diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17851 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 2544,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011792452830188679,
+      "grad_norm": 52.72592939639735,
+      "learning_rate": 0.0,
+      "loss": 11.2614,
+      "step": 1
+    },
+    {
+      "epoch": 0.0023584905660377358,
+      "grad_norm": 52.44856988830773,
+      "learning_rate": 1.9607843137254904e-07,
+      "loss": 11.23,
+      "step": 2
+    },
+    {
+      "epoch": 0.003537735849056604,
+      "grad_norm": 54.12721690800911,
+      "learning_rate": 3.921568627450981e-07,
+      "loss": 11.2839,
+      "step": 3
+    },
+    {
+      "epoch": 0.0047169811320754715,
+      "grad_norm": 52.780578768983275,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 11.2233,
+      "step": 4
+    },
+    {
+      "epoch": 0.00589622641509434,
+      "grad_norm": 53.91399030099141,
+      "learning_rate": 7.843137254901962e-07,
+      "loss": 11.1155,
+      "step": 5
+    },
+    {
+      "epoch": 0.007075471698113208,
+      "grad_norm": 54.21730126159324,
+      "learning_rate": 9.80392156862745e-07,
+      "loss": 11.0729,
+      "step": 6
+    },
+    {
+      "epoch": 0.008254716981132075,
+      "grad_norm": 56.070307783860805,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 11.046,
+      "step": 7
+    },
+    {
+      "epoch": 0.009433962264150943,
+      "grad_norm": 60.68878768102106,
+      "learning_rate": 1.3725490196078432e-06,
+      "loss": 10.8062,
+      "step": 8
+    },
+    {
+      "epoch": 0.01061320754716981,
+      "grad_norm": 57.28310541455512,
+      "learning_rate": 1.5686274509803923e-06,
+      "loss": 10.8488,
+      "step": 9
+    },
+    {
+      "epoch": 0.01179245283018868,
+      "grad_norm": 59.89815937247923,
+      "learning_rate": 1.7647058823529412e-06,
+      "loss": 10.7459,
+      "step": 10
+    },
+    {
+      "epoch": 0.012971698113207548,
+      "grad_norm": 82.51283087217136,
+      "learning_rate": 1.96078431372549e-06,
+      "loss": 9.5469,
+      "step": 11
+    },
+    {
+      "epoch": 0.014150943396226415,
+      "grad_norm": 81.78269585480245,
+      "learning_rate": 2.1568627450980393e-06,
+      "loss": 9.6243,
+      "step": 12
+    },
+    {
+      "epoch": 0.015330188679245283,
+      "grad_norm": 94.56482157285834,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 9.2047,
+      "step": 13
+    },
+    {
+      "epoch": 0.01650943396226415,
+      "grad_norm": 107.54352507218434,
+      "learning_rate": 2.549019607843137e-06,
+      "loss": 5.4186,
+      "step": 14
+    },
+    {
+      "epoch": 0.01768867924528302,
+      "grad_norm": 65.14244110636824,
+      "learning_rate": 2.7450980392156863e-06,
+      "loss": 3.8615,
+      "step": 15
+    },
+    {
+      "epoch": 0.018867924528301886,
+      "grad_norm": 56.522969880457524,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 3.4723,
+      "step": 16
+    },
+    {
+      "epoch": 0.020047169811320754,
+      "grad_norm": 39.687375416866544,
+      "learning_rate": 3.1372549019607846e-06,
+      "loss": 2.7613,
+      "step": 17
+    },
+    {
+      "epoch": 0.02122641509433962,
+      "grad_norm": 34.853305236518054,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 2.774,
+      "step": 18
+    },
+    {
+      "epoch": 0.02240566037735849,
+      "grad_norm": 11.134357159763422,
+      "learning_rate": 3.5294117647058825e-06,
+      "loss": 1.848,
+      "step": 19
+    },
+    {
+      "epoch": 0.02358490566037736,
+      "grad_norm": 7.326268202325733,
+      "learning_rate": 3.7254901960784316e-06,
+      "loss": 1.521,
+      "step": 20
+    },
+    {
+      "epoch": 0.024764150943396228,
+      "grad_norm": 5.269250601116452,
+      "learning_rate": 3.92156862745098e-06,
+      "loss": 1.4373,
+      "step": 21
+    },
+    {
+      "epoch": 0.025943396226415096,
+      "grad_norm": 4.32188069057022,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 1.4667,
+      "step": 22
+    },
+    {
+      "epoch": 0.027122641509433963,
+      "grad_norm": 3.2226157285235044,
+      "learning_rate": 4.313725490196079e-06,
+      "loss": 1.3178,
+      "step": 23
+    },
+    {
+      "epoch": 0.02830188679245283,
+      "grad_norm": 2.608975775320711,
+      "learning_rate": 4.509803921568627e-06,
+      "loss": 1.2731,
+      "step": 24
+    },
+    {
+      "epoch": 0.0294811320754717,
+      "grad_norm": 2.142046778223866,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 1.2115,
+      "step": 25
+    },
+    {
+      "epoch": 0.030660377358490566,
+      "grad_norm": 1.5886254924847858,
+      "learning_rate": 4.901960784313726e-06,
+      "loss": 1.1567,
+      "step": 26
+    },
+    {
+      "epoch": 0.031839622641509434,
+      "grad_norm": 3.62968187828875,
+      "learning_rate": 5.098039215686274e-06,
+      "loss": 1.0813,
+      "step": 27
+    },
+    {
+      "epoch": 0.0330188679245283,
+      "grad_norm": 8.746999443686875,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 1.0072,
+      "step": 28
+    },
+    {
+      "epoch": 0.03419811320754717,
+      "grad_norm": 1.4893943463788462,
+      "learning_rate": 5.490196078431373e-06,
+      "loss": 1.0095,
+      "step": 29
+    },
+    {
+      "epoch": 0.03537735849056604,
+      "grad_norm": 1.2060990167082555,
+      "learning_rate": 5.686274509803921e-06,
+      "loss": 1.0807,
+      "step": 30
+    },
+    {
+      "epoch": 0.036556603773584904,
+      "grad_norm": 1.0065249539642935,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 1.0103,
+      "step": 31
+    },
+    {
+      "epoch": 0.03773584905660377,
+      "grad_norm": 0.881022330208351,
+      "learning_rate": 6.07843137254902e-06,
+      "loss": 0.9292,
+      "step": 32
+    },
+    {
+      "epoch": 0.03891509433962264,
+      "grad_norm": 0.848548154499319,
+      "learning_rate": 6.274509803921569e-06,
+      "loss": 0.9174,
+      "step": 33
+    },
+    {
+      "epoch": 0.04009433962264151,
+      "grad_norm": 0.8050012523786563,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 0.9559,
+      "step": 34
+    },
+    {
+      "epoch": 0.041273584905660375,
+      "grad_norm": 0.7996754833388382,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.8686,
+      "step": 35
+    },
+    {
+      "epoch": 0.04245283018867924,
+      "grad_norm": 0.680307846077049,
+      "learning_rate": 6.862745098039216e-06,
+      "loss": 0.9173,
+      "step": 36
+    },
+    {
+      "epoch": 0.04363207547169811,
+      "grad_norm": 0.609424689632246,
+      "learning_rate": 7.058823529411765e-06,
+      "loss": 0.8853,
+      "step": 37
+    },
+    {
+      "epoch": 0.04481132075471698,
+      "grad_norm": 0.6479491283555069,
+      "learning_rate": 7.2549019607843145e-06,
+      "loss": 0.9124,
+      "step": 38
+    },
+    {
+      "epoch": 0.045990566037735846,
+      "grad_norm": 0.6474386794973557,
+      "learning_rate": 7.450980392156863e-06,
+      "loss": 0.8959,
+      "step": 39
+    },
+    {
+      "epoch": 0.04716981132075472,
+      "grad_norm": 0.5970881346996586,
+      "learning_rate": 7.647058823529413e-06,
+      "loss": 0.8328,
+      "step": 40
+    },
+    {
+      "epoch": 0.04834905660377359,
+      "grad_norm": 0.5769178947433148,
+      "learning_rate": 7.84313725490196e-06,
+      "loss": 0.8161,
+      "step": 41
+    },
+    {
+      "epoch": 0.049528301886792456,
+      "grad_norm": 0.5585493974034338,
+      "learning_rate": 8.03921568627451e-06,
+      "loss": 0.8169,
+      "step": 42
+    },
+    {
+      "epoch": 0.05070754716981132,
+      "grad_norm": 0.46026061774711374,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.7802,
+      "step": 43
+    },
+    {
+      "epoch": 0.05188679245283019,
+      "grad_norm": 0.46431836397404985,
+      "learning_rate": 8.43137254901961e-06,
+      "loss": 0.8303,
+      "step": 44
+    },
+    {
+      "epoch": 0.05306603773584906,
+      "grad_norm": 0.4850580623632229,
+      "learning_rate": 8.627450980392157e-06,
+      "loss": 0.8131,
+      "step": 45
+    },
+    {
+      "epoch": 0.054245283018867926,
+      "grad_norm": 0.45192933876797614,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.8159,
+      "step": 46
+    },
+    {
+      "epoch": 0.055424528301886794,
+      "grad_norm": 0.48930795796756626,
+      "learning_rate": 9.019607843137255e-06,
+      "loss": 0.7778,
+      "step": 47
+    },
+    {
+      "epoch": 0.05660377358490566,
+      "grad_norm": 0.45750750778838317,
+      "learning_rate": 9.215686274509804e-06,
+      "loss": 0.791,
+      "step": 48
+    },
+    {
+      "epoch": 0.05778301886792453,
+      "grad_norm": 0.379750046518956,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.7572,
+      "step": 49
+    },
+    {
+      "epoch": 0.0589622641509434,
+      "grad_norm": 0.3735186130159981,
+      "learning_rate": 9.607843137254903e-06,
+      "loss": 0.7219,
+      "step": 50
+    },
+    {
+      "epoch": 0.060141509433962265,
+      "grad_norm": 0.43283756197357337,
+      "learning_rate": 9.803921568627451e-06,
+      "loss": 0.733,
+      "step": 51
+    },
+    {
+      "epoch": 0.06132075471698113,
+      "grad_norm": 0.43256622124663757,
+      "learning_rate": 1e-05,
+      "loss": 0.74,
+      "step": 52
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 0.37433449430725996,
+      "learning_rate": 1.0196078431372549e-05,
+      "loss": 0.7659,
+      "step": 53
+    },
+    {
+      "epoch": 0.06367924528301887,
+      "grad_norm": 0.3238293094838614,
+      "learning_rate": 1.0392156862745098e-05,
+      "loss": 0.73,
+      "step": 54
+    },
+    {
+      "epoch": 0.06485849056603774,
+      "grad_norm": 0.353827805104361,
+      "learning_rate": 1.0588235294117648e-05,
+      "loss": 0.7188,
+      "step": 55
+    },
+    {
+      "epoch": 0.0660377358490566,
+      "grad_norm": 0.3374847325061807,
+      "learning_rate": 1.0784313725490197e-05,
+      "loss": 0.7253,
+      "step": 56
+    },
+    {
+      "epoch": 0.06721698113207547,
+      "grad_norm": 0.32552024088645254,
+      "learning_rate": 1.0980392156862745e-05,
+      "loss": 0.7442,
+      "step": 57
+    },
+    {
+      "epoch": 0.06839622641509434,
+      "grad_norm": 0.30398050209596367,
+      "learning_rate": 1.1176470588235295e-05,
+      "loss": 0.7324,
+      "step": 58
+    },
+    {
+      "epoch": 0.0695754716981132,
+      "grad_norm": 0.3256669272532338,
+      "learning_rate": 1.1372549019607843e-05,
+      "loss": 0.7212,
+      "step": 59
+    },
+    {
+      "epoch": 0.07075471698113207,
+      "grad_norm": 0.28793884891473576,
+      "learning_rate": 1.1568627450980394e-05,
+      "loss": 0.7002,
+      "step": 60
+    },
+    {
+      "epoch": 0.07193396226415094,
+      "grad_norm": 0.2822574460158667,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 0.6505,
+      "step": 61
+    },
+    {
+      "epoch": 0.07311320754716981,
+      "grad_norm": 0.32451786011226375,
+      "learning_rate": 1.1960784313725491e-05,
+      "loss": 0.7057,
+      "step": 62
+    },
+    {
+      "epoch": 0.07429245283018868,
+      "grad_norm": 0.3173494178559797,
+      "learning_rate": 1.215686274509804e-05,
+      "loss": 0.6942,
+      "step": 63
+    },
+    {
+      "epoch": 0.07547169811320754,
+      "grad_norm": 0.2769121467253274,
+      "learning_rate": 1.2352941176470589e-05,
+      "loss": 0.7104,
+      "step": 64
+    },
+    {
+      "epoch": 0.07665094339622641,
+      "grad_norm": 0.3113300950916788,
+      "learning_rate": 1.2549019607843138e-05,
+      "loss": 0.6962,
+      "step": 65
+    },
+    {
+      "epoch": 0.07783018867924528,
+      "grad_norm": 0.33563793969658845,
+      "learning_rate": 1.2745098039215686e-05,
+      "loss": 0.7695,
+      "step": 66
+    },
+    {
+      "epoch": 0.07900943396226415,
+      "grad_norm": 0.29682356684115546,
+      "learning_rate": 1.2941176470588238e-05,
+      "loss": 0.7102,
+      "step": 67
+    },
+    {
+      "epoch": 0.08018867924528301,
+      "grad_norm": 0.2768732800186212,
+      "learning_rate": 1.3137254901960785e-05,
+      "loss": 0.6945,
+      "step": 68
+    },
+    {
+      "epoch": 0.08136792452830188,
+      "grad_norm": 0.26465500494463334,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6835,
+      "step": 69
+    },
+    {
+      "epoch": 0.08254716981132075,
+      "grad_norm": 0.28553518188103943,
+      "learning_rate": 1.3529411764705883e-05,
+      "loss": 0.6938,
+      "step": 70
+    },
+    {
+      "epoch": 0.08372641509433962,
+      "grad_norm": 0.29860603249757367,
+      "learning_rate": 1.3725490196078432e-05,
+      "loss": 0.6719,
+      "step": 71
+    },
+    {
+      "epoch": 0.08490566037735849,
+      "grad_norm": 0.30461642109298254,
+      "learning_rate": 1.3921568627450982e-05,
+      "loss": 0.6718,
+      "step": 72
+    },
+    {
+      "epoch": 0.08608490566037735,
+      "grad_norm": 0.28971032262018376,
+      "learning_rate": 1.411764705882353e-05,
+      "loss": 0.6658,
+      "step": 73
+    },
+    {
+      "epoch": 0.08726415094339622,
+      "grad_norm": 0.24752829566653028,
+      "learning_rate": 1.4313725490196078e-05,
+      "loss": 0.7009,
+      "step": 74
+    },
+    {
+      "epoch": 0.08844339622641509,
+      "grad_norm": 0.2878149125447009,
+      "learning_rate": 1.4509803921568629e-05,
+      "loss": 0.7236,
+      "step": 75
+    },
+    {
+      "epoch": 0.08962264150943396,
+      "grad_norm": 0.2798116279538163,
+      "learning_rate": 1.4705882352941177e-05,
+      "loss": 0.6758,
+      "step": 76
+    },
+    {
+      "epoch": 0.09080188679245282,
+      "grad_norm": 0.25984213728983124,
+      "learning_rate": 1.4901960784313726e-05,
+      "loss": 0.6714,
+      "step": 77
+    },
+    {
+      "epoch": 0.09198113207547169,
+      "grad_norm": 0.29247536578204714,
+      "learning_rate": 1.5098039215686274e-05,
+      "loss": 0.6643,
+      "step": 78
+    },
+    {
+      "epoch": 0.09316037735849056,
+      "grad_norm": 0.2671693236833337,
+      "learning_rate": 1.5294117647058826e-05,
+      "loss": 0.6464,
+      "step": 79
+    },
+    {
+      "epoch": 0.09433962264150944,
+      "grad_norm": 0.24006204870669592,
+      "learning_rate": 1.5490196078431373e-05,
+      "loss": 0.6301,
+      "step": 80
+    },
+    {
+      "epoch": 0.09551886792452831,
+      "grad_norm": 0.2903578504228764,
+      "learning_rate": 1.568627450980392e-05,
+      "loss": 0.6611,
+      "step": 81
+    },
+    {
+      "epoch": 0.09669811320754718,
+      "grad_norm": 0.2628745429413269,
+      "learning_rate": 1.588235294117647e-05,
+      "loss": 0.7069,
+      "step": 82
+    },
+    {
+      "epoch": 0.09787735849056604,
+      "grad_norm": 0.2611007957099753,
+      "learning_rate": 1.607843137254902e-05,
+      "loss": 0.6981,
+      "step": 83
+    },
+    {
+      "epoch": 0.09905660377358491,
+      "grad_norm": 0.2608153326168798,
+      "learning_rate": 1.627450980392157e-05,
+      "loss": 0.6589,
+      "step": 84
+    },
+    {
+      "epoch": 0.10023584905660378,
+      "grad_norm": 0.2692321164607793,
+      "learning_rate": 1.647058823529412e-05,
+      "loss": 0.6736,
+      "step": 85
+    },
+    {
+      "epoch": 0.10141509433962265,
+      "grad_norm": 0.2613578673140556,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.6812,
+      "step": 86
+    },
+    {
+      "epoch": 0.10259433962264151,
+      "grad_norm": 0.2738197185034184,
+      "learning_rate": 1.686274509803922e-05,
+      "loss": 0.6989,
+      "step": 87
+    },
+    {
+      "epoch": 0.10377358490566038,
+      "grad_norm": 0.3019586243003191,
+      "learning_rate": 1.7058823529411767e-05,
+      "loss": 0.6629,
+      "step": 88
+    },
+    {
+      "epoch": 0.10495283018867925,
+      "grad_norm": 0.2804950919214591,
+      "learning_rate": 1.7254901960784314e-05,
+      "loss": 0.6421,
+      "step": 89
+    },
+    {
+      "epoch": 0.10613207547169812,
+      "grad_norm": 0.2567324188211,
+      "learning_rate": 1.7450980392156862e-05,
+      "loss": 0.662,
+      "step": 90
+    },
+    {
+      "epoch": 0.10731132075471699,
+      "grad_norm": 0.2552689219716839,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 0.66,
+      "step": 91
+    },
+    {
+      "epoch": 0.10849056603773585,
+      "grad_norm": 0.27721230008883196,
+      "learning_rate": 1.784313725490196e-05,
+      "loss": 0.6463,
+      "step": 92
+    },
+    {
+      "epoch": 0.10966981132075472,
+      "grad_norm": 0.24653953582568974,
+      "learning_rate": 1.803921568627451e-05,
+      "loss": 0.6234,
+      "step": 93
+    },
+    {
+      "epoch": 0.11084905660377359,
+      "grad_norm": 0.2473262101299466,
+      "learning_rate": 1.8235294117647057e-05,
+      "loss": 0.5961,
+      "step": 94
+    },
+    {
+      "epoch": 0.11202830188679246,
+      "grad_norm": 0.27001085154557897,
+      "learning_rate": 1.843137254901961e-05,
+      "loss": 0.6792,
+      "step": 95
+    },
+    {
+      "epoch": 0.11320754716981132,
+      "grad_norm": 0.2638549858747966,
+      "learning_rate": 1.862745098039216e-05,
+      "loss": 0.6023,
+      "step": 96
+    },
+    {
+      "epoch": 0.11438679245283019,
+      "grad_norm": 0.2773892354363382,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 0.6371,
+      "step": 97
+    },
+    {
+      "epoch": 0.11556603773584906,
+      "grad_norm": 0.2585428639391871,
+      "learning_rate": 1.9019607843137255e-05,
+      "loss": 0.6713,
+      "step": 98
+    },
+    {
+      "epoch": 0.11674528301886793,
+      "grad_norm": 0.2616973466851501,
+      "learning_rate": 1.9215686274509807e-05,
+      "loss": 0.6204,
+      "step": 99
+    },
+    {
+      "epoch": 0.1179245283018868,
+      "grad_norm": 0.27011523399869225,
+      "learning_rate": 1.9411764705882355e-05,
+      "loss": 0.6199,
+      "step": 100
+    },
+    {
+      "epoch": 0.11910377358490566,
+      "grad_norm": 0.300588505461326,
+      "learning_rate": 1.9607843137254903e-05,
+      "loss": 0.6646,
+      "step": 101
+    },
+    {
+      "epoch": 0.12028301886792453,
+      "grad_norm": 0.25999006115282974,
+      "learning_rate": 1.980392156862745e-05,
+      "loss": 0.6565,
+      "step": 102
+    },
+    {
+      "epoch": 0.1214622641509434,
+      "grad_norm": 0.3023016170821517,
+      "learning_rate": 2e-05,
+      "loss": 0.6385,
+      "step": 103
+    },
+    {
+      "epoch": 0.12264150943396226,
+      "grad_norm": 0.2435233100624575,
+      "learning_rate": 2.019607843137255e-05,
+      "loss": 0.5963,
+      "step": 104
+    },
+    {
+      "epoch": 0.12382075471698113,
+      "grad_norm": 0.26966018960719396,
+      "learning_rate": 2.0392156862745097e-05,
+      "loss": 0.6066,
+      "step": 105
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.2420412036829079,
+      "learning_rate": 2.058823529411765e-05,
+      "loss": 0.6275,
+      "step": 106
+    },
+    {
+      "epoch": 0.12617924528301888,
+      "grad_norm": 0.28076833893372555,
+      "learning_rate": 2.0784313725490197e-05,
+      "loss": 0.6362,
+      "step": 107
+    },
+    {
+      "epoch": 0.12735849056603774,
+      "grad_norm": 0.27972893991176323,
+      "learning_rate": 2.0980392156862748e-05,
+      "loss": 0.5995,
+      "step": 108
+    },
+    {
+      "epoch": 0.12853773584905662,
+      "grad_norm": 0.270535793579532,
+      "learning_rate": 2.1176470588235296e-05,
+      "loss": 0.6121,
+      "step": 109
+    },
+    {
+      "epoch": 0.12971698113207547,
+      "grad_norm": 0.26571697040062514,
+      "learning_rate": 2.1372549019607844e-05,
+      "loss": 0.6199,
+      "step": 110
+    },
+    {
+      "epoch": 0.13089622641509435,
+      "grad_norm": 0.2977365369298319,
+      "learning_rate": 2.1568627450980395e-05,
+      "loss": 0.6206,
+      "step": 111
+    },
+    {
+      "epoch": 0.1320754716981132,
+      "grad_norm": 0.27350608050213876,
+      "learning_rate": 2.1764705882352943e-05,
+      "loss": 0.6099,
+      "step": 112
+    },
+    {
+      "epoch": 0.1332547169811321,
+      "grad_norm": 0.2592725941817156,
+      "learning_rate": 2.196078431372549e-05,
+      "loss": 0.5972,
+      "step": 113
+    },
+    {
+      "epoch": 0.13443396226415094,
+      "grad_norm": 0.3085280971473181,
+      "learning_rate": 2.215686274509804e-05,
+      "loss": 0.6024,
+      "step": 114
+    },
+    {
+      "epoch": 0.13561320754716982,
+      "grad_norm": 0.2665740771040603,
+      "learning_rate": 2.235294117647059e-05,
+      "loss": 0.6048,
+      "step": 115
+    },
+    {
+      "epoch": 0.13679245283018868,
+      "grad_norm": 0.28849378622113,
+      "learning_rate": 2.2549019607843138e-05,
+      "loss": 0.6,
+      "step": 116
+    },
+    {
+      "epoch": 0.13797169811320756,
+      "grad_norm": 0.29093662824635036,
+      "learning_rate": 2.2745098039215685e-05,
+      "loss": 0.6237,
+      "step": 117
+    },
+    {
+      "epoch": 0.1391509433962264,
+      "grad_norm": 0.29349215678674867,
+      "learning_rate": 2.2941176470588237e-05,
+      "loss": 0.6013,
+      "step": 118
+    },
+    {
+      "epoch": 0.1403301886792453,
+      "grad_norm": 0.29041627803702685,
+      "learning_rate": 2.3137254901960788e-05,
+      "loss": 0.6153,
+      "step": 119
+    },
+    {
+      "epoch": 0.14150943396226415,
+      "grad_norm": 0.30149294839952323,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.6298,
+      "step": 120
+    },
+    {
+      "epoch": 0.14268867924528303,
+      "grad_norm": 0.38737866567507945,
+      "learning_rate": 2.3529411764705884e-05,
+      "loss": 0.6209,
+      "step": 121
+    },
+    {
+      "epoch": 0.14386792452830188,
+      "grad_norm": 0.31748438350245217,
+      "learning_rate": 2.372549019607843e-05,
+      "loss": 0.6144,
+      "step": 122
+    },
+    {
+      "epoch": 0.14504716981132076,
+      "grad_norm": 0.320525457260337,
+      "learning_rate": 2.3921568627450983e-05,
+      "loss": 0.6189,
+      "step": 123
+    },
+    {
+      "epoch": 0.14622641509433962,
+      "grad_norm": 0.3278305817657651,
+      "learning_rate": 2.411764705882353e-05,
+      "loss": 0.6037,
+      "step": 124
+    },
+    {
+      "epoch": 0.1474056603773585,
+      "grad_norm": 0.31367557135687635,
+      "learning_rate": 2.431372549019608e-05,
+      "loss": 0.6488,
+      "step": 125
+    },
+    {
+      "epoch": 0.14858490566037735,
+      "grad_norm": 0.32067309797470683,
+      "learning_rate": 2.4509803921568626e-05,
+      "loss": 0.603,
+      "step": 126
+    },
+    {
+      "epoch": 0.14976415094339623,
+      "grad_norm": 0.2910780084224769,
+      "learning_rate": 2.4705882352941178e-05,
+      "loss": 0.5935,
+      "step": 127
+    },
+    {
+      "epoch": 0.1509433962264151,
+      "grad_norm": 0.3612240113533813,
+      "learning_rate": 2.4901960784313726e-05,
+      "loss": 0.636,
+      "step": 128
+    },
+    {
+      "epoch": 0.15212264150943397,
+      "grad_norm": 0.31528429428516114,
+      "learning_rate": 2.5098039215686277e-05,
+      "loss": 0.6306,
+      "step": 129
+    },
+    {
+      "epoch": 0.15330188679245282,
+      "grad_norm": 0.30560099618040315,
+      "learning_rate": 2.5294117647058825e-05,
+      "loss": 0.5962,
+      "step": 130
+    },
+    {
+      "epoch": 0.1544811320754717,
+      "grad_norm": 0.2837967466311176,
+      "learning_rate": 2.5490196078431373e-05,
+      "loss": 0.5609,
+      "step": 131
+    },
+    {
+      "epoch": 0.15566037735849056,
+      "grad_norm": 0.37600219141297575,
+      "learning_rate": 2.568627450980392e-05,
+      "loss": 0.6103,
+      "step": 132
+    },
+    {
+      "epoch": 0.15683962264150944,
+      "grad_norm": 0.3544398089194838,
+      "learning_rate": 2.5882352941176475e-05,
+      "loss": 0.6287,
+      "step": 133
+    },
+    {
+      "epoch": 0.1580188679245283,
+      "grad_norm": 0.36498153064878025,
+      "learning_rate": 2.6078431372549023e-05,
+      "loss": 0.5946,
+      "step": 134
+    },
+    {
+      "epoch": 0.15919811320754718,
+      "grad_norm": 0.3046952533623325,
+      "learning_rate": 2.627450980392157e-05,
+      "loss": 0.6204,
+      "step": 135
+    },
+    {
+      "epoch": 0.16037735849056603,
+      "grad_norm": 0.32958231690461604,
+      "learning_rate": 2.647058823529412e-05,
+      "loss": 0.6149,
+      "step": 136
+    },
+    {
+      "epoch": 0.1615566037735849,
+      "grad_norm": 0.25386434332121066,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5433,
+      "step": 137
+    },
+    {
+      "epoch": 0.16273584905660377,
+      "grad_norm": 0.2866152511804096,
+      "learning_rate": 2.6862745098039214e-05,
+      "loss": 0.563,
+      "step": 138
+    },
+    {
+      "epoch": 0.16391509433962265,
+      "grad_norm": 0.27604666660051896,
+      "learning_rate": 2.7058823529411766e-05,
+      "loss": 0.5827,
+      "step": 139
+    },
+    {
+      "epoch": 0.1650943396226415,
+      "grad_norm": 0.2790668508989892,
+      "learning_rate": 2.7254901960784314e-05,
+      "loss": 0.5939,
+      "step": 140
+    },
+    {
+      "epoch": 0.16627358490566038,
+      "grad_norm": 0.2728709069404598,
+      "learning_rate": 2.7450980392156865e-05,
+      "loss": 0.5762,
+      "step": 141
+    },
+    {
+      "epoch": 0.16745283018867924,
+      "grad_norm": 0.29382080188529913,
+      "learning_rate": 2.7647058823529416e-05,
+      "loss": 0.5768,
+      "step": 142
+    },
+    {
+      "epoch": 0.16863207547169812,
+      "grad_norm": 0.3102492795213314,
+      "learning_rate": 2.7843137254901964e-05,
+      "loss": 0.6011,
+      "step": 143
+    },
+    {
+      "epoch": 0.16981132075471697,
+      "grad_norm": 0.2957406064461956,
+      "learning_rate": 2.8039215686274512e-05,
+      "loss": 0.6058,
+      "step": 144
+    },
+    {
+      "epoch": 0.17099056603773585,
+      "grad_norm": 0.2687749256242989,
+      "learning_rate": 2.823529411764706e-05,
+      "loss": 0.5862,
+      "step": 145
+    },
+    {
+      "epoch": 0.1721698113207547,
+      "grad_norm": 0.308695360819304,
+      "learning_rate": 2.8431372549019608e-05,
+      "loss": 0.569,
+      "step": 146
+    },
+    {
+      "epoch": 0.1733490566037736,
+      "grad_norm": 0.2523615267040781,
+      "learning_rate": 2.8627450980392155e-05,
+      "loss": 0.6047,
+      "step": 147
+    },
+    {
+      "epoch": 0.17452830188679244,
+      "grad_norm": 0.34775414736242244,
+      "learning_rate": 2.8823529411764703e-05,
+      "loss": 0.5545,
+      "step": 148
+    },
+    {
+      "epoch": 0.17570754716981132,
+      "grad_norm": 0.2868050434462969,
+      "learning_rate": 2.9019607843137258e-05,
+      "loss": 0.5899,
+      "step": 149
+    },
+    {
+      "epoch": 0.17688679245283018,
+      "grad_norm": 0.31123474070251195,
+      "learning_rate": 2.9215686274509806e-05,
+      "loss": 0.5489,
+      "step": 150
+    },
+    {
+      "epoch": 0.17806603773584906,
+      "grad_norm": 0.3466380178616281,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 0.6043,
+      "step": 151
+    },
+    {
+      "epoch": 0.1792452830188679,
+      "grad_norm": 0.27187796878322257,
+      "learning_rate": 2.9607843137254905e-05,
+      "loss": 0.592,
+      "step": 152
+    },
+    {
+      "epoch": 0.1804245283018868,
+      "grad_norm": 0.36094542423571174,
+      "learning_rate": 2.9803921568627453e-05,
+      "loss": 0.613,
+      "step": 153
+    },
+    {
+      "epoch": 0.18160377358490565,
+      "grad_norm": 0.3639564752753993,
+      "learning_rate": 3e-05,
+      "loss": 0.5742,
+      "step": 154
+    },
+    {
+      "epoch": 0.18278301886792453,
+      "grad_norm": 0.30040882697993804,
+      "learning_rate": 3.019607843137255e-05,
+      "loss": 0.5698,
+      "step": 155
+    },
+    {
+      "epoch": 0.18396226415094338,
+      "grad_norm": 0.3442904950422494,
+      "learning_rate": 3.0392156862745097e-05,
+      "loss": 0.592,
+      "step": 156
+    },
+    {
+      "epoch": 0.18514150943396226,
+      "grad_norm": 0.36073121530929464,
+      "learning_rate": 3.058823529411765e-05,
+      "loss": 0.5908,
+      "step": 157
+    },
+    {
+      "epoch": 0.18632075471698112,
+      "grad_norm": 0.29757834692618373,
+      "learning_rate": 3.07843137254902e-05,
+      "loss": 0.5878,
+      "step": 158
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 0.3151369774932867,
+      "learning_rate": 3.098039215686275e-05,
+      "loss": 0.5929,
+      "step": 159
+    },
+    {
+      "epoch": 0.18867924528301888,
+      "grad_norm": 0.3261517677524137,
+      "learning_rate": 3.11764705882353e-05,
+      "loss": 0.5722,
+      "step": 160
+    },
+    {
+      "epoch": 0.18985849056603774,
+      "grad_norm": 0.353366772889164,
+      "learning_rate": 3.137254901960784e-05,
+      "loss": 0.6156,
+      "step": 161
+    },
+    {
+      "epoch": 0.19103773584905662,
+      "grad_norm": 0.30622152411935233,
+      "learning_rate": 3.1568627450980394e-05,
+      "loss": 0.5793,
+      "step": 162
+    },
+    {
+      "epoch": 0.19221698113207547,
+      "grad_norm": 0.36698536794178416,
+      "learning_rate": 3.176470588235294e-05,
+      "loss": 0.6217,
+      "step": 163
+    },
+    {
+      "epoch": 0.19339622641509435,
+      "grad_norm": 0.3362124783732646,
+      "learning_rate": 3.196078431372549e-05,
+      "loss": 0.6106,
+      "step": 164
+    },
+    {
+      "epoch": 0.1945754716981132,
+      "grad_norm": 0.2899637305406634,
+      "learning_rate": 3.215686274509804e-05,
+      "loss": 0.5849,
+      "step": 165
+    },
+    {
+      "epoch": 0.1957547169811321,
+      "grad_norm": 0.3206144307518075,
+      "learning_rate": 3.235294117647059e-05,
+      "loss": 0.5465,
+      "step": 166
+    },
+    {
+      "epoch": 0.19693396226415094,
+      "grad_norm": 0.31006103974168564,
+      "learning_rate": 3.254901960784314e-05,
+      "loss": 0.5548,
+      "step": 167
+    },
+    {
+      "epoch": 0.19811320754716982,
+      "grad_norm": 0.3321586928747005,
+      "learning_rate": 3.274509803921569e-05,
+      "loss": 0.631,
+      "step": 168
+    },
+    {
+      "epoch": 0.19929245283018868,
+      "grad_norm": 0.4095688833748799,
+      "learning_rate": 3.294117647058824e-05,
+      "loss": 0.6026,
+      "step": 169
+    },
+    {
+      "epoch": 0.20047169811320756,
+      "grad_norm": 0.34395514890471596,
+      "learning_rate": 3.3137254901960784e-05,
+      "loss": 0.5567,
+      "step": 170
+    },
+    {
+      "epoch": 0.2016509433962264,
+      "grad_norm": 0.35205325247024194,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.5852,
+      "step": 171
+    },
+    {
+      "epoch": 0.2028301886792453,
+      "grad_norm": 0.412211694125209,
+      "learning_rate": 3.352941176470588e-05,
+      "loss": 0.5815,
+      "step": 172
+    },
+    {
+      "epoch": 0.20400943396226415,
+      "grad_norm": 0.3699137359171901,
+      "learning_rate": 3.372549019607844e-05,
+      "loss": 0.5598,
+      "step": 173
+    },
+    {
+      "epoch": 0.20518867924528303,
+      "grad_norm": 0.3300187726439282,
+      "learning_rate": 3.392156862745098e-05,
+      "loss": 0.5813,
+      "step": 174
+    },
+    {
+      "epoch": 0.20636792452830188,
+      "grad_norm": 0.400114788883625,
+      "learning_rate": 3.411764705882353e-05,
+      "loss": 0.5938,
+      "step": 175
+    },
+    {
+      "epoch": 0.20754716981132076,
+      "grad_norm": 0.3793812643176407,
+      "learning_rate": 3.431372549019608e-05,
+      "loss": 0.5862,
+      "step": 176
+    },
+    {
+      "epoch": 0.20872641509433962,
+      "grad_norm": 0.2861341090960564,
+      "learning_rate": 3.450980392156863e-05,
+      "loss": 0.6215,
+      "step": 177
+    },
+    {
+      "epoch": 0.2099056603773585,
+      "grad_norm": 0.39026344700494076,
+      "learning_rate": 3.470588235294118e-05,
+      "loss": 0.5822,
+      "step": 178
+    },
+    {
+      "epoch": 0.21108490566037735,
+      "grad_norm": 0.347515557536463,
+      "learning_rate": 3.4901960784313725e-05,
+      "loss": 0.5564,
+      "step": 179
+    },
+    {
+      "epoch": 0.21226415094339623,
+      "grad_norm": 0.31357484840960886,
+      "learning_rate": 3.5098039215686276e-05,
+      "loss": 0.5696,
+      "step": 180
+    },
+    {
+      "epoch": 0.2134433962264151,
+      "grad_norm": 0.4580783510528136,
+      "learning_rate": 3.529411764705883e-05,
+      "loss": 0.5634,
+      "step": 181
+    },
+    {
+      "epoch": 0.21462264150943397,
+      "grad_norm": 0.4656582529694432,
+      "learning_rate": 3.549019607843138e-05,
+      "loss": 0.5961,
+      "step": 182
+    },
+    {
+      "epoch": 0.21580188679245282,
+      "grad_norm": 0.3378464627638342,
+      "learning_rate": 3.568627450980392e-05,
+      "loss": 0.5957,
+      "step": 183
+    },
+    {
+      "epoch": 0.2169811320754717,
+      "grad_norm": 0.5990471547688463,
+      "learning_rate": 3.5882352941176474e-05,
+      "loss": 0.5857,
+      "step": 184
+    },
+    {
+      "epoch": 0.21816037735849056,
+      "grad_norm": 0.40540234994556373,
+      "learning_rate": 3.607843137254902e-05,
+      "loss": 0.5512,
+      "step": 185
+    },
+    {
+      "epoch": 0.21933962264150944,
+      "grad_norm": 0.3990194678195888,
+      "learning_rate": 3.627450980392157e-05,
+      "loss": 0.5796,
+      "step": 186
+    },
+    {
+      "epoch": 0.2205188679245283,
+      "grad_norm": 0.5261368800410511,
+      "learning_rate": 3.6470588235294114e-05,
+      "loss": 0.5702,
+      "step": 187
+    },
+    {
+      "epoch": 0.22169811320754718,
+      "grad_norm": 0.3256012589890071,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.5616,
+      "step": 188
+    },
+    {
+      "epoch": 0.22287735849056603,
+      "grad_norm": 0.38501924324156434,
+      "learning_rate": 3.686274509803922e-05,
+      "loss": 0.5842,
+      "step": 189
+    },
+    {
+      "epoch": 0.2240566037735849,
+      "grad_norm": 0.3979730977289346,
+      "learning_rate": 3.705882352941177e-05,
+      "loss": 0.586,
+      "step": 190
+    },
+    {
+      "epoch": 0.22523584905660377,
+      "grad_norm": 0.32924423792747387,
+      "learning_rate": 3.725490196078432e-05,
+      "loss": 0.583,
+      "step": 191
+    },
+    {
+      "epoch": 0.22641509433962265,
+      "grad_norm": 0.4298626945547182,
+      "learning_rate": 3.7450980392156864e-05,
+      "loss": 0.5939,
+      "step": 192
+    },
+    {
+      "epoch": 0.2275943396226415,
+      "grad_norm": 0.39678738123541807,
+      "learning_rate": 3.7647058823529415e-05,
+      "loss": 0.5855,
+      "step": 193
+    },
+    {
+      "epoch": 0.22877358490566038,
+      "grad_norm": 0.3785530288617002,
+      "learning_rate": 3.784313725490196e-05,
+      "loss": 0.5429,
+      "step": 194
+    },
+    {
+      "epoch": 0.22995283018867924,
+      "grad_norm": 0.32502237655613264,
+      "learning_rate": 3.803921568627451e-05,
+      "loss": 0.5609,
+      "step": 195
+    },
+    {
+      "epoch": 0.23113207547169812,
+      "grad_norm": 0.28408226345867754,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.5642,
+      "step": 196
+    },
+    {
+      "epoch": 0.23231132075471697,
+      "grad_norm": 0.37233944481515463,
+      "learning_rate": 3.8431372549019614e-05,
+      "loss": 0.5897,
+      "step": 197
+    },
+    {
+      "epoch": 0.23349056603773585,
+      "grad_norm": 0.33169975034089944,
+      "learning_rate": 3.862745098039216e-05,
+      "loss": 0.6019,
+      "step": 198
+    },
+    {
+      "epoch": 0.2346698113207547,
+      "grad_norm": 0.33922338645706834,
+      "learning_rate": 3.882352941176471e-05,
+      "loss": 0.5625,
+      "step": 199
+    },
+    {
+      "epoch": 0.2358490566037736,
+      "grad_norm": 0.3905110044421479,
+      "learning_rate": 3.9019607843137254e-05,
+      "loss": 0.5885,
+      "step": 200
+    },
+    {
+      "epoch": 0.23702830188679244,
+      "grad_norm": 0.36111979412690576,
+      "learning_rate": 3.9215686274509805e-05,
+      "loss": 0.5763,
+      "step": 201
+    },
+    {
+      "epoch": 0.23820754716981132,
+      "grad_norm": 0.30772624814623173,
+      "learning_rate": 3.9411764705882356e-05,
+      "loss": 0.5566,
+      "step": 202
+    },
+    {
+      "epoch": 0.23938679245283018,
+      "grad_norm": 0.3859532115992298,
+      "learning_rate": 3.96078431372549e-05,
+      "loss": 0.56,
+      "step": 203
+    },
+    {
+      "epoch": 0.24056603773584906,
+      "grad_norm": 0.329603614841811,
+      "learning_rate": 3.980392156862745e-05,
+      "loss": 0.5426,
+      "step": 204
+    },
+    {
+      "epoch": 0.2417452830188679,
+      "grad_norm": 0.35175814757716395,
+      "learning_rate": 4e-05,
+      "loss": 0.5668,
+      "step": 205
+    },
+    {
+      "epoch": 0.2429245283018868,
+      "grad_norm": 0.3787854074702761,
+      "learning_rate": 4.0196078431372555e-05,
+      "loss": 0.5576,
+      "step": 206
+    },
+    {
+      "epoch": 0.24410377358490565,
+      "grad_norm": 0.28749074102993205,
+      "learning_rate": 4.03921568627451e-05,
+      "loss": 0.5452,
+      "step": 207
+    },
+    {
+      "epoch": 0.24528301886792453,
+      "grad_norm": 0.388945319114422,
+      "learning_rate": 4.058823529411765e-05,
+      "loss": 0.5622,
+      "step": 208
+    },
+    {
+      "epoch": 0.24646226415094338,
+      "grad_norm": 0.37735894020060534,
+      "learning_rate": 4.0784313725490195e-05,
+      "loss": 0.5833,
+      "step": 209
+    },
+    {
+      "epoch": 0.24764150943396226,
+      "grad_norm": 0.3660851678615854,
+      "learning_rate": 4.0980392156862746e-05,
+      "loss": 0.5321,
+      "step": 210
+    },
+    {
+      "epoch": 0.24882075471698112,
+      "grad_norm": 0.4129245617113434,
+      "learning_rate": 4.11764705882353e-05,
+      "loss": 0.5924,
+      "step": 211
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.32470818038222043,
+      "learning_rate": 4.137254901960784e-05,
+      "loss": 0.5612,
+      "step": 212
+    },
+    {
+      "epoch": 0.2511792452830189,
+      "grad_norm": 0.32887018750480307,
+      "learning_rate": 4.156862745098039e-05,
+      "loss": 0.5565,
+      "step": 213
+    },
+    {
+      "epoch": 0.25235849056603776,
+      "grad_norm": 0.3685234841927023,
+      "learning_rate": 4.1764705882352944e-05,
+      "loss": 0.5595,
+      "step": 214
+    },
+    {
+      "epoch": 0.2535377358490566,
+      "grad_norm": 0.35604272584025953,
+      "learning_rate": 4.1960784313725496e-05,
+      "loss": 0.574,
+      "step": 215
+    },
+    {
+      "epoch": 0.25471698113207547,
+      "grad_norm": 0.3456679645268957,
+      "learning_rate": 4.215686274509804e-05,
+      "loss": 0.5625,
+      "step": 216
+    },
+    {
+      "epoch": 0.25589622641509435,
+      "grad_norm": 0.39913031554345235,
+      "learning_rate": 4.235294117647059e-05,
+      "loss": 0.5537,
+      "step": 217
+    },
+    {
+      "epoch": 0.25707547169811323,
+      "grad_norm": 0.30857870932592635,
+      "learning_rate": 4.2549019607843136e-05,
+      "loss": 0.5593,
+      "step": 218
+    },
+    {
+      "epoch": 0.25825471698113206,
+      "grad_norm": 0.38815719928350045,
+      "learning_rate": 4.274509803921569e-05,
+      "loss": 0.5501,
+      "step": 219
+    },
+    {
+      "epoch": 0.25943396226415094,
+      "grad_norm": 0.33979438830739794,
+      "learning_rate": 4.294117647058823e-05,
+      "loss": 0.5745,
+      "step": 220
+    },
+    {
+      "epoch": 0.2606132075471698,
+      "grad_norm": 0.33316871288159844,
+      "learning_rate": 4.313725490196079e-05,
+      "loss": 0.5578,
+      "step": 221
+    },
+    {
+      "epoch": 0.2617924528301887,
+      "grad_norm": 0.35237535229859496,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.5503,
+      "step": 222
+    },
+    {
+      "epoch": 0.26297169811320753,
+      "grad_norm": 0.3005848521454557,
+      "learning_rate": 4.3529411764705885e-05,
+      "loss": 0.5307,
+      "step": 223
+    },
+    {
+      "epoch": 0.2641509433962264,
+      "grad_norm": 0.34906842320131487,
+      "learning_rate": 4.3725490196078437e-05,
+      "loss": 0.5656,
+      "step": 224
+    },
+    {
+      "epoch": 0.2653301886792453,
+      "grad_norm": 0.37065596862544675,
+      "learning_rate": 4.392156862745098e-05,
+      "loss": 0.5659,
+      "step": 225
+    },
+    {
+      "epoch": 0.2665094339622642,
+      "grad_norm": 0.3113922594915071,
+      "learning_rate": 4.411764705882353e-05,
+      "loss": 0.5513,
+      "step": 226
+    },
+    {
+      "epoch": 0.267688679245283,
+      "grad_norm": 0.349075826782398,
+      "learning_rate": 4.431372549019608e-05,
+      "loss": 0.5386,
+      "step": 227
+    },
+    {
+      "epoch": 0.2688679245283019,
+      "grad_norm": 0.2975531506900559,
+      "learning_rate": 4.450980392156863e-05,
+      "loss": 0.5989,
+      "step": 228
+    },
+    {
+      "epoch": 0.27004716981132076,
+      "grad_norm": 0.3310975264171541,
+      "learning_rate": 4.470588235294118e-05,
+      "loss": 0.5393,
+      "step": 229
+    },
+    {
+      "epoch": 0.27122641509433965,
+      "grad_norm": 0.35331340755683055,
+      "learning_rate": 4.490196078431373e-05,
+      "loss": 0.5724,
+      "step": 230
+    },
+    {
+      "epoch": 0.27240566037735847,
+      "grad_norm": 0.3435667607323172,
+      "learning_rate": 4.5098039215686275e-05,
+      "loss": 0.5441,
+      "step": 231
+    },
+    {
+      "epoch": 0.27358490566037735,
+      "grad_norm": 0.3406102906798467,
+      "learning_rate": 4.5294117647058826e-05,
+      "loss": 0.5737,
+      "step": 232
+    },
+    {
+      "epoch": 0.27476415094339623,
+      "grad_norm": 0.3592255641400085,
+      "learning_rate": 4.549019607843137e-05,
+      "loss": 0.5349,
+      "step": 233
+    },
+    {
+      "epoch": 0.2759433962264151,
+      "grad_norm": 0.30277787333312645,
+      "learning_rate": 4.568627450980392e-05,
+      "loss": 0.538,
+      "step": 234
+    },
+    {
+      "epoch": 0.27712264150943394,
+      "grad_norm": 0.34883499039655014,
+      "learning_rate": 4.588235294117647e-05,
+      "loss": 0.5359,
+      "step": 235
+    },
+    {
+      "epoch": 0.2783018867924528,
+      "grad_norm": 0.3151091543966709,
+      "learning_rate": 4.607843137254902e-05,
+      "loss": 0.5245,
+      "step": 236
+    },
+    {
+      "epoch": 0.2794811320754717,
+      "grad_norm": 0.35907455969622387,
+      "learning_rate": 4.6274509803921576e-05,
+      "loss": 0.6,
+      "step": 237
+    },
+    {
+      "epoch": 0.2806603773584906,
+      "grad_norm": 0.3032818041603703,
+      "learning_rate": 4.647058823529412e-05,
+      "loss": 0.552,
+      "step": 238
+    },
+    {
+      "epoch": 0.2818396226415094,
+      "grad_norm": 0.35282824658487727,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.5418,
+      "step": 239
+    },
+    {
+      "epoch": 0.2830188679245283,
+      "grad_norm": 0.2920832412484557,
+      "learning_rate": 4.6862745098039216e-05,
+      "loss": 0.5614,
+      "step": 240
+    },
+    {
+      "epoch": 0.2841981132075472,
+      "grad_norm": 0.3435992121563552,
+      "learning_rate": 4.705882352941177e-05,
+      "loss": 0.5494,
+      "step": 241
+    },
+    {
+      "epoch": 0.28537735849056606,
+      "grad_norm": 0.32047998142918005,
+      "learning_rate": 4.725490196078431e-05,
+      "loss": 0.5452,
+      "step": 242
+    },
+    {
+      "epoch": 0.2865566037735849,
+      "grad_norm": 0.4019038012750058,
+      "learning_rate": 4.745098039215686e-05,
+      "loss": 0.5525,
+      "step": 243
+    },
+    {
+      "epoch": 0.28773584905660377,
+      "grad_norm": 0.3432644374575133,
+      "learning_rate": 4.7647058823529414e-05,
+      "loss": 0.545,
+      "step": 244
+    },
+    {
+      "epoch": 0.28891509433962265,
+      "grad_norm": 0.5285900491128185,
+      "learning_rate": 4.7843137254901966e-05,
+      "loss": 0.5503,
+      "step": 245
+    },
+    {
+      "epoch": 0.29009433962264153,
+      "grad_norm": 0.29134542644610206,
+      "learning_rate": 4.803921568627452e-05,
+      "loss": 0.5401,
+      "step": 246
+    },
+    {
+      "epoch": 0.29127358490566035,
+      "grad_norm": 0.4635148372260288,
+      "learning_rate": 4.823529411764706e-05,
+      "loss": 0.5803,
+      "step": 247
+    },
+    {
+      "epoch": 0.29245283018867924,
+      "grad_norm": 0.3030524544715968,
+      "learning_rate": 4.843137254901961e-05,
+      "loss": 0.5624,
+      "step": 248
+    },
+    {
+      "epoch": 0.2936320754716981,
+      "grad_norm": 0.40275184473600484,
+      "learning_rate": 4.862745098039216e-05,
+      "loss": 0.5713,
+      "step": 249
+    },
+    {
+      "epoch": 0.294811320754717,
+      "grad_norm": 0.3118951554360199,
+      "learning_rate": 4.882352941176471e-05,
+      "loss": 0.5522,
+      "step": 250
+    },
+    {
+      "epoch": 0.2959905660377358,
+      "grad_norm": 0.32579582500388415,
+      "learning_rate": 4.901960784313725e-05,
+      "loss": 0.5246,
+      "step": 251
+    },
+    {
+      "epoch": 0.2971698113207547,
+      "grad_norm": 0.3033155103292917,
+      "learning_rate": 4.9215686274509804e-05,
+      "loss": 0.5633,
+      "step": 252
+    },
+    {
+      "epoch": 0.2983490566037736,
+      "grad_norm": 0.3049336228203138,
+      "learning_rate": 4.9411764705882355e-05,
+      "loss": 0.5243,
+      "step": 253
+    },
+    {
+      "epoch": 0.29952830188679247,
+      "grad_norm": 0.36373919092208606,
+      "learning_rate": 4.960784313725491e-05,
+      "loss": 0.5417,
+      "step": 254
+    },
+    {
+      "epoch": 0.3007075471698113,
+      "grad_norm": 0.37167866504897323,
+      "learning_rate": 4.980392156862745e-05,
+      "loss": 0.5456,
+      "step": 255
+    },
+    {
+      "epoch": 0.3018867924528302,
+      "grad_norm": 0.3105010766366475,
+      "learning_rate": 5e-05,
+      "loss": 0.5794,
+      "step": 256
+    },
+    {
+      "epoch": 0.30306603773584906,
+      "grad_norm": 0.4244351615233353,
+      "learning_rate": 4.9978156400174754e-05,
+      "loss": 0.5642,
+      "step": 257
+    },
+    {
+      "epoch": 0.30424528301886794,
+      "grad_norm": 0.36490127605628797,
+      "learning_rate": 4.99563128003495e-05,
+      "loss": 0.5337,
+      "step": 258
+    },
+    {
+      "epoch": 0.30542452830188677,
+      "grad_norm": 0.40736314132332574,
+      "learning_rate": 4.9934469200524245e-05,
+      "loss": 0.5562,
+      "step": 259
+    },
+    {
+      "epoch": 0.30660377358490565,
+      "grad_norm": 0.36578406941621544,
+      "learning_rate": 4.9912625600699e-05,
+      "loss": 0.5511,
+      "step": 260
+    },
+    {
+      "epoch": 0.30778301886792453,
+      "grad_norm": 0.4440539081028125,
+      "learning_rate": 4.989078200087375e-05,
+      "loss": 0.551,
+      "step": 261
+    },
+    {
+      "epoch": 0.3089622641509434,
+      "grad_norm": 0.3520247037802827,
+      "learning_rate": 4.9868938401048495e-05,
+      "loss": 0.556,
+      "step": 262
+    },
+    {
+      "epoch": 0.31014150943396224,
+      "grad_norm": 0.4235808519845711,
+      "learning_rate": 4.984709480122324e-05,
+      "loss": 0.5554,
+      "step": 263
+    },
+    {
+      "epoch": 0.3113207547169811,
+      "grad_norm": 0.4864485623060006,
+      "learning_rate": 4.982525120139799e-05,
+      "loss": 0.5436,
+      "step": 264
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.3415754461802786,
+      "learning_rate": 4.9803407601572744e-05,
+      "loss": 0.5448,
+      "step": 265
+    },
+    {
+      "epoch": 0.3136792452830189,
+      "grad_norm": 0.4111870336007634,
+      "learning_rate": 4.978156400174749e-05,
+      "loss": 0.5763,
+      "step": 266
+    },
+    {
+      "epoch": 0.31485849056603776,
+      "grad_norm": 0.396879220907095,
+      "learning_rate": 4.975972040192224e-05,
+      "loss": 0.5716,
+      "step": 267
+    },
+    {
+      "epoch": 0.3160377358490566,
+      "grad_norm": 0.3635812786478751,
+      "learning_rate": 4.973787680209699e-05,
+      "loss": 0.5107,
+      "step": 268
+    },
+    {
+      "epoch": 0.31721698113207547,
+      "grad_norm": 0.4253526896428924,
+      "learning_rate": 4.971603320227174e-05,
+      "loss": 0.5746,
+      "step": 269
+    },
+    {
+      "epoch": 0.31839622641509435,
+      "grad_norm": 0.37107715043693323,
+      "learning_rate": 4.9694189602446484e-05,
+      "loss": 0.5482,
+      "step": 270
+    },
+    {
+      "epoch": 0.31957547169811323,
+      "grad_norm": 0.3284887985916158,
+      "learning_rate": 4.9672346002621236e-05,
+      "loss": 0.5252,
+      "step": 271
+    },
+    {
+      "epoch": 0.32075471698113206,
+      "grad_norm": 0.32193873352000796,
+      "learning_rate": 4.965050240279598e-05,
+      "loss": 0.5805,
+      "step": 272
+    },
+    {
+      "epoch": 0.32193396226415094,
+      "grad_norm": 0.4082422463910078,
+      "learning_rate": 4.9628658802970734e-05,
+      "loss": 0.5676,
+      "step": 273
+    },
+    {
+      "epoch": 0.3231132075471698,
+      "grad_norm": 0.28548955444833474,
+      "learning_rate": 4.960681520314548e-05,
+      "loss": 0.5274,
+      "step": 274
+    },
+    {
+      "epoch": 0.3242924528301887,
+      "grad_norm": 0.3748130745365143,
+      "learning_rate": 4.958497160332023e-05,
+      "loss": 0.539,
+      "step": 275
+    },
+    {
+      "epoch": 0.32547169811320753,
+      "grad_norm": 0.2979631195497302,
+      "learning_rate": 4.9563128003494976e-05,
+      "loss": 0.5712,
+      "step": 276
+    },
+    {
+      "epoch": 0.3266509433962264,
+      "grad_norm": 0.315109215120387,
+      "learning_rate": 4.954128440366973e-05,
+      "loss": 0.5696,
+      "step": 277
+    },
+    {
+      "epoch": 0.3278301886792453,
+      "grad_norm": 0.32655980609053575,
+      "learning_rate": 4.9519440803844474e-05,
+      "loss": 0.5462,
+      "step": 278
+    },
+    {
+      "epoch": 0.3290094339622642,
+      "grad_norm": 0.344665006335857,
+      "learning_rate": 4.9497597204019226e-05,
+      "loss": 0.5511,
+      "step": 279
+    },
+    {
+      "epoch": 0.330188679245283,
+      "grad_norm": 0.3286374689187085,
+      "learning_rate": 4.947575360419398e-05,
+      "loss": 0.5613,
+      "step": 280
+    },
+    {
+      "epoch": 0.3313679245283019,
+      "grad_norm": 0.33729231557462946,
+      "learning_rate": 4.945391000436872e-05,
+      "loss": 0.532,
+      "step": 281
+    },
+    {
+      "epoch": 0.33254716981132076,
+      "grad_norm": 0.32373727577838135,
+      "learning_rate": 4.943206640454347e-05,
+      "loss": 0.5907,
+      "step": 282
+    },
+    {
+      "epoch": 0.33372641509433965,
+      "grad_norm": 0.33482282831438476,
+      "learning_rate": 4.941022280471822e-05,
+      "loss": 0.5546,
+      "step": 283
+    },
+    {
+      "epoch": 0.33490566037735847,
+      "grad_norm": 0.3202335748431458,
+      "learning_rate": 4.938837920489297e-05,
+      "loss": 0.5735,
+      "step": 284
+    },
+    {
+      "epoch": 0.33608490566037735,
+      "grad_norm": 0.3465318516125321,
+      "learning_rate": 4.936653560506772e-05,
+      "loss": 0.5469,
+      "step": 285
+    },
+    {
+      "epoch": 0.33726415094339623,
+      "grad_norm": 0.3660181563789553,
+      "learning_rate": 4.934469200524246e-05,
+      "loss": 0.5615,
+      "step": 286
+    },
+    {
+      "epoch": 0.3384433962264151,
+      "grad_norm": 0.30831007587160425,
+      "learning_rate": 4.9322848405417215e-05,
+      "loss": 0.5324,
+      "step": 287
+    },
+    {
+      "epoch": 0.33962264150943394,
+      "grad_norm": 0.3291048653447946,
+      "learning_rate": 4.930100480559197e-05,
+      "loss": 0.5731,
+      "step": 288
+    },
+    {
+      "epoch": 0.3408018867924528,
+      "grad_norm": 0.3366115740131569,
+      "learning_rate": 4.927916120576671e-05,
+      "loss": 0.5438,
+      "step": 289
+    },
+    {
+      "epoch": 0.3419811320754717,
+      "grad_norm": 0.33005113056155644,
+      "learning_rate": 4.925731760594146e-05,
+      "loss": 0.5263,
+      "step": 290
+    },
+    {
+      "epoch": 0.3431603773584906,
+      "grad_norm": 0.30587785430407993,
+      "learning_rate": 4.923547400611621e-05,
+      "loss": 0.5279,
+      "step": 291
+    },
+    {
+      "epoch": 0.3443396226415094,
+      "grad_norm": 0.2707802297809873,
+      "learning_rate": 4.921363040629096e-05,
+      "loss": 0.5431,
+      "step": 292
+    },
+    {
+      "epoch": 0.3455188679245283,
+      "grad_norm": 0.32462843717912787,
+      "learning_rate": 4.919178680646571e-05,
+      "loss": 0.5745,
+      "step": 293
+    },
+    {
+      "epoch": 0.3466981132075472,
+      "grad_norm": 0.33720149340702477,
+      "learning_rate": 4.916994320664046e-05,
+      "loss": 0.5542,
+      "step": 294
+    },
+    {
+      "epoch": 0.34787735849056606,
+      "grad_norm": 0.3012561881506545,
+      "learning_rate": 4.9148099606815205e-05,
+      "loss": 0.5491,
+      "step": 295
+    },
+    {
+      "epoch": 0.3490566037735849,
+      "grad_norm": 0.35056440877362316,
+      "learning_rate": 4.912625600698996e-05,
+      "loss": 0.5334,
+      "step": 296
+    },
+    {
+      "epoch": 0.35023584905660377,
+      "grad_norm": 0.27696203340371517,
+      "learning_rate": 4.91044124071647e-05,
+      "loss": 0.5198,
+      "step": 297
+    },
+    {
+      "epoch": 0.35141509433962265,
+      "grad_norm": 0.3093704886211973,
+      "learning_rate": 4.9082568807339454e-05,
+      "loss": 0.5149,
+      "step": 298
+    },
+    {
+      "epoch": 0.35259433962264153,
+      "grad_norm": 0.2715029004960904,
+      "learning_rate": 4.90607252075142e-05,
+      "loss": 0.5351,
+      "step": 299
+    },
+    {
+      "epoch": 0.35377358490566035,
+      "grad_norm": 0.36387837113759924,
+      "learning_rate": 4.9038881607688945e-05,
+      "loss": 0.5219,
+      "step": 300
+    },
+    {
+      "epoch": 0.35495283018867924,
+      "grad_norm": 0.2937952313277803,
+      "learning_rate": 4.90170380078637e-05,
+      "loss": 0.5621,
+      "step": 301
+    },
+    {
+      "epoch": 0.3561320754716981,
+      "grad_norm": 0.3333901002208202,
+      "learning_rate": 4.899519440803845e-05,
+      "loss": 0.5219,
+      "step": 302
+    },
+    {
+      "epoch": 0.357311320754717,
+      "grad_norm": 0.3235852638636914,
+      "learning_rate": 4.8973350808213194e-05,
+      "loss": 0.55,
+      "step": 303
+    },
+    {
+      "epoch": 0.3584905660377358,
+      "grad_norm": 0.2980940685533884,
+      "learning_rate": 4.895150720838794e-05,
+      "loss": 0.5548,
+      "step": 304
+    },
+    {
+      "epoch": 0.3596698113207547,
+      "grad_norm": 0.3020264858126994,
+      "learning_rate": 4.892966360856269e-05,
+      "loss": 0.5321,
+      "step": 305
+    },
+    {
+      "epoch": 0.3608490566037736,
+      "grad_norm": 0.2809772696071854,
+      "learning_rate": 4.8907820008737444e-05,
+      "loss": 0.5186,
+      "step": 306
+    },
+    {
+      "epoch": 0.36202830188679247,
+      "grad_norm": 0.34827633120614243,
+      "learning_rate": 4.8885976408912196e-05,
+      "loss": 0.5293,
+      "step": 307
+    },
+    {
+      "epoch": 0.3632075471698113,
+      "grad_norm": 0.2915547520920253,
+      "learning_rate": 4.8864132809086935e-05,
+      "loss": 0.5205,
+      "step": 308
+    },
+    {
+      "epoch": 0.3643867924528302,
+      "grad_norm": 0.3133013285517287,
+      "learning_rate": 4.884228920926169e-05,
+      "loss": 0.5269,
+      "step": 309
+    },
+    {
+      "epoch": 0.36556603773584906,
+      "grad_norm": 0.271738840462795,
+      "learning_rate": 4.882044560943644e-05,
+      "loss": 0.5518,
+      "step": 310
+    },
+    {
+      "epoch": 0.36674528301886794,
+      "grad_norm": 0.3091272021832932,
+      "learning_rate": 4.879860200961119e-05,
+      "loss": 0.5401,
+      "step": 311
+    },
+    {
+      "epoch": 0.36792452830188677,
+      "grad_norm": 0.2959015905071043,
+      "learning_rate": 4.8776758409785936e-05,
+      "loss": 0.5295,
+      "step": 312
+    },
+    {
+      "epoch": 0.36910377358490565,
+      "grad_norm": 0.2983896113823104,
+      "learning_rate": 4.875491480996068e-05,
+      "loss": 0.5402,
+      "step": 313
+    },
+    {
+      "epoch": 0.37028301886792453,
+      "grad_norm": 0.249892385163274,
+      "learning_rate": 4.8733071210135434e-05,
+      "loss": 0.5486,
+      "step": 314
+    },
+    {
+      "epoch": 0.3714622641509434,
+      "grad_norm": 0.3521508546158798,
+      "learning_rate": 4.8711227610310186e-05,
+      "loss": 0.5577,
+      "step": 315
+    },
+    {
+      "epoch": 0.37264150943396224,
+      "grad_norm": 0.30287493483253297,
+      "learning_rate": 4.868938401048493e-05,
+      "loss": 0.5472,
+      "step": 316
+    },
+    {
+      "epoch": 0.3738207547169811,
+      "grad_norm": 0.31028440009690905,
+      "learning_rate": 4.8667540410659676e-05,
+      "loss": 0.5331,
+      "step": 317
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.31094507066705473,
+      "learning_rate": 4.864569681083443e-05,
+      "loss": 0.5357,
+      "step": 318
+    },
+    {
+      "epoch": 0.3761792452830189,
+      "grad_norm": 0.2849688969381232,
+      "learning_rate": 4.862385321100918e-05,
+      "loss": 0.5417,
+      "step": 319
+    },
+    {
+      "epoch": 0.37735849056603776,
+      "grad_norm": 0.2868612092734886,
+      "learning_rate": 4.8602009611183926e-05,
+      "loss": 0.5266,
+      "step": 320
+    },
+    {
+      "epoch": 0.3785377358490566,
+      "grad_norm": 0.3080940969988439,
+      "learning_rate": 4.858016601135867e-05,
+      "loss": 0.5449,
+      "step": 321
+    },
+    {
+      "epoch": 0.37971698113207547,
+      "grad_norm": 0.32493931387502994,
+      "learning_rate": 4.855832241153342e-05,
+      "loss": 0.5389,
+      "step": 322
+    },
+    {
+      "epoch": 0.38089622641509435,
+      "grad_norm": 0.32796208560872914,
+      "learning_rate": 4.853647881170817e-05,
+      "loss": 0.5321,
+      "step": 323
+    },
+    {
+      "epoch": 0.38207547169811323,
+      "grad_norm": 0.3401241285590959,
+      "learning_rate": 4.851463521188292e-05,
+      "loss": 0.5331,
+      "step": 324
+    },
+    {
+      "epoch": 0.38325471698113206,
+      "grad_norm": 0.32022208375449956,
+      "learning_rate": 4.849279161205767e-05,
+      "loss": 0.5246,
+      "step": 325
+    },
+    {
+      "epoch": 0.38443396226415094,
+      "grad_norm": 0.3700555582660161,
+      "learning_rate": 4.847094801223242e-05,
+      "loss": 0.5562,
+      "step": 326
+    },
+    {
+      "epoch": 0.3856132075471698,
+      "grad_norm": 0.3333300260112938,
+      "learning_rate": 4.844910441240716e-05,
+      "loss": 0.5449,
+      "step": 327
+    },
+    {
+      "epoch": 0.3867924528301887,
+      "grad_norm": 0.338662439886321,
+      "learning_rate": 4.8427260812581915e-05,
+      "loss": 0.5346,
+      "step": 328
+    },
+    {
+      "epoch": 0.38797169811320753,
+      "grad_norm": 0.3019719038624798,
+      "learning_rate": 4.840541721275667e-05,
+      "loss": 0.5267,
+      "step": 329
+    },
+    {
+      "epoch": 0.3891509433962264,
+      "grad_norm": 0.3308744356140444,
+      "learning_rate": 4.838357361293141e-05,
+      "loss": 0.5579,
+      "step": 330
+    },
+    {
+      "epoch": 0.3903301886792453,
+      "grad_norm": 0.30626156864580834,
+      "learning_rate": 4.836173001310616e-05,
+      "loss": 0.5553,
+      "step": 331
+    },
+    {
+      "epoch": 0.3915094339622642,
+      "grad_norm": 0.3089573287454703,
+      "learning_rate": 4.833988641328091e-05,
+      "loss": 0.5457,
+      "step": 332
+    },
+    {
+      "epoch": 0.392688679245283,
+      "grad_norm": 0.3150250719575896,
+      "learning_rate": 4.831804281345566e-05,
+      "loss": 0.5272,
+      "step": 333
+    },
+    {
+      "epoch": 0.3938679245283019,
+      "grad_norm": 0.35683229585382326,
+      "learning_rate": 4.8296199213630414e-05,
+      "loss": 0.5339,
+      "step": 334
+    },
+    {
+      "epoch": 0.39504716981132076,
+      "grad_norm": 0.25994948465813517,
+      "learning_rate": 4.827435561380515e-05,
+      "loss": 0.548,
+      "step": 335
+    },
+    {
+      "epoch": 0.39622641509433965,
+      "grad_norm": 0.32957719448111233,
+      "learning_rate": 4.8252512013979905e-05,
+      "loss": 0.5504,
+      "step": 336
+    },
+    {
+      "epoch": 0.39740566037735847,
+      "grad_norm": 0.33155434466706185,
+      "learning_rate": 4.823066841415466e-05,
+      "loss": 0.5527,
+      "step": 337
+    },
+    {
+      "epoch": 0.39858490566037735,
+      "grad_norm": 0.3227727526168797,
+      "learning_rate": 4.820882481432941e-05,
+      "loss": 0.542,
+      "step": 338
+    },
+    {
+      "epoch": 0.39976415094339623,
+      "grad_norm": 0.25874317817422376,
+      "learning_rate": 4.8186981214504154e-05,
+      "loss": 0.52,
+      "step": 339
+    },
+    {
+      "epoch": 0.4009433962264151,
+      "grad_norm": 0.3230571629413525,
+      "learning_rate": 4.81651376146789e-05,
+      "loss": 0.5436,
+      "step": 340
+    },
+    {
+      "epoch": 0.40212264150943394,
+      "grad_norm": 0.3203215338019937,
+      "learning_rate": 4.814329401485365e-05,
+      "loss": 0.5309,
+      "step": 341
+    },
+    {
+      "epoch": 0.4033018867924528,
+      "grad_norm": 0.3143710910398086,
+      "learning_rate": 4.81214504150284e-05,
+      "loss": 0.5587,
+      "step": 342
+    },
+    {
+      "epoch": 0.4044811320754717,
+      "grad_norm": 0.2662118913546137,
+      "learning_rate": 4.809960681520315e-05,
+      "loss": 0.5122,
+      "step": 343
+    },
+    {
+      "epoch": 0.4056603773584906,
+      "grad_norm": 0.31097668421670677,
+      "learning_rate": 4.8077763215377894e-05,
+      "loss": 0.5434,
+      "step": 344
+    },
+    {
+      "epoch": 0.4068396226415094,
+      "grad_norm": 0.3085568268272095,
+      "learning_rate": 4.8055919615552646e-05,
+      "loss": 0.5223,
+      "step": 345
+    },
+    {
+      "epoch": 0.4080188679245283,
+      "grad_norm": 0.25470244774475226,
+      "learning_rate": 4.803407601572739e-05,
+      "loss": 0.529,
+      "step": 346
+    },
+    {
+      "epoch": 0.4091981132075472,
+      "grad_norm": 0.3225194200011961,
+      "learning_rate": 4.8012232415902144e-05,
+      "loss": 0.5153,
+      "step": 347
+    },
+    {
+      "epoch": 0.41037735849056606,
+      "grad_norm": 0.2819969012589933,
+      "learning_rate": 4.799038881607689e-05,
+      "loss": 0.5481,
+      "step": 348
+    },
+    {
+      "epoch": 0.4115566037735849,
+      "grad_norm": 0.3072032354663631,
+      "learning_rate": 4.796854521625164e-05,
+      "loss": 0.534,
+      "step": 349
+    },
+    {
+      "epoch": 0.41273584905660377,
+      "grad_norm": 0.266329017685968,
+      "learning_rate": 4.7946701616426387e-05,
+      "loss": 0.5215,
+      "step": 350
+    },
+    {
+      "epoch": 0.41391509433962265,
+      "grad_norm": 0.37456231364680326,
+      "learning_rate": 4.792485801660114e-05,
+      "loss": 0.5427,
+      "step": 351
+    },
+    {
+      "epoch": 0.41509433962264153,
+      "grad_norm": 0.28857303821707214,
+      "learning_rate": 4.790301441677589e-05,
+      "loss": 0.5139,
+      "step": 352
+    },
+    {
+      "epoch": 0.41627358490566035,
+      "grad_norm": 0.3107397475609488,
+      "learning_rate": 4.7881170816950636e-05,
+      "loss": 0.5479,
+      "step": 353
+    },
+    {
+      "epoch": 0.41745283018867924,
+      "grad_norm": 0.3343870089925237,
+      "learning_rate": 4.785932721712538e-05,
+      "loss": 0.5207,
+      "step": 354
+    },
+    {
+      "epoch": 0.4186320754716981,
+      "grad_norm": 0.2961022046033557,
+      "learning_rate": 4.7837483617300133e-05,
+      "loss": 0.5358,
+      "step": 355
+    },
+    {
+      "epoch": 0.419811320754717,
+      "grad_norm": 0.38234573002072186,
+      "learning_rate": 4.7815640017474885e-05,
+      "loss": 0.5034,
+      "step": 356
+    },
+    {
+      "epoch": 0.4209905660377358,
+      "grad_norm": 0.3114281284406419,
+      "learning_rate": 4.779379641764963e-05,
+      "loss": 0.5557,
+      "step": 357
+    },
+    {
+      "epoch": 0.4221698113207547,
+      "grad_norm": 0.36771118291222166,
+      "learning_rate": 4.7771952817824376e-05,
+      "loss": 0.5183,
+      "step": 358
+    },
+    {
+      "epoch": 0.4233490566037736,
+      "grad_norm": 0.27927815962581154,
+      "learning_rate": 4.775010921799913e-05,
+      "loss": 0.4995,
+      "step": 359
+    },
+    {
+      "epoch": 0.42452830188679247,
+      "grad_norm": 0.42683614069434733,
+      "learning_rate": 4.772826561817388e-05,
+      "loss": 0.5214,
+      "step": 360
+    },
+    {
+      "epoch": 0.4257075471698113,
+      "grad_norm": 0.2939701135414953,
+      "learning_rate": 4.7706422018348626e-05,
+      "loss": 0.5354,
+      "step": 361
+    },
+    {
+      "epoch": 0.4268867924528302,
+      "grad_norm": 0.39699865299542914,
+      "learning_rate": 4.768457841852337e-05,
+      "loss": 0.5345,
+      "step": 362
+    },
+    {
+      "epoch": 0.42806603773584906,
+      "grad_norm": 0.28094008256503344,
+      "learning_rate": 4.766273481869812e-05,
+      "loss": 0.5686,
+      "step": 363
+    },
+    {
+      "epoch": 0.42924528301886794,
+      "grad_norm": 0.3511378415179953,
+      "learning_rate": 4.7640891218872875e-05,
+      "loss": 0.5093,
+      "step": 364
+    },
+    {
+      "epoch": 0.43042452830188677,
+      "grad_norm": 0.30120576791228043,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.5315,
+      "step": 365
+    },
+    {
+      "epoch": 0.43160377358490565,
+      "grad_norm": 0.3978515527462058,
+      "learning_rate": 4.7597204019222366e-05,
+      "loss": 0.5414,
+      "step": 366
+    },
+    {
+      "epoch": 0.43278301886792453,
+      "grad_norm": 0.2961521316723327,
+      "learning_rate": 4.757536041939712e-05,
+      "loss": 0.5284,
+      "step": 367
+    },
+    {
+      "epoch": 0.4339622641509434,
+      "grad_norm": 0.4632665516816651,
+      "learning_rate": 4.755351681957187e-05,
+      "loss": 0.531,
+      "step": 368
+    },
+    {
+      "epoch": 0.43514150943396224,
+      "grad_norm": 0.3229591738801333,
+      "learning_rate": 4.7531673219746615e-05,
+      "loss": 0.5363,
+      "step": 369
+    },
+    {
+      "epoch": 0.4363207547169811,
+      "grad_norm": 0.4368759792244959,
+      "learning_rate": 4.750982961992137e-05,
+      "loss": 0.5192,
+      "step": 370
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 0.33102476059298974,
+      "learning_rate": 4.748798602009611e-05,
+      "loss": 0.5176,
+      "step": 371
+    },
+    {
+      "epoch": 0.4386792452830189,
+      "grad_norm": 0.42289814893398486,
+      "learning_rate": 4.7466142420270865e-05,
+      "loss": 0.5566,
+      "step": 372
+    },
+    {
+      "epoch": 0.43985849056603776,
+      "grad_norm": 0.33949456559936203,
+      "learning_rate": 4.744429882044561e-05,
+      "loss": 0.5366,
+      "step": 373
+    },
+    {
+      "epoch": 0.4410377358490566,
+      "grad_norm": 0.41657881320226897,
+      "learning_rate": 4.742245522062036e-05,
+      "loss": 0.5408,
+      "step": 374
+    },
+    {
+      "epoch": 0.44221698113207547,
+      "grad_norm": 0.28559105912959254,
+      "learning_rate": 4.740061162079511e-05,
+      "loss": 0.5436,
+      "step": 375
+    },
+    {
+      "epoch": 0.44339622641509435,
+      "grad_norm": 0.36612709752424666,
+      "learning_rate": 4.737876802096986e-05,
+      "loss": 0.4976,
+      "step": 376
+    },
+    {
+      "epoch": 0.44457547169811323,
+      "grad_norm": 0.26695280840025026,
+      "learning_rate": 4.7356924421144605e-05,
+      "loss": 0.5253,
+      "step": 377
+    },
+    {
+      "epoch": 0.44575471698113206,
+      "grad_norm": 0.2960695663925585,
+      "learning_rate": 4.733508082131936e-05,
+      "loss": 0.5294,
+      "step": 378
+    },
+    {
+      "epoch": 0.44693396226415094,
+      "grad_norm": 0.26521643814710716,
+      "learning_rate": 4.731323722149411e-05,
+      "loss": 0.5431,
+      "step": 379
+    },
+    {
+      "epoch": 0.4481132075471698,
+      "grad_norm": 0.3319958661518853,
+      "learning_rate": 4.7291393621668854e-05,
+      "loss": 0.5242,
+      "step": 380
+    },
+    {
+      "epoch": 0.4492924528301887,
+      "grad_norm": 0.2699716670193801,
+      "learning_rate": 4.72695500218436e-05,
+      "loss": 0.556,
+      "step": 381
+    },
+    {
+      "epoch": 0.45047169811320753,
+      "grad_norm": 0.30779646112237813,
+      "learning_rate": 4.724770642201835e-05,
+      "loss": 0.5276,
+      "step": 382
+    },
+    {
+      "epoch": 0.4516509433962264,
+      "grad_norm": 0.2958760835830514,
+      "learning_rate": 4.7225862822193104e-05,
+      "loss": 0.5106,
+      "step": 383
+    },
+    {
+      "epoch": 0.4528301886792453,
+      "grad_norm": 0.2900591028954035,
+      "learning_rate": 4.720401922236785e-05,
+      "loss": 0.5319,
+      "step": 384
+    },
+    {
+      "epoch": 0.4540094339622642,
+      "grad_norm": 0.30679464652410204,
+      "learning_rate": 4.7182175622542594e-05,
+      "loss": 0.5245,
+      "step": 385
+    },
+    {
+      "epoch": 0.455188679245283,
+      "grad_norm": 0.26774977668970545,
+      "learning_rate": 4.7160332022717346e-05,
+      "loss": 0.5228,
+      "step": 386
+    },
+    {
+      "epoch": 0.4563679245283019,
+      "grad_norm": 0.27873633718721863,
+      "learning_rate": 4.71384884228921e-05,
+      "loss": 0.5203,
+      "step": 387
+    },
+    {
+      "epoch": 0.45754716981132076,
+      "grad_norm": 0.24356347086191166,
+      "learning_rate": 4.7116644823066844e-05,
+      "loss": 0.5349,
+      "step": 388
+    },
+    {
+      "epoch": 0.45872641509433965,
+      "grad_norm": 0.2534304699436805,
+      "learning_rate": 4.709480122324159e-05,
+      "loss": 0.4955,
+      "step": 389
+    },
+    {
+      "epoch": 0.45990566037735847,
+      "grad_norm": 0.27498577283977593,
+      "learning_rate": 4.707295762341634e-05,
+      "loss": 0.5184,
+      "step": 390
+    },
+    {
+      "epoch": 0.46108490566037735,
+      "grad_norm": 0.262640030640518,
+      "learning_rate": 4.705111402359109e-05,
+      "loss": 0.5324,
+      "step": 391
+    },
+    {
+      "epoch": 0.46226415094339623,
+      "grad_norm": 0.28861785365500653,
+      "learning_rate": 4.702927042376584e-05,
+      "loss": 0.512,
+      "step": 392
+    },
+    {
+      "epoch": 0.4634433962264151,
+      "grad_norm": 0.27578925464834636,
+      "learning_rate": 4.7007426823940584e-05,
+      "loss": 0.5416,
+      "step": 393
+    },
+    {
+      "epoch": 0.46462264150943394,
+      "grad_norm": 0.28295778863192367,
+      "learning_rate": 4.6985583224115336e-05,
+      "loss": 0.518,
+      "step": 394
+    },
+    {
+      "epoch": 0.4658018867924528,
+      "grad_norm": 0.29139055830232186,
+      "learning_rate": 4.696373962429009e-05,
+      "loss": 0.5532,
+      "step": 395
+    },
+    {
+      "epoch": 0.4669811320754717,
+      "grad_norm": 0.28693459759705536,
+      "learning_rate": 4.694189602446483e-05,
+      "loss": 0.5248,
+      "step": 396
+    },
+    {
+      "epoch": 0.4681603773584906,
+      "grad_norm": 0.25800527838560716,
+      "learning_rate": 4.6920052424639585e-05,
+      "loss": 0.5223,
+      "step": 397
+    },
+    {
+      "epoch": 0.4693396226415094,
+      "grad_norm": 0.25955156486065434,
+      "learning_rate": 4.689820882481433e-05,
+      "loss": 0.5266,
+      "step": 398
+    },
+    {
+      "epoch": 0.4705188679245283,
+      "grad_norm": 0.2517430644014706,
+      "learning_rate": 4.687636522498908e-05,
+      "loss": 0.5281,
+      "step": 399
+    },
+    {
+      "epoch": 0.4716981132075472,
+      "grad_norm": 0.26538790343867724,
+      "learning_rate": 4.685452162516383e-05,
+      "loss": 0.5127,
+      "step": 400
+    },
+    {
+      "epoch": 0.47287735849056606,
+      "grad_norm": 0.2794874029921559,
+      "learning_rate": 4.683267802533858e-05,
+      "loss": 0.5102,
+      "step": 401
+    },
+    {
+      "epoch": 0.4740566037735849,
+      "grad_norm": 0.2526417292194699,
+      "learning_rate": 4.6810834425513325e-05,
+      "loss": 0.5287,
+      "step": 402
+    },
+    {
+      "epoch": 0.47523584905660377,
+      "grad_norm": 0.30311422774966934,
+      "learning_rate": 4.678899082568808e-05,
+      "loss": 0.5196,
+      "step": 403
+    },
+    {
+      "epoch": 0.47641509433962265,
+      "grad_norm": 0.25885515300741013,
+      "learning_rate": 4.676714722586282e-05,
+      "loss": 0.531,
+      "step": 404
+    },
+    {
+      "epoch": 0.47759433962264153,
+      "grad_norm": 0.3063612772858509,
+      "learning_rate": 4.6745303626037575e-05,
+      "loss": 0.5157,
+      "step": 405
+    },
+    {
+      "epoch": 0.47877358490566035,
+      "grad_norm": 0.2907280379125239,
+      "learning_rate": 4.672346002621232e-05,
+      "loss": 0.5317,
+      "step": 406
+    },
+    {
+      "epoch": 0.47995283018867924,
+      "grad_norm": 0.27405325473950426,
+      "learning_rate": 4.6701616426387066e-05,
+      "loss": 0.5365,
+      "step": 407
+    },
+    {
+      "epoch": 0.4811320754716981,
+      "grad_norm": 0.2504324876011709,
+      "learning_rate": 4.667977282656182e-05,
+      "loss": 0.5152,
+      "step": 408
+    },
+    {
+      "epoch": 0.482311320754717,
+      "grad_norm": 0.3025304805938269,
+      "learning_rate": 4.665792922673657e-05,
+      "loss": 0.4879,
+      "step": 409
+    },
+    {
+      "epoch": 0.4834905660377358,
+      "grad_norm": 0.26789066496163716,
+      "learning_rate": 4.663608562691132e-05,
+      "loss": 0.528,
+      "step": 410
+    },
+    {
+      "epoch": 0.4846698113207547,
+      "grad_norm": 0.2914464421825445,
+      "learning_rate": 4.661424202708606e-05,
+      "loss": 0.5065,
+      "step": 411
+    },
+    {
+      "epoch": 0.4858490566037736,
+      "grad_norm": 0.3121272362740641,
+      "learning_rate": 4.659239842726081e-05,
+      "loss": 0.5483,
+      "step": 412
+    },
+    {
+      "epoch": 0.48702830188679247,
+      "grad_norm": 0.2924447729532895,
+      "learning_rate": 4.6570554827435565e-05,
+      "loss": 0.5194,
+      "step": 413
+    },
+    {
+      "epoch": 0.4882075471698113,
+      "grad_norm": 0.34173039840814984,
+      "learning_rate": 4.6548711227610317e-05,
+      "loss": 0.5473,
+      "step": 414
+    },
+    {
+      "epoch": 0.4893867924528302,
+      "grad_norm": 0.26128488759645097,
+      "learning_rate": 4.652686762778506e-05,
+      "loss": 0.5048,
+      "step": 415
+    },
+    {
+      "epoch": 0.49056603773584906,
+      "grad_norm": 0.3448433726123372,
+      "learning_rate": 4.650502402795981e-05,
+      "loss": 0.5129,
+      "step": 416
+    },
+    {
+      "epoch": 0.49174528301886794,
+      "grad_norm": 0.3517821249726224,
+      "learning_rate": 4.648318042813456e-05,
+      "loss": 0.5022,
+      "step": 417
+    },
+    {
+      "epoch": 0.49292452830188677,
+      "grad_norm": 0.28955878571498517,
+      "learning_rate": 4.646133682830931e-05,
+      "loss": 0.5085,
+      "step": 418
+    },
+    {
+      "epoch": 0.49410377358490565,
+      "grad_norm": 0.31006492547051273,
+      "learning_rate": 4.643949322848406e-05,
+      "loss": 0.5191,
+      "step": 419
+    },
+    {
+      "epoch": 0.49528301886792453,
+      "grad_norm": 0.24757931311014925,
+      "learning_rate": 4.64176496286588e-05,
+      "loss": 0.5126,
+      "step": 420
+    },
+    {
+      "epoch": 0.4964622641509434,
+      "grad_norm": 0.29856728192179655,
+      "learning_rate": 4.6395806028833554e-05,
+      "loss": 0.5184,
+      "step": 421
+    },
+    {
+      "epoch": 0.49764150943396224,
+      "grad_norm": 0.28962634276242777,
+      "learning_rate": 4.6373962429008306e-05,
+      "loss": 0.5276,
+      "step": 422
+    },
+    {
+      "epoch": 0.4988207547169811,
+      "grad_norm": 0.28265537272907354,
+      "learning_rate": 4.635211882918305e-05,
+      "loss": 0.5191,
+      "step": 423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2482772518678899,
+      "learning_rate": 4.6330275229357804e-05,
+      "loss": 0.5269,
+      "step": 424
+    },
+    {
+      "epoch": 0.5011792452830188,
+      "grad_norm": 0.27424191614970855,
+      "learning_rate": 4.630843162953255e-05,
+      "loss": 0.5334,
+      "step": 425
+    },
+    {
+      "epoch": 0.5023584905660378,
+      "grad_norm": 0.26145199985459167,
+      "learning_rate": 4.62865880297073e-05,
+      "loss": 0.513,
+      "step": 426
+    },
+    {
+      "epoch": 0.5035377358490566,
+      "grad_norm": 0.27241918606009785,
+      "learning_rate": 4.6264744429882046e-05,
+      "loss": 0.5427,
+      "step": 427
+    },
+    {
+      "epoch": 0.5047169811320755,
+      "grad_norm": 0.26783184022496637,
+      "learning_rate": 4.62429008300568e-05,
+      "loss": 0.5192,
+      "step": 428
+    },
+    {
+      "epoch": 0.5058962264150944,
+      "grad_norm": 0.2719599836110263,
+      "learning_rate": 4.6221057230231544e-05,
+      "loss": 0.5547,
+      "step": 429
+    },
+    {
+      "epoch": 0.5070754716981132,
+      "grad_norm": 0.2728064462992442,
+      "learning_rate": 4.619921363040629e-05,
+      "loss": 0.5152,
+      "step": 430
+    },
+    {
+      "epoch": 0.5082547169811321,
+      "grad_norm": 0.2577183774472052,
+      "learning_rate": 4.617737003058104e-05,
+      "loss": 0.5237,
+      "step": 431
+    },
+    {
+      "epoch": 0.5094339622641509,
+      "grad_norm": 0.2951448299640186,
+      "learning_rate": 4.615552643075579e-05,
+      "loss": 0.5288,
+      "step": 432
+    },
+    {
+      "epoch": 0.5106132075471698,
+      "grad_norm": 0.32700939825991465,
+      "learning_rate": 4.613368283093054e-05,
+      "loss": 0.5099,
+      "step": 433
+    },
+    {
+      "epoch": 0.5117924528301887,
+      "grad_norm": 0.28266284773355566,
+      "learning_rate": 4.6111839231105284e-05,
+      "loss": 0.5265,
+      "step": 434
+    },
+    {
+      "epoch": 0.5129716981132075,
+      "grad_norm": 0.3100911144503355,
+      "learning_rate": 4.6089995631280036e-05,
+      "loss": 0.5507,
+      "step": 435
+    },
+    {
+      "epoch": 0.5141509433962265,
+      "grad_norm": 0.2606090604665932,
+      "learning_rate": 4.606815203145479e-05,
+      "loss": 0.4933,
+      "step": 436
+    },
+    {
+      "epoch": 0.5153301886792453,
+      "grad_norm": 0.2886251823061575,
+      "learning_rate": 4.604630843162954e-05,
+      "loss": 0.5222,
+      "step": 437
+    },
+    {
+      "epoch": 0.5165094339622641,
+      "grad_norm": 0.256892313897739,
+      "learning_rate": 4.602446483180428e-05,
+      "loss": 0.5137,
+      "step": 438
+    },
+    {
+      "epoch": 0.5176886792452831,
+      "grad_norm": 0.26141306072246684,
+      "learning_rate": 4.600262123197903e-05,
+      "loss": 0.52,
+      "step": 439
+    },
+    {
+      "epoch": 0.5188679245283019,
+      "grad_norm": 0.2805386044027373,
+      "learning_rate": 4.598077763215378e-05,
+      "loss": 0.4942,
+      "step": 440
+    },
+    {
+      "epoch": 0.5200471698113207,
+      "grad_norm": 0.26269935601124783,
+      "learning_rate": 4.5958934032328535e-05,
+      "loss": 0.5384,
+      "step": 441
+    },
+    {
+      "epoch": 0.5212264150943396,
+      "grad_norm": 0.300554543076909,
+      "learning_rate": 4.593709043250328e-05,
+      "loss": 0.5256,
+      "step": 442
+    },
+    {
+      "epoch": 0.5224056603773585,
+      "grad_norm": 0.25750063776798665,
+      "learning_rate": 4.5915246832678025e-05,
+      "loss": 0.495,
+      "step": 443
+    },
+    {
+      "epoch": 0.5235849056603774,
+      "grad_norm": 0.2790691281783888,
+      "learning_rate": 4.589340323285278e-05,
+      "loss": 0.491,
+      "step": 444
+    },
+    {
+      "epoch": 0.5247641509433962,
+      "grad_norm": 0.27657898063492725,
+      "learning_rate": 4.587155963302753e-05,
+      "loss": 0.4848,
+      "step": 445
+    },
+    {
+      "epoch": 0.5259433962264151,
+      "grad_norm": 0.2712132743691986,
+      "learning_rate": 4.5849716033202275e-05,
+      "loss": 0.527,
+      "step": 446
+    },
+    {
+      "epoch": 0.527122641509434,
+      "grad_norm": 0.3109908351147375,
+      "learning_rate": 4.582787243337702e-05,
+      "loss": 0.5411,
+      "step": 447
+    },
+    {
+      "epoch": 0.5283018867924528,
+      "grad_norm": 0.24529239381565646,
+      "learning_rate": 4.580602883355177e-05,
+      "loss": 0.5061,
+      "step": 448
+    },
+    {
+      "epoch": 0.5294811320754716,
+      "grad_norm": 0.3131019304119496,
+      "learning_rate": 4.578418523372652e-05,
+      "loss": 0.5075,
+      "step": 449
+    },
+    {
+      "epoch": 0.5306603773584906,
+      "grad_norm": 0.2649807643456328,
+      "learning_rate": 4.576234163390127e-05,
+      "loss": 0.5035,
+      "step": 450
+    },
+    {
+      "epoch": 0.5318396226415094,
+      "grad_norm": 0.3325145601868781,
+      "learning_rate": 4.5740498034076015e-05,
+      "loss": 0.5244,
+      "step": 451
+    },
+    {
+      "epoch": 0.5330188679245284,
+      "grad_norm": 0.25006420290227555,
+      "learning_rate": 4.571865443425077e-05,
+      "loss": 0.5374,
+      "step": 452
+    },
+    {
+      "epoch": 0.5341981132075472,
+      "grad_norm": 0.34300122637624275,
+      "learning_rate": 4.569681083442551e-05,
+      "loss": 0.5497,
+      "step": 453
+    },
+    {
+      "epoch": 0.535377358490566,
+      "grad_norm": 0.2441218676541257,
+      "learning_rate": 4.5674967234600264e-05,
+      "loss": 0.5093,
+      "step": 454
+    },
+    {
+      "epoch": 0.5365566037735849,
+      "grad_norm": 0.31900264882464985,
+      "learning_rate": 4.5653123634775017e-05,
+      "loss": 0.5128,
+      "step": 455
+    },
+    {
+      "epoch": 0.5377358490566038,
+      "grad_norm": 0.26647041420393697,
+      "learning_rate": 4.563128003494976e-05,
+      "loss": 0.5197,
+      "step": 456
+    },
+    {
+      "epoch": 0.5389150943396226,
+      "grad_norm": 0.2762784691094774,
+      "learning_rate": 4.560943643512451e-05,
+      "loss": 0.5045,
+      "step": 457
+    },
+    {
+      "epoch": 0.5400943396226415,
+      "grad_norm": 0.26970067569470174,
+      "learning_rate": 4.558759283529926e-05,
+      "loss": 0.5573,
+      "step": 458
+    },
+    {
+      "epoch": 0.5412735849056604,
+      "grad_norm": 0.27296033448222384,
+      "learning_rate": 4.556574923547401e-05,
+      "loss": 0.4889,
+      "step": 459
+    },
+    {
+      "epoch": 0.5424528301886793,
+      "grad_norm": 0.2669782963452363,
+      "learning_rate": 4.5543905635648757e-05,
+      "loss": 0.5193,
+      "step": 460
+    },
+    {
+      "epoch": 0.5436320754716981,
+      "grad_norm": 0.2420839869194232,
+      "learning_rate": 4.55220620358235e-05,
+      "loss": 0.5066,
+      "step": 461
+    },
+    {
+      "epoch": 0.5448113207547169,
+      "grad_norm": 0.30075908192816014,
+      "learning_rate": 4.5500218435998254e-05,
+      "loss": 0.5209,
+      "step": 462
+    },
+    {
+      "epoch": 0.5459905660377359,
+      "grad_norm": 0.2616885502770783,
+      "learning_rate": 4.5478374836173006e-05,
+      "loss": 0.5005,
+      "step": 463
+    },
+    {
+      "epoch": 0.5471698113207547,
+      "grad_norm": 0.2748334340116257,
+      "learning_rate": 4.545653123634775e-05,
+      "loss": 0.5064,
+      "step": 464
+    },
+    {
+      "epoch": 0.5483490566037735,
+      "grad_norm": 0.26973231336067605,
+      "learning_rate": 4.54346876365225e-05,
+      "loss": 0.528,
+      "step": 465
+    },
+    {
+      "epoch": 0.5495283018867925,
+      "grad_norm": 0.25998315263430305,
+      "learning_rate": 4.541284403669725e-05,
+      "loss": 0.5366,
+      "step": 466
+    },
+    {
+      "epoch": 0.5507075471698113,
+      "grad_norm": 0.2820166882206177,
+      "learning_rate": 4.5391000436872e-05,
+      "loss": 0.5163,
+      "step": 467
+    },
+    {
+      "epoch": 0.5518867924528302,
+      "grad_norm": 0.28217752310684024,
+      "learning_rate": 4.536915683704675e-05,
+      "loss": 0.5257,
+      "step": 468
+    },
+    {
+      "epoch": 0.5530660377358491,
+      "grad_norm": 0.2773929929521568,
+      "learning_rate": 4.534731323722149e-05,
+      "loss": 0.5148,
+      "step": 469
+    },
+    {
+      "epoch": 0.5542452830188679,
+      "grad_norm": 0.28411543850154314,
+      "learning_rate": 4.5325469637396244e-05,
+      "loss": 0.5278,
+      "step": 470
+    },
+    {
+      "epoch": 0.5554245283018868,
+      "grad_norm": 0.279622349490722,
+      "learning_rate": 4.5303626037570996e-05,
+      "loss": 0.5132,
+      "step": 471
+    },
+    {
+      "epoch": 0.5566037735849056,
+      "grad_norm": 0.31096874115948614,
+      "learning_rate": 4.528178243774574e-05,
+      "loss": 0.5188,
+      "step": 472
+    },
+    {
+      "epoch": 0.5577830188679245,
+      "grad_norm": 0.26076428661828516,
+      "learning_rate": 4.525993883792049e-05,
+      "loss": 0.5151,
+      "step": 473
+    },
+    {
+      "epoch": 0.5589622641509434,
+      "grad_norm": 0.3330795332993912,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.5365,
+      "step": 474
+    },
+    {
+      "epoch": 0.5601415094339622,
+      "grad_norm": 0.2799085156011805,
+      "learning_rate": 4.521625163826999e-05,
+      "loss": 0.5185,
+      "step": 475
+    },
+    {
+      "epoch": 0.5613207547169812,
+      "grad_norm": 0.28259602526046196,
+      "learning_rate": 4.5194408038444736e-05,
+      "loss": 0.5247,
+      "step": 476
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.2594489957877634,
+      "learning_rate": 4.517256443861949e-05,
+      "loss": 0.5228,
+      "step": 477
+    },
+    {
+      "epoch": 0.5636792452830188,
+      "grad_norm": 0.26250268299035523,
+      "learning_rate": 4.515072083879423e-05,
+      "loss": 0.5278,
+      "step": 478
+    },
+    {
+      "epoch": 0.5648584905660378,
+      "grad_norm": 0.2887462101686738,
+      "learning_rate": 4.5128877238968985e-05,
+      "loss": 0.5397,
+      "step": 479
+    },
+    {
+      "epoch": 0.5660377358490566,
+      "grad_norm": 0.29582731677273394,
+      "learning_rate": 4.510703363914373e-05,
+      "loss": 0.5246,
+      "step": 480
+    },
+    {
+      "epoch": 0.5672169811320755,
+      "grad_norm": 0.3060531359871034,
+      "learning_rate": 4.508519003931848e-05,
+      "loss": 0.518,
+      "step": 481
+    },
+    {
+      "epoch": 0.5683962264150944,
+      "grad_norm": 0.3067665272673759,
+      "learning_rate": 4.5063346439493235e-05,
+      "loss": 0.5265,
+      "step": 482
+    },
+    {
+      "epoch": 0.5695754716981132,
+      "grad_norm": 0.2799032831799505,
+      "learning_rate": 4.504150283966798e-05,
+      "loss": 0.532,
+      "step": 483
+    },
+    {
+      "epoch": 0.5707547169811321,
+      "grad_norm": 0.2777526373186499,
+      "learning_rate": 4.5019659239842725e-05,
+      "loss": 0.5031,
+      "step": 484
+    },
+    {
+      "epoch": 0.5719339622641509,
+      "grad_norm": 0.2951780529323735,
+      "learning_rate": 4.499781564001748e-05,
+      "loss": 0.5002,
+      "step": 485
+    },
+    {
+      "epoch": 0.5731132075471698,
+      "grad_norm": 0.2739525693842679,
+      "learning_rate": 4.497597204019223e-05,
+      "loss": 0.5237,
+      "step": 486
+    },
+    {
+      "epoch": 0.5742924528301887,
+      "grad_norm": 0.2916650856029222,
+      "learning_rate": 4.4954128440366975e-05,
+      "loss": 0.5312,
+      "step": 487
+    },
+    {
+      "epoch": 0.5754716981132075,
+      "grad_norm": 0.2955966208775429,
+      "learning_rate": 4.493228484054172e-05,
+      "loss": 0.5011,
+      "step": 488
+    },
+    {
+      "epoch": 0.5766509433962265,
+      "grad_norm": 0.29476047670066957,
+      "learning_rate": 4.491044124071647e-05,
+      "loss": 0.5365,
+      "step": 489
+    },
+    {
+      "epoch": 0.5778301886792453,
+      "grad_norm": 0.27238123796033986,
+      "learning_rate": 4.4888597640891224e-05,
+      "loss": 0.486,
+      "step": 490
+    },
+    {
+      "epoch": 0.5790094339622641,
+      "grad_norm": 0.2534718562111071,
+      "learning_rate": 4.486675404106597e-05,
+      "loss": 0.506,
+      "step": 491
+    },
+    {
+      "epoch": 0.5801886792452831,
+      "grad_norm": 0.25172271549049324,
+      "learning_rate": 4.4844910441240715e-05,
+      "loss": 0.4992,
+      "step": 492
+    },
+    {
+      "epoch": 0.5813679245283019,
+      "grad_norm": 0.29657123538548724,
+      "learning_rate": 4.482306684141547e-05,
+      "loss": 0.5141,
+      "step": 493
+    },
+    {
+      "epoch": 0.5825471698113207,
+      "grad_norm": 0.2627838294317313,
+      "learning_rate": 4.480122324159022e-05,
+      "loss": 0.5333,
+      "step": 494
+    },
+    {
+      "epoch": 0.5837264150943396,
+      "grad_norm": 0.2868630459631176,
+      "learning_rate": 4.4779379641764964e-05,
+      "loss": 0.5046,
+      "step": 495
+    },
+    {
+      "epoch": 0.5849056603773585,
+      "grad_norm": 0.2898288915072131,
+      "learning_rate": 4.475753604193971e-05,
+      "loss": 0.5275,
+      "step": 496
+    },
+    {
+      "epoch": 0.5860849056603774,
+      "grad_norm": 0.27875110733515107,
+      "learning_rate": 4.473569244211446e-05,
+      "loss": 0.5007,
+      "step": 497
+    },
+    {
+      "epoch": 0.5872641509433962,
+      "grad_norm": 0.2959626202056171,
+      "learning_rate": 4.4713848842289214e-05,
+      "loss": 0.4887,
+      "step": 498
+    },
+    {
+      "epoch": 0.5884433962264151,
+      "grad_norm": 0.3089036795097389,
+      "learning_rate": 4.469200524246396e-05,
+      "loss": 0.5035,
+      "step": 499
+    },
+    {
+      "epoch": 0.589622641509434,
+      "grad_norm": 0.30684195482698545,
+      "learning_rate": 4.467016164263871e-05,
+      "loss": 0.5144,
+      "step": 500
+    },
+    {
+      "epoch": 0.5908018867924528,
+      "grad_norm": 0.28021783777308745,
+      "learning_rate": 4.4648318042813456e-05,
+      "loss": 0.5073,
+      "step": 501
+    },
+    {
+      "epoch": 0.5919811320754716,
+      "grad_norm": 0.29522691945216856,
+      "learning_rate": 4.462647444298821e-05,
+      "loss": 0.5316,
+      "step": 502
+    },
+    {
+      "epoch": 0.5931603773584906,
+      "grad_norm": 0.25326758265051136,
+      "learning_rate": 4.4604630843162954e-05,
+      "loss": 0.5118,
+      "step": 503
+    },
+    {
+      "epoch": 0.5943396226415094,
+      "grad_norm": 0.3144190380453841,
+      "learning_rate": 4.4582787243337706e-05,
+      "loss": 0.5049,
+      "step": 504
+    },
+    {
+      "epoch": 0.5955188679245284,
+      "grad_norm": 0.2567817299059237,
+      "learning_rate": 4.456094364351245e-05,
+      "loss": 0.5166,
+      "step": 505
+    },
+    {
+      "epoch": 0.5966981132075472,
+      "grad_norm": 0.2816605903635317,
+      "learning_rate": 4.45391000436872e-05,
+      "loss": 0.5117,
+      "step": 506
+    },
+    {
+      "epoch": 0.597877358490566,
+      "grad_norm": 0.2691810459007978,
+      "learning_rate": 4.451725644386195e-05,
+      "loss": 0.533,
+      "step": 507
+    },
+    {
+      "epoch": 0.5990566037735849,
+      "grad_norm": 0.2592649610992042,
+      "learning_rate": 4.44954128440367e-05,
+      "loss": 0.4638,
+      "step": 508
+    },
+    {
+      "epoch": 0.6002358490566038,
+      "grad_norm": 0.30985044413798773,
+      "learning_rate": 4.4473569244211446e-05,
+      "loss": 0.5029,
+      "step": 509
+    },
+    {
+      "epoch": 0.6014150943396226,
+      "grad_norm": 0.27700884730015984,
+      "learning_rate": 4.44517256443862e-05,
+      "loss": 0.4999,
+      "step": 510
+    },
+    {
+      "epoch": 0.6025943396226415,
+      "grad_norm": 0.3092777296146725,
+      "learning_rate": 4.4429882044560943e-05,
+      "loss": 0.5117,
+      "step": 511
+    },
+    {
+      "epoch": 0.6037735849056604,
+      "grad_norm": 0.2734175270692199,
+      "learning_rate": 4.4408038444735696e-05,
+      "loss": 0.503,
+      "step": 512
+    },
+    {
+      "epoch": 0.6049528301886793,
+      "grad_norm": 0.29544030216739003,
+      "learning_rate": 4.438619484491045e-05,
+      "loss": 0.5047,
+      "step": 513
+    },
+    {
+      "epoch": 0.6061320754716981,
+      "grad_norm": 0.3388271202700074,
+      "learning_rate": 4.4364351245085186e-05,
+      "loss": 0.5153,
+      "step": 514
+    },
+    {
+      "epoch": 0.6073113207547169,
+      "grad_norm": 0.3084420673835403,
+      "learning_rate": 4.434250764525994e-05,
+      "loss": 0.5149,
+      "step": 515
+    },
+    {
+      "epoch": 0.6084905660377359,
+      "grad_norm": 0.3738211650585104,
+      "learning_rate": 4.432066404543469e-05,
+      "loss": 0.5143,
+      "step": 516
+    },
+    {
+      "epoch": 0.6096698113207547,
+      "grad_norm": 0.37143047274685254,
+      "learning_rate": 4.429882044560944e-05,
+      "loss": 0.5161,
+      "step": 517
+    },
+    {
+      "epoch": 0.6108490566037735,
+      "grad_norm": 0.3279346978541537,
+      "learning_rate": 4.427697684578419e-05,
+      "loss": 0.4986,
+      "step": 518
+    },
+    {
+      "epoch": 0.6120283018867925,
+      "grad_norm": 0.3499921711754944,
+      "learning_rate": 4.425513324595893e-05,
+      "loss": 0.4962,
+      "step": 519
+    },
+    {
+      "epoch": 0.6132075471698113,
+      "grad_norm": 0.24981256061091087,
+      "learning_rate": 4.4233289646133685e-05,
+      "loss": 0.5086,
+      "step": 520
+    },
+    {
+      "epoch": 0.6143867924528302,
+      "grad_norm": 0.3627511471458732,
+      "learning_rate": 4.421144604630844e-05,
+      "loss": 0.5188,
+      "step": 521
+    },
+    {
+      "epoch": 0.6155660377358491,
+      "grad_norm": 0.2360377019039375,
+      "learning_rate": 4.418960244648318e-05,
+      "loss": 0.5151,
+      "step": 522
+    },
+    {
+      "epoch": 0.6167452830188679,
+      "grad_norm": 0.31807588971932416,
+      "learning_rate": 4.416775884665793e-05,
+      "loss": 0.5189,
+      "step": 523
+    },
+    {
+      "epoch": 0.6179245283018868,
+      "grad_norm": 0.2790258320330484,
+      "learning_rate": 4.414591524683268e-05,
+      "loss": 0.5081,
+      "step": 524
+    },
+    {
+      "epoch": 0.6191037735849056,
+      "grad_norm": 0.2812698157295098,
+      "learning_rate": 4.412407164700743e-05,
+      "loss": 0.4711,
+      "step": 525
+    },
+    {
+      "epoch": 0.6202830188679245,
+      "grad_norm": 0.2732600387967395,
+      "learning_rate": 4.410222804718218e-05,
+      "loss": 0.4725,
+      "step": 526
+    },
+    {
+      "epoch": 0.6214622641509434,
+      "grad_norm": 0.30848121736473383,
+      "learning_rate": 4.408038444735693e-05,
+      "loss": 0.5463,
+      "step": 527
+    },
+    {
+      "epoch": 0.6226415094339622,
+      "grad_norm": 0.27255547006989567,
+      "learning_rate": 4.4058540847531675e-05,
+      "loss": 0.4955,
+      "step": 528
+    },
+    {
+      "epoch": 0.6238207547169812,
+      "grad_norm": 0.3206505647697594,
+      "learning_rate": 4.403669724770643e-05,
+      "loss": 0.5489,
+      "step": 529
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.2682010420211527,
+      "learning_rate": 4.401485364788117e-05,
+      "loss": 0.5366,
+      "step": 530
+    },
+    {
+      "epoch": 0.6261792452830188,
+      "grad_norm": 0.31254640718277493,
+      "learning_rate": 4.3993010048055924e-05,
+      "loss": 0.5073,
+      "step": 531
+    },
+    {
+      "epoch": 0.6273584905660378,
+      "grad_norm": 0.2597008675719424,
+      "learning_rate": 4.397116644823067e-05,
+      "loss": 0.5176,
+      "step": 532
+    },
+    {
+      "epoch": 0.6285377358490566,
+      "grad_norm": 0.23040968036176274,
+      "learning_rate": 4.394932284840542e-05,
+      "loss": 0.4743,
+      "step": 533
+    },
+    {
+      "epoch": 0.6297169811320755,
+      "grad_norm": 0.31233139039432,
+      "learning_rate": 4.392747924858017e-05,
+      "loss": 0.5111,
+      "step": 534
+    },
+    {
+      "epoch": 0.6308962264150944,
+      "grad_norm": 0.2843468579184076,
+      "learning_rate": 4.390563564875492e-05,
+      "loss": 0.5312,
+      "step": 535
+    },
+    {
+      "epoch": 0.6320754716981132,
+      "grad_norm": 0.27743803201533385,
+      "learning_rate": 4.3883792048929664e-05,
+      "loss": 0.4972,
+      "step": 536
+    },
+    {
+      "epoch": 0.6332547169811321,
+      "grad_norm": 0.3272611205925401,
+      "learning_rate": 4.386194844910441e-05,
+      "loss": 0.5119,
+      "step": 537
+    },
+    {
+      "epoch": 0.6344339622641509,
+      "grad_norm": 0.2785556775773219,
+      "learning_rate": 4.384010484927916e-05,
+      "loss": 0.4995,
+      "step": 538
+    },
+    {
+      "epoch": 0.6356132075471698,
+      "grad_norm": 0.3145357178904373,
+      "learning_rate": 4.3818261249453914e-05,
+      "loss": 0.4788,
+      "step": 539
+    },
+    {
+      "epoch": 0.6367924528301887,
+      "grad_norm": 0.32887110754967924,
+      "learning_rate": 4.3796417649628666e-05,
+      "loss": 0.498,
+      "step": 540
+    },
+    {
+      "epoch": 0.6379716981132075,
+      "grad_norm": 0.2699554968797453,
+      "learning_rate": 4.3774574049803404e-05,
+      "loss": 0.5052,
+      "step": 541
+    },
+    {
+      "epoch": 0.6391509433962265,
+      "grad_norm": 0.3021008945393018,
+      "learning_rate": 4.3752730449978156e-05,
+      "loss": 0.5201,
+      "step": 542
+    },
+    {
+      "epoch": 0.6403301886792453,
+      "grad_norm": 0.22809332745169184,
+      "learning_rate": 4.373088685015291e-05,
+      "loss": 0.4728,
+      "step": 543
+    },
+    {
+      "epoch": 0.6415094339622641,
+      "grad_norm": 0.2771793913395451,
+      "learning_rate": 4.370904325032766e-05,
+      "loss": 0.5001,
+      "step": 544
+    },
+    {
+      "epoch": 0.6426886792452831,
+      "grad_norm": 0.23770279993523855,
+      "learning_rate": 4.3687199650502406e-05,
+      "loss": 0.5163,
+      "step": 545
+    },
+    {
+      "epoch": 0.6438679245283019,
+      "grad_norm": 0.29082883550991223,
+      "learning_rate": 4.366535605067715e-05,
+      "loss": 0.5304,
+      "step": 546
+    },
+    {
+      "epoch": 0.6450471698113207,
+      "grad_norm": 0.2781201595422939,
+      "learning_rate": 4.36435124508519e-05,
+      "loss": 0.5066,
+      "step": 547
+    },
+    {
+      "epoch": 0.6462264150943396,
+      "grad_norm": 0.2575565279322435,
+      "learning_rate": 4.3621668851026655e-05,
+      "loss": 0.4954,
+      "step": 548
+    },
+    {
+      "epoch": 0.6474056603773585,
+      "grad_norm": 0.2643909809039024,
+      "learning_rate": 4.35998252512014e-05,
+      "loss": 0.5135,
+      "step": 549
+    },
+    {
+      "epoch": 0.6485849056603774,
+      "grad_norm": 0.25135696901136934,
+      "learning_rate": 4.3577981651376146e-05,
+      "loss": 0.5277,
+      "step": 550
+    },
+    {
+      "epoch": 0.6497641509433962,
+      "grad_norm": 0.2640900075653264,
+      "learning_rate": 4.35561380515509e-05,
+      "loss": 0.5134,
+      "step": 551
+    },
+    {
+      "epoch": 0.6509433962264151,
+      "grad_norm": 0.2717972824881742,
+      "learning_rate": 4.353429445172565e-05,
+      "loss": 0.5038,
+      "step": 552
+    },
+    {
+      "epoch": 0.652122641509434,
+      "grad_norm": 0.2798642182490913,
+      "learning_rate": 4.3512450851900395e-05,
+      "loss": 0.4971,
+      "step": 553
+    },
+    {
+      "epoch": 0.6533018867924528,
+      "grad_norm": 0.27410116224771286,
+      "learning_rate": 4.349060725207514e-05,
+      "loss": 0.4912,
+      "step": 554
+    },
+    {
+      "epoch": 0.6544811320754716,
+      "grad_norm": 0.28766988860278325,
+      "learning_rate": 4.346876365224989e-05,
+      "loss": 0.5015,
+      "step": 555
+    },
+    {
+      "epoch": 0.6556603773584906,
+      "grad_norm": 0.256619722379234,
+      "learning_rate": 4.344692005242464e-05,
+      "loss": 0.5027,
+      "step": 556
+    },
+    {
+      "epoch": 0.6568396226415094,
+      "grad_norm": 0.24456497407395852,
+      "learning_rate": 4.342507645259939e-05,
+      "loss": 0.4809,
+      "step": 557
+    },
+    {
+      "epoch": 0.6580188679245284,
+      "grad_norm": 0.2590067680797023,
+      "learning_rate": 4.340323285277414e-05,
+      "loss": 0.5131,
+      "step": 558
+    },
+    {
+      "epoch": 0.6591981132075472,
+      "grad_norm": 0.27695021593828395,
+      "learning_rate": 4.338138925294889e-05,
+      "loss": 0.4933,
+      "step": 559
+    },
+    {
+      "epoch": 0.660377358490566,
+      "grad_norm": 0.25400063354283453,
+      "learning_rate": 4.335954565312363e-05,
+      "loss": 0.4805,
+      "step": 560
+    },
+    {
+      "epoch": 0.6615566037735849,
+      "grad_norm": 0.2830615969762543,
+      "learning_rate": 4.3337702053298385e-05,
+      "loss": 0.5053,
+      "step": 561
+    },
+    {
+      "epoch": 0.6627358490566038,
+      "grad_norm": 0.23911955430050397,
+      "learning_rate": 4.331585845347314e-05,
+      "loss": 0.4937,
+      "step": 562
+    },
+    {
+      "epoch": 0.6639150943396226,
+      "grad_norm": 0.2652899961692889,
+      "learning_rate": 4.329401485364788e-05,
+      "loss": 0.4873,
+      "step": 563
+    },
+    {
+      "epoch": 0.6650943396226415,
+      "grad_norm": 0.2559779130702805,
+      "learning_rate": 4.327217125382263e-05,
+      "loss": 0.5014,
+      "step": 564
+    },
+    {
+      "epoch": 0.6662735849056604,
+      "grad_norm": 0.26094530321324066,
+      "learning_rate": 4.325032765399738e-05,
+      "loss": 0.5006,
+      "step": 565
+    },
+    {
+      "epoch": 0.6674528301886793,
+      "grad_norm": 0.26243639453065026,
+      "learning_rate": 4.322848405417213e-05,
+      "loss": 0.5095,
+      "step": 566
+    },
+    {
+      "epoch": 0.6686320754716981,
+      "grad_norm": 0.2818870320975374,
+      "learning_rate": 4.3206640454346884e-05,
+      "loss": 0.5111,
+      "step": 567
+    },
+    {
+      "epoch": 0.6698113207547169,
+      "grad_norm": 0.2678259481237977,
+      "learning_rate": 4.318479685452162e-05,
+      "loss": 0.5284,
+      "step": 568
+    },
+    {
+      "epoch": 0.6709905660377359,
+      "grad_norm": 0.26727573093468215,
+      "learning_rate": 4.3162953254696375e-05,
+      "loss": 0.5103,
+      "step": 569
+    },
+    {
+      "epoch": 0.6721698113207547,
+      "grad_norm": 0.2552532802886714,
+      "learning_rate": 4.314110965487113e-05,
+      "loss": 0.525,
+      "step": 570
+    },
+    {
+      "epoch": 0.6733490566037735,
+      "grad_norm": 0.28505526058066244,
+      "learning_rate": 4.311926605504588e-05,
+      "loss": 0.5127,
+      "step": 571
+    },
+    {
+      "epoch": 0.6745283018867925,
+      "grad_norm": 0.2738879294541768,
+      "learning_rate": 4.3097422455220624e-05,
+      "loss": 0.481,
+      "step": 572
+    },
+    {
+      "epoch": 0.6757075471698113,
+      "grad_norm": 0.27418973346616005,
+      "learning_rate": 4.307557885539537e-05,
+      "loss": 0.5148,
+      "step": 573
+    },
+    {
+      "epoch": 0.6768867924528302,
+      "grad_norm": 0.267413174392808,
+      "learning_rate": 4.305373525557012e-05,
+      "loss": 0.5104,
+      "step": 574
+    },
+    {
+      "epoch": 0.6780660377358491,
+      "grad_norm": 0.2792581621973113,
+      "learning_rate": 4.3031891655744874e-05,
+      "loss": 0.5018,
+      "step": 575
+    },
+    {
+      "epoch": 0.6792452830188679,
+      "grad_norm": 0.28147460571056965,
+      "learning_rate": 4.301004805591962e-05,
+      "loss": 0.5304,
+      "step": 576
+    },
+    {
+      "epoch": 0.6804245283018868,
+      "grad_norm": 0.30049642812193145,
+      "learning_rate": 4.2988204456094364e-05,
+      "loss": 0.5158,
+      "step": 577
+    },
+    {
+      "epoch": 0.6816037735849056,
+      "grad_norm": 0.26938846485188966,
+      "learning_rate": 4.2966360856269116e-05,
+      "loss": 0.502,
+      "step": 578
+    },
+    {
+      "epoch": 0.6827830188679245,
+      "grad_norm": 0.2819870328440458,
+      "learning_rate": 4.294451725644386e-05,
+      "loss": 0.4925,
+      "step": 579
+    },
+    {
+      "epoch": 0.6839622641509434,
+      "grad_norm": 0.3118133320372834,
+      "learning_rate": 4.2922673656618614e-05,
+      "loss": 0.4898,
+      "step": 580
+    },
+    {
+      "epoch": 0.6851415094339622,
+      "grad_norm": 0.255144014392819,
+      "learning_rate": 4.290083005679336e-05,
+      "loss": 0.4997,
+      "step": 581
+    },
+    {
+      "epoch": 0.6863207547169812,
+      "grad_norm": 0.2991856946525077,
+      "learning_rate": 4.287898645696811e-05,
+      "loss": 0.5182,
+      "step": 582
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.2552400651173224,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.5013,
+      "step": 583
+    },
+    {
+      "epoch": 0.6886792452830188,
+      "grad_norm": 0.26200314577611217,
+      "learning_rate": 4.283529925731761e-05,
+      "loss": 0.5018,
+      "step": 584
+    },
+    {
+      "epoch": 0.6898584905660378,
+      "grad_norm": 0.2717669894451444,
+      "learning_rate": 4.281345565749236e-05,
+      "loss": 0.5119,
+      "step": 585
+    },
+    {
+      "epoch": 0.6910377358490566,
+      "grad_norm": 0.22531758336078928,
+      "learning_rate": 4.2791612057667106e-05,
+      "loss": 0.4882,
+      "step": 586
+    },
+    {
+      "epoch": 0.6922169811320755,
+      "grad_norm": 0.26803714949054,
+      "learning_rate": 4.276976845784185e-05,
+      "loss": 0.5329,
+      "step": 587
+    },
+    {
+      "epoch": 0.6933962264150944,
+      "grad_norm": 0.2583859970460127,
+      "learning_rate": 4.27479248580166e-05,
+      "loss": 0.5324,
+      "step": 588
+    },
+    {
+      "epoch": 0.6945754716981132,
+      "grad_norm": 0.4059176539071615,
+      "learning_rate": 4.2726081258191355e-05,
+      "loss": 0.5005,
+      "step": 589
+    },
+    {
+      "epoch": 0.6957547169811321,
+      "grad_norm": 0.3155785687249483,
+      "learning_rate": 4.27042376583661e-05,
+      "loss": 0.5207,
+      "step": 590
+    },
+    {
+      "epoch": 0.6969339622641509,
+      "grad_norm": 0.24936114903790027,
+      "learning_rate": 4.2682394058540846e-05,
+      "loss": 0.5023,
+      "step": 591
+    },
+    {
+      "epoch": 0.6981132075471698,
+      "grad_norm": 0.26418411816745624,
+      "learning_rate": 4.26605504587156e-05,
+      "loss": 0.4936,
+      "step": 592
+    },
+    {
+      "epoch": 0.6992924528301887,
+      "grad_norm": 0.2749790198690333,
+      "learning_rate": 4.263870685889035e-05,
+      "loss": 0.5082,
+      "step": 593
+    },
+    {
+      "epoch": 0.7004716981132075,
+      "grad_norm": 0.2146653745114201,
+      "learning_rate": 4.2616863259065095e-05,
+      "loss": 0.4857,
+      "step": 594
+    },
+    {
+      "epoch": 0.7016509433962265,
+      "grad_norm": 0.24291347395173268,
+      "learning_rate": 4.259501965923984e-05,
+      "loss": 0.4634,
+      "step": 595
+    },
+    {
+      "epoch": 0.7028301886792453,
+      "grad_norm": 0.25191450742631893,
+      "learning_rate": 4.257317605941459e-05,
+      "loss": 0.5317,
+      "step": 596
+    },
+    {
+      "epoch": 0.7040094339622641,
+      "grad_norm": 0.228218721995259,
+      "learning_rate": 4.2551332459589345e-05,
+      "loss": 0.4894,
+      "step": 597
+    },
+    {
+      "epoch": 0.7051886792452831,
+      "grad_norm": 0.2545000545269755,
+      "learning_rate": 4.25294888597641e-05,
+      "loss": 0.5038,
+      "step": 598
+    },
+    {
+      "epoch": 0.7063679245283019,
+      "grad_norm": 0.2328904783830322,
+      "learning_rate": 4.2507645259938835e-05,
+      "loss": 0.4928,
+      "step": 599
+    },
+    {
+      "epoch": 0.7075471698113207,
+      "grad_norm": 0.2523663348143776,
+      "learning_rate": 4.248580166011359e-05,
+      "loss": 0.5122,
+      "step": 600
+    },
+    {
+      "epoch": 0.7087264150943396,
+      "grad_norm": 0.24955303440512616,
+      "learning_rate": 4.246395806028834e-05,
+      "loss": 0.5182,
+      "step": 601
+    },
+    {
+      "epoch": 0.7099056603773585,
+      "grad_norm": 0.24133356806206482,
+      "learning_rate": 4.2442114460463085e-05,
+      "loss": 0.4943,
+      "step": 602
+    },
+    {
+      "epoch": 0.7110849056603774,
+      "grad_norm": 0.2624303955147311,
+      "learning_rate": 4.242027086063784e-05,
+      "loss": 0.5064,
+      "step": 603
+    },
+    {
+      "epoch": 0.7122641509433962,
+      "grad_norm": 0.26505680976514995,
+      "learning_rate": 4.239842726081258e-05,
+      "loss": 0.5438,
+      "step": 604
+    },
+    {
+      "epoch": 0.7134433962264151,
+      "grad_norm": 0.26090173907555286,
+      "learning_rate": 4.2376583660987334e-05,
+      "loss": 0.5047,
+      "step": 605
+    },
+    {
+      "epoch": 0.714622641509434,
+      "grad_norm": 0.23247037861377723,
+      "learning_rate": 4.235474006116208e-05,
+      "loss": 0.5287,
+      "step": 606
+    },
+    {
+      "epoch": 0.7158018867924528,
+      "grad_norm": 0.24566726151092622,
+      "learning_rate": 4.233289646133683e-05,
+      "loss": 0.4777,
+      "step": 607
+    },
+    {
+      "epoch": 0.7169811320754716,
+      "grad_norm": 0.23444437665071782,
+      "learning_rate": 4.231105286151158e-05,
+      "loss": 0.4992,
+      "step": 608
+    },
+    {
+      "epoch": 0.7181603773584906,
+      "grad_norm": 0.2208204538534351,
+      "learning_rate": 4.228920926168633e-05,
+      "loss": 0.4937,
+      "step": 609
+    },
+    {
+      "epoch": 0.7193396226415094,
+      "grad_norm": 0.28987265958525427,
+      "learning_rate": 4.2267365661861074e-05,
+      "loss": 0.5459,
+      "step": 610
+    },
+    {
+      "epoch": 0.7205188679245284,
+      "grad_norm": 0.2274846712895149,
+      "learning_rate": 4.2245522062035827e-05,
+      "loss": 0.5009,
+      "step": 611
+    },
+    {
+      "epoch": 0.7216981132075472,
+      "grad_norm": 0.24395038174975997,
+      "learning_rate": 4.222367846221057e-05,
+      "loss": 0.5021,
+      "step": 612
+    },
+    {
+      "epoch": 0.722877358490566,
+      "grad_norm": 0.27895528458379504,
+      "learning_rate": 4.2201834862385324e-05,
+      "loss": 0.5058,
+      "step": 613
+    },
+    {
+      "epoch": 0.7240566037735849,
+      "grad_norm": 0.23877081132170094,
+      "learning_rate": 4.217999126256007e-05,
+      "loss": 0.5191,
+      "step": 614
+    },
+    {
+      "epoch": 0.7252358490566038,
+      "grad_norm": 0.28921946713597996,
+      "learning_rate": 4.215814766273482e-05,
+      "loss": 0.4654,
+      "step": 615
+    },
+    {
+      "epoch": 0.7264150943396226,
+      "grad_norm": 0.2604275990304178,
+      "learning_rate": 4.2136304062909573e-05,
+      "loss": 0.5447,
+      "step": 616
+    },
+    {
+      "epoch": 0.7275943396226415,
+      "grad_norm": 0.2834986366212253,
+      "learning_rate": 4.211446046308432e-05,
+      "loss": 0.4871,
+      "step": 617
+    },
+    {
+      "epoch": 0.7287735849056604,
+      "grad_norm": 0.24814609728479958,
+      "learning_rate": 4.2092616863259064e-05,
+      "loss": 0.5052,
+      "step": 618
+    },
+    {
+      "epoch": 0.7299528301886793,
+      "grad_norm": 0.2732157664365053,
+      "learning_rate": 4.2070773263433816e-05,
+      "loss": 0.512,
+      "step": 619
+    },
+    {
+      "epoch": 0.7311320754716981,
+      "grad_norm": 0.2836601784682677,
+      "learning_rate": 4.204892966360857e-05,
+      "loss": 0.5032,
+      "step": 620
+    },
+    {
+      "epoch": 0.7323113207547169,
+      "grad_norm": 0.2747463140608284,
+      "learning_rate": 4.2027086063783313e-05,
+      "loss": 0.5173,
+      "step": 621
+    },
+    {
+      "epoch": 0.7334905660377359,
+      "grad_norm": 0.26247329047404183,
+      "learning_rate": 4.200524246395806e-05,
+      "loss": 0.5143,
+      "step": 622
+    },
+    {
+      "epoch": 0.7346698113207547,
+      "grad_norm": 0.2737343099223215,
+      "learning_rate": 4.198339886413281e-05,
+      "loss": 0.5291,
+      "step": 623
+    },
+    {
+      "epoch": 0.7358490566037735,
+      "grad_norm": 0.2455359017090087,
+      "learning_rate": 4.196155526430756e-05,
+      "loss": 0.4858,
+      "step": 624
+    },
+    {
+      "epoch": 0.7370283018867925,
+      "grad_norm": 0.290659535087624,
+      "learning_rate": 4.193971166448231e-05,
+      "loss": 0.5177,
+      "step": 625
+    },
+    {
+      "epoch": 0.7382075471698113,
+      "grad_norm": 0.2759552781706607,
+      "learning_rate": 4.1917868064657054e-05,
+      "loss": 0.4839,
+      "step": 626
+    },
+    {
+      "epoch": 0.7393867924528302,
+      "grad_norm": 0.270686717835934,
+      "learning_rate": 4.1896024464831806e-05,
+      "loss": 0.5088,
+      "step": 627
+    },
+    {
+      "epoch": 0.7405660377358491,
+      "grad_norm": 0.26479024357482384,
+      "learning_rate": 4.187418086500656e-05,
+      "loss": 0.5171,
+      "step": 628
+    },
+    {
+      "epoch": 0.7417452830188679,
+      "grad_norm": 0.2508882633219807,
+      "learning_rate": 4.18523372651813e-05,
+      "loss": 0.5111,
+      "step": 629
+    },
+    {
+      "epoch": 0.7429245283018868,
+      "grad_norm": 0.3072771230532909,
+      "learning_rate": 4.1830493665356055e-05,
+      "loss": 0.5203,
+      "step": 630
+    },
+    {
+      "epoch": 0.7441037735849056,
+      "grad_norm": 0.24542159689894488,
+      "learning_rate": 4.18086500655308e-05,
+      "loss": 0.4937,
+      "step": 631
+    },
+    {
+      "epoch": 0.7452830188679245,
+      "grad_norm": 0.2586499685773474,
+      "learning_rate": 4.178680646570555e-05,
+      "loss": 0.5095,
+      "step": 632
+    },
+    {
+      "epoch": 0.7464622641509434,
+      "grad_norm": 0.2522822577421208,
+      "learning_rate": 4.17649628658803e-05,
+      "loss": 0.4962,
+      "step": 633
+    },
+    {
+      "epoch": 0.7476415094339622,
+      "grad_norm": 0.26818480965181823,
+      "learning_rate": 4.174311926605505e-05,
+      "loss": 0.4919,
+      "step": 634
+    },
+    {
+      "epoch": 0.7488207547169812,
+      "grad_norm": 0.23935946126636562,
+      "learning_rate": 4.1721275666229795e-05,
+      "loss": 0.5113,
+      "step": 635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24948420118060155,
+      "learning_rate": 4.169943206640455e-05,
+      "loss": 0.514,
+      "step": 636
+    },
+    {
+      "epoch": 0.7511792452830188,
+      "grad_norm": 0.2254944570520605,
+      "learning_rate": 4.167758846657929e-05,
+      "loss": 0.5029,
+      "step": 637
+    },
+    {
+      "epoch": 0.7523584905660378,
+      "grad_norm": 0.2718236470988199,
+      "learning_rate": 4.1655744866754045e-05,
+      "loss": 0.5207,
+      "step": 638
+    },
+    {
+      "epoch": 0.7535377358490566,
+      "grad_norm": 0.2397021808718291,
+      "learning_rate": 4.163390126692879e-05,
+      "loss": 0.5019,
+      "step": 639
+    },
+    {
+      "epoch": 0.7547169811320755,
+      "grad_norm": 0.2347604429762889,
+      "learning_rate": 4.161205766710354e-05,
+      "loss": 0.4738,
+      "step": 640
+    },
+    {
+      "epoch": 0.7558962264150944,
+      "grad_norm": 0.2541915342398753,
+      "learning_rate": 4.159021406727829e-05,
+      "loss": 0.5043,
+      "step": 641
+    },
+    {
+      "epoch": 0.7570754716981132,
+      "grad_norm": 0.22768268739291994,
+      "learning_rate": 4.156837046745304e-05,
+      "loss": 0.4983,
+      "step": 642
+    },
+    {
+      "epoch": 0.7582547169811321,
+      "grad_norm": 0.25304167587643683,
+      "learning_rate": 4.154652686762779e-05,
+      "loss": 0.487,
+      "step": 643
+    },
+    {
+      "epoch": 0.7594339622641509,
+      "grad_norm": 0.2562597738965785,
+      "learning_rate": 4.152468326780253e-05,
+      "loss": 0.5016,
+      "step": 644
+    },
+    {
+      "epoch": 0.7606132075471698,
+      "grad_norm": 0.2429403222622004,
+      "learning_rate": 4.150283966797728e-05,
+      "loss": 0.5047,
+      "step": 645
+    },
+    {
+      "epoch": 0.7617924528301887,
+      "grad_norm": 0.2564398221182257,
+      "learning_rate": 4.1480996068152034e-05,
+      "loss": 0.5101,
+      "step": 646
+    },
+    {
+      "epoch": 0.7629716981132075,
+      "grad_norm": 0.23109813957827627,
+      "learning_rate": 4.1459152468326786e-05,
+      "loss": 0.4817,
+      "step": 647
+    },
+    {
+      "epoch": 0.7641509433962265,
+      "grad_norm": 0.23556655647422345,
+      "learning_rate": 4.143730886850153e-05,
+      "loss": 0.5056,
+      "step": 648
+    },
+    {
+      "epoch": 0.7653301886792453,
+      "grad_norm": 0.2439306848883772,
+      "learning_rate": 4.141546526867628e-05,
+      "loss": 0.5153,
+      "step": 649
+    },
+    {
+      "epoch": 0.7665094339622641,
+      "grad_norm": 0.22385931114682797,
+      "learning_rate": 4.139362166885103e-05,
+      "loss": 0.4963,
+      "step": 650
+    },
+    {
+      "epoch": 0.7676886792452831,
+      "grad_norm": 0.2479181059999666,
+      "learning_rate": 4.137177806902578e-05,
+      "loss": 0.4946,
+      "step": 651
+    },
+    {
+      "epoch": 0.7688679245283019,
+      "grad_norm": 0.22682893788241226,
+      "learning_rate": 4.1349934469200526e-05,
+      "loss": 0.5059,
+      "step": 652
+    },
+    {
+      "epoch": 0.7700471698113207,
+      "grad_norm": 0.25480910643762983,
+      "learning_rate": 4.132809086937527e-05,
+      "loss": 0.5064,
+      "step": 653
+    },
+    {
+      "epoch": 0.7712264150943396,
+      "grad_norm": 0.24226850209039316,
+      "learning_rate": 4.1306247269550024e-05,
+      "loss": 0.501,
+      "step": 654
+    },
+    {
+      "epoch": 0.7724056603773585,
+      "grad_norm": 0.2300853003154503,
+      "learning_rate": 4.1284403669724776e-05,
+      "loss": 0.4798,
+      "step": 655
+    },
+    {
+      "epoch": 0.7735849056603774,
+      "grad_norm": 0.23171147909216228,
+      "learning_rate": 4.126256006989952e-05,
+      "loss": 0.5016,
+      "step": 656
+    },
+    {
+      "epoch": 0.7747641509433962,
+      "grad_norm": 0.2244466407010123,
+      "learning_rate": 4.1240716470074267e-05,
+      "loss": 0.5028,
+      "step": 657
+    },
+    {
+      "epoch": 0.7759433962264151,
+      "grad_norm": 0.2282089085207076,
+      "learning_rate": 4.121887287024902e-05,
+      "loss": 0.51,
+      "step": 658
+    },
+    {
+      "epoch": 0.777122641509434,
+      "grad_norm": 0.2266481807773254,
+      "learning_rate": 4.119702927042377e-05,
+      "loss": 0.4782,
+      "step": 659
+    },
+    {
+      "epoch": 0.7783018867924528,
+      "grad_norm": 0.22555792807357655,
+      "learning_rate": 4.1175185670598516e-05,
+      "loss": 0.5081,
+      "step": 660
+    },
+    {
+      "epoch": 0.7794811320754716,
+      "grad_norm": 0.22828883360405255,
+      "learning_rate": 4.115334207077327e-05,
+      "loss": 0.4871,
+      "step": 661
+    },
+    {
+      "epoch": 0.7806603773584906,
+      "grad_norm": 0.24136842279832876,
+      "learning_rate": 4.1131498470948013e-05,
+      "loss": 0.5221,
+      "step": 662
+    },
+    {
+      "epoch": 0.7818396226415094,
+      "grad_norm": 0.2480443907027707,
+      "learning_rate": 4.110965487112276e-05,
+      "loss": 0.4882,
+      "step": 663
+    },
+    {
+      "epoch": 0.7830188679245284,
+      "grad_norm": 0.2879712871390392,
+      "learning_rate": 4.108781127129751e-05,
+      "loss": 0.5308,
+      "step": 664
+    },
+    {
+      "epoch": 0.7841981132075472,
+      "grad_norm": 0.2355318095641946,
+      "learning_rate": 4.106596767147226e-05,
+      "loss": 0.5027,
+      "step": 665
+    },
+    {
+      "epoch": 0.785377358490566,
+      "grad_norm": 0.2637926256192344,
+      "learning_rate": 4.104412407164701e-05,
+      "loss": 0.5,
+      "step": 666
+    },
+    {
+      "epoch": 0.7865566037735849,
+      "grad_norm": 0.2555381169153006,
+      "learning_rate": 4.1022280471821753e-05,
+      "loss": 0.5126,
+      "step": 667
+    },
+    {
+      "epoch": 0.7877358490566038,
+      "grad_norm": 0.24883324820067274,
+      "learning_rate": 4.1000436871996506e-05,
+      "loss": 0.5411,
+      "step": 668
+    },
+    {
+      "epoch": 0.7889150943396226,
+      "grad_norm": 0.29814929958401626,
+      "learning_rate": 4.097859327217126e-05,
+      "loss": 0.4811,
+      "step": 669
+    },
+    {
+      "epoch": 0.7900943396226415,
+      "grad_norm": 0.2699705099111919,
+      "learning_rate": 4.095674967234601e-05,
+      "loss": 0.5026,
+      "step": 670
+    },
+    {
+      "epoch": 0.7912735849056604,
+      "grad_norm": 0.2711521401604636,
+      "learning_rate": 4.093490607252075e-05,
+      "loss": 0.52,
+      "step": 671
+    },
+    {
+      "epoch": 0.7924528301886793,
+      "grad_norm": 0.2823994270234719,
+      "learning_rate": 4.09130624726955e-05,
+      "loss": 0.5166,
+      "step": 672
+    },
+    {
+      "epoch": 0.7936320754716981,
+      "grad_norm": 0.28003812919391935,
+      "learning_rate": 4.089121887287025e-05,
+      "loss": 0.4893,
+      "step": 673
+    },
+    {
+      "epoch": 0.7948113207547169,
+      "grad_norm": 0.2832080644800241,
+      "learning_rate": 4.0869375273045005e-05,
+      "loss": 0.4921,
+      "step": 674
+    },
+    {
+      "epoch": 0.7959905660377359,
+      "grad_norm": 0.2573481554499683,
+      "learning_rate": 4.084753167321975e-05,
+      "loss": 0.4712,
+      "step": 675
+    },
+    {
+      "epoch": 0.7971698113207547,
+      "grad_norm": 0.25699711764398914,
+      "learning_rate": 4.0825688073394495e-05,
+      "loss": 0.4712,
+      "step": 676
+    },
+    {
+      "epoch": 0.7983490566037735,
+      "grad_norm": 0.25032558462629795,
+      "learning_rate": 4.080384447356925e-05,
+      "loss": 0.4668,
+      "step": 677
+    },
+    {
+      "epoch": 0.7995283018867925,
+      "grad_norm": 0.3183457778892485,
+      "learning_rate": 4.0782000873744e-05,
+      "loss": 0.5432,
+      "step": 678
+    },
+    {
+      "epoch": 0.8007075471698113,
+      "grad_norm": 0.2688323668113325,
+      "learning_rate": 4.0760157273918745e-05,
+      "loss": 0.5103,
+      "step": 679
+    },
+    {
+      "epoch": 0.8018867924528302,
+      "grad_norm": 0.2604465483450334,
+      "learning_rate": 4.073831367409349e-05,
+      "loss": 0.503,
+      "step": 680
+    },
+    {
+      "epoch": 0.8030660377358491,
+      "grad_norm": 0.25792246582422473,
+      "learning_rate": 4.071647007426824e-05,
+      "loss": 0.4961,
+      "step": 681
+    },
+    {
+      "epoch": 0.8042452830188679,
+      "grad_norm": 0.23237432729230614,
+      "learning_rate": 4.0694626474442994e-05,
+      "loss": 0.5055,
+      "step": 682
+    },
+    {
+      "epoch": 0.8054245283018868,
+      "grad_norm": 0.2489991052564345,
+      "learning_rate": 4.067278287461774e-05,
+      "loss": 0.4855,
+      "step": 683
+    },
+    {
+      "epoch": 0.8066037735849056,
+      "grad_norm": 0.2672742420995334,
+      "learning_rate": 4.0650939274792485e-05,
+      "loss": 0.5199,
+      "step": 684
+    },
+    {
+      "epoch": 0.8077830188679245,
+      "grad_norm": 0.24339991359605592,
+      "learning_rate": 4.062909567496724e-05,
+      "loss": 0.503,
+      "step": 685
+    },
+    {
+      "epoch": 0.8089622641509434,
+      "grad_norm": 0.26080158250553054,
+      "learning_rate": 4.060725207514198e-05,
+      "loss": 0.5058,
+      "step": 686
+    },
+    {
+      "epoch": 0.8101415094339622,
+      "grad_norm": 0.22850722317077063,
+      "learning_rate": 4.0585408475316734e-05,
+      "loss": 0.5008,
+      "step": 687
+    },
+    {
+      "epoch": 0.8113207547169812,
+      "grad_norm": 0.23243796868231534,
+      "learning_rate": 4.0563564875491486e-05,
+      "loss": 0.5116,
+      "step": 688
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 0.24179837258194165,
+      "learning_rate": 4.054172127566623e-05,
+      "loss": 0.4782,
+      "step": 689
+    },
+    {
+      "epoch": 0.8136792452830188,
+      "grad_norm": 0.2395648709827474,
+      "learning_rate": 4.051987767584098e-05,
+      "loss": 0.5039,
+      "step": 690
+    },
+    {
+      "epoch": 0.8148584905660378,
+      "grad_norm": 0.2461091131478281,
+      "learning_rate": 4.049803407601573e-05,
+      "loss": 0.4915,
+      "step": 691
+    },
+    {
+      "epoch": 0.8160377358490566,
+      "grad_norm": 0.25716787549934317,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 0.5081,
+      "step": 692
+    },
+    {
+      "epoch": 0.8172169811320755,
+      "grad_norm": 0.23840878170091137,
+      "learning_rate": 4.0454346876365226e-05,
+      "loss": 0.5017,
+      "step": 693
+    },
+    {
+      "epoch": 0.8183962264150944,
+      "grad_norm": 0.23023004836087224,
+      "learning_rate": 4.043250327653997e-05,
+      "loss": 0.4741,
+      "step": 694
+    },
+    {
+      "epoch": 0.8195754716981132,
+      "grad_norm": 0.2359568601453405,
+      "learning_rate": 4.0410659676714724e-05,
+      "loss": 0.4859,
+      "step": 695
+    },
+    {
+      "epoch": 0.8207547169811321,
+      "grad_norm": 0.2506832590114902,
+      "learning_rate": 4.0388816076889476e-05,
+      "loss": 0.5135,
+      "step": 696
+    },
+    {
+      "epoch": 0.8219339622641509,
+      "grad_norm": 0.2112059448543871,
+      "learning_rate": 4.036697247706422e-05,
+      "loss": 0.5051,
+      "step": 697
+    },
+    {
+      "epoch": 0.8231132075471698,
+      "grad_norm": 0.24122637795774843,
+      "learning_rate": 4.0345128877238966e-05,
+      "loss": 0.4734,
+      "step": 698
+    },
+    {
+      "epoch": 0.8242924528301887,
+      "grad_norm": 0.23163007435280125,
+      "learning_rate": 4.032328527741372e-05,
+      "loss": 0.5116,
+      "step": 699
+    },
+    {
+      "epoch": 0.8254716981132075,
+      "grad_norm": 0.2333552177907508,
+      "learning_rate": 4.030144167758847e-05,
+      "loss": 0.4998,
+      "step": 700
+    },
+    {
+      "epoch": 0.8266509433962265,
+      "grad_norm": 0.21150599298363407,
+      "learning_rate": 4.027959807776322e-05,
+      "loss": 0.5076,
+      "step": 701
+    },
+    {
+      "epoch": 0.8278301886792453,
+      "grad_norm": 0.24767079096428044,
+      "learning_rate": 4.025775447793796e-05,
+      "loss": 0.4875,
+      "step": 702
+    },
+    {
+      "epoch": 0.8290094339622641,
+      "grad_norm": 0.22392750628143399,
+      "learning_rate": 4.023591087811271e-05,
+      "loss": 0.5062,
+      "step": 703
+    },
+    {
+      "epoch": 0.8301886792452831,
+      "grad_norm": 0.2354411402286094,
+      "learning_rate": 4.0214067278287465e-05,
+      "loss": 0.452,
+      "step": 704
+    },
+    {
+      "epoch": 0.8313679245283019,
+      "grad_norm": 0.23843389197680495,
+      "learning_rate": 4.019222367846222e-05,
+      "loss": 0.502,
+      "step": 705
+    },
+    {
+      "epoch": 0.8325471698113207,
+      "grad_norm": 0.22800434233805478,
+      "learning_rate": 4.017038007863696e-05,
+      "loss": 0.5228,
+      "step": 706
+    },
+    {
+      "epoch": 0.8337264150943396,
+      "grad_norm": 0.255928185516414,
+      "learning_rate": 4.014853647881171e-05,
+      "loss": 0.475,
+      "step": 707
+    },
+    {
+      "epoch": 0.8349056603773585,
+      "grad_norm": 0.23924994109356257,
+      "learning_rate": 4.012669287898646e-05,
+      "loss": 0.4816,
+      "step": 708
+    },
+    {
+      "epoch": 0.8360849056603774,
+      "grad_norm": 0.2907726453365134,
+      "learning_rate": 4.0104849279161205e-05,
+      "loss": 0.493,
+      "step": 709
+    },
+    {
+      "epoch": 0.8372641509433962,
+      "grad_norm": 0.23148519934700681,
+      "learning_rate": 4.008300567933596e-05,
+      "loss": 0.4976,
+      "step": 710
+    },
+    {
+      "epoch": 0.8384433962264151,
+      "grad_norm": 0.25978148641472215,
+      "learning_rate": 4.00611620795107e-05,
+      "loss": 0.5126,
+      "step": 711
+    },
+    {
+      "epoch": 0.839622641509434,
+      "grad_norm": 0.25132733145725283,
+      "learning_rate": 4.0039318479685455e-05,
+      "loss": 0.4734,
+      "step": 712
+    },
+    {
+      "epoch": 0.8408018867924528,
+      "grad_norm": 0.2689527905670351,
+      "learning_rate": 4.00174748798602e-05,
+      "loss": 0.49,
+      "step": 713
+    },
+    {
+      "epoch": 0.8419811320754716,
+      "grad_norm": 0.25254725086390944,
+      "learning_rate": 3.999563128003495e-05,
+      "loss": 0.5103,
+      "step": 714
+    },
+    {
+      "epoch": 0.8431603773584906,
+      "grad_norm": 0.2556013484710927,
+      "learning_rate": 3.9973787680209704e-05,
+      "loss": 0.4994,
+      "step": 715
+    },
+    {
+      "epoch": 0.8443396226415094,
+      "grad_norm": 0.23699059857989502,
+      "learning_rate": 3.995194408038445e-05,
+      "loss": 0.4924,
+      "step": 716
+    },
+    {
+      "epoch": 0.8455188679245284,
+      "grad_norm": 0.24040299806386425,
+      "learning_rate": 3.9930100480559195e-05,
+      "loss": 0.5084,
+      "step": 717
+    },
+    {
+      "epoch": 0.8466981132075472,
+      "grad_norm": 0.2743275760524044,
+      "learning_rate": 3.990825688073395e-05,
+      "loss": 0.4732,
+      "step": 718
+    },
+    {
+      "epoch": 0.847877358490566,
+      "grad_norm": 0.21960401735475218,
+      "learning_rate": 3.98864132809087e-05,
+      "loss": 0.4814,
+      "step": 719
+    },
+    {
+      "epoch": 0.8490566037735849,
+      "grad_norm": 0.2501074546454782,
+      "learning_rate": 3.9864569681083445e-05,
+      "loss": 0.4749,
+      "step": 720
+    },
+    {
+      "epoch": 0.8502358490566038,
+      "grad_norm": 0.21964211313465964,
+      "learning_rate": 3.984272608125819e-05,
+      "loss": 0.4893,
+      "step": 721
+    },
+    {
+      "epoch": 0.8514150943396226,
+      "grad_norm": 0.23292042782662897,
+      "learning_rate": 3.982088248143294e-05,
+      "loss": 0.5064,
+      "step": 722
+    },
+    {
+      "epoch": 0.8525943396226415,
+      "grad_norm": 0.2723640415908324,
+      "learning_rate": 3.9799038881607694e-05,
+      "loss": 0.5065,
+      "step": 723
+    },
+    {
+      "epoch": 0.8537735849056604,
+      "grad_norm": 0.23589095556601464,
+      "learning_rate": 3.977719528178244e-05,
+      "loss": 0.5217,
+      "step": 724
+    },
+    {
+      "epoch": 0.8549528301886793,
+      "grad_norm": 0.2573984068774592,
+      "learning_rate": 3.9755351681957185e-05,
+      "loss": 0.4861,
+      "step": 725
+    },
+    {
+      "epoch": 0.8561320754716981,
+      "grad_norm": 0.23890774758637937,
+      "learning_rate": 3.973350808213194e-05,
+      "loss": 0.4893,
+      "step": 726
+    },
+    {
+      "epoch": 0.8573113207547169,
+      "grad_norm": 0.25684000794203543,
+      "learning_rate": 3.971166448230669e-05,
+      "loss": 0.4966,
+      "step": 727
+    },
+    {
+      "epoch": 0.8584905660377359,
+      "grad_norm": 0.2709287164587681,
+      "learning_rate": 3.9689820882481434e-05,
+      "loss": 0.4899,
+      "step": 728
+    },
+    {
+      "epoch": 0.8596698113207547,
+      "grad_norm": 0.24159644634890484,
+      "learning_rate": 3.966797728265618e-05,
+      "loss": 0.4834,
+      "step": 729
+    },
+    {
+      "epoch": 0.8608490566037735,
+      "grad_norm": 0.2957173154453446,
+      "learning_rate": 3.964613368283093e-05,
+      "loss": 0.4657,
+      "step": 730
+    },
+    {
+      "epoch": 0.8620283018867925,
+      "grad_norm": 0.26474313536639654,
+      "learning_rate": 3.9624290083005684e-05,
+      "loss": 0.5232,
+      "step": 731
+    },
+    {
+      "epoch": 0.8632075471698113,
+      "grad_norm": 0.292615863635337,
+      "learning_rate": 3.960244648318043e-05,
+      "loss": 0.4808,
+      "step": 732
+    },
+    {
+      "epoch": 0.8643867924528302,
+      "grad_norm": 0.24881109473472302,
+      "learning_rate": 3.958060288335518e-05,
+      "loss": 0.4826,
+      "step": 733
+    },
+    {
+      "epoch": 0.8655660377358491,
+      "grad_norm": 0.2797164721094761,
+      "learning_rate": 3.9558759283529926e-05,
+      "loss": 0.4872,
+      "step": 734
+    },
+    {
+      "epoch": 0.8667452830188679,
+      "grad_norm": 0.27342903454955014,
+      "learning_rate": 3.953691568370468e-05,
+      "loss": 0.4997,
+      "step": 735
+    },
+    {
+      "epoch": 0.8679245283018868,
+      "grad_norm": 0.23304499264725798,
+      "learning_rate": 3.9515072083879424e-05,
+      "loss": 0.4922,
+      "step": 736
+    },
+    {
+      "epoch": 0.8691037735849056,
+      "grad_norm": 0.2639933703690393,
+      "learning_rate": 3.9493228484054176e-05,
+      "loss": 0.4838,
+      "step": 737
+    },
+    {
+      "epoch": 0.8702830188679245,
+      "grad_norm": 0.2561267186117552,
+      "learning_rate": 3.947138488422892e-05,
+      "loss": 0.5028,
+      "step": 738
+    },
+    {
+      "epoch": 0.8714622641509434,
+      "grad_norm": 0.2586659467639545,
+      "learning_rate": 3.944954128440367e-05,
+      "loss": 0.4978,
+      "step": 739
+    },
+    {
+      "epoch": 0.8726415094339622,
+      "grad_norm": 0.23430100397389453,
+      "learning_rate": 3.942769768457842e-05,
+      "loss": 0.4869,
+      "step": 740
+    },
+    {
+      "epoch": 0.8738207547169812,
+      "grad_norm": 0.31986881234540376,
+      "learning_rate": 3.940585408475317e-05,
+      "loss": 0.5121,
+      "step": 741
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.2816891669702893,
+      "learning_rate": 3.9384010484927916e-05,
+      "loss": 0.4871,
+      "step": 742
+    },
+    {
+      "epoch": 0.8761792452830188,
+      "grad_norm": 0.2930343475929812,
+      "learning_rate": 3.936216688510267e-05,
+      "loss": 0.4873,
+      "step": 743
+    },
+    {
+      "epoch": 0.8773584905660378,
+      "grad_norm": 0.3346623003391378,
+      "learning_rate": 3.934032328527741e-05,
+      "loss": 0.5271,
+      "step": 744
+    },
+    {
+      "epoch": 0.8785377358490566,
+      "grad_norm": 0.24357857116347278,
+      "learning_rate": 3.9318479685452165e-05,
+      "loss": 0.4699,
+      "step": 745
+    },
+    {
+      "epoch": 0.8797169811320755,
+      "grad_norm": 0.2767237904315916,
+      "learning_rate": 3.929663608562692e-05,
+      "loss": 0.515,
+      "step": 746
+    },
+    {
+      "epoch": 0.8808962264150944,
+      "grad_norm": 0.2638133200812286,
+      "learning_rate": 3.927479248580166e-05,
+      "loss": 0.4966,
+      "step": 747
+    },
+    {
+      "epoch": 0.8820754716981132,
+      "grad_norm": 0.24357425889682338,
+      "learning_rate": 3.925294888597641e-05,
+      "loss": 0.4996,
+      "step": 748
+    },
+    {
+      "epoch": 0.8832547169811321,
+      "grad_norm": 0.27209634194119214,
+      "learning_rate": 3.923110528615116e-05,
+      "loss": 0.4839,
+      "step": 749
+    },
+    {
+      "epoch": 0.8844339622641509,
+      "grad_norm": 0.2536227763175075,
+      "learning_rate": 3.920926168632591e-05,
+      "loss": 0.4932,
+      "step": 750
+    },
+    {
+      "epoch": 0.8856132075471698,
+      "grad_norm": 0.28076915025406957,
+      "learning_rate": 3.918741808650066e-05,
+      "loss": 0.4965,
+      "step": 751
+    },
+    {
+      "epoch": 0.8867924528301887,
+      "grad_norm": 0.24838585574886424,
+      "learning_rate": 3.91655744866754e-05,
+      "loss": 0.4922,
+      "step": 752
+    },
+    {
+      "epoch": 0.8879716981132075,
+      "grad_norm": 0.2537977265623713,
+      "learning_rate": 3.9143730886850155e-05,
+      "loss": 0.5187,
+      "step": 753
+    },
+    {
+      "epoch": 0.8891509433962265,
+      "grad_norm": 0.2533039847703217,
+      "learning_rate": 3.912188728702491e-05,
+      "loss": 0.5126,
+      "step": 754
+    },
+    {
+      "epoch": 0.8903301886792453,
+      "grad_norm": 0.26811017357126354,
+      "learning_rate": 3.910004368719965e-05,
+      "loss": 0.4896,
+      "step": 755
+    },
+    {
+      "epoch": 0.8915094339622641,
+      "grad_norm": 0.2587535913999112,
+      "learning_rate": 3.90782000873744e-05,
+      "loss": 0.4906,
+      "step": 756
+    },
+    {
+      "epoch": 0.8926886792452831,
+      "grad_norm": 0.2559039885592712,
+      "learning_rate": 3.905635648754915e-05,
+      "loss": 0.5067,
+      "step": 757
+    },
+    {
+      "epoch": 0.8938679245283019,
+      "grad_norm": 0.2501998046332605,
+      "learning_rate": 3.90345128877239e-05,
+      "loss": 0.4723,
+      "step": 758
+    },
+    {
+      "epoch": 0.8950471698113207,
+      "grad_norm": 0.23530968082338627,
+      "learning_rate": 3.901266928789865e-05,
+      "loss": 0.4697,
+      "step": 759
+    },
+    {
+      "epoch": 0.8962264150943396,
+      "grad_norm": 0.23672776337639312,
+      "learning_rate": 3.89908256880734e-05,
+      "loss": 0.5049,
+      "step": 760
+    },
+    {
+      "epoch": 0.8974056603773585,
+      "grad_norm": 0.22427560309793967,
+      "learning_rate": 3.8968982088248144e-05,
+      "loss": 0.4836,
+      "step": 761
+    },
+    {
+      "epoch": 0.8985849056603774,
+      "grad_norm": 0.26104860156620097,
+      "learning_rate": 3.8947138488422896e-05,
+      "loss": 0.4978,
+      "step": 762
+    },
+    {
+      "epoch": 0.8997641509433962,
+      "grad_norm": 0.23528295385165157,
+      "learning_rate": 3.892529488859764e-05,
+      "loss": 0.4685,
+      "step": 763
+    },
+    {
+      "epoch": 0.9009433962264151,
+      "grad_norm": 0.2821317503726348,
+      "learning_rate": 3.8903451288772394e-05,
+      "loss": 0.5152,
+      "step": 764
+    },
+    {
+      "epoch": 0.902122641509434,
+      "grad_norm": 0.22978916288586523,
+      "learning_rate": 3.888160768894714e-05,
+      "loss": 0.4725,
+      "step": 765
+    },
+    {
+      "epoch": 0.9033018867924528,
+      "grad_norm": 0.2834160581606795,
+      "learning_rate": 3.885976408912189e-05,
+      "loss": 0.4864,
+      "step": 766
+    },
+    {
+      "epoch": 0.9044811320754716,
+      "grad_norm": 0.28835562544152565,
+      "learning_rate": 3.8837920489296637e-05,
+      "loss": 0.4991,
+      "step": 767
+    },
+    {
+      "epoch": 0.9056603773584906,
+      "grad_norm": 0.2574647906018368,
+      "learning_rate": 3.881607688947139e-05,
+      "loss": 0.499,
+      "step": 768
+    },
+    {
+      "epoch": 0.9068396226415094,
+      "grad_norm": 0.23015210376619902,
+      "learning_rate": 3.8794233289646134e-05,
+      "loss": 0.4779,
+      "step": 769
+    },
+    {
+      "epoch": 0.9080188679245284,
+      "grad_norm": 0.2809108173601727,
+      "learning_rate": 3.877238968982088e-05,
+      "loss": 0.4884,
+      "step": 770
+    },
+    {
+      "epoch": 0.9091981132075472,
+      "grad_norm": 0.2541663039451844,
+      "learning_rate": 3.875054608999563e-05,
+      "loss": 0.4994,
+      "step": 771
+    },
+    {
+      "epoch": 0.910377358490566,
+      "grad_norm": 0.24731906233418693,
+      "learning_rate": 3.8728702490170383e-05,
+      "loss": 0.4887,
+      "step": 772
+    },
+    {
+      "epoch": 0.9115566037735849,
+      "grad_norm": 0.2549259707246151,
+      "learning_rate": 3.8706858890345136e-05,
+      "loss": 0.4802,
+      "step": 773
+    },
+    {
+      "epoch": 0.9127358490566038,
+      "grad_norm": 0.26111467742959227,
+      "learning_rate": 3.8685015290519874e-05,
+      "loss": 0.4879,
+      "step": 774
+    },
+    {
+      "epoch": 0.9139150943396226,
+      "grad_norm": 0.2275661905911339,
+      "learning_rate": 3.8663171690694626e-05,
+      "loss": 0.4611,
+      "step": 775
+    },
+    {
+      "epoch": 0.9150943396226415,
+      "grad_norm": 0.2552214156223457,
+      "learning_rate": 3.864132809086938e-05,
+      "loss": 0.4806,
+      "step": 776
+    },
+    {
+      "epoch": 0.9162735849056604,
+      "grad_norm": 0.2345640537624651,
+      "learning_rate": 3.861948449104413e-05,
+      "loss": 0.4859,
+      "step": 777
+    },
+    {
+      "epoch": 0.9174528301886793,
+      "grad_norm": 0.252875250933215,
+      "learning_rate": 3.8597640891218876e-05,
+      "loss": 0.4827,
+      "step": 778
+    },
+    {
+      "epoch": 0.9186320754716981,
+      "grad_norm": 0.2647435618350338,
+      "learning_rate": 3.857579729139362e-05,
+      "loss": 0.5163,
+      "step": 779
+    },
+    {
+      "epoch": 0.9198113207547169,
+      "grad_norm": 0.25079126477025565,
+      "learning_rate": 3.855395369156837e-05,
+      "loss": 0.4937,
+      "step": 780
+    },
+    {
+      "epoch": 0.9209905660377359,
+      "grad_norm": 0.27145546356604167,
+      "learning_rate": 3.8532110091743125e-05,
+      "loss": 0.4845,
+      "step": 781
+    },
+    {
+      "epoch": 0.9221698113207547,
+      "grad_norm": 0.2950056670844647,
+      "learning_rate": 3.851026649191787e-05,
+      "loss": 0.5013,
+      "step": 782
+    },
+    {
+      "epoch": 0.9233490566037735,
+      "grad_norm": 0.2861592904495696,
+      "learning_rate": 3.8488422892092616e-05,
+      "loss": 0.5176,
+      "step": 783
+    },
+    {
+      "epoch": 0.9245283018867925,
+      "grad_norm": 0.2833055591131783,
+      "learning_rate": 3.846657929226737e-05,
+      "loss": 0.4653,
+      "step": 784
+    },
+    {
+      "epoch": 0.9257075471698113,
+      "grad_norm": 0.26147403072669234,
+      "learning_rate": 3.844473569244212e-05,
+      "loss": 0.4753,
+      "step": 785
+    },
+    {
+      "epoch": 0.9268867924528302,
+      "grad_norm": 0.2578440998293417,
+      "learning_rate": 3.8422892092616865e-05,
+      "loss": 0.4808,
+      "step": 786
+    },
+    {
+      "epoch": 0.9280660377358491,
+      "grad_norm": 0.2488196717876031,
+      "learning_rate": 3.840104849279161e-05,
+      "loss": 0.4842,
+      "step": 787
+    },
+    {
+      "epoch": 0.9292452830188679,
+      "grad_norm": 0.2999521532837809,
+      "learning_rate": 3.837920489296636e-05,
+      "loss": 0.4952,
+      "step": 788
+    },
+    {
+      "epoch": 0.9304245283018868,
+      "grad_norm": 0.23124078170786094,
+      "learning_rate": 3.8357361293141115e-05,
+      "loss": 0.4897,
+      "step": 789
+    },
+    {
+      "epoch": 0.9316037735849056,
+      "grad_norm": 0.2689785119212323,
+      "learning_rate": 3.833551769331586e-05,
+      "loss": 0.4746,
+      "step": 790
+    },
+    {
+      "epoch": 0.9327830188679245,
+      "grad_norm": 0.26305138591095656,
+      "learning_rate": 3.831367409349061e-05,
+      "loss": 0.4924,
+      "step": 791
+    },
+    {
+      "epoch": 0.9339622641509434,
+      "grad_norm": 0.25333618109375183,
+      "learning_rate": 3.829183049366536e-05,
+      "loss": 0.528,
+      "step": 792
+    },
+    {
+      "epoch": 0.9351415094339622,
+      "grad_norm": 0.2594508612204989,
+      "learning_rate": 3.82699868938401e-05,
+      "loss": 0.4947,
+      "step": 793
+    },
+    {
+      "epoch": 0.9363207547169812,
+      "grad_norm": 0.2580258504686285,
+      "learning_rate": 3.8248143294014855e-05,
+      "loss": 0.5243,
+      "step": 794
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.2923604820824771,
+      "learning_rate": 3.822629969418961e-05,
+      "loss": 0.4987,
+      "step": 795
+    },
+    {
+      "epoch": 0.9386792452830188,
+      "grad_norm": 0.248837794989193,
+      "learning_rate": 3.820445609436435e-05,
+      "loss": 0.4835,
+      "step": 796
+    },
+    {
+      "epoch": 0.9398584905660378,
+      "grad_norm": 0.26552683587557185,
+      "learning_rate": 3.81826124945391e-05,
+      "loss": 0.4685,
+      "step": 797
+    },
+    {
+      "epoch": 0.9410377358490566,
+      "grad_norm": 0.23136526078746436,
+      "learning_rate": 3.816076889471385e-05,
+      "loss": 0.4744,
+      "step": 798
+    },
+    {
+      "epoch": 0.9422169811320755,
+      "grad_norm": 0.247164442821578,
+      "learning_rate": 3.81389252948886e-05,
+      "loss": 0.5079,
+      "step": 799
+    },
+    {
+      "epoch": 0.9433962264150944,
+      "grad_norm": 0.2478073328007951,
+      "learning_rate": 3.811708169506335e-05,
+      "loss": 0.5048,
+      "step": 800
+    },
+    {
+      "epoch": 0.9445754716981132,
+      "grad_norm": 0.26401264950152825,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 0.508,
+      "step": 801
+    },
+    {
+      "epoch": 0.9457547169811321,
+      "grad_norm": 0.24603615816510574,
+      "learning_rate": 3.8073394495412844e-05,
+      "loss": 0.4809,
+      "step": 802
+    },
+    {
+      "epoch": 0.9469339622641509,
+      "grad_norm": 0.21475924354162468,
+      "learning_rate": 3.8051550895587596e-05,
+      "loss": 0.4591,
+      "step": 803
+    },
+    {
+      "epoch": 0.9481132075471698,
+      "grad_norm": 0.23746891094153094,
+      "learning_rate": 3.802970729576235e-05,
+      "loss": 0.484,
+      "step": 804
+    },
+    {
+      "epoch": 0.9492924528301887,
+      "grad_norm": 0.2580025876096533,
+      "learning_rate": 3.8007863695937094e-05,
+      "loss": 0.4857,
+      "step": 805
+    },
+    {
+      "epoch": 0.9504716981132075,
+      "grad_norm": 0.2446969943374308,
+      "learning_rate": 3.798602009611184e-05,
+      "loss": 0.4719,
+      "step": 806
+    },
+    {
+      "epoch": 0.9516509433962265,
+      "grad_norm": 0.25915299023286803,
+      "learning_rate": 3.796417649628659e-05,
+      "loss": 0.4844,
+      "step": 807
+    },
+    {
+      "epoch": 0.9528301886792453,
+      "grad_norm": 0.22687920230562006,
+      "learning_rate": 3.794233289646134e-05,
+      "loss": 0.4992,
+      "step": 808
+    },
+    {
+      "epoch": 0.9540094339622641,
+      "grad_norm": 0.27690417429167696,
+      "learning_rate": 3.792048929663609e-05,
+      "loss": 0.5065,
+      "step": 809
+    },
+    {
+      "epoch": 0.9551886792452831,
+      "grad_norm": 0.23116489931290019,
+      "learning_rate": 3.7898645696810834e-05,
+      "loss": 0.5003,
+      "step": 810
+    },
+    {
+      "epoch": 0.9563679245283019,
+      "grad_norm": 0.2513480787602484,
+      "learning_rate": 3.7876802096985586e-05,
+      "loss": 0.4941,
+      "step": 811
+    },
+    {
+      "epoch": 0.9575471698113207,
+      "grad_norm": 0.25093559298435386,
+      "learning_rate": 3.785495849716034e-05,
+      "loss": 0.4794,
+      "step": 812
+    },
+    {
+      "epoch": 0.9587264150943396,
+      "grad_norm": 0.23581304230248934,
+      "learning_rate": 3.783311489733508e-05,
+      "loss": 0.4829,
+      "step": 813
+    },
+    {
+      "epoch": 0.9599056603773585,
+      "grad_norm": 0.24138445527594582,
+      "learning_rate": 3.781127129750983e-05,
+      "loss": 0.4893,
+      "step": 814
+    },
+    {
+      "epoch": 0.9610849056603774,
+      "grad_norm": 0.25985422402835623,
+      "learning_rate": 3.778942769768458e-05,
+      "loss": 0.5397,
+      "step": 815
+    },
+    {
+      "epoch": 0.9622641509433962,
+      "grad_norm": 0.2963855768420568,
+      "learning_rate": 3.7767584097859326e-05,
+      "loss": 0.5382,
+      "step": 816
+    },
+    {
+      "epoch": 0.9634433962264151,
+      "grad_norm": 0.24439449878726668,
+      "learning_rate": 3.774574049803408e-05,
+      "loss": 0.4984,
+      "step": 817
+    },
+    {
+      "epoch": 0.964622641509434,
+      "grad_norm": 0.26374035888641406,
+      "learning_rate": 3.772389689820883e-05,
+      "loss": 0.5053,
+      "step": 818
+    },
+    {
+      "epoch": 0.9658018867924528,
+      "grad_norm": 0.22397621341693386,
+      "learning_rate": 3.7702053298383576e-05,
+      "loss": 0.4757,
+      "step": 819
+    },
+    {
+      "epoch": 0.9669811320754716,
+      "grad_norm": 0.25921178518254484,
+      "learning_rate": 3.768020969855832e-05,
+      "loss": 0.4654,
+      "step": 820
+    },
+    {
+      "epoch": 0.9681603773584906,
+      "grad_norm": 0.22183776706465486,
+      "learning_rate": 3.765836609873307e-05,
+      "loss": 0.4785,
+      "step": 821
+    },
+    {
+      "epoch": 0.9693396226415094,
+      "grad_norm": 0.24643896137021398,
+      "learning_rate": 3.7636522498907825e-05,
+      "loss": 0.5108,
+      "step": 822
+    },
+    {
+      "epoch": 0.9705188679245284,
+      "grad_norm": 0.2371452497797819,
+      "learning_rate": 3.761467889908257e-05,
+      "loss": 0.4885,
+      "step": 823
+    },
+    {
+      "epoch": 0.9716981132075472,
+      "grad_norm": 0.22319896207524803,
+      "learning_rate": 3.7592835299257316e-05,
+      "loss": 0.4758,
+      "step": 824
+    },
+    {
+      "epoch": 0.972877358490566,
+      "grad_norm": 0.24671355037016787,
+      "learning_rate": 3.757099169943207e-05,
+      "loss": 0.4869,
+      "step": 825
+    },
+    {
+      "epoch": 0.9740566037735849,
+      "grad_norm": 0.25496441395400243,
+      "learning_rate": 3.754914809960682e-05,
+      "loss": 0.5007,
+      "step": 826
+    },
+    {
+      "epoch": 0.9752358490566038,
+      "grad_norm": 0.2306269342411826,
+      "learning_rate": 3.7527304499781565e-05,
+      "loss": 0.4985,
+      "step": 827
+    },
+    {
+      "epoch": 0.9764150943396226,
+      "grad_norm": 0.24757327672519336,
+      "learning_rate": 3.750546089995631e-05,
+      "loss": 0.459,
+      "step": 828
+    },
+    {
+      "epoch": 0.9775943396226415,
+      "grad_norm": 0.24989441763400477,
+      "learning_rate": 3.748361730013106e-05,
+      "loss": 0.4811,
+      "step": 829
+    },
+    {
+      "epoch": 0.9787735849056604,
+      "grad_norm": 0.25396490758730084,
+      "learning_rate": 3.7461773700305815e-05,
+      "loss": 0.4597,
+      "step": 830
+    },
+    {
+      "epoch": 0.9799528301886793,
+      "grad_norm": 0.2924779682952035,
+      "learning_rate": 3.743993010048057e-05,
+      "loss": 0.521,
+      "step": 831
+    },
+    {
+      "epoch": 0.9811320754716981,
+      "grad_norm": 0.2629982679747208,
+      "learning_rate": 3.7418086500655305e-05,
+      "loss": 0.5021,
+      "step": 832
+    },
+    {
+      "epoch": 0.9823113207547169,
+      "grad_norm": 0.25886094717112285,
+      "learning_rate": 3.739624290083006e-05,
+      "loss": 0.4963,
+      "step": 833
+    },
+    {
+      "epoch": 0.9834905660377359,
+      "grad_norm": 0.2604412729325055,
+      "learning_rate": 3.737439930100481e-05,
+      "loss": 0.4834,
+      "step": 834
+    },
+    {
+      "epoch": 0.9846698113207547,
+      "grad_norm": 0.2425614744473973,
+      "learning_rate": 3.7352555701179555e-05,
+      "loss": 0.4889,
+      "step": 835
+    },
+    {
+      "epoch": 0.9858490566037735,
+      "grad_norm": 0.23294222743435747,
+      "learning_rate": 3.733071210135431e-05,
+      "loss": 0.4709,
+      "step": 836
+    },
+    {
+      "epoch": 0.9870283018867925,
+      "grad_norm": 0.22780763228682688,
+      "learning_rate": 3.730886850152905e-05,
+      "loss": 0.4617,
+      "step": 837
+    },
+    {
+      "epoch": 0.9882075471698113,
+      "grad_norm": 0.22476812851651537,
+      "learning_rate": 3.7287024901703804e-05,
+      "loss": 0.5181,
+      "step": 838
+    },
+    {
+      "epoch": 0.9893867924528302,
+      "grad_norm": 0.21557080362917241,
+      "learning_rate": 3.726518130187855e-05,
+      "loss": 0.4835,
+      "step": 839
+    },
+    {
+      "epoch": 0.9905660377358491,
+      "grad_norm": 0.21721898282705407,
+      "learning_rate": 3.72433377020533e-05,
+      "loss": 0.4942,
+      "step": 840
+    },
+    {
+      "epoch": 0.9917452830188679,
+      "grad_norm": 0.22391834351966772,
+      "learning_rate": 3.722149410222805e-05,
+      "loss": 0.4688,
+      "step": 841
+    },
+    {
+      "epoch": 0.9929245283018868,
+      "grad_norm": 0.2262369199104073,
+      "learning_rate": 3.71996505024028e-05,
+      "loss": 0.4798,
+      "step": 842
+    },
+    {
+      "epoch": 0.9941037735849056,
+      "grad_norm": 0.2373804320347337,
+      "learning_rate": 3.7177806902577544e-05,
+      "loss": 0.5086,
+      "step": 843
+    },
+    {
+      "epoch": 0.9952830188679245,
+      "grad_norm": 0.2301720824922742,
+      "learning_rate": 3.7155963302752296e-05,
+      "loss": 0.4545,
+      "step": 844
+    },
+    {
+      "epoch": 0.9964622641509434,
+      "grad_norm": 0.2211609687911312,
+      "learning_rate": 3.713411970292704e-05,
+      "loss": 0.4806,
+      "step": 845
+    },
+    {
+      "epoch": 0.9976415094339622,
+      "grad_norm": 0.2598895747375138,
+      "learning_rate": 3.7112276103101794e-05,
+      "loss": 0.4858,
+      "step": 846
+    },
+    {
+      "epoch": 0.9988207547169812,
+      "grad_norm": 0.21208021565647933,
+      "learning_rate": 3.709043250327654e-05,
+      "loss": 0.4869,
+      "step": 847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25975653068892846,
+      "learning_rate": 3.706858890345129e-05,
+      "loss": 0.4809,
+      "step": 848
+    },
+    {
+      "epoch": 1.0011792452830188,
+      "grad_norm": 0.3026979757082602,
+      "learning_rate": 3.704674530362604e-05,
+      "loss": 0.4021,
+      "step": 849
+    },
+    {
+      "epoch": 1.0023584905660377,
+      "grad_norm": 0.2699251824411265,
+      "learning_rate": 3.702490170380079e-05,
+      "loss": 0.4401,
+      "step": 850
+    },
+    {
+      "epoch": 1.0035377358490567,
+      "grad_norm": 0.23377138170904477,
+      "learning_rate": 3.7003058103975534e-05,
+      "loss": 0.4204,
+      "step": 851
+    },
+    {
+      "epoch": 1.0047169811320755,
+      "grad_norm": 0.3089462259433967,
+      "learning_rate": 3.6981214504150286e-05,
+      "loss": 0.4316,
+      "step": 852
+    },
+    {
+      "epoch": 1.0058962264150944,
+      "grad_norm": 0.2576318836757543,
+      "learning_rate": 3.695937090432504e-05,
+      "loss": 0.4343,
+      "step": 853
+    },
+    {
+      "epoch": 1.0070754716981132,
+      "grad_norm": 0.3119264952258992,
+      "learning_rate": 3.693752730449978e-05,
+      "loss": 0.4215,
+      "step": 854
+    },
+    {
+      "epoch": 1.008254716981132,
+      "grad_norm": 0.2669968709036758,
+      "learning_rate": 3.691568370467453e-05,
+      "loss": 0.4133,
+      "step": 855
+    },
+    {
+      "epoch": 1.009433962264151,
+      "grad_norm": 0.24340569051837158,
+      "learning_rate": 3.689384010484928e-05,
+      "loss": 0.4303,
+      "step": 856
+    },
+    {
+      "epoch": 1.0106132075471699,
+      "grad_norm": 0.2611930034927872,
+      "learning_rate": 3.687199650502403e-05,
+      "loss": 0.4217,
+      "step": 857
+    },
+    {
+      "epoch": 1.0117924528301887,
+      "grad_norm": 0.2255198088294251,
+      "learning_rate": 3.685015290519878e-05,
+      "loss": 0.4056,
+      "step": 858
+    },
+    {
+      "epoch": 1.0129716981132075,
+      "grad_norm": 0.25897259917068055,
+      "learning_rate": 3.682830930537352e-05,
+      "loss": 0.4127,
+      "step": 859
+    },
+    {
+      "epoch": 1.0141509433962264,
+      "grad_norm": 0.24427211727192194,
+      "learning_rate": 3.6806465705548275e-05,
+      "loss": 0.439,
+      "step": 860
+    },
+    {
+      "epoch": 1.0153301886792452,
+      "grad_norm": 0.2579713071670691,
+      "learning_rate": 3.678462210572303e-05,
+      "loss": 0.4444,
+      "step": 861
+    },
+    {
+      "epoch": 1.0165094339622642,
+      "grad_norm": 0.24318116965094672,
+      "learning_rate": 3.676277850589777e-05,
+      "loss": 0.4223,
+      "step": 862
+    },
+    {
+      "epoch": 1.017688679245283,
+      "grad_norm": 0.24661995582806065,
+      "learning_rate": 3.6740934906072525e-05,
+      "loss": 0.4254,
+      "step": 863
+    },
+    {
+      "epoch": 1.0188679245283019,
+      "grad_norm": 0.24940149591802208,
+      "learning_rate": 3.671909130624727e-05,
+      "loss": 0.4155,
+      "step": 864
+    },
+    {
+      "epoch": 1.0200471698113207,
+      "grad_norm": 0.25164537232647,
+      "learning_rate": 3.669724770642202e-05,
+      "loss": 0.4034,
+      "step": 865
+    },
+    {
+      "epoch": 1.0212264150943395,
+      "grad_norm": 0.24440435087433665,
+      "learning_rate": 3.667540410659677e-05,
+      "loss": 0.4099,
+      "step": 866
+    },
+    {
+      "epoch": 1.0224056603773586,
+      "grad_norm": 0.25051144341393644,
+      "learning_rate": 3.665356050677152e-05,
+      "loss": 0.4375,
+      "step": 867
+    },
+    {
+      "epoch": 1.0235849056603774,
+      "grad_norm": 0.2575087856074404,
+      "learning_rate": 3.6631716906946265e-05,
+      "loss": 0.4375,
+      "step": 868
+    },
+    {
+      "epoch": 1.0247641509433962,
+      "grad_norm": 0.24995640529144472,
+      "learning_rate": 3.660987330712102e-05,
+      "loss": 0.4073,
+      "step": 869
+    },
+    {
+      "epoch": 1.025943396226415,
+      "grad_norm": 0.25264088812882957,
+      "learning_rate": 3.658802970729576e-05,
+      "loss": 0.4096,
+      "step": 870
+    },
+    {
+      "epoch": 1.0271226415094339,
+      "grad_norm": 0.2545009240947338,
+      "learning_rate": 3.6566186107470514e-05,
+      "loss": 0.4312,
+      "step": 871
+    },
+    {
+      "epoch": 1.028301886792453,
+      "grad_norm": 0.21863879204723632,
+      "learning_rate": 3.654434250764526e-05,
+      "loss": 0.4107,
+      "step": 872
+    },
+    {
+      "epoch": 1.0294811320754718,
+      "grad_norm": 0.25112720883355705,
+      "learning_rate": 3.652249890782001e-05,
+      "loss": 0.4154,
+      "step": 873
+    },
+    {
+      "epoch": 1.0306603773584906,
+      "grad_norm": 0.2771557721051068,
+      "learning_rate": 3.650065530799476e-05,
+      "loss": 0.4166,
+      "step": 874
+    },
+    {
+      "epoch": 1.0318396226415094,
+      "grad_norm": 0.23925929498258777,
+      "learning_rate": 3.647881170816951e-05,
+      "loss": 0.3921,
+      "step": 875
+    },
+    {
+      "epoch": 1.0330188679245282,
+      "grad_norm": 0.23562800658188263,
+      "learning_rate": 3.645696810834426e-05,
+      "loss": 0.4013,
+      "step": 876
+    },
+    {
+      "epoch": 1.034198113207547,
+      "grad_norm": 0.2643051310130577,
+      "learning_rate": 3.6435124508519e-05,
+      "loss": 0.4222,
+      "step": 877
+    },
+    {
+      "epoch": 1.0353773584905661,
+      "grad_norm": 0.24310909907941217,
+      "learning_rate": 3.641328090869375e-05,
+      "loss": 0.3753,
+      "step": 878
+    },
+    {
+      "epoch": 1.036556603773585,
+      "grad_norm": 0.2574990717037648,
+      "learning_rate": 3.6391437308868504e-05,
+      "loss": 0.4264,
+      "step": 879
+    },
+    {
+      "epoch": 1.0377358490566038,
+      "grad_norm": 0.2621292344688516,
+      "learning_rate": 3.6369593709043256e-05,
+      "loss": 0.413,
+      "step": 880
+    },
+    {
+      "epoch": 1.0389150943396226,
+      "grad_norm": 0.19373142974289476,
+      "learning_rate": 3.6347750109218e-05,
+      "loss": 0.3982,
+      "step": 881
+    },
+    {
+      "epoch": 1.0400943396226414,
+      "grad_norm": 0.22847860034154588,
+      "learning_rate": 3.632590650939275e-05,
+      "loss": 0.4079,
+      "step": 882
+    },
+    {
+      "epoch": 1.0412735849056605,
+      "grad_norm": 0.21541355895908165,
+      "learning_rate": 3.63040629095675e-05,
+      "loss": 0.4173,
+      "step": 883
+    },
+    {
+      "epoch": 1.0424528301886793,
+      "grad_norm": 0.2172794287564027,
+      "learning_rate": 3.628221930974225e-05,
+      "loss": 0.3952,
+      "step": 884
+    },
+    {
+      "epoch": 1.0436320754716981,
+      "grad_norm": 0.2272158415255476,
+      "learning_rate": 3.6260375709916996e-05,
+      "loss": 0.4172,
+      "step": 885
+    },
+    {
+      "epoch": 1.044811320754717,
+      "grad_norm": 0.20265114903288872,
+      "learning_rate": 3.623853211009174e-05,
+      "loss": 0.3991,
+      "step": 886
+    },
+    {
+      "epoch": 1.0459905660377358,
+      "grad_norm": 0.244614566811056,
+      "learning_rate": 3.6216688510266494e-05,
+      "loss": 0.4232,
+      "step": 887
+    },
+    {
+      "epoch": 1.0471698113207548,
+      "grad_norm": 0.22050306443216294,
+      "learning_rate": 3.6194844910441246e-05,
+      "loss": 0.4194,
+      "step": 888
+    },
+    {
+      "epoch": 1.0483490566037736,
+      "grad_norm": 0.22791074158601177,
+      "learning_rate": 3.617300131061599e-05,
+      "loss": 0.3977,
+      "step": 889
+    },
+    {
+      "epoch": 1.0495283018867925,
+      "grad_norm": 0.2191426097867157,
+      "learning_rate": 3.6151157710790736e-05,
+      "loss": 0.4221,
+      "step": 890
+    },
+    {
+      "epoch": 1.0507075471698113,
+      "grad_norm": 0.21260779349431103,
+      "learning_rate": 3.612931411096549e-05,
+      "loss": 0.4148,
+      "step": 891
+    },
+    {
+      "epoch": 1.0518867924528301,
+      "grad_norm": 0.2435559846841907,
+      "learning_rate": 3.610747051114024e-05,
+      "loss": 0.4021,
+      "step": 892
+    },
+    {
+      "epoch": 1.053066037735849,
+      "grad_norm": 0.20770456272020502,
+      "learning_rate": 3.6085626911314986e-05,
+      "loss": 0.4122,
+      "step": 893
+    },
+    {
+      "epoch": 1.054245283018868,
+      "grad_norm": 0.2544627616184649,
+      "learning_rate": 3.606378331148974e-05,
+      "loss": 0.4261,
+      "step": 894
+    },
+    {
+      "epoch": 1.0554245283018868,
+      "grad_norm": 0.2523565400437521,
+      "learning_rate": 3.604193971166448e-05,
+      "loss": 0.4229,
+      "step": 895
+    },
+    {
+      "epoch": 1.0566037735849056,
+      "grad_norm": 0.22556627091502435,
+      "learning_rate": 3.6020096111839235e-05,
+      "loss": 0.4118,
+      "step": 896
+    },
+    {
+      "epoch": 1.0577830188679245,
+      "grad_norm": 0.2480279259764831,
+      "learning_rate": 3.599825251201398e-05,
+      "loss": 0.4372,
+      "step": 897
+    },
+    {
+      "epoch": 1.0589622641509433,
+      "grad_norm": 0.2157222481483112,
+      "learning_rate": 3.597640891218873e-05,
+      "loss": 0.4026,
+      "step": 898
+    },
+    {
+      "epoch": 1.0601415094339623,
+      "grad_norm": 0.22554034367975878,
+      "learning_rate": 3.595456531236348e-05,
+      "loss": 0.4232,
+      "step": 899
+    },
+    {
+      "epoch": 1.0613207547169812,
+      "grad_norm": 0.21222093580269316,
+      "learning_rate": 3.593272171253822e-05,
+      "loss": 0.4003,
+      "step": 900
+    },
+    {
+      "epoch": 1.0625,
+      "grad_norm": 0.22375933128301076,
+      "learning_rate": 3.5910878112712975e-05,
+      "loss": 0.414,
+      "step": 901
+    },
+    {
+      "epoch": 1.0636792452830188,
+      "grad_norm": 0.23011539972895415,
+      "learning_rate": 3.588903451288773e-05,
+      "loss": 0.4119,
+      "step": 902
+    },
+    {
+      "epoch": 1.0648584905660377,
+      "grad_norm": 0.2360743692569845,
+      "learning_rate": 3.586719091306248e-05,
+      "loss": 0.431,
+      "step": 903
+    },
+    {
+      "epoch": 1.0660377358490567,
+      "grad_norm": 0.2483450453829572,
+      "learning_rate": 3.584534731323722e-05,
+      "loss": 0.4103,
+      "step": 904
+    },
+    {
+      "epoch": 1.0672169811320755,
+      "grad_norm": 0.22633069072054823,
+      "learning_rate": 3.582350371341197e-05,
+      "loss": 0.4403,
+      "step": 905
+    },
+    {
+      "epoch": 1.0683962264150944,
+      "grad_norm": 0.24047737148648296,
+      "learning_rate": 3.580166011358672e-05,
+      "loss": 0.4157,
+      "step": 906
+    },
+    {
+      "epoch": 1.0695754716981132,
+      "grad_norm": 0.2344821912791983,
+      "learning_rate": 3.5779816513761474e-05,
+      "loss": 0.3998,
+      "step": 907
+    },
+    {
+      "epoch": 1.070754716981132,
+      "grad_norm": 0.22910003517080066,
+      "learning_rate": 3.575797291393622e-05,
+      "loss": 0.3945,
+      "step": 908
+    },
+    {
+      "epoch": 1.071933962264151,
+      "grad_norm": 0.21769262056461972,
+      "learning_rate": 3.5736129314110965e-05,
+      "loss": 0.4217,
+      "step": 909
+    },
+    {
+      "epoch": 1.0731132075471699,
+      "grad_norm": 0.23348756064826967,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.4063,
+      "step": 910
+    },
+    {
+      "epoch": 1.0742924528301887,
+      "grad_norm": 0.2428212314398451,
+      "learning_rate": 3.569244211446047e-05,
+      "loss": 0.3984,
+      "step": 911
+    },
+    {
+      "epoch": 1.0754716981132075,
+      "grad_norm": 0.24696321328173076,
+      "learning_rate": 3.5670598514635214e-05,
+      "loss": 0.4185,
+      "step": 912
+    },
+    {
+      "epoch": 1.0766509433962264,
+      "grad_norm": 0.22630415658402947,
+      "learning_rate": 3.564875491480996e-05,
+      "loss": 0.426,
+      "step": 913
+    },
+    {
+      "epoch": 1.0778301886792452,
+      "grad_norm": 0.2200096333782469,
+      "learning_rate": 3.562691131498471e-05,
+      "loss": 0.4062,
+      "step": 914
+    },
+    {
+      "epoch": 1.0790094339622642,
+      "grad_norm": 0.22044071806730298,
+      "learning_rate": 3.5605067715159464e-05,
+      "loss": 0.4415,
+      "step": 915
+    },
+    {
+      "epoch": 1.080188679245283,
+      "grad_norm": 0.24922470896384297,
+      "learning_rate": 3.558322411533421e-05,
+      "loss": 0.4041,
+      "step": 916
+    },
+    {
+      "epoch": 1.0813679245283019,
+      "grad_norm": 0.22806225985386658,
+      "learning_rate": 3.5561380515508954e-05,
+      "loss": 0.4075,
+      "step": 917
+    },
+    {
+      "epoch": 1.0825471698113207,
+      "grad_norm": 0.20010269461233385,
+      "learning_rate": 3.5539536915683707e-05,
+      "loss": 0.4081,
+      "step": 918
+    },
+    {
+      "epoch": 1.0837264150943395,
+      "grad_norm": 0.22381084941120963,
+      "learning_rate": 3.551769331585846e-05,
+      "loss": 0.4104,
+      "step": 919
+    },
+    {
+      "epoch": 1.0849056603773586,
+      "grad_norm": 0.21896245288115604,
+      "learning_rate": 3.5495849716033204e-05,
+      "loss": 0.388,
+      "step": 920
+    },
+    {
+      "epoch": 1.0860849056603774,
+      "grad_norm": 0.19639279373817617,
+      "learning_rate": 3.5474006116207956e-05,
+      "loss": 0.397,
+      "step": 921
+    },
+    {
+      "epoch": 1.0872641509433962,
+      "grad_norm": 0.24409749631657615,
+      "learning_rate": 3.54521625163827e-05,
+      "loss": 0.4145,
+      "step": 922
+    },
+    {
+      "epoch": 1.088443396226415,
+      "grad_norm": 0.24194737372402503,
+      "learning_rate": 3.543031891655745e-05,
+      "loss": 0.4331,
+      "step": 923
+    },
+    {
+      "epoch": 1.0896226415094339,
+      "grad_norm": 0.22136160211486697,
+      "learning_rate": 3.54084753167322e-05,
+      "loss": 0.4186,
+      "step": 924
+    },
+    {
+      "epoch": 1.0908018867924527,
+      "grad_norm": 0.22297081840037847,
+      "learning_rate": 3.538663171690695e-05,
+      "loss": 0.3983,
+      "step": 925
+    },
+    {
+      "epoch": 1.0919811320754718,
+      "grad_norm": 0.2232181910545884,
+      "learning_rate": 3.5364788117081696e-05,
+      "loss": 0.4259,
+      "step": 926
+    },
+    {
+      "epoch": 1.0931603773584906,
+      "grad_norm": 0.2114231725565784,
+      "learning_rate": 3.534294451725644e-05,
+      "loss": 0.4318,
+      "step": 927
+    },
+    {
+      "epoch": 1.0943396226415094,
+      "grad_norm": 0.21570060315608092,
+      "learning_rate": 3.5321100917431193e-05,
+      "loss": 0.4096,
+      "step": 928
+    },
+    {
+      "epoch": 1.0955188679245282,
+      "grad_norm": 0.21302875835079219,
+      "learning_rate": 3.5299257317605946e-05,
+      "loss": 0.4187,
+      "step": 929
+    },
+    {
+      "epoch": 1.0966981132075473,
+      "grad_norm": 0.20924239545631723,
+      "learning_rate": 3.527741371778069e-05,
+      "loss": 0.4039,
+      "step": 930
+    },
+    {
+      "epoch": 1.0978773584905661,
+      "grad_norm": 0.21512035836153612,
+      "learning_rate": 3.5255570117955436e-05,
+      "loss": 0.4356,
+      "step": 931
+    },
+    {
+      "epoch": 1.099056603773585,
+      "grad_norm": 0.2486007182735829,
+      "learning_rate": 3.523372651813019e-05,
+      "loss": 0.4087,
+      "step": 932
+    },
+    {
+      "epoch": 1.1002358490566038,
+      "grad_norm": 0.22934276031314774,
+      "learning_rate": 3.521188291830494e-05,
+      "loss": 0.4236,
+      "step": 933
+    },
+    {
+      "epoch": 1.1014150943396226,
+      "grad_norm": 0.2334012486701856,
+      "learning_rate": 3.519003931847969e-05,
+      "loss": 0.3987,
+      "step": 934
+    },
+    {
+      "epoch": 1.1025943396226414,
+      "grad_norm": 0.24959441053396011,
+      "learning_rate": 3.516819571865443e-05,
+      "loss": 0.3821,
+      "step": 935
+    },
+    {
+      "epoch": 1.1037735849056605,
+      "grad_norm": 0.24208877064876277,
+      "learning_rate": 3.514635211882918e-05,
+      "loss": 0.4183,
+      "step": 936
+    },
+    {
+      "epoch": 1.1049528301886793,
+      "grad_norm": 0.2194354939037756,
+      "learning_rate": 3.5124508519003935e-05,
+      "loss": 0.4073,
+      "step": 937
+    },
+    {
+      "epoch": 1.1061320754716981,
+      "grad_norm": 0.21736918690327697,
+      "learning_rate": 3.510266491917869e-05,
+      "loss": 0.3997,
+      "step": 938
+    },
+    {
+      "epoch": 1.107311320754717,
+      "grad_norm": 0.23233968126321417,
+      "learning_rate": 3.508082131935343e-05,
+      "loss": 0.4171,
+      "step": 939
+    },
+    {
+      "epoch": 1.1084905660377358,
+      "grad_norm": 0.2320808182824319,
+      "learning_rate": 3.505897771952818e-05,
+      "loss": 0.4329,
+      "step": 940
+    },
+    {
+      "epoch": 1.1096698113207548,
+      "grad_norm": 0.21160946926760762,
+      "learning_rate": 3.503713411970293e-05,
+      "loss": 0.4351,
+      "step": 941
+    },
+    {
+      "epoch": 1.1108490566037736,
+      "grad_norm": 0.23669619875320522,
+      "learning_rate": 3.5015290519877675e-05,
+      "loss": 0.4439,
+      "step": 942
+    },
+    {
+      "epoch": 1.1120283018867925,
+      "grad_norm": 0.23033546690867743,
+      "learning_rate": 3.499344692005243e-05,
+      "loss": 0.4403,
+      "step": 943
+    },
+    {
+      "epoch": 1.1132075471698113,
+      "grad_norm": 0.19627312029637123,
+      "learning_rate": 3.497160332022717e-05,
+      "loss": 0.3849,
+      "step": 944
+    },
+    {
+      "epoch": 1.1143867924528301,
+      "grad_norm": 0.23710315554106845,
+      "learning_rate": 3.4949759720401925e-05,
+      "loss": 0.4141,
+      "step": 945
+    },
+    {
+      "epoch": 1.115566037735849,
+      "grad_norm": 0.23067796404852123,
+      "learning_rate": 3.492791612057667e-05,
+      "loss": 0.3952,
+      "step": 946
+    },
+    {
+      "epoch": 1.116745283018868,
+      "grad_norm": 0.21853311974733883,
+      "learning_rate": 3.490607252075142e-05,
+      "loss": 0.4254,
+      "step": 947
+    },
+    {
+      "epoch": 1.1179245283018868,
+      "grad_norm": 0.2171570925030938,
+      "learning_rate": 3.4884228920926174e-05,
+      "loss": 0.4335,
+      "step": 948
+    },
+    {
+      "epoch": 1.1191037735849056,
+      "grad_norm": 0.22221856963560851,
+      "learning_rate": 3.486238532110092e-05,
+      "loss": 0.3995,
+      "step": 949
+    },
+    {
+      "epoch": 1.1202830188679245,
+      "grad_norm": 0.23591573448152003,
+      "learning_rate": 3.4840541721275665e-05,
+      "loss": 0.4223,
+      "step": 950
+    },
+    {
+      "epoch": 1.1214622641509433,
+      "grad_norm": 0.21726031860781705,
+      "learning_rate": 3.481869812145042e-05,
+      "loss": 0.4213,
+      "step": 951
+    },
+    {
+      "epoch": 1.1226415094339623,
+      "grad_norm": 0.2345818359228329,
+      "learning_rate": 3.479685452162517e-05,
+      "loss": 0.4153,
+      "step": 952
+    },
+    {
+      "epoch": 1.1238207547169812,
+      "grad_norm": 0.25818821183326407,
+      "learning_rate": 3.4775010921799914e-05,
+      "loss": 0.4278,
+      "step": 953
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.2238814728237969,
+      "learning_rate": 3.475316732197466e-05,
+      "loss": 0.4359,
+      "step": 954
+    },
+    {
+      "epoch": 1.1261792452830188,
+      "grad_norm": 0.25397829866569566,
+      "learning_rate": 3.473132372214941e-05,
+      "loss": 0.4027,
+      "step": 955
+    },
+    {
+      "epoch": 1.1273584905660377,
+      "grad_norm": 0.2620335713951043,
+      "learning_rate": 3.4709480122324164e-05,
+      "loss": 0.3987,
+      "step": 956
+    },
+    {
+      "epoch": 1.1285377358490567,
+      "grad_norm": 0.19450300949986937,
+      "learning_rate": 3.468763652249891e-05,
+      "loss": 0.4154,
+      "step": 957
+    },
+    {
+      "epoch": 1.1297169811320755,
+      "grad_norm": 0.24198737424840266,
+      "learning_rate": 3.4665792922673654e-05,
+      "loss": 0.428,
+      "step": 958
+    },
+    {
+      "epoch": 1.1308962264150944,
+      "grad_norm": 0.2512950734080483,
+      "learning_rate": 3.4643949322848406e-05,
+      "loss": 0.4473,
+      "step": 959
+    },
+    {
+      "epoch": 1.1320754716981132,
+      "grad_norm": 0.21132413004451991,
+      "learning_rate": 3.462210572302316e-05,
+      "loss": 0.4034,
+      "step": 960
+    },
+    {
+      "epoch": 1.133254716981132,
+      "grad_norm": 0.24974313355720326,
+      "learning_rate": 3.460026212319791e-05,
+      "loss": 0.4581,
+      "step": 961
+    },
+    {
+      "epoch": 1.134433962264151,
+      "grad_norm": 0.24184955562130983,
+      "learning_rate": 3.457841852337265e-05,
+      "loss": 0.3989,
+      "step": 962
+    },
+    {
+      "epoch": 1.1356132075471699,
+      "grad_norm": 0.23240296194054316,
+      "learning_rate": 3.45565749235474e-05,
+      "loss": 0.4136,
+      "step": 963
+    },
+    {
+      "epoch": 1.1367924528301887,
+      "grad_norm": 0.26110883639283866,
+      "learning_rate": 3.453473132372215e-05,
+      "loss": 0.4187,
+      "step": 964
+    },
+    {
+      "epoch": 1.1379716981132075,
+      "grad_norm": 0.2231482144770287,
+      "learning_rate": 3.45128877238969e-05,
+      "loss": 0.397,
+      "step": 965
+    },
+    {
+      "epoch": 1.1391509433962264,
+      "grad_norm": 0.2491193717203104,
+      "learning_rate": 3.449104412407165e-05,
+      "loss": 0.4134,
+      "step": 966
+    },
+    {
+      "epoch": 1.1403301886792452,
+      "grad_norm": 0.21091551458094684,
+      "learning_rate": 3.4469200524246396e-05,
+      "loss": 0.4215,
+      "step": 967
+    },
+    {
+      "epoch": 1.1415094339622642,
+      "grad_norm": 0.2270353396903535,
+      "learning_rate": 3.444735692442115e-05,
+      "loss": 0.4224,
+      "step": 968
+    },
+    {
+      "epoch": 1.142688679245283,
+      "grad_norm": 0.2215959466238593,
+      "learning_rate": 3.442551332459589e-05,
+      "loss": 0.3911,
+      "step": 969
+    },
+    {
+      "epoch": 1.1438679245283019,
+      "grad_norm": 0.20342045246332588,
+      "learning_rate": 3.4403669724770645e-05,
+      "loss": 0.4131,
+      "step": 970
+    },
+    {
+      "epoch": 1.1450471698113207,
+      "grad_norm": 0.2201313865388485,
+      "learning_rate": 3.438182612494539e-05,
+      "loss": 0.3976,
+      "step": 971
+    },
+    {
+      "epoch": 1.1462264150943395,
+      "grad_norm": 0.22475125950381664,
+      "learning_rate": 3.435998252512014e-05,
+      "loss": 0.4211,
+      "step": 972
+    },
+    {
+      "epoch": 1.1474056603773586,
+      "grad_norm": 0.2509267885869613,
+      "learning_rate": 3.433813892529489e-05,
+      "loss": 0.4287,
+      "step": 973
+    },
+    {
+      "epoch": 1.1485849056603774,
+      "grad_norm": 0.21894575501817715,
+      "learning_rate": 3.431629532546964e-05,
+      "loss": 0.4479,
+      "step": 974
+    },
+    {
+      "epoch": 1.1497641509433962,
+      "grad_norm": 0.22153312520421983,
+      "learning_rate": 3.4294451725644386e-05,
+      "loss": 0.4075,
+      "step": 975
+    },
+    {
+      "epoch": 1.150943396226415,
+      "grad_norm": 0.22332768252013702,
+      "learning_rate": 3.427260812581914e-05,
+      "loss": 0.4117,
+      "step": 976
+    },
+    {
+      "epoch": 1.1521226415094339,
+      "grad_norm": 0.19762286316303287,
+      "learning_rate": 3.425076452599388e-05,
+      "loss": 0.4192,
+      "step": 977
+    },
+    {
+      "epoch": 1.1533018867924527,
+      "grad_norm": 0.21137993114746936,
+      "learning_rate": 3.4228920926168635e-05,
+      "loss": 0.4167,
+      "step": 978
+    },
+    {
+      "epoch": 1.1544811320754718,
+      "grad_norm": 0.21466542301585587,
+      "learning_rate": 3.420707732634339e-05,
+      "loss": 0.4176,
+      "step": 979
+    },
+    {
+      "epoch": 1.1556603773584906,
+      "grad_norm": 0.21649072331121802,
+      "learning_rate": 3.418523372651813e-05,
+      "loss": 0.4293,
+      "step": 980
+    },
+    {
+      "epoch": 1.1568396226415094,
+      "grad_norm": 0.21502231792903864,
+      "learning_rate": 3.416339012669288e-05,
+      "loss": 0.4039,
+      "step": 981
+    },
+    {
+      "epoch": 1.1580188679245282,
+      "grad_norm": 0.2066566694757468,
+      "learning_rate": 3.414154652686763e-05,
+      "loss": 0.408,
+      "step": 982
+    },
+    {
+      "epoch": 1.1591981132075473,
+      "grad_norm": 0.24893495675051047,
+      "learning_rate": 3.411970292704238e-05,
+      "loss": 0.4043,
+      "step": 983
+    },
+    {
+      "epoch": 1.1603773584905661,
+      "grad_norm": 0.23962991862025937,
+      "learning_rate": 3.409785932721713e-05,
+      "loss": 0.4157,
+      "step": 984
+    },
+    {
+      "epoch": 1.161556603773585,
+      "grad_norm": 0.22986352827672185,
+      "learning_rate": 3.407601572739187e-05,
+      "loss": 0.4058,
+      "step": 985
+    },
+    {
+      "epoch": 1.1627358490566038,
+      "grad_norm": 0.24057830210035722,
+      "learning_rate": 3.4054172127566625e-05,
+      "loss": 0.4238,
+      "step": 986
+    },
+    {
+      "epoch": 1.1639150943396226,
+      "grad_norm": 0.23021709634603632,
+      "learning_rate": 3.403232852774138e-05,
+      "loss": 0.4373,
+      "step": 987
+    },
+    {
+      "epoch": 1.1650943396226414,
+      "grad_norm": 0.21047186954123018,
+      "learning_rate": 3.401048492791612e-05,
+      "loss": 0.4028,
+      "step": 988
+    },
+    {
+      "epoch": 1.1662735849056605,
+      "grad_norm": 0.2418626698134968,
+      "learning_rate": 3.398864132809087e-05,
+      "loss": 0.4,
+      "step": 989
+    },
+    {
+      "epoch": 1.1674528301886793,
+      "grad_norm": 0.20215092837297166,
+      "learning_rate": 3.396679772826562e-05,
+      "loss": 0.4078,
+      "step": 990
+    },
+    {
+      "epoch": 1.1686320754716981,
+      "grad_norm": 0.22961966283322793,
+      "learning_rate": 3.394495412844037e-05,
+      "loss": 0.4073,
+      "step": 991
+    },
+    {
+      "epoch": 1.169811320754717,
+      "grad_norm": 0.20897711725590074,
+      "learning_rate": 3.392311052861512e-05,
+      "loss": 0.3995,
+      "step": 992
+    },
+    {
+      "epoch": 1.1709905660377358,
+      "grad_norm": 0.2345691259138141,
+      "learning_rate": 3.390126692878986e-05,
+      "loss": 0.4353,
+      "step": 993
+    },
+    {
+      "epoch": 1.1721698113207548,
+      "grad_norm": 0.2189663949268016,
+      "learning_rate": 3.3879423328964614e-05,
+      "loss": 0.4403,
+      "step": 994
+    },
+    {
+      "epoch": 1.1733490566037736,
+      "grad_norm": 0.22948939645377653,
+      "learning_rate": 3.3857579729139366e-05,
+      "loss": 0.4007,
+      "step": 995
+    },
+    {
+      "epoch": 1.1745283018867925,
+      "grad_norm": 0.2265026190426907,
+      "learning_rate": 3.383573612931411e-05,
+      "loss": 0.4358,
+      "step": 996
+    },
+    {
+      "epoch": 1.1757075471698113,
+      "grad_norm": 0.2242473535849221,
+      "learning_rate": 3.3813892529488864e-05,
+      "loss": 0.429,
+      "step": 997
+    },
+    {
+      "epoch": 1.1768867924528301,
+      "grad_norm": 0.2242839081883014,
+      "learning_rate": 3.379204892966361e-05,
+      "loss": 0.3842,
+      "step": 998
+    },
+    {
+      "epoch": 1.178066037735849,
+      "grad_norm": 0.21947240039489727,
+      "learning_rate": 3.377020532983836e-05,
+      "loss": 0.4021,
+      "step": 999
+    },
+    {
+      "epoch": 1.179245283018868,
+      "grad_norm": 0.21863120596541222,
+      "learning_rate": 3.3748361730013106e-05,
+      "loss": 0.4061,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1804245283018868,
+      "grad_norm": 0.2442909934093241,
+      "learning_rate": 3.372651813018786e-05,
+      "loss": 0.4478,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1816037735849056,
+      "grad_norm": 0.22476413549457236,
+      "learning_rate": 3.3704674530362604e-05,
+      "loss": 0.4062,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1827830188679245,
+      "grad_norm": 0.2589744433899884,
+      "learning_rate": 3.3682830930537356e-05,
+      "loss": 0.4636,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1839622641509433,
+      "grad_norm": 0.22651418094771644,
+      "learning_rate": 3.36609873307121e-05,
+      "loss": 0.4336,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1851415094339623,
+      "grad_norm": 0.2596314826852634,
+      "learning_rate": 3.363914373088685e-05,
+      "loss": 0.4195,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1863207547169812,
+      "grad_norm": 0.2500772984863732,
+      "learning_rate": 3.3617300131061605e-05,
+      "loss": 0.4184,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1875,
+      "grad_norm": 0.22904926250539473,
+      "learning_rate": 3.3595456531236344e-05,
+      "loss": 0.4074,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1886792452830188,
+      "grad_norm": 0.21485118471219933,
+      "learning_rate": 3.3573612931411096e-05,
+      "loss": 0.4086,
+      "step": 1008
+    },
+    {
+      "epoch": 1.1898584905660377,
+      "grad_norm": 0.2126835515282088,
+      "learning_rate": 3.355176933158585e-05,
+      "loss": 0.3963,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1910377358490567,
+      "grad_norm": 0.25321948534115873,
+      "learning_rate": 3.35299257317606e-05,
+      "loss": 0.4243,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1922169811320755,
+      "grad_norm": 0.20454034184926043,
+      "learning_rate": 3.3508082131935345e-05,
+      "loss": 0.3947,
+      "step": 1011
+    },
+    {
+      "epoch": 1.1933962264150944,
+      "grad_norm": 0.2515301641120097,
+      "learning_rate": 3.348623853211009e-05,
+      "loss": 0.4127,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1945754716981132,
+      "grad_norm": 0.21780071948467974,
+      "learning_rate": 3.346439493228484e-05,
+      "loss": 0.4074,
+      "step": 1013
+    },
+    {
+      "epoch": 1.195754716981132,
+      "grad_norm": 0.21996689087102791,
+      "learning_rate": 3.3442551332459595e-05,
+      "loss": 0.3942,
+      "step": 1014
+    },
+    {
+      "epoch": 1.196933962264151,
+      "grad_norm": 0.21739502656751997,
+      "learning_rate": 3.342070773263434e-05,
+      "loss": 0.3817,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1981132075471699,
+      "grad_norm": 0.21786696467668035,
+      "learning_rate": 3.3398864132809085e-05,
+      "loss": 0.4096,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1992924528301887,
+      "grad_norm": 0.21214698954985847,
+      "learning_rate": 3.337702053298384e-05,
+      "loss": 0.4208,
+      "step": 1017
+    },
+    {
+      "epoch": 1.2004716981132075,
+      "grad_norm": 0.4769803058871747,
+      "learning_rate": 3.335517693315859e-05,
+      "loss": 0.4036,
+      "step": 1018
+    },
+    {
+      "epoch": 1.2016509433962264,
+      "grad_norm": 0.24369273296115573,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.4591,
+      "step": 1019
+    },
+    {
+      "epoch": 1.2028301886792452,
+      "grad_norm": 0.19720830923078247,
+      "learning_rate": 3.331148973350808e-05,
+      "loss": 0.4191,
+      "step": 1020
+    },
+    {
+      "epoch": 1.2040094339622642,
+      "grad_norm": 0.19995037419735143,
+      "learning_rate": 3.328964613368283e-05,
+      "loss": 0.4097,
+      "step": 1021
+    },
+    {
+      "epoch": 1.205188679245283,
+      "grad_norm": 0.21735368588479498,
+      "learning_rate": 3.3267802533857584e-05,
+      "loss": 0.423,
+      "step": 1022
+    },
+    {
+      "epoch": 1.2063679245283019,
+      "grad_norm": 0.2193821784871924,
+      "learning_rate": 3.324595893403233e-05,
+      "loss": 0.4077,
+      "step": 1023
+    },
+    {
+      "epoch": 1.2075471698113207,
+      "grad_norm": 0.2099978232984502,
+      "learning_rate": 3.322411533420708e-05,
+      "loss": 0.4252,
+      "step": 1024
+    },
+    {
+      "epoch": 1.2087264150943395,
+      "grad_norm": 0.22556440932476454,
+      "learning_rate": 3.320227173438183e-05,
+      "loss": 0.422,
+      "step": 1025
+    },
+    {
+      "epoch": 1.2099056603773586,
+      "grad_norm": 0.25032684094109714,
+      "learning_rate": 3.318042813455658e-05,
+      "loss": 0.4282,
+      "step": 1026
+    },
+    {
+      "epoch": 1.2110849056603774,
+      "grad_norm": 0.2254763785262832,
+      "learning_rate": 3.3158584534731324e-05,
+      "loss": 0.4106,
+      "step": 1027
+    },
+    {
+      "epoch": 1.2122641509433962,
+      "grad_norm": 0.24153785012873127,
+      "learning_rate": 3.3136740934906077e-05,
+      "loss": 0.422,
+      "step": 1028
+    },
+    {
+      "epoch": 1.213443396226415,
+      "grad_norm": 0.21869124238887114,
+      "learning_rate": 3.311489733508082e-05,
+      "loss": 0.3955,
+      "step": 1029
+    },
+    {
+      "epoch": 1.2146226415094339,
+      "grad_norm": 0.22262594606125383,
+      "learning_rate": 3.309305373525557e-05,
+      "loss": 0.3924,
+      "step": 1030
+    },
+    {
+      "epoch": 1.2158018867924527,
+      "grad_norm": 0.2687513901720489,
+      "learning_rate": 3.307121013543032e-05,
+      "loss": 0.4338,
+      "step": 1031
+    },
+    {
+      "epoch": 1.2169811320754718,
+      "grad_norm": 0.2413375299118909,
+      "learning_rate": 3.304936653560507e-05,
+      "loss": 0.4236,
+      "step": 1032
+    },
+    {
+      "epoch": 1.2181603773584906,
+      "grad_norm": 0.2171339813887172,
+      "learning_rate": 3.302752293577982e-05,
+      "loss": 0.4186,
+      "step": 1033
+    },
+    {
+      "epoch": 1.2193396226415094,
+      "grad_norm": 0.2636259378417185,
+      "learning_rate": 3.300567933595456e-05,
+      "loss": 0.4351,
+      "step": 1034
+    },
+    {
+      "epoch": 1.2205188679245282,
+      "grad_norm": 0.2549674923328108,
+      "learning_rate": 3.2983835736129314e-05,
+      "loss": 0.4227,
+      "step": 1035
+    },
+    {
+      "epoch": 1.2216981132075473,
+      "grad_norm": 0.21357497778166554,
+      "learning_rate": 3.2961992136304066e-05,
+      "loss": 0.4134,
+      "step": 1036
+    },
+    {
+      "epoch": 1.2228773584905661,
+      "grad_norm": 0.2552236621449751,
+      "learning_rate": 3.294014853647882e-05,
+      "loss": 0.4168,
+      "step": 1037
+    },
+    {
+      "epoch": 1.224056603773585,
+      "grad_norm": 0.26926500902347517,
+      "learning_rate": 3.291830493665356e-05,
+      "loss": 0.4187,
+      "step": 1038
+    },
+    {
+      "epoch": 1.2252358490566038,
+      "grad_norm": 0.2337395513239789,
+      "learning_rate": 3.289646133682831e-05,
+      "loss": 0.4114,
+      "step": 1039
+    },
+    {
+      "epoch": 1.2264150943396226,
+      "grad_norm": 0.2223739399076446,
+      "learning_rate": 3.287461773700306e-05,
+      "loss": 0.4106,
+      "step": 1040
+    },
+    {
+      "epoch": 1.2275943396226414,
+      "grad_norm": 0.246641179547852,
+      "learning_rate": 3.285277413717781e-05,
+      "loss": 0.4092,
+      "step": 1041
+    },
+    {
+      "epoch": 1.2287735849056605,
+      "grad_norm": 0.25121859762904725,
+      "learning_rate": 3.283093053735256e-05,
+      "loss": 0.4034,
+      "step": 1042
+    },
+    {
+      "epoch": 1.2299528301886793,
+      "grad_norm": 0.23250455561388367,
+      "learning_rate": 3.2809086937527304e-05,
+      "loss": 0.4087,
+      "step": 1043
+    },
+    {
+      "epoch": 1.2311320754716981,
+      "grad_norm": 0.2245855252964798,
+      "learning_rate": 3.2787243337702056e-05,
+      "loss": 0.4116,
+      "step": 1044
+    },
+    {
+      "epoch": 1.232311320754717,
+      "grad_norm": 0.21738991066960126,
+      "learning_rate": 3.276539973787681e-05,
+      "loss": 0.4116,
+      "step": 1045
+    },
+    {
+      "epoch": 1.2334905660377358,
+      "grad_norm": 0.23620929228767304,
+      "learning_rate": 3.274355613805155e-05,
+      "loss": 0.4339,
+      "step": 1046
+    },
+    {
+      "epoch": 1.2346698113207548,
+      "grad_norm": 0.2117923869801937,
+      "learning_rate": 3.27217125382263e-05,
+      "loss": 0.4189,
+      "step": 1047
+    },
+    {
+      "epoch": 1.2358490566037736,
+      "grad_norm": 0.20905637529253512,
+      "learning_rate": 3.269986893840105e-05,
+      "loss": 0.4168,
+      "step": 1048
+    },
+    {
+      "epoch": 1.2370283018867925,
+      "grad_norm": 0.2095097567962844,
+      "learning_rate": 3.2678025338575796e-05,
+      "loss": 0.4121,
+      "step": 1049
+    },
+    {
+      "epoch": 1.2382075471698113,
+      "grad_norm": 0.2162062155094128,
+      "learning_rate": 3.265618173875055e-05,
+      "loss": 0.4165,
+      "step": 1050
+    },
+    {
+      "epoch": 1.2393867924528301,
+      "grad_norm": 0.22847860243598767,
+      "learning_rate": 3.26343381389253e-05,
+      "loss": 0.4184,
+      "step": 1051
+    },
+    {
+      "epoch": 1.240566037735849,
+      "grad_norm": 0.22696290173009662,
+      "learning_rate": 3.2612494539100045e-05,
+      "loss": 0.4106,
+      "step": 1052
+    },
+    {
+      "epoch": 1.241745283018868,
+      "grad_norm": 0.21494911684378043,
+      "learning_rate": 3.259065093927479e-05,
+      "loss": 0.411,
+      "step": 1053
+    },
+    {
+      "epoch": 1.2429245283018868,
+      "grad_norm": 0.21721005475472557,
+      "learning_rate": 3.256880733944954e-05,
+      "loss": 0.4228,
+      "step": 1054
+    },
+    {
+      "epoch": 1.2441037735849056,
+      "grad_norm": 0.1985844482694683,
+      "learning_rate": 3.2546963739624295e-05,
+      "loss": 0.4012,
+      "step": 1055
+    },
+    {
+      "epoch": 1.2452830188679245,
+      "grad_norm": 0.2394317165292965,
+      "learning_rate": 3.252512013979904e-05,
+      "loss": 0.4196,
+      "step": 1056
+    },
+    {
+      "epoch": 1.2464622641509433,
+      "grad_norm": 0.20612692237738908,
+      "learning_rate": 3.2503276539973785e-05,
+      "loss": 0.4113,
+      "step": 1057
+    },
+    {
+      "epoch": 1.2476415094339623,
+      "grad_norm": 0.2177435693658179,
+      "learning_rate": 3.248143294014854e-05,
+      "loss": 0.4147,
+      "step": 1058
+    },
+    {
+      "epoch": 1.2488207547169812,
+      "grad_norm": 0.21162431535332576,
+      "learning_rate": 3.245958934032329e-05,
+      "loss": 0.4014,
+      "step": 1059
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.21552576236853646,
+      "learning_rate": 3.2437745740498035e-05,
+      "loss": 0.4072,
+      "step": 1060
+    },
+    {
+      "epoch": 1.2511792452830188,
+      "grad_norm": 0.21502848979293437,
+      "learning_rate": 3.241590214067278e-05,
+      "loss": 0.397,
+      "step": 1061
+    },
+    {
+      "epoch": 1.2523584905660377,
+      "grad_norm": 0.2540899237923299,
+      "learning_rate": 3.239405854084753e-05,
+      "loss": 0.4434,
+      "step": 1062
+    },
+    {
+      "epoch": 1.2535377358490565,
+      "grad_norm": 0.2400206230374796,
+      "learning_rate": 3.2372214941022284e-05,
+      "loss": 0.4311,
+      "step": 1063
+    },
+    {
+      "epoch": 1.2547169811320755,
+      "grad_norm": 0.2178779663373802,
+      "learning_rate": 3.2350371341197036e-05,
+      "loss": 0.4256,
+      "step": 1064
+    },
+    {
+      "epoch": 1.2558962264150944,
+      "grad_norm": 0.22594960160257838,
+      "learning_rate": 3.2328527741371775e-05,
+      "loss": 0.4111,
+      "step": 1065
+    },
+    {
+      "epoch": 1.2570754716981132,
+      "grad_norm": 0.24036363605551977,
+      "learning_rate": 3.230668414154653e-05,
+      "loss": 0.4191,
+      "step": 1066
+    },
+    {
+      "epoch": 1.258254716981132,
+      "grad_norm": 0.2302931359546029,
+      "learning_rate": 3.228484054172128e-05,
+      "loss": 0.4116,
+      "step": 1067
+    },
+    {
+      "epoch": 1.259433962264151,
+      "grad_norm": 0.25952126055386066,
+      "learning_rate": 3.226299694189603e-05,
+      "loss": 0.4528,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2606132075471699,
+      "grad_norm": 0.24594400166785893,
+      "learning_rate": 3.2241153342070776e-05,
+      "loss": 0.417,
+      "step": 1069
+    },
+    {
+      "epoch": 1.2617924528301887,
+      "grad_norm": 0.2550412883576241,
+      "learning_rate": 3.221930974224552e-05,
+      "loss": 0.422,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2629716981132075,
+      "grad_norm": 0.28920845370496895,
+      "learning_rate": 3.2197466142420274e-05,
+      "loss": 0.4008,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2641509433962264,
+      "grad_norm": 0.21897468968860365,
+      "learning_rate": 3.217562254259502e-05,
+      "loss": 0.3932,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2653301886792452,
+      "grad_norm": 0.24654577985251894,
+      "learning_rate": 3.215377894276977e-05,
+      "loss": 0.4095,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2665094339622642,
+      "grad_norm": 0.20662448906140657,
+      "learning_rate": 3.2131935342944517e-05,
+      "loss": 0.4059,
+      "step": 1074
+    },
+    {
+      "epoch": 1.267688679245283,
+      "grad_norm": 0.2221341237349744,
+      "learning_rate": 3.211009174311927e-05,
+      "loss": 0.4119,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2688679245283019,
+      "grad_norm": 0.2525718502206223,
+      "learning_rate": 3.2088248143294014e-05,
+      "loss": 0.4383,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2700471698113207,
+      "grad_norm": 0.26046690548612217,
+      "learning_rate": 3.2066404543468766e-05,
+      "loss": 0.402,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2712264150943398,
+      "grad_norm": 0.2635634146904999,
+      "learning_rate": 3.204456094364351e-05,
+      "loss": 0.4142,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2724056603773586,
+      "grad_norm": 0.23913527021594191,
+      "learning_rate": 3.2022717343818263e-05,
+      "loss": 0.3989,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2735849056603774,
+      "grad_norm": 0.265201248179354,
+      "learning_rate": 3.200087374399301e-05,
+      "loss": 0.4121,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2747641509433962,
+      "grad_norm": 0.22504762563248468,
+      "learning_rate": 3.197903014416776e-05,
+      "loss": 0.39,
+      "step": 1081
+    },
+    {
+      "epoch": 1.275943396226415,
+      "grad_norm": 0.24003607891681036,
+      "learning_rate": 3.195718654434251e-05,
+      "loss": 0.4064,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2771226415094339,
+      "grad_norm": 0.22565277179048526,
+      "learning_rate": 3.193534294451726e-05,
+      "loss": 0.4072,
+      "step": 1083
+    },
+    {
+      "epoch": 1.2783018867924527,
+      "grad_norm": 0.20274426360494877,
+      "learning_rate": 3.1913499344692004e-05,
+      "loss": 0.411,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2794811320754718,
+      "grad_norm": 0.2229585359667357,
+      "learning_rate": 3.1891655744866756e-05,
+      "loss": 0.4189,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2806603773584906,
+      "grad_norm": 0.24218890147957048,
+      "learning_rate": 3.186981214504151e-05,
+      "loss": 0.4105,
+      "step": 1086
+    },
+    {
+      "epoch": 1.2818396226415094,
+      "grad_norm": 0.20804427385516983,
+      "learning_rate": 3.184796854521625e-05,
+      "loss": 0.4214,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2830188679245282,
+      "grad_norm": 0.21722736342605142,
+      "learning_rate": 3.1826124945391e-05,
+      "loss": 0.4021,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2841981132075473,
+      "grad_norm": 0.23329911381942678,
+      "learning_rate": 3.180428134556575e-05,
+      "loss": 0.4349,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2853773584905661,
+      "grad_norm": 0.24174279136236873,
+      "learning_rate": 3.17824377457405e-05,
+      "loss": 0.4431,
+      "step": 1090
+    },
+    {
+      "epoch": 1.286556603773585,
+      "grad_norm": 0.20387308959846118,
+      "learning_rate": 3.176059414591525e-05,
+      "loss": 0.4055,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2877358490566038,
+      "grad_norm": 0.2288444532374385,
+      "learning_rate": 3.173875054608999e-05,
+      "loss": 0.4009,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2889150943396226,
+      "grad_norm": 0.2427921008128372,
+      "learning_rate": 3.1716906946264745e-05,
+      "loss": 0.4292,
+      "step": 1093
+    },
+    {
+      "epoch": 1.2900943396226414,
+      "grad_norm": 0.19975768041996433,
+      "learning_rate": 3.16950633464395e-05,
+      "loss": 0.3778,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2912735849056602,
+      "grad_norm": 0.24760102770115847,
+      "learning_rate": 3.167321974661424e-05,
+      "loss": 0.4206,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2924528301886793,
+      "grad_norm": 0.23938594686973963,
+      "learning_rate": 3.1651376146788995e-05,
+      "loss": 0.424,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2936320754716981,
+      "grad_norm": 0.24203804844983895,
+      "learning_rate": 3.162953254696374e-05,
+      "loss": 0.4222,
+      "step": 1097
+    },
+    {
+      "epoch": 1.294811320754717,
+      "grad_norm": 0.20663289875932084,
+      "learning_rate": 3.160768894713849e-05,
+      "loss": 0.406,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2959905660377358,
+      "grad_norm": 0.269721984447071,
+      "learning_rate": 3.158584534731324e-05,
+      "loss": 0.4071,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2971698113207548,
+      "grad_norm": 0.2350698250527083,
+      "learning_rate": 3.156400174748799e-05,
+      "loss": 0.3984,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2983490566037736,
+      "grad_norm": 0.24359081126993667,
+      "learning_rate": 3.1542158147662735e-05,
+      "loss": 0.4108,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2995283018867925,
+      "grad_norm": 0.25919980025334965,
+      "learning_rate": 3.152031454783749e-05,
+      "loss": 0.3999,
+      "step": 1102
+    },
+    {
+      "epoch": 1.3007075471698113,
+      "grad_norm": 0.22644665241871645,
+      "learning_rate": 3.149847094801223e-05,
+      "loss": 0.4165,
+      "step": 1103
+    },
+    {
+      "epoch": 1.3018867924528301,
+      "grad_norm": 0.21640355564800734,
+      "learning_rate": 3.1476627348186984e-05,
+      "loss": 0.4151,
+      "step": 1104
+    },
+    {
+      "epoch": 1.303066037735849,
+      "grad_norm": 0.21044750475263166,
+      "learning_rate": 3.145478374836173e-05,
+      "loss": 0.4123,
+      "step": 1105
+    },
+    {
+      "epoch": 1.304245283018868,
+      "grad_norm": 0.2596684273426906,
+      "learning_rate": 3.143294014853648e-05,
+      "loss": 0.4533,
+      "step": 1106
+    },
+    {
+      "epoch": 1.3054245283018868,
+      "grad_norm": 0.21778031884000562,
+      "learning_rate": 3.141109654871123e-05,
+      "loss": 0.4348,
+      "step": 1107
+    },
+    {
+      "epoch": 1.3066037735849056,
+      "grad_norm": 0.2203410587923888,
+      "learning_rate": 3.138925294888598e-05,
+      "loss": 0.3861,
+      "step": 1108
+    },
+    {
+      "epoch": 1.3077830188679245,
+      "grad_norm": 0.23704242531073683,
+      "learning_rate": 3.136740934906073e-05,
+      "loss": 0.3967,
+      "step": 1109
+    },
+    {
+      "epoch": 1.3089622641509435,
+      "grad_norm": 0.22574710630487824,
+      "learning_rate": 3.1345565749235476e-05,
+      "loss": 0.3979,
+      "step": 1110
+    },
+    {
+      "epoch": 1.3101415094339623,
+      "grad_norm": 0.22499475536730915,
+      "learning_rate": 3.132372214941022e-05,
+      "loss": 0.4057,
+      "step": 1111
+    },
+    {
+      "epoch": 1.3113207547169812,
+      "grad_norm": 0.24748961224987165,
+      "learning_rate": 3.1301878549584974e-05,
+      "loss": 0.4376,
+      "step": 1112
+    },
+    {
+      "epoch": 1.3125,
+      "grad_norm": 0.20805081945804674,
+      "learning_rate": 3.1280034949759726e-05,
+      "loss": 0.3989,
+      "step": 1113
+    },
+    {
+      "epoch": 1.3136792452830188,
+      "grad_norm": 0.23877522701781242,
+      "learning_rate": 3.125819134993447e-05,
+      "loss": 0.4328,
+      "step": 1114
+    },
+    {
+      "epoch": 1.3148584905660377,
+      "grad_norm": 0.21861501107625828,
+      "learning_rate": 3.1236347750109216e-05,
+      "loss": 0.4066,
+      "step": 1115
+    },
+    {
+      "epoch": 1.3160377358490565,
+      "grad_norm": 0.2213445213821968,
+      "learning_rate": 3.121450415028397e-05,
+      "loss": 0.415,
+      "step": 1116
+    },
+    {
+      "epoch": 1.3172169811320755,
+      "grad_norm": 0.21920748106059837,
+      "learning_rate": 3.119266055045872e-05,
+      "loss": 0.4027,
+      "step": 1117
+    },
+    {
+      "epoch": 1.3183962264150944,
+      "grad_norm": 0.2205863856595905,
+      "learning_rate": 3.1170816950633466e-05,
+      "loss": 0.3865,
+      "step": 1118
+    },
+    {
+      "epoch": 1.3195754716981132,
+      "grad_norm": 0.23645466906136384,
+      "learning_rate": 3.114897335080821e-05,
+      "loss": 0.4364,
+      "step": 1119
+    },
+    {
+      "epoch": 1.320754716981132,
+      "grad_norm": 0.1940203671146209,
+      "learning_rate": 3.112712975098296e-05,
+      "loss": 0.4012,
+      "step": 1120
+    },
+    {
+      "epoch": 1.321933962264151,
+      "grad_norm": 0.29195500914442635,
+      "learning_rate": 3.1105286151157715e-05,
+      "loss": 0.4186,
+      "step": 1121
+    },
+    {
+      "epoch": 1.3231132075471699,
+      "grad_norm": 0.20963762014057083,
+      "learning_rate": 3.108344255133246e-05,
+      "loss": 0.3893,
+      "step": 1122
+    },
+    {
+      "epoch": 1.3242924528301887,
+      "grad_norm": 0.22705631075745752,
+      "learning_rate": 3.1061598951507206e-05,
+      "loss": 0.406,
+      "step": 1123
+    },
+    {
+      "epoch": 1.3254716981132075,
+      "grad_norm": 0.2352334247953276,
+      "learning_rate": 3.103975535168196e-05,
+      "loss": 0.3997,
+      "step": 1124
+    },
+    {
+      "epoch": 1.3266509433962264,
+      "grad_norm": 0.20236757693286403,
+      "learning_rate": 3.101791175185671e-05,
+      "loss": 0.3871,
+      "step": 1125
+    },
+    {
+      "epoch": 1.3278301886792452,
+      "grad_norm": 0.22227741892024316,
+      "learning_rate": 3.0996068152031456e-05,
+      "loss": 0.3856,
+      "step": 1126
+    },
+    {
+      "epoch": 1.3290094339622642,
+      "grad_norm": 0.2264957885294324,
+      "learning_rate": 3.097422455220621e-05,
+      "loss": 0.4036,
+      "step": 1127
+    },
+    {
+      "epoch": 1.330188679245283,
+      "grad_norm": 0.2156636005227406,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 0.4086,
+      "step": 1128
+    },
+    {
+      "epoch": 1.3313679245283019,
+      "grad_norm": 0.22819101801770172,
+      "learning_rate": 3.0930537352555705e-05,
+      "loss": 0.4438,
+      "step": 1129
+    },
+    {
+      "epoch": 1.3325471698113207,
+      "grad_norm": 0.23871428091141658,
+      "learning_rate": 3.090869375273045e-05,
+      "loss": 0.4256,
+      "step": 1130
+    },
+    {
+      "epoch": 1.3337264150943398,
+      "grad_norm": 0.22662412061632314,
+      "learning_rate": 3.08868501529052e-05,
+      "loss": 0.3872,
+      "step": 1131
+    },
+    {
+      "epoch": 1.3349056603773586,
+      "grad_norm": 0.20728948185196747,
+      "learning_rate": 3.086500655307995e-05,
+      "loss": 0.4271,
+      "step": 1132
+    },
+    {
+      "epoch": 1.3360849056603774,
+      "grad_norm": 0.2681663120803152,
+      "learning_rate": 3.08431629532547e-05,
+      "loss": 0.4108,
+      "step": 1133
+    },
+    {
+      "epoch": 1.3372641509433962,
+      "grad_norm": 0.21632811131384935,
+      "learning_rate": 3.0821319353429445e-05,
+      "loss": 0.4027,
+      "step": 1134
+    },
+    {
+      "epoch": 1.338443396226415,
+      "grad_norm": 0.2163203102147421,
+      "learning_rate": 3.07994757536042e-05,
+      "loss": 0.4065,
+      "step": 1135
+    },
+    {
+      "epoch": 1.3396226415094339,
+      "grad_norm": 0.2507229400797401,
+      "learning_rate": 3.077763215377894e-05,
+      "loss": 0.4231,
+      "step": 1136
+    },
+    {
+      "epoch": 1.3408018867924527,
+      "grad_norm": 0.24856261427420429,
+      "learning_rate": 3.075578855395369e-05,
+      "loss": 0.4182,
+      "step": 1137
+    },
+    {
+      "epoch": 1.3419811320754718,
+      "grad_norm": 0.20656230141492646,
+      "learning_rate": 3.073394495412844e-05,
+      "loss": 0.4204,
+      "step": 1138
+    },
+    {
+      "epoch": 1.3431603773584906,
+      "grad_norm": 0.2130486608100856,
+      "learning_rate": 3.071210135430319e-05,
+      "loss": 0.4166,
+      "step": 1139
+    },
+    {
+      "epoch": 1.3443396226415094,
+      "grad_norm": 0.23254619428244877,
+      "learning_rate": 3.0690257754477944e-05,
+      "loss": 0.4059,
+      "step": 1140
+    },
+    {
+      "epoch": 1.3455188679245282,
+      "grad_norm": 0.24964959148864335,
+      "learning_rate": 3.066841415465269e-05,
+      "loss": 0.4178,
+      "step": 1141
+    },
+    {
+      "epoch": 1.3466981132075473,
+      "grad_norm": 0.20473780476420295,
+      "learning_rate": 3.0646570554827435e-05,
+      "loss": 0.379,
+      "step": 1142
+    },
+    {
+      "epoch": 1.3478773584905661,
+      "grad_norm": 0.26669775917647015,
+      "learning_rate": 3.062472695500219e-05,
+      "loss": 0.4089,
+      "step": 1143
+    },
+    {
+      "epoch": 1.349056603773585,
+      "grad_norm": 0.2626805701507444,
+      "learning_rate": 3.060288335517694e-05,
+      "loss": 0.4316,
+      "step": 1144
+    },
+    {
+      "epoch": 1.3502358490566038,
+      "grad_norm": 0.1983747933868541,
+      "learning_rate": 3.0581039755351684e-05,
+      "loss": 0.428,
+      "step": 1145
+    },
+    {
+      "epoch": 1.3514150943396226,
+      "grad_norm": 0.2521189734831692,
+      "learning_rate": 3.055919615552643e-05,
+      "loss": 0.4109,
+      "step": 1146
+    },
+    {
+      "epoch": 1.3525943396226414,
+      "grad_norm": 0.22450470095906516,
+      "learning_rate": 3.053735255570118e-05,
+      "loss": 0.3949,
+      "step": 1147
+    },
+    {
+      "epoch": 1.3537735849056602,
+      "grad_norm": 0.2272506671296636,
+      "learning_rate": 3.0515508955875934e-05,
+      "loss": 0.424,
+      "step": 1148
+    },
+    {
+      "epoch": 1.3549528301886793,
+      "grad_norm": 0.23439796618049022,
+      "learning_rate": 3.0493665356050676e-05,
+      "loss": 0.3823,
+      "step": 1149
+    },
+    {
+      "epoch": 1.3561320754716981,
+      "grad_norm": 0.21179540823840506,
+      "learning_rate": 3.0471821756225428e-05,
+      "loss": 0.3868,
+      "step": 1150
+    },
+    {
+      "epoch": 1.357311320754717,
+      "grad_norm": 0.2514948841761971,
+      "learning_rate": 3.0449978156400176e-05,
+      "loss": 0.425,
+      "step": 1151
+    },
+    {
+      "epoch": 1.3584905660377358,
+      "grad_norm": 0.21770946544533817,
+      "learning_rate": 3.042813455657493e-05,
+      "loss": 0.3885,
+      "step": 1152
+    },
+    {
+      "epoch": 1.3596698113207548,
+      "grad_norm": 0.2183907340992172,
+      "learning_rate": 3.0406290956749674e-05,
+      "loss": 0.4339,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3608490566037736,
+      "grad_norm": 0.24160807680309979,
+      "learning_rate": 3.0384447356924422e-05,
+      "loss": 0.4004,
+      "step": 1154
+    },
+    {
+      "epoch": 1.3620283018867925,
+      "grad_norm": 0.2247905059187668,
+      "learning_rate": 3.0362603757099174e-05,
+      "loss": 0.4006,
+      "step": 1155
+    },
+    {
+      "epoch": 1.3632075471698113,
+      "grad_norm": 0.2222645477943424,
+      "learning_rate": 3.0340760157273916e-05,
+      "loss": 0.4206,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3643867924528301,
+      "grad_norm": 0.24331039884078204,
+      "learning_rate": 3.031891655744867e-05,
+      "loss": 0.415,
+      "step": 1157
+    },
+    {
+      "epoch": 1.365566037735849,
+      "grad_norm": 0.2244619925350708,
+      "learning_rate": 3.0297072957623417e-05,
+      "loss": 0.4164,
+      "step": 1158
+    },
+    {
+      "epoch": 1.366745283018868,
+      "grad_norm": 0.20041603151246618,
+      "learning_rate": 3.027522935779817e-05,
+      "loss": 0.4033,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3679245283018868,
+      "grad_norm": 0.22544962420164721,
+      "learning_rate": 3.0253385757972915e-05,
+      "loss": 0.4031,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3691037735849056,
+      "grad_norm": 0.2467522553564424,
+      "learning_rate": 3.0231542158147663e-05,
+      "loss": 0.421,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3702830188679245,
+      "grad_norm": 0.20769792608005763,
+      "learning_rate": 3.0209698558322412e-05,
+      "loss": 0.3692,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3714622641509435,
+      "grad_norm": 0.23334734847370314,
+      "learning_rate": 3.0187854958497164e-05,
+      "loss": 0.4279,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3726415094339623,
+      "grad_norm": 0.19781875389876216,
+      "learning_rate": 3.016601135867191e-05,
+      "loss": 0.4031,
+      "step": 1164
+    },
+    {
+      "epoch": 1.3738207547169812,
+      "grad_norm": 0.22335064991637626,
+      "learning_rate": 3.0144167758846658e-05,
+      "loss": 0.4308,
+      "step": 1165
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 0.20457324469400257,
+      "learning_rate": 3.012232415902141e-05,
+      "loss": 0.4099,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3761792452830188,
+      "grad_norm": 0.2130012886480549,
+      "learning_rate": 3.010048055919616e-05,
+      "loss": 0.4099,
+      "step": 1167
+    },
+    {
+      "epoch": 1.3773584905660377,
+      "grad_norm": 0.21554914458870472,
+      "learning_rate": 3.0078636959370904e-05,
+      "loss": 0.4313,
+      "step": 1168
+    },
+    {
+      "epoch": 1.3785377358490565,
+      "grad_norm": 0.22563536746684412,
+      "learning_rate": 3.0056793359545653e-05,
+      "loss": 0.4293,
+      "step": 1169
+    },
+    {
+      "epoch": 1.3797169811320755,
+      "grad_norm": 0.21539723189091176,
+      "learning_rate": 3.0034949759720405e-05,
+      "loss": 0.4217,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3808962264150944,
+      "grad_norm": 0.20216601910373092,
+      "learning_rate": 3.0013106159895154e-05,
+      "loss": 0.4138,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3820754716981132,
+      "grad_norm": 0.2082324466440655,
+      "learning_rate": 2.99912625600699e-05,
+      "loss": 0.4402,
+      "step": 1172
+    },
+    {
+      "epoch": 1.383254716981132,
+      "grad_norm": 0.20928093925780542,
+      "learning_rate": 2.996941896024465e-05,
+      "loss": 0.4149,
+      "step": 1173
+    },
+    {
+      "epoch": 1.384433962264151,
+      "grad_norm": 0.23572973167688563,
+      "learning_rate": 2.99475753604194e-05,
+      "loss": 0.4221,
+      "step": 1174
+    },
+    {
+      "epoch": 1.3856132075471699,
+      "grad_norm": 0.20847886034131094,
+      "learning_rate": 2.9925731760594152e-05,
+      "loss": 0.4102,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3867924528301887,
+      "grad_norm": 0.21670573597889825,
+      "learning_rate": 2.9903888160768894e-05,
+      "loss": 0.4286,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3879716981132075,
+      "grad_norm": 0.23523765736139765,
+      "learning_rate": 2.9882044560943646e-05,
+      "loss": 0.3979,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3891509433962264,
+      "grad_norm": 0.2202168943575381,
+      "learning_rate": 2.9860200961118394e-05,
+      "loss": 0.435,
+      "step": 1178
+    },
+    {
+      "epoch": 1.3903301886792452,
+      "grad_norm": 0.2106768445784913,
+      "learning_rate": 2.983835736129314e-05,
+      "loss": 0.4265,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3915094339622642,
+      "grad_norm": 0.21781193330958606,
+      "learning_rate": 2.9816513761467892e-05,
+      "loss": 0.4285,
+      "step": 1180
+    },
+    {
+      "epoch": 1.392688679245283,
+      "grad_norm": 0.22848843055984788,
+      "learning_rate": 2.979467016164264e-05,
+      "loss": 0.4223,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3938679245283019,
+      "grad_norm": 0.23357079098523334,
+      "learning_rate": 2.977282656181739e-05,
+      "loss": 0.4268,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3950471698113207,
+      "grad_norm": 0.2169085152421391,
+      "learning_rate": 2.9750982961992135e-05,
+      "loss": 0.4082,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3962264150943398,
+      "grad_norm": 0.2248445131870522,
+      "learning_rate": 2.9729139362166887e-05,
+      "loss": 0.3971,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3974056603773586,
+      "grad_norm": 0.22761982101547149,
+      "learning_rate": 2.9707295762341635e-05,
+      "loss": 0.4157,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3985849056603774,
+      "grad_norm": 0.22674103738583629,
+      "learning_rate": 2.9685452162516387e-05,
+      "loss": 0.4237,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3997641509433962,
+      "grad_norm": 0.23401223041187905,
+      "learning_rate": 2.966360856269113e-05,
+      "loss": 0.4116,
+      "step": 1187
+    },
+    {
+      "epoch": 1.400943396226415,
+      "grad_norm": 0.21174438889195843,
+      "learning_rate": 2.964176496286588e-05,
+      "loss": 0.4183,
+      "step": 1188
+    },
+    {
+      "epoch": 1.4021226415094339,
+      "grad_norm": 0.23284393408664572,
+      "learning_rate": 2.961992136304063e-05,
+      "loss": 0.4128,
+      "step": 1189
+    },
+    {
+      "epoch": 1.4033018867924527,
+      "grad_norm": 0.23769220204045302,
+      "learning_rate": 2.9598077763215382e-05,
+      "loss": 0.4162,
+      "step": 1190
+    },
+    {
+      "epoch": 1.4044811320754718,
+      "grad_norm": 0.22892352349051076,
+      "learning_rate": 2.9576234163390127e-05,
+      "loss": 0.4041,
+      "step": 1191
+    },
+    {
+      "epoch": 1.4056603773584906,
+      "grad_norm": 0.21767973139941368,
+      "learning_rate": 2.9554390563564876e-05,
+      "loss": 0.4318,
+      "step": 1192
+    },
+    {
+      "epoch": 1.4068396226415094,
+      "grad_norm": 0.2245307327905914,
+      "learning_rate": 2.9532546963739628e-05,
+      "loss": 0.4143,
+      "step": 1193
+    },
+    {
+      "epoch": 1.4080188679245282,
+      "grad_norm": 0.21276501991358118,
+      "learning_rate": 2.9510703363914377e-05,
+      "loss": 0.4179,
+      "step": 1194
+    },
+    {
+      "epoch": 1.4091981132075473,
+      "grad_norm": 0.2184193987510409,
+      "learning_rate": 2.9488859764089122e-05,
+      "loss": 0.4074,
+      "step": 1195
+    },
+    {
+      "epoch": 1.4103773584905661,
+      "grad_norm": 0.21965573611436523,
+      "learning_rate": 2.946701616426387e-05,
+      "loss": 0.4222,
+      "step": 1196
+    },
+    {
+      "epoch": 1.411556603773585,
+      "grad_norm": 0.22602369306827297,
+      "learning_rate": 2.9445172564438623e-05,
+      "loss": 0.4225,
+      "step": 1197
+    },
+    {
+      "epoch": 1.4127358490566038,
+      "grad_norm": 0.2224538723674266,
+      "learning_rate": 2.942332896461337e-05,
+      "loss": 0.4422,
+      "step": 1198
+    },
+    {
+      "epoch": 1.4139150943396226,
+      "grad_norm": 0.2449057609587921,
+      "learning_rate": 2.9401485364788117e-05,
+      "loss": 0.4225,
+      "step": 1199
+    },
+    {
+      "epoch": 1.4150943396226414,
+      "grad_norm": 0.2430327026502991,
+      "learning_rate": 2.937964176496287e-05,
+      "loss": 0.3994,
+      "step": 1200
+    },
+    {
+      "epoch": 1.4162735849056602,
+      "grad_norm": 0.2417690933179807,
+      "learning_rate": 2.9357798165137618e-05,
+      "loss": 0.4054,
+      "step": 1201
+    },
+    {
+      "epoch": 1.4174528301886793,
+      "grad_norm": 0.20457342164313377,
+      "learning_rate": 2.9335954565312363e-05,
+      "loss": 0.4075,
+      "step": 1202
+    },
+    {
+      "epoch": 1.4186320754716981,
+      "grad_norm": 0.20444037734689097,
+      "learning_rate": 2.9314110965487112e-05,
+      "loss": 0.398,
+      "step": 1203
+    },
+    {
+      "epoch": 1.419811320754717,
+      "grad_norm": 0.24386128531552634,
+      "learning_rate": 2.9292267365661864e-05,
+      "loss": 0.3936,
+      "step": 1204
+    },
+    {
+      "epoch": 1.4209905660377358,
+      "grad_norm": 0.2354642115339831,
+      "learning_rate": 2.9270423765836613e-05,
+      "loss": 0.3978,
+      "step": 1205
+    },
+    {
+      "epoch": 1.4221698113207548,
+      "grad_norm": 0.2080090519225196,
+      "learning_rate": 2.9248580166011358e-05,
+      "loss": 0.4021,
+      "step": 1206
+    },
+    {
+      "epoch": 1.4233490566037736,
+      "grad_norm": 0.23178177765576122,
+      "learning_rate": 2.9226736566186107e-05,
+      "loss": 0.3804,
+      "step": 1207
+    },
+    {
+      "epoch": 1.4245283018867925,
+      "grad_norm": 0.24146723708388626,
+      "learning_rate": 2.920489296636086e-05,
+      "loss": 0.4173,
+      "step": 1208
+    },
+    {
+      "epoch": 1.4257075471698113,
+      "grad_norm": 0.24083873403430023,
+      "learning_rate": 2.9183049366535607e-05,
+      "loss": 0.3986,
+      "step": 1209
+    },
+    {
+      "epoch": 1.4268867924528301,
+      "grad_norm": 0.2198240828287477,
+      "learning_rate": 2.9161205766710353e-05,
+      "loss": 0.3982,
+      "step": 1210
+    },
+    {
+      "epoch": 1.428066037735849,
+      "grad_norm": 0.2207082276811831,
+      "learning_rate": 2.9139362166885105e-05,
+      "loss": 0.4086,
+      "step": 1211
+    },
+    {
+      "epoch": 1.429245283018868,
+      "grad_norm": 0.20168133090077064,
+      "learning_rate": 2.9117518567059853e-05,
+      "loss": 0.3973,
+      "step": 1212
+    },
+    {
+      "epoch": 1.4304245283018868,
+      "grad_norm": 0.21247662800264422,
+      "learning_rate": 2.9095674967234606e-05,
+      "loss": 0.4101,
+      "step": 1213
+    },
+    {
+      "epoch": 1.4316037735849056,
+      "grad_norm": 0.21663793009251478,
+      "learning_rate": 2.9073831367409347e-05,
+      "loss": 0.4215,
+      "step": 1214
+    },
+    {
+      "epoch": 1.4327830188679245,
+      "grad_norm": 0.21005689415248643,
+      "learning_rate": 2.90519877675841e-05,
+      "loss": 0.3884,
+      "step": 1215
+    },
+    {
+      "epoch": 1.4339622641509435,
+      "grad_norm": 0.2261696356271968,
+      "learning_rate": 2.9030144167758848e-05,
+      "loss": 0.4139,
+      "step": 1216
+    },
+    {
+      "epoch": 1.4351415094339623,
+      "grad_norm": 0.19461677714612402,
+      "learning_rate": 2.90083005679336e-05,
+      "loss": 0.3972,
+      "step": 1217
+    },
+    {
+      "epoch": 1.4363207547169812,
+      "grad_norm": 0.24740204280560768,
+      "learning_rate": 2.8986456968108346e-05,
+      "loss": 0.3964,
+      "step": 1218
+    },
+    {
+      "epoch": 1.4375,
+      "grad_norm": 0.21511328583691422,
+      "learning_rate": 2.8964613368283094e-05,
+      "loss": 0.3883,
+      "step": 1219
+    },
+    {
+      "epoch": 1.4386792452830188,
+      "grad_norm": 0.23597616187957676,
+      "learning_rate": 2.8942769768457843e-05,
+      "loss": 0.408,
+      "step": 1220
+    },
+    {
+      "epoch": 1.4398584905660377,
+      "grad_norm": 0.22914132191471998,
+      "learning_rate": 2.892092616863259e-05,
+      "loss": 0.4302,
+      "step": 1221
+    },
+    {
+      "epoch": 1.4410377358490565,
+      "grad_norm": 0.21866274201597038,
+      "learning_rate": 2.889908256880734e-05,
+      "loss": 0.4174,
+      "step": 1222
+    },
+    {
+      "epoch": 1.4422169811320755,
+      "grad_norm": 0.23996863702599655,
+      "learning_rate": 2.887723896898209e-05,
+      "loss": 0.3914,
+      "step": 1223
+    },
+    {
+      "epoch": 1.4433962264150944,
+      "grad_norm": 0.21896454452835618,
+      "learning_rate": 2.885539536915684e-05,
+      "loss": 0.3949,
+      "step": 1224
+    },
+    {
+      "epoch": 1.4445754716981132,
+      "grad_norm": 0.209472986492276,
+      "learning_rate": 2.8833551769331587e-05,
+      "loss": 0.4012,
+      "step": 1225
+    },
+    {
+      "epoch": 1.445754716981132,
+      "grad_norm": 0.21680184056749827,
+      "learning_rate": 2.8811708169506335e-05,
+      "loss": 0.4018,
+      "step": 1226
+    },
+    {
+      "epoch": 1.446933962264151,
+      "grad_norm": 0.20089233663698924,
+      "learning_rate": 2.8789864569681084e-05,
+      "loss": 0.3867,
+      "step": 1227
+    },
+    {
+      "epoch": 1.4481132075471699,
+      "grad_norm": 0.21958465251311016,
+      "learning_rate": 2.8768020969855836e-05,
+      "loss": 0.399,
+      "step": 1228
+    },
+    {
+      "epoch": 1.4492924528301887,
+      "grad_norm": 0.20395005806438185,
+      "learning_rate": 2.874617737003058e-05,
+      "loss": 0.4176,
+      "step": 1229
+    },
+    {
+      "epoch": 1.4504716981132075,
+      "grad_norm": 0.22100851906890784,
+      "learning_rate": 2.872433377020533e-05,
+      "loss": 0.3913,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4516509433962264,
+      "grad_norm": 0.23085587241091715,
+      "learning_rate": 2.8702490170380082e-05,
+      "loss": 0.3903,
+      "step": 1231
+    },
+    {
+      "epoch": 1.4528301886792452,
+      "grad_norm": 0.2041094404731893,
+      "learning_rate": 2.868064657055483e-05,
+      "loss": 0.3945,
+      "step": 1232
+    },
+    {
+      "epoch": 1.4540094339622642,
+      "grad_norm": 0.23507916888151262,
+      "learning_rate": 2.8658802970729576e-05,
+      "loss": 0.4062,
+      "step": 1233
+    },
+    {
+      "epoch": 1.455188679245283,
+      "grad_norm": 0.22911415278238498,
+      "learning_rate": 2.8636959370904325e-05,
+      "loss": 0.425,
+      "step": 1234
+    },
+    {
+      "epoch": 1.4563679245283019,
+      "grad_norm": 0.2146946099288813,
+      "learning_rate": 2.8615115771079077e-05,
+      "loss": 0.4163,
+      "step": 1235
+    },
+    {
+      "epoch": 1.4575471698113207,
+      "grad_norm": 0.25003350590801404,
+      "learning_rate": 2.8593272171253826e-05,
+      "loss": 0.4232,
+      "step": 1236
+    },
+    {
+      "epoch": 1.4587264150943398,
+      "grad_norm": 0.24030858064362737,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.4324,
+      "step": 1237
+    },
+    {
+      "epoch": 1.4599056603773586,
+      "grad_norm": 0.20561801098512333,
+      "learning_rate": 2.8549584971603323e-05,
+      "loss": 0.3941,
+      "step": 1238
+    },
+    {
+      "epoch": 1.4610849056603774,
+      "grad_norm": 0.2486502958220263,
+      "learning_rate": 2.852774137177807e-05,
+      "loss": 0.3853,
+      "step": 1239
+    },
+    {
+      "epoch": 1.4622641509433962,
+      "grad_norm": 0.21468116450689015,
+      "learning_rate": 2.850589777195282e-05,
+      "loss": 0.3869,
+      "step": 1240
+    },
+    {
+      "epoch": 1.463443396226415,
+      "grad_norm": 0.20510275498142055,
+      "learning_rate": 2.8484054172127566e-05,
+      "loss": 0.4081,
+      "step": 1241
+    },
+    {
+      "epoch": 1.4646226415094339,
+      "grad_norm": 0.22856064836861462,
+      "learning_rate": 2.8462210572302318e-05,
+      "loss": 0.4124,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4658018867924527,
+      "grad_norm": 0.21410598292559924,
+      "learning_rate": 2.8440366972477066e-05,
+      "loss": 0.4284,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4669811320754718,
+      "grad_norm": 0.22639153920391838,
+      "learning_rate": 2.8418523372651812e-05,
+      "loss": 0.4264,
+      "step": 1244
+    },
+    {
+      "epoch": 1.4681603773584906,
+      "grad_norm": 0.2139515316842224,
+      "learning_rate": 2.839667977282656e-05,
+      "loss": 0.3964,
+      "step": 1245
+    },
+    {
+      "epoch": 1.4693396226415094,
+      "grad_norm": 0.22342295202892448,
+      "learning_rate": 2.8374836173001313e-05,
+      "loss": 0.4136,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4705188679245282,
+      "grad_norm": 0.20942969092853156,
+      "learning_rate": 2.835299257317606e-05,
+      "loss": 0.4021,
+      "step": 1247
+    },
+    {
+      "epoch": 1.4716981132075473,
+      "grad_norm": 0.2268743739438781,
+      "learning_rate": 2.8331148973350807e-05,
+      "loss": 0.4034,
+      "step": 1248
+    },
+    {
+      "epoch": 1.4728773584905661,
+      "grad_norm": 0.21937665639585302,
+      "learning_rate": 2.830930537352556e-05,
+      "loss": 0.3978,
+      "step": 1249
+    },
+    {
+      "epoch": 1.474056603773585,
+      "grad_norm": 0.21991768330386674,
+      "learning_rate": 2.8287461773700307e-05,
+      "loss": 0.4096,
+      "step": 1250
+    },
+    {
+      "epoch": 1.4752358490566038,
+      "grad_norm": 0.22547411302778128,
+      "learning_rate": 2.826561817387506e-05,
+      "loss": 0.4352,
+      "step": 1251
+    },
+    {
+      "epoch": 1.4764150943396226,
+      "grad_norm": 0.20652929795572805,
+      "learning_rate": 2.82437745740498e-05,
+      "loss": 0.4258,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4775943396226414,
+      "grad_norm": 0.21929401185180966,
+      "learning_rate": 2.8221930974224553e-05,
+      "loss": 0.4101,
+      "step": 1253
+    },
+    {
+      "epoch": 1.4787735849056602,
+      "grad_norm": 0.20457005797909533,
+      "learning_rate": 2.8200087374399302e-05,
+      "loss": 0.3883,
+      "step": 1254
+    },
+    {
+      "epoch": 1.4799528301886793,
+      "grad_norm": 0.20654234803512725,
+      "learning_rate": 2.8178243774574054e-05,
+      "loss": 0.3729,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4811320754716981,
+      "grad_norm": 0.20442574732073074,
+      "learning_rate": 2.81564001747488e-05,
+      "loss": 0.4007,
+      "step": 1256
+    },
+    {
+      "epoch": 1.482311320754717,
+      "grad_norm": 0.22697880518685118,
+      "learning_rate": 2.8134556574923548e-05,
+      "loss": 0.4186,
+      "step": 1257
+    },
+    {
+      "epoch": 1.4834905660377358,
+      "grad_norm": 0.2126800661835353,
+      "learning_rate": 2.81127129750983e-05,
+      "loss": 0.4043,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4846698113207548,
+      "grad_norm": 0.20589292197849315,
+      "learning_rate": 2.809086937527305e-05,
+      "loss": 0.4312,
+      "step": 1259
+    },
+    {
+      "epoch": 1.4858490566037736,
+      "grad_norm": 0.21508678859969108,
+      "learning_rate": 2.8069025775447794e-05,
+      "loss": 0.4132,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4870283018867925,
+      "grad_norm": 0.20099776208291142,
+      "learning_rate": 2.8047182175622543e-05,
+      "loss": 0.404,
+      "step": 1261
+    },
+    {
+      "epoch": 1.4882075471698113,
+      "grad_norm": 0.2265047385693361,
+      "learning_rate": 2.8025338575797295e-05,
+      "loss": 0.4058,
+      "step": 1262
+    },
+    {
+      "epoch": 1.4893867924528301,
+      "grad_norm": 0.23373053864403623,
+      "learning_rate": 2.800349497597204e-05,
+      "loss": 0.4087,
+      "step": 1263
+    },
+    {
+      "epoch": 1.490566037735849,
+      "grad_norm": 0.20085489654072947,
+      "learning_rate": 2.798165137614679e-05,
+      "loss": 0.4177,
+      "step": 1264
+    },
+    {
+      "epoch": 1.491745283018868,
+      "grad_norm": 0.22979369530457386,
+      "learning_rate": 2.7959807776321538e-05,
+      "loss": 0.4087,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4929245283018868,
+      "grad_norm": 0.2016193170433445,
+      "learning_rate": 2.793796417649629e-05,
+      "loss": 0.4046,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4941037735849056,
+      "grad_norm": 0.20731633574914662,
+      "learning_rate": 2.7916120576671035e-05,
+      "loss": 0.415,
+      "step": 1267
+    },
+    {
+      "epoch": 1.4952830188679245,
+      "grad_norm": 0.21399062257150217,
+      "learning_rate": 2.7894276976845784e-05,
+      "loss": 0.4129,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4964622641509435,
+      "grad_norm": 0.24371866684733842,
+      "learning_rate": 2.7872433377020536e-05,
+      "loss": 0.395,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4976415094339623,
+      "grad_norm": 0.218285387039673,
+      "learning_rate": 2.7850589777195285e-05,
+      "loss": 0.4236,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4988207547169812,
+      "grad_norm": 0.3017102894752362,
+      "learning_rate": 2.782874617737003e-05,
+      "loss": 0.417,
+      "step": 1271
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.23325490009111427,
+      "learning_rate": 2.780690257754478e-05,
+      "loss": 0.4414,
+      "step": 1272
+    },
+    {
+      "epoch": 1.5011792452830188,
+      "grad_norm": 0.22573139650054141,
+      "learning_rate": 2.778505897771953e-05,
+      "loss": 0.4143,
+      "step": 1273
+    },
+    {
+      "epoch": 1.5023584905660377,
+      "grad_norm": 0.2112434457109376,
+      "learning_rate": 2.776321537789428e-05,
+      "loss": 0.4099,
+      "step": 1274
+    },
+    {
+      "epoch": 1.5035377358490565,
+      "grad_norm": 0.2404786884529957,
+      "learning_rate": 2.7741371778069025e-05,
+      "loss": 0.3834,
+      "step": 1275
+    },
+    {
+      "epoch": 1.5047169811320755,
+      "grad_norm": 0.21778863121995937,
+      "learning_rate": 2.7719528178243777e-05,
+      "loss": 0.4204,
+      "step": 1276
+    },
+    {
+      "epoch": 1.5058962264150944,
+      "grad_norm": 0.25065716965489604,
+      "learning_rate": 2.7697684578418525e-05,
+      "loss": 0.4018,
+      "step": 1277
+    },
+    {
+      "epoch": 1.5070754716981132,
+      "grad_norm": 0.22280498274360175,
+      "learning_rate": 2.7675840978593278e-05,
+      "loss": 0.4438,
+      "step": 1278
+    },
+    {
+      "epoch": 1.5082547169811322,
+      "grad_norm": 0.22917987368517873,
+      "learning_rate": 2.765399737876802e-05,
+      "loss": 0.3963,
+      "step": 1279
+    },
+    {
+      "epoch": 1.509433962264151,
+      "grad_norm": 0.19910669000360742,
+      "learning_rate": 2.763215377894277e-05,
+      "loss": 0.3877,
+      "step": 1280
+    },
+    {
+      "epoch": 1.5106132075471699,
+      "grad_norm": 0.2075320870660181,
+      "learning_rate": 2.761031017911752e-05,
+      "loss": 0.4136,
+      "step": 1281
+    },
+    {
+      "epoch": 1.5117924528301887,
+      "grad_norm": 0.22022771369485117,
+      "learning_rate": 2.7588466579292272e-05,
+      "loss": 0.4048,
+      "step": 1282
+    },
+    {
+      "epoch": 1.5129716981132075,
+      "grad_norm": 0.2050375951511324,
+      "learning_rate": 2.7566622979467018e-05,
+      "loss": 0.3825,
+      "step": 1283
+    },
+    {
+      "epoch": 1.5141509433962264,
+      "grad_norm": 0.2024238630403348,
+      "learning_rate": 2.7544779379641766e-05,
+      "loss": 0.4067,
+      "step": 1284
+    },
+    {
+      "epoch": 1.5153301886792452,
+      "grad_norm": 0.2384330182665047,
+      "learning_rate": 2.7522935779816515e-05,
+      "loss": 0.4094,
+      "step": 1285
+    },
+    {
+      "epoch": 1.516509433962264,
+      "grad_norm": 0.2416171764364662,
+      "learning_rate": 2.750109217999126e-05,
+      "loss": 0.4076,
+      "step": 1286
+    },
+    {
+      "epoch": 1.517688679245283,
+      "grad_norm": 0.2342918306592882,
+      "learning_rate": 2.7479248580166012e-05,
+      "loss": 0.4336,
+      "step": 1287
+    },
+    {
+      "epoch": 1.5188679245283019,
+      "grad_norm": 0.20196938982710516,
+      "learning_rate": 2.745740498034076e-05,
+      "loss": 0.3826,
+      "step": 1288
+    },
+    {
+      "epoch": 1.5200471698113207,
+      "grad_norm": 0.2589797487848365,
+      "learning_rate": 2.7435561380515513e-05,
+      "loss": 0.418,
+      "step": 1289
+    },
+    {
+      "epoch": 1.5212264150943398,
+      "grad_norm": 0.22088691306946423,
+      "learning_rate": 2.7413717780690255e-05,
+      "loss": 0.4114,
+      "step": 1290
+    },
+    {
+      "epoch": 1.5224056603773586,
+      "grad_norm": 0.19645590006676264,
+      "learning_rate": 2.7391874180865007e-05,
+      "loss": 0.4178,
+      "step": 1291
+    },
+    {
+      "epoch": 1.5235849056603774,
+      "grad_norm": 0.22551822324368495,
+      "learning_rate": 2.7370030581039756e-05,
+      "loss": 0.4155,
+      "step": 1292
+    },
+    {
+      "epoch": 1.5247641509433962,
+      "grad_norm": 0.20908969135386687,
+      "learning_rate": 2.7348186981214508e-05,
+      "loss": 0.4218,
+      "step": 1293
+    },
+    {
+      "epoch": 1.525943396226415,
+      "grad_norm": 0.20437560964025317,
+      "learning_rate": 2.7326343381389253e-05,
+      "loss": 0.4365,
+      "step": 1294
+    },
+    {
+      "epoch": 1.5271226415094339,
+      "grad_norm": 0.215948786336936,
+      "learning_rate": 2.7304499781564002e-05,
+      "loss": 0.4074,
+      "step": 1295
+    },
+    {
+      "epoch": 1.5283018867924527,
+      "grad_norm": 0.19031902273466564,
+      "learning_rate": 2.7282656181738754e-05,
+      "loss": 0.3922,
+      "step": 1296
+    },
+    {
+      "epoch": 1.5294811320754715,
+      "grad_norm": 0.2076364369313589,
+      "learning_rate": 2.7260812581913503e-05,
+      "loss": 0.3862,
+      "step": 1297
+    },
+    {
+      "epoch": 1.5306603773584906,
+      "grad_norm": 0.22626701150751616,
+      "learning_rate": 2.7238968982088248e-05,
+      "loss": 0.4393,
+      "step": 1298
+    },
+    {
+      "epoch": 1.5318396226415094,
+      "grad_norm": 0.2205716303106275,
+      "learning_rate": 2.7217125382262997e-05,
+      "loss": 0.379,
+      "step": 1299
+    },
+    {
+      "epoch": 1.5330188679245285,
+      "grad_norm": 0.24061639064982476,
+      "learning_rate": 2.719528178243775e-05,
+      "loss": 0.4136,
+      "step": 1300
+    },
+    {
+      "epoch": 1.5341981132075473,
+      "grad_norm": 0.21377985513813186,
+      "learning_rate": 2.7173438182612498e-05,
+      "loss": 0.401,
+      "step": 1301
+    },
+    {
+      "epoch": 1.5353773584905661,
+      "grad_norm": 0.23768790534505693,
+      "learning_rate": 2.7151594582787243e-05,
+      "loss": 0.4169,
+      "step": 1302
+    },
+    {
+      "epoch": 1.536556603773585,
+      "grad_norm": 0.24570610302978826,
+      "learning_rate": 2.7129750982961995e-05,
+      "loss": 0.4189,
+      "step": 1303
+    },
+    {
+      "epoch": 1.5377358490566038,
+      "grad_norm": 0.5101657429976868,
+      "learning_rate": 2.7107907383136744e-05,
+      "loss": 0.4185,
+      "step": 1304
+    },
+    {
+      "epoch": 1.5389150943396226,
+      "grad_norm": 0.2282321995254729,
+      "learning_rate": 2.708606378331149e-05,
+      "loss": 0.3784,
+      "step": 1305
+    },
+    {
+      "epoch": 1.5400943396226414,
+      "grad_norm": 0.20137450058439763,
+      "learning_rate": 2.7064220183486238e-05,
+      "loss": 0.4017,
+      "step": 1306
+    },
+    {
+      "epoch": 1.5412735849056602,
+      "grad_norm": 0.24662803453242976,
+      "learning_rate": 2.704237658366099e-05,
+      "loss": 0.4104,
+      "step": 1307
+    },
+    {
+      "epoch": 1.5424528301886793,
+      "grad_norm": 0.2327488195492099,
+      "learning_rate": 2.702053298383574e-05,
+      "loss": 0.3924,
+      "step": 1308
+    },
+    {
+      "epoch": 1.5436320754716981,
+      "grad_norm": 0.21647335126791142,
+      "learning_rate": 2.6998689384010484e-05,
+      "loss": 0.3982,
+      "step": 1309
+    },
+    {
+      "epoch": 1.544811320754717,
+      "grad_norm": 0.23080075457924176,
+      "learning_rate": 2.6976845784185232e-05,
+      "loss": 0.4102,
+      "step": 1310
+    },
+    {
+      "epoch": 1.545990566037736,
+      "grad_norm": 0.24624106834561066,
+      "learning_rate": 2.6955002184359984e-05,
+      "loss": 0.4181,
+      "step": 1311
+    },
+    {
+      "epoch": 1.5471698113207548,
+      "grad_norm": 0.2169069570052017,
+      "learning_rate": 2.6933158584534733e-05,
+      "loss": 0.4309,
+      "step": 1312
+    },
+    {
+      "epoch": 1.5483490566037736,
+      "grad_norm": 0.22647876208320675,
+      "learning_rate": 2.691131498470948e-05,
+      "loss": 0.4051,
+      "step": 1313
+    },
+    {
+      "epoch": 1.5495283018867925,
+      "grad_norm": 0.19860906872344797,
+      "learning_rate": 2.688947138488423e-05,
+      "loss": 0.3862,
+      "step": 1314
+    },
+    {
+      "epoch": 1.5507075471698113,
+      "grad_norm": 0.22122612341866357,
+      "learning_rate": 2.686762778505898e-05,
+      "loss": 0.4161,
+      "step": 1315
+    },
+    {
+      "epoch": 1.5518867924528301,
+      "grad_norm": 0.24330732642335434,
+      "learning_rate": 2.684578418523373e-05,
+      "loss": 0.3907,
+      "step": 1316
+    },
+    {
+      "epoch": 1.553066037735849,
+      "grad_norm": 0.20772634207286245,
+      "learning_rate": 2.6823940585408473e-05,
+      "loss": 0.4162,
+      "step": 1317
+    },
+    {
+      "epoch": 1.5542452830188678,
+      "grad_norm": 0.2029526056439857,
+      "learning_rate": 2.6802096985583225e-05,
+      "loss": 0.398,
+      "step": 1318
+    },
+    {
+      "epoch": 1.5554245283018868,
+      "grad_norm": 0.24427922535093669,
+      "learning_rate": 2.6780253385757974e-05,
+      "loss": 0.4313,
+      "step": 1319
+    },
+    {
+      "epoch": 1.5566037735849056,
+      "grad_norm": 0.22481608987357404,
+      "learning_rate": 2.6758409785932726e-05,
+      "loss": 0.4087,
+      "step": 1320
+    },
+    {
+      "epoch": 1.5577830188679245,
+      "grad_norm": 0.2277508247603496,
+      "learning_rate": 2.673656618610747e-05,
+      "loss": 0.4299,
+      "step": 1321
+    },
+    {
+      "epoch": 1.5589622641509435,
+      "grad_norm": 0.2404837181971092,
+      "learning_rate": 2.671472258628222e-05,
+      "loss": 0.4391,
+      "step": 1322
+    },
+    {
+      "epoch": 1.5601415094339623,
+      "grad_norm": 0.20572851742077666,
+      "learning_rate": 2.6692878986456972e-05,
+      "loss": 0.3874,
+      "step": 1323
+    },
+    {
+      "epoch": 1.5613207547169812,
+      "grad_norm": 0.21856258410235366,
+      "learning_rate": 2.667103538663172e-05,
+      "loss": 0.4044,
+      "step": 1324
+    },
+    {
+      "epoch": 1.5625,
+      "grad_norm": 0.25681359618643723,
+      "learning_rate": 2.6649191786806466e-05,
+      "loss": 0.4034,
+      "step": 1325
+    },
+    {
+      "epoch": 1.5636792452830188,
+      "grad_norm": 0.20587282221601869,
+      "learning_rate": 2.6627348186981215e-05,
+      "loss": 0.4222,
+      "step": 1326
+    },
+    {
+      "epoch": 1.5648584905660377,
+      "grad_norm": 0.2310961459239284,
+      "learning_rate": 2.6605504587155967e-05,
+      "loss": 0.4109,
+      "step": 1327
+    },
+    {
+      "epoch": 1.5660377358490565,
+      "grad_norm": 0.21698565792287322,
+      "learning_rate": 2.6583660987330712e-05,
+      "loss": 0.4105,
+      "step": 1328
+    },
+    {
+      "epoch": 1.5672169811320755,
+      "grad_norm": 0.20951703936282065,
+      "learning_rate": 2.656181738750546e-05,
+      "loss": 0.3964,
+      "step": 1329
+    },
+    {
+      "epoch": 1.5683962264150944,
+      "grad_norm": 0.221250660474374,
+      "learning_rate": 2.653997378768021e-05,
+      "loss": 0.4124,
+      "step": 1330
+    },
+    {
+      "epoch": 1.5695754716981132,
+      "grad_norm": 0.2262036600737205,
+      "learning_rate": 2.6518130187854962e-05,
+      "loss": 0.4138,
+      "step": 1331
+    },
+    {
+      "epoch": 1.5707547169811322,
+      "grad_norm": 0.2233304512956643,
+      "learning_rate": 2.6496286588029707e-05,
+      "loss": 0.4034,
+      "step": 1332
+    },
+    {
+      "epoch": 1.571933962264151,
+      "grad_norm": 0.21375270546268346,
+      "learning_rate": 2.6474442988204456e-05,
+      "loss": 0.4192,
+      "step": 1333
+    },
+    {
+      "epoch": 1.5731132075471699,
+      "grad_norm": 0.23980774480276704,
+      "learning_rate": 2.6452599388379208e-05,
+      "loss": 0.4427,
+      "step": 1334
+    },
+    {
+      "epoch": 1.5742924528301887,
+      "grad_norm": 0.23500341508980846,
+      "learning_rate": 2.6430755788553957e-05,
+      "loss": 0.3989,
+      "step": 1335
+    },
+    {
+      "epoch": 1.5754716981132075,
+      "grad_norm": 0.19536719999687333,
+      "learning_rate": 2.6408912188728702e-05,
+      "loss": 0.3894,
+      "step": 1336
+    },
+    {
+      "epoch": 1.5766509433962264,
+      "grad_norm": 0.26716711708217916,
+      "learning_rate": 2.638706858890345e-05,
+      "loss": 0.4228,
+      "step": 1337
+    },
+    {
+      "epoch": 1.5778301886792452,
+      "grad_norm": 0.2201248479841448,
+      "learning_rate": 2.6365224989078203e-05,
+      "loss": 0.4027,
+      "step": 1338
+    },
+    {
+      "epoch": 1.579009433962264,
+      "grad_norm": 0.23204178394591046,
+      "learning_rate": 2.634338138925295e-05,
+      "loss": 0.3873,
+      "step": 1339
+    },
+    {
+      "epoch": 1.580188679245283,
+      "grad_norm": 0.2037214007922871,
+      "learning_rate": 2.6321537789427697e-05,
+      "loss": 0.3863,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5813679245283019,
+      "grad_norm": 0.6846561410704591,
+      "learning_rate": 2.629969418960245e-05,
+      "loss": 0.4316,
+      "step": 1341
+    },
+    {
+      "epoch": 1.5825471698113207,
+      "grad_norm": 0.1996460544391412,
+      "learning_rate": 2.6277850589777197e-05,
+      "loss": 0.4009,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5837264150943398,
+      "grad_norm": 0.20397193516153456,
+      "learning_rate": 2.625600698995195e-05,
+      "loss": 0.3897,
+      "step": 1343
+    },
+    {
+      "epoch": 1.5849056603773586,
+      "grad_norm": 0.22472429977774186,
+      "learning_rate": 2.623416339012669e-05,
+      "loss": 0.4313,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5860849056603774,
+      "grad_norm": 0.21011275352511258,
+      "learning_rate": 2.6212319790301444e-05,
+      "loss": 0.393,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5872641509433962,
+      "grad_norm": 0.22314726612075503,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 0.4228,
+      "step": 1346
+    },
+    {
+      "epoch": 1.588443396226415,
+      "grad_norm": 0.22326141907915306,
+      "learning_rate": 2.6168632590650944e-05,
+      "loss": 0.4068,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5896226415094339,
+      "grad_norm": 0.34608664721267224,
+      "learning_rate": 2.614678899082569e-05,
+      "loss": 0.4089,
+      "step": 1348
+    },
+    {
+      "epoch": 1.5908018867924527,
+      "grad_norm": 0.22520828950814906,
+      "learning_rate": 2.6124945391000438e-05,
+      "loss": 0.4128,
+      "step": 1349
+    },
+    {
+      "epoch": 1.5919811320754715,
+      "grad_norm": 0.20828332533623958,
+      "learning_rate": 2.6103101791175187e-05,
+      "loss": 0.3856,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5931603773584906,
+      "grad_norm": 0.22075449968220864,
+      "learning_rate": 2.6081258191349932e-05,
+      "loss": 0.3946,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5943396226415094,
+      "grad_norm": 0.2451071623855063,
+      "learning_rate": 2.6059414591524684e-05,
+      "loss": 0.4081,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5955188679245285,
+      "grad_norm": 0.24089817145489392,
+      "learning_rate": 2.6037570991699433e-05,
+      "loss": 0.4067,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5966981132075473,
+      "grad_norm": 0.222625236897842,
+      "learning_rate": 2.6015727391874185e-05,
+      "loss": 0.4056,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5978773584905661,
+      "grad_norm": 0.21763610766423558,
+      "learning_rate": 2.5993883792048927e-05,
+      "loss": 0.392,
+      "step": 1355
+    },
+    {
+      "epoch": 1.599056603773585,
+      "grad_norm": 0.2228214451854881,
+      "learning_rate": 2.597204019222368e-05,
+      "loss": 0.4307,
+      "step": 1356
+    },
+    {
+      "epoch": 1.6002358490566038,
+      "grad_norm": 0.23732944902864894,
+      "learning_rate": 2.5950196592398428e-05,
+      "loss": 0.4151,
+      "step": 1357
+    },
+    {
+      "epoch": 1.6014150943396226,
+      "grad_norm": 0.21708950897751622,
+      "learning_rate": 2.592835299257318e-05,
+      "loss": 0.4144,
+      "step": 1358
+    },
+    {
+      "epoch": 1.6025943396226414,
+      "grad_norm": 0.21740764034757898,
+      "learning_rate": 2.5906509392747925e-05,
+      "loss": 0.4216,
+      "step": 1359
+    },
+    {
+      "epoch": 1.6037735849056602,
+      "grad_norm": 0.20768328054686858,
+      "learning_rate": 2.5884665792922674e-05,
+      "loss": 0.4139,
+      "step": 1360
+    },
+    {
+      "epoch": 1.6049528301886793,
+      "grad_norm": 0.20555968975633174,
+      "learning_rate": 2.5862822193097426e-05,
+      "loss": 0.3858,
+      "step": 1361
+    },
+    {
+      "epoch": 1.6061320754716981,
+      "grad_norm": 0.20620177182137409,
+      "learning_rate": 2.5840978593272175e-05,
+      "loss": 0.4115,
+      "step": 1362
+    },
+    {
+      "epoch": 1.607311320754717,
+      "grad_norm": 0.1991209831338197,
+      "learning_rate": 2.581913499344692e-05,
+      "loss": 0.4246,
+      "step": 1363
+    },
+    {
+      "epoch": 1.608490566037736,
+      "grad_norm": 0.20355169134061105,
+      "learning_rate": 2.579729139362167e-05,
+      "loss": 0.414,
+      "step": 1364
+    },
+    {
+      "epoch": 1.6096698113207548,
+      "grad_norm": 0.2162492284648667,
+      "learning_rate": 2.577544779379642e-05,
+      "loss": 0.4088,
+      "step": 1365
+    },
+    {
+      "epoch": 1.6108490566037736,
+      "grad_norm": 0.22726527711498448,
+      "learning_rate": 2.575360419397117e-05,
+      "loss": 0.4246,
+      "step": 1366
+    },
+    {
+      "epoch": 1.6120283018867925,
+      "grad_norm": 0.1856703030493701,
+      "learning_rate": 2.5731760594145915e-05,
+      "loss": 0.4177,
+      "step": 1367
+    },
+    {
+      "epoch": 1.6132075471698113,
+      "grad_norm": 0.21193388996056942,
+      "learning_rate": 2.5709916994320667e-05,
+      "loss": 0.4201,
+      "step": 1368
+    },
+    {
+      "epoch": 1.6143867924528301,
+      "grad_norm": 0.20500671277934554,
+      "learning_rate": 2.5688073394495416e-05,
+      "loss": 0.4033,
+      "step": 1369
+    },
+    {
+      "epoch": 1.615566037735849,
+      "grad_norm": 0.21746130175189868,
+      "learning_rate": 2.566622979467016e-05,
+      "loss": 0.3843,
+      "step": 1370
+    },
+    {
+      "epoch": 1.6167452830188678,
+      "grad_norm": 0.22412154104181223,
+      "learning_rate": 2.564438619484491e-05,
+      "loss": 0.412,
+      "step": 1371
+    },
+    {
+      "epoch": 1.6179245283018868,
+      "grad_norm": 0.20566260393070215,
+      "learning_rate": 2.562254259501966e-05,
+      "loss": 0.4107,
+      "step": 1372
+    },
+    {
+      "epoch": 1.6191037735849056,
+      "grad_norm": 0.20265652576908705,
+      "learning_rate": 2.560069899519441e-05,
+      "loss": 0.399,
+      "step": 1373
+    },
+    {
+      "epoch": 1.6202830188679245,
+      "grad_norm": 0.19359727968350354,
+      "learning_rate": 2.5578855395369156e-05,
+      "loss": 0.4091,
+      "step": 1374
+    },
+    {
+      "epoch": 1.6214622641509435,
+      "grad_norm": 0.22425065724836318,
+      "learning_rate": 2.5557011795543904e-05,
+      "loss": 0.4059,
+      "step": 1375
+    },
+    {
+      "epoch": 1.6226415094339623,
+      "grad_norm": 0.21349624510654777,
+      "learning_rate": 2.5535168195718656e-05,
+      "loss": 0.3949,
+      "step": 1376
+    },
+    {
+      "epoch": 1.6238207547169812,
+      "grad_norm": 0.22837274286613238,
+      "learning_rate": 2.5513324595893405e-05,
+      "loss": 0.4053,
+      "step": 1377
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 0.2104213314712281,
+      "learning_rate": 2.549148099606815e-05,
+      "loss": 0.4093,
+      "step": 1378
+    },
+    {
+      "epoch": 1.6261792452830188,
+      "grad_norm": 0.21024215902608887,
+      "learning_rate": 2.5469637396242903e-05,
+      "loss": 0.4049,
+      "step": 1379
+    },
+    {
+      "epoch": 1.6273584905660377,
+      "grad_norm": 0.2052824571562388,
+      "learning_rate": 2.544779379641765e-05,
+      "loss": 0.4121,
+      "step": 1380
+    },
+    {
+      "epoch": 1.6285377358490565,
+      "grad_norm": 0.21366609076732598,
+      "learning_rate": 2.5425950196592403e-05,
+      "loss": 0.4184,
+      "step": 1381
+    },
+    {
+      "epoch": 1.6297169811320755,
+      "grad_norm": 0.20572139889811183,
+      "learning_rate": 2.5404106596767145e-05,
+      "loss": 0.3903,
+      "step": 1382
+    },
+    {
+      "epoch": 1.6308962264150944,
+      "grad_norm": 0.2223247500512629,
+      "learning_rate": 2.5382262996941897e-05,
+      "loss": 0.402,
+      "step": 1383
+    },
+    {
+      "epoch": 1.6320754716981132,
+      "grad_norm": 0.20655947043770684,
+      "learning_rate": 2.5360419397116646e-05,
+      "loss": 0.38,
+      "step": 1384
+    },
+    {
+      "epoch": 1.6332547169811322,
+      "grad_norm": 0.22067263463438555,
+      "learning_rate": 2.5338575797291398e-05,
+      "loss": 0.3915,
+      "step": 1385
+    },
+    {
+      "epoch": 1.634433962264151,
+      "grad_norm": 0.22395694218915477,
+      "learning_rate": 2.5316732197466143e-05,
+      "loss": 0.4229,
+      "step": 1386
+    },
+    {
+      "epoch": 1.6356132075471699,
+      "grad_norm": 0.2038796331995424,
+      "learning_rate": 2.5294888597640892e-05,
+      "loss": 0.3938,
+      "step": 1387
+    },
+    {
+      "epoch": 1.6367924528301887,
+      "grad_norm": 0.22976132451928274,
+      "learning_rate": 2.527304499781564e-05,
+      "loss": 0.4018,
+      "step": 1388
+    },
+    {
+      "epoch": 1.6379716981132075,
+      "grad_norm": 0.2444239508188354,
+      "learning_rate": 2.5251201397990393e-05,
+      "loss": 0.3907,
+      "step": 1389
+    },
+    {
+      "epoch": 1.6391509433962264,
+      "grad_norm": 0.27030009380361064,
+      "learning_rate": 2.5229357798165138e-05,
+      "loss": 0.4231,
+      "step": 1390
+    },
+    {
+      "epoch": 1.6403301886792452,
+      "grad_norm": 0.19178090358470962,
+      "learning_rate": 2.5207514198339887e-05,
+      "loss": 0.4046,
+      "step": 1391
+    },
+    {
+      "epoch": 1.641509433962264,
+      "grad_norm": 0.20484005591835108,
+      "learning_rate": 2.518567059851464e-05,
+      "loss": 0.3883,
+      "step": 1392
+    },
+    {
+      "epoch": 1.642688679245283,
+      "grad_norm": 0.2594607114886826,
+      "learning_rate": 2.5163826998689384e-05,
+      "loss": 0.3933,
+      "step": 1393
+    },
+    {
+      "epoch": 1.6438679245283019,
+      "grad_norm": 0.22960293804866308,
+      "learning_rate": 2.5141983398864133e-05,
+      "loss": 0.4261,
+      "step": 1394
+    },
+    {
+      "epoch": 1.6450471698113207,
+      "grad_norm": 0.19916339697766117,
+      "learning_rate": 2.512013979903888e-05,
+      "loss": 0.4234,
+      "step": 1395
+    },
+    {
+      "epoch": 1.6462264150943398,
+      "grad_norm": 0.22465805650055193,
+      "learning_rate": 2.5098296199213634e-05,
+      "loss": 0.42,
+      "step": 1396
+    },
+    {
+      "epoch": 1.6474056603773586,
+      "grad_norm": 0.24722126321810117,
+      "learning_rate": 2.507645259938838e-05,
+      "loss": 0.413,
+      "step": 1397
+    },
+    {
+      "epoch": 1.6485849056603774,
+      "grad_norm": 0.2248210211575688,
+      "learning_rate": 2.5054608999563128e-05,
+      "loss": 0.4009,
+      "step": 1398
+    },
+    {
+      "epoch": 1.6497641509433962,
+      "grad_norm": 0.20885315007117716,
+      "learning_rate": 2.503276539973788e-05,
+      "loss": 0.4047,
+      "step": 1399
+    },
+    {
+      "epoch": 1.650943396226415,
+      "grad_norm": 0.247704864160704,
+      "learning_rate": 2.501092179991263e-05,
+      "loss": 0.4329,
+      "step": 1400
+    },
+    {
+      "epoch": 1.6521226415094339,
+      "grad_norm": 0.2240866671175147,
+      "learning_rate": 2.4989078200087377e-05,
+      "loss": 0.3873,
+      "step": 1401
+    },
+    {
+      "epoch": 1.6533018867924527,
+      "grad_norm": 0.21584076206617087,
+      "learning_rate": 2.4967234600262123e-05,
+      "loss": 0.4138,
+      "step": 1402
+    },
+    {
+      "epoch": 1.6544811320754715,
+      "grad_norm": 0.22456594837740573,
+      "learning_rate": 2.4945391000436875e-05,
+      "loss": 0.4037,
+      "step": 1403
+    },
+    {
+      "epoch": 1.6556603773584906,
+      "grad_norm": 0.2258079192500056,
+      "learning_rate": 2.492354740061162e-05,
+      "loss": 0.4164,
+      "step": 1404
+    },
+    {
+      "epoch": 1.6568396226415094,
+      "grad_norm": 0.23433695026672222,
+      "learning_rate": 2.4901703800786372e-05,
+      "loss": 0.4205,
+      "step": 1405
+    },
+    {
+      "epoch": 1.6580188679245285,
+      "grad_norm": 0.2394224243133106,
+      "learning_rate": 2.487986020096112e-05,
+      "loss": 0.3926,
+      "step": 1406
+    },
+    {
+      "epoch": 1.6591981132075473,
+      "grad_norm": 0.25266501420619314,
+      "learning_rate": 2.485801660113587e-05,
+      "loss": 0.4151,
+      "step": 1407
+    },
+    {
+      "epoch": 1.6603773584905661,
+      "grad_norm": 0.23351781591081178,
+      "learning_rate": 2.4836173001310618e-05,
+      "loss": 0.4215,
+      "step": 1408
+    },
+    {
+      "epoch": 1.661556603773585,
+      "grad_norm": 0.2316187736865331,
+      "learning_rate": 2.4814329401485367e-05,
+      "loss": 0.4182,
+      "step": 1409
+    },
+    {
+      "epoch": 1.6627358490566038,
+      "grad_norm": 0.20591828024870676,
+      "learning_rate": 2.4792485801660115e-05,
+      "loss": 0.4288,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6639150943396226,
+      "grad_norm": 0.20822142219804787,
+      "learning_rate": 2.4770642201834864e-05,
+      "loss": 0.374,
+      "step": 1411
+    },
+    {
+      "epoch": 1.6650943396226414,
+      "grad_norm": 0.2197090052832817,
+      "learning_rate": 2.4748798602009613e-05,
+      "loss": 0.3838,
+      "step": 1412
+    },
+    {
+      "epoch": 1.6662735849056602,
+      "grad_norm": 0.2186221230207099,
+      "learning_rate": 2.472695500218436e-05,
+      "loss": 0.4214,
+      "step": 1413
+    },
+    {
+      "epoch": 1.6674528301886793,
+      "grad_norm": 0.21755344907970256,
+      "learning_rate": 2.470511140235911e-05,
+      "loss": 0.4218,
+      "step": 1414
+    },
+    {
+      "epoch": 1.6686320754716981,
+      "grad_norm": 0.19832560323237136,
+      "learning_rate": 2.468326780253386e-05,
+      "loss": 0.4117,
+      "step": 1415
+    },
+    {
+      "epoch": 1.669811320754717,
+      "grad_norm": 0.20434942504874115,
+      "learning_rate": 2.4661424202708608e-05,
+      "loss": 0.3957,
+      "step": 1416
+    },
+    {
+      "epoch": 1.670990566037736,
+      "grad_norm": 0.20571852475217003,
+      "learning_rate": 2.4639580602883356e-05,
+      "loss": 0.4148,
+      "step": 1417
+    },
+    {
+      "epoch": 1.6721698113207548,
+      "grad_norm": 0.20286321793331072,
+      "learning_rate": 2.4617737003058105e-05,
+      "loss": 0.4198,
+      "step": 1418
+    },
+    {
+      "epoch": 1.6733490566037736,
+      "grad_norm": 0.18479337429791398,
+      "learning_rate": 2.4595893403232854e-05,
+      "loss": 0.3941,
+      "step": 1419
+    },
+    {
+      "epoch": 1.6745283018867925,
+      "grad_norm": 0.2174168795356865,
+      "learning_rate": 2.4574049803407602e-05,
+      "loss": 0.4241,
+      "step": 1420
+    },
+    {
+      "epoch": 1.6757075471698113,
+      "grad_norm": 0.21503098741152646,
+      "learning_rate": 2.455220620358235e-05,
+      "loss": 0.4166,
+      "step": 1421
+    },
+    {
+      "epoch": 1.6768867924528301,
+      "grad_norm": 0.23666801024149361,
+      "learning_rate": 2.45303626037571e-05,
+      "loss": 0.4159,
+      "step": 1422
+    },
+    {
+      "epoch": 1.678066037735849,
+      "grad_norm": 0.22362139581714074,
+      "learning_rate": 2.450851900393185e-05,
+      "loss": 0.4063,
+      "step": 1423
+    },
+    {
+      "epoch": 1.6792452830188678,
+      "grad_norm": 0.2427404234071747,
+      "learning_rate": 2.4486675404106597e-05,
+      "loss": 0.3868,
+      "step": 1424
+    },
+    {
+      "epoch": 1.6804245283018868,
+      "grad_norm": 0.25051270537980486,
+      "learning_rate": 2.4464831804281346e-05,
+      "loss": 0.4043,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6816037735849056,
+      "grad_norm": 0.2114026433326264,
+      "learning_rate": 2.4442988204456098e-05,
+      "loss": 0.3908,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6827830188679245,
+      "grad_norm": 0.22630961054718832,
+      "learning_rate": 2.4421144604630843e-05,
+      "loss": 0.3812,
+      "step": 1427
+    },
+    {
+      "epoch": 1.6839622641509435,
+      "grad_norm": 0.2027409456596229,
+      "learning_rate": 2.4399301004805595e-05,
+      "loss": 0.4057,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6851415094339623,
+      "grad_norm": 0.19700546924006174,
+      "learning_rate": 2.437745740498034e-05,
+      "loss": 0.3932,
+      "step": 1429
+    },
+    {
+      "epoch": 1.6863207547169812,
+      "grad_norm": 0.20774223962659658,
+      "learning_rate": 2.4355613805155093e-05,
+      "loss": 0.4284,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6875,
+      "grad_norm": 0.20470513048564787,
+      "learning_rate": 2.4333770205329838e-05,
+      "loss": 0.4242,
+      "step": 1431
+    },
+    {
+      "epoch": 1.6886792452830188,
+      "grad_norm": 0.2169322379281303,
+      "learning_rate": 2.431192660550459e-05,
+      "loss": 0.413,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6898584905660377,
+      "grad_norm": 0.19421611718673626,
+      "learning_rate": 2.4290083005679335e-05,
+      "loss": 0.4022,
+      "step": 1433
+    },
+    {
+      "epoch": 1.6910377358490565,
+      "grad_norm": 0.21989456392084616,
+      "learning_rate": 2.4268239405854084e-05,
+      "loss": 0.4401,
+      "step": 1434
+    },
+    {
+      "epoch": 1.6922169811320755,
+      "grad_norm": 0.2027110101652188,
+      "learning_rate": 2.4246395806028836e-05,
+      "loss": 0.3753,
+      "step": 1435
+    },
+    {
+      "epoch": 1.6933962264150944,
+      "grad_norm": 0.21737178954064565,
+      "learning_rate": 2.422455220620358e-05,
+      "loss": 0.4012,
+      "step": 1436
+    },
+    {
+      "epoch": 1.6945754716981132,
+      "grad_norm": 0.22509948088708803,
+      "learning_rate": 2.4202708606378334e-05,
+      "loss": 0.401,
+      "step": 1437
+    },
+    {
+      "epoch": 1.6957547169811322,
+      "grad_norm": 0.22856611121940773,
+      "learning_rate": 2.418086500655308e-05,
+      "loss": 0.4074,
+      "step": 1438
+    },
+    {
+      "epoch": 1.696933962264151,
+      "grad_norm": 0.20183259412126245,
+      "learning_rate": 2.415902140672783e-05,
+      "loss": 0.3915,
+      "step": 1439
+    },
+    {
+      "epoch": 1.6981132075471699,
+      "grad_norm": 0.2269970880291317,
+      "learning_rate": 2.4137177806902576e-05,
+      "loss": 0.4175,
+      "step": 1440
+    },
+    {
+      "epoch": 1.6992924528301887,
+      "grad_norm": 0.21775137163901243,
+      "learning_rate": 2.411533420707733e-05,
+      "loss": 0.3923,
+      "step": 1441
+    },
+    {
+      "epoch": 1.7004716981132075,
+      "grad_norm": 0.17808939393110113,
+      "learning_rate": 2.4093490607252077e-05,
+      "loss": 0.3915,
+      "step": 1442
+    },
+    {
+      "epoch": 1.7016509433962264,
+      "grad_norm": 0.21553681388332016,
+      "learning_rate": 2.4071647007426826e-05,
+      "loss": 0.4031,
+      "step": 1443
+    },
+    {
+      "epoch": 1.7028301886792452,
+      "grad_norm": 0.22068250064248476,
+      "learning_rate": 2.4049803407601575e-05,
+      "loss": 0.4324,
+      "step": 1444
+    },
+    {
+      "epoch": 1.704009433962264,
+      "grad_norm": 0.1956826840509377,
+      "learning_rate": 2.4027959807776323e-05,
+      "loss": 0.3981,
+      "step": 1445
+    },
+    {
+      "epoch": 1.705188679245283,
+      "grad_norm": 0.1995407223296442,
+      "learning_rate": 2.4006116207951072e-05,
+      "loss": 0.3961,
+      "step": 1446
+    },
+    {
+      "epoch": 1.7063679245283019,
+      "grad_norm": 0.1879763400297516,
+      "learning_rate": 2.398427260812582e-05,
+      "loss": 0.3903,
+      "step": 1447
+    },
+    {
+      "epoch": 1.7075471698113207,
+      "grad_norm": 0.19993788979388288,
+      "learning_rate": 2.396242900830057e-05,
+      "loss": 0.3889,
+      "step": 1448
+    },
+    {
+      "epoch": 1.7087264150943398,
+      "grad_norm": 0.2204284809694649,
+      "learning_rate": 2.3940585408475318e-05,
+      "loss": 0.4216,
+      "step": 1449
+    },
+    {
+      "epoch": 1.7099056603773586,
+      "grad_norm": 0.18950466870485377,
+      "learning_rate": 2.3918741808650067e-05,
+      "loss": 0.4037,
+      "step": 1450
+    },
+    {
+      "epoch": 1.7110849056603774,
+      "grad_norm": 0.18091771814080904,
+      "learning_rate": 2.3896898208824815e-05,
+      "loss": 0.3971,
+      "step": 1451
+    },
+    {
+      "epoch": 1.7122641509433962,
+      "grad_norm": 0.1918875359341879,
+      "learning_rate": 2.3875054608999564e-05,
+      "loss": 0.4321,
+      "step": 1452
+    },
+    {
+      "epoch": 1.713443396226415,
+      "grad_norm": 0.22362746328003025,
+      "learning_rate": 2.3853211009174313e-05,
+      "loss": 0.4113,
+      "step": 1453
+    },
+    {
+      "epoch": 1.7146226415094339,
+      "grad_norm": 0.20385197137158048,
+      "learning_rate": 2.383136740934906e-05,
+      "loss": 0.4047,
+      "step": 1454
+    },
+    {
+      "epoch": 1.7158018867924527,
+      "grad_norm": 1.0658260618080246,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.4241,
+      "step": 1455
+    },
+    {
+      "epoch": 1.7169811320754715,
+      "grad_norm": 0.19222989956220848,
+      "learning_rate": 2.378768020969856e-05,
+      "loss": 0.3902,
+      "step": 1456
+    },
+    {
+      "epoch": 1.7181603773584906,
+      "grad_norm": 0.21539778577701668,
+      "learning_rate": 2.3765836609873308e-05,
+      "loss": 0.4369,
+      "step": 1457
+    },
+    {
+      "epoch": 1.7193396226415094,
+      "grad_norm": 0.20621815884606154,
+      "learning_rate": 2.3743993010048056e-05,
+      "loss": 0.392,
+      "step": 1458
+    },
+    {
+      "epoch": 1.7205188679245285,
+      "grad_norm": 0.21075372356561448,
+      "learning_rate": 2.3722149410222805e-05,
+      "loss": 0.4091,
+      "step": 1459
+    },
+    {
+      "epoch": 1.7216981132075473,
+      "grad_norm": 0.19735794864643422,
+      "learning_rate": 2.3700305810397554e-05,
+      "loss": 0.3958,
+      "step": 1460
+    },
+    {
+      "epoch": 1.7228773584905661,
+      "grad_norm": 0.21410222237106835,
+      "learning_rate": 2.3678462210572302e-05,
+      "loss": 0.4347,
+      "step": 1461
+    },
+    {
+      "epoch": 1.724056603773585,
+      "grad_norm": 0.22392391494642636,
+      "learning_rate": 2.3656618610747054e-05,
+      "loss": 0.4203,
+      "step": 1462
+    },
+    {
+      "epoch": 1.7252358490566038,
+      "grad_norm": 0.2060985140494863,
+      "learning_rate": 2.36347750109218e-05,
+      "loss": 0.3946,
+      "step": 1463
+    },
+    {
+      "epoch": 1.7264150943396226,
+      "grad_norm": 0.21156810384599467,
+      "learning_rate": 2.3612931411096552e-05,
+      "loss": 0.4198,
+      "step": 1464
+    },
+    {
+      "epoch": 1.7275943396226414,
+      "grad_norm": 0.19722883997597312,
+      "learning_rate": 2.3591087811271297e-05,
+      "loss": 0.4196,
+      "step": 1465
+    },
+    {
+      "epoch": 1.7287735849056602,
+      "grad_norm": 0.19580260838624886,
+      "learning_rate": 2.356924421144605e-05,
+      "loss": 0.3909,
+      "step": 1466
+    },
+    {
+      "epoch": 1.7299528301886793,
+      "grad_norm": 0.20504833413937154,
+      "learning_rate": 2.3547400611620795e-05,
+      "loss": 0.4136,
+      "step": 1467
+    },
+    {
+      "epoch": 1.7311320754716981,
+      "grad_norm": 0.20273139910734633,
+      "learning_rate": 2.3525557011795547e-05,
+      "loss": 0.3965,
+      "step": 1468
+    },
+    {
+      "epoch": 1.732311320754717,
+      "grad_norm": 0.20523478001379203,
+      "learning_rate": 2.3503713411970292e-05,
+      "loss": 0.4139,
+      "step": 1469
+    },
+    {
+      "epoch": 1.733490566037736,
+      "grad_norm": 0.2101351054745724,
+      "learning_rate": 2.3481869812145044e-05,
+      "loss": 0.3881,
+      "step": 1470
+    },
+    {
+      "epoch": 1.7346698113207548,
+      "grad_norm": 0.20776461151685613,
+      "learning_rate": 2.3460026212319793e-05,
+      "loss": 0.4159,
+      "step": 1471
+    },
+    {
+      "epoch": 1.7358490566037736,
+      "grad_norm": 0.1994305860459061,
+      "learning_rate": 2.343818261249454e-05,
+      "loss": 0.3949,
+      "step": 1472
+    },
+    {
+      "epoch": 1.7370283018867925,
+      "grad_norm": 0.2244640598032367,
+      "learning_rate": 2.341633901266929e-05,
+      "loss": 0.3907,
+      "step": 1473
+    },
+    {
+      "epoch": 1.7382075471698113,
+      "grad_norm": 0.23530410246930233,
+      "learning_rate": 2.339449541284404e-05,
+      "loss": 0.4055,
+      "step": 1474
+    },
+    {
+      "epoch": 1.7393867924528301,
+      "grad_norm": 0.20423410140901893,
+      "learning_rate": 2.3372651813018787e-05,
+      "loss": 0.413,
+      "step": 1475
+    },
+    {
+      "epoch": 1.740566037735849,
+      "grad_norm": 0.2065556372587226,
+      "learning_rate": 2.3350808213193533e-05,
+      "loss": 0.3868,
+      "step": 1476
+    },
+    {
+      "epoch": 1.7417452830188678,
+      "grad_norm": 0.2434474746184842,
+      "learning_rate": 2.3328964613368285e-05,
+      "loss": 0.4091,
+      "step": 1477
+    },
+    {
+      "epoch": 1.7429245283018868,
+      "grad_norm": 0.20473914281768427,
+      "learning_rate": 2.330712101354303e-05,
+      "loss": 0.4098,
+      "step": 1478
+    },
+    {
+      "epoch": 1.7441037735849056,
+      "grad_norm": 0.21616974967798855,
+      "learning_rate": 2.3285277413717782e-05,
+      "loss": 0.4002,
+      "step": 1479
+    },
+    {
+      "epoch": 1.7452830188679245,
+      "grad_norm": 0.22036684904237563,
+      "learning_rate": 2.326343381389253e-05,
+      "loss": 0.4068,
+      "step": 1480
+    },
+    {
+      "epoch": 1.7464622641509435,
+      "grad_norm": 0.20617290984866213,
+      "learning_rate": 2.324159021406728e-05,
+      "loss": 0.3932,
+      "step": 1481
+    },
+    {
+      "epoch": 1.7476415094339623,
+      "grad_norm": 0.20974506392306264,
+      "learning_rate": 2.321974661424203e-05,
+      "loss": 0.4,
+      "step": 1482
+    },
+    {
+      "epoch": 1.7488207547169812,
+      "grad_norm": 0.22801550528431883,
+      "learning_rate": 2.3197903014416777e-05,
+      "loss": 0.4047,
+      "step": 1483
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.22656697101225418,
+      "learning_rate": 2.3176059414591526e-05,
+      "loss": 0.3943,
+      "step": 1484
+    },
+    {
+      "epoch": 1.7511792452830188,
+      "grad_norm": 0.22832257386816826,
+      "learning_rate": 2.3154215814766274e-05,
+      "loss": 0.4103,
+      "step": 1485
+    },
+    {
+      "epoch": 1.7523584905660377,
+      "grad_norm": 0.22430837670515139,
+      "learning_rate": 2.3132372214941023e-05,
+      "loss": 0.3848,
+      "step": 1486
+    },
+    {
+      "epoch": 1.7535377358490565,
+      "grad_norm": 0.20434050990458644,
+      "learning_rate": 2.3110528615115772e-05,
+      "loss": 0.3919,
+      "step": 1487
+    },
+    {
+      "epoch": 1.7547169811320755,
+      "grad_norm": 0.22351684966590535,
+      "learning_rate": 2.308868501529052e-05,
+      "loss": 0.4014,
+      "step": 1488
+    },
+    {
+      "epoch": 1.7558962264150944,
+      "grad_norm": 0.21474951146643748,
+      "learning_rate": 2.306684141546527e-05,
+      "loss": 0.3921,
+      "step": 1489
+    },
+    {
+      "epoch": 1.7570754716981132,
+      "grad_norm": 0.2237113123227196,
+      "learning_rate": 2.3044997815640018e-05,
+      "loss": 0.3996,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7582547169811322,
+      "grad_norm": 0.19793548112507214,
+      "learning_rate": 2.302315421581477e-05,
+      "loss": 0.4177,
+      "step": 1491
+    },
+    {
+      "epoch": 1.759433962264151,
+      "grad_norm": 0.20270947107841533,
+      "learning_rate": 2.3001310615989515e-05,
+      "loss": 0.4105,
+      "step": 1492
+    },
+    {
+      "epoch": 1.7606132075471699,
+      "grad_norm": 0.24791341691339058,
+      "learning_rate": 2.2979467016164267e-05,
+      "loss": 0.3993,
+      "step": 1493
+    },
+    {
+      "epoch": 1.7617924528301887,
+      "grad_norm": 0.2130022844808386,
+      "learning_rate": 2.2957623416339013e-05,
+      "loss": 0.3906,
+      "step": 1494
+    },
+    {
+      "epoch": 1.7629716981132075,
+      "grad_norm": 0.2092712226312565,
+      "learning_rate": 2.2935779816513765e-05,
+      "loss": 0.4044,
+      "step": 1495
+    },
+    {
+      "epoch": 1.7641509433962264,
+      "grad_norm": 0.263216371009863,
+      "learning_rate": 2.291393621668851e-05,
+      "loss": 0.4385,
+      "step": 1496
+    },
+    {
+      "epoch": 1.7653301886792452,
+      "grad_norm": 0.20902057801833185,
+      "learning_rate": 2.289209261686326e-05,
+      "loss": 0.3902,
+      "step": 1497
+    },
+    {
+      "epoch": 1.766509433962264,
+      "grad_norm": 0.21052058398141946,
+      "learning_rate": 2.2870249017038007e-05,
+      "loss": 0.4068,
+      "step": 1498
+    },
+    {
+      "epoch": 1.767688679245283,
+      "grad_norm": 0.20458078655376435,
+      "learning_rate": 2.2848405417212756e-05,
+      "loss": 0.4008,
+      "step": 1499
+    },
+    {
+      "epoch": 1.7688679245283019,
+      "grad_norm": 0.2180917030455745,
+      "learning_rate": 2.2826561817387508e-05,
+      "loss": 0.4005,
+      "step": 1500
+    },
+    {
+      "epoch": 1.7700471698113207,
+      "grad_norm": 0.20851069625902627,
+      "learning_rate": 2.2804718217562254e-05,
+      "loss": 0.3958,
+      "step": 1501
+    },
+    {
+      "epoch": 1.7712264150943398,
+      "grad_norm": 0.19784525464987718,
+      "learning_rate": 2.2782874617737006e-05,
+      "loss": 0.3942,
+      "step": 1502
+    },
+    {
+      "epoch": 1.7724056603773586,
+      "grad_norm": 0.2088183731165048,
+      "learning_rate": 2.276103101791175e-05,
+      "loss": 0.4085,
+      "step": 1503
+    },
+    {
+      "epoch": 1.7735849056603774,
+      "grad_norm": 0.24997439902701463,
+      "learning_rate": 2.2739187418086503e-05,
+      "loss": 0.4066,
+      "step": 1504
+    },
+    {
+      "epoch": 1.7747641509433962,
+      "grad_norm": 0.2164794235402129,
+      "learning_rate": 2.271734381826125e-05,
+      "loss": 0.3765,
+      "step": 1505
+    },
+    {
+      "epoch": 1.775943396226415,
+      "grad_norm": 0.21510783055759888,
+      "learning_rate": 2.2695500218436e-05,
+      "loss": 0.3987,
+      "step": 1506
+    },
+    {
+      "epoch": 1.7771226415094339,
+      "grad_norm": 0.2046662745714208,
+      "learning_rate": 2.2673656618610746e-05,
+      "loss": 0.4053,
+      "step": 1507
+    },
+    {
+      "epoch": 1.7783018867924527,
+      "grad_norm": 0.23550003675768222,
+      "learning_rate": 2.2651813018785498e-05,
+      "loss": 0.4303,
+      "step": 1508
+    },
+    {
+      "epoch": 1.7794811320754715,
+      "grad_norm": 0.21157469259112704,
+      "learning_rate": 2.2629969418960247e-05,
+      "loss": 0.398,
+      "step": 1509
+    },
+    {
+      "epoch": 1.7806603773584906,
+      "grad_norm": 0.2084734495808188,
+      "learning_rate": 2.2608125819134995e-05,
+      "loss": 0.3975,
+      "step": 1510
+    },
+    {
+      "epoch": 1.7818396226415094,
+      "grad_norm": 0.2104723006232279,
+      "learning_rate": 2.2586282219309744e-05,
+      "loss": 0.3975,
+      "step": 1511
+    },
+    {
+      "epoch": 1.7830188679245285,
+      "grad_norm": 0.19359445118788865,
+      "learning_rate": 2.2564438619484493e-05,
+      "loss": 0.3909,
+      "step": 1512
+    },
+    {
+      "epoch": 1.7841981132075473,
+      "grad_norm": 0.21005278822778925,
+      "learning_rate": 2.254259501965924e-05,
+      "loss": 0.4166,
+      "step": 1513
+    },
+    {
+      "epoch": 1.7853773584905661,
+      "grad_norm": 0.2246212257419986,
+      "learning_rate": 2.252075141983399e-05,
+      "loss": 0.3994,
+      "step": 1514
+    },
+    {
+      "epoch": 1.786556603773585,
+      "grad_norm": 0.20601942643528962,
+      "learning_rate": 2.249890782000874e-05,
+      "loss": 0.4062,
+      "step": 1515
+    },
+    {
+      "epoch": 1.7877358490566038,
+      "grad_norm": 0.20181609798178546,
+      "learning_rate": 2.2477064220183487e-05,
+      "loss": 0.4028,
+      "step": 1516
+    },
+    {
+      "epoch": 1.7889150943396226,
+      "grad_norm": 0.2222602189603544,
+      "learning_rate": 2.2455220620358236e-05,
+      "loss": 0.3927,
+      "step": 1517
+    },
+    {
+      "epoch": 1.7900943396226414,
+      "grad_norm": 0.21515384458873246,
+      "learning_rate": 2.2433377020532985e-05,
+      "loss": 0.4247,
+      "step": 1518
+    },
+    {
+      "epoch": 1.7912735849056602,
+      "grad_norm": 0.20966324701180167,
+      "learning_rate": 2.2411533420707733e-05,
+      "loss": 0.3985,
+      "step": 1519
+    },
+    {
+      "epoch": 1.7924528301886793,
+      "grad_norm": 0.19412588224398103,
+      "learning_rate": 2.2389689820882482e-05,
+      "loss": 0.3852,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7936320754716981,
+      "grad_norm": 0.2019444801709361,
+      "learning_rate": 2.236784622105723e-05,
+      "loss": 0.4048,
+      "step": 1521
+    },
+    {
+      "epoch": 1.794811320754717,
+      "grad_norm": 0.20337374974774441,
+      "learning_rate": 2.234600262123198e-05,
+      "loss": 0.3915,
+      "step": 1522
+    },
+    {
+      "epoch": 1.795990566037736,
+      "grad_norm": 0.19252665139498673,
+      "learning_rate": 2.2324159021406728e-05,
+      "loss": 0.3741,
+      "step": 1523
+    },
+    {
+      "epoch": 1.7971698113207548,
+      "grad_norm": 0.19907054314947462,
+      "learning_rate": 2.2302315421581477e-05,
+      "loss": 0.3988,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7983490566037736,
+      "grad_norm": 0.19789128333416464,
+      "learning_rate": 2.2280471821756226e-05,
+      "loss": 0.3919,
+      "step": 1525
+    },
+    {
+      "epoch": 1.7995283018867925,
+      "grad_norm": 0.1732762625692499,
+      "learning_rate": 2.2258628221930974e-05,
+      "loss": 0.387,
+      "step": 1526
+    },
+    {
+      "epoch": 1.8007075471698113,
+      "grad_norm": 0.22711432154700156,
+      "learning_rate": 2.2236784622105723e-05,
+      "loss": 0.4522,
+      "step": 1527
+    },
+    {
+      "epoch": 1.8018867924528301,
+      "grad_norm": 0.19301680902371104,
+      "learning_rate": 2.2214941022280472e-05,
+      "loss": 0.4315,
+      "step": 1528
+    },
+    {
+      "epoch": 1.803066037735849,
+      "grad_norm": 0.21448855452313392,
+      "learning_rate": 2.2193097422455224e-05,
+      "loss": 0.4218,
+      "step": 1529
+    },
+    {
+      "epoch": 1.8042452830188678,
+      "grad_norm": 0.2082126986304676,
+      "learning_rate": 2.217125382262997e-05,
+      "loss": 0.4155,
+      "step": 1530
+    },
+    {
+      "epoch": 1.8054245283018868,
+      "grad_norm": 0.19179808770549617,
+      "learning_rate": 2.214941022280472e-05,
+      "loss": 0.3827,
+      "step": 1531
+    },
+    {
+      "epoch": 1.8066037735849056,
+      "grad_norm": 0.2113532930907674,
+      "learning_rate": 2.2127566622979467e-05,
+      "loss": 0.399,
+      "step": 1532
+    },
+    {
+      "epoch": 1.8077830188679245,
+      "grad_norm": 0.20428774801449043,
+      "learning_rate": 2.210572302315422e-05,
+      "loss": 0.4019,
+      "step": 1533
+    },
+    {
+      "epoch": 1.8089622641509435,
+      "grad_norm": 0.18933742203501783,
+      "learning_rate": 2.2083879423328964e-05,
+      "loss": 0.3994,
+      "step": 1534
+    },
+    {
+      "epoch": 1.8101415094339623,
+      "grad_norm": 0.20688069488645622,
+      "learning_rate": 2.2062035823503716e-05,
+      "loss": 0.3891,
+      "step": 1535
+    },
+    {
+      "epoch": 1.8113207547169812,
+      "grad_norm": 0.19129061924346183,
+      "learning_rate": 2.2040192223678465e-05,
+      "loss": 0.3733,
+      "step": 1536
+    },
+    {
+      "epoch": 1.8125,
+      "grad_norm": 0.2007187311250329,
+      "learning_rate": 2.2018348623853213e-05,
+      "loss": 0.4044,
+      "step": 1537
+    },
+    {
+      "epoch": 1.8136792452830188,
+      "grad_norm": 0.19396365251608347,
+      "learning_rate": 2.1996505024027962e-05,
+      "loss": 0.3965,
+      "step": 1538
+    },
+    {
+      "epoch": 1.8148584905660377,
+      "grad_norm": 0.2013515280774341,
+      "learning_rate": 2.197466142420271e-05,
+      "loss": 0.3963,
+      "step": 1539
+    },
+    {
+      "epoch": 1.8160377358490565,
+      "grad_norm": 0.20605149149387775,
+      "learning_rate": 2.195281782437746e-05,
+      "loss": 0.3823,
+      "step": 1540
+    },
+    {
+      "epoch": 1.8172169811320755,
+      "grad_norm": 0.1961832230270824,
+      "learning_rate": 2.1930974224552205e-05,
+      "loss": 0.3791,
+      "step": 1541
+    },
+    {
+      "epoch": 1.8183962264150944,
+      "grad_norm": 0.20649728598694303,
+      "learning_rate": 2.1909130624726957e-05,
+      "loss": 0.3939,
+      "step": 1542
+    },
+    {
+      "epoch": 1.8195754716981132,
+      "grad_norm": 0.2121712696306857,
+      "learning_rate": 2.1887287024901702e-05,
+      "loss": 0.3837,
+      "step": 1543
+    },
+    {
+      "epoch": 1.8207547169811322,
+      "grad_norm": 0.2074666643659278,
+      "learning_rate": 2.1865443425076454e-05,
+      "loss": 0.4065,
+      "step": 1544
+    },
+    {
+      "epoch": 1.821933962264151,
+      "grad_norm": 0.20548914744595748,
+      "learning_rate": 2.1843599825251203e-05,
+      "loss": 0.4208,
+      "step": 1545
+    },
+    {
+      "epoch": 1.8231132075471699,
+      "grad_norm": 0.21422907007780323,
+      "learning_rate": 2.182175622542595e-05,
+      "loss": 0.3696,
+      "step": 1546
+    },
+    {
+      "epoch": 1.8242924528301887,
+      "grad_norm": 0.2136705435703334,
+      "learning_rate": 2.17999126256007e-05,
+      "loss": 0.399,
+      "step": 1547
+    },
+    {
+      "epoch": 1.8254716981132075,
+      "grad_norm": 0.2043456317987832,
+      "learning_rate": 2.177806902577545e-05,
+      "loss": 0.3715,
+      "step": 1548
+    },
+    {
+      "epoch": 1.8266509433962264,
+      "grad_norm": 0.20627261938990618,
+      "learning_rate": 2.1756225425950198e-05,
+      "loss": 0.4007,
+      "step": 1549
+    },
+    {
+      "epoch": 1.8278301886792452,
+      "grad_norm": 0.19654856291030567,
+      "learning_rate": 2.1734381826124946e-05,
+      "loss": 0.3822,
+      "step": 1550
+    },
+    {
+      "epoch": 1.829009433962264,
+      "grad_norm": 0.22464656143404804,
+      "learning_rate": 2.1712538226299695e-05,
+      "loss": 0.3985,
+      "step": 1551
+    },
+    {
+      "epoch": 1.830188679245283,
+      "grad_norm": 0.22477894673805532,
+      "learning_rate": 2.1690694626474444e-05,
+      "loss": 0.4252,
+      "step": 1552
+    },
+    {
+      "epoch": 1.8313679245283019,
+      "grad_norm": 0.21168956289578536,
+      "learning_rate": 2.1668851026649192e-05,
+      "loss": 0.3991,
+      "step": 1553
+    },
+    {
+      "epoch": 1.8325471698113207,
+      "grad_norm": 0.25360412688449907,
+      "learning_rate": 2.164700742682394e-05,
+      "loss": 0.4178,
+      "step": 1554
+    },
+    {
+      "epoch": 1.8337264150943398,
+      "grad_norm": 0.23302805966866005,
+      "learning_rate": 2.162516382699869e-05,
+      "loss": 0.4259,
+      "step": 1555
+    },
+    {
+      "epoch": 1.8349056603773586,
+      "grad_norm": 0.20445075711291585,
+      "learning_rate": 2.1603320227173442e-05,
+      "loss": 0.3789,
+      "step": 1556
+    },
+    {
+      "epoch": 1.8360849056603774,
+      "grad_norm": 0.23493088709940918,
+      "learning_rate": 2.1581476627348187e-05,
+      "loss": 0.4046,
+      "step": 1557
+    },
+    {
+      "epoch": 1.8372641509433962,
+      "grad_norm": 0.21635401096282536,
+      "learning_rate": 2.155963302752294e-05,
+      "loss": 0.3913,
+      "step": 1558
+    },
+    {
+      "epoch": 1.838443396226415,
+      "grad_norm": 0.22134483944098965,
+      "learning_rate": 2.1537789427697685e-05,
+      "loss": 0.3795,
+      "step": 1559
+    },
+    {
+      "epoch": 1.8396226415094339,
+      "grad_norm": 0.20321448112354382,
+      "learning_rate": 2.1515945827872437e-05,
+      "loss": 0.4119,
+      "step": 1560
+    },
+    {
+      "epoch": 1.8408018867924527,
+      "grad_norm": 0.2200716581914572,
+      "learning_rate": 2.1494102228047182e-05,
+      "loss": 0.3878,
+      "step": 1561
+    },
+    {
+      "epoch": 1.8419811320754715,
+      "grad_norm": 0.23675683596260863,
+      "learning_rate": 2.147225862822193e-05,
+      "loss": 0.3811,
+      "step": 1562
+    },
+    {
+      "epoch": 1.8431603773584906,
+      "grad_norm": 0.22214066415604247,
+      "learning_rate": 2.145041502839668e-05,
+      "loss": 0.4102,
+      "step": 1563
+    },
+    {
+      "epoch": 1.8443396226415094,
+      "grad_norm": 0.20029961022796033,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.3962,
+      "step": 1564
+    },
+    {
+      "epoch": 1.8455188679245285,
+      "grad_norm": 0.21970748070443055,
+      "learning_rate": 2.140672782874618e-05,
+      "loss": 0.3991,
+      "step": 1565
+    },
+    {
+      "epoch": 1.8466981132075473,
+      "grad_norm": 0.18897125335554477,
+      "learning_rate": 2.1384884228920926e-05,
+      "loss": 0.3969,
+      "step": 1566
+    },
+    {
+      "epoch": 1.8478773584905661,
+      "grad_norm": 0.21228081712606958,
+      "learning_rate": 2.1363040629095678e-05,
+      "loss": 0.4076,
+      "step": 1567
+    },
+    {
+      "epoch": 1.849056603773585,
+      "grad_norm": 0.2009065546039051,
+      "learning_rate": 2.1341197029270423e-05,
+      "loss": 0.3943,
+      "step": 1568
+    },
+    {
+      "epoch": 1.8502358490566038,
+      "grad_norm": 0.22547517049885538,
+      "learning_rate": 2.1319353429445175e-05,
+      "loss": 0.398,
+      "step": 1569
+    },
+    {
+      "epoch": 1.8514150943396226,
+      "grad_norm": 0.19221917804562508,
+      "learning_rate": 2.129750982961992e-05,
+      "loss": 0.3643,
+      "step": 1570
+    },
+    {
+      "epoch": 1.8525943396226414,
+      "grad_norm": 0.1942229995586528,
+      "learning_rate": 2.1275666229794672e-05,
+      "loss": 0.3776,
+      "step": 1571
+    },
+    {
+      "epoch": 1.8537735849056602,
+      "grad_norm": 0.21145914632732032,
+      "learning_rate": 2.1253822629969418e-05,
+      "loss": 0.398,
+      "step": 1572
+    },
+    {
+      "epoch": 1.8549528301886793,
+      "grad_norm": 0.22201952955487372,
+      "learning_rate": 2.123197903014417e-05,
+      "loss": 0.4013,
+      "step": 1573
+    },
+    {
+      "epoch": 1.8561320754716981,
+      "grad_norm": 0.21246981261588418,
+      "learning_rate": 2.121013543031892e-05,
+      "loss": 0.4072,
+      "step": 1574
+    },
+    {
+      "epoch": 1.857311320754717,
+      "grad_norm": 0.2106384418982402,
+      "learning_rate": 2.1188291830493667e-05,
+      "loss": 0.4112,
+      "step": 1575
+    },
+    {
+      "epoch": 1.858490566037736,
+      "grad_norm": 0.22530475602460645,
+      "learning_rate": 2.1166448230668416e-05,
+      "loss": 0.3983,
+      "step": 1576
+    },
+    {
+      "epoch": 1.8596698113207548,
+      "grad_norm": 0.2086546267728454,
+      "learning_rate": 2.1144604630843165e-05,
+      "loss": 0.3827,
+      "step": 1577
+    },
+    {
+      "epoch": 1.8608490566037736,
+      "grad_norm": 0.19778804905671218,
+      "learning_rate": 2.1122761031017913e-05,
+      "loss": 0.3928,
+      "step": 1578
+    },
+    {
+      "epoch": 1.8620283018867925,
+      "grad_norm": 0.2279818365177272,
+      "learning_rate": 2.1100917431192662e-05,
+      "loss": 0.4151,
+      "step": 1579
+    },
+    {
+      "epoch": 1.8632075471698113,
+      "grad_norm": 0.2277852463078636,
+      "learning_rate": 2.107907383136741e-05,
+      "loss": 0.4021,
+      "step": 1580
+    },
+    {
+      "epoch": 1.8643867924528301,
+      "grad_norm": 0.19196365740496807,
+      "learning_rate": 2.105723023154216e-05,
+      "loss": 0.392,
+      "step": 1581
+    },
+    {
+      "epoch": 1.865566037735849,
+      "grad_norm": 0.20779369191304728,
+      "learning_rate": 2.1035386631716908e-05,
+      "loss": 0.4125,
+      "step": 1582
+    },
+    {
+      "epoch": 1.8667452830188678,
+      "grad_norm": 0.22002521619935264,
+      "learning_rate": 2.1013543031891657e-05,
+      "loss": 0.4007,
+      "step": 1583
+    },
+    {
+      "epoch": 1.8679245283018868,
+      "grad_norm": 0.22456697258155267,
+      "learning_rate": 2.0991699432066405e-05,
+      "loss": 0.3858,
+      "step": 1584
+    },
+    {
+      "epoch": 1.8691037735849056,
+      "grad_norm": 0.2146245680384488,
+      "learning_rate": 2.0969855832241154e-05,
+      "loss": 0.421,
+      "step": 1585
+    },
+    {
+      "epoch": 1.8702830188679245,
+      "grad_norm": 0.21393861079725812,
+      "learning_rate": 2.0948012232415903e-05,
+      "loss": 0.4015,
+      "step": 1586
+    },
+    {
+      "epoch": 1.8714622641509435,
+      "grad_norm": 0.22020224434618496,
+      "learning_rate": 2.092616863259065e-05,
+      "loss": 0.423,
+      "step": 1587
+    },
+    {
+      "epoch": 1.8726415094339623,
+      "grad_norm": 0.22175408424061566,
+      "learning_rate": 2.09043250327654e-05,
+      "loss": 0.4139,
+      "step": 1588
+    },
+    {
+      "epoch": 1.8738207547169812,
+      "grad_norm": 0.19633285391246932,
+      "learning_rate": 2.088248143294015e-05,
+      "loss": 0.3668,
+      "step": 1589
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.205308912329877,
+      "learning_rate": 2.0860637833114898e-05,
+      "loss": 0.403,
+      "step": 1590
+    },
+    {
+      "epoch": 1.8761792452830188,
+      "grad_norm": 0.21606153059172256,
+      "learning_rate": 2.0838794233289646e-05,
+      "loss": 0.4002,
+      "step": 1591
+    },
+    {
+      "epoch": 1.8773584905660377,
+      "grad_norm": 0.1899179701021968,
+      "learning_rate": 2.0816950633464395e-05,
+      "loss": 0.3924,
+      "step": 1592
+    },
+    {
+      "epoch": 1.8785377358490565,
+      "grad_norm": 0.20261049557251687,
+      "learning_rate": 2.0795107033639144e-05,
+      "loss": 0.4125,
+      "step": 1593
+    },
+    {
+      "epoch": 1.8797169811320755,
+      "grad_norm": 0.20293487922999356,
+      "learning_rate": 2.0773263433813896e-05,
+      "loss": 0.4102,
+      "step": 1594
+    },
+    {
+      "epoch": 1.8808962264150944,
+      "grad_norm": 0.1934891306779378,
+      "learning_rate": 2.075141983398864e-05,
+      "loss": 0.4132,
+      "step": 1595
+    },
+    {
+      "epoch": 1.8820754716981132,
+      "grad_norm": 0.2067892273858679,
+      "learning_rate": 2.0729576234163393e-05,
+      "loss": 0.395,
+      "step": 1596
+    },
+    {
+      "epoch": 1.8832547169811322,
+      "grad_norm": 0.20966082701920696,
+      "learning_rate": 2.070773263433814e-05,
+      "loss": 0.4319,
+      "step": 1597
+    },
+    {
+      "epoch": 1.884433962264151,
+      "grad_norm": 0.2023642889732131,
+      "learning_rate": 2.068588903451289e-05,
+      "loss": 0.414,
+      "step": 1598
+    },
+    {
+      "epoch": 1.8856132075471699,
+      "grad_norm": 0.21632959617934547,
+      "learning_rate": 2.0664045434687636e-05,
+      "loss": 0.4203,
+      "step": 1599
+    },
+    {
+      "epoch": 1.8867924528301887,
+      "grad_norm": 0.1898871133764119,
+      "learning_rate": 2.0642201834862388e-05,
+      "loss": 0.3835,
+      "step": 1600
+    },
+    {
+      "epoch": 1.8879716981132075,
+      "grad_norm": 0.23625385417428868,
+      "learning_rate": 2.0620358235037133e-05,
+      "loss": 0.3978,
+      "step": 1601
+    },
+    {
+      "epoch": 1.8891509433962264,
+      "grad_norm": 0.20552134771346262,
+      "learning_rate": 2.0598514635211885e-05,
+      "loss": 0.3855,
+      "step": 1602
+    },
+    {
+      "epoch": 1.8903301886792452,
+      "grad_norm": 0.38726867966688044,
+      "learning_rate": 2.0576671035386634e-05,
+      "loss": 0.3951,
+      "step": 1603
+    },
+    {
+      "epoch": 1.891509433962264,
+      "grad_norm": 0.2020624662362772,
+      "learning_rate": 2.055482743556138e-05,
+      "loss": 0.3955,
+      "step": 1604
+    },
+    {
+      "epoch": 1.892688679245283,
+      "grad_norm": 0.2178365165474841,
+      "learning_rate": 2.053298383573613e-05,
+      "loss": 0.4032,
+      "step": 1605
+    },
+    {
+      "epoch": 1.8938679245283019,
+      "grad_norm": 0.21370177380967498,
+      "learning_rate": 2.0511140235910877e-05,
+      "loss": 0.399,
+      "step": 1606
+    },
+    {
+      "epoch": 1.8950471698113207,
+      "grad_norm": 0.21775720209538,
+      "learning_rate": 2.048929663608563e-05,
+      "loss": 0.4073,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8962264150943398,
+      "grad_norm": 0.194469358540329,
+      "learning_rate": 2.0467453036260374e-05,
+      "loss": 0.398,
+      "step": 1608
+    },
+    {
+      "epoch": 1.8974056603773586,
+      "grad_norm": 0.36591913428450845,
+      "learning_rate": 2.0445609436435126e-05,
+      "loss": 0.4026,
+      "step": 1609
+    },
+    {
+      "epoch": 1.8985849056603774,
+      "grad_norm": 0.21431988537750618,
+      "learning_rate": 2.0423765836609875e-05,
+      "loss": 0.3915,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8997641509433962,
+      "grad_norm": 0.19199134665880424,
+      "learning_rate": 2.0401922236784624e-05,
+      "loss": 0.4009,
+      "step": 1611
+    },
+    {
+      "epoch": 1.900943396226415,
+      "grad_norm": 0.2103758448999669,
+      "learning_rate": 2.0380078636959372e-05,
+      "loss": 0.4091,
+      "step": 1612
+    },
+    {
+      "epoch": 1.9021226415094339,
+      "grad_norm": 0.2214541790794102,
+      "learning_rate": 2.035823503713412e-05,
+      "loss": 0.4087,
+      "step": 1613
+    },
+    {
+      "epoch": 1.9033018867924527,
+      "grad_norm": 0.19408548419668703,
+      "learning_rate": 2.033639143730887e-05,
+      "loss": 0.3947,
+      "step": 1614
+    },
+    {
+      "epoch": 1.9044811320754715,
+      "grad_norm": 0.209886563659498,
+      "learning_rate": 2.031454783748362e-05,
+      "loss": 0.4028,
+      "step": 1615
+    },
+    {
+      "epoch": 1.9056603773584906,
+      "grad_norm": 0.2012601685244372,
+      "learning_rate": 2.0292704237658367e-05,
+      "loss": 0.3897,
+      "step": 1616
+    },
+    {
+      "epoch": 1.9068396226415094,
+      "grad_norm": 0.19527633187126142,
+      "learning_rate": 2.0270860637833116e-05,
+      "loss": 0.3893,
+      "step": 1617
+    },
+    {
+      "epoch": 1.9080188679245285,
+      "grad_norm": 0.20622715485954643,
+      "learning_rate": 2.0249017038007864e-05,
+      "loss": 0.3935,
+      "step": 1618
+    },
+    {
+      "epoch": 1.9091981132075473,
+      "grad_norm": 0.19186136344177357,
+      "learning_rate": 2.0227173438182613e-05,
+      "loss": 0.3807,
+      "step": 1619
+    },
+    {
+      "epoch": 1.9103773584905661,
+      "grad_norm": 0.2154410770974301,
+      "learning_rate": 2.0205329838357362e-05,
+      "loss": 0.4128,
+      "step": 1620
+    },
+    {
+      "epoch": 1.911556603773585,
+      "grad_norm": 0.1985202080071459,
+      "learning_rate": 2.018348623853211e-05,
+      "loss": 0.4287,
+      "step": 1621
+    },
+    {
+      "epoch": 1.9127358490566038,
+      "grad_norm": 0.20591567476868125,
+      "learning_rate": 2.016164263870686e-05,
+      "loss": 0.4174,
+      "step": 1622
+    },
+    {
+      "epoch": 1.9139150943396226,
+      "grad_norm": 0.23335640805344268,
+      "learning_rate": 2.013979903888161e-05,
+      "loss": 0.3973,
+      "step": 1623
+    },
+    {
+      "epoch": 1.9150943396226414,
+      "grad_norm": 0.2066975953120477,
+      "learning_rate": 2.0117955439056357e-05,
+      "loss": 0.398,
+      "step": 1624
+    },
+    {
+      "epoch": 1.9162735849056602,
+      "grad_norm": 0.2001667616828653,
+      "learning_rate": 2.009611183923111e-05,
+      "loss": 0.3861,
+      "step": 1625
+    },
+    {
+      "epoch": 1.9174528301886793,
+      "grad_norm": 0.2428732458903274,
+      "learning_rate": 2.0074268239405854e-05,
+      "loss": 0.4229,
+      "step": 1626
+    },
+    {
+      "epoch": 1.9186320754716981,
+      "grad_norm": 0.21139183924728439,
+      "learning_rate": 2.0052424639580603e-05,
+      "loss": 0.3957,
+      "step": 1627
+    },
+    {
+      "epoch": 1.919811320754717,
+      "grad_norm": 0.22953740980107015,
+      "learning_rate": 2.003058103975535e-05,
+      "loss": 0.3927,
+      "step": 1628
+    },
+    {
+      "epoch": 1.920990566037736,
+      "grad_norm": 0.23982171709172134,
+      "learning_rate": 2.00087374399301e-05,
+      "loss": 0.4091,
+      "step": 1629
+    },
+    {
+      "epoch": 1.9221698113207548,
+      "grad_norm": 0.2130063022476788,
+      "learning_rate": 1.9986893840104852e-05,
+      "loss": 0.4019,
+      "step": 1630
+    },
+    {
+      "epoch": 1.9233490566037736,
+      "grad_norm": 0.22370974353676495,
+      "learning_rate": 1.9965050240279598e-05,
+      "loss": 0.3868,
+      "step": 1631
+    },
+    {
+      "epoch": 1.9245283018867925,
+      "grad_norm": 0.23097425035633193,
+      "learning_rate": 1.994320664045435e-05,
+      "loss": 0.4306,
+      "step": 1632
+    },
+    {
+      "epoch": 1.9257075471698113,
+      "grad_norm": 0.1913493884357088,
+      "learning_rate": 1.9921363040629095e-05,
+      "loss": 0.3616,
+      "step": 1633
+    },
+    {
+      "epoch": 1.9268867924528301,
+      "grad_norm": 0.2150554242948543,
+      "learning_rate": 1.9899519440803847e-05,
+      "loss": 0.4046,
+      "step": 1634
+    },
+    {
+      "epoch": 1.928066037735849,
+      "grad_norm": 0.24631350035479818,
+      "learning_rate": 1.9877675840978592e-05,
+      "loss": 0.4321,
+      "step": 1635
+    },
+    {
+      "epoch": 1.9292452830188678,
+      "grad_norm": 0.19557510232208575,
+      "learning_rate": 1.9855832241153344e-05,
+      "loss": 0.4035,
+      "step": 1636
+    },
+    {
+      "epoch": 1.9304245283018868,
+      "grad_norm": 0.21253158099460714,
+      "learning_rate": 1.983398864132809e-05,
+      "loss": 0.4095,
+      "step": 1637
+    },
+    {
+      "epoch": 1.9316037735849056,
+      "grad_norm": 0.1987825015108815,
+      "learning_rate": 1.9812145041502842e-05,
+      "loss": 0.4184,
+      "step": 1638
+    },
+    {
+      "epoch": 1.9327830188679245,
+      "grad_norm": 0.21741882979098168,
+      "learning_rate": 1.979030144167759e-05,
+      "loss": 0.3799,
+      "step": 1639
+    },
+    {
+      "epoch": 1.9339622641509435,
+      "grad_norm": 0.19486399493657705,
+      "learning_rate": 1.976845784185234e-05,
+      "loss": 0.3781,
+      "step": 1640
+    },
+    {
+      "epoch": 1.9351415094339623,
+      "grad_norm": 0.2148531866351998,
+      "learning_rate": 1.9746614242027088e-05,
+      "loss": 0.4013,
+      "step": 1641
+    },
+    {
+      "epoch": 1.9363207547169812,
+      "grad_norm": 0.20782578344216127,
+      "learning_rate": 1.9724770642201837e-05,
+      "loss": 0.4021,
+      "step": 1642
+    },
+    {
+      "epoch": 1.9375,
+      "grad_norm": 0.2183171403917148,
+      "learning_rate": 1.9702927042376585e-05,
+      "loss": 0.4049,
+      "step": 1643
+    },
+    {
+      "epoch": 1.9386792452830188,
+      "grad_norm": 0.19588431302381387,
+      "learning_rate": 1.9681083442551334e-05,
+      "loss": 0.4029,
+      "step": 1644
+    },
+    {
+      "epoch": 1.9398584905660377,
+      "grad_norm": 0.23029037710961614,
+      "learning_rate": 1.9659239842726083e-05,
+      "loss": 0.4116,
+      "step": 1645
+    },
+    {
+      "epoch": 1.9410377358490565,
+      "grad_norm": 0.20695984614246754,
+      "learning_rate": 1.963739624290083e-05,
+      "loss": 0.414,
+      "step": 1646
+    },
+    {
+      "epoch": 1.9422169811320755,
+      "grad_norm": 0.21307995639092164,
+      "learning_rate": 1.961555264307558e-05,
+      "loss": 0.4106,
+      "step": 1647
+    },
+    {
+      "epoch": 1.9433962264150944,
+      "grad_norm": 0.2658699558792974,
+      "learning_rate": 1.959370904325033e-05,
+      "loss": 0.4064,
+      "step": 1648
+    },
+    {
+      "epoch": 1.9445754716981132,
+      "grad_norm": 0.198071540706148,
+      "learning_rate": 1.9571865443425077e-05,
+      "loss": 0.3964,
+      "step": 1649
+    },
+    {
+      "epoch": 1.9457547169811322,
+      "grad_norm": 0.21927532264886201,
+      "learning_rate": 1.9550021843599826e-05,
+      "loss": 0.4182,
+      "step": 1650
+    },
+    {
+      "epoch": 1.946933962264151,
+      "grad_norm": 0.22195498440662775,
+      "learning_rate": 1.9528178243774575e-05,
+      "loss": 0.3912,
+      "step": 1651
+    },
+    {
+      "epoch": 1.9481132075471699,
+      "grad_norm": 0.2020787610361101,
+      "learning_rate": 1.9506334643949324e-05,
+      "loss": 0.3681,
+      "step": 1652
+    },
+    {
+      "epoch": 1.9492924528301887,
+      "grad_norm": 0.18532069275779592,
+      "learning_rate": 1.9484491044124072e-05,
+      "loss": 0.4044,
+      "step": 1653
+    },
+    {
+      "epoch": 1.9504716981132075,
+      "grad_norm": 0.21125258707316552,
+      "learning_rate": 1.946264744429882e-05,
+      "loss": 0.4221,
+      "step": 1654
+    },
+    {
+      "epoch": 1.9516509433962264,
+      "grad_norm": 0.20753870772639366,
+      "learning_rate": 1.944080384447357e-05,
+      "loss": 0.4198,
+      "step": 1655
+    },
+    {
+      "epoch": 1.9528301886792452,
+      "grad_norm": 0.19687745728149714,
+      "learning_rate": 1.9418960244648318e-05,
+      "loss": 0.3878,
+      "step": 1656
+    },
+    {
+      "epoch": 1.954009433962264,
+      "grad_norm": 0.20195557839577002,
+      "learning_rate": 1.9397116644823067e-05,
+      "loss": 0.3979,
+      "step": 1657
+    },
+    {
+      "epoch": 1.955188679245283,
+      "grad_norm": 0.2139871090644139,
+      "learning_rate": 1.9375273044997816e-05,
+      "loss": 0.3914,
+      "step": 1658
+    },
+    {
+      "epoch": 1.9563679245283019,
+      "grad_norm": 0.18549967152745428,
+      "learning_rate": 1.9353429445172568e-05,
+      "loss": 0.4074,
+      "step": 1659
+    },
+    {
+      "epoch": 1.9575471698113207,
+      "grad_norm": 0.22119924685121353,
+      "learning_rate": 1.9331585845347313e-05,
+      "loss": 0.3919,
+      "step": 1660
+    },
+    {
+      "epoch": 1.9587264150943398,
+      "grad_norm": 0.211181581575686,
+      "learning_rate": 1.9309742245522065e-05,
+      "loss": 0.3987,
+      "step": 1661
+    },
+    {
+      "epoch": 1.9599056603773586,
+      "grad_norm": 0.21065377826032625,
+      "learning_rate": 1.928789864569681e-05,
+      "loss": 0.4115,
+      "step": 1662
+    },
+    {
+      "epoch": 1.9610849056603774,
+      "grad_norm": 0.18447099099744121,
+      "learning_rate": 1.9266055045871563e-05,
+      "loss": 0.3915,
+      "step": 1663
+    },
+    {
+      "epoch": 1.9622641509433962,
+      "grad_norm": 0.20044313449266218,
+      "learning_rate": 1.9244211446046308e-05,
+      "loss": 0.3918,
+      "step": 1664
+    },
+    {
+      "epoch": 1.963443396226415,
+      "grad_norm": 0.21865374842225674,
+      "learning_rate": 1.922236784622106e-05,
+      "loss": 0.4246,
+      "step": 1665
+    },
+    {
+      "epoch": 1.9646226415094339,
+      "grad_norm": 0.2094733080347252,
+      "learning_rate": 1.9200524246395805e-05,
+      "loss": 0.3963,
+      "step": 1666
+    },
+    {
+      "epoch": 1.9658018867924527,
+      "grad_norm": 0.18399958241896142,
+      "learning_rate": 1.9178680646570557e-05,
+      "loss": 0.3939,
+      "step": 1667
+    },
+    {
+      "epoch": 1.9669811320754715,
+      "grad_norm": 0.2183878382204506,
+      "learning_rate": 1.9156837046745306e-05,
+      "loss": 0.3995,
+      "step": 1668
+    },
+    {
+      "epoch": 1.9681603773584906,
+      "grad_norm": 0.22150985881853558,
+      "learning_rate": 1.913499344692005e-05,
+      "loss": 0.3913,
+      "step": 1669
+    },
+    {
+      "epoch": 1.9693396226415094,
+      "grad_norm": 0.1957248821764151,
+      "learning_rate": 1.9113149847094803e-05,
+      "loss": 0.3941,
+      "step": 1670
+    },
+    {
+      "epoch": 1.9705188679245285,
+      "grad_norm": 0.2060468245529492,
+      "learning_rate": 1.909130624726955e-05,
+      "loss": 0.4096,
+      "step": 1671
+    },
+    {
+      "epoch": 1.9716981132075473,
+      "grad_norm": 0.20917493929197786,
+      "learning_rate": 1.90694626474443e-05,
+      "loss": 0.3953,
+      "step": 1672
+    },
+    {
+      "epoch": 1.9728773584905661,
+      "grad_norm": 0.2263180229485947,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 0.4123,
+      "step": 1673
+    },
+    {
+      "epoch": 1.974056603773585,
+      "grad_norm": 0.21270934328617838,
+      "learning_rate": 1.9025775447793798e-05,
+      "loss": 0.4044,
+      "step": 1674
+    },
+    {
+      "epoch": 1.9752358490566038,
+      "grad_norm": 0.20252691856757646,
+      "learning_rate": 1.9003931847968547e-05,
+      "loss": 0.396,
+      "step": 1675
+    },
+    {
+      "epoch": 1.9764150943396226,
+      "grad_norm": 0.1964744360321122,
+      "learning_rate": 1.8982088248143296e-05,
+      "loss": 0.3971,
+      "step": 1676
+    },
+    {
+      "epoch": 1.9775943396226414,
+      "grad_norm": 0.21474961681825946,
+      "learning_rate": 1.8960244648318044e-05,
+      "loss": 0.4146,
+      "step": 1677
+    },
+    {
+      "epoch": 1.9787735849056602,
+      "grad_norm": 0.2354095902158232,
+      "learning_rate": 1.8938401048492793e-05,
+      "loss": 0.4259,
+      "step": 1678
+    },
+    {
+      "epoch": 1.9799528301886793,
+      "grad_norm": 0.2067186549862809,
+      "learning_rate": 1.891655744866754e-05,
+      "loss": 0.4019,
+      "step": 1679
+    },
+    {
+      "epoch": 1.9811320754716981,
+      "grad_norm": 0.19797616142985944,
+      "learning_rate": 1.889471384884229e-05,
+      "loss": 0.4018,
+      "step": 1680
+    },
+    {
+      "epoch": 1.982311320754717,
+      "grad_norm": 0.2053507310649329,
+      "learning_rate": 1.887287024901704e-05,
+      "loss": 0.4115,
+      "step": 1681
+    },
+    {
+      "epoch": 1.983490566037736,
+      "grad_norm": 0.19418969594670255,
+      "learning_rate": 1.8851026649191788e-05,
+      "loss": 0.391,
+      "step": 1682
+    },
+    {
+      "epoch": 1.9846698113207548,
+      "grad_norm": 0.19907069543560957,
+      "learning_rate": 1.8829183049366536e-05,
+      "loss": 0.3957,
+      "step": 1683
+    },
+    {
+      "epoch": 1.9858490566037736,
+      "grad_norm": 0.20350674207067418,
+      "learning_rate": 1.8807339449541285e-05,
+      "loss": 0.3985,
+      "step": 1684
+    },
+    {
+      "epoch": 1.9870283018867925,
+      "grad_norm": 0.23243007534205815,
+      "learning_rate": 1.8785495849716034e-05,
+      "loss": 0.4217,
+      "step": 1685
+    },
+    {
+      "epoch": 1.9882075471698113,
+      "grad_norm": 0.18616877757469605,
+      "learning_rate": 1.8763652249890783e-05,
+      "loss": 0.407,
+      "step": 1686
+    },
+    {
+      "epoch": 1.9893867924528301,
+      "grad_norm": 0.1979274026555626,
+      "learning_rate": 1.874180865006553e-05,
+      "loss": 0.3789,
+      "step": 1687
+    },
+    {
+      "epoch": 1.990566037735849,
+      "grad_norm": 0.20719228098270076,
+      "learning_rate": 1.8719965050240283e-05,
+      "loss": 0.4043,
+      "step": 1688
+    },
+    {
+      "epoch": 1.9917452830188678,
+      "grad_norm": 0.1928725552239057,
+      "learning_rate": 1.869812145041503e-05,
+      "loss": 0.4058,
+      "step": 1689
+    },
+    {
+      "epoch": 1.9929245283018868,
+      "grad_norm": 0.19134329418298013,
+      "learning_rate": 1.8676277850589777e-05,
+      "loss": 0.391,
+      "step": 1690
+    },
+    {
+      "epoch": 1.9941037735849056,
+      "grad_norm": 0.18151531899713402,
+      "learning_rate": 1.8654434250764526e-05,
+      "loss": 0.3892,
+      "step": 1691
+    },
+    {
+      "epoch": 1.9952830188679245,
+      "grad_norm": 0.19966818216985052,
+      "learning_rate": 1.8632590650939275e-05,
+      "loss": 0.3843,
+      "step": 1692
+    },
+    {
+      "epoch": 1.9964622641509435,
+      "grad_norm": 0.1958745330942674,
+      "learning_rate": 1.8610747051114023e-05,
+      "loss": 0.396,
+      "step": 1693
+    },
+    {
+      "epoch": 1.9976415094339623,
+      "grad_norm": 0.18924304320808616,
+      "learning_rate": 1.8588903451288772e-05,
+      "loss": 0.3945,
+      "step": 1694
+    },
+    {
+      "epoch": 1.9988207547169812,
+      "grad_norm": 0.18800931090130785,
+      "learning_rate": 1.856705985146352e-05,
+      "loss": 0.4044,
+      "step": 1695
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.1917054945829851,
+      "learning_rate": 1.854521625163827e-05,
+      "loss": 0.3876,
+      "step": 1696
+    },
+    {
+      "epoch": 2.001179245283019,
+      "grad_norm": 0.28739258327657086,
+      "learning_rate": 1.852337265181302e-05,
+      "loss": 0.3218,
+      "step": 1697
+    },
+    {
+      "epoch": 2.0023584905660377,
+      "grad_norm": 0.22863474421012883,
+      "learning_rate": 1.8501529051987767e-05,
+      "loss": 0.3234,
+      "step": 1698
+    },
+    {
+      "epoch": 2.0035377358490565,
+      "grad_norm": 0.2166975811341777,
+      "learning_rate": 1.847968545216252e-05,
+      "loss": 0.3377,
+      "step": 1699
+    },
+    {
+      "epoch": 2.0047169811320753,
+      "grad_norm": 0.2843403360456305,
+      "learning_rate": 1.8457841852337264e-05,
+      "loss": 0.3052,
+      "step": 1700
+    },
+    {
+      "epoch": 2.005896226415094,
+      "grad_norm": 0.2613010373099317,
+      "learning_rate": 1.8435998252512016e-05,
+      "loss": 0.3157,
+      "step": 1701
+    },
+    {
+      "epoch": 2.0070754716981134,
+      "grad_norm": 0.20464704840448744,
+      "learning_rate": 1.841415465268676e-05,
+      "loss": 0.3056,
+      "step": 1702
+    },
+    {
+      "epoch": 2.0082547169811322,
+      "grad_norm": 0.2596334725243723,
+      "learning_rate": 1.8392311052861514e-05,
+      "loss": 0.3032,
+      "step": 1703
+    },
+    {
+      "epoch": 2.009433962264151,
+      "grad_norm": 0.24867259148704193,
+      "learning_rate": 1.8370467453036262e-05,
+      "loss": 0.3027,
+      "step": 1704
+    },
+    {
+      "epoch": 2.01061320754717,
+      "grad_norm": 0.21580446989316396,
+      "learning_rate": 1.834862385321101e-05,
+      "loss": 0.3124,
+      "step": 1705
+    },
+    {
+      "epoch": 2.0117924528301887,
+      "grad_norm": 0.23635955131371922,
+      "learning_rate": 1.832678025338576e-05,
+      "loss": 0.3212,
+      "step": 1706
+    },
+    {
+      "epoch": 2.0129716981132075,
+      "grad_norm": 0.2223064872994047,
+      "learning_rate": 1.830493665356051e-05,
+      "loss": 0.3177,
+      "step": 1707
+    },
+    {
+      "epoch": 2.0141509433962264,
+      "grad_norm": 0.21402955079134658,
+      "learning_rate": 1.8283093053735257e-05,
+      "loss": 0.3209,
+      "step": 1708
+    },
+    {
+      "epoch": 2.015330188679245,
+      "grad_norm": 0.19209462085633416,
+      "learning_rate": 1.8261249453910006e-05,
+      "loss": 0.309,
+      "step": 1709
+    },
+    {
+      "epoch": 2.016509433962264,
+      "grad_norm": 0.20045440086764485,
+      "learning_rate": 1.8239405854084755e-05,
+      "loss": 0.3009,
+      "step": 1710
+    },
+    {
+      "epoch": 2.017688679245283,
+      "grad_norm": 0.2175309625966113,
+      "learning_rate": 1.82175622542595e-05,
+      "loss": 0.3028,
+      "step": 1711
+    },
+    {
+      "epoch": 2.018867924528302,
+      "grad_norm": 0.2072991308303101,
+      "learning_rate": 1.8195718654434252e-05,
+      "loss": 0.2904,
+      "step": 1712
+    },
+    {
+      "epoch": 2.020047169811321,
+      "grad_norm": 0.20117897395288753,
+      "learning_rate": 1.8173875054609e-05,
+      "loss": 0.3023,
+      "step": 1713
+    },
+    {
+      "epoch": 2.0212264150943398,
+      "grad_norm": 0.21121246176832634,
+      "learning_rate": 1.815203145478375e-05,
+      "loss": 0.3045,
+      "step": 1714
+    },
+    {
+      "epoch": 2.0224056603773586,
+      "grad_norm": 0.22072384697006373,
+      "learning_rate": 1.8130187854958498e-05,
+      "loss": 0.3093,
+      "step": 1715
+    },
+    {
+      "epoch": 2.0235849056603774,
+      "grad_norm": 0.200343406923079,
+      "learning_rate": 1.8108344255133247e-05,
+      "loss": 0.3092,
+      "step": 1716
+    },
+    {
+      "epoch": 2.0247641509433962,
+      "grad_norm": 0.2019336549219962,
+      "learning_rate": 1.8086500655307995e-05,
+      "loss": 0.3041,
+      "step": 1717
+    },
+    {
+      "epoch": 2.025943396226415,
+      "grad_norm": 0.21543705813239897,
+      "learning_rate": 1.8064657055482744e-05,
+      "loss": 0.3354,
+      "step": 1718
+    },
+    {
+      "epoch": 2.027122641509434,
+      "grad_norm": 0.2028198166375768,
+      "learning_rate": 1.8042813455657493e-05,
+      "loss": 0.3213,
+      "step": 1719
+    },
+    {
+      "epoch": 2.0283018867924527,
+      "grad_norm": 0.22084474678865051,
+      "learning_rate": 1.802096985583224e-05,
+      "loss": 0.3078,
+      "step": 1720
+    },
+    {
+      "epoch": 2.0294811320754715,
+      "grad_norm": 0.19784830227795758,
+      "learning_rate": 1.799912625600699e-05,
+      "loss": 0.3198,
+      "step": 1721
+    },
+    {
+      "epoch": 2.0306603773584904,
+      "grad_norm": 0.18964819460109433,
+      "learning_rate": 1.797728265618174e-05,
+      "loss": 0.3049,
+      "step": 1722
+    },
+    {
+      "epoch": 2.0318396226415096,
+      "grad_norm": 0.20640081012699119,
+      "learning_rate": 1.7955439056356488e-05,
+      "loss": 0.3048,
+      "step": 1723
+    },
+    {
+      "epoch": 2.0330188679245285,
+      "grad_norm": 0.2019428063371375,
+      "learning_rate": 1.793359545653124e-05,
+      "loss": 0.3025,
+      "step": 1724
+    },
+    {
+      "epoch": 2.0341981132075473,
+      "grad_norm": 0.18740484920131212,
+      "learning_rate": 1.7911751856705985e-05,
+      "loss": 0.2932,
+      "step": 1725
+    },
+    {
+      "epoch": 2.035377358490566,
+      "grad_norm": 0.19394496325598373,
+      "learning_rate": 1.7889908256880737e-05,
+      "loss": 0.3207,
+      "step": 1726
+    },
+    {
+      "epoch": 2.036556603773585,
+      "grad_norm": 0.18542495284974972,
+      "learning_rate": 1.7868064657055482e-05,
+      "loss": 0.291,
+      "step": 1727
+    },
+    {
+      "epoch": 2.0377358490566038,
+      "grad_norm": 0.19858580352195107,
+      "learning_rate": 1.7846221057230235e-05,
+      "loss": 0.3123,
+      "step": 1728
+    },
+    {
+      "epoch": 2.0389150943396226,
+      "grad_norm": 0.21045103075964564,
+      "learning_rate": 1.782437745740498e-05,
+      "loss": 0.3242,
+      "step": 1729
+    },
+    {
+      "epoch": 2.0400943396226414,
+      "grad_norm": 0.19540156673718334,
+      "learning_rate": 1.7802533857579732e-05,
+      "loss": 0.3177,
+      "step": 1730
+    },
+    {
+      "epoch": 2.0412735849056602,
+      "grad_norm": 0.19082698486801683,
+      "learning_rate": 1.7780690257754477e-05,
+      "loss": 0.3121,
+      "step": 1731
+    },
+    {
+      "epoch": 2.042452830188679,
+      "grad_norm": 0.2089097439921359,
+      "learning_rate": 1.775884665792923e-05,
+      "loss": 0.3062,
+      "step": 1732
+    },
+    {
+      "epoch": 2.043632075471698,
+      "grad_norm": 0.19962499856514407,
+      "learning_rate": 1.7737003058103978e-05,
+      "loss": 0.3069,
+      "step": 1733
+    },
+    {
+      "epoch": 2.044811320754717,
+      "grad_norm": 0.21963125314590395,
+      "learning_rate": 1.7715159458278723e-05,
+      "loss": 0.3268,
+      "step": 1734
+    },
+    {
+      "epoch": 2.045990566037736,
+      "grad_norm": 0.22186599460850634,
+      "learning_rate": 1.7693315858453475e-05,
+      "loss": 0.3113,
+      "step": 1735
+    },
+    {
+      "epoch": 2.047169811320755,
+      "grad_norm": 0.19280744359286997,
+      "learning_rate": 1.767147225862822e-05,
+      "loss": 0.3074,
+      "step": 1736
+    },
+    {
+      "epoch": 2.0483490566037736,
+      "grad_norm": 0.18485834220820893,
+      "learning_rate": 1.7649628658802973e-05,
+      "loss": 0.3017,
+      "step": 1737
+    },
+    {
+      "epoch": 2.0495283018867925,
+      "grad_norm": 0.20033118870048802,
+      "learning_rate": 1.7627785058977718e-05,
+      "loss": 0.301,
+      "step": 1738
+    },
+    {
+      "epoch": 2.0507075471698113,
+      "grad_norm": 0.20079509812392987,
+      "learning_rate": 1.760594145915247e-05,
+      "loss": 0.3046,
+      "step": 1739
+    },
+    {
+      "epoch": 2.05188679245283,
+      "grad_norm": 0.20890232769959055,
+      "learning_rate": 1.7584097859327215e-05,
+      "loss": 0.318,
+      "step": 1740
+    },
+    {
+      "epoch": 2.053066037735849,
+      "grad_norm": 0.20200718083733624,
+      "learning_rate": 1.7562254259501968e-05,
+      "loss": 0.3074,
+      "step": 1741
+    },
+    {
+      "epoch": 2.0542452830188678,
+      "grad_norm": 0.20020071657861455,
+      "learning_rate": 1.7540410659676716e-05,
+      "loss": 0.3163,
+      "step": 1742
+    },
+    {
+      "epoch": 2.0554245283018866,
+      "grad_norm": 0.2136194243462434,
+      "learning_rate": 1.7518567059851465e-05,
+      "loss": 0.3235,
+      "step": 1743
+    },
+    {
+      "epoch": 2.056603773584906,
+      "grad_norm": 0.19897262657208223,
+      "learning_rate": 1.7496723460026214e-05,
+      "loss": 0.3157,
+      "step": 1744
+    },
+    {
+      "epoch": 2.0577830188679247,
+      "grad_norm": 0.19985235510806515,
+      "learning_rate": 1.7474879860200962e-05,
+      "loss": 0.3055,
+      "step": 1745
+    },
+    {
+      "epoch": 2.0589622641509435,
+      "grad_norm": 0.1998010851180384,
+      "learning_rate": 1.745303626037571e-05,
+      "loss": 0.3315,
+      "step": 1746
+    },
+    {
+      "epoch": 2.0601415094339623,
+      "grad_norm": 0.2004048769846835,
+      "learning_rate": 1.743119266055046e-05,
+      "loss": 0.3089,
+      "step": 1747
+    },
+    {
+      "epoch": 2.061320754716981,
+      "grad_norm": 0.19159612638895795,
+      "learning_rate": 1.740934906072521e-05,
+      "loss": 0.2895,
+      "step": 1748
+    },
+    {
+      "epoch": 2.0625,
+      "grad_norm": 0.19042562112480202,
+      "learning_rate": 1.7387505460899957e-05,
+      "loss": 0.2904,
+      "step": 1749
+    },
+    {
+      "epoch": 2.063679245283019,
+      "grad_norm": 0.2043542832083641,
+      "learning_rate": 1.7365661861074706e-05,
+      "loss": 0.3167,
+      "step": 1750
+    },
+    {
+      "epoch": 2.0648584905660377,
+      "grad_norm": 0.1886567594748555,
+      "learning_rate": 1.7343818261249455e-05,
+      "loss": 0.3088,
+      "step": 1751
+    },
+    {
+      "epoch": 2.0660377358490565,
+      "grad_norm": 0.18482212491631905,
+      "learning_rate": 1.7321974661424203e-05,
+      "loss": 0.3056,
+      "step": 1752
+    },
+    {
+      "epoch": 2.0672169811320753,
+      "grad_norm": 0.19038083881625473,
+      "learning_rate": 1.7300131061598955e-05,
+      "loss": 0.3117,
+      "step": 1753
+    },
+    {
+      "epoch": 2.068396226415094,
+      "grad_norm": 0.2087071828669557,
+      "learning_rate": 1.72782874617737e-05,
+      "loss": 0.3141,
+      "step": 1754
+    },
+    {
+      "epoch": 2.0695754716981134,
+      "grad_norm": 0.1893065220433546,
+      "learning_rate": 1.725644386194845e-05,
+      "loss": 0.3029,
+      "step": 1755
+    },
+    {
+      "epoch": 2.0707547169811322,
+      "grad_norm": 0.2007010247233874,
+      "learning_rate": 1.7234600262123198e-05,
+      "loss": 0.2937,
+      "step": 1756
+    },
+    {
+      "epoch": 2.071933962264151,
+      "grad_norm": 0.20349639440643436,
+      "learning_rate": 1.7212756662297947e-05,
+      "loss": 0.3064,
+      "step": 1757
+    },
+    {
+      "epoch": 2.07311320754717,
+      "grad_norm": 0.20178487016431365,
+      "learning_rate": 1.7190913062472695e-05,
+      "loss": 0.3144,
+      "step": 1758
+    },
+    {
+      "epoch": 2.0742924528301887,
+      "grad_norm": 0.2109722701087339,
+      "learning_rate": 1.7169069462647444e-05,
+      "loss": 0.3115,
+      "step": 1759
+    },
+    {
+      "epoch": 2.0754716981132075,
+      "grad_norm": 0.21993638930498485,
+      "learning_rate": 1.7147225862822193e-05,
+      "loss": 0.33,
+      "step": 1760
+    },
+    {
+      "epoch": 2.0766509433962264,
+      "grad_norm": 0.1831675938950534,
+      "learning_rate": 1.712538226299694e-05,
+      "loss": 0.3196,
+      "step": 1761
+    },
+    {
+      "epoch": 2.077830188679245,
+      "grad_norm": 0.20635964819153654,
+      "learning_rate": 1.7103538663171694e-05,
+      "loss": 0.31,
+      "step": 1762
+    },
+    {
+      "epoch": 2.079009433962264,
+      "grad_norm": 0.19531250069573033,
+      "learning_rate": 1.708169506334644e-05,
+      "loss": 0.2939,
+      "step": 1763
+    },
+    {
+      "epoch": 2.080188679245283,
+      "grad_norm": 0.19838143611538558,
+      "learning_rate": 1.705985146352119e-05,
+      "loss": 0.3057,
+      "step": 1764
+    },
+    {
+      "epoch": 2.081367924528302,
+      "grad_norm": 0.18648449617415286,
+      "learning_rate": 1.7038007863695936e-05,
+      "loss": 0.2924,
+      "step": 1765
+    },
+    {
+      "epoch": 2.082547169811321,
+      "grad_norm": 0.1874151448753245,
+      "learning_rate": 1.701616426387069e-05,
+      "loss": 0.2991,
+      "step": 1766
+    },
+    {
+      "epoch": 2.0837264150943398,
+      "grad_norm": 0.202981971595534,
+      "learning_rate": 1.6994320664045434e-05,
+      "loss": 0.2977,
+      "step": 1767
+    },
+    {
+      "epoch": 2.0849056603773586,
+      "grad_norm": 0.19195917660460346,
+      "learning_rate": 1.6972477064220186e-05,
+      "loss": 0.3062,
+      "step": 1768
+    },
+    {
+      "epoch": 2.0860849056603774,
+      "grad_norm": 0.18651773270290645,
+      "learning_rate": 1.695063346439493e-05,
+      "loss": 0.2898,
+      "step": 1769
+    },
+    {
+      "epoch": 2.0872641509433962,
+      "grad_norm": 0.1944266110840758,
+      "learning_rate": 1.6928789864569683e-05,
+      "loss": 0.3096,
+      "step": 1770
+    },
+    {
+      "epoch": 2.088443396226415,
+      "grad_norm": 0.19477711602331677,
+      "learning_rate": 1.6906946264744432e-05,
+      "loss": 0.3067,
+      "step": 1771
+    },
+    {
+      "epoch": 2.089622641509434,
+      "grad_norm": 0.1959193803136535,
+      "learning_rate": 1.688510266491918e-05,
+      "loss": 0.3094,
+      "step": 1772
+    },
+    {
+      "epoch": 2.0908018867924527,
+      "grad_norm": 0.19517097735169084,
+      "learning_rate": 1.686325906509393e-05,
+      "loss": 0.3089,
+      "step": 1773
+    },
+    {
+      "epoch": 2.0919811320754715,
+      "grad_norm": 0.19928231944493874,
+      "learning_rate": 1.6841415465268678e-05,
+      "loss": 0.3243,
+      "step": 1774
+    },
+    {
+      "epoch": 2.0931603773584904,
+      "grad_norm": 0.1942096933755466,
+      "learning_rate": 1.6819571865443427e-05,
+      "loss": 0.3037,
+      "step": 1775
+    },
+    {
+      "epoch": 2.0943396226415096,
+      "grad_norm": 0.19345704047800316,
+      "learning_rate": 1.6797728265618172e-05,
+      "loss": 0.2841,
+      "step": 1776
+    },
+    {
+      "epoch": 2.0955188679245285,
+      "grad_norm": 0.18701982485818444,
+      "learning_rate": 1.6775884665792924e-05,
+      "loss": 0.2933,
+      "step": 1777
+    },
+    {
+      "epoch": 2.0966981132075473,
+      "grad_norm": 0.20682107128196978,
+      "learning_rate": 1.6754041065967673e-05,
+      "loss": 0.3157,
+      "step": 1778
+    },
+    {
+      "epoch": 2.097877358490566,
+      "grad_norm": 0.1984750541076218,
+      "learning_rate": 1.673219746614242e-05,
+      "loss": 0.3058,
+      "step": 1779
+    },
+    {
+      "epoch": 2.099056603773585,
+      "grad_norm": 0.20847665006056154,
+      "learning_rate": 1.671035386631717e-05,
+      "loss": 0.3174,
+      "step": 1780
+    },
+    {
+      "epoch": 2.1002358490566038,
+      "grad_norm": 0.2089120285037029,
+      "learning_rate": 1.668851026649192e-05,
+      "loss": 0.318,
+      "step": 1781
+    },
+    {
+      "epoch": 2.1014150943396226,
+      "grad_norm": 0.19347080540992007,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.2996,
+      "step": 1782
+    },
+    {
+      "epoch": 2.1025943396226414,
+      "grad_norm": 0.21289175270701036,
+      "learning_rate": 1.6644823066841416e-05,
+      "loss": 0.3022,
+      "step": 1783
+    },
+    {
+      "epoch": 2.1037735849056602,
+      "grad_norm": 0.1936221674612755,
+      "learning_rate": 1.6622979467016165e-05,
+      "loss": 0.3002,
+      "step": 1784
+    },
+    {
+      "epoch": 2.104952830188679,
+      "grad_norm": 0.1927174547819021,
+      "learning_rate": 1.6601135867190914e-05,
+      "loss": 0.305,
+      "step": 1785
+    },
+    {
+      "epoch": 2.106132075471698,
+      "grad_norm": 0.2156001600371975,
+      "learning_rate": 1.6579292267365662e-05,
+      "loss": 0.3057,
+      "step": 1786
+    },
+    {
+      "epoch": 2.107311320754717,
+      "grad_norm": 0.1857863094239874,
+      "learning_rate": 1.655744866754041e-05,
+      "loss": 0.3072,
+      "step": 1787
+    },
+    {
+      "epoch": 2.108490566037736,
+      "grad_norm": 0.1961362921510552,
+      "learning_rate": 1.653560506771516e-05,
+      "loss": 0.296,
+      "step": 1788
+    },
+    {
+      "epoch": 2.109669811320755,
+      "grad_norm": 0.19533757433997687,
+      "learning_rate": 1.651376146788991e-05,
+      "loss": 0.3202,
+      "step": 1789
+    },
+    {
+      "epoch": 2.1108490566037736,
+      "grad_norm": 0.20313411216761265,
+      "learning_rate": 1.6491917868064657e-05,
+      "loss": 0.3061,
+      "step": 1790
+    },
+    {
+      "epoch": 2.1120283018867925,
+      "grad_norm": 0.18742847735612364,
+      "learning_rate": 1.647007426823941e-05,
+      "loss": 0.3069,
+      "step": 1791
+    },
+    {
+      "epoch": 2.1132075471698113,
+      "grad_norm": 0.1967884095390451,
+      "learning_rate": 1.6448230668414154e-05,
+      "loss": 0.3238,
+      "step": 1792
+    },
+    {
+      "epoch": 2.11438679245283,
+      "grad_norm": 0.20805970208406702,
+      "learning_rate": 1.6426387068588907e-05,
+      "loss": 0.3177,
+      "step": 1793
+    },
+    {
+      "epoch": 2.115566037735849,
+      "grad_norm": 0.19313300000396869,
+      "learning_rate": 1.6404543468763652e-05,
+      "loss": 0.2971,
+      "step": 1794
+    },
+    {
+      "epoch": 2.1167452830188678,
+      "grad_norm": 0.1979006276267319,
+      "learning_rate": 1.6382699868938404e-05,
+      "loss": 0.3123,
+      "step": 1795
+    },
+    {
+      "epoch": 2.1179245283018866,
+      "grad_norm": 0.21191940663524814,
+      "learning_rate": 1.636085626911315e-05,
+      "loss": 0.3214,
+      "step": 1796
+    },
+    {
+      "epoch": 2.119103773584906,
+      "grad_norm": 0.1919264120674389,
+      "learning_rate": 1.6339012669287898e-05,
+      "loss": 0.2998,
+      "step": 1797
+    },
+    {
+      "epoch": 2.1202830188679247,
+      "grad_norm": 0.20361857544452674,
+      "learning_rate": 1.631716906946265e-05,
+      "loss": 0.3251,
+      "step": 1798
+    },
+    {
+      "epoch": 2.1214622641509435,
+      "grad_norm": 0.20461932127160584,
+      "learning_rate": 1.6295325469637395e-05,
+      "loss": 0.3009,
+      "step": 1799
+    },
+    {
+      "epoch": 2.1226415094339623,
+      "grad_norm": 0.19464372674492428,
+      "learning_rate": 1.6273481869812147e-05,
+      "loss": 0.3095,
+      "step": 1800
+    },
+    {
+      "epoch": 2.123820754716981,
+      "grad_norm": 0.1937187958099221,
+      "learning_rate": 1.6251638269986893e-05,
+      "loss": 0.3085,
+      "step": 1801
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 0.1983728705029803,
+      "learning_rate": 1.6229794670161645e-05,
+      "loss": 0.3276,
+      "step": 1802
+    },
+    {
+      "epoch": 2.126179245283019,
+      "grad_norm": 0.1995783174851144,
+      "learning_rate": 1.620795107033639e-05,
+      "loss": 0.3093,
+      "step": 1803
+    },
+    {
+      "epoch": 2.1273584905660377,
+      "grad_norm": 0.18588388096789343,
+      "learning_rate": 1.6186107470511142e-05,
+      "loss": 0.3081,
+      "step": 1804
+    },
+    {
+      "epoch": 2.1285377358490565,
+      "grad_norm": 0.20079057417739254,
+      "learning_rate": 1.6164263870685887e-05,
+      "loss": 0.3062,
+      "step": 1805
+    },
+    {
+      "epoch": 2.1297169811320753,
+      "grad_norm": 0.1953253196489552,
+      "learning_rate": 1.614242027086064e-05,
+      "loss": 0.3015,
+      "step": 1806
+    },
+    {
+      "epoch": 2.1308962264150946,
+      "grad_norm": 0.20182604290464234,
+      "learning_rate": 1.6120576671035388e-05,
+      "loss": 0.3033,
+      "step": 1807
+    },
+    {
+      "epoch": 2.1320754716981134,
+      "grad_norm": 0.20056592792170877,
+      "learning_rate": 1.6098733071210137e-05,
+      "loss": 0.308,
+      "step": 1808
+    },
+    {
+      "epoch": 2.1332547169811322,
+      "grad_norm": 0.20276579043765605,
+      "learning_rate": 1.6076889471384886e-05,
+      "loss": 0.3074,
+      "step": 1809
+    },
+    {
+      "epoch": 2.134433962264151,
+      "grad_norm": 0.20718419754717135,
+      "learning_rate": 1.6055045871559634e-05,
+      "loss": 0.3198,
+      "step": 1810
+    },
+    {
+      "epoch": 2.13561320754717,
+      "grad_norm": 0.20358119609691455,
+      "learning_rate": 1.6033202271734383e-05,
+      "loss": 0.3092,
+      "step": 1811
+    },
+    {
+      "epoch": 2.1367924528301887,
+      "grad_norm": 0.20295203588960425,
+      "learning_rate": 1.6011358671909132e-05,
+      "loss": 0.3234,
+      "step": 1812
+    },
+    {
+      "epoch": 2.1379716981132075,
+      "grad_norm": 0.22717784841149538,
+      "learning_rate": 1.598951507208388e-05,
+      "loss": 0.3101,
+      "step": 1813
+    },
+    {
+      "epoch": 2.1391509433962264,
+      "grad_norm": 0.1953691442221245,
+      "learning_rate": 1.596767147225863e-05,
+      "loss": 0.3099,
+      "step": 1814
+    },
+    {
+      "epoch": 2.140330188679245,
+      "grad_norm": 0.20077077111458974,
+      "learning_rate": 1.5945827872433378e-05,
+      "loss": 0.3242,
+      "step": 1815
+    },
+    {
+      "epoch": 2.141509433962264,
+      "grad_norm": 0.2050288953035635,
+      "learning_rate": 1.5923984272608126e-05,
+      "loss": 0.3162,
+      "step": 1816
+    },
+    {
+      "epoch": 2.142688679245283,
+      "grad_norm": 0.22577500966300465,
+      "learning_rate": 1.5902140672782875e-05,
+      "loss": 0.3046,
+      "step": 1817
+    },
+    {
+      "epoch": 2.143867924528302,
+      "grad_norm": 0.20230083715979796,
+      "learning_rate": 1.5880297072957624e-05,
+      "loss": 0.3061,
+      "step": 1818
+    },
+    {
+      "epoch": 2.145047169811321,
+      "grad_norm": 0.24154657556425735,
+      "learning_rate": 1.5858453473132373e-05,
+      "loss": 0.3216,
+      "step": 1819
+    },
+    {
+      "epoch": 2.1462264150943398,
+      "grad_norm": 0.1994037357482287,
+      "learning_rate": 1.583660987330712e-05,
+      "loss": 0.2876,
+      "step": 1820
+    },
+    {
+      "epoch": 2.1474056603773586,
+      "grad_norm": 0.19492556504395167,
+      "learning_rate": 1.581476627348187e-05,
+      "loss": 0.3236,
+      "step": 1821
+    },
+    {
+      "epoch": 2.1485849056603774,
+      "grad_norm": 0.20832818247287835,
+      "learning_rate": 1.579292267365662e-05,
+      "loss": 0.3155,
+      "step": 1822
+    },
+    {
+      "epoch": 2.1497641509433962,
+      "grad_norm": 0.21216984274734071,
+      "learning_rate": 1.5771079073831367e-05,
+      "loss": 0.3156,
+      "step": 1823
+    },
+    {
+      "epoch": 2.150943396226415,
+      "grad_norm": 0.20036421363572335,
+      "learning_rate": 1.5749235474006116e-05,
+      "loss": 0.3124,
+      "step": 1824
+    },
+    {
+      "epoch": 2.152122641509434,
+      "grad_norm": 0.1981876421176524,
+      "learning_rate": 1.5727391874180865e-05,
+      "loss": 0.2919,
+      "step": 1825
+    },
+    {
+      "epoch": 2.1533018867924527,
+      "grad_norm": 0.20085806212503035,
+      "learning_rate": 1.5705548274355613e-05,
+      "loss": 0.3388,
+      "step": 1826
+    },
+    {
+      "epoch": 2.1544811320754715,
+      "grad_norm": 0.21722716691618096,
+      "learning_rate": 1.5683704674530366e-05,
+      "loss": 0.3157,
+      "step": 1827
+    },
+    {
+      "epoch": 2.1556603773584904,
+      "grad_norm": 0.2040980776465905,
+      "learning_rate": 1.566186107470511e-05,
+      "loss": 0.3227,
+      "step": 1828
+    },
+    {
+      "epoch": 2.1568396226415096,
+      "grad_norm": 0.20760402369859443,
+      "learning_rate": 1.5640017474879863e-05,
+      "loss": 0.3238,
+      "step": 1829
+    },
+    {
+      "epoch": 2.1580188679245285,
+      "grad_norm": 0.22182260464874928,
+      "learning_rate": 1.5618173875054608e-05,
+      "loss": 0.3147,
+      "step": 1830
+    },
+    {
+      "epoch": 2.1591981132075473,
+      "grad_norm": 0.19484070254310915,
+      "learning_rate": 1.559633027522936e-05,
+      "loss": 0.298,
+      "step": 1831
+    },
+    {
+      "epoch": 2.160377358490566,
+      "grad_norm": 0.2094906470658297,
+      "learning_rate": 1.5574486675404106e-05,
+      "loss": 0.3145,
+      "step": 1832
+    },
+    {
+      "epoch": 2.161556603773585,
+      "grad_norm": 0.18973838574899685,
+      "learning_rate": 1.5552643075578858e-05,
+      "loss": 0.296,
+      "step": 1833
+    },
+    {
+      "epoch": 2.1627358490566038,
+      "grad_norm": 0.18967826975202445,
+      "learning_rate": 1.5530799475753603e-05,
+      "loss": 0.3034,
+      "step": 1834
+    },
+    {
+      "epoch": 2.1639150943396226,
+      "grad_norm": 0.19441243152352533,
+      "learning_rate": 1.5508955875928355e-05,
+      "loss": 0.313,
+      "step": 1835
+    },
+    {
+      "epoch": 2.1650943396226414,
+      "grad_norm": 0.21158174065308855,
+      "learning_rate": 1.5487112276103104e-05,
+      "loss": 0.3161,
+      "step": 1836
+    },
+    {
+      "epoch": 2.1662735849056602,
+      "grad_norm": 0.19545727580361338,
+      "learning_rate": 1.5465268676277852e-05,
+      "loss": 0.3086,
+      "step": 1837
+    },
+    {
+      "epoch": 2.167452830188679,
+      "grad_norm": 0.18457542769231092,
+      "learning_rate": 1.54434250764526e-05,
+      "loss": 0.3192,
+      "step": 1838
+    },
+    {
+      "epoch": 2.168632075471698,
+      "grad_norm": 0.1995868564203529,
+      "learning_rate": 1.542158147662735e-05,
+      "loss": 0.3174,
+      "step": 1839
+    },
+    {
+      "epoch": 2.169811320754717,
+      "grad_norm": 0.20216516265562806,
+      "learning_rate": 1.53997378768021e-05,
+      "loss": 0.3259,
+      "step": 1840
+    },
+    {
+      "epoch": 2.170990566037736,
+      "grad_norm": 0.18758785726208943,
+      "learning_rate": 1.5377894276976844e-05,
+      "loss": 0.3155,
+      "step": 1841
+    },
+    {
+      "epoch": 2.172169811320755,
+      "grad_norm": 0.1976158085198784,
+      "learning_rate": 1.5356050677151596e-05,
+      "loss": 0.328,
+      "step": 1842
+    },
+    {
+      "epoch": 2.1733490566037736,
+      "grad_norm": 0.19610950321277784,
+      "learning_rate": 1.5334207077326345e-05,
+      "loss": 0.304,
+      "step": 1843
+    },
+    {
+      "epoch": 2.1745283018867925,
+      "grad_norm": 0.21223289428002642,
+      "learning_rate": 1.5312363477501093e-05,
+      "loss": 0.3161,
+      "step": 1844
+    },
+    {
+      "epoch": 2.1757075471698113,
+      "grad_norm": 0.1942997818844832,
+      "learning_rate": 1.5290519877675842e-05,
+      "loss": 0.3052,
+      "step": 1845
+    },
+    {
+      "epoch": 2.17688679245283,
+      "grad_norm": 0.19377756386469455,
+      "learning_rate": 1.526867627785059e-05,
+      "loss": 0.3063,
+      "step": 1846
+    },
+    {
+      "epoch": 2.178066037735849,
+      "grad_norm": 0.19976213329733555,
+      "learning_rate": 1.5246832678025338e-05,
+      "loss": 0.3134,
+      "step": 1847
+    },
+    {
+      "epoch": 2.1792452830188678,
+      "grad_norm": 0.18888107096096166,
+      "learning_rate": 1.5224989078200088e-05,
+      "loss": 0.3156,
+      "step": 1848
+    },
+    {
+      "epoch": 2.1804245283018866,
+      "grad_norm": 0.19994139884311737,
+      "learning_rate": 1.5203145478374837e-05,
+      "loss": 0.3109,
+      "step": 1849
+    },
+    {
+      "epoch": 2.1816037735849054,
+      "grad_norm": 0.18861076333280682,
+      "learning_rate": 1.5181301878549587e-05,
+      "loss": 0.2978,
+      "step": 1850
+    },
+    {
+      "epoch": 2.1827830188679247,
+      "grad_norm": 0.19469157111220367,
+      "learning_rate": 1.5159458278724334e-05,
+      "loss": 0.3191,
+      "step": 1851
+    },
+    {
+      "epoch": 2.1839622641509435,
+      "grad_norm": 0.19015769777699426,
+      "learning_rate": 1.5137614678899085e-05,
+      "loss": 0.3006,
+      "step": 1852
+    },
+    {
+      "epoch": 2.1851415094339623,
+      "grad_norm": 0.19053274419196858,
+      "learning_rate": 1.5115771079073832e-05,
+      "loss": 0.3215,
+      "step": 1853
+    },
+    {
+      "epoch": 2.186320754716981,
+      "grad_norm": 0.2012261411130295,
+      "learning_rate": 1.5093927479248582e-05,
+      "loss": 0.3066,
+      "step": 1854
+    },
+    {
+      "epoch": 2.1875,
+      "grad_norm": 0.20325242817963268,
+      "learning_rate": 1.5072083879423329e-05,
+      "loss": 0.3141,
+      "step": 1855
+    },
+    {
+      "epoch": 2.188679245283019,
+      "grad_norm": 0.1896651924492433,
+      "learning_rate": 1.505024027959808e-05,
+      "loss": 0.303,
+      "step": 1856
+    },
+    {
+      "epoch": 2.1898584905660377,
+      "grad_norm": 0.23205184724123581,
+      "learning_rate": 1.5028396679772826e-05,
+      "loss": 0.3051,
+      "step": 1857
+    },
+    {
+      "epoch": 2.1910377358490565,
+      "grad_norm": 0.19750155966968402,
+      "learning_rate": 1.5006553079947577e-05,
+      "loss": 0.3203,
+      "step": 1858
+    },
+    {
+      "epoch": 2.1922169811320753,
+      "grad_norm": 0.2030797589950051,
+      "learning_rate": 1.4984709480122325e-05,
+      "loss": 0.3282,
+      "step": 1859
+    },
+    {
+      "epoch": 2.1933962264150946,
+      "grad_norm": 0.1992534671970061,
+      "learning_rate": 1.4962865880297076e-05,
+      "loss": 0.3255,
+      "step": 1860
+    },
+    {
+      "epoch": 2.1945754716981134,
+      "grad_norm": 0.1965115227940344,
+      "learning_rate": 1.4941022280471823e-05,
+      "loss": 0.3087,
+      "step": 1861
+    },
+    {
+      "epoch": 2.1957547169811322,
+      "grad_norm": 0.20478907082325687,
+      "learning_rate": 1.491917868064657e-05,
+      "loss": 0.321,
+      "step": 1862
+    },
+    {
+      "epoch": 2.196933962264151,
+      "grad_norm": 0.19370354876367532,
+      "learning_rate": 1.489733508082132e-05,
+      "loss": 0.3245,
+      "step": 1863
+    },
+    {
+      "epoch": 2.19811320754717,
+      "grad_norm": 0.2006595609032095,
+      "learning_rate": 1.4875491480996067e-05,
+      "loss": 0.3166,
+      "step": 1864
+    },
+    {
+      "epoch": 2.1992924528301887,
+      "grad_norm": 0.21205929661010273,
+      "learning_rate": 1.4853647881170818e-05,
+      "loss": 0.3166,
+      "step": 1865
+    },
+    {
+      "epoch": 2.2004716981132075,
+      "grad_norm": 0.19918265992419137,
+      "learning_rate": 1.4831804281345565e-05,
+      "loss": 0.3034,
+      "step": 1866
+    },
+    {
+      "epoch": 2.2016509433962264,
+      "grad_norm": 0.1973366432683058,
+      "learning_rate": 1.4809960681520315e-05,
+      "loss": 0.3001,
+      "step": 1867
+    },
+    {
+      "epoch": 2.202830188679245,
+      "grad_norm": 0.18680588946197943,
+      "learning_rate": 1.4788117081695064e-05,
+      "loss": 0.2909,
+      "step": 1868
+    },
+    {
+      "epoch": 2.204009433962264,
+      "grad_norm": 0.19568368483537285,
+      "learning_rate": 1.4766273481869814e-05,
+      "loss": 0.3142,
+      "step": 1869
+    },
+    {
+      "epoch": 2.205188679245283,
+      "grad_norm": 0.18940932203122723,
+      "learning_rate": 1.4744429882044561e-05,
+      "loss": 0.3105,
+      "step": 1870
+    },
+    {
+      "epoch": 2.206367924528302,
+      "grad_norm": 0.19913401420308222,
+      "learning_rate": 1.4722586282219312e-05,
+      "loss": 0.3079,
+      "step": 1871
+    },
+    {
+      "epoch": 2.207547169811321,
+      "grad_norm": 0.1799449674855431,
+      "learning_rate": 1.4700742682394059e-05,
+      "loss": 0.3058,
+      "step": 1872
+    },
+    {
+      "epoch": 2.2087264150943398,
+      "grad_norm": 0.1875688927805789,
+      "learning_rate": 1.4678899082568809e-05,
+      "loss": 0.3165,
+      "step": 1873
+    },
+    {
+      "epoch": 2.2099056603773586,
+      "grad_norm": 0.19385539495155205,
+      "learning_rate": 1.4657055482743556e-05,
+      "loss": 0.3114,
+      "step": 1874
+    },
+    {
+      "epoch": 2.2110849056603774,
+      "grad_norm": 0.19059067454943165,
+      "learning_rate": 1.4635211882918306e-05,
+      "loss": 0.3149,
+      "step": 1875
+    },
+    {
+      "epoch": 2.2122641509433962,
+      "grad_norm": 0.19016557025236822,
+      "learning_rate": 1.4613368283093053e-05,
+      "loss": 0.2934,
+      "step": 1876
+    },
+    {
+      "epoch": 2.213443396226415,
+      "grad_norm": 0.19247887087157758,
+      "learning_rate": 1.4591524683267804e-05,
+      "loss": 0.3113,
+      "step": 1877
+    },
+    {
+      "epoch": 2.214622641509434,
+      "grad_norm": 0.1968753954539843,
+      "learning_rate": 1.4569681083442552e-05,
+      "loss": 0.3115,
+      "step": 1878
+    },
+    {
+      "epoch": 2.2158018867924527,
+      "grad_norm": 0.19131611469416995,
+      "learning_rate": 1.4547837483617303e-05,
+      "loss": 0.3167,
+      "step": 1879
+    },
+    {
+      "epoch": 2.2169811320754715,
+      "grad_norm": 0.1948940434713202,
+      "learning_rate": 1.452599388379205e-05,
+      "loss": 0.3024,
+      "step": 1880
+    },
+    {
+      "epoch": 2.2181603773584904,
+      "grad_norm": 0.1854588889897325,
+      "learning_rate": 1.45041502839668e-05,
+      "loss": 0.2956,
+      "step": 1881
+    },
+    {
+      "epoch": 2.2193396226415096,
+      "grad_norm": 0.19261701516892485,
+      "learning_rate": 1.4482306684141547e-05,
+      "loss": 0.3188,
+      "step": 1882
+    },
+    {
+      "epoch": 2.2205188679245285,
+      "grad_norm": 0.1834387500992822,
+      "learning_rate": 1.4460463084316294e-05,
+      "loss": 0.3111,
+      "step": 1883
+    },
+    {
+      "epoch": 2.2216981132075473,
+      "grad_norm": 0.18686867317110228,
+      "learning_rate": 1.4438619484491045e-05,
+      "loss": 0.3302,
+      "step": 1884
+    },
+    {
+      "epoch": 2.222877358490566,
+      "grad_norm": 0.25992927127119025,
+      "learning_rate": 1.4416775884665793e-05,
+      "loss": 0.312,
+      "step": 1885
+    },
+    {
+      "epoch": 2.224056603773585,
+      "grad_norm": 0.18561002541982863,
+      "learning_rate": 1.4394932284840542e-05,
+      "loss": 0.3177,
+      "step": 1886
+    },
+    {
+      "epoch": 2.2252358490566038,
+      "grad_norm": 0.1943297682574138,
+      "learning_rate": 1.437308868501529e-05,
+      "loss": 0.3192,
+      "step": 1887
+    },
+    {
+      "epoch": 2.2264150943396226,
+      "grad_norm": 0.20547454693804443,
+      "learning_rate": 1.4351245085190041e-05,
+      "loss": 0.3062,
+      "step": 1888
+    },
+    {
+      "epoch": 2.2275943396226414,
+      "grad_norm": 0.19616408302141755,
+      "learning_rate": 1.4329401485364788e-05,
+      "loss": 0.3238,
+      "step": 1889
+    },
+    {
+      "epoch": 2.2287735849056602,
+      "grad_norm": 0.1906993177623927,
+      "learning_rate": 1.4307557885539538e-05,
+      "loss": 0.3182,
+      "step": 1890
+    },
+    {
+      "epoch": 2.229952830188679,
+      "grad_norm": 0.20942546193686407,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.2989,
+      "step": 1891
+    },
+    {
+      "epoch": 2.231132075471698,
+      "grad_norm": 0.20416753147053354,
+      "learning_rate": 1.4263870685889036e-05,
+      "loss": 0.3072,
+      "step": 1892
+    },
+    {
+      "epoch": 2.232311320754717,
+      "grad_norm": 0.19427344926280343,
+      "learning_rate": 1.4242027086063783e-05,
+      "loss": 0.3136,
+      "step": 1893
+    },
+    {
+      "epoch": 2.233490566037736,
+      "grad_norm": 0.20425481843091928,
+      "learning_rate": 1.4220183486238533e-05,
+      "loss": 0.3024,
+      "step": 1894
+    },
+    {
+      "epoch": 2.234669811320755,
+      "grad_norm": 0.1890734735124647,
+      "learning_rate": 1.419833988641328e-05,
+      "loss": 0.3126,
+      "step": 1895
+    },
+    {
+      "epoch": 2.2358490566037736,
+      "grad_norm": 0.19507409514455792,
+      "learning_rate": 1.417649628658803e-05,
+      "loss": 0.314,
+      "step": 1896
+    },
+    {
+      "epoch": 2.2370283018867925,
+      "grad_norm": 0.18673173836580023,
+      "learning_rate": 1.415465268676278e-05,
+      "loss": 0.3022,
+      "step": 1897
+    },
+    {
+      "epoch": 2.2382075471698113,
+      "grad_norm": 0.22476465850297941,
+      "learning_rate": 1.413280908693753e-05,
+      "loss": 0.3222,
+      "step": 1898
+    },
+    {
+      "epoch": 2.23938679245283,
+      "grad_norm": 0.2387626997785429,
+      "learning_rate": 1.4110965487112277e-05,
+      "loss": 0.3143,
+      "step": 1899
+    },
+    {
+      "epoch": 2.240566037735849,
+      "grad_norm": 0.20866037498074663,
+      "learning_rate": 1.4089121887287027e-05,
+      "loss": 0.3364,
+      "step": 1900
+    },
+    {
+      "epoch": 2.2417452830188678,
+      "grad_norm": 0.20219919412137202,
+      "learning_rate": 1.4067278287461774e-05,
+      "loss": 0.3237,
+      "step": 1901
+    },
+    {
+      "epoch": 2.2429245283018866,
+      "grad_norm": 0.19638194086011765,
+      "learning_rate": 1.4045434687636524e-05,
+      "loss": 0.3232,
+      "step": 1902
+    },
+    {
+      "epoch": 2.2441037735849054,
+      "grad_norm": 0.19397566513144676,
+      "learning_rate": 1.4023591087811271e-05,
+      "loss": 0.3059,
+      "step": 1903
+    },
+    {
+      "epoch": 2.2452830188679247,
+      "grad_norm": 0.19989278622933637,
+      "learning_rate": 1.400174748798602e-05,
+      "loss": 0.3062,
+      "step": 1904
+    },
+    {
+      "epoch": 2.2464622641509435,
+      "grad_norm": 0.22310867144173435,
+      "learning_rate": 1.3979903888160769e-05,
+      "loss": 0.313,
+      "step": 1905
+    },
+    {
+      "epoch": 2.2476415094339623,
+      "grad_norm": 0.19585061553542923,
+      "learning_rate": 1.3958060288335518e-05,
+      "loss": 0.3166,
+      "step": 1906
+    },
+    {
+      "epoch": 2.248820754716981,
+      "grad_norm": 0.1854365298948458,
+      "learning_rate": 1.3936216688510268e-05,
+      "loss": 0.3004,
+      "step": 1907
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.18454317288718752,
+      "learning_rate": 1.3914373088685015e-05,
+      "loss": 0.2929,
+      "step": 1908
+    },
+    {
+      "epoch": 2.251179245283019,
+      "grad_norm": 0.20946408399824287,
+      "learning_rate": 1.3892529488859765e-05,
+      "loss": 0.3193,
+      "step": 1909
+    },
+    {
+      "epoch": 2.2523584905660377,
+      "grad_norm": 0.20210504730772288,
+      "learning_rate": 1.3870685889034512e-05,
+      "loss": 0.3205,
+      "step": 1910
+    },
+    {
+      "epoch": 2.2535377358490565,
+      "grad_norm": 0.19361723009664847,
+      "learning_rate": 1.3848842289209263e-05,
+      "loss": 0.3127,
+      "step": 1911
+    },
+    {
+      "epoch": 2.2547169811320753,
+      "grad_norm": 0.18915058806159887,
+      "learning_rate": 1.382699868938401e-05,
+      "loss": 0.3129,
+      "step": 1912
+    },
+    {
+      "epoch": 2.2558962264150946,
+      "grad_norm": 0.2088339961074856,
+      "learning_rate": 1.380515508955876e-05,
+      "loss": 0.3057,
+      "step": 1913
+    },
+    {
+      "epoch": 2.2570754716981134,
+      "grad_norm": 0.2144273404632206,
+      "learning_rate": 1.3783311489733509e-05,
+      "loss": 0.3329,
+      "step": 1914
+    },
+    {
+      "epoch": 2.2582547169811322,
+      "grad_norm": 0.21298688972127067,
+      "learning_rate": 1.3761467889908258e-05,
+      "loss": 0.3318,
+      "step": 1915
+    },
+    {
+      "epoch": 2.259433962264151,
+      "grad_norm": 0.20655062564469,
+      "learning_rate": 1.3739624290083006e-05,
+      "loss": 0.3032,
+      "step": 1916
+    },
+    {
+      "epoch": 2.26061320754717,
+      "grad_norm": 0.20898718818974577,
+      "learning_rate": 1.3717780690257757e-05,
+      "loss": 0.3041,
+      "step": 1917
+    },
+    {
+      "epoch": 2.2617924528301887,
+      "grad_norm": 0.20834153502821318,
+      "learning_rate": 1.3695937090432504e-05,
+      "loss": 0.3114,
+      "step": 1918
+    },
+    {
+      "epoch": 2.2629716981132075,
+      "grad_norm": 0.1952777472248904,
+      "learning_rate": 1.3674093490607254e-05,
+      "loss": 0.3091,
+      "step": 1919
+    },
+    {
+      "epoch": 2.2641509433962264,
+      "grad_norm": 0.2031986799497434,
+      "learning_rate": 1.3652249890782001e-05,
+      "loss": 0.3144,
+      "step": 1920
+    },
+    {
+      "epoch": 2.265330188679245,
+      "grad_norm": 0.18682152181478287,
+      "learning_rate": 1.3630406290956751e-05,
+      "loss": 0.3155,
+      "step": 1921
+    },
+    {
+      "epoch": 2.266509433962264,
+      "grad_norm": 0.195387362553787,
+      "learning_rate": 1.3608562691131498e-05,
+      "loss": 0.2944,
+      "step": 1922
+    },
+    {
+      "epoch": 2.267688679245283,
+      "grad_norm": 0.20013445043214204,
+      "learning_rate": 1.3586719091306249e-05,
+      "loss": 0.3298,
+      "step": 1923
+    },
+    {
+      "epoch": 2.268867924528302,
+      "grad_norm": 0.17788713528984887,
+      "learning_rate": 1.3564875491480997e-05,
+      "loss": 0.2966,
+      "step": 1924
+    },
+    {
+      "epoch": 2.270047169811321,
+      "grad_norm": 0.19675314230832333,
+      "learning_rate": 1.3543031891655744e-05,
+      "loss": 0.3346,
+      "step": 1925
+    },
+    {
+      "epoch": 2.2712264150943398,
+      "grad_norm": 0.1987336102725448,
+      "learning_rate": 1.3521188291830495e-05,
+      "loss": 0.306,
+      "step": 1926
+    },
+    {
+      "epoch": 2.2724056603773586,
+      "grad_norm": 0.1966403956286768,
+      "learning_rate": 1.3499344692005242e-05,
+      "loss": 0.3176,
+      "step": 1927
+    },
+    {
+      "epoch": 2.2735849056603774,
+      "grad_norm": 0.1917415322340461,
+      "learning_rate": 1.3477501092179992e-05,
+      "loss": 0.3139,
+      "step": 1928
+    },
+    {
+      "epoch": 2.2747641509433962,
+      "grad_norm": 0.19928132965214085,
+      "learning_rate": 1.345565749235474e-05,
+      "loss": 0.3221,
+      "step": 1929
+    },
+    {
+      "epoch": 2.275943396226415,
+      "grad_norm": 0.1946387590364937,
+      "learning_rate": 1.343381389252949e-05,
+      "loss": 0.2981,
+      "step": 1930
+    },
+    {
+      "epoch": 2.277122641509434,
+      "grad_norm": 0.19485877029706677,
+      "learning_rate": 1.3411970292704237e-05,
+      "loss": 0.3156,
+      "step": 1931
+    },
+    {
+      "epoch": 2.2783018867924527,
+      "grad_norm": 0.20339550029837003,
+      "learning_rate": 1.3390126692878987e-05,
+      "loss": 0.3112,
+      "step": 1932
+    },
+    {
+      "epoch": 2.2794811320754715,
+      "grad_norm": 0.18416469875437455,
+      "learning_rate": 1.3368283093053736e-05,
+      "loss": 0.2967,
+      "step": 1933
+    },
+    {
+      "epoch": 2.2806603773584904,
+      "grad_norm": 0.2023800512377023,
+      "learning_rate": 1.3346439493228486e-05,
+      "loss": 0.3209,
+      "step": 1934
+    },
+    {
+      "epoch": 2.2818396226415096,
+      "grad_norm": 0.18150331617684445,
+      "learning_rate": 1.3324595893403233e-05,
+      "loss": 0.2955,
+      "step": 1935
+    },
+    {
+      "epoch": 2.2830188679245285,
+      "grad_norm": 0.20144055562843655,
+      "learning_rate": 1.3302752293577984e-05,
+      "loss": 0.3108,
+      "step": 1936
+    },
+    {
+      "epoch": 2.2841981132075473,
+      "grad_norm": 0.2033138691472472,
+      "learning_rate": 1.328090869375273e-05,
+      "loss": 0.302,
+      "step": 1937
+    },
+    {
+      "epoch": 2.285377358490566,
+      "grad_norm": 0.2012251887971359,
+      "learning_rate": 1.3259065093927481e-05,
+      "loss": 0.3057,
+      "step": 1938
+    },
+    {
+      "epoch": 2.286556603773585,
+      "grad_norm": 0.1870850006129352,
+      "learning_rate": 1.3237221494102228e-05,
+      "loss": 0.2957,
+      "step": 1939
+    },
+    {
+      "epoch": 2.2877358490566038,
+      "grad_norm": 0.19703725834996355,
+      "learning_rate": 1.3215377894276978e-05,
+      "loss": 0.2964,
+      "step": 1940
+    },
+    {
+      "epoch": 2.2889150943396226,
+      "grad_norm": 0.21263370748721558,
+      "learning_rate": 1.3193534294451725e-05,
+      "loss": 0.3207,
+      "step": 1941
+    },
+    {
+      "epoch": 2.2900943396226414,
+      "grad_norm": 0.20065596377197523,
+      "learning_rate": 1.3171690694626476e-05,
+      "loss": 0.3162,
+      "step": 1942
+    },
+    {
+      "epoch": 2.2912735849056602,
+      "grad_norm": 0.18380947699161723,
+      "learning_rate": 1.3149847094801224e-05,
+      "loss": 0.2885,
+      "step": 1943
+    },
+    {
+      "epoch": 2.292452830188679,
+      "grad_norm": 0.2000074175340788,
+      "learning_rate": 1.3128003494975975e-05,
+      "loss": 0.3162,
+      "step": 1944
+    },
+    {
+      "epoch": 2.293632075471698,
+      "grad_norm": 0.19558996090067,
+      "learning_rate": 1.3106159895150722e-05,
+      "loss": 0.302,
+      "step": 1945
+    },
+    {
+      "epoch": 2.294811320754717,
+      "grad_norm": 0.19607389083654606,
+      "learning_rate": 1.3084316295325472e-05,
+      "loss": 0.3159,
+      "step": 1946
+    },
+    {
+      "epoch": 2.295990566037736,
+      "grad_norm": 0.20272647066475172,
+      "learning_rate": 1.3062472695500219e-05,
+      "loss": 0.3118,
+      "step": 1947
+    },
+    {
+      "epoch": 2.297169811320755,
+      "grad_norm": 0.19837640620749628,
+      "learning_rate": 1.3040629095674966e-05,
+      "loss": 0.3145,
+      "step": 1948
+    },
+    {
+      "epoch": 2.2983490566037736,
+      "grad_norm": 0.1839559104079096,
+      "learning_rate": 1.3018785495849717e-05,
+      "loss": 0.2937,
+      "step": 1949
+    },
+    {
+      "epoch": 2.2995283018867925,
+      "grad_norm": 0.21912604900388394,
+      "learning_rate": 1.2996941896024464e-05,
+      "loss": 0.3276,
+      "step": 1950
+    },
+    {
+      "epoch": 2.3007075471698113,
+      "grad_norm": 0.21476941037359004,
+      "learning_rate": 1.2975098296199214e-05,
+      "loss": 0.3185,
+      "step": 1951
+    },
+    {
+      "epoch": 2.30188679245283,
+      "grad_norm": 0.20264254929311565,
+      "learning_rate": 1.2953254696373963e-05,
+      "loss": 0.3165,
+      "step": 1952
+    },
+    {
+      "epoch": 2.303066037735849,
+      "grad_norm": 0.2017975737832373,
+      "learning_rate": 1.2931411096548713e-05,
+      "loss": 0.3126,
+      "step": 1953
+    },
+    {
+      "epoch": 2.3042452830188678,
+      "grad_norm": 0.18491039653047547,
+      "learning_rate": 1.290956749672346e-05,
+      "loss": 0.2943,
+      "step": 1954
+    },
+    {
+      "epoch": 2.3054245283018866,
+      "grad_norm": 0.19926856719919772,
+      "learning_rate": 1.288772389689821e-05,
+      "loss": 0.3186,
+      "step": 1955
+    },
+    {
+      "epoch": 2.3066037735849054,
+      "grad_norm": 0.19848446881517257,
+      "learning_rate": 1.2865880297072957e-05,
+      "loss": 0.3155,
+      "step": 1956
+    },
+    {
+      "epoch": 2.3077830188679247,
+      "grad_norm": 0.1972881623080172,
+      "learning_rate": 1.2844036697247708e-05,
+      "loss": 0.3164,
+      "step": 1957
+    },
+    {
+      "epoch": 2.3089622641509435,
+      "grad_norm": 0.19975961772526488,
+      "learning_rate": 1.2822193097422455e-05,
+      "loss": 0.3247,
+      "step": 1958
+    },
+    {
+      "epoch": 2.3101415094339623,
+      "grad_norm": 0.2028205970378534,
+      "learning_rate": 1.2800349497597205e-05,
+      "loss": 0.3064,
+      "step": 1959
+    },
+    {
+      "epoch": 2.311320754716981,
+      "grad_norm": 0.19304885149729414,
+      "learning_rate": 1.2778505897771952e-05,
+      "loss": 0.3061,
+      "step": 1960
+    },
+    {
+      "epoch": 2.3125,
+      "grad_norm": 0.187371198538623,
+      "learning_rate": 1.2756662297946703e-05,
+      "loss": 0.301,
+      "step": 1961
+    },
+    {
+      "epoch": 2.313679245283019,
+      "grad_norm": 0.2041831391963872,
+      "learning_rate": 1.2734818698121451e-05,
+      "loss": 0.3148,
+      "step": 1962
+    },
+    {
+      "epoch": 2.3148584905660377,
+      "grad_norm": 0.19925233202267498,
+      "learning_rate": 1.2712975098296202e-05,
+      "loss": 0.2999,
+      "step": 1963
+    },
+    {
+      "epoch": 2.3160377358490565,
+      "grad_norm": 0.19188454780775782,
+      "learning_rate": 1.2691131498470949e-05,
+      "loss": 0.2993,
+      "step": 1964
+    },
+    {
+      "epoch": 2.3172169811320753,
+      "grad_norm": 0.21549107246125326,
+      "learning_rate": 1.2669287898645699e-05,
+      "loss": 0.3046,
+      "step": 1965
+    },
+    {
+      "epoch": 2.3183962264150946,
+      "grad_norm": 0.1846799635306901,
+      "learning_rate": 1.2647444298820446e-05,
+      "loss": 0.3064,
+      "step": 1966
+    },
+    {
+      "epoch": 2.3195754716981134,
+      "grad_norm": 0.20669127965561931,
+      "learning_rate": 1.2625600698995196e-05,
+      "loss": 0.317,
+      "step": 1967
+    },
+    {
+      "epoch": 2.3207547169811322,
+      "grad_norm": 0.20221689161278708,
+      "learning_rate": 1.2603757099169943e-05,
+      "loss": 0.2869,
+      "step": 1968
+    },
+    {
+      "epoch": 2.321933962264151,
+      "grad_norm": 0.1949018556490392,
+      "learning_rate": 1.2581913499344692e-05,
+      "loss": 0.2971,
+      "step": 1969
+    },
+    {
+      "epoch": 2.32311320754717,
+      "grad_norm": 0.20680028379220755,
+      "learning_rate": 1.256006989951944e-05,
+      "loss": 0.3178,
+      "step": 1970
+    },
+    {
+      "epoch": 2.3242924528301887,
+      "grad_norm": 0.19283322466630975,
+      "learning_rate": 1.253822629969419e-05,
+      "loss": 0.3142,
+      "step": 1971
+    },
+    {
+      "epoch": 2.3254716981132075,
+      "grad_norm": 0.21934978923218287,
+      "learning_rate": 1.251638269986894e-05,
+      "loss": 0.331,
+      "step": 1972
+    },
+    {
+      "epoch": 2.3266509433962264,
+      "grad_norm": 0.1975238553600828,
+      "learning_rate": 1.2494539100043689e-05,
+      "loss": 0.31,
+      "step": 1973
+    },
+    {
+      "epoch": 2.327830188679245,
+      "grad_norm": 0.19550054602995573,
+      "learning_rate": 1.2472695500218437e-05,
+      "loss": 0.3096,
+      "step": 1974
+    },
+    {
+      "epoch": 2.329009433962264,
+      "grad_norm": 0.18625196189249027,
+      "learning_rate": 1.2450851900393186e-05,
+      "loss": 0.3056,
+      "step": 1975
+    },
+    {
+      "epoch": 2.330188679245283,
+      "grad_norm": 0.19121919870898604,
+      "learning_rate": 1.2429008300567935e-05,
+      "loss": 0.306,
+      "step": 1976
+    },
+    {
+      "epoch": 2.331367924528302,
+      "grad_norm": 0.20048688206859577,
+      "learning_rate": 1.2407164700742683e-05,
+      "loss": 0.2759,
+      "step": 1977
+    },
+    {
+      "epoch": 2.332547169811321,
+      "grad_norm": 0.20350553937922167,
+      "learning_rate": 1.2385321100917432e-05,
+      "loss": 0.3019,
+      "step": 1978
+    },
+    {
+      "epoch": 2.3337264150943398,
+      "grad_norm": 0.1965790695294183,
+      "learning_rate": 1.236347750109218e-05,
+      "loss": 0.3048,
+      "step": 1979
+    },
+    {
+      "epoch": 2.3349056603773586,
+      "grad_norm": 0.1931354914312612,
+      "learning_rate": 1.234163390126693e-05,
+      "loss": 0.3035,
+      "step": 1980
+    },
+    {
+      "epoch": 2.3360849056603774,
+      "grad_norm": 0.19815625174353566,
+      "learning_rate": 1.2319790301441678e-05,
+      "loss": 0.3191,
+      "step": 1981
+    },
+    {
+      "epoch": 2.3372641509433962,
+      "grad_norm": 0.2016431049277712,
+      "learning_rate": 1.2297946701616427e-05,
+      "loss": 0.3104,
+      "step": 1982
+    },
+    {
+      "epoch": 2.338443396226415,
+      "grad_norm": 0.1941485806911156,
+      "learning_rate": 1.2276103101791176e-05,
+      "loss": 0.3031,
+      "step": 1983
+    },
+    {
+      "epoch": 2.339622641509434,
+      "grad_norm": 0.19933696117828348,
+      "learning_rate": 1.2254259501965924e-05,
+      "loss": 0.303,
+      "step": 1984
+    },
+    {
+      "epoch": 2.3408018867924527,
+      "grad_norm": 0.19547313322951276,
+      "learning_rate": 1.2232415902140673e-05,
+      "loss": 0.3034,
+      "step": 1985
+    },
+    {
+      "epoch": 2.3419811320754715,
+      "grad_norm": 0.20920339426117138,
+      "learning_rate": 1.2210572302315422e-05,
+      "loss": 0.3129,
+      "step": 1986
+    },
+    {
+      "epoch": 2.3431603773584904,
+      "grad_norm": 0.20299503907349972,
+      "learning_rate": 1.218872870249017e-05,
+      "loss": 0.3125,
+      "step": 1987
+    },
+    {
+      "epoch": 2.3443396226415096,
+      "grad_norm": 0.1954679999890336,
+      "learning_rate": 1.2166885102664919e-05,
+      "loss": 0.3211,
+      "step": 1988
+    },
+    {
+      "epoch": 2.3455188679245285,
+      "grad_norm": 0.19261523278741946,
+      "learning_rate": 1.2145041502839668e-05,
+      "loss": 0.3112,
+      "step": 1989
+    },
+    {
+      "epoch": 2.3466981132075473,
+      "grad_norm": 0.19702849343798712,
+      "learning_rate": 1.2123197903014418e-05,
+      "loss": 0.3119,
+      "step": 1990
+    },
+    {
+      "epoch": 2.347877358490566,
+      "grad_norm": 0.20031122093878354,
+      "learning_rate": 1.2101354303189167e-05,
+      "loss": 0.29,
+      "step": 1991
+    },
+    {
+      "epoch": 2.349056603773585,
+      "grad_norm": 0.20492461026807285,
+      "learning_rate": 1.2079510703363916e-05,
+      "loss": 0.2955,
+      "step": 1992
+    },
+    {
+      "epoch": 2.3502358490566038,
+      "grad_norm": 0.19742445415587875,
+      "learning_rate": 1.2057667103538664e-05,
+      "loss": 0.3166,
+      "step": 1993
+    },
+    {
+      "epoch": 2.3514150943396226,
+      "grad_norm": 0.20005824302633762,
+      "learning_rate": 1.2035823503713413e-05,
+      "loss": 0.3178,
+      "step": 1994
+    },
+    {
+      "epoch": 2.3525943396226414,
+      "grad_norm": 0.1831883572348429,
+      "learning_rate": 1.2013979903888162e-05,
+      "loss": 0.2965,
+      "step": 1995
+    },
+    {
+      "epoch": 2.3537735849056602,
+      "grad_norm": 0.19522703560793492,
+      "learning_rate": 1.199213630406291e-05,
+      "loss": 0.3106,
+      "step": 1996
+    },
+    {
+      "epoch": 2.354952830188679,
+      "grad_norm": 0.19919271835707458,
+      "learning_rate": 1.1970292704237659e-05,
+      "loss": 0.3193,
+      "step": 1997
+    },
+    {
+      "epoch": 2.356132075471698,
+      "grad_norm": 0.2108033172581886,
+      "learning_rate": 1.1948449104412408e-05,
+      "loss": 0.3193,
+      "step": 1998
+    },
+    {
+      "epoch": 2.357311320754717,
+      "grad_norm": 0.20697124335486997,
+      "learning_rate": 1.1926605504587156e-05,
+      "loss": 0.3155,
+      "step": 1999
+    },
+    {
+      "epoch": 2.358490566037736,
+      "grad_norm": 0.19616033843848504,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 0.3072,
+      "step": 2000
+    },
+    {
+      "epoch": 2.359669811320755,
+      "grad_norm": 0.19885238312456885,
+      "learning_rate": 1.1882918304936654e-05,
+      "loss": 0.3187,
+      "step": 2001
+    },
+    {
+      "epoch": 2.3608490566037736,
+      "grad_norm": 0.20886468939421285,
+      "learning_rate": 1.1861074705111402e-05,
+      "loss": 0.3037,
+      "step": 2002
+    },
+    {
+      "epoch": 2.3620283018867925,
+      "grad_norm": 0.2079269173574578,
+      "learning_rate": 1.1839231105286151e-05,
+      "loss": 0.3239,
+      "step": 2003
+    },
+    {
+      "epoch": 2.3632075471698113,
+      "grad_norm": 0.1892195753064003,
+      "learning_rate": 1.18173875054609e-05,
+      "loss": 0.2963,
+      "step": 2004
+    },
+    {
+      "epoch": 2.36438679245283,
+      "grad_norm": 0.1971430251975,
+      "learning_rate": 1.1795543905635649e-05,
+      "loss": 0.3096,
+      "step": 2005
+    },
+    {
+      "epoch": 2.365566037735849,
+      "grad_norm": 0.19936795993590892,
+      "learning_rate": 1.1773700305810397e-05,
+      "loss": 0.3178,
+      "step": 2006
+    },
+    {
+      "epoch": 2.3667452830188678,
+      "grad_norm": 0.19866181470378763,
+      "learning_rate": 1.1751856705985146e-05,
+      "loss": 0.328,
+      "step": 2007
+    },
+    {
+      "epoch": 2.3679245283018866,
+      "grad_norm": 0.18946763297127844,
+      "learning_rate": 1.1730013106159896e-05,
+      "loss": 0.3085,
+      "step": 2008
+    },
+    {
+      "epoch": 2.3691037735849054,
+      "grad_norm": 0.2043276656523459,
+      "learning_rate": 1.1708169506334645e-05,
+      "loss": 0.313,
+      "step": 2009
+    },
+    {
+      "epoch": 2.3702830188679247,
+      "grad_norm": 0.18551195276687002,
+      "learning_rate": 1.1686325906509394e-05,
+      "loss": 0.3047,
+      "step": 2010
+    },
+    {
+      "epoch": 2.3714622641509435,
+      "grad_norm": 0.2009593869345854,
+      "learning_rate": 1.1664482306684142e-05,
+      "loss": 0.3034,
+      "step": 2011
+    },
+    {
+      "epoch": 2.3726415094339623,
+      "grad_norm": 0.18939053442848844,
+      "learning_rate": 1.1642638706858891e-05,
+      "loss": 0.3109,
+      "step": 2012
+    },
+    {
+      "epoch": 2.373820754716981,
+      "grad_norm": 0.19810635474784083,
+      "learning_rate": 1.162079510703364e-05,
+      "loss": 0.3164,
+      "step": 2013
+    },
+    {
+      "epoch": 2.375,
+      "grad_norm": 0.1982561014836671,
+      "learning_rate": 1.1598951507208389e-05,
+      "loss": 0.3029,
+      "step": 2014
+    },
+    {
+      "epoch": 2.376179245283019,
+      "grad_norm": 0.19663492983921044,
+      "learning_rate": 1.1577107907383137e-05,
+      "loss": 0.31,
+      "step": 2015
+    },
+    {
+      "epoch": 2.3773584905660377,
+      "grad_norm": 0.18389983077221983,
+      "learning_rate": 1.1555264307557886e-05,
+      "loss": 0.3134,
+      "step": 2016
+    },
+    {
+      "epoch": 2.3785377358490565,
+      "grad_norm": 0.20657714312111985,
+      "learning_rate": 1.1533420707732635e-05,
+      "loss": 0.3178,
+      "step": 2017
+    },
+    {
+      "epoch": 2.3797169811320753,
+      "grad_norm": 0.20142808285994854,
+      "learning_rate": 1.1511577107907385e-05,
+      "loss": 0.3112,
+      "step": 2018
+    },
+    {
+      "epoch": 2.3808962264150946,
+      "grad_norm": 0.1904368238811153,
+      "learning_rate": 1.1489733508082134e-05,
+      "loss": 0.32,
+      "step": 2019
+    },
+    {
+      "epoch": 2.3820754716981134,
+      "grad_norm": 0.18783714736801382,
+      "learning_rate": 1.1467889908256882e-05,
+      "loss": 0.314,
+      "step": 2020
+    },
+    {
+      "epoch": 2.3832547169811322,
+      "grad_norm": 0.18804644794811018,
+      "learning_rate": 1.144604630843163e-05,
+      "loss": 0.3192,
+      "step": 2021
+    },
+    {
+      "epoch": 2.384433962264151,
+      "grad_norm": 0.18765232092236517,
+      "learning_rate": 1.1424202708606378e-05,
+      "loss": 0.302,
+      "step": 2022
+    },
+    {
+      "epoch": 2.38561320754717,
+      "grad_norm": 0.19163813169518115,
+      "learning_rate": 1.1402359108781127e-05,
+      "loss": 0.295,
+      "step": 2023
+    },
+    {
+      "epoch": 2.3867924528301887,
+      "grad_norm": 0.1994868355963002,
+      "learning_rate": 1.1380515508955875e-05,
+      "loss": 0.3121,
+      "step": 2024
+    },
+    {
+      "epoch": 2.3879716981132075,
+      "grad_norm": 0.20294996706656157,
+      "learning_rate": 1.1358671909130624e-05,
+      "loss": 0.324,
+      "step": 2025
+    },
+    {
+      "epoch": 2.3891509433962264,
+      "grad_norm": 0.20285097410858,
+      "learning_rate": 1.1336828309305373e-05,
+      "loss": 0.3206,
+      "step": 2026
+    },
+    {
+      "epoch": 2.390330188679245,
+      "grad_norm": 0.19152558121462768,
+      "learning_rate": 1.1314984709480123e-05,
+      "loss": 0.3161,
+      "step": 2027
+    },
+    {
+      "epoch": 2.391509433962264,
+      "grad_norm": 0.19461703777752304,
+      "learning_rate": 1.1293141109654872e-05,
+      "loss": 0.3204,
+      "step": 2028
+    },
+    {
+      "epoch": 2.392688679245283,
+      "grad_norm": 0.17904523577062287,
+      "learning_rate": 1.127129750982962e-05,
+      "loss": 0.3016,
+      "step": 2029
+    },
+    {
+      "epoch": 2.393867924528302,
+      "grad_norm": 0.2067778312761449,
+      "learning_rate": 1.124945391000437e-05,
+      "loss": 0.3154,
+      "step": 2030
+    },
+    {
+      "epoch": 2.395047169811321,
+      "grad_norm": 0.18274564339461588,
+      "learning_rate": 1.1227610310179118e-05,
+      "loss": 0.3026,
+      "step": 2031
+    },
+    {
+      "epoch": 2.3962264150943398,
+      "grad_norm": 0.18182227947986937,
+      "learning_rate": 1.1205766710353867e-05,
+      "loss": 0.3087,
+      "step": 2032
+    },
+    {
+      "epoch": 2.3974056603773586,
+      "grad_norm": 0.21113764089577589,
+      "learning_rate": 1.1183923110528615e-05,
+      "loss": 0.3299,
+      "step": 2033
+    },
+    {
+      "epoch": 2.3985849056603774,
+      "grad_norm": 0.2053874116495947,
+      "learning_rate": 1.1162079510703364e-05,
+      "loss": 0.3235,
+      "step": 2034
+    },
+    {
+      "epoch": 2.3997641509433962,
+      "grad_norm": 0.192147222262657,
+      "learning_rate": 1.1140235910878113e-05,
+      "loss": 0.3273,
+      "step": 2035
+    },
+    {
+      "epoch": 2.400943396226415,
+      "grad_norm": 0.19693952429584277,
+      "learning_rate": 1.1118392311052862e-05,
+      "loss": 0.3231,
+      "step": 2036
+    },
+    {
+      "epoch": 2.402122641509434,
+      "grad_norm": 0.2013339510460299,
+      "learning_rate": 1.1096548711227612e-05,
+      "loss": 0.3025,
+      "step": 2037
+    },
+    {
+      "epoch": 2.4033018867924527,
+      "grad_norm": 0.18664456827262657,
+      "learning_rate": 1.107470511140236e-05,
+      "loss": 0.3202,
+      "step": 2038
+    },
+    {
+      "epoch": 2.4044811320754715,
+      "grad_norm": 0.17458518041004067,
+      "learning_rate": 1.105286151157711e-05,
+      "loss": 0.3038,
+      "step": 2039
+    },
+    {
+      "epoch": 2.4056603773584904,
+      "grad_norm": 0.1966569401114815,
+      "learning_rate": 1.1031017911751858e-05,
+      "loss": 0.3168,
+      "step": 2040
+    },
+    {
+      "epoch": 2.4068396226415096,
+      "grad_norm": 0.1998412903744102,
+      "learning_rate": 1.1009174311926607e-05,
+      "loss": 0.3152,
+      "step": 2041
+    },
+    {
+      "epoch": 2.4080188679245285,
+      "grad_norm": 0.18919322306030947,
+      "learning_rate": 1.0987330712101355e-05,
+      "loss": 0.3343,
+      "step": 2042
+    },
+    {
+      "epoch": 2.4091981132075473,
+      "grad_norm": 0.19036663145677973,
+      "learning_rate": 1.0965487112276102e-05,
+      "loss": 0.3115,
+      "step": 2043
+    },
+    {
+      "epoch": 2.410377358490566,
+      "grad_norm": 0.19018195653092537,
+      "learning_rate": 1.0943643512450851e-05,
+      "loss": 0.3088,
+      "step": 2044
+    },
+    {
+      "epoch": 2.411556603773585,
+      "grad_norm": 0.19704656373993695,
+      "learning_rate": 1.0921799912625601e-05,
+      "loss": 0.2937,
+      "step": 2045
+    },
+    {
+      "epoch": 2.4127358490566038,
+      "grad_norm": 0.19557962058756373,
+      "learning_rate": 1.089995631280035e-05,
+      "loss": 0.3275,
+      "step": 2046
+    },
+    {
+      "epoch": 2.4139150943396226,
+      "grad_norm": 0.1898911723897887,
+      "learning_rate": 1.0878112712975099e-05,
+      "loss": 0.3042,
+      "step": 2047
+    },
+    {
+      "epoch": 2.4150943396226414,
+      "grad_norm": 0.19505469178006987,
+      "learning_rate": 1.0856269113149848e-05,
+      "loss": 0.3054,
+      "step": 2048
+    },
+    {
+      "epoch": 2.4162735849056602,
+      "grad_norm": 0.1802770081806493,
+      "learning_rate": 1.0834425513324596e-05,
+      "loss": 0.3016,
+      "step": 2049
+    },
+    {
+      "epoch": 2.417452830188679,
+      "grad_norm": 0.203856552314741,
+      "learning_rate": 1.0812581913499345e-05,
+      "loss": 0.3211,
+      "step": 2050
+    },
+    {
+      "epoch": 2.418632075471698,
+      "grad_norm": 0.1927288412316893,
+      "learning_rate": 1.0790738313674094e-05,
+      "loss": 0.303,
+      "step": 2051
+    },
+    {
+      "epoch": 2.419811320754717,
+      "grad_norm": 0.1967127793107974,
+      "learning_rate": 1.0768894713848842e-05,
+      "loss": 0.3329,
+      "step": 2052
+    },
+    {
+      "epoch": 2.420990566037736,
+      "grad_norm": 0.185382729419658,
+      "learning_rate": 1.0747051114023591e-05,
+      "loss": 0.3163,
+      "step": 2053
+    },
+    {
+      "epoch": 2.422169811320755,
+      "grad_norm": 0.19797988035772227,
+      "learning_rate": 1.072520751419834e-05,
+      "loss": 0.3157,
+      "step": 2054
+    },
+    {
+      "epoch": 2.4233490566037736,
+      "grad_norm": 0.18673395751968688,
+      "learning_rate": 1.070336391437309e-05,
+      "loss": 0.3003,
+      "step": 2055
+    },
+    {
+      "epoch": 2.4245283018867925,
+      "grad_norm": 0.17981604926512862,
+      "learning_rate": 1.0681520314547839e-05,
+      "loss": 0.293,
+      "step": 2056
+    },
+    {
+      "epoch": 2.4257075471698113,
+      "grad_norm": 0.18595981335215248,
+      "learning_rate": 1.0659676714722588e-05,
+      "loss": 0.3008,
+      "step": 2057
+    },
+    {
+      "epoch": 2.42688679245283,
+      "grad_norm": 0.2015633136260089,
+      "learning_rate": 1.0637833114897336e-05,
+      "loss": 0.3025,
+      "step": 2058
+    },
+    {
+      "epoch": 2.428066037735849,
+      "grad_norm": 0.17848078788570546,
+      "learning_rate": 1.0615989515072085e-05,
+      "loss": 0.292,
+      "step": 2059
+    },
+    {
+      "epoch": 2.4292452830188678,
+      "grad_norm": 0.18262810033156854,
+      "learning_rate": 1.0594145915246834e-05,
+      "loss": 0.3087,
+      "step": 2060
+    },
+    {
+      "epoch": 2.4304245283018866,
+      "grad_norm": 0.18678912106388662,
+      "learning_rate": 1.0572302315421582e-05,
+      "loss": 0.3179,
+      "step": 2061
+    },
+    {
+      "epoch": 2.4316037735849054,
+      "grad_norm": 0.1986361786079406,
+      "learning_rate": 1.0550458715596331e-05,
+      "loss": 0.3058,
+      "step": 2062
+    },
+    {
+      "epoch": 2.4327830188679247,
+      "grad_norm": 0.1911497518289234,
+      "learning_rate": 1.052861511577108e-05,
+      "loss": 0.3231,
+      "step": 2063
+    },
+    {
+      "epoch": 2.4339622641509435,
+      "grad_norm": 0.18610544251106834,
+      "learning_rate": 1.0506771515945828e-05,
+      "loss": 0.3079,
+      "step": 2064
+    },
+    {
+      "epoch": 2.4351415094339623,
+      "grad_norm": 0.19814952082782622,
+      "learning_rate": 1.0484927916120577e-05,
+      "loss": 0.3199,
+      "step": 2065
+    },
+    {
+      "epoch": 2.436320754716981,
+      "grad_norm": 0.1940086015823025,
+      "learning_rate": 1.0463084316295326e-05,
+      "loss": 0.3283,
+      "step": 2066
+    },
+    {
+      "epoch": 2.4375,
+      "grad_norm": 0.18595063977319343,
+      "learning_rate": 1.0441240716470074e-05,
+      "loss": 0.3225,
+      "step": 2067
+    },
+    {
+      "epoch": 2.438679245283019,
+      "grad_norm": 0.17963693561750205,
+      "learning_rate": 1.0419397116644823e-05,
+      "loss": 0.3086,
+      "step": 2068
+    },
+    {
+      "epoch": 2.4398584905660377,
+      "grad_norm": 0.18679928340239646,
+      "learning_rate": 1.0397553516819572e-05,
+      "loss": 0.3296,
+      "step": 2069
+    },
+    {
+      "epoch": 2.4410377358490565,
+      "grad_norm": 0.19421561821559358,
+      "learning_rate": 1.037570991699432e-05,
+      "loss": 0.3109,
+      "step": 2070
+    },
+    {
+      "epoch": 2.4422169811320753,
+      "grad_norm": 0.2213160125786252,
+      "learning_rate": 1.035386631716907e-05,
+      "loss": 0.3275,
+      "step": 2071
+    },
+    {
+      "epoch": 2.4433962264150946,
+      "grad_norm": 0.18864162662440068,
+      "learning_rate": 1.0332022717343818e-05,
+      "loss": 0.3253,
+      "step": 2072
+    },
+    {
+      "epoch": 2.4445754716981134,
+      "grad_norm": 0.1837302716435305,
+      "learning_rate": 1.0310179117518567e-05,
+      "loss": 0.2943,
+      "step": 2073
+    },
+    {
+      "epoch": 2.4457547169811322,
+      "grad_norm": 0.18858961422679346,
+      "learning_rate": 1.0288335517693317e-05,
+      "loss": 0.32,
+      "step": 2074
+    },
+    {
+      "epoch": 2.446933962264151,
+      "grad_norm": 0.19331437296470316,
+      "learning_rate": 1.0266491917868066e-05,
+      "loss": 0.3058,
+      "step": 2075
+    },
+    {
+      "epoch": 2.44811320754717,
+      "grad_norm": 0.19046971399025775,
+      "learning_rate": 1.0244648318042814e-05,
+      "loss": 0.2979,
+      "step": 2076
+    },
+    {
+      "epoch": 2.4492924528301887,
+      "grad_norm": 0.1887436873149276,
+      "learning_rate": 1.0222804718217563e-05,
+      "loss": 0.3028,
+      "step": 2077
+    },
+    {
+      "epoch": 2.4504716981132075,
+      "grad_norm": 0.1824084807648471,
+      "learning_rate": 1.0200961118392312e-05,
+      "loss": 0.3002,
+      "step": 2078
+    },
+    {
+      "epoch": 2.4516509433962264,
+      "grad_norm": 0.18976437319850098,
+      "learning_rate": 1.017911751856706e-05,
+      "loss": 0.3152,
+      "step": 2079
+    },
+    {
+      "epoch": 2.452830188679245,
+      "grad_norm": 0.19744248395345695,
+      "learning_rate": 1.015727391874181e-05,
+      "loss": 0.321,
+      "step": 2080
+    },
+    {
+      "epoch": 2.454009433962264,
+      "grad_norm": 0.18761088970807466,
+      "learning_rate": 1.0135430318916558e-05,
+      "loss": 0.3049,
+      "step": 2081
+    },
+    {
+      "epoch": 2.455188679245283,
+      "grad_norm": 0.1900550052699813,
+      "learning_rate": 1.0113586719091307e-05,
+      "loss": 0.305,
+      "step": 2082
+    },
+    {
+      "epoch": 2.456367924528302,
+      "grad_norm": 0.1881520553119906,
+      "learning_rate": 1.0091743119266055e-05,
+      "loss": 0.313,
+      "step": 2083
+    },
+    {
+      "epoch": 2.457547169811321,
+      "grad_norm": 0.19647740255535784,
+      "learning_rate": 1.0069899519440806e-05,
+      "loss": 0.3095,
+      "step": 2084
+    },
+    {
+      "epoch": 2.4587264150943398,
+      "grad_norm": 0.1928921942135357,
+      "learning_rate": 1.0048055919615554e-05,
+      "loss": 0.3079,
+      "step": 2085
+    },
+    {
+      "epoch": 2.4599056603773586,
+      "grad_norm": 0.17907865866607706,
+      "learning_rate": 1.0026212319790301e-05,
+      "loss": 0.2975,
+      "step": 2086
+    },
+    {
+      "epoch": 2.4610849056603774,
+      "grad_norm": 0.19515100460188364,
+      "learning_rate": 1.000436871996505e-05,
+      "loss": 0.3075,
+      "step": 2087
+    },
+    {
+      "epoch": 2.4622641509433962,
+      "grad_norm": 0.19626585358894813,
+      "learning_rate": 9.982525120139799e-06,
+      "loss": 0.3175,
+      "step": 2088
+    },
+    {
+      "epoch": 2.463443396226415,
+      "grad_norm": 0.18175989753686542,
+      "learning_rate": 9.960681520314547e-06,
+      "loss": 0.2967,
+      "step": 2089
+    },
+    {
+      "epoch": 2.464622641509434,
+      "grad_norm": 0.21848912430258843,
+      "learning_rate": 9.938837920489296e-06,
+      "loss": 0.3343,
+      "step": 2090
+    },
+    {
+      "epoch": 2.4658018867924527,
+      "grad_norm": 0.20936586548216438,
+      "learning_rate": 9.916994320664045e-06,
+      "loss": 0.3012,
+      "step": 2091
+    },
+    {
+      "epoch": 2.4669811320754715,
+      "grad_norm": 0.17949404913824213,
+      "learning_rate": 9.895150720838795e-06,
+      "loss": 0.3068,
+      "step": 2092
+    },
+    {
+      "epoch": 2.4681603773584904,
+      "grad_norm": 0.2032556594635456,
+      "learning_rate": 9.873307121013544e-06,
+      "loss": 0.3255,
+      "step": 2093
+    },
+    {
+      "epoch": 2.4693396226415096,
+      "grad_norm": 0.19166835070735352,
+      "learning_rate": 9.851463521188293e-06,
+      "loss": 0.2989,
+      "step": 2094
+    },
+    {
+      "epoch": 2.4705188679245285,
+      "grad_norm": 0.18139630778480897,
+      "learning_rate": 9.829619921363041e-06,
+      "loss": 0.3124,
+      "step": 2095
+    },
+    {
+      "epoch": 2.4716981132075473,
+      "grad_norm": 0.18755774464991426,
+      "learning_rate": 9.80777632153779e-06,
+      "loss": 0.3196,
+      "step": 2096
+    },
+    {
+      "epoch": 2.472877358490566,
+      "grad_norm": 0.17482240084008582,
+      "learning_rate": 9.785932721712539e-06,
+      "loss": 0.2981,
+      "step": 2097
+    },
+    {
+      "epoch": 2.474056603773585,
+      "grad_norm": 0.18443480926342176,
+      "learning_rate": 9.764089121887287e-06,
+      "loss": 0.299,
+      "step": 2098
+    },
+    {
+      "epoch": 2.4752358490566038,
+      "grad_norm": 0.1873559423697444,
+      "learning_rate": 9.742245522062036e-06,
+      "loss": 0.2965,
+      "step": 2099
+    },
+    {
+      "epoch": 2.4764150943396226,
+      "grad_norm": 0.2190482957871864,
+      "learning_rate": 9.720401922236785e-06,
+      "loss": 0.3164,
+      "step": 2100
+    },
+    {
+      "epoch": 2.4775943396226414,
+      "grad_norm": 0.19086939998724958,
+      "learning_rate": 9.698558322411533e-06,
+      "loss": 0.3018,
+      "step": 2101
+    },
+    {
+      "epoch": 2.4787735849056602,
+      "grad_norm": 0.19142879019453185,
+      "learning_rate": 9.676714722586284e-06,
+      "loss": 0.3199,
+      "step": 2102
+    },
+    {
+      "epoch": 2.479952830188679,
+      "grad_norm": 0.1896898304334053,
+      "learning_rate": 9.654871122761033e-06,
+      "loss": 0.2963,
+      "step": 2103
+    },
+    {
+      "epoch": 2.481132075471698,
+      "grad_norm": 0.20289685668285223,
+      "learning_rate": 9.633027522935781e-06,
+      "loss": 0.3196,
+      "step": 2104
+    },
+    {
+      "epoch": 2.482311320754717,
+      "grad_norm": 0.19667270300049164,
+      "learning_rate": 9.61118392311053e-06,
+      "loss": 0.3227,
+      "step": 2105
+    },
+    {
+      "epoch": 2.483490566037736,
+      "grad_norm": 0.1959344036263567,
+      "learning_rate": 9.589340323285279e-06,
+      "loss": 0.313,
+      "step": 2106
+    },
+    {
+      "epoch": 2.484669811320755,
+      "grad_norm": 0.2152487986423751,
+      "learning_rate": 9.567496723460026e-06,
+      "loss": 0.3307,
+      "step": 2107
+    },
+    {
+      "epoch": 2.4858490566037736,
+      "grad_norm": 0.19176448269198962,
+      "learning_rate": 9.545653123634774e-06,
+      "loss": 0.3289,
+      "step": 2108
+    },
+    {
+      "epoch": 2.4870283018867925,
+      "grad_norm": 0.18937560597574432,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.3174,
+      "step": 2109
+    },
+    {
+      "epoch": 2.4882075471698113,
+      "grad_norm": 0.19797933639674067,
+      "learning_rate": 9.501965923984273e-06,
+      "loss": 0.3308,
+      "step": 2110
+    },
+    {
+      "epoch": 2.48938679245283,
+      "grad_norm": 0.20068778391035696,
+      "learning_rate": 9.480122324159022e-06,
+      "loss": 0.3121,
+      "step": 2111
+    },
+    {
+      "epoch": 2.490566037735849,
+      "grad_norm": 0.18141254516036356,
+      "learning_rate": 9.45827872433377e-06,
+      "loss": 0.2897,
+      "step": 2112
+    },
+    {
+      "epoch": 2.4917452830188678,
+      "grad_norm": 0.1823523196803717,
+      "learning_rate": 9.43643512450852e-06,
+      "loss": 0.3175,
+      "step": 2113
+    },
+    {
+      "epoch": 2.4929245283018866,
+      "grad_norm": 0.20301030268754022,
+      "learning_rate": 9.414591524683268e-06,
+      "loss": 0.3172,
+      "step": 2114
+    },
+    {
+      "epoch": 2.4941037735849054,
+      "grad_norm": 0.18486001544313976,
+      "learning_rate": 9.392747924858017e-06,
+      "loss": 0.3078,
+      "step": 2115
+    },
+    {
+      "epoch": 2.4952830188679247,
+      "grad_norm": 0.19319113557662798,
+      "learning_rate": 9.370904325032766e-06,
+      "loss": 0.3163,
+      "step": 2116
+    },
+    {
+      "epoch": 2.4964622641509435,
+      "grad_norm": 0.18987621181990952,
+      "learning_rate": 9.349060725207514e-06,
+      "loss": 0.3155,
+      "step": 2117
+    },
+    {
+      "epoch": 2.4976415094339623,
+      "grad_norm": 0.20372335797980107,
+      "learning_rate": 9.327217125382263e-06,
+      "loss": 0.3245,
+      "step": 2118
+    },
+    {
+      "epoch": 2.498820754716981,
+      "grad_norm": 0.18388982872893378,
+      "learning_rate": 9.305373525557012e-06,
+      "loss": 0.2996,
+      "step": 2119
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.19485223375251565,
+      "learning_rate": 9.28352992573176e-06,
+      "loss": 0.3156,
+      "step": 2120
+    },
+    {
+      "epoch": 2.501179245283019,
+      "grad_norm": 0.18843372431396313,
+      "learning_rate": 9.26168632590651e-06,
+      "loss": 0.3173,
+      "step": 2121
+    },
+    {
+      "epoch": 2.5023584905660377,
+      "grad_norm": 0.19586883610774788,
+      "learning_rate": 9.23984272608126e-06,
+      "loss": 0.3114,
+      "step": 2122
+    },
+    {
+      "epoch": 2.5035377358490565,
+      "grad_norm": 0.19859496443868235,
+      "learning_rate": 9.217999126256008e-06,
+      "loss": 0.312,
+      "step": 2123
+    },
+    {
+      "epoch": 2.5047169811320753,
+      "grad_norm": 0.18448407814429837,
+      "learning_rate": 9.196155526430757e-06,
+      "loss": 0.2898,
+      "step": 2124
+    },
+    {
+      "epoch": 2.5058962264150946,
+      "grad_norm": 0.1929831596561828,
+      "learning_rate": 9.174311926605506e-06,
+      "loss": 0.3174,
+      "step": 2125
+    },
+    {
+      "epoch": 2.507075471698113,
+      "grad_norm": 0.2008759748940534,
+      "learning_rate": 9.152468326780254e-06,
+      "loss": 0.3071,
+      "step": 2126
+    },
+    {
+      "epoch": 2.5082547169811322,
+      "grad_norm": 0.19892621752497816,
+      "learning_rate": 9.130624726955003e-06,
+      "loss": 0.3,
+      "step": 2127
+    },
+    {
+      "epoch": 2.509433962264151,
+      "grad_norm": 0.19350193096815252,
+      "learning_rate": 9.10878112712975e-06,
+      "loss": 0.2945,
+      "step": 2128
+    },
+    {
+      "epoch": 2.51061320754717,
+      "grad_norm": 0.19783423635804215,
+      "learning_rate": 9.0869375273045e-06,
+      "loss": 0.3116,
+      "step": 2129
+    },
+    {
+      "epoch": 2.5117924528301887,
+      "grad_norm": 0.20264900433007627,
+      "learning_rate": 9.065093927479249e-06,
+      "loss": 0.3003,
+      "step": 2130
+    },
+    {
+      "epoch": 2.5129716981132075,
+      "grad_norm": 0.2176781994931378,
+      "learning_rate": 9.043250327653998e-06,
+      "loss": 0.3379,
+      "step": 2131
+    },
+    {
+      "epoch": 2.5141509433962264,
+      "grad_norm": 0.2075175881630815,
+      "learning_rate": 9.021406727828746e-06,
+      "loss": 0.3171,
+      "step": 2132
+    },
+    {
+      "epoch": 2.515330188679245,
+      "grad_norm": 0.1888352522813758,
+      "learning_rate": 8.999563128003495e-06,
+      "loss": 0.3066,
+      "step": 2133
+    },
+    {
+      "epoch": 2.516509433962264,
+      "grad_norm": 0.20348238481654496,
+      "learning_rate": 8.977719528178244e-06,
+      "loss": 0.3103,
+      "step": 2134
+    },
+    {
+      "epoch": 2.517688679245283,
+      "grad_norm": 0.18484021198867595,
+      "learning_rate": 8.955875928352993e-06,
+      "loss": 0.2976,
+      "step": 2135
+    },
+    {
+      "epoch": 2.518867924528302,
+      "grad_norm": 0.18824954439589786,
+      "learning_rate": 8.934032328527741e-06,
+      "loss": 0.3167,
+      "step": 2136
+    },
+    {
+      "epoch": 2.5200471698113205,
+      "grad_norm": 0.19590402139329557,
+      "learning_rate": 8.91218872870249e-06,
+      "loss": 0.3347,
+      "step": 2137
+    },
+    {
+      "epoch": 2.5212264150943398,
+      "grad_norm": 0.20325512070583318,
+      "learning_rate": 8.890345128877239e-06,
+      "loss": 0.334,
+      "step": 2138
+    },
+    {
+      "epoch": 2.5224056603773586,
+      "grad_norm": 0.1990175856932863,
+      "learning_rate": 8.868501529051989e-06,
+      "loss": 0.3257,
+      "step": 2139
+    },
+    {
+      "epoch": 2.5235849056603774,
+      "grad_norm": 0.18558705589054883,
+      "learning_rate": 8.846657929226738e-06,
+      "loss": 0.2992,
+      "step": 2140
+    },
+    {
+      "epoch": 2.5247641509433962,
+      "grad_norm": 0.17777529671596856,
+      "learning_rate": 8.824814329401486e-06,
+      "loss": 0.2987,
+      "step": 2141
+    },
+    {
+      "epoch": 2.525943396226415,
+      "grad_norm": 0.20337319258019323,
+      "learning_rate": 8.802970729576235e-06,
+      "loss": 0.2984,
+      "step": 2142
+    },
+    {
+      "epoch": 2.527122641509434,
+      "grad_norm": 0.19263593461418027,
+      "learning_rate": 8.781127129750984e-06,
+      "loss": 0.3181,
+      "step": 2143
+    },
+    {
+      "epoch": 2.5283018867924527,
+      "grad_norm": 0.19289530837651161,
+      "learning_rate": 8.759283529925732e-06,
+      "loss": 0.3108,
+      "step": 2144
+    },
+    {
+      "epoch": 2.5294811320754715,
+      "grad_norm": 0.18546000914980734,
+      "learning_rate": 8.737439930100481e-06,
+      "loss": 0.3059,
+      "step": 2145
+    },
+    {
+      "epoch": 2.5306603773584904,
+      "grad_norm": 0.18781891618188698,
+      "learning_rate": 8.71559633027523e-06,
+      "loss": 0.3066,
+      "step": 2146
+    },
+    {
+      "epoch": 2.5318396226415096,
+      "grad_norm": 0.1916366538530401,
+      "learning_rate": 8.693752730449979e-06,
+      "loss": 0.3126,
+      "step": 2147
+    },
+    {
+      "epoch": 2.5330188679245285,
+      "grad_norm": 0.18414433676479142,
+      "learning_rate": 8.671909130624727e-06,
+      "loss": 0.3116,
+      "step": 2148
+    },
+    {
+      "epoch": 2.5341981132075473,
+      "grad_norm": 0.18357052323836953,
+      "learning_rate": 8.650065530799478e-06,
+      "loss": 0.301,
+      "step": 2149
+    },
+    {
+      "epoch": 2.535377358490566,
+      "grad_norm": 0.19353160552301463,
+      "learning_rate": 8.628221930974225e-06,
+      "loss": 0.3321,
+      "step": 2150
+    },
+    {
+      "epoch": 2.536556603773585,
+      "grad_norm": 0.18600388735409526,
+      "learning_rate": 8.606378331148973e-06,
+      "loss": 0.2965,
+      "step": 2151
+    },
+    {
+      "epoch": 2.5377358490566038,
+      "grad_norm": 0.2798422712779728,
+      "learning_rate": 8.584534731323722e-06,
+      "loss": 0.3128,
+      "step": 2152
+    },
+    {
+      "epoch": 2.5389150943396226,
+      "grad_norm": 0.19260593805331494,
+      "learning_rate": 8.56269113149847e-06,
+      "loss": 0.3023,
+      "step": 2153
+    },
+    {
+      "epoch": 2.5400943396226414,
+      "grad_norm": 0.19340308342389448,
+      "learning_rate": 8.54084753167322e-06,
+      "loss": 0.3123,
+      "step": 2154
+    },
+    {
+      "epoch": 2.5412735849056602,
+      "grad_norm": 0.19197847131539988,
+      "learning_rate": 8.519003931847968e-06,
+      "loss": 0.3125,
+      "step": 2155
+    },
+    {
+      "epoch": 2.5424528301886795,
+      "grad_norm": 0.17900660901935325,
+      "learning_rate": 8.497160332022717e-06,
+      "loss": 0.2951,
+      "step": 2156
+    },
+    {
+      "epoch": 2.543632075471698,
+      "grad_norm": 0.18699977979581936,
+      "learning_rate": 8.475316732197466e-06,
+      "loss": 0.3094,
+      "step": 2157
+    },
+    {
+      "epoch": 2.544811320754717,
+      "grad_norm": 0.1915352460298986,
+      "learning_rate": 8.453473132372216e-06,
+      "loss": 0.3064,
+      "step": 2158
+    },
+    {
+      "epoch": 2.545990566037736,
+      "grad_norm": 0.19122944584709378,
+      "learning_rate": 8.431629532546965e-06,
+      "loss": 0.3145,
+      "step": 2159
+    },
+    {
+      "epoch": 2.547169811320755,
+      "grad_norm": 0.19902127539372313,
+      "learning_rate": 8.409785932721713e-06,
+      "loss": 0.338,
+      "step": 2160
+    },
+    {
+      "epoch": 2.5483490566037736,
+      "grad_norm": 0.18170107131094554,
+      "learning_rate": 8.387942332896462e-06,
+      "loss": 0.3057,
+      "step": 2161
+    },
+    {
+      "epoch": 2.5495283018867925,
+      "grad_norm": 0.1872699314154282,
+      "learning_rate": 8.36609873307121e-06,
+      "loss": 0.2867,
+      "step": 2162
+    },
+    {
+      "epoch": 2.5507075471698113,
+      "grad_norm": 0.19096029392930228,
+      "learning_rate": 8.34425513324596e-06,
+      "loss": 0.2972,
+      "step": 2163
+    },
+    {
+      "epoch": 2.55188679245283,
+      "grad_norm": 0.17996955091766964,
+      "learning_rate": 8.322411533420708e-06,
+      "loss": 0.2948,
+      "step": 2164
+    },
+    {
+      "epoch": 2.553066037735849,
+      "grad_norm": 0.17766202252625235,
+      "learning_rate": 8.300567933595457e-06,
+      "loss": 0.3073,
+      "step": 2165
+    },
+    {
+      "epoch": 2.5542452830188678,
+      "grad_norm": 0.17701071695175405,
+      "learning_rate": 8.278724333770205e-06,
+      "loss": 0.2976,
+      "step": 2166
+    },
+    {
+      "epoch": 2.555424528301887,
+      "grad_norm": 0.1903538946378906,
+      "learning_rate": 8.256880733944954e-06,
+      "loss": 0.3328,
+      "step": 2167
+    },
+    {
+      "epoch": 2.5566037735849054,
+      "grad_norm": 0.1875279755187838,
+      "learning_rate": 8.235037134119705e-06,
+      "loss": 0.3169,
+      "step": 2168
+    },
+    {
+      "epoch": 2.5577830188679247,
+      "grad_norm": 0.19305675805263067,
+      "learning_rate": 8.213193534294453e-06,
+      "loss": 0.2863,
+      "step": 2169
+    },
+    {
+      "epoch": 2.5589622641509435,
+      "grad_norm": 0.185760226925305,
+      "learning_rate": 8.191349934469202e-06,
+      "loss": 0.296,
+      "step": 2170
+    },
+    {
+      "epoch": 2.5601415094339623,
+      "grad_norm": 0.1837116704289445,
+      "learning_rate": 8.169506334643949e-06,
+      "loss": 0.2953,
+      "step": 2171
+    },
+    {
+      "epoch": 2.561320754716981,
+      "grad_norm": 0.1884202022930756,
+      "learning_rate": 8.147662734818698e-06,
+      "loss": 0.3151,
+      "step": 2172
+    },
+    {
+      "epoch": 2.5625,
+      "grad_norm": 0.18108313707818158,
+      "learning_rate": 8.125819134993446e-06,
+      "loss": 0.2856,
+      "step": 2173
+    },
+    {
+      "epoch": 2.563679245283019,
+      "grad_norm": 0.2011652258477509,
+      "learning_rate": 8.103975535168195e-06,
+      "loss": 0.3153,
+      "step": 2174
+    },
+    {
+      "epoch": 2.5648584905660377,
+      "grad_norm": 0.18250310532315214,
+      "learning_rate": 8.082131935342944e-06,
+      "loss": 0.2948,
+      "step": 2175
+    },
+    {
+      "epoch": 2.5660377358490565,
+      "grad_norm": 0.18806905833600585,
+      "learning_rate": 8.060288335517694e-06,
+      "loss": 0.3231,
+      "step": 2176
+    },
+    {
+      "epoch": 2.5672169811320753,
+      "grad_norm": 0.17839682127600728,
+      "learning_rate": 8.038444735692443e-06,
+      "loss": 0.3041,
+      "step": 2177
+    },
+    {
+      "epoch": 2.5683962264150946,
+      "grad_norm": 0.18176345431894117,
+      "learning_rate": 8.016601135867192e-06,
+      "loss": 0.2996,
+      "step": 2178
+    },
+    {
+      "epoch": 2.569575471698113,
+      "grad_norm": 0.17835736094729368,
+      "learning_rate": 7.99475753604194e-06,
+      "loss": 0.3015,
+      "step": 2179
+    },
+    {
+      "epoch": 2.5707547169811322,
+      "grad_norm": 0.19650371247690035,
+      "learning_rate": 7.972913936216689e-06,
+      "loss": 0.3146,
+      "step": 2180
+    },
+    {
+      "epoch": 2.571933962264151,
+      "grad_norm": 0.1815819917040991,
+      "learning_rate": 7.951070336391438e-06,
+      "loss": 0.3278,
+      "step": 2181
+    },
+    {
+      "epoch": 2.57311320754717,
+      "grad_norm": 0.17955794235333458,
+      "learning_rate": 7.929226736566186e-06,
+      "loss": 0.3075,
+      "step": 2182
+    },
+    {
+      "epoch": 2.5742924528301887,
+      "grad_norm": 0.184995080390407,
+      "learning_rate": 7.907383136740935e-06,
+      "loss": 0.326,
+      "step": 2183
+    },
+    {
+      "epoch": 2.5754716981132075,
+      "grad_norm": 0.18850906586354046,
+      "learning_rate": 7.885539536915684e-06,
+      "loss": 0.3221,
+      "step": 2184
+    },
+    {
+      "epoch": 2.5766509433962264,
+      "grad_norm": 0.1878932010851018,
+      "learning_rate": 7.863695937090432e-06,
+      "loss": 0.309,
+      "step": 2185
+    },
+    {
+      "epoch": 2.577830188679245,
+      "grad_norm": 0.1788322160771785,
+      "learning_rate": 7.841852337265183e-06,
+      "loss": 0.2944,
+      "step": 2186
+    },
+    {
+      "epoch": 2.579009433962264,
+      "grad_norm": 0.17422483735976851,
+      "learning_rate": 7.820008737439931e-06,
+      "loss": 0.3019,
+      "step": 2187
+    },
+    {
+      "epoch": 2.580188679245283,
+      "grad_norm": 0.19407464304447028,
+      "learning_rate": 7.79816513761468e-06,
+      "loss": 0.3173,
+      "step": 2188
+    },
+    {
+      "epoch": 2.581367924528302,
+      "grad_norm": 0.1914606018135022,
+      "learning_rate": 7.776321537789429e-06,
+      "loss": 0.3096,
+      "step": 2189
+    },
+    {
+      "epoch": 2.5825471698113205,
+      "grad_norm": 0.19688392982035668,
+      "learning_rate": 7.754477937964178e-06,
+      "loss": 0.3245,
+      "step": 2190
+    },
+    {
+      "epoch": 2.5837264150943398,
+      "grad_norm": 0.18038805184740703,
+      "learning_rate": 7.732634338138926e-06,
+      "loss": 0.3071,
+      "step": 2191
+    },
+    {
+      "epoch": 2.5849056603773586,
+      "grad_norm": 0.20539482270828197,
+      "learning_rate": 7.710790738313675e-06,
+      "loss": 0.3324,
+      "step": 2192
+    },
+    {
+      "epoch": 2.5860849056603774,
+      "grad_norm": 0.19451456442076623,
+      "learning_rate": 7.688947138488422e-06,
+      "loss": 0.3182,
+      "step": 2193
+    },
+    {
+      "epoch": 2.5872641509433962,
+      "grad_norm": 0.19654141769652497,
+      "learning_rate": 7.667103538663172e-06,
+      "loss": 0.3179,
+      "step": 2194
+    },
+    {
+      "epoch": 2.588443396226415,
+      "grad_norm": 0.18650991902574948,
+      "learning_rate": 7.645259938837921e-06,
+      "loss": 0.3091,
+      "step": 2195
+    },
+    {
+      "epoch": 2.589622641509434,
+      "grad_norm": 0.19061508336061328,
+      "learning_rate": 7.623416339012669e-06,
+      "loss": 0.2979,
+      "step": 2196
+    },
+    {
+      "epoch": 2.5908018867924527,
+      "grad_norm": 0.17353121313585296,
+      "learning_rate": 7.601572739187418e-06,
+      "loss": 0.2953,
+      "step": 2197
+    },
+    {
+      "epoch": 2.5919811320754715,
+      "grad_norm": 0.1771033888220001,
+      "learning_rate": 7.579729139362167e-06,
+      "loss": 0.2915,
+      "step": 2198
+    },
+    {
+      "epoch": 2.5931603773584904,
+      "grad_norm": 0.20490176904521823,
+      "learning_rate": 7.557885539536916e-06,
+      "loss": 0.3155,
+      "step": 2199
+    },
+    {
+      "epoch": 2.5943396226415096,
+      "grad_norm": 0.2036019368986814,
+      "learning_rate": 7.5360419397116645e-06,
+      "loss": 0.3089,
+      "step": 2200
+    },
+    {
+      "epoch": 2.5955188679245285,
+      "grad_norm": 0.20089540655395413,
+      "learning_rate": 7.514198339886413e-06,
+      "loss": 0.3059,
+      "step": 2201
+    },
+    {
+      "epoch": 2.5966981132075473,
+      "grad_norm": 0.18377843869830607,
+      "learning_rate": 7.492354740061163e-06,
+      "loss": 0.3236,
+      "step": 2202
+    },
+    {
+      "epoch": 2.597877358490566,
+      "grad_norm": 0.44104552420260307,
+      "learning_rate": 7.4705111402359114e-06,
+      "loss": 0.3033,
+      "step": 2203
+    },
+    {
+      "epoch": 2.599056603773585,
+      "grad_norm": 0.18209395692161032,
+      "learning_rate": 7.44866754041066e-06,
+      "loss": 0.3072,
+      "step": 2204
+    },
+    {
+      "epoch": 2.6002358490566038,
+      "grad_norm": 0.1977001161793314,
+      "learning_rate": 7.426823940585409e-06,
+      "loss": 0.3124,
+      "step": 2205
+    },
+    {
+      "epoch": 2.6014150943396226,
+      "grad_norm": 0.1890211913826179,
+      "learning_rate": 7.4049803407601575e-06,
+      "loss": 0.2951,
+      "step": 2206
+    },
+    {
+      "epoch": 2.6025943396226414,
+      "grad_norm": 0.18393194486306272,
+      "learning_rate": 7.383136740934907e-06,
+      "loss": 0.3074,
+      "step": 2207
+    },
+    {
+      "epoch": 2.6037735849056602,
+      "grad_norm": 0.18041566601015868,
+      "learning_rate": 7.361293141109656e-06,
+      "loss": 0.3115,
+      "step": 2208
+    },
+    {
+      "epoch": 2.6049528301886795,
+      "grad_norm": 0.18340538231440048,
+      "learning_rate": 7.3394495412844045e-06,
+      "loss": 0.3139,
+      "step": 2209
+    },
+    {
+      "epoch": 2.606132075471698,
+      "grad_norm": 0.1935945176345643,
+      "learning_rate": 7.317605941459153e-06,
+      "loss": 0.3309,
+      "step": 2210
+    },
+    {
+      "epoch": 2.607311320754717,
+      "grad_norm": 0.19821315728817174,
+      "learning_rate": 7.295762341633902e-06,
+      "loss": 0.3165,
+      "step": 2211
+    },
+    {
+      "epoch": 2.608490566037736,
+      "grad_norm": 0.1883553746637234,
+      "learning_rate": 7.273918741808651e-06,
+      "loss": 0.3113,
+      "step": 2212
+    },
+    {
+      "epoch": 2.609669811320755,
+      "grad_norm": 0.17854218840811667,
+      "learning_rate": 7.2520751419834e-06,
+      "loss": 0.2851,
+      "step": 2213
+    },
+    {
+      "epoch": 2.6108490566037736,
+      "grad_norm": 0.1859389827470776,
+      "learning_rate": 7.230231542158147e-06,
+      "loss": 0.314,
+      "step": 2214
+    },
+    {
+      "epoch": 2.6120283018867925,
+      "grad_norm": 0.1898561312100963,
+      "learning_rate": 7.208387942332897e-06,
+      "loss": 0.3267,
+      "step": 2215
+    },
+    {
+      "epoch": 2.6132075471698113,
+      "grad_norm": 0.1901641734246072,
+      "learning_rate": 7.186544342507645e-06,
+      "loss": 0.3269,
+      "step": 2216
+    },
+    {
+      "epoch": 2.61438679245283,
+      "grad_norm": 0.19290825560297406,
+      "learning_rate": 7.164700742682394e-06,
+      "loss": 0.3092,
+      "step": 2217
+    },
+    {
+      "epoch": 2.615566037735849,
+      "grad_norm": 0.20253425582358137,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.3233,
+      "step": 2218
+    },
+    {
+      "epoch": 2.6167452830188678,
+      "grad_norm": 0.1795122909076603,
+      "learning_rate": 7.121013543031891e-06,
+      "loss": 0.3169,
+      "step": 2219
+    },
+    {
+      "epoch": 2.617924528301887,
+      "grad_norm": 0.1797509891334342,
+      "learning_rate": 7.09916994320664e-06,
+      "loss": 0.3077,
+      "step": 2220
+    },
+    {
+      "epoch": 2.6191037735849054,
+      "grad_norm": 0.19628505377966074,
+      "learning_rate": 7.07732634338139e-06,
+      "loss": 0.306,
+      "step": 2221
+    },
+    {
+      "epoch": 2.6202830188679247,
+      "grad_norm": 0.19271141986964963,
+      "learning_rate": 7.055482743556138e-06,
+      "loss": 0.2983,
+      "step": 2222
+    },
+    {
+      "epoch": 2.6214622641509435,
+      "grad_norm": 0.18128032152073398,
+      "learning_rate": 7.033639143730887e-06,
+      "loss": 0.3008,
+      "step": 2223
+    },
+    {
+      "epoch": 2.6226415094339623,
+      "grad_norm": 0.19118762478436666,
+      "learning_rate": 7.011795543905636e-06,
+      "loss": 0.3201,
+      "step": 2224
+    },
+    {
+      "epoch": 2.623820754716981,
+      "grad_norm": 0.19610104053638822,
+      "learning_rate": 6.9899519440803844e-06,
+      "loss": 0.3334,
+      "step": 2225
+    },
+    {
+      "epoch": 2.625,
+      "grad_norm": 0.18866003173365284,
+      "learning_rate": 6.968108344255134e-06,
+      "loss": 0.3182,
+      "step": 2226
+    },
+    {
+      "epoch": 2.626179245283019,
+      "grad_norm": 0.2032170312500724,
+      "learning_rate": 6.946264744429883e-06,
+      "loss": 0.3331,
+      "step": 2227
+    },
+    {
+      "epoch": 2.6273584905660377,
+      "grad_norm": 0.18416173469238484,
+      "learning_rate": 6.924421144604631e-06,
+      "loss": 0.2888,
+      "step": 2228
+    },
+    {
+      "epoch": 2.6285377358490565,
+      "grad_norm": 0.18484778031106144,
+      "learning_rate": 6.90257754477938e-06,
+      "loss": 0.3207,
+      "step": 2229
+    },
+    {
+      "epoch": 2.6297169811320753,
+      "grad_norm": 0.1815055375431982,
+      "learning_rate": 6.880733944954129e-06,
+      "loss": 0.3069,
+      "step": 2230
+    },
+    {
+      "epoch": 2.6308962264150946,
+      "grad_norm": 0.18422182187274638,
+      "learning_rate": 6.858890345128878e-06,
+      "loss": 0.3029,
+      "step": 2231
+    },
+    {
+      "epoch": 2.632075471698113,
+      "grad_norm": 0.18429725956448295,
+      "learning_rate": 6.837046745303627e-06,
+      "loss": 0.3061,
+      "step": 2232
+    },
+    {
+      "epoch": 2.6332547169811322,
+      "grad_norm": 0.19007340400010345,
+      "learning_rate": 6.815203145478376e-06,
+      "loss": 0.3184,
+      "step": 2233
+    },
+    {
+      "epoch": 2.634433962264151,
+      "grad_norm": 0.1870094218659709,
+      "learning_rate": 6.793359545653124e-06,
+      "loss": 0.292,
+      "step": 2234
+    },
+    {
+      "epoch": 2.63561320754717,
+      "grad_norm": 0.18945494885933414,
+      "learning_rate": 6.771515945827872e-06,
+      "loss": 0.3116,
+      "step": 2235
+    },
+    {
+      "epoch": 2.6367924528301887,
+      "grad_norm": 0.18976346828352916,
+      "learning_rate": 6.749672346002621e-06,
+      "loss": 0.3067,
+      "step": 2236
+    },
+    {
+      "epoch": 2.6379716981132075,
+      "grad_norm": 0.19095148102521062,
+      "learning_rate": 6.72782874617737e-06,
+      "loss": 0.3146,
+      "step": 2237
+    },
+    {
+      "epoch": 2.6391509433962264,
+      "grad_norm": 0.1946532475250486,
+      "learning_rate": 6.705985146352118e-06,
+      "loss": 0.3185,
+      "step": 2238
+    },
+    {
+      "epoch": 2.640330188679245,
+      "grad_norm": 0.20658434809496945,
+      "learning_rate": 6.684141546526868e-06,
+      "loss": 0.3178,
+      "step": 2239
+    },
+    {
+      "epoch": 2.641509433962264,
+      "grad_norm": 0.1907317403019871,
+      "learning_rate": 6.6622979467016166e-06,
+      "loss": 0.3066,
+      "step": 2240
+    },
+    {
+      "epoch": 2.642688679245283,
+      "grad_norm": 0.18503929021485738,
+      "learning_rate": 6.640454346876365e-06,
+      "loss": 0.2997,
+      "step": 2241
+    },
+    {
+      "epoch": 2.643867924528302,
+      "grad_norm": 0.18682409380695533,
+      "learning_rate": 6.618610747051114e-06,
+      "loss": 0.3015,
+      "step": 2242
+    },
+    {
+      "epoch": 2.6450471698113205,
+      "grad_norm": 0.18039932487734295,
+      "learning_rate": 6.596767147225863e-06,
+      "loss": 0.2975,
+      "step": 2243
+    },
+    {
+      "epoch": 2.6462264150943398,
+      "grad_norm": 0.19748760971859744,
+      "learning_rate": 6.574923547400612e-06,
+      "loss": 0.3088,
+      "step": 2244
+    },
+    {
+      "epoch": 2.6474056603773586,
+      "grad_norm": 0.18204134332768546,
+      "learning_rate": 6.553079947575361e-06,
+      "loss": 0.3002,
+      "step": 2245
+    },
+    {
+      "epoch": 2.6485849056603774,
+      "grad_norm": 0.18920056552205333,
+      "learning_rate": 6.5312363477501096e-06,
+      "loss": 0.3232,
+      "step": 2246
+    },
+    {
+      "epoch": 2.6497641509433962,
+      "grad_norm": 0.18146355461902952,
+      "learning_rate": 6.509392747924858e-06,
+      "loss": 0.3107,
+      "step": 2247
+    },
+    {
+      "epoch": 2.650943396226415,
+      "grad_norm": 0.18491598728748118,
+      "learning_rate": 6.487549148099607e-06,
+      "loss": 0.3027,
+      "step": 2248
+    },
+    {
+      "epoch": 2.652122641509434,
+      "grad_norm": 0.19049485106219685,
+      "learning_rate": 6.4657055482743565e-06,
+      "loss": 0.318,
+      "step": 2249
+    },
+    {
+      "epoch": 2.6533018867924527,
+      "grad_norm": 0.19271548102405361,
+      "learning_rate": 6.443861948449105e-06,
+      "loss": 0.3053,
+      "step": 2250
+    },
+    {
+      "epoch": 2.6544811320754715,
+      "grad_norm": 0.18306777875413202,
+      "learning_rate": 6.422018348623854e-06,
+      "loss": 0.3106,
+      "step": 2251
+    },
+    {
+      "epoch": 2.6556603773584904,
+      "grad_norm": 3.4303652481196854,
+      "learning_rate": 6.400174748798603e-06,
+      "loss": 0.3549,
+      "step": 2252
+    },
+    {
+      "epoch": 2.6568396226415096,
+      "grad_norm": 0.18447766377692654,
+      "learning_rate": 6.378331148973351e-06,
+      "loss": 0.2949,
+      "step": 2253
+    },
+    {
+      "epoch": 2.6580188679245285,
+      "grad_norm": 0.18478864467760703,
+      "learning_rate": 6.356487549148101e-06,
+      "loss": 0.2977,
+      "step": 2254
+    },
+    {
+      "epoch": 2.6591981132075473,
+      "grad_norm": 0.18748567983983838,
+      "learning_rate": 6.3346439493228495e-06,
+      "loss": 0.3058,
+      "step": 2255
+    },
+    {
+      "epoch": 2.660377358490566,
+      "grad_norm": 0.18869103916534413,
+      "learning_rate": 6.312800349497598e-06,
+      "loss": 0.2989,
+      "step": 2256
+    },
+    {
+      "epoch": 2.661556603773585,
+      "grad_norm": 0.19697395965669462,
+      "learning_rate": 6.290956749672346e-06,
+      "loss": 0.3352,
+      "step": 2257
+    },
+    {
+      "epoch": 2.6627358490566038,
+      "grad_norm": 0.1897753175546308,
+      "learning_rate": 6.269113149847095e-06,
+      "loss": 0.3071,
+      "step": 2258
+    },
+    {
+      "epoch": 2.6639150943396226,
+      "grad_norm": 0.18759729173643586,
+      "learning_rate": 6.247269550021844e-06,
+      "loss": 0.3063,
+      "step": 2259
+    },
+    {
+      "epoch": 2.6650943396226414,
+      "grad_norm": 0.19448381269815043,
+      "learning_rate": 6.225425950196593e-06,
+      "loss": 0.3184,
+      "step": 2260
+    },
+    {
+      "epoch": 2.6662735849056602,
+      "grad_norm": 0.17564883659524053,
+      "learning_rate": 6.203582350371342e-06,
+      "loss": 0.2896,
+      "step": 2261
+    },
+    {
+      "epoch": 2.6674528301886795,
+      "grad_norm": 0.18187140072066255,
+      "learning_rate": 6.18173875054609e-06,
+      "loss": 0.3101,
+      "step": 2262
+    },
+    {
+      "epoch": 2.668632075471698,
+      "grad_norm": 0.17853105617373732,
+      "learning_rate": 6.159895150720839e-06,
+      "loss": 0.3121,
+      "step": 2263
+    },
+    {
+      "epoch": 2.669811320754717,
+      "grad_norm": 0.17587056578455568,
+      "learning_rate": 6.138051550895588e-06,
+      "loss": 0.3061,
+      "step": 2264
+    },
+    {
+      "epoch": 2.670990566037736,
+      "grad_norm": 0.17990012409022968,
+      "learning_rate": 6.1162079510703365e-06,
+      "loss": 0.2989,
+      "step": 2265
+    },
+    {
+      "epoch": 2.672169811320755,
+      "grad_norm": 0.1840513198410727,
+      "learning_rate": 6.094364351245085e-06,
+      "loss": 0.3135,
+      "step": 2266
+    },
+    {
+      "epoch": 2.6733490566037736,
+      "grad_norm": 0.18146582021412921,
+      "learning_rate": 6.072520751419834e-06,
+      "loss": 0.3106,
+      "step": 2267
+    },
+    {
+      "epoch": 2.6745283018867925,
+      "grad_norm": 0.19208774813580046,
+      "learning_rate": 6.050677151594583e-06,
+      "loss": 0.3068,
+      "step": 2268
+    },
+    {
+      "epoch": 2.6757075471698113,
+      "grad_norm": 0.17655916045384493,
+      "learning_rate": 6.028833551769332e-06,
+      "loss": 0.2942,
+      "step": 2269
+    },
+    {
+      "epoch": 2.67688679245283,
+      "grad_norm": 0.19530613741812614,
+      "learning_rate": 6.006989951944081e-06,
+      "loss": 0.3028,
+      "step": 2270
+    },
+    {
+      "epoch": 2.678066037735849,
+      "grad_norm": 0.18831853999716958,
+      "learning_rate": 5.9851463521188295e-06,
+      "loss": 0.3405,
+      "step": 2271
+    },
+    {
+      "epoch": 2.6792452830188678,
+      "grad_norm": 0.18880023342091487,
+      "learning_rate": 5.963302752293578e-06,
+      "loss": 0.3043,
+      "step": 2272
+    },
+    {
+      "epoch": 2.680424528301887,
+      "grad_norm": 0.18923223948372422,
+      "learning_rate": 5.941459152468327e-06,
+      "loss": 0.3153,
+      "step": 2273
+    },
+    {
+      "epoch": 2.6816037735849054,
+      "grad_norm": 0.17795434099075522,
+      "learning_rate": 5.919615552643076e-06,
+      "loss": 0.2995,
+      "step": 2274
+    },
+    {
+      "epoch": 2.6827830188679247,
+      "grad_norm": 0.190890478000946,
+      "learning_rate": 5.897771952817824e-06,
+      "loss": 0.3144,
+      "step": 2275
+    },
+    {
+      "epoch": 2.6839622641509435,
+      "grad_norm": 0.18357954202115695,
+      "learning_rate": 5.875928352992573e-06,
+      "loss": 0.3252,
+      "step": 2276
+    },
+    {
+      "epoch": 2.6851415094339623,
+      "grad_norm": 0.18297568260497815,
+      "learning_rate": 5.8540847531673225e-06,
+      "loss": 0.3133,
+      "step": 2277
+    },
+    {
+      "epoch": 2.686320754716981,
+      "grad_norm": 0.17821542122136924,
+      "learning_rate": 5.832241153342071e-06,
+      "loss": 0.3077,
+      "step": 2278
+    },
+    {
+      "epoch": 2.6875,
+      "grad_norm": 0.17825584808446093,
+      "learning_rate": 5.81039755351682e-06,
+      "loss": 0.2934,
+      "step": 2279
+    },
+    {
+      "epoch": 2.688679245283019,
+      "grad_norm": 0.17645725385858294,
+      "learning_rate": 5.788553953691569e-06,
+      "loss": 0.3079,
+      "step": 2280
+    },
+    {
+      "epoch": 2.6898584905660377,
+      "grad_norm": 0.1853482558891638,
+      "learning_rate": 5.766710353866317e-06,
+      "loss": 0.3108,
+      "step": 2281
+    },
+    {
+      "epoch": 2.6910377358490565,
+      "grad_norm": 0.1876417213301573,
+      "learning_rate": 5.744866754041067e-06,
+      "loss": 0.3069,
+      "step": 2282
+    },
+    {
+      "epoch": 2.6922169811320753,
+      "grad_norm": 0.17962666866239108,
+      "learning_rate": 5.723023154215815e-06,
+      "loss": 0.2969,
+      "step": 2283
+    },
+    {
+      "epoch": 2.6933962264150946,
+      "grad_norm": 0.17795128589569875,
+      "learning_rate": 5.701179554390563e-06,
+      "loss": 0.3037,
+      "step": 2284
+    },
+    {
+      "epoch": 2.694575471698113,
+      "grad_norm": 0.19856112809052115,
+      "learning_rate": 5.679335954565312e-06,
+      "loss": 0.3172,
+      "step": 2285
+    },
+    {
+      "epoch": 2.6957547169811322,
+      "grad_norm": 0.18817458383613442,
+      "learning_rate": 5.657492354740062e-06,
+      "loss": 0.3247,
+      "step": 2286
+    },
+    {
+      "epoch": 2.696933962264151,
+      "grad_norm": 0.18187237803069464,
+      "learning_rate": 5.63564875491481e-06,
+      "loss": 0.2992,
+      "step": 2287
+    },
+    {
+      "epoch": 2.69811320754717,
+      "grad_norm": 0.19644897529860433,
+      "learning_rate": 5.613805155089559e-06,
+      "loss": 0.3085,
+      "step": 2288
+    },
+    {
+      "epoch": 2.6992924528301887,
+      "grad_norm": 0.1791929897343325,
+      "learning_rate": 5.591961555264308e-06,
+      "loss": 0.3027,
+      "step": 2289
+    },
+    {
+      "epoch": 2.7004716981132075,
+      "grad_norm": 0.1874586497859773,
+      "learning_rate": 5.570117955439056e-06,
+      "loss": 0.317,
+      "step": 2290
+    },
+    {
+      "epoch": 2.7016509433962264,
+      "grad_norm": 0.1879526984607769,
+      "learning_rate": 5.548274355613806e-06,
+      "loss": 0.2876,
+      "step": 2291
+    },
+    {
+      "epoch": 2.702830188679245,
+      "grad_norm": 0.1861975635620161,
+      "learning_rate": 5.526430755788555e-06,
+      "loss": 0.3131,
+      "step": 2292
+    },
+    {
+      "epoch": 2.704009433962264,
+      "grad_norm": 0.18896595267128305,
+      "learning_rate": 5.504587155963303e-06,
+      "loss": 0.3092,
+      "step": 2293
+    },
+    {
+      "epoch": 2.705188679245283,
+      "grad_norm": 0.17708615486302332,
+      "learning_rate": 5.482743556138051e-06,
+      "loss": 0.291,
+      "step": 2294
+    },
+    {
+      "epoch": 2.706367924528302,
+      "grad_norm": 0.1878689440044866,
+      "learning_rate": 5.460899956312801e-06,
+      "loss": 0.312,
+      "step": 2295
+    },
+    {
+      "epoch": 2.7075471698113205,
+      "grad_norm": 0.17998271248348088,
+      "learning_rate": 5.4390563564875494e-06,
+      "loss": 0.307,
+      "step": 2296
+    },
+    {
+      "epoch": 2.7087264150943398,
+      "grad_norm": 0.17875458746427525,
+      "learning_rate": 5.417212756662298e-06,
+      "loss": 0.3139,
+      "step": 2297
+    },
+    {
+      "epoch": 2.7099056603773586,
+      "grad_norm": 0.1783902805827607,
+      "learning_rate": 5.395369156837047e-06,
+      "loss": 0.3082,
+      "step": 2298
+    },
+    {
+      "epoch": 2.7110849056603774,
+      "grad_norm": 0.18458911092136399,
+      "learning_rate": 5.3735255570117955e-06,
+      "loss": 0.3175,
+      "step": 2299
+    },
+    {
+      "epoch": 2.7122641509433962,
+      "grad_norm": 0.18679440300612335,
+      "learning_rate": 5.351681957186545e-06,
+      "loss": 0.3098,
+      "step": 2300
+    },
+    {
+      "epoch": 2.713443396226415,
+      "grad_norm": 0.18121921843049066,
+      "learning_rate": 5.329838357361294e-06,
+      "loss": 0.2945,
+      "step": 2301
+    },
+    {
+      "epoch": 2.714622641509434,
+      "grad_norm": 0.1769154809210729,
+      "learning_rate": 5.3079947575360424e-06,
+      "loss": 0.3009,
+      "step": 2302
+    },
+    {
+      "epoch": 2.7158018867924527,
+      "grad_norm": 0.18079183927996015,
+      "learning_rate": 5.286151157710791e-06,
+      "loss": 0.308,
+      "step": 2303
+    },
+    {
+      "epoch": 2.7169811320754715,
+      "grad_norm": 0.19583814348504033,
+      "learning_rate": 5.26430755788554e-06,
+      "loss": 0.3037,
+      "step": 2304
+    },
+    {
+      "epoch": 2.7181603773584904,
+      "grad_norm": 0.1941875248159903,
+      "learning_rate": 5.2424639580602885e-06,
+      "loss": 0.3334,
+      "step": 2305
+    },
+    {
+      "epoch": 2.7193396226415096,
+      "grad_norm": 0.1837624264310947,
+      "learning_rate": 5.220620358235037e-06,
+      "loss": 0.3024,
+      "step": 2306
+    },
+    {
+      "epoch": 2.7205188679245285,
+      "grad_norm": 0.17640616584208088,
+      "learning_rate": 5.198776758409786e-06,
+      "loss": 0.2983,
+      "step": 2307
+    },
+    {
+      "epoch": 2.7216981132075473,
+      "grad_norm": 0.18779051439569733,
+      "learning_rate": 5.176933158584535e-06,
+      "loss": 0.3171,
+      "step": 2308
+    },
+    {
+      "epoch": 2.722877358490566,
+      "grad_norm": 0.18552242890347997,
+      "learning_rate": 5.155089558759283e-06,
+      "loss": 0.2983,
+      "step": 2309
+    },
+    {
+      "epoch": 2.724056603773585,
+      "grad_norm": 0.1780929496822298,
+      "learning_rate": 5.133245958934033e-06,
+      "loss": 0.3029,
+      "step": 2310
+    },
+    {
+      "epoch": 2.7252358490566038,
+      "grad_norm": 0.18692339692491067,
+      "learning_rate": 5.1114023591087816e-06,
+      "loss": 0.3107,
+      "step": 2311
+    },
+    {
+      "epoch": 2.7264150943396226,
+      "grad_norm": 0.19294990824193195,
+      "learning_rate": 5.08955875928353e-06,
+      "loss": 0.3125,
+      "step": 2312
+    },
+    {
+      "epoch": 2.7275943396226414,
+      "grad_norm": 0.18735826166178288,
+      "learning_rate": 5.067715159458279e-06,
+      "loss": 0.3078,
+      "step": 2313
+    },
+    {
+      "epoch": 2.7287735849056602,
+      "grad_norm": 0.18579835478419543,
+      "learning_rate": 5.045871559633028e-06,
+      "loss": 0.3092,
+      "step": 2314
+    },
+    {
+      "epoch": 2.7299528301886795,
+      "grad_norm": 0.19740136050521268,
+      "learning_rate": 5.024027959807777e-06,
+      "loss": 0.3052,
+      "step": 2315
+    },
+    {
+      "epoch": 2.731132075471698,
+      "grad_norm": 0.18504094656890205,
+      "learning_rate": 5.002184359982525e-06,
+      "loss": 0.3133,
+      "step": 2316
+    },
+    {
+      "epoch": 2.732311320754717,
+      "grad_norm": 0.17694236856669165,
+      "learning_rate": 4.980340760157274e-06,
+      "loss": 0.3017,
+      "step": 2317
+    },
+    {
+      "epoch": 2.733490566037736,
+      "grad_norm": 0.18303721701299,
+      "learning_rate": 4.958497160332022e-06,
+      "loss": 0.292,
+      "step": 2318
+    },
+    {
+      "epoch": 2.734669811320755,
+      "grad_norm": 0.17947261480054413,
+      "learning_rate": 4.936653560506772e-06,
+      "loss": 0.3057,
+      "step": 2319
+    },
+    {
+      "epoch": 2.7358490566037736,
+      "grad_norm": 0.18739566767548785,
+      "learning_rate": 4.914809960681521e-06,
+      "loss": 0.3117,
+      "step": 2320
+    },
+    {
+      "epoch": 2.7370283018867925,
+      "grad_norm": 0.18357011248329652,
+      "learning_rate": 4.892966360856269e-06,
+      "loss": 0.3274,
+      "step": 2321
+    },
+    {
+      "epoch": 2.7382075471698113,
+      "grad_norm": 0.17770517010512088,
+      "learning_rate": 4.871122761031018e-06,
+      "loss": 0.2952,
+      "step": 2322
+    },
+    {
+      "epoch": 2.73938679245283,
+      "grad_norm": 0.18698405053361986,
+      "learning_rate": 4.849279161205767e-06,
+      "loss": 0.3184,
+      "step": 2323
+    },
+    {
+      "epoch": 2.740566037735849,
+      "grad_norm": 0.5543602148586059,
+      "learning_rate": 4.827435561380516e-06,
+      "loss": 0.3167,
+      "step": 2324
+    },
+    {
+      "epoch": 2.7417452830188678,
+      "grad_norm": 0.1850744335095528,
+      "learning_rate": 4.805591961555265e-06,
+      "loss": 0.2963,
+      "step": 2325
+    },
+    {
+      "epoch": 2.742924528301887,
+      "grad_norm": 0.17811975978607755,
+      "learning_rate": 4.783748361730013e-06,
+      "loss": 0.3034,
+      "step": 2326
+    },
+    {
+      "epoch": 2.7441037735849054,
+      "grad_norm": 0.17235221893865804,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 0.3034,
+      "step": 2327
+    },
+    {
+      "epoch": 2.7452830188679247,
+      "grad_norm": 0.17824591253750358,
+      "learning_rate": 4.740061162079511e-06,
+      "loss": 0.3105,
+      "step": 2328
+    },
+    {
+      "epoch": 2.7464622641509435,
+      "grad_norm": 0.17463193396439447,
+      "learning_rate": 4.71821756225426e-06,
+      "loss": 0.2957,
+      "step": 2329
+    },
+    {
+      "epoch": 2.7476415094339623,
+      "grad_norm": 0.19189827055072967,
+      "learning_rate": 4.6963739624290085e-06,
+      "loss": 0.3176,
+      "step": 2330
+    },
+    {
+      "epoch": 2.748820754716981,
+      "grad_norm": 0.1746808794124796,
+      "learning_rate": 4.674530362603757e-06,
+      "loss": 0.2978,
+      "step": 2331
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.18534972099059002,
+      "learning_rate": 4.652686762778506e-06,
+      "loss": 0.3236,
+      "step": 2332
+    },
+    {
+      "epoch": 2.751179245283019,
+      "grad_norm": 0.1859416196045105,
+      "learning_rate": 4.630843162953255e-06,
+      "loss": 0.3181,
+      "step": 2333
+    },
+    {
+      "epoch": 2.7523584905660377,
+      "grad_norm": 0.17977032751040462,
+      "learning_rate": 4.608999563128004e-06,
+      "loss": 0.3035,
+      "step": 2334
+    },
+    {
+      "epoch": 2.7535377358490565,
+      "grad_norm": 0.1810595985558168,
+      "learning_rate": 4.587155963302753e-06,
+      "loss": 0.3122,
+      "step": 2335
+    },
+    {
+      "epoch": 2.7547169811320753,
+      "grad_norm": 0.1847373123312656,
+      "learning_rate": 4.5653123634775015e-06,
+      "loss": 0.3185,
+      "step": 2336
+    },
+    {
+      "epoch": 2.7558962264150946,
+      "grad_norm": 0.1869979448513391,
+      "learning_rate": 4.54346876365225e-06,
+      "loss": 0.3024,
+      "step": 2337
+    },
+    {
+      "epoch": 2.757075471698113,
+      "grad_norm": 0.1791477042865141,
+      "learning_rate": 4.521625163826999e-06,
+      "loss": 0.3078,
+      "step": 2338
+    },
+    {
+      "epoch": 2.7582547169811322,
+      "grad_norm": 0.17700317859834772,
+      "learning_rate": 4.4997815640017476e-06,
+      "loss": 0.298,
+      "step": 2339
+    },
+    {
+      "epoch": 2.759433962264151,
+      "grad_norm": 0.17446225646018568,
+      "learning_rate": 4.477937964176496e-06,
+      "loss": 0.3014,
+      "step": 2340
+    },
+    {
+      "epoch": 2.76061320754717,
+      "grad_norm": 0.1980809579883177,
+      "learning_rate": 4.456094364351245e-06,
+      "loss": 0.3302,
+      "step": 2341
+    },
+    {
+      "epoch": 2.7617924528301887,
+      "grad_norm": 0.1786095850723116,
+      "learning_rate": 4.4342507645259945e-06,
+      "loss": 0.3094,
+      "step": 2342
+    },
+    {
+      "epoch": 2.7629716981132075,
+      "grad_norm": 0.1838807602962587,
+      "learning_rate": 4.412407164700743e-06,
+      "loss": 0.3119,
+      "step": 2343
+    },
+    {
+      "epoch": 2.7641509433962264,
+      "grad_norm": 0.18474508374229698,
+      "learning_rate": 4.390563564875492e-06,
+      "loss": 0.3008,
+      "step": 2344
+    },
+    {
+      "epoch": 2.765330188679245,
+      "grad_norm": 0.17804994045224432,
+      "learning_rate": 4.368719965050241e-06,
+      "loss": 0.3086,
+      "step": 2345
+    },
+    {
+      "epoch": 2.766509433962264,
+      "grad_norm": 0.18751652464390905,
+      "learning_rate": 4.346876365224989e-06,
+      "loss": 0.3024,
+      "step": 2346
+    },
+    {
+      "epoch": 2.767688679245283,
+      "grad_norm": 0.1860150913633458,
+      "learning_rate": 4.325032765399739e-06,
+      "loss": 0.308,
+      "step": 2347
+    },
+    {
+      "epoch": 2.768867924528302,
+      "grad_norm": 0.18598992123885855,
+      "learning_rate": 4.303189165574487e-06,
+      "loss": 0.326,
+      "step": 2348
+    },
+    {
+      "epoch": 2.7700471698113205,
+      "grad_norm": 0.1823449758743092,
+      "learning_rate": 4.281345565749235e-06,
+      "loss": 0.2971,
+      "step": 2349
+    },
+    {
+      "epoch": 2.7712264150943398,
+      "grad_norm": 0.18666089549727938,
+      "learning_rate": 4.259501965923984e-06,
+      "loss": 0.2898,
+      "step": 2350
+    },
+    {
+      "epoch": 2.7724056603773586,
+      "grad_norm": 0.18508743698903984,
+      "learning_rate": 4.237658366098733e-06,
+      "loss": 0.3056,
+      "step": 2351
+    },
+    {
+      "epoch": 2.7735849056603774,
+      "grad_norm": 0.17757388711589883,
+      "learning_rate": 4.215814766273482e-06,
+      "loss": 0.2973,
+      "step": 2352
+    },
+    {
+      "epoch": 2.7747641509433962,
+      "grad_norm": 0.1845496110494431,
+      "learning_rate": 4.193971166448231e-06,
+      "loss": 0.3137,
+      "step": 2353
+    },
+    {
+      "epoch": 2.775943396226415,
+      "grad_norm": 0.18467605797588024,
+      "learning_rate": 4.17212756662298e-06,
+      "loss": 0.3038,
+      "step": 2354
+    },
+    {
+      "epoch": 2.777122641509434,
+      "grad_norm": 0.1832864359953904,
+      "learning_rate": 4.150283966797728e-06,
+      "loss": 0.2946,
+      "step": 2355
+    },
+    {
+      "epoch": 2.7783018867924527,
+      "grad_norm": 0.1818349775837569,
+      "learning_rate": 4.128440366972477e-06,
+      "loss": 0.3126,
+      "step": 2356
+    },
+    {
+      "epoch": 2.7794811320754715,
+      "grad_norm": 0.18335383818392686,
+      "learning_rate": 4.106596767147227e-06,
+      "loss": 0.3053,
+      "step": 2357
+    },
+    {
+      "epoch": 2.7806603773584904,
+      "grad_norm": 0.18232912925263148,
+      "learning_rate": 4.0847531673219745e-06,
+      "loss": 0.2912,
+      "step": 2358
+    },
+    {
+      "epoch": 2.7818396226415096,
+      "grad_norm": 0.1811775411268928,
+      "learning_rate": 4.062909567496723e-06,
+      "loss": 0.3159,
+      "step": 2359
+    },
+    {
+      "epoch": 2.7830188679245285,
+      "grad_norm": 0.17983317314226246,
+      "learning_rate": 4.041065967671472e-06,
+      "loss": 0.2935,
+      "step": 2360
+    },
+    {
+      "epoch": 2.7841981132075473,
+      "grad_norm": 0.19415843582403586,
+      "learning_rate": 4.019222367846221e-06,
+      "loss": 0.2999,
+      "step": 2361
+    },
+    {
+      "epoch": 2.785377358490566,
+      "grad_norm": 0.1875008696726478,
+      "learning_rate": 3.99737876802097e-06,
+      "loss": 0.3118,
+      "step": 2362
+    },
+    {
+      "epoch": 2.786556603773585,
+      "grad_norm": 0.18038987730619058,
+      "learning_rate": 3.975535168195719e-06,
+      "loss": 0.3096,
+      "step": 2363
+    },
+    {
+      "epoch": 2.7877358490566038,
+      "grad_norm": 0.17684425118577185,
+      "learning_rate": 3.9536915683704675e-06,
+      "loss": 0.2912,
+      "step": 2364
+    },
+    {
+      "epoch": 2.7889150943396226,
+      "grad_norm": 0.18340921278159458,
+      "learning_rate": 3.931847968545216e-06,
+      "loss": 0.3011,
+      "step": 2365
+    },
+    {
+      "epoch": 2.7900943396226414,
+      "grad_norm": 0.18445152112117244,
+      "learning_rate": 3.910004368719966e-06,
+      "loss": 0.313,
+      "step": 2366
+    },
+    {
+      "epoch": 2.7912735849056602,
+      "grad_norm": 0.18354445610409342,
+      "learning_rate": 3.8881607688947144e-06,
+      "loss": 0.3225,
+      "step": 2367
+    },
+    {
+      "epoch": 2.7924528301886795,
+      "grad_norm": 0.17750337634122243,
+      "learning_rate": 3.866317169069463e-06,
+      "loss": 0.3073,
+      "step": 2368
+    },
+    {
+      "epoch": 2.793632075471698,
+      "grad_norm": 0.17902192655932994,
+      "learning_rate": 3.844473569244211e-06,
+      "loss": 0.2967,
+      "step": 2369
+    },
+    {
+      "epoch": 2.794811320754717,
+      "grad_norm": 0.17996145364288807,
+      "learning_rate": 3.8226299694189605e-06,
+      "loss": 0.3167,
+      "step": 2370
+    },
+    {
+      "epoch": 2.795990566037736,
+      "grad_norm": 0.18253376747780675,
+      "learning_rate": 3.800786369593709e-06,
+      "loss": 0.2938,
+      "step": 2371
+    },
+    {
+      "epoch": 2.797169811320755,
+      "grad_norm": 0.1842438379069429,
+      "learning_rate": 3.778942769768458e-06,
+      "loss": 0.3008,
+      "step": 2372
+    },
+    {
+      "epoch": 2.7983490566037736,
+      "grad_norm": 0.17702339555730523,
+      "learning_rate": 3.7570991699432066e-06,
+      "loss": 0.3056,
+      "step": 2373
+    },
+    {
+      "epoch": 2.7995283018867925,
+      "grad_norm": 0.18275223423761253,
+      "learning_rate": 3.7352555701179557e-06,
+      "loss": 0.3039,
+      "step": 2374
+    },
+    {
+      "epoch": 2.8007075471698113,
+      "grad_norm": 0.1866411586104698,
+      "learning_rate": 3.7134119702927044e-06,
+      "loss": 0.3065,
+      "step": 2375
+    },
+    {
+      "epoch": 2.80188679245283,
+      "grad_norm": 0.18274273532054738,
+      "learning_rate": 3.6915683704674535e-06,
+      "loss": 0.2974,
+      "step": 2376
+    },
+    {
+      "epoch": 2.803066037735849,
+      "grad_norm": 0.18047396536996288,
+      "learning_rate": 3.6697247706422022e-06,
+      "loss": 0.3097,
+      "step": 2377
+    },
+    {
+      "epoch": 2.8042452830188678,
+      "grad_norm": 0.17834751667188417,
+      "learning_rate": 3.647881170816951e-06,
+      "loss": 0.2868,
+      "step": 2378
+    },
+    {
+      "epoch": 2.805424528301887,
+      "grad_norm": 0.18163738658284528,
+      "learning_rate": 3.6260375709917e-06,
+      "loss": 0.3077,
+      "step": 2379
+    },
+    {
+      "epoch": 2.8066037735849054,
+      "grad_norm": 0.1820902189797999,
+      "learning_rate": 3.6041939711664483e-06,
+      "loss": 0.304,
+      "step": 2380
+    },
+    {
+      "epoch": 2.8077830188679247,
+      "grad_norm": 0.17387226653859322,
+      "learning_rate": 3.582350371341197e-06,
+      "loss": 0.301,
+      "step": 2381
+    },
+    {
+      "epoch": 2.8089622641509435,
+      "grad_norm": 0.17939884074782794,
+      "learning_rate": 3.5605067715159457e-06,
+      "loss": 0.3118,
+      "step": 2382
+    },
+    {
+      "epoch": 2.8101415094339623,
+      "grad_norm": 0.18816649779017636,
+      "learning_rate": 3.538663171690695e-06,
+      "loss": 0.3136,
+      "step": 2383
+    },
+    {
+      "epoch": 2.811320754716981,
+      "grad_norm": 0.17700909486865823,
+      "learning_rate": 3.5168195718654435e-06,
+      "loss": 0.2914,
+      "step": 2384
+    },
+    {
+      "epoch": 2.8125,
+      "grad_norm": 0.17286190374914093,
+      "learning_rate": 3.4949759720401922e-06,
+      "loss": 0.2986,
+      "step": 2385
+    },
+    {
+      "epoch": 2.813679245283019,
+      "grad_norm": 0.17894395957738396,
+      "learning_rate": 3.4731323722149413e-06,
+      "loss": 0.2979,
+      "step": 2386
+    },
+    {
+      "epoch": 2.8148584905660377,
+      "grad_norm": 0.18089812360670215,
+      "learning_rate": 3.45128877238969e-06,
+      "loss": 0.3103,
+      "step": 2387
+    },
+    {
+      "epoch": 2.8160377358490565,
+      "grad_norm": 0.18200247085407942,
+      "learning_rate": 3.429445172564439e-06,
+      "loss": 0.3205,
+      "step": 2388
+    },
+    {
+      "epoch": 2.8172169811320753,
+      "grad_norm": 0.2242462777013315,
+      "learning_rate": 3.407601572739188e-06,
+      "loss": 0.3384,
+      "step": 2389
+    },
+    {
+      "epoch": 2.8183962264150946,
+      "grad_norm": 0.18526162238680527,
+      "learning_rate": 3.385757972913936e-06,
+      "loss": 0.3053,
+      "step": 2390
+    },
+    {
+      "epoch": 2.819575471698113,
+      "grad_norm": 0.1779419860517371,
+      "learning_rate": 3.363914373088685e-06,
+      "loss": 0.3076,
+      "step": 2391
+    },
+    {
+      "epoch": 2.8207547169811322,
+      "grad_norm": 0.19146208766991515,
+      "learning_rate": 3.342070773263434e-06,
+      "loss": 0.3229,
+      "step": 2392
+    },
+    {
+      "epoch": 2.821933962264151,
+      "grad_norm": 0.18004096583008616,
+      "learning_rate": 3.3202271734381826e-06,
+      "loss": 0.3057,
+      "step": 2393
+    },
+    {
+      "epoch": 2.82311320754717,
+      "grad_norm": 0.18077831830317162,
+      "learning_rate": 3.2983835736129313e-06,
+      "loss": 0.3119,
+      "step": 2394
+    },
+    {
+      "epoch": 2.8242924528301887,
+      "grad_norm": 0.17622187718334073,
+      "learning_rate": 3.2765399737876804e-06,
+      "loss": 0.3073,
+      "step": 2395
+    },
+    {
+      "epoch": 2.8254716981132075,
+      "grad_norm": 0.17912700210565446,
+      "learning_rate": 3.254696373962429e-06,
+      "loss": 0.3292,
+      "step": 2396
+    },
+    {
+      "epoch": 2.8266509433962264,
+      "grad_norm": 0.17911645039486324,
+      "learning_rate": 3.2328527741371783e-06,
+      "loss": 0.3031,
+      "step": 2397
+    },
+    {
+      "epoch": 2.827830188679245,
+      "grad_norm": 0.2043523654054939,
+      "learning_rate": 3.211009174311927e-06,
+      "loss": 0.3161,
+      "step": 2398
+    },
+    {
+      "epoch": 2.829009433962264,
+      "grad_norm": 0.18077871349060776,
+      "learning_rate": 3.1891655744866756e-06,
+      "loss": 0.3089,
+      "step": 2399
+    },
+    {
+      "epoch": 2.830188679245283,
+      "grad_norm": 0.17602761373636347,
+      "learning_rate": 3.1673219746614248e-06,
+      "loss": 0.2958,
+      "step": 2400
+    },
+    {
+      "epoch": 2.831367924528302,
+      "grad_norm": 0.18682927093125923,
+      "learning_rate": 3.145478374836173e-06,
+      "loss": 0.3233,
+      "step": 2401
+    },
+    {
+      "epoch": 2.8325471698113205,
+      "grad_norm": 0.17359483369502934,
+      "learning_rate": 3.123634775010922e-06,
+      "loss": 0.298,
+      "step": 2402
+    },
+    {
+      "epoch": 2.8337264150943398,
+      "grad_norm": 0.17610884980380961,
+      "learning_rate": 3.101791175185671e-06,
+      "loss": 0.2967,
+      "step": 2403
+    },
+    {
+      "epoch": 2.8349056603773586,
+      "grad_norm": 0.18113062434146604,
+      "learning_rate": 3.0799475753604195e-06,
+      "loss": 0.3158,
+      "step": 2404
+    },
+    {
+      "epoch": 2.8360849056603774,
+      "grad_norm": 0.18325617894044766,
+      "learning_rate": 3.0581039755351682e-06,
+      "loss": 0.3033,
+      "step": 2405
+    },
+    {
+      "epoch": 2.8372641509433962,
+      "grad_norm": 0.18714520595536147,
+      "learning_rate": 3.036260375709917e-06,
+      "loss": 0.2978,
+      "step": 2406
+    },
+    {
+      "epoch": 2.838443396226415,
+      "grad_norm": 0.18794165670163032,
+      "learning_rate": 3.014416775884666e-06,
+      "loss": 0.3197,
+      "step": 2407
+    },
+    {
+      "epoch": 2.839622641509434,
+      "grad_norm": 0.1887544725466422,
+      "learning_rate": 2.9925731760594148e-06,
+      "loss": 0.3143,
+      "step": 2408
+    },
+    {
+      "epoch": 2.8408018867924527,
+      "grad_norm": 0.1751522818756046,
+      "learning_rate": 2.9707295762341634e-06,
+      "loss": 0.291,
+      "step": 2409
+    },
+    {
+      "epoch": 2.8419811320754715,
+      "grad_norm": 0.18690282415861223,
+      "learning_rate": 2.948885976408912e-06,
+      "loss": 0.3118,
+      "step": 2410
+    },
+    {
+      "epoch": 2.8431603773584904,
+      "grad_norm": 0.17630736296690822,
+      "learning_rate": 2.9270423765836613e-06,
+      "loss": 0.3059,
+      "step": 2411
+    },
+    {
+      "epoch": 2.8443396226415096,
+      "grad_norm": 0.17603722923351758,
+      "learning_rate": 2.90519877675841e-06,
+      "loss": 0.301,
+      "step": 2412
+    },
+    {
+      "epoch": 2.8455188679245285,
+      "grad_norm": 0.18303796055407293,
+      "learning_rate": 2.8833551769331587e-06,
+      "loss": 0.3296,
+      "step": 2413
+    },
+    {
+      "epoch": 2.8466981132075473,
+      "grad_norm": 0.18026416533556505,
+      "learning_rate": 2.8615115771079073e-06,
+      "loss": 0.3065,
+      "step": 2414
+    },
+    {
+      "epoch": 2.847877358490566,
+      "grad_norm": 0.18289528262263882,
+      "learning_rate": 2.839667977282656e-06,
+      "loss": 0.3035,
+      "step": 2415
+    },
+    {
+      "epoch": 2.849056603773585,
+      "grad_norm": 0.2529614177848074,
+      "learning_rate": 2.817824377457405e-06,
+      "loss": 0.3319,
+      "step": 2416
+    },
+    {
+      "epoch": 2.8502358490566038,
+      "grad_norm": 0.18126425234406848,
+      "learning_rate": 2.795980777632154e-06,
+      "loss": 0.3103,
+      "step": 2417
+    },
+    {
+      "epoch": 2.8514150943396226,
+      "grad_norm": 0.17723391645007922,
+      "learning_rate": 2.774137177806903e-06,
+      "loss": 0.3078,
+      "step": 2418
+    },
+    {
+      "epoch": 2.8525943396226414,
+      "grad_norm": 0.1836514743842032,
+      "learning_rate": 2.7522935779816517e-06,
+      "loss": 0.3058,
+      "step": 2419
+    },
+    {
+      "epoch": 2.8537735849056602,
+      "grad_norm": 0.1836659641345924,
+      "learning_rate": 2.7304499781564004e-06,
+      "loss": 0.3383,
+      "step": 2420
+    },
+    {
+      "epoch": 2.8549528301886795,
+      "grad_norm": 0.1904961191728377,
+      "learning_rate": 2.708606378331149e-06,
+      "loss": 0.3079,
+      "step": 2421
+    },
+    {
+      "epoch": 2.856132075471698,
+      "grad_norm": 0.18655045764404765,
+      "learning_rate": 2.6867627785058978e-06,
+      "loss": 0.3284,
+      "step": 2422
+    },
+    {
+      "epoch": 2.857311320754717,
+      "grad_norm": 0.1773392136269016,
+      "learning_rate": 2.664919178680647e-06,
+      "loss": 0.3133,
+      "step": 2423
+    },
+    {
+      "epoch": 2.858490566037736,
+      "grad_norm": 0.18113648640496413,
+      "learning_rate": 2.6430755788553956e-06,
+      "loss": 0.3186,
+      "step": 2424
+    },
+    {
+      "epoch": 2.859669811320755,
+      "grad_norm": 0.17354790727659802,
+      "learning_rate": 2.6212319790301443e-06,
+      "loss": 0.3013,
+      "step": 2425
+    },
+    {
+      "epoch": 2.8608490566037736,
+      "grad_norm": 0.17896784407495717,
+      "learning_rate": 2.599388379204893e-06,
+      "loss": 0.3136,
+      "step": 2426
+    },
+    {
+      "epoch": 2.8620283018867925,
+      "grad_norm": 0.18235010556670325,
+      "learning_rate": 2.5775447793796417e-06,
+      "loss": 0.3181,
+      "step": 2427
+    },
+    {
+      "epoch": 2.8632075471698113,
+      "grad_norm": 0.17945102036784494,
+      "learning_rate": 2.5557011795543908e-06,
+      "loss": 0.3192,
+      "step": 2428
+    },
+    {
+      "epoch": 2.86438679245283,
+      "grad_norm": 0.18153021476631767,
+      "learning_rate": 2.5338575797291395e-06,
+      "loss": 0.2987,
+      "step": 2429
+    },
+    {
+      "epoch": 2.865566037735849,
+      "grad_norm": 0.1756733508160426,
+      "learning_rate": 2.5120139799038886e-06,
+      "loss": 0.3023,
+      "step": 2430
+    },
+    {
+      "epoch": 2.8667452830188678,
+      "grad_norm": 0.17559553399042704,
+      "learning_rate": 2.490170380078637e-06,
+      "loss": 0.3143,
+      "step": 2431
+    },
+    {
+      "epoch": 2.867924528301887,
+      "grad_norm": 0.17990581195760857,
+      "learning_rate": 2.468326780253386e-06,
+      "loss": 0.3196,
+      "step": 2432
+    },
+    {
+      "epoch": 2.8691037735849054,
+      "grad_norm": 0.17617380228846796,
+      "learning_rate": 2.4464831804281347e-06,
+      "loss": 0.309,
+      "step": 2433
+    },
+    {
+      "epoch": 2.8702830188679247,
+      "grad_norm": 0.17660078049244599,
+      "learning_rate": 2.4246395806028834e-06,
+      "loss": 0.3153,
+      "step": 2434
+    },
+    {
+      "epoch": 2.8714622641509435,
+      "grad_norm": 0.1961170149142407,
+      "learning_rate": 2.4027959807776325e-06,
+      "loss": 0.3231,
+      "step": 2435
+    },
+    {
+      "epoch": 2.8726415094339623,
+      "grad_norm": 0.17207064436614733,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 0.3014,
+      "step": 2436
+    },
+    {
+      "epoch": 2.873820754716981,
+      "grad_norm": 0.18171027702037518,
+      "learning_rate": 2.35910878112713e-06,
+      "loss": 0.3017,
+      "step": 2437
+    },
+    {
+      "epoch": 2.875,
+      "grad_norm": 0.1822737009601447,
+      "learning_rate": 2.3372651813018786e-06,
+      "loss": 0.3098,
+      "step": 2438
+    },
+    {
+      "epoch": 2.876179245283019,
+      "grad_norm": 0.18116174624852444,
+      "learning_rate": 2.3154215814766277e-06,
+      "loss": 0.3233,
+      "step": 2439
+    },
+    {
+      "epoch": 2.8773584905660377,
+      "grad_norm": 0.1764633228619726,
+      "learning_rate": 2.2935779816513764e-06,
+      "loss": 0.3014,
+      "step": 2440
+    },
+    {
+      "epoch": 2.8785377358490565,
+      "grad_norm": 0.18455495847726117,
+      "learning_rate": 2.271734381826125e-06,
+      "loss": 0.3072,
+      "step": 2441
+    },
+    {
+      "epoch": 2.8797169811320753,
+      "grad_norm": 0.17393672547685504,
+      "learning_rate": 2.2498907820008738e-06,
+      "loss": 0.2894,
+      "step": 2442
+    },
+    {
+      "epoch": 2.8808962264150946,
+      "grad_norm": 0.181588623360161,
+      "learning_rate": 2.2280471821756225e-06,
+      "loss": 0.3171,
+      "step": 2443
+    },
+    {
+      "epoch": 2.882075471698113,
+      "grad_norm": 0.32117181924421945,
+      "learning_rate": 2.2062035823503716e-06,
+      "loss": 0.3216,
+      "step": 2444
+    },
+    {
+      "epoch": 2.8832547169811322,
+      "grad_norm": 0.17605458832435372,
+      "learning_rate": 2.1843599825251203e-06,
+      "loss": 0.304,
+      "step": 2445
+    },
+    {
+      "epoch": 2.884433962264151,
+      "grad_norm": 0.18620893662467478,
+      "learning_rate": 2.1625163826998694e-06,
+      "loss": 0.3059,
+      "step": 2446
+    },
+    {
+      "epoch": 2.88561320754717,
+      "grad_norm": 0.18313929771645862,
+      "learning_rate": 2.1406727828746177e-06,
+      "loss": 0.3322,
+      "step": 2447
+    },
+    {
+      "epoch": 2.8867924528301887,
+      "grad_norm": 0.21394045863542901,
+      "learning_rate": 2.1188291830493664e-06,
+      "loss": 0.3101,
+      "step": 2448
+    },
+    {
+      "epoch": 2.8879716981132075,
+      "grad_norm": 0.17674572803882183,
+      "learning_rate": 2.0969855832241155e-06,
+      "loss": 0.309,
+      "step": 2449
+    },
+    {
+      "epoch": 2.8891509433962264,
+      "grad_norm": 0.1789351815602257,
+      "learning_rate": 2.075141983398864e-06,
+      "loss": 0.3061,
+      "step": 2450
+    },
+    {
+      "epoch": 2.890330188679245,
+      "grad_norm": 0.1770496796587849,
+      "learning_rate": 2.0532983835736133e-06,
+      "loss": 0.3052,
+      "step": 2451
+    },
+    {
+      "epoch": 2.891509433962264,
+      "grad_norm": 0.17985467931135846,
+      "learning_rate": 2.0314547837483616e-06,
+      "loss": 0.3119,
+      "step": 2452
+    },
+    {
+      "epoch": 2.892688679245283,
+      "grad_norm": 0.18056381595907023,
+      "learning_rate": 2.0096111839231107e-06,
+      "loss": 0.3102,
+      "step": 2453
+    },
+    {
+      "epoch": 2.893867924528302,
+      "grad_norm": 0.1699837296195604,
+      "learning_rate": 1.9877675840978594e-06,
+      "loss": 0.3027,
+      "step": 2454
+    },
+    {
+      "epoch": 2.8950471698113205,
+      "grad_norm": 0.17178551850953075,
+      "learning_rate": 1.965923984272608e-06,
+      "loss": 0.3049,
+      "step": 2455
+    },
+    {
+      "epoch": 2.8962264150943398,
+      "grad_norm": 0.17139274492093895,
+      "learning_rate": 1.9440803844473572e-06,
+      "loss": 0.2945,
+      "step": 2456
+    },
+    {
+      "epoch": 2.8974056603773586,
+      "grad_norm": 1.4139687382239683,
+      "learning_rate": 1.9222367846221055e-06,
+      "loss": 0.3327,
+      "step": 2457
+    },
+    {
+      "epoch": 2.8985849056603774,
+      "grad_norm": 0.17813014130898164,
+      "learning_rate": 1.9003931847968546e-06,
+      "loss": 0.3122,
+      "step": 2458
+    },
+    {
+      "epoch": 2.8997641509433962,
+      "grad_norm": 0.1785535340672431,
+      "learning_rate": 1.8785495849716033e-06,
+      "loss": 0.3012,
+      "step": 2459
+    },
+    {
+      "epoch": 2.900943396226415,
+      "grad_norm": 0.17866653260700685,
+      "learning_rate": 1.8567059851463522e-06,
+      "loss": 0.3073,
+      "step": 2460
+    },
+    {
+      "epoch": 2.902122641509434,
+      "grad_norm": 0.17489747313336276,
+      "learning_rate": 1.8348623853211011e-06,
+      "loss": 0.2968,
+      "step": 2461
+    },
+    {
+      "epoch": 2.9033018867924527,
+      "grad_norm": 0.17533354314615884,
+      "learning_rate": 1.81301878549585e-06,
+      "loss": 0.3026,
+      "step": 2462
+    },
+    {
+      "epoch": 2.9044811320754715,
+      "grad_norm": 0.1726298571888876,
+      "learning_rate": 1.7911751856705985e-06,
+      "loss": 0.2996,
+      "step": 2463
+    },
+    {
+      "epoch": 2.9056603773584904,
+      "grad_norm": 0.17782797850006998,
+      "learning_rate": 1.7693315858453474e-06,
+      "loss": 0.301,
+      "step": 2464
+    },
+    {
+      "epoch": 2.9068396226415096,
+      "grad_norm": 0.17927665097077458,
+      "learning_rate": 1.7474879860200961e-06,
+      "loss": 0.3241,
+      "step": 2465
+    },
+    {
+      "epoch": 2.9080188679245285,
+      "grad_norm": 0.17158258355618458,
+      "learning_rate": 1.725644386194845e-06,
+      "loss": 0.2942,
+      "step": 2466
+    },
+    {
+      "epoch": 2.9091981132075473,
+      "grad_norm": 0.17014161243227532,
+      "learning_rate": 1.703800786369594e-06,
+      "loss": 0.2867,
+      "step": 2467
+    },
+    {
+      "epoch": 2.910377358490566,
+      "grad_norm": 0.17967220565867034,
+      "learning_rate": 1.6819571865443424e-06,
+      "loss": 0.312,
+      "step": 2468
+    },
+    {
+      "epoch": 2.911556603773585,
+      "grad_norm": 0.18008635809611737,
+      "learning_rate": 1.6601135867190913e-06,
+      "loss": 0.3005,
+      "step": 2469
+    },
+    {
+      "epoch": 2.9127358490566038,
+      "grad_norm": 0.18252865401266094,
+      "learning_rate": 1.6382699868938402e-06,
+      "loss": 0.2858,
+      "step": 2470
+    },
+    {
+      "epoch": 2.9139150943396226,
+      "grad_norm": 0.17702705760107526,
+      "learning_rate": 1.6164263870685891e-06,
+      "loss": 0.3027,
+      "step": 2471
+    },
+    {
+      "epoch": 2.9150943396226414,
+      "grad_norm": 0.17899285380762192,
+      "learning_rate": 1.5945827872433378e-06,
+      "loss": 0.3094,
+      "step": 2472
+    },
+    {
+      "epoch": 2.9162735849056602,
+      "grad_norm": 0.17549250801109176,
+      "learning_rate": 1.5727391874180865e-06,
+      "loss": 0.3122,
+      "step": 2473
+    },
+    {
+      "epoch": 2.9174528301886795,
+      "grad_norm": 0.18643955049709746,
+      "learning_rate": 1.5508955875928354e-06,
+      "loss": 0.3098,
+      "step": 2474
+    },
+    {
+      "epoch": 2.918632075471698,
+      "grad_norm": 0.175863298528974,
+      "learning_rate": 1.5290519877675841e-06,
+      "loss": 0.3081,
+      "step": 2475
+    },
+    {
+      "epoch": 2.919811320754717,
+      "grad_norm": 0.17939291218900172,
+      "learning_rate": 1.507208387942333e-06,
+      "loss": 0.3215,
+      "step": 2476
+    },
+    {
+      "epoch": 2.920990566037736,
+      "grad_norm": 0.17352572554844495,
+      "learning_rate": 1.4853647881170817e-06,
+      "loss": 0.2965,
+      "step": 2477
+    },
+    {
+      "epoch": 2.922169811320755,
+      "grad_norm": 0.17932648869911344,
+      "learning_rate": 1.4635211882918306e-06,
+      "loss": 0.314,
+      "step": 2478
+    },
+    {
+      "epoch": 2.9233490566037736,
+      "grad_norm": 0.18300870900406985,
+      "learning_rate": 1.4416775884665793e-06,
+      "loss": 0.3225,
+      "step": 2479
+    },
+    {
+      "epoch": 2.9245283018867925,
+      "grad_norm": 0.17208807913591603,
+      "learning_rate": 1.419833988641328e-06,
+      "loss": 0.3009,
+      "step": 2480
+    },
+    {
+      "epoch": 2.9257075471698113,
+      "grad_norm": 0.1748396310213249,
+      "learning_rate": 1.397990388816077e-06,
+      "loss": 0.2985,
+      "step": 2481
+    },
+    {
+      "epoch": 2.92688679245283,
+      "grad_norm": 0.1812579673096089,
+      "learning_rate": 1.3761467889908258e-06,
+      "loss": 0.3207,
+      "step": 2482
+    },
+    {
+      "epoch": 2.928066037735849,
+      "grad_norm": 0.1751494600060829,
+      "learning_rate": 1.3543031891655745e-06,
+      "loss": 0.3082,
+      "step": 2483
+    },
+    {
+      "epoch": 2.9292452830188678,
+      "grad_norm": 0.20330271042937156,
+      "learning_rate": 1.3324595893403234e-06,
+      "loss": 0.3146,
+      "step": 2484
+    },
+    {
+      "epoch": 2.930424528301887,
+      "grad_norm": 0.1782270279449391,
+      "learning_rate": 1.3106159895150721e-06,
+      "loss": 0.3082,
+      "step": 2485
+    },
+    {
+      "epoch": 2.9316037735849054,
+      "grad_norm": 0.1707615455094258,
+      "learning_rate": 1.2887723896898208e-06,
+      "loss": 0.2988,
+      "step": 2486
+    },
+    {
+      "epoch": 2.9327830188679247,
+      "grad_norm": 0.1714045397724636,
+      "learning_rate": 1.2669287898645697e-06,
+      "loss": 0.3064,
+      "step": 2487
+    },
+    {
+      "epoch": 2.9339622641509435,
+      "grad_norm": 0.18187522364892267,
+      "learning_rate": 1.2450851900393184e-06,
+      "loss": 0.3116,
+      "step": 2488
+    },
+    {
+      "epoch": 2.9351415094339623,
+      "grad_norm": 0.1734800605140774,
+      "learning_rate": 1.2232415902140673e-06,
+      "loss": 0.3074,
+      "step": 2489
+    },
+    {
+      "epoch": 2.936320754716981,
+      "grad_norm": 0.17373796287555546,
+      "learning_rate": 1.2013979903888162e-06,
+      "loss": 0.3088,
+      "step": 2490
+    },
+    {
+      "epoch": 2.9375,
+      "grad_norm": 0.17683315129147673,
+      "learning_rate": 1.179554390563565e-06,
+      "loss": 0.2938,
+      "step": 2491
+    },
+    {
+      "epoch": 2.938679245283019,
+      "grad_norm": 0.172022115567748,
+      "learning_rate": 1.1577107907383138e-06,
+      "loss": 0.2984,
+      "step": 2492
+    },
+    {
+      "epoch": 2.9398584905660377,
+      "grad_norm": 0.17273551024760692,
+      "learning_rate": 1.1358671909130625e-06,
+      "loss": 0.2924,
+      "step": 2493
+    },
+    {
+      "epoch": 2.9410377358490565,
+      "grad_norm": 0.1853509311944495,
+      "learning_rate": 1.1140235910878112e-06,
+      "loss": 0.3162,
+      "step": 2494
+    },
+    {
+      "epoch": 2.9422169811320753,
+      "grad_norm": 0.17921977952137783,
+      "learning_rate": 1.0921799912625601e-06,
+      "loss": 0.2969,
+      "step": 2495
+    },
+    {
+      "epoch": 2.9433962264150946,
+      "grad_norm": 0.18341167290156046,
+      "learning_rate": 1.0703363914373088e-06,
+      "loss": 0.3163,
+      "step": 2496
+    },
+    {
+      "epoch": 2.944575471698113,
+      "grad_norm": 0.1778817008018014,
+      "learning_rate": 1.0484927916120577e-06,
+      "loss": 0.3177,
+      "step": 2497
+    },
+    {
+      "epoch": 2.9457547169811322,
+      "grad_norm": 0.17878509748957608,
+      "learning_rate": 1.0266491917868067e-06,
+      "loss": 0.3145,
+      "step": 2498
+    },
+    {
+      "epoch": 2.946933962264151,
+      "grad_norm": 0.20496756275731176,
+      "learning_rate": 1.0048055919615554e-06,
+      "loss": 0.3125,
+      "step": 2499
+    },
+    {
+      "epoch": 2.94811320754717,
+      "grad_norm": 0.23053632795550388,
+      "learning_rate": 9.82961992136304e-07,
+      "loss": 0.3164,
+      "step": 2500
+    },
+    {
+      "epoch": 2.9492924528301887,
+      "grad_norm": 0.1884267973435929,
+      "learning_rate": 9.611183923110527e-07,
+      "loss": 0.3192,
+      "step": 2501
+    },
+    {
+      "epoch": 2.9504716981132075,
+      "grad_norm": 0.17684910119185598,
+      "learning_rate": 9.392747924858016e-07,
+      "loss": 0.3047,
+      "step": 2502
+    },
+    {
+      "epoch": 2.9516509433962264,
+      "grad_norm": 0.16869543630782027,
+      "learning_rate": 9.174311926605506e-07,
+      "loss": 0.2995,
+      "step": 2503
+    },
+    {
+      "epoch": 2.952830188679245,
+      "grad_norm": 0.17744644715566724,
+      "learning_rate": 8.955875928352993e-07,
+      "loss": 0.3091,
+      "step": 2504
+    },
+    {
+      "epoch": 2.954009433962264,
+      "grad_norm": 0.17200855103158258,
+      "learning_rate": 8.737439930100481e-07,
+      "loss": 0.3073,
+      "step": 2505
+    },
+    {
+      "epoch": 2.955188679245283,
+      "grad_norm": 0.17829553143641072,
+      "learning_rate": 8.51900393184797e-07,
+      "loss": 0.3022,
+      "step": 2506
+    },
+    {
+      "epoch": 2.956367924528302,
+      "grad_norm": 0.1748459078141117,
+      "learning_rate": 8.300567933595457e-07,
+      "loss": 0.3085,
+      "step": 2507
+    },
+    {
+      "epoch": 2.9575471698113205,
+      "grad_norm": 0.17749552901001048,
+      "learning_rate": 8.082131935342946e-07,
+      "loss": 0.3142,
+      "step": 2508
+    },
+    {
+      "epoch": 2.9587264150943398,
+      "grad_norm": 0.17931480984800935,
+      "learning_rate": 7.863695937090433e-07,
+      "loss": 0.3043,
+      "step": 2509
+    },
+    {
+      "epoch": 2.9599056603773586,
+      "grad_norm": 0.19163975475985795,
+      "learning_rate": 7.645259938837921e-07,
+      "loss": 0.3135,
+      "step": 2510
+    },
+    {
+      "epoch": 2.9610849056603774,
+      "grad_norm": 0.1785597885610294,
+      "learning_rate": 7.426823940585409e-07,
+      "loss": 0.3093,
+      "step": 2511
+    },
+    {
+      "epoch": 2.9622641509433962,
+      "grad_norm": 0.18050226915689238,
+      "learning_rate": 7.208387942332897e-07,
+      "loss": 0.315,
+      "step": 2512
+    },
+    {
+      "epoch": 2.963443396226415,
+      "grad_norm": 0.21766622490686485,
+      "learning_rate": 6.989951944080385e-07,
+      "loss": 0.33,
+      "step": 2513
+    },
+    {
+      "epoch": 2.964622641509434,
+      "grad_norm": 0.1771519071324037,
+      "learning_rate": 6.771515945827873e-07,
+      "loss": 0.3019,
+      "step": 2514
+    },
+    {
+      "epoch": 2.9658018867924527,
+      "grad_norm": 0.18448833558359257,
+      "learning_rate": 6.553079947575361e-07,
+      "loss": 0.3143,
+      "step": 2515
+    },
+    {
+      "epoch": 2.9669811320754715,
+      "grad_norm": 0.1769682110348559,
+      "learning_rate": 6.334643949322849e-07,
+      "loss": 0.2931,
+      "step": 2516
+    },
+    {
+      "epoch": 2.9681603773584904,
+      "grad_norm": 0.17439127766935392,
+      "learning_rate": 6.116207951070337e-07,
+      "loss": 0.2918,
+      "step": 2517
+    },
+    {
+      "epoch": 2.9693396226415096,
+      "grad_norm": 0.17902772389964206,
+      "learning_rate": 5.897771952817825e-07,
+      "loss": 0.3038,
+      "step": 2518
+    },
+    {
+      "epoch": 2.9705188679245285,
+      "grad_norm": 0.17865003456027628,
+      "learning_rate": 5.679335954565313e-07,
+      "loss": 0.2998,
+      "step": 2519
+    },
+    {
+      "epoch": 2.9716981132075473,
+      "grad_norm": 0.1765625077988113,
+      "learning_rate": 5.460899956312801e-07,
+      "loss": 0.2975,
+      "step": 2520
+    },
+    {
+      "epoch": 2.972877358490566,
+      "grad_norm": 0.1831001173635964,
+      "learning_rate": 5.242463958060289e-07,
+      "loss": 0.3225,
+      "step": 2521
+    },
+    {
+      "epoch": 2.974056603773585,
+      "grad_norm": 0.18359331944650628,
+      "learning_rate": 5.024027959807777e-07,
+      "loss": 0.3053,
+      "step": 2522
+    },
+    {
+      "epoch": 2.9752358490566038,
+      "grad_norm": 0.17115693217167863,
+      "learning_rate": 4.805591961555264e-07,
+      "loss": 0.2969,
+      "step": 2523
+    },
+    {
+      "epoch": 2.9764150943396226,
+      "grad_norm": 0.1696232403102028,
+      "learning_rate": 4.587155963302753e-07,
+      "loss": 0.2947,
+      "step": 2524
+    },
+    {
+      "epoch": 2.9775943396226414,
+      "grad_norm": 0.178123046773488,
+      "learning_rate": 4.3687199650502403e-07,
+      "loss": 0.307,
+      "step": 2525
+    },
+    {
+      "epoch": 2.9787735849056602,
+      "grad_norm": 0.16688006273768935,
+      "learning_rate": 4.1502839667977283e-07,
+      "loss": 0.2895,
+      "step": 2526
+    },
+    {
+      "epoch": 2.9799528301886795,
+      "grad_norm": 0.18112894152766457,
+      "learning_rate": 3.9318479685452163e-07,
+      "loss": 0.3091,
+      "step": 2527
+    },
+    {
+      "epoch": 2.981132075471698,
+      "grad_norm": 0.16680446101230792,
+      "learning_rate": 3.7134119702927043e-07,
+      "loss": 0.2887,
+      "step": 2528
+    },
+    {
+      "epoch": 2.982311320754717,
+      "grad_norm": 0.17099251644773078,
+      "learning_rate": 3.4949759720401923e-07,
+      "loss": 0.2953,
+      "step": 2529
+    },
+    {
+      "epoch": 2.983490566037736,
+      "grad_norm": 0.16955325432054616,
+      "learning_rate": 3.2765399737876803e-07,
+      "loss": 0.2891,
+      "step": 2530
+    },
+    {
+      "epoch": 2.984669811320755,
+      "grad_norm": 0.18443223952062773,
+      "learning_rate": 3.0581039755351683e-07,
+      "loss": 0.3076,
+      "step": 2531
+    },
+    {
+      "epoch": 2.9858490566037736,
+      "grad_norm": 0.1746280595579024,
+      "learning_rate": 2.8396679772826564e-07,
+      "loss": 0.2911,
+      "step": 2532
+    },
+    {
+      "epoch": 2.9870283018867925,
+      "grad_norm": 0.17391551319123555,
+      "learning_rate": 2.6212319790301444e-07,
+      "loss": 0.3144,
+      "step": 2533
+    },
+    {
+      "epoch": 2.9882075471698113,
+      "grad_norm": 0.17643751755912135,
+      "learning_rate": 2.402795980777632e-07,
+      "loss": 0.3069,
+      "step": 2534
+    },
+    {
+      "epoch": 2.98938679245283,
+      "grad_norm": 0.17535564731205386,
+      "learning_rate": 2.1843599825251201e-07,
+      "loss": 0.307,
+      "step": 2535
+    },
+    {
+      "epoch": 2.990566037735849,
+      "grad_norm": 0.17197828472844037,
+      "learning_rate": 1.9659239842726081e-07,
+      "loss": 0.296,
+      "step": 2536
+    },
+    {
+      "epoch": 2.9917452830188678,
+      "grad_norm": 0.18029021894372535,
+      "learning_rate": 1.7474879860200962e-07,
+      "loss": 0.3185,
+      "step": 2537
+    },
+    {
+      "epoch": 2.992924528301887,
+      "grad_norm": 0.17569550946943568,
+      "learning_rate": 1.5290519877675842e-07,
+      "loss": 0.3036,
+      "step": 2538
+    },
+    {
+      "epoch": 2.9941037735849054,
+      "grad_norm": 0.1822133019688892,
+      "learning_rate": 1.3106159895150722e-07,
+      "loss": 0.3084,
+      "step": 2539
+    },
+    {
+      "epoch": 2.9952830188679247,
+      "grad_norm": 0.1756721218049293,
+      "learning_rate": 1.0921799912625601e-07,
+      "loss": 0.3042,
+      "step": 2540
+    },
+    {
+      "epoch": 2.9964622641509435,
+      "grad_norm": 0.17163186303515487,
+      "learning_rate": 8.737439930100481e-08,
+      "loss": 0.3054,
+      "step": 2541
+    },
+    {
+      "epoch": 2.9976415094339623,
+      "grad_norm": 0.17447244692297156,
+      "learning_rate": 6.553079947575361e-08,
+      "loss": 0.3049,
+      "step": 2542
+    },
+    {
+      "epoch": 2.998820754716981,
+      "grad_norm": 0.16601292825818093,
+      "learning_rate": 4.3687199650502404e-08,
+      "loss": 0.2797,
+      "step": 2543
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.1772075545853887,
+      "learning_rate": 2.1843599825251202e-08,
+      "loss": 0.3076,
+      "step": 2544
+    },
+    {
+      "epoch": 3.0,
+      "step": 2544,
+      "total_flos": 2.8292565162912645e+19,
+      "train_loss": 0.4820789343226443,
+      "train_runtime": 73019.0476,
+      "train_samples_per_second": 0.557,
+      "train_steps_per_second": 0.035
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2544,
+  "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": 2.8292565162912645e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}