364,
+      "grad_norm": 0.3177638053894043,
+      "learning_rate": 9.182023143434193e-09,
+      "loss": 0.2977,
+      "step": 6338
+    },
+    {
+      "epoch": 2.9479150519299333,
+      "grad_norm": 0.31308892369270325,
+      "learning_rate": 9.018839519902012e-09,
+      "loss": 0.3232,
+      "step": 6339
+    },
+    {
+      "epoch": 2.94838009610913,
+      "grad_norm": 0.3451671600341797,
+      "learning_rate": 8.857117673681804e-09,
+      "loss": 0.3357,
+      "step": 6340
+    },
+    {
+      "epoch": 2.948845140288327,
+      "grad_norm": 0.32019293308258057,
+      "learning_rate": 8.696857652140677e-09,
+      "loss": 0.3324,
+      "step": 6341
+    },
+    {
+      "epoch": 2.9493101844675245,
+      "grad_norm": 0.33214157819747925,
+      "learning_rate": 8.538059502214979e-09,
+      "loss": 0.2954,
+      "step": 6342
+    },
+    {
+      "epoch": 2.9497752286467214,
+      "grad_norm": 0.33576342463493347,
+      "learning_rate": 8.380723270414726e-09,
+      "loss": 0.3511,
+      "step": 6343
+    },
+    {
+      "epoch": 2.9502402728259183,
+      "grad_norm": 0.32539987564086914,
+      "learning_rate": 8.224849002820834e-09,
+      "loss": 0.3342,
+      "step": 6344
+    },
+    {
+      "epoch": 2.9507053170051156,
+      "grad_norm": 0.31811603903770447,
+      "learning_rate": 8.07043674508623e-09,
+      "loss": 0.3285,
+      "step": 6345
+    },
+    {
+      "epoch": 2.9511703611843125,
+      "grad_norm": 0.30037549138069153,
+      "learning_rate": 7.917486542436404e-09,
+      "loss": 0.3136,
+      "step": 6346
+    },
+    {
+      "epoch": 2.9516354053635094,
+      "grad_norm": 0.3149525225162506,
+      "learning_rate": 7.765998439667743e-09,
+      "loss": 0.3187,
+      "step": 6347
+    },
+    {
+      "epoch": 2.952100449542707,
+      "grad_norm": 0.3301972448825836,
+      "learning_rate": 7.615972481148094e-09,
+      "loss": 0.329,
+      "step": 6348
+    },
+    {
+      "epoch": 2.9525654937219037,
+      "grad_norm": 0.3264675736427307,
+      "learning_rate": 7.467408710818414e-09,
+      "loss": 0.335,
+      "step": 6349
+    },
+    {
+      "epoch": 2.9530305379011006,
+      "grad_norm": 0.32105743885040283,
+      "learning_rate": 7.320307172190011e-09,
+      "loss": 0.3006,
+      "step": 6350
+    },
+    {
+      "epoch": 2.9534955820802975,
+      "grad_norm": 0.3135838806629181,
+      "learning_rate": 7.174667908346755e-09,
+      "loss": 0.3276,
+      "step": 6351
+    },
+    {
+      "epoch": 2.9539606262594944,
+      "grad_norm": 0.36441266536712646,
+      "learning_rate": 7.0304909619439695e-09,
+      "loss": 0.347,
+      "step": 6352
+    },
+    {
+      "epoch": 2.9544256704386918,
+      "grad_norm": 0.32082051038742065,
+      "learning_rate": 6.887776375208433e-09,
+      "loss": 0.3151,
+      "step": 6353
+    },
+    {
+      "epoch": 2.9548907146178887,
+      "grad_norm": 0.3240443170070648,
+      "learning_rate": 6.74652418994004e-09,
+      "loss": 0.3199,
+      "step": 6354
+    },
+    {
+      "epoch": 2.9553557587970856,
+      "grad_norm": 0.31917044520378113,
+      "learning_rate": 6.606734447507923e-09,
+      "loss": 0.306,
+      "step": 6355
+    },
+    {
+      "epoch": 2.955820802976283,
+      "grad_norm": 0.3240707218647003,
+      "learning_rate": 6.4684071888554415e-09,
+      "loss": 0.3439,
+      "step": 6356
+    },
+    {
+      "epoch": 2.95628584715548,
+      "grad_norm": 0.3343864679336548,
+      "learning_rate": 6.331542454495188e-09,
+      "loss": 0.299,
+      "step": 6357
+    },
+    {
+      "epoch": 2.9567508913346767,
+      "grad_norm": 0.3145192563533783,
+      "learning_rate": 6.19614028451343e-09,
+      "loss": 0.2969,
+      "step": 6358
+    },
+    {
+      "epoch": 2.957215935513874,
+      "grad_norm": 0.3098326027393341,
+      "learning_rate": 6.062200718567335e-09,
+      "loss": 0.3037,
+      "step": 6359
+    },
+    {
+      "epoch": 2.957680979693071,
+      "grad_norm": 0.3226139545440674,
+      "learning_rate": 5.929723795884967e-09,
+      "loss": 0.303,
+      "step": 6360
+    },
+    {
+      "epoch": 2.958146023872268,
+      "grad_norm": 0.31032806634902954,
+      "learning_rate": 5.798709555266957e-09,
+      "loss": 0.3072,
+      "step": 6361
+    },
+    {
+      "epoch": 2.958611068051465,
+      "grad_norm": 0.3300245702266693,
+      "learning_rate": 5.669158035085387e-09,
+      "loss": 0.3254,
+      "step": 6362
+    },
+    {
+      "epoch": 2.9590761122306617,
+      "grad_norm": 0.30949413776397705,
+      "learning_rate": 5.5410692732837946e-09,
+      "loss": 0.3046,
+      "step": 6363
+    },
+    {
+      "epoch": 2.959541156409859,
+      "grad_norm": 0.32829898595809937,
+      "learning_rate": 5.414443307377171e-09,
+      "loss": 0.3468,
+      "step": 6364
+    },
+    {
+      "epoch": 2.960006200589056,
+      "grad_norm": 0.3059823215007782,
+      "learning_rate": 5.2892801744525154e-09,
+      "loss": 0.3097,
+      "step": 6365
+    },
+    {
+      "epoch": 2.960471244768253,
+      "grad_norm": 0.3185559809207916,
+      "learning_rate": 5.165579911167729e-09,
+      "loss": 0.304,
+      "step": 6366
+    },
+    {
+      "epoch": 2.96093628894745,
+      "grad_norm": 0.3140532076358795,
+      "learning_rate": 5.043342553752717e-09,
+      "loss": 0.2994,
+      "step": 6367
+    },
+    {
+      "epoch": 2.961401333126647,
+      "grad_norm": 0.3156139552593231,
+      "learning_rate": 4.922568138008843e-09,
+      "loss": 0.3252,
+      "step": 6368
+    },
+    {
+      "epoch": 2.961866377305844,
+      "grad_norm": 0.3156965374946594,
+      "learning_rate": 4.803256699308923e-09,
+      "loss": 0.3367,
+      "step": 6369
+    },
+    {
+      "epoch": 2.962331421485041,
+      "grad_norm": 0.3089465796947479,
+      "learning_rate": 4.685408272597225e-09,
+      "loss": 0.3097,
+      "step": 6370
+    },
+    {
+      "epoch": 2.9627964656642383,
+      "grad_norm": 0.3329167664051056,
+      "learning_rate": 4.5690228923894744e-09,
+      "loss": 0.3361,
+      "step": 6371
+    },
+    {
+      "epoch": 2.963261509843435,
+      "grad_norm": 0.3192375600337982,
+      "learning_rate": 4.454100592773958e-09,
+      "loss": 0.3245,
+      "step": 6372
+    },
+    {
+      "epoch": 2.963726554022632,
+      "grad_norm": 0.31667426228523254,
+      "learning_rate": 4.340641407408752e-09,
+      "loss": 0.3076,
+      "step": 6373
+    },
+    {
+      "epoch": 2.964191598201829,
+      "grad_norm": 0.3276313841342926,
+      "learning_rate": 4.228645369523943e-09,
+      "loss": 0.3068,
+      "step": 6374
+    },
+    {
+      "epoch": 2.9646566423810263,
+      "grad_norm": 0.3252321481704712,
+      "learning_rate": 4.1181125119221785e-09,
+      "loss": 0.3306,
+      "step": 6375
+    },
+    {
+      "epoch": 2.9651216865602232,
+      "grad_norm": 0.3393605351448059,
+      "learning_rate": 4.009042866976454e-09,
+      "loss": 0.3405,
+      "step": 6376
+    },
+    {
+      "epoch": 2.96558673073942,
+      "grad_norm": 0.3192119002342224,
+      "learning_rate": 3.901436466631215e-09,
+      "loss": 0.3125,
+      "step": 6377
+    },
+    {
+      "epoch": 2.9660517749186175,
+      "grad_norm": 0.32561612129211426,
+      "learning_rate": 3.795293342402917e-09,
+      "loss": 0.331,
+      "step": 6378
+    },
+    {
+      "epoch": 2.9665168190978144,
+      "grad_norm": 0.29712536931037903,
+      "learning_rate": 3.690613525379472e-09,
+      "loss": 0.3094,
+      "step": 6379
+    },
+    {
+      "epoch": 2.9669818632770113,
+      "grad_norm": 0.3177908658981323,
+      "learning_rate": 3.587397046219132e-09,
+      "loss": 0.32,
+      "step": 6380
+    },
+    {
+      "epoch": 2.967446907456208,
+      "grad_norm": 0.334224134683609,
+      "learning_rate": 3.485643935152716e-09,
+      "loss": 0.3327,
+      "step": 6381
+    },
+    {
+      "epoch": 2.967911951635405,
+      "grad_norm": 0.3135557174682617,
+      "learning_rate": 3.3853542219819403e-09,
+      "loss": 0.3077,
+      "step": 6382
+    },
+    {
+      "epoch": 2.9683769958146025,
+      "grad_norm": 0.3370562195777893,
+      "learning_rate": 3.286527936079975e-09,
+      "loss": 0.3069,
+      "step": 6383
+    },
+    {
+      "epoch": 2.9688420399937994,
+      "grad_norm": 0.2906818091869354,
+      "learning_rate": 3.1891651063920003e-09,
+      "loss": 0.2884,
+      "step": 6384
+    },
+    {
+      "epoch": 2.9693070841729963,
+      "grad_norm": 0.3011315166950226,
+      "learning_rate": 3.0932657614329843e-09,
+      "loss": 0.3154,
+      "step": 6385
+    },
+    {
+      "epoch": 2.9697721283521936,
+      "grad_norm": 0.3322399854660034,
+      "learning_rate": 2.998829929291569e-09,
+      "loss": 0.3232,
+      "step": 6386
+    },
+    {
+      "epoch": 2.9702371725313905,
+      "grad_norm": 0.31616201996803284,
+      "learning_rate": 2.905857637625076e-09,
+      "loss": 0.338,
+      "step": 6387
+    },
+    {
+      "epoch": 2.9707022167105874,
+      "grad_norm": 0.2996443510055542,
+      "learning_rate": 2.8143489136650547e-09,
+      "loss": 0.2643,
+      "step": 6388
+    },
+    {
+      "epoch": 2.971167260889785,
+      "grad_norm": 0.330763578414917,
+      "learning_rate": 2.724303784211735e-09,
+      "loss": 0.3372,
+      "step": 6389
+    },
+    {
+      "epoch": 2.9716323050689817,
+      "grad_norm": 0.3177904486656189,
+      "learning_rate": 2.635722275638464e-09,
+      "loss": 0.3079,
+      "step": 6390
+    },
+    {
+      "epoch": 2.9720973492481786,
+      "grad_norm": 0.32477807998657227,
+      "learning_rate": 2.548604413888933e-09,
+      "loss": 0.3184,
+      "step": 6391
+    },
+    {
+      "epoch": 2.9725623934273755,
+      "grad_norm": 0.29389679431915283,
+      "learning_rate": 2.462950224478844e-09,
+      "loss": 0.2963,
+      "step": 6392
+    },
+    {
+      "epoch": 2.9730274376065724,
+      "grad_norm": 0.3086054027080536,
+      "learning_rate": 2.3787597324947953e-09,
+      "loss": 0.3402,
+      "step": 6393
+    },
+    {
+      "epoch": 2.9734924817857697,
+      "grad_norm": 0.2977517247200012,
+      "learning_rate": 2.2960329625953957e-09,
+      "loss": 0.3352,
+      "step": 6394
+    },
+    {
+      "epoch": 2.9739575259649667,
+      "grad_norm": 0.31417447328567505,
+      "learning_rate": 2.2147699390090425e-09,
+      "loss": 0.3128,
+      "step": 6395
+    },
+    {
+      "epoch": 2.9744225701441636,
+      "grad_norm": 0.3330288827419281,
+      "learning_rate": 2.134970685536697e-09,
+      "loss": 0.3592,
+      "step": 6396
+    },
+    {
+      "epoch": 2.974887614323361,
+      "grad_norm": 0.30559176206588745,
+      "learning_rate": 2.056635225550219e-09,
+      "loss": 0.2894,
+      "step": 6397
+    },
+    {
+      "epoch": 2.975352658502558,
+      "grad_norm": 0.31366166472435,
+      "learning_rate": 1.9797635819934768e-09,
+      "loss": 0.3098,
+      "step": 6398
+    },
+    {
+      "epoch": 2.9758177026817547,
+      "grad_norm": 0.30758044123649597,
+      "learning_rate": 1.904355777379574e-09,
+      "loss": 0.2947,
+      "step": 6399
+    },
+    {
+      "epoch": 2.976282746860952,
+      "grad_norm": 0.2907547652721405,
+      "learning_rate": 1.830411833795287e-09,
+      "loss": 0.298,
+      "step": 6400
+    },
+    {
+      "epoch": 2.976747791040149,
+      "grad_norm": 0.3252081274986267,
+      "learning_rate": 1.7579317728977363e-09,
+      "loss": 0.3325,
+      "step": 6401
+    },
+    {
+      "epoch": 2.977212835219346,
+      "grad_norm": 0.3026246428489685,
+      "learning_rate": 1.6869156159143861e-09,
+      "loss": 0.3322,
+      "step": 6402
+    },
+    {
+      "epoch": 2.977677879398543,
+      "grad_norm": 0.3425191640853882,
+      "learning_rate": 1.617363383645265e-09,
+      "loss": 0.3364,
+      "step": 6403
+    },
+    {
+      "epoch": 2.9781429235777397,
+      "grad_norm": 0.32005518674850464,
+      "learning_rate": 1.549275096460745e-09,
+      "loss": 0.332,
+      "step": 6404
+    },
+    {
+      "epoch": 2.978607967756937,
+      "grad_norm": 0.2961218059062958,
+      "learning_rate": 1.4826507743032071e-09,
+      "loss": 0.2848,
+      "step": 6405
+    },
+    {
+      "epoch": 2.979073011936134,
+      "grad_norm": 0.3002302944660187,
+      "learning_rate": 1.417490436685376e-09,
+      "loss": 0.3021,
+      "step": 6406
+    },
+    {
+      "epoch": 2.979538056115331,
+      "grad_norm": 0.3076646625995636,
+      "learning_rate": 1.3537941026914302e-09,
+      "loss": 0.3196,
+      "step": 6407
+    },
+    {
+      "epoch": 2.980003100294528,
+      "grad_norm": 0.29915478825569153,
+      "learning_rate": 1.291561790978113e-09,
+      "loss": 0.283,
+      "step": 6408
+    },
+    {
+      "epoch": 2.980468144473725,
+      "grad_norm": 0.313909113407135,
+      "learning_rate": 1.2307935197708453e-09,
+      "loss": 0.3223,
+      "step": 6409
+    },
+    {
+      "epoch": 2.980933188652922,
+      "grad_norm": 0.33077502250671387,
+      "learning_rate": 1.1714893068687228e-09,
+      "loss": 0.3171,
+      "step": 6410
+    },
+    {
+      "epoch": 2.981398232832119,
+      "grad_norm": 0.3245851397514343,
+      "learning_rate": 1.1136491696406293e-09,
+      "loss": 0.3136,
+      "step": 6411
+    },
+    {
+      "epoch": 2.981863277011316,
+      "grad_norm": 0.3183169364929199,
+      "learning_rate": 1.057273125026903e-09,
+      "loss": 0.3079,
+      "step": 6412
+    },
+    {
+      "epoch": 2.982328321190513,
+      "grad_norm": 0.3176499605178833,
+      "learning_rate": 1.0023611895393358e-09,
+      "loss": 0.3185,
+      "step": 6413
+    },
+    {
+      "epoch": 2.98279336536971,
+      "grad_norm": 0.29598644375801086,
+      "learning_rate": 9.489133792611738e-10,
+      "loss": 0.2992,
+      "step": 6414
+    },
+    {
+      "epoch": 2.983258409548907,
+      "grad_norm": 0.32176917791366577,
+      "learning_rate": 8.96929709845451e-10,
+      "loss": 0.3118,
+      "step": 6415
+    },
+    {
+      "epoch": 2.9837234537281043,
+      "grad_norm": 0.32616156339645386,
+      "learning_rate": 8.464101965177662e-10,
+      "loss": 0.342,
+      "step": 6416
+    },
+    {
+      "epoch": 2.9841884979073012,
+      "grad_norm": 0.3436664640903473,
+      "learning_rate": 7.973548540740616e-10,
+      "loss": 0.323,
+      "step": 6417
+    },
+    {
+      "epoch": 2.984653542086498,
+      "grad_norm": 0.3062160313129425,
+      "learning_rate": 7.497636968828436e-10,
+      "loss": 0.2861,
+      "step": 6418
+    },
+    {
+      "epoch": 2.9851185862656955,
+      "grad_norm": 0.3244827389717102,
+      "learning_rate": 7.036367388824073e-10,
+      "loss": 0.3261,
+      "step": 6419
+    },
+    {
+      "epoch": 2.9855836304448924,
+      "grad_norm": 0.3119675815105438,
+      "learning_rate": 6.589739935819461e-10,
+      "loss": 0.3165,
+      "step": 6420
+    },
+    {
+      "epoch": 2.9860486746240893,
+      "grad_norm": 0.3089711666107178,
+      "learning_rate": 6.157754740632183e-10,
+      "loss": 0.3039,
+      "step": 6421
+    },
+    {
+      "epoch": 2.986513718803286,
+      "grad_norm": 0.3472222089767456,
+      "learning_rate": 5.7404119297777e-10,
+      "loss": 0.2805,
+      "step": 6422
+    },
+    {
+      "epoch": 2.986978762982483,
+      "grad_norm": 0.343313604593277,
+      "learning_rate": 5.337711625497122e-10,
+      "loss": 0.3385,
+      "step": 6423
+    },
+    {
+      "epoch": 2.9874438071616805,
+      "grad_norm": 0.34118393063545227,
+      "learning_rate": 4.949653945723886e-10,
+      "loss": 0.3172,
+      "step": 6424
+    },
+    {
+      "epoch": 2.9879088513408774,
+      "grad_norm": 0.331533282995224,
+      "learning_rate": 4.576239004122629e-10,
+      "loss": 0.3628,
+      "step": 6425
+    },
+    {
+      "epoch": 2.9883738955200743,
+      "grad_norm": 0.30840861797332764,
+      "learning_rate": 4.2174669100558673e-10,
+      "loss": 0.2938,
+      "step": 6426
+    },
+    {
+      "epoch": 2.9888389396992716,
+      "grad_norm": 0.3454267382621765,
+      "learning_rate": 3.8733377686062115e-10,
+      "loss": 0.3555,
+      "step": 6427
+    },
+    {
+      "epoch": 2.9893039838784685,
+      "grad_norm": 0.3079105019569397,
+      "learning_rate": 3.5438516805597067e-10,
+      "loss": 0.3275,
+      "step": 6428
+    },
+    {
+      "epoch": 2.9897690280576654,
+      "grad_norm": 0.32878822088241577,
+      "learning_rate": 3.229008742416939e-10,
+      "loss": 0.3502,
+      "step": 6429
+    },
+    {
+      "epoch": 2.9902340722368628,
+      "grad_norm": 0.3611544072628021,
+      "learning_rate": 2.928809046398584e-10,
+      "loss": 0.3244,
+      "step": 6430
+    },
+    {
+      "epoch": 2.9906991164160597,
+      "grad_norm": 0.3192159831523895,
+      "learning_rate": 2.6432526804176517e-10,
+      "loss": 0.3126,
+      "step": 6431
+    },
+    {
+      "epoch": 2.9911641605952566,
+      "grad_norm": 0.353327214717865,
+      "learning_rate": 2.3723397281127937e-10,
+      "loss": 0.3284,
+      "step": 6432
+    },
+    {
+      "epoch": 2.9916292047744535,
+      "grad_norm": 0.31189650297164917,
+      "learning_rate": 2.1160702688260981e-10,
+      "loss": 0.3192,
+      "step": 6433
+    },
+    {
+      "epoch": 2.9920942489536504,
+      "grad_norm": 0.31306004524230957,
+      "learning_rate": 1.8744443776252952e-10,
+      "loss": 0.3168,
+      "step": 6434
+    },
+    {
+      "epoch": 2.9925592931328477,
+      "grad_norm": 0.3070635497570038,
+      "learning_rate": 1.6474621252704494e-10,
+      "loss": 0.3385,
+      "step": 6435
+    },
+    {
+      "epoch": 2.9930243373120446,
+      "grad_norm": 0.32101500034332275,
+      "learning_rate": 1.435123578241715e-10,
+      "loss": 0.3346,
+      "step": 6436
+    },
+    {
+      "epoch": 2.9934893814912416,
+      "grad_norm": 0.3093291223049164,
+      "learning_rate": 1.2374287987337864e-10,
+      "loss": 0.3378,
+      "step": 6437
+    },
+    {
+      "epoch": 2.993954425670439,
+      "grad_norm": 0.3259122371673584,
+      "learning_rate": 1.0543778446392427e-10,
+      "loss": 0.3228,
+      "step": 6438
+    },
+    {
+      "epoch": 2.994419469849636,
+      "grad_norm": 0.31066158413887024,
+      "learning_rate": 8.859707695818564e-11,
+      "loss": 0.29,
+      "step": 6439
+    },
+    {
+      "epoch": 2.9948845140288327,
+      "grad_norm": 0.31667444109916687,
+      "learning_rate": 7.322076228777342e-11,
+      "loss": 0.3338,
+      "step": 6440
+    },
+    {
+      "epoch": 2.9953495582080296,
+      "grad_norm": 0.3047296404838562,
+      "learning_rate": 5.93088449568624e-11,
+      "loss": 0.2968,
+      "step": 6441
+    },
+    {
+      "epoch": 2.9958146023872265,
+      "grad_norm": 0.3278631567955017,
+      "learning_rate": 4.686132903886087e-11,
+      "loss": 0.3333,
+      "step": 6442
+    },
+    {
+      "epoch": 2.996279646566424,
+      "grad_norm": 0.3081776201725006,
+      "learning_rate": 3.587821818085146e-11,
+      "loss": 0.2953,
+      "step": 6443
+    },
+    {
+      "epoch": 2.9967446907456208,
+      "grad_norm": 0.3233565092086792,
+      "learning_rate": 2.6359515598595174e-11,
+      "loss": 0.3309,
+      "step": 6444
+    },
+    {
+      "epoch": 2.9972097349248177,
+      "grad_norm": 0.29509052634239197,
+      "learning_rate": 1.8305224079862015e-11,
+      "loss": 0.3168,
+      "step": 6445
+    },
+    {
+      "epoch": 2.997674779104015,
+      "grad_norm": 0.3146141469478607,
+      "learning_rate": 1.1715345984431026e-11,
+      "loss": 0.3229,
+      "step": 6446
+    },
+    {
+      "epoch": 2.998139823283212,
+      "grad_norm": 0.3134753406047821,
+      "learning_rate": 6.5898832424249324e-12,
+      "loss": 0.3339,
+      "step": 6447
+    },
+    {
+      "epoch": 2.998604867462409,
+      "grad_norm": 0.30789443850517273,
+      "learning_rate": 2.928837353755043e-12,
+      "loss": 0.3139,
+      "step": 6448
+    },
+    {
+      "epoch": 2.999069911641606,
+      "grad_norm": 0.31371957063674927,
+      "learning_rate": 7.322093920070217e-13,
+      "loss": 0.3199,
+      "step": 6449
+    },
+    {
+      "epoch": 2.999534955820803,
+      "grad_norm": 0.3083919286727905,
+      "learning_rate": 0.0,
+      "loss": 0.3054,
+      "step": 6450
+    },
+    {
+      "epoch": 2.999534955820803,
+      "step": 6450,
+      "total_flos": 6039468541263872.0,
+      "train_loss": 0.3618491206889929,
+      "train_runtime": 197008.5242,
+      "train_samples_per_second": 3.143,
+      "train_steps_per_second": 0.033
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6450,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6039468541263872.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}