diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,28560 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 4074,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007363770250368188,
+      "grad_norm": 5.638559341430664,
+      "learning_rate": 2.450980392156863e-08,
+      "loss": 0.8206,
+      "step": 1
+    },
+    {
+      "epoch": 0.0014727540500736377,
+      "grad_norm": 5.574512004852295,
+      "learning_rate": 4.901960784313726e-08,
+      "loss": 0.8179,
+      "step": 2
+    },
+    {
+      "epoch": 0.0022091310751104565,
+      "grad_norm": 5.833616733551025,
+      "learning_rate": 7.352941176470589e-08,
+      "loss": 0.8298,
+      "step": 3
+    },
+    {
+      "epoch": 0.0029455081001472753,
+      "grad_norm": 5.760316848754883,
+      "learning_rate": 9.803921568627452e-08,
+      "loss": 0.8612,
+      "step": 4
+    },
+    {
+      "epoch": 0.003681885125184094,
+      "grad_norm": 5.582140922546387,
+      "learning_rate": 1.2254901960784314e-07,
+      "loss": 0.8597,
+      "step": 5
+    },
+    {
+      "epoch": 0.004418262150220913,
+      "grad_norm": 5.870736122131348,
+      "learning_rate": 1.4705882352941178e-07,
+      "loss": 0.8647,
+      "step": 6
+    },
+    {
+      "epoch": 0.005154639175257732,
+      "grad_norm": 5.794524192810059,
+      "learning_rate": 1.7156862745098042e-07,
+      "loss": 0.8709,
+      "step": 7
+    },
+    {
+      "epoch": 0.005891016200294551,
+      "grad_norm": 5.833931922912598,
+      "learning_rate": 1.9607843137254904e-07,
+      "loss": 0.8438,
+      "step": 8
+    },
+    {
+      "epoch": 0.0066273932253313695,
+      "grad_norm": 5.879095554351807,
+      "learning_rate": 2.2058823529411768e-07,
+      "loss": 0.8583,
+      "step": 9
+    },
+    {
+      "epoch": 0.007363770250368188,
+      "grad_norm": 5.551036357879639,
+      "learning_rate": 2.4509803921568627e-07,
+      "loss": 0.8273,
+      "step": 10
+    },
+    {
+      "epoch": 0.008100147275405008,
+      "grad_norm": 5.67064094543457,
+      "learning_rate": 2.696078431372549e-07,
+      "loss": 0.8328,
+      "step": 11
+    },
+    {
+      "epoch": 0.008836524300441826,
+      "grad_norm": 5.897095203399658,
+      "learning_rate": 2.9411764705882356e-07,
+      "loss": 0.8742,
+      "step": 12
+    },
+    {
+      "epoch": 0.009572901325478646,
+      "grad_norm": 5.685695648193359,
+      "learning_rate": 3.1862745098039215e-07,
+      "loss": 0.8389,
+      "step": 13
+    },
+    {
+      "epoch": 0.010309278350515464,
+      "grad_norm": 5.742497444152832,
+      "learning_rate": 3.4313725490196084e-07,
+      "loss": 0.8695,
+      "step": 14
+    },
+    {
+      "epoch": 0.011045655375552283,
+      "grad_norm": 5.5533037185668945,
+      "learning_rate": 3.6764705882352943e-07,
+      "loss": 0.8515,
+      "step": 15
+    },
+    {
+      "epoch": 0.011782032400589101,
+      "grad_norm": 5.345178604125977,
+      "learning_rate": 3.921568627450981e-07,
+      "loss": 0.816,
+      "step": 16
+    },
+    {
+      "epoch": 0.012518409425625921,
+      "grad_norm": 5.715511322021484,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 0.9075,
+      "step": 17
+    },
+    {
+      "epoch": 0.013254786450662739,
+      "grad_norm": 5.173557758331299,
+      "learning_rate": 4.4117647058823536e-07,
+      "loss": 0.8311,
+      "step": 18
+    },
+    {
+      "epoch": 0.013991163475699559,
+      "grad_norm": 5.542694568634033,
+      "learning_rate": 4.6568627450980395e-07,
+      "loss": 0.8385,
+      "step": 19
+    },
+    {
+      "epoch": 0.014727540500736377,
+      "grad_norm": 4.2676191329956055,
+      "learning_rate": 4.901960784313725e-07,
+      "loss": 0.8088,
+      "step": 20
+    },
+    {
+      "epoch": 0.015463917525773196,
+      "grad_norm": 4.362669944763184,
+      "learning_rate": 5.147058823529412e-07,
+      "loss": 0.8291,
+      "step": 21
+    },
+    {
+      "epoch": 0.016200294550810016,
+      "grad_norm": 4.3511643409729,
+      "learning_rate": 5.392156862745098e-07,
+      "loss": 0.8043,
+      "step": 22
+    },
+    {
+      "epoch": 0.016936671575846832,
+      "grad_norm": 4.051642894744873,
+      "learning_rate": 5.637254901960785e-07,
+      "loss": 0.744,
+      "step": 23
+    },
+    {
+      "epoch": 0.017673048600883652,
+      "grad_norm": 4.032516002655029,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 0.7975,
+      "step": 24
+    },
+    {
+      "epoch": 0.018409425625920472,
+      "grad_norm": 3.7897393703460693,
+      "learning_rate": 6.127450980392157e-07,
+      "loss": 0.7686,
+      "step": 25
+    },
+    {
+      "epoch": 0.01914580265095729,
+      "grad_norm": 3.99767804145813,
+      "learning_rate": 6.372549019607843e-07,
+      "loss": 0.777,
+      "step": 26
+    },
+    {
+      "epoch": 0.019882179675994108,
+      "grad_norm": 2.312422037124634,
+      "learning_rate": 6.61764705882353e-07,
+      "loss": 0.7323,
+      "step": 27
+    },
+    {
+      "epoch": 0.020618556701030927,
+      "grad_norm": 2.436676025390625,
+      "learning_rate": 6.862745098039217e-07,
+      "loss": 0.7402,
+      "step": 28
+    },
+    {
+      "epoch": 0.021354933726067747,
+      "grad_norm": 2.346937417984009,
+      "learning_rate": 7.107843137254903e-07,
+      "loss": 0.7634,
+      "step": 29
+    },
+    {
+      "epoch": 0.022091310751104567,
+      "grad_norm": 2.329108476638794,
+      "learning_rate": 7.352941176470589e-07,
+      "loss": 0.8077,
+      "step": 30
+    },
+    {
+      "epoch": 0.022827687776141383,
+      "grad_norm": 2.1621270179748535,
+      "learning_rate": 7.598039215686275e-07,
+      "loss": 0.7406,
+      "step": 31
+    },
+    {
+      "epoch": 0.023564064801178203,
+      "grad_norm": 2.004992723464966,
+      "learning_rate": 7.843137254901962e-07,
+      "loss": 0.7373,
+      "step": 32
+    },
+    {
+      "epoch": 0.024300441826215022,
+      "grad_norm": 1.9387824535369873,
+      "learning_rate": 8.088235294117648e-07,
+      "loss": 0.6983,
+      "step": 33
+    },
+    {
+      "epoch": 0.025036818851251842,
+      "grad_norm": 1.8947597742080688,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.7385,
+      "step": 34
+    },
+    {
+      "epoch": 0.02577319587628866,
+      "grad_norm": 1.8148812055587769,
+      "learning_rate": 8.57843137254902e-07,
+      "loss": 0.7547,
+      "step": 35
+    },
+    {
+      "epoch": 0.026509572901325478,
+      "grad_norm": 1.3978095054626465,
+      "learning_rate": 8.823529411764707e-07,
+      "loss": 0.721,
+      "step": 36
+    },
+    {
+      "epoch": 0.027245949926362298,
+      "grad_norm": 1.5461397171020508,
+      "learning_rate": 9.068627450980393e-07,
+      "loss": 0.703,
+      "step": 37
+    },
+    {
+      "epoch": 0.027982326951399118,
+      "grad_norm": 1.797391414642334,
+      "learning_rate": 9.313725490196079e-07,
+      "loss": 0.6854,
+      "step": 38
+    },
+    {
+      "epoch": 0.028718703976435934,
+      "grad_norm": 2.147284746170044,
+      "learning_rate": 9.558823529411764e-07,
+      "loss": 0.736,
+      "step": 39
+    },
+    {
+      "epoch": 0.029455081001472753,
+      "grad_norm": 2.267225980758667,
+      "learning_rate": 9.80392156862745e-07,
+      "loss": 0.7235,
+      "step": 40
+    },
+    {
+      "epoch": 0.030191458026509573,
+      "grad_norm": 2.0823991298675537,
+      "learning_rate": 1.0049019607843138e-06,
+      "loss": 0.7262,
+      "step": 41
+    },
+    {
+      "epoch": 0.030927835051546393,
+      "grad_norm": 2.126589775085449,
+      "learning_rate": 1.0294117647058825e-06,
+      "loss": 0.7298,
+      "step": 42
+    },
+    {
+      "epoch": 0.03166421207658321,
+      "grad_norm": 1.8871612548828125,
+      "learning_rate": 1.0539215686274512e-06,
+      "loss": 0.7012,
+      "step": 43
+    },
+    {
+      "epoch": 0.03240058910162003,
+      "grad_norm": 1.6312057971954346,
+      "learning_rate": 1.0784313725490197e-06,
+      "loss": 0.6583,
+      "step": 44
+    },
+    {
+      "epoch": 0.03313696612665685,
+      "grad_norm": 1.5015575885772705,
+      "learning_rate": 1.1029411764705884e-06,
+      "loss": 0.6593,
+      "step": 45
+    },
+    {
+      "epoch": 0.033873343151693665,
+      "grad_norm": 1.2276663780212402,
+      "learning_rate": 1.127450980392157e-06,
+      "loss": 0.6918,
+      "step": 46
+    },
+    {
+      "epoch": 0.03460972017673049,
+      "grad_norm": 1.0622628927230835,
+      "learning_rate": 1.1519607843137255e-06,
+      "loss": 0.6918,
+      "step": 47
+    },
+    {
+      "epoch": 0.035346097201767304,
+      "grad_norm": 0.9096408486366272,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 0.6847,
+      "step": 48
+    },
+    {
+      "epoch": 0.03608247422680412,
+      "grad_norm": 0.8563209772109985,
+      "learning_rate": 1.200980392156863e-06,
+      "loss": 0.6646,
+      "step": 49
+    },
+    {
+      "epoch": 0.036818851251840944,
+      "grad_norm": 0.9862013459205627,
+      "learning_rate": 1.2254901960784314e-06,
+      "loss": 0.6732,
+      "step": 50
+    },
+    {
+      "epoch": 0.03755522827687776,
+      "grad_norm": 1.099507212638855,
+      "learning_rate": 1.25e-06,
+      "loss": 0.6882,
+      "step": 51
+    },
+    {
+      "epoch": 0.03829160530191458,
+      "grad_norm": 1.1145635843276978,
+      "learning_rate": 1.2745098039215686e-06,
+      "loss": 0.6658,
+      "step": 52
+    },
+    {
+      "epoch": 0.0390279823269514,
+      "grad_norm": 1.0199302434921265,
+      "learning_rate": 1.2990196078431375e-06,
+      "loss": 0.6439,
+      "step": 53
+    },
+    {
+      "epoch": 0.039764359351988215,
+      "grad_norm": 0.8398745059967041,
+      "learning_rate": 1.323529411764706e-06,
+      "loss": 0.641,
+      "step": 54
+    },
+    {
+      "epoch": 0.04050073637702504,
+      "grad_norm": 0.7679091095924377,
+      "learning_rate": 1.3480392156862745e-06,
+      "loss": 0.6067,
+      "step": 55
+    },
+    {
+      "epoch": 0.041237113402061855,
+      "grad_norm": 0.7991195321083069,
+      "learning_rate": 1.3725490196078434e-06,
+      "loss": 0.677,
+      "step": 56
+    },
+    {
+      "epoch": 0.04197349042709867,
+      "grad_norm": 0.7389347553253174,
+      "learning_rate": 1.3970588235294119e-06,
+      "loss": 0.6504,
+      "step": 57
+    },
+    {
+      "epoch": 0.042709867452135494,
+      "grad_norm": 0.7039929628372192,
+      "learning_rate": 1.4215686274509805e-06,
+      "loss": 0.6002,
+      "step": 58
+    },
+    {
+      "epoch": 0.04344624447717231,
+      "grad_norm": 0.7065069675445557,
+      "learning_rate": 1.4460784313725492e-06,
+      "loss": 0.581,
+      "step": 59
+    },
+    {
+      "epoch": 0.044182621502209134,
+      "grad_norm": 0.6952792406082153,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 0.6523,
+      "step": 60
+    },
+    {
+      "epoch": 0.04491899852724595,
+      "grad_norm": 0.6492939591407776,
+      "learning_rate": 1.4950980392156864e-06,
+      "loss": 0.6587,
+      "step": 61
+    },
+    {
+      "epoch": 0.045655375552282766,
+      "grad_norm": 0.6436308026313782,
+      "learning_rate": 1.519607843137255e-06,
+      "loss": 0.6048,
+      "step": 62
+    },
+    {
+      "epoch": 0.04639175257731959,
+      "grad_norm": 0.6987003087997437,
+      "learning_rate": 1.5441176470588238e-06,
+      "loss": 0.6353,
+      "step": 63
+    },
+    {
+      "epoch": 0.047128129602356406,
+      "grad_norm": 0.698427677154541,
+      "learning_rate": 1.5686274509803923e-06,
+      "loss": 0.63,
+      "step": 64
+    },
+    {
+      "epoch": 0.04786450662739323,
+      "grad_norm": 0.7047044634819031,
+      "learning_rate": 1.5931372549019608e-06,
+      "loss": 0.6048,
+      "step": 65
+    },
+    {
+      "epoch": 0.048600883652430045,
+      "grad_norm": 0.6246572732925415,
+      "learning_rate": 1.6176470588235297e-06,
+      "loss": 0.6264,
+      "step": 66
+    },
+    {
+      "epoch": 0.04933726067746686,
+      "grad_norm": 0.6060186624526978,
+      "learning_rate": 1.6421568627450982e-06,
+      "loss": 0.6277,
+      "step": 67
+    },
+    {
+      "epoch": 0.050073637702503684,
+      "grad_norm": 0.5113458037376404,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.5985,
+      "step": 68
+    },
+    {
+      "epoch": 0.0508100147275405,
+      "grad_norm": 0.6172787547111511,
+      "learning_rate": 1.6911764705882356e-06,
+      "loss": 0.622,
+      "step": 69
+    },
+    {
+      "epoch": 0.05154639175257732,
+      "grad_norm": 0.4991741180419922,
+      "learning_rate": 1.715686274509804e-06,
+      "loss": 0.6048,
+      "step": 70
+    },
+    {
+      "epoch": 0.05228276877761414,
+      "grad_norm": 0.5184860825538635,
+      "learning_rate": 1.7401960784313725e-06,
+      "loss": 0.6033,
+      "step": 71
+    },
+    {
+      "epoch": 0.053019145802650956,
+      "grad_norm": 0.4985435903072357,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 0.5792,
+      "step": 72
+    },
+    {
+      "epoch": 0.05375552282768778,
+      "grad_norm": 0.5244366526603699,
+      "learning_rate": 1.78921568627451e-06,
+      "loss": 0.6168,
+      "step": 73
+    },
+    {
+      "epoch": 0.054491899852724596,
+      "grad_norm": 0.6064889430999756,
+      "learning_rate": 1.8137254901960786e-06,
+      "loss": 0.6034,
+      "step": 74
+    },
+    {
+      "epoch": 0.05522827687776141,
+      "grad_norm": 0.5819352269172668,
+      "learning_rate": 1.8382352941176473e-06,
+      "loss": 0.5942,
+      "step": 75
+    },
+    {
+      "epoch": 0.055964653902798235,
+      "grad_norm": 0.47975030541419983,
+      "learning_rate": 1.8627450980392158e-06,
+      "loss": 0.5817,
+      "step": 76
+    },
+    {
+      "epoch": 0.05670103092783505,
+      "grad_norm": 0.47679486870765686,
+      "learning_rate": 1.8872549019607845e-06,
+      "loss": 0.5789,
+      "step": 77
+    },
+    {
+      "epoch": 0.05743740795287187,
+      "grad_norm": 0.44839027523994446,
+      "learning_rate": 1.9117647058823528e-06,
+      "loss": 0.595,
+      "step": 78
+    },
+    {
+      "epoch": 0.05817378497790869,
+      "grad_norm": 0.43815600872039795,
+      "learning_rate": 1.9362745098039217e-06,
+      "loss": 0.5648,
+      "step": 79
+    },
+    {
+      "epoch": 0.05891016200294551,
+      "grad_norm": 0.457387775182724,
+      "learning_rate": 1.96078431372549e-06,
+      "loss": 0.6067,
+      "step": 80
+    },
+    {
+      "epoch": 0.05964653902798233,
+      "grad_norm": 0.4818171560764313,
+      "learning_rate": 1.985294117647059e-06,
+      "loss": 0.61,
+      "step": 81
+    },
+    {
+      "epoch": 0.060382916053019146,
+      "grad_norm": 0.4079191982746124,
+      "learning_rate": 2.0098039215686276e-06,
+      "loss": 0.5663,
+      "step": 82
+    },
+    {
+      "epoch": 0.06111929307805596,
+      "grad_norm": 0.4442485272884369,
+      "learning_rate": 2.034313725490196e-06,
+      "loss": 0.5761,
+      "step": 83
+    },
+    {
+      "epoch": 0.061855670103092786,
+      "grad_norm": 0.4076555073261261,
+      "learning_rate": 2.058823529411765e-06,
+      "loss": 0.5667,
+      "step": 84
+    },
+    {
+      "epoch": 0.0625920471281296,
+      "grad_norm": 0.4406167268753052,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.586,
+      "step": 85
+    },
+    {
+      "epoch": 0.06332842415316642,
+      "grad_norm": 0.4781548082828522,
+      "learning_rate": 2.1078431372549023e-06,
+      "loss": 0.5387,
+      "step": 86
+    },
+    {
+      "epoch": 0.06406480117820323,
+      "grad_norm": 0.4489077925682068,
+      "learning_rate": 2.132352941176471e-06,
+      "loss": 0.5346,
+      "step": 87
+    },
+    {
+      "epoch": 0.06480117820324006,
+      "grad_norm": 0.47296687960624695,
+      "learning_rate": 2.1568627450980393e-06,
+      "loss": 0.611,
+      "step": 88
+    },
+    {
+      "epoch": 0.06553755522827688,
+      "grad_norm": 0.4460233151912689,
+      "learning_rate": 2.1813725490196082e-06,
+      "loss": 0.5806,
+      "step": 89
+    },
+    {
+      "epoch": 0.0662739322533137,
+      "grad_norm": 0.45212772488594055,
+      "learning_rate": 2.2058823529411767e-06,
+      "loss": 0.5761,
+      "step": 90
+    },
+    {
+      "epoch": 0.06701030927835051,
+      "grad_norm": 0.4867306053638458,
+      "learning_rate": 2.2303921568627456e-06,
+      "loss": 0.555,
+      "step": 91
+    },
+    {
+      "epoch": 0.06774668630338733,
+      "grad_norm": 0.44260451197624207,
+      "learning_rate": 2.254901960784314e-06,
+      "loss": 0.5333,
+      "step": 92
+    },
+    {
+      "epoch": 0.06848306332842416,
+      "grad_norm": 0.4509826898574829,
+      "learning_rate": 2.2794117647058826e-06,
+      "loss": 0.5817,
+      "step": 93
+    },
+    {
+      "epoch": 0.06921944035346098,
+      "grad_norm": 0.422654926776886,
+      "learning_rate": 2.303921568627451e-06,
+      "loss": 0.5797,
+      "step": 94
+    },
+    {
+      "epoch": 0.06995581737849779,
+      "grad_norm": 0.40386420488357544,
+      "learning_rate": 2.32843137254902e-06,
+      "loss": 0.5972,
+      "step": 95
+    },
+    {
+      "epoch": 0.07069219440353461,
+      "grad_norm": 0.4382508099079132,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.588,
+      "step": 96
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 0.4391435980796814,
+      "learning_rate": 2.377450980392157e-06,
+      "loss": 0.5434,
+      "step": 97
+    },
+    {
+      "epoch": 0.07216494845360824,
+      "grad_norm": 0.39785510301589966,
+      "learning_rate": 2.401960784313726e-06,
+      "loss": 0.55,
+      "step": 98
+    },
+    {
+      "epoch": 0.07290132547864507,
+      "grad_norm": 0.4807189404964447,
+      "learning_rate": 2.4264705882352943e-06,
+      "loss": 0.5642,
+      "step": 99
+    },
+    {
+      "epoch": 0.07363770250368189,
+      "grad_norm": 0.40981948375701904,
+      "learning_rate": 2.450980392156863e-06,
+      "loss": 0.5257,
+      "step": 100
+    },
+    {
+      "epoch": 0.0743740795287187,
+      "grad_norm": 0.4581270217895508,
+      "learning_rate": 2.4754901960784317e-06,
+      "loss": 0.5676,
+      "step": 101
+    },
+    {
+      "epoch": 0.07511045655375552,
+      "grad_norm": 0.4791598320007324,
+      "learning_rate": 2.5e-06,
+      "loss": 0.5589,
+      "step": 102
+    },
+    {
+      "epoch": 0.07584683357879234,
+      "grad_norm": 0.4747304916381836,
+      "learning_rate": 2.5245098039215687e-06,
+      "loss": 0.573,
+      "step": 103
+    },
+    {
+      "epoch": 0.07658321060382917,
+      "grad_norm": 0.501209557056427,
+      "learning_rate": 2.549019607843137e-06,
+      "loss": 0.5354,
+      "step": 104
+    },
+    {
+      "epoch": 0.07731958762886598,
+      "grad_norm": 0.42233791947364807,
+      "learning_rate": 2.5735294117647057e-06,
+      "loss": 0.5822,
+      "step": 105
+    },
+    {
+      "epoch": 0.0780559646539028,
+      "grad_norm": 0.44912800192832947,
+      "learning_rate": 2.598039215686275e-06,
+      "loss": 0.5279,
+      "step": 106
+    },
+    {
+      "epoch": 0.07879234167893961,
+      "grad_norm": 0.47332027554512024,
+      "learning_rate": 2.6225490196078435e-06,
+      "loss": 0.5641,
+      "step": 107
+    },
+    {
+      "epoch": 0.07952871870397643,
+      "grad_norm": 0.44078969955444336,
+      "learning_rate": 2.647058823529412e-06,
+      "loss": 0.555,
+      "step": 108
+    },
+    {
+      "epoch": 0.08026509572901326,
+      "grad_norm": 0.4458158016204834,
+      "learning_rate": 2.6715686274509804e-06,
+      "loss": 0.5706,
+      "step": 109
+    },
+    {
+      "epoch": 0.08100147275405008,
+      "grad_norm": 0.4494592249393463,
+      "learning_rate": 2.696078431372549e-06,
+      "loss": 0.5742,
+      "step": 110
+    },
+    {
+      "epoch": 0.0817378497790869,
+      "grad_norm": 0.47519755363464355,
+      "learning_rate": 2.720588235294118e-06,
+      "loss": 0.5553,
+      "step": 111
+    },
+    {
+      "epoch": 0.08247422680412371,
+      "grad_norm": 0.4567236006259918,
+      "learning_rate": 2.7450980392156867e-06,
+      "loss": 0.5677,
+      "step": 112
+    },
+    {
+      "epoch": 0.08321060382916053,
+      "grad_norm": 0.45444926619529724,
+      "learning_rate": 2.7696078431372552e-06,
+      "loss": 0.537,
+      "step": 113
+    },
+    {
+      "epoch": 0.08394698085419734,
+      "grad_norm": 0.5289820432662964,
+      "learning_rate": 2.7941176470588237e-06,
+      "loss": 0.5701,
+      "step": 114
+    },
+    {
+      "epoch": 0.08468335787923417,
+      "grad_norm": 0.5139638781547546,
+      "learning_rate": 2.818627450980392e-06,
+      "loss": 0.5588,
+      "step": 115
+    },
+    {
+      "epoch": 0.08541973490427099,
+      "grad_norm": 0.42048802971839905,
+      "learning_rate": 2.843137254901961e-06,
+      "loss": 0.5342,
+      "step": 116
+    },
+    {
+      "epoch": 0.0861561119293078,
+      "grad_norm": 0.43646562099456787,
+      "learning_rate": 2.8676470588235296e-06,
+      "loss": 0.5506,
+      "step": 117
+    },
+    {
+      "epoch": 0.08689248895434462,
+      "grad_norm": 0.4417460560798645,
+      "learning_rate": 2.8921568627450985e-06,
+      "loss": 0.5493,
+      "step": 118
+    },
+    {
+      "epoch": 0.08762886597938144,
+      "grad_norm": 0.46676769852638245,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 0.5831,
+      "step": 119
+    },
+    {
+      "epoch": 0.08836524300441827,
+      "grad_norm": 0.45995548367500305,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.5807,
+      "step": 120
+    },
+    {
+      "epoch": 0.08910162002945508,
+      "grad_norm": 0.42616021633148193,
+      "learning_rate": 2.9656862745098044e-06,
+      "loss": 0.5562,
+      "step": 121
+    },
+    {
+      "epoch": 0.0898379970544919,
+      "grad_norm": 0.419491708278656,
+      "learning_rate": 2.990196078431373e-06,
+      "loss": 0.5285,
+      "step": 122
+    },
+    {
+      "epoch": 0.09057437407952872,
+      "grad_norm": 0.4764864146709442,
+      "learning_rate": 3.0147058823529413e-06,
+      "loss": 0.5761,
+      "step": 123
+    },
+    {
+      "epoch": 0.09131075110456553,
+      "grad_norm": 0.4339980185031891,
+      "learning_rate": 3.03921568627451e-06,
+      "loss": 0.5707,
+      "step": 124
+    },
+    {
+      "epoch": 0.09204712812960236,
+      "grad_norm": 0.41494664549827576,
+      "learning_rate": 3.0637254901960787e-06,
+      "loss": 0.5294,
+      "step": 125
+    },
+    {
+      "epoch": 0.09278350515463918,
+      "grad_norm": 0.43163448572158813,
+      "learning_rate": 3.0882352941176476e-06,
+      "loss": 0.5539,
+      "step": 126
+    },
+    {
+      "epoch": 0.093519882179676,
+      "grad_norm": 0.42778855562210083,
+      "learning_rate": 3.112745098039216e-06,
+      "loss": 0.5625,
+      "step": 127
+    },
+    {
+      "epoch": 0.09425625920471281,
+      "grad_norm": 0.4978928565979004,
+      "learning_rate": 3.1372549019607846e-06,
+      "loss": 0.5107,
+      "step": 128
+    },
+    {
+      "epoch": 0.09499263622974963,
+      "grad_norm": 0.4967725872993469,
+      "learning_rate": 3.161764705882353e-06,
+      "loss": 0.5434,
+      "step": 129
+    },
+    {
+      "epoch": 0.09572901325478646,
+      "grad_norm": 0.4463275372982025,
+      "learning_rate": 3.1862745098039216e-06,
+      "loss": 0.553,
+      "step": 130
+    },
+    {
+      "epoch": 0.09646539027982327,
+      "grad_norm": 0.5262526869773865,
+      "learning_rate": 3.210784313725491e-06,
+      "loss": 0.5464,
+      "step": 131
+    },
+    {
+      "epoch": 0.09720176730486009,
+      "grad_norm": 0.4471772015094757,
+      "learning_rate": 3.2352941176470594e-06,
+      "loss": 0.5089,
+      "step": 132
+    },
+    {
+      "epoch": 0.0979381443298969,
+      "grad_norm": 0.41978949308395386,
+      "learning_rate": 3.259803921568628e-06,
+      "loss": 0.4965,
+      "step": 133
+    },
+    {
+      "epoch": 0.09867452135493372,
+      "grad_norm": 0.4757583439350128,
+      "learning_rate": 3.2843137254901964e-06,
+      "loss": 0.5509,
+      "step": 134
+    },
+    {
+      "epoch": 0.09941089837997054,
+      "grad_norm": 0.4464501738548279,
+      "learning_rate": 3.308823529411765e-06,
+      "loss": 0.5032,
+      "step": 135
+    },
+    {
+      "epoch": 0.10014727540500737,
+      "grad_norm": 0.46477508544921875,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.5429,
+      "step": 136
+    },
+    {
+      "epoch": 0.10088365243004419,
+      "grad_norm": 0.426889032125473,
+      "learning_rate": 3.357843137254902e-06,
+      "loss": 0.5572,
+      "step": 137
+    },
+    {
+      "epoch": 0.101620029455081,
+      "grad_norm": 0.42896464467048645,
+      "learning_rate": 3.382352941176471e-06,
+      "loss": 0.541,
+      "step": 138
+    },
+    {
+      "epoch": 0.10235640648011782,
+      "grad_norm": 0.4436034560203552,
+      "learning_rate": 3.4068627450980396e-06,
+      "loss": 0.5648,
+      "step": 139
+    },
+    {
+      "epoch": 0.10309278350515463,
+      "grad_norm": 0.4158652722835541,
+      "learning_rate": 3.431372549019608e-06,
+      "loss": 0.5607,
+      "step": 140
+    },
+    {
+      "epoch": 0.10382916053019146,
+      "grad_norm": 0.4582897424697876,
+      "learning_rate": 3.4558823529411766e-06,
+      "loss": 0.5745,
+      "step": 141
+    },
+    {
+      "epoch": 0.10456553755522828,
+      "grad_norm": 0.4275548756122589,
+      "learning_rate": 3.480392156862745e-06,
+      "loss": 0.5147,
+      "step": 142
+    },
+    {
+      "epoch": 0.1053019145802651,
+      "grad_norm": 0.4628273546695709,
+      "learning_rate": 3.504901960784314e-06,
+      "loss": 0.5305,
+      "step": 143
+    },
+    {
+      "epoch": 0.10603829160530191,
+      "grad_norm": 0.4495219588279724,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 0.563,
+      "step": 144
+    },
+    {
+      "epoch": 0.10677466863033873,
+      "grad_norm": 0.48380184173583984,
+      "learning_rate": 3.5539215686274514e-06,
+      "loss": 0.562,
+      "step": 145
+    },
+    {
+      "epoch": 0.10751104565537556,
+      "grad_norm": 0.4028421938419342,
+      "learning_rate": 3.57843137254902e-06,
+      "loss": 0.5137,
+      "step": 146
+    },
+    {
+      "epoch": 0.10824742268041238,
+      "grad_norm": 0.4970363974571228,
+      "learning_rate": 3.6029411764705883e-06,
+      "loss": 0.5268,
+      "step": 147
+    },
+    {
+      "epoch": 0.10898379970544919,
+      "grad_norm": 0.46763232350349426,
+      "learning_rate": 3.6274509803921573e-06,
+      "loss": 0.542,
+      "step": 148
+    },
+    {
+      "epoch": 0.10972017673048601,
+      "grad_norm": 0.457398921251297,
+      "learning_rate": 3.6519607843137257e-06,
+      "loss": 0.5289,
+      "step": 149
+    },
+    {
+      "epoch": 0.11045655375552282,
+      "grad_norm": 0.44986864924430847,
+      "learning_rate": 3.6764705882352946e-06,
+      "loss": 0.549,
+      "step": 150
+    },
+    {
+      "epoch": 0.11119293078055964,
+      "grad_norm": 0.46221205592155457,
+      "learning_rate": 3.700980392156863e-06,
+      "loss": 0.5005,
+      "step": 151
+    },
+    {
+      "epoch": 0.11192930780559647,
+      "grad_norm": 0.5250900983810425,
+      "learning_rate": 3.7254901960784316e-06,
+      "loss": 0.5468,
+      "step": 152
+    },
+    {
+      "epoch": 0.11266568483063329,
+      "grad_norm": 0.434471994638443,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.5307,
+      "step": 153
+    },
+    {
+      "epoch": 0.1134020618556701,
+      "grad_norm": 0.5031168460845947,
+      "learning_rate": 3.774509803921569e-06,
+      "loss": 0.5383,
+      "step": 154
+    },
+    {
+      "epoch": 0.11413843888070692,
+      "grad_norm": 0.4508562684059143,
+      "learning_rate": 3.7990196078431375e-06,
+      "loss": 0.5359,
+      "step": 155
+    },
+    {
+      "epoch": 0.11487481590574374,
+      "grad_norm": 0.4205038845539093,
+      "learning_rate": 3.8235294117647055e-06,
+      "loss": 0.5315,
+      "step": 156
+    },
+    {
+      "epoch": 0.11561119293078057,
+      "grad_norm": 0.40232402086257935,
+      "learning_rate": 3.848039215686275e-06,
+      "loss": 0.5075,
+      "step": 157
+    },
+    {
+      "epoch": 0.11634756995581738,
+      "grad_norm": 0.45643943548202515,
+      "learning_rate": 3.872549019607843e-06,
+      "loss": 0.5481,
+      "step": 158
+    },
+    {
+      "epoch": 0.1170839469808542,
+      "grad_norm": 0.4849976599216461,
+      "learning_rate": 3.897058823529412e-06,
+      "loss": 0.5068,
+      "step": 159
+    },
+    {
+      "epoch": 0.11782032400589101,
+      "grad_norm": 0.4546898901462555,
+      "learning_rate": 3.92156862745098e-06,
+      "loss": 0.5489,
+      "step": 160
+    },
+    {
+      "epoch": 0.11855670103092783,
+      "grad_norm": 0.4682084023952484,
+      "learning_rate": 3.946078431372549e-06,
+      "loss": 0.5448,
+      "step": 161
+    },
+    {
+      "epoch": 0.11929307805596466,
+      "grad_norm": 0.4213881492614746,
+      "learning_rate": 3.970588235294118e-06,
+      "loss": 0.4813,
+      "step": 162
+    },
+    {
+      "epoch": 0.12002945508100148,
+      "grad_norm": 0.4240805208683014,
+      "learning_rate": 3.995098039215687e-06,
+      "loss": 0.5164,
+      "step": 163
+    },
+    {
+      "epoch": 0.12076583210603829,
+      "grad_norm": 0.5331303477287292,
+      "learning_rate": 4.019607843137255e-06,
+      "loss": 0.5201,
+      "step": 164
+    },
+    {
+      "epoch": 0.12150220913107511,
+      "grad_norm": 0.4699327349662781,
+      "learning_rate": 4.044117647058824e-06,
+      "loss": 0.527,
+      "step": 165
+    },
+    {
+      "epoch": 0.12223858615611193,
+      "grad_norm": 0.44539308547973633,
+      "learning_rate": 4.068627450980392e-06,
+      "loss": 0.4908,
+      "step": 166
+    },
+    {
+      "epoch": 0.12297496318114874,
+      "grad_norm": 0.4506961703300476,
+      "learning_rate": 4.093137254901961e-06,
+      "loss": 0.5314,
+      "step": 167
+    },
+    {
+      "epoch": 0.12371134020618557,
+      "grad_norm": 0.44389674067497253,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.506,
+      "step": 168
+    },
+    {
+      "epoch": 0.12444771723122239,
+      "grad_norm": 0.445320725440979,
+      "learning_rate": 4.142156862745099e-06,
+      "loss": 0.5236,
+      "step": 169
+    },
+    {
+      "epoch": 0.1251840942562592,
+      "grad_norm": 0.44341355562210083,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.571,
+      "step": 170
+    },
+    {
+      "epoch": 0.12592047128129602,
+      "grad_norm": 0.42870596051216125,
+      "learning_rate": 4.191176470588236e-06,
+      "loss": 0.5204,
+      "step": 171
+    },
+    {
+      "epoch": 0.12665684830633284,
+      "grad_norm": 0.4404330849647522,
+      "learning_rate": 4.215686274509805e-06,
+      "loss": 0.5623,
+      "step": 172
+    },
+    {
+      "epoch": 0.12739322533136965,
+      "grad_norm": 0.542121946811676,
+      "learning_rate": 4.240196078431373e-06,
+      "loss": 0.5301,
+      "step": 173
+    },
+    {
+      "epoch": 0.12812960235640647,
+      "grad_norm": 0.5006332397460938,
+      "learning_rate": 4.264705882352942e-06,
+      "loss": 0.5237,
+      "step": 174
+    },
+    {
+      "epoch": 0.12886597938144329,
+      "grad_norm": 0.4504947066307068,
+      "learning_rate": 4.28921568627451e-06,
+      "loss": 0.5052,
+      "step": 175
+    },
+    {
+      "epoch": 0.12960235640648013,
+      "grad_norm": 0.4694218635559082,
+      "learning_rate": 4.313725490196079e-06,
+      "loss": 0.5012,
+      "step": 176
+    },
+    {
+      "epoch": 0.13033873343151695,
+      "grad_norm": 0.47194620966911316,
+      "learning_rate": 4.3382352941176475e-06,
+      "loss": 0.5049,
+      "step": 177
+    },
+    {
+      "epoch": 0.13107511045655376,
+      "grad_norm": 0.5346331596374512,
+      "learning_rate": 4.3627450980392164e-06,
+      "loss": 0.5118,
+      "step": 178
+    },
+    {
+      "epoch": 0.13181148748159058,
+      "grad_norm": 0.5172557234764099,
+      "learning_rate": 4.3872549019607845e-06,
+      "loss": 0.4994,
+      "step": 179
+    },
+    {
+      "epoch": 0.1325478645066274,
+      "grad_norm": 0.4675532579421997,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 0.5519,
+      "step": 180
+    },
+    {
+      "epoch": 0.1332842415316642,
+      "grad_norm": 0.4985281527042389,
+      "learning_rate": 4.4362745098039215e-06,
+      "loss": 0.5058,
+      "step": 181
+    },
+    {
+      "epoch": 0.13402061855670103,
+      "grad_norm": 0.46521127223968506,
+      "learning_rate": 4.460784313725491e-06,
+      "loss": 0.5559,
+      "step": 182
+    },
+    {
+      "epoch": 0.13475699558173784,
+      "grad_norm": 0.48757219314575195,
+      "learning_rate": 4.485294117647059e-06,
+      "loss": 0.5249,
+      "step": 183
+    },
+    {
+      "epoch": 0.13549337260677466,
+      "grad_norm": 0.5091550946235657,
+      "learning_rate": 4.509803921568628e-06,
+      "loss": 0.5223,
+      "step": 184
+    },
+    {
+      "epoch": 0.13622974963181148,
+      "grad_norm": 0.4934118688106537,
+      "learning_rate": 4.534313725490196e-06,
+      "loss": 0.5419,
+      "step": 185
+    },
+    {
+      "epoch": 0.13696612665684832,
+      "grad_norm": 0.5085634589195251,
+      "learning_rate": 4.558823529411765e-06,
+      "loss": 0.516,
+      "step": 186
+    },
+    {
+      "epoch": 0.13770250368188514,
+      "grad_norm": 0.46844273805618286,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 0.5093,
+      "step": 187
+    },
+    {
+      "epoch": 0.13843888070692195,
+      "grad_norm": 0.45088204741477966,
+      "learning_rate": 4.607843137254902e-06,
+      "loss": 0.5236,
+      "step": 188
+    },
+    {
+      "epoch": 0.13917525773195877,
+      "grad_norm": 0.4446581304073334,
+      "learning_rate": 4.632352941176471e-06,
+      "loss": 0.5117,
+      "step": 189
+    },
+    {
+      "epoch": 0.13991163475699558,
+      "grad_norm": 0.47169244289398193,
+      "learning_rate": 4.65686274509804e-06,
+      "loss": 0.5264,
+      "step": 190
+    },
+    {
+      "epoch": 0.1406480117820324,
+      "grad_norm": 0.4336180090904236,
+      "learning_rate": 4.681372549019608e-06,
+      "loss": 0.527,
+      "step": 191
+    },
+    {
+      "epoch": 0.14138438880706922,
+      "grad_norm": 0.4745936989784241,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.5235,
+      "step": 192
+    },
+    {
+      "epoch": 0.14212076583210603,
+      "grad_norm": 0.49792900681495667,
+      "learning_rate": 4.730392156862745e-06,
+      "loss": 0.5006,
+      "step": 193
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.503589391708374,
+      "learning_rate": 4.754901960784314e-06,
+      "loss": 0.5314,
+      "step": 194
+    },
+    {
+      "epoch": 0.14359351988217967,
+      "grad_norm": 0.4816850423812866,
+      "learning_rate": 4.779411764705883e-06,
+      "loss": 0.5176,
+      "step": 195
+    },
+    {
+      "epoch": 0.14432989690721648,
+      "grad_norm": 0.5058243870735168,
+      "learning_rate": 4.803921568627452e-06,
+      "loss": 0.4917,
+      "step": 196
+    },
+    {
+      "epoch": 0.14506627393225333,
+      "grad_norm": 0.4948558509349823,
+      "learning_rate": 4.82843137254902e-06,
+      "loss": 0.5227,
+      "step": 197
+    },
+    {
+      "epoch": 0.14580265095729014,
+      "grad_norm": 0.43871647119522095,
+      "learning_rate": 4.852941176470589e-06,
+      "loss": 0.5413,
+      "step": 198
+    },
+    {
+      "epoch": 0.14653902798232696,
+      "grad_norm": 0.45166996121406555,
+      "learning_rate": 4.8774509803921576e-06,
+      "loss": 0.5645,
+      "step": 199
+    },
+    {
+      "epoch": 0.14727540500736377,
+      "grad_norm": 0.5076332688331604,
+      "learning_rate": 4.901960784313726e-06,
+      "loss": 0.5222,
+      "step": 200
+    },
+    {
+      "epoch": 0.1480117820324006,
+      "grad_norm": 0.46355342864990234,
+      "learning_rate": 4.9264705882352945e-06,
+      "loss": 0.4985,
+      "step": 201
+    },
+    {
+      "epoch": 0.1487481590574374,
+      "grad_norm": 0.473640114068985,
+      "learning_rate": 4.9509803921568634e-06,
+      "loss": 0.5339,
+      "step": 202
+    },
+    {
+      "epoch": 0.14948453608247422,
+      "grad_norm": 0.43335700035095215,
+      "learning_rate": 4.9754901960784315e-06,
+      "loss": 0.5236,
+      "step": 203
+    },
+    {
+      "epoch": 0.15022091310751104,
+      "grad_norm": 0.5285197496414185,
+      "learning_rate": 5e-06,
+      "loss": 0.5102,
+      "step": 204
+    },
+    {
+      "epoch": 0.15095729013254786,
+      "grad_norm": 0.4338807463645935,
+      "learning_rate": 5.024509803921569e-06,
+      "loss": 0.5243,
+      "step": 205
+    },
+    {
+      "epoch": 0.15169366715758467,
+      "grad_norm": 0.5603867769241333,
+      "learning_rate": 5.049019607843137e-06,
+      "loss": 0.5269,
+      "step": 206
+    },
+    {
+      "epoch": 0.15243004418262152,
+      "grad_norm": 0.547127366065979,
+      "learning_rate": 5.073529411764706e-06,
+      "loss": 0.5376,
+      "step": 207
+    },
+    {
+      "epoch": 0.15316642120765833,
+      "grad_norm": 0.44914114475250244,
+      "learning_rate": 5.098039215686274e-06,
+      "loss": 0.4983,
+      "step": 208
+    },
+    {
+      "epoch": 0.15390279823269515,
+      "grad_norm": 0.48467424511909485,
+      "learning_rate": 5.122549019607843e-06,
+      "loss": 0.4931,
+      "step": 209
+    },
+    {
+      "epoch": 0.15463917525773196,
+      "grad_norm": 0.445891797542572,
+      "learning_rate": 5.147058823529411e-06,
+      "loss": 0.5086,
+      "step": 210
+    },
+    {
+      "epoch": 0.15537555228276878,
+      "grad_norm": 0.44430914521217346,
+      "learning_rate": 5.171568627450981e-06,
+      "loss": 0.5023,
+      "step": 211
+    },
+    {
+      "epoch": 0.1561119293078056,
+      "grad_norm": 0.5020100474357605,
+      "learning_rate": 5.19607843137255e-06,
+      "loss": 0.492,
+      "step": 212
+    },
+    {
+      "epoch": 0.1568483063328424,
+      "grad_norm": 0.4869686961174011,
+      "learning_rate": 5.220588235294118e-06,
+      "loss": 0.4876,
+      "step": 213
+    },
+    {
+      "epoch": 0.15758468335787923,
+      "grad_norm": 0.48111408948898315,
+      "learning_rate": 5.245098039215687e-06,
+      "loss": 0.5117,
+      "step": 214
+    },
+    {
+      "epoch": 0.15832106038291605,
+      "grad_norm": 0.5187819004058838,
+      "learning_rate": 5.269607843137256e-06,
+      "loss": 0.5188,
+      "step": 215
+    },
+    {
+      "epoch": 0.15905743740795286,
+      "grad_norm": 0.510890007019043,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 0.545,
+      "step": 216
+    },
+    {
+      "epoch": 0.15979381443298968,
+      "grad_norm": 0.5489839315414429,
+      "learning_rate": 5.318627450980393e-06,
+      "loss": 0.508,
+      "step": 217
+    },
+    {
+      "epoch": 0.16053019145802652,
+      "grad_norm": 0.4811173677444458,
+      "learning_rate": 5.343137254901961e-06,
+      "loss": 0.4955,
+      "step": 218
+    },
+    {
+      "epoch": 0.16126656848306334,
+      "grad_norm": 0.49445685744285583,
+      "learning_rate": 5.36764705882353e-06,
+      "loss": 0.4678,
+      "step": 219
+    },
+    {
+      "epoch": 0.16200294550810015,
+      "grad_norm": 0.44502392411231995,
+      "learning_rate": 5.392156862745098e-06,
+      "loss": 0.5166,
+      "step": 220
+    },
+    {
+      "epoch": 0.16273932253313697,
+      "grad_norm": 0.46069005131721497,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 0.5279,
+      "step": 221
+    },
+    {
+      "epoch": 0.1634756995581738,
+      "grad_norm": 0.4491000175476074,
+      "learning_rate": 5.441176470588236e-06,
+      "loss": 0.5093,
+      "step": 222
+    },
+    {
+      "epoch": 0.1642120765832106,
+      "grad_norm": 0.5266180634498596,
+      "learning_rate": 5.465686274509804e-06,
+      "loss": 0.5221,
+      "step": 223
+    },
+    {
+      "epoch": 0.16494845360824742,
+      "grad_norm": 0.48982885479927063,
+      "learning_rate": 5.4901960784313735e-06,
+      "loss": 0.4804,
+      "step": 224
+    },
+    {
+      "epoch": 0.16568483063328424,
+      "grad_norm": 0.46362483501434326,
+      "learning_rate": 5.514705882352942e-06,
+      "loss": 0.524,
+      "step": 225
+    },
+    {
+      "epoch": 0.16642120765832105,
+      "grad_norm": 0.480934202671051,
+      "learning_rate": 5.5392156862745104e-06,
+      "loss": 0.5063,
+      "step": 226
+    },
+    {
+      "epoch": 0.16715758468335787,
+      "grad_norm": 0.4418233633041382,
+      "learning_rate": 5.563725490196079e-06,
+      "loss": 0.485,
+      "step": 227
+    },
+    {
+      "epoch": 0.16789396170839468,
+      "grad_norm": 0.4766086935997009,
+      "learning_rate": 5.588235294117647e-06,
+      "loss": 0.4859,
+      "step": 228
+    },
+    {
+      "epoch": 0.16863033873343153,
+      "grad_norm": 0.5201296806335449,
+      "learning_rate": 5.612745098039216e-06,
+      "loss": 0.5128,
+      "step": 229
+    },
+    {
+      "epoch": 0.16936671575846834,
+      "grad_norm": 0.4526103734970093,
+      "learning_rate": 5.637254901960784e-06,
+      "loss": 0.5115,
+      "step": 230
+    },
+    {
+      "epoch": 0.17010309278350516,
+      "grad_norm": 0.4606441557407379,
+      "learning_rate": 5.661764705882353e-06,
+      "loss": 0.5074,
+      "step": 231
+    },
+    {
+      "epoch": 0.17083946980854198,
+      "grad_norm": 0.4658742845058441,
+      "learning_rate": 5.686274509803922e-06,
+      "loss": 0.5212,
+      "step": 232
+    },
+    {
+      "epoch": 0.1715758468335788,
+      "grad_norm": 0.45116207003593445,
+      "learning_rate": 5.71078431372549e-06,
+      "loss": 0.5178,
+      "step": 233
+    },
+    {
+      "epoch": 0.1723122238586156,
+      "grad_norm": 0.40379467606544495,
+      "learning_rate": 5.735294117647059e-06,
+      "loss": 0.4907,
+      "step": 234
+    },
+    {
+      "epoch": 0.17304860088365243,
+      "grad_norm": 0.461535781621933,
+      "learning_rate": 5.759803921568627e-06,
+      "loss": 0.4812,
+      "step": 235
+    },
+    {
+      "epoch": 0.17378497790868924,
+      "grad_norm": 0.40763628482818604,
+      "learning_rate": 5.784313725490197e-06,
+      "loss": 0.4861,
+      "step": 236
+    },
+    {
+      "epoch": 0.17452135493372606,
+      "grad_norm": 0.4869963228702545,
+      "learning_rate": 5.808823529411766e-06,
+      "loss": 0.5126,
+      "step": 237
+    },
+    {
+      "epoch": 0.17525773195876287,
+      "grad_norm": 0.43191322684288025,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.5194,
+      "step": 238
+    },
+    {
+      "epoch": 0.17599410898379972,
+      "grad_norm": 0.4565574824810028,
+      "learning_rate": 5.857843137254903e-06,
+      "loss": 0.5108,
+      "step": 239
+    },
+    {
+      "epoch": 0.17673048600883653,
+      "grad_norm": 0.4109247326850891,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.4885,
+      "step": 240
+    },
+    {
+      "epoch": 0.17746686303387335,
+      "grad_norm": 0.5282112956047058,
+      "learning_rate": 5.90686274509804e-06,
+      "loss": 0.538,
+      "step": 241
+    },
+    {
+      "epoch": 0.17820324005891017,
+      "grad_norm": 0.4757918417453766,
+      "learning_rate": 5.931372549019609e-06,
+      "loss": 0.5143,
+      "step": 242
+    },
+    {
+      "epoch": 0.17893961708394698,
+      "grad_norm": 0.4788837134838104,
+      "learning_rate": 5.955882352941177e-06,
+      "loss": 0.4996,
+      "step": 243
+    },
+    {
+      "epoch": 0.1796759941089838,
+      "grad_norm": 0.4603605568408966,
+      "learning_rate": 5.980392156862746e-06,
+      "loss": 0.4856,
+      "step": 244
+    },
+    {
+      "epoch": 0.18041237113402062,
+      "grad_norm": 0.48925524950027466,
+      "learning_rate": 6.004901960784314e-06,
+      "loss": 0.5062,
+      "step": 245
+    },
+    {
+      "epoch": 0.18114874815905743,
+      "grad_norm": 0.4555787742137909,
+      "learning_rate": 6.029411764705883e-06,
+      "loss": 0.4869,
+      "step": 246
+    },
+    {
+      "epoch": 0.18188512518409425,
+      "grad_norm": 0.4337728023529053,
+      "learning_rate": 6.053921568627451e-06,
+      "loss": 0.5095,
+      "step": 247
+    },
+    {
+      "epoch": 0.18262150220913106,
+      "grad_norm": 0.4596711993217468,
+      "learning_rate": 6.07843137254902e-06,
+      "loss": 0.5334,
+      "step": 248
+    },
+    {
+      "epoch": 0.18335787923416788,
+      "grad_norm": 0.55954509973526,
+      "learning_rate": 6.102941176470589e-06,
+      "loss": 0.542,
+      "step": 249
+    },
+    {
+      "epoch": 0.18409425625920472,
+      "grad_norm": 0.4391202926635742,
+      "learning_rate": 6.1274509803921575e-06,
+      "loss": 0.5085,
+      "step": 250
+    },
+    {
+      "epoch": 0.18483063328424154,
+      "grad_norm": 0.5165490508079529,
+      "learning_rate": 6.151960784313726e-06,
+      "loss": 0.5267,
+      "step": 251
+    },
+    {
+      "epoch": 0.18556701030927836,
+      "grad_norm": 0.5765572786331177,
+      "learning_rate": 6.176470588235295e-06,
+      "loss": 0.5302,
+      "step": 252
+    },
+    {
+      "epoch": 0.18630338733431517,
+      "grad_norm": 0.4425363540649414,
+      "learning_rate": 6.200980392156863e-06,
+      "loss": 0.5294,
+      "step": 253
+    },
+    {
+      "epoch": 0.187039764359352,
+      "grad_norm": 0.5258336663246155,
+      "learning_rate": 6.225490196078432e-06,
+      "loss": 0.4794,
+      "step": 254
+    },
+    {
+      "epoch": 0.1877761413843888,
+      "grad_norm": 0.4908381998538971,
+      "learning_rate": 6.25e-06,
+      "loss": 0.4942,
+      "step": 255
+    },
+    {
+      "epoch": 0.18851251840942562,
+      "grad_norm": 0.480133980512619,
+      "learning_rate": 6.274509803921569e-06,
+      "loss": 0.4991,
+      "step": 256
+    },
+    {
+      "epoch": 0.18924889543446244,
+      "grad_norm": 0.48387065529823303,
+      "learning_rate": 6.299019607843137e-06,
+      "loss": 0.513,
+      "step": 257
+    },
+    {
+      "epoch": 0.18998527245949925,
+      "grad_norm": 0.5518890619277954,
+      "learning_rate": 6.323529411764706e-06,
+      "loss": 0.522,
+      "step": 258
+    },
+    {
+      "epoch": 0.19072164948453607,
+      "grad_norm": 0.5575649738311768,
+      "learning_rate": 6.348039215686275e-06,
+      "loss": 0.5046,
+      "step": 259
+    },
+    {
+      "epoch": 0.19145802650957292,
+      "grad_norm": 0.5497487187385559,
+      "learning_rate": 6.372549019607843e-06,
+      "loss": 0.517,
+      "step": 260
+    },
+    {
+      "epoch": 0.19219440353460973,
+      "grad_norm": 0.613700807094574,
+      "learning_rate": 6.397058823529412e-06,
+      "loss": 0.4874,
+      "step": 261
+    },
+    {
+      "epoch": 0.19293078055964655,
+      "grad_norm": 0.5236343741416931,
+      "learning_rate": 6.421568627450982e-06,
+      "loss": 0.4959,
+      "step": 262
+    },
+    {
+      "epoch": 0.19366715758468336,
+      "grad_norm": 0.5656614303588867,
+      "learning_rate": 6.44607843137255e-06,
+      "loss": 0.5192,
+      "step": 263
+    },
+    {
+      "epoch": 0.19440353460972018,
+      "grad_norm": 0.4702565371990204,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 0.4974,
+      "step": 264
+    },
+    {
+      "epoch": 0.195139911634757,
+      "grad_norm": 0.5624386668205261,
+      "learning_rate": 6.495098039215687e-06,
+      "loss": 0.5131,
+      "step": 265
+    },
+    {
+      "epoch": 0.1958762886597938,
+      "grad_norm": 0.5100853443145752,
+      "learning_rate": 6.519607843137256e-06,
+      "loss": 0.4945,
+      "step": 266
+    },
+    {
+      "epoch": 0.19661266568483063,
+      "grad_norm": 0.48825833201408386,
+      "learning_rate": 6.544117647058824e-06,
+      "loss": 0.4909,
+      "step": 267
+    },
+    {
+      "epoch": 0.19734904270986744,
+      "grad_norm": 0.4945085942745209,
+      "learning_rate": 6.568627450980393e-06,
+      "loss": 0.5235,
+      "step": 268
+    },
+    {
+      "epoch": 0.19808541973490426,
+      "grad_norm": 0.46311086416244507,
+      "learning_rate": 6.593137254901962e-06,
+      "loss": 0.4975,
+      "step": 269
+    },
+    {
+      "epoch": 0.19882179675994108,
+      "grad_norm": 0.4879513680934906,
+      "learning_rate": 6.61764705882353e-06,
+      "loss": 0.494,
+      "step": 270
+    },
+    {
+      "epoch": 0.19955817378497792,
+      "grad_norm": 0.43951448798179626,
+      "learning_rate": 6.642156862745099e-06,
+      "loss": 0.5052,
+      "step": 271
+    },
+    {
+      "epoch": 0.20029455081001474,
+      "grad_norm": 0.4860183596611023,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.506,
+      "step": 272
+    },
+    {
+      "epoch": 0.20103092783505155,
+      "grad_norm": 0.5143676400184631,
+      "learning_rate": 6.6911764705882356e-06,
+      "loss": 0.5197,
+      "step": 273
+    },
+    {
+      "epoch": 0.20176730486008837,
+      "grad_norm": 0.4551714062690735,
+      "learning_rate": 6.715686274509804e-06,
+      "loss": 0.522,
+      "step": 274
+    },
+    {
+      "epoch": 0.2025036818851252,
+      "grad_norm": 0.49370667338371277,
+      "learning_rate": 6.740196078431373e-06,
+      "loss": 0.4651,
+      "step": 275
+    },
+    {
+      "epoch": 0.203240058910162,
+      "grad_norm": 0.42197269201278687,
+      "learning_rate": 6.764705882352942e-06,
+      "loss": 0.4675,
+      "step": 276
+    },
+    {
+      "epoch": 0.20397643593519882,
+      "grad_norm": 0.48403897881507874,
+      "learning_rate": 6.78921568627451e-06,
+      "loss": 0.4831,
+      "step": 277
+    },
+    {
+      "epoch": 0.20471281296023564,
+      "grad_norm": 0.5002971291542053,
+      "learning_rate": 6.813725490196079e-06,
+      "loss": 0.502,
+      "step": 278
+    },
+    {
+      "epoch": 0.20544918998527245,
+      "grad_norm": 0.49652308225631714,
+      "learning_rate": 6.838235294117648e-06,
+      "loss": 0.5341,
+      "step": 279
+    },
+    {
+      "epoch": 0.20618556701030927,
+      "grad_norm": 0.4934190809726715,
+      "learning_rate": 6.862745098039216e-06,
+      "loss": 0.5091,
+      "step": 280
+    },
+    {
+      "epoch": 0.20692194403534608,
+      "grad_norm": 0.5390608310699463,
+      "learning_rate": 6.887254901960785e-06,
+      "loss": 0.4987,
+      "step": 281
+    },
+    {
+      "epoch": 0.20765832106038293,
+      "grad_norm": 0.5566352605819702,
+      "learning_rate": 6.911764705882353e-06,
+      "loss": 0.5024,
+      "step": 282
+    },
+    {
+      "epoch": 0.20839469808541974,
+      "grad_norm": 0.5296184420585632,
+      "learning_rate": 6.936274509803922e-06,
+      "loss": 0.5055,
+      "step": 283
+    },
+    {
+      "epoch": 0.20913107511045656,
+      "grad_norm": 0.4719228744506836,
+      "learning_rate": 6.96078431372549e-06,
+      "loss": 0.506,
+      "step": 284
+    },
+    {
+      "epoch": 0.20986745213549338,
+      "grad_norm": 0.49798110127449036,
+      "learning_rate": 6.985294117647059e-06,
+      "loss": 0.5311,
+      "step": 285
+    },
+    {
+      "epoch": 0.2106038291605302,
+      "grad_norm": 0.5432643890380859,
+      "learning_rate": 7.009803921568628e-06,
+      "loss": 0.4817,
+      "step": 286
+    },
+    {
+      "epoch": 0.211340206185567,
+      "grad_norm": 0.5015427470207214,
+      "learning_rate": 7.034313725490197e-06,
+      "loss": 0.4787,
+      "step": 287
+    },
+    {
+      "epoch": 0.21207658321060383,
+      "grad_norm": 0.4968355596065521,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 0.5051,
+      "step": 288
+    },
+    {
+      "epoch": 0.21281296023564064,
+      "grad_norm": 0.6130073666572571,
+      "learning_rate": 7.083333333333335e-06,
+      "loss": 0.4917,
+      "step": 289
+    },
+    {
+      "epoch": 0.21354933726067746,
+      "grad_norm": 0.4766775369644165,
+      "learning_rate": 7.107843137254903e-06,
+      "loss": 0.4849,
+      "step": 290
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 0.5826199650764465,
+      "learning_rate": 7.132352941176472e-06,
+      "loss": 0.4868,
+      "step": 291
+    },
+    {
+      "epoch": 0.21502209131075112,
+      "grad_norm": 0.43983742594718933,
+      "learning_rate": 7.15686274509804e-06,
+      "loss": 0.4949,
+      "step": 292
+    },
+    {
+      "epoch": 0.21575846833578793,
+      "grad_norm": 0.561597466468811,
+      "learning_rate": 7.181372549019609e-06,
+      "loss": 0.4796,
+      "step": 293
+    },
+    {
+      "epoch": 0.21649484536082475,
+      "grad_norm": 0.5399706363677979,
+      "learning_rate": 7.205882352941177e-06,
+      "loss": 0.4824,
+      "step": 294
+    },
+    {
+      "epoch": 0.21723122238586157,
+      "grad_norm": 0.45450007915496826,
+      "learning_rate": 7.230392156862746e-06,
+      "loss": 0.5083,
+      "step": 295
+    },
+    {
+      "epoch": 0.21796759941089838,
+      "grad_norm": 0.5827359557151794,
+      "learning_rate": 7.2549019607843145e-06,
+      "loss": 0.5013,
+      "step": 296
+    },
+    {
+      "epoch": 0.2187039764359352,
+      "grad_norm": 0.5125693678855896,
+      "learning_rate": 7.2794117647058826e-06,
+      "loss": 0.505,
+      "step": 297
+    },
+    {
+      "epoch": 0.21944035346097202,
+      "grad_norm": 0.5245058536529541,
+      "learning_rate": 7.3039215686274515e-06,
+      "loss": 0.4918,
+      "step": 298
+    },
+    {
+      "epoch": 0.22017673048600883,
+      "grad_norm": 0.5342023372650146,
+      "learning_rate": 7.3284313725490195e-06,
+      "loss": 0.4898,
+      "step": 299
+    },
+    {
+      "epoch": 0.22091310751104565,
+      "grad_norm": 0.4772011339664459,
+      "learning_rate": 7.352941176470589e-06,
+      "loss": 0.4958,
+      "step": 300
+    },
+    {
+      "epoch": 0.22164948453608246,
+      "grad_norm": 0.5259323716163635,
+      "learning_rate": 7.377450980392158e-06,
+      "loss": 0.4956,
+      "step": 301
+    },
+    {
+      "epoch": 0.22238586156111928,
+      "grad_norm": 0.5168774724006653,
+      "learning_rate": 7.401960784313726e-06,
+      "loss": 0.4735,
+      "step": 302
+    },
+    {
+      "epoch": 0.22312223858615612,
+      "grad_norm": 0.4500476121902466,
+      "learning_rate": 7.426470588235295e-06,
+      "loss": 0.5039,
+      "step": 303
+    },
+    {
+      "epoch": 0.22385861561119294,
+      "grad_norm": 0.49780377745628357,
+      "learning_rate": 7.450980392156863e-06,
+      "loss": 0.4629,
+      "step": 304
+    },
+    {
+      "epoch": 0.22459499263622976,
+      "grad_norm": 0.5209547281265259,
+      "learning_rate": 7.475490196078432e-06,
+      "loss": 0.4927,
+      "step": 305
+    },
+    {
+      "epoch": 0.22533136966126657,
+      "grad_norm": 0.6013898849487305,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.5053,
+      "step": 306
+    },
+    {
+      "epoch": 0.2260677466863034,
+      "grad_norm": 0.6150967478752136,
+      "learning_rate": 7.524509803921569e-06,
+      "loss": 0.4872,
+      "step": 307
+    },
+    {
+      "epoch": 0.2268041237113402,
+      "grad_norm": 0.48117902874946594,
+      "learning_rate": 7.549019607843138e-06,
+      "loss": 0.5077,
+      "step": 308
+    },
+    {
+      "epoch": 0.22754050073637702,
+      "grad_norm": 0.6058593392372131,
+      "learning_rate": 7.573529411764706e-06,
+      "loss": 0.4854,
+      "step": 309
+    },
+    {
+      "epoch": 0.22827687776141384,
+      "grad_norm": 0.5079674124717712,
+      "learning_rate": 7.598039215686275e-06,
+      "loss": 0.4977,
+      "step": 310
+    },
+    {
+      "epoch": 0.22901325478645065,
+      "grad_norm": 0.5175543427467346,
+      "learning_rate": 7.622549019607843e-06,
+      "loss": 0.5181,
+      "step": 311
+    },
+    {
+      "epoch": 0.22974963181148747,
+      "grad_norm": 0.5423833131790161,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 0.5364,
+      "step": 312
+    },
+    {
+      "epoch": 0.23048600883652431,
+      "grad_norm": 0.49087536334991455,
+      "learning_rate": 7.671568627450981e-06,
+      "loss": 0.4943,
+      "step": 313
+    },
+    {
+      "epoch": 0.23122238586156113,
+      "grad_norm": 0.5285847783088684,
+      "learning_rate": 7.69607843137255e-06,
+      "loss": 0.5098,
+      "step": 314
+    },
+    {
+      "epoch": 0.23195876288659795,
+      "grad_norm": 0.4606481194496155,
+      "learning_rate": 7.720588235294119e-06,
+      "loss": 0.477,
+      "step": 315
+    },
+    {
+      "epoch": 0.23269513991163476,
+      "grad_norm": 0.5595932006835938,
+      "learning_rate": 7.745098039215687e-06,
+      "loss": 0.5141,
+      "step": 316
+    },
+    {
+      "epoch": 0.23343151693667158,
+      "grad_norm": 0.4855089783668518,
+      "learning_rate": 7.769607843137256e-06,
+      "loss": 0.5224,
+      "step": 317
+    },
+    {
+      "epoch": 0.2341678939617084,
+      "grad_norm": 0.49227041006088257,
+      "learning_rate": 7.794117647058825e-06,
+      "loss": 0.4859,
+      "step": 318
+    },
+    {
+      "epoch": 0.2349042709867452,
+      "grad_norm": 0.46864691376686096,
+      "learning_rate": 7.818627450980393e-06,
+      "loss": 0.4907,
+      "step": 319
+    },
+    {
+      "epoch": 0.23564064801178203,
+      "grad_norm": 0.5469391942024231,
+      "learning_rate": 7.84313725490196e-06,
+      "loss": 0.4839,
+      "step": 320
+    },
+    {
+      "epoch": 0.23637702503681884,
+      "grad_norm": 0.5292018055915833,
+      "learning_rate": 7.86764705882353e-06,
+      "loss": 0.4977,
+      "step": 321
+    },
+    {
+      "epoch": 0.23711340206185566,
+      "grad_norm": 0.5337371230125427,
+      "learning_rate": 7.892156862745098e-06,
+      "loss": 0.4871,
+      "step": 322
+    },
+    {
+      "epoch": 0.23784977908689248,
+      "grad_norm": 0.5429027080535889,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 0.4995,
+      "step": 323
+    },
+    {
+      "epoch": 0.23858615611192932,
+      "grad_norm": 0.5223814249038696,
+      "learning_rate": 7.941176470588236e-06,
+      "loss": 0.5147,
+      "step": 324
+    },
+    {
+      "epoch": 0.23932253313696614,
+      "grad_norm": 0.4621415436267853,
+      "learning_rate": 7.965686274509804e-06,
+      "loss": 0.4943,
+      "step": 325
+    },
+    {
+      "epoch": 0.24005891016200295,
+      "grad_norm": 0.6054308414459229,
+      "learning_rate": 7.990196078431374e-06,
+      "loss": 0.535,
+      "step": 326
+    },
+    {
+      "epoch": 0.24079528718703977,
+      "grad_norm": 0.5081130266189575,
+      "learning_rate": 8.014705882352942e-06,
+      "loss": 0.4994,
+      "step": 327
+    },
+    {
+      "epoch": 0.24153166421207659,
+      "grad_norm": 0.5109507441520691,
+      "learning_rate": 8.03921568627451e-06,
+      "loss": 0.4765,
+      "step": 328
+    },
+    {
+      "epoch": 0.2422680412371134,
+      "grad_norm": 0.5108455419540405,
+      "learning_rate": 8.06372549019608e-06,
+      "loss": 0.4965,
+      "step": 329
+    },
+    {
+      "epoch": 0.24300441826215022,
+      "grad_norm": 0.4857231378555298,
+      "learning_rate": 8.088235294117648e-06,
+      "loss": 0.4663,
+      "step": 330
+    },
+    {
+      "epoch": 0.24374079528718703,
+      "grad_norm": 0.5149396061897278,
+      "learning_rate": 8.112745098039216e-06,
+      "loss": 0.517,
+      "step": 331
+    },
+    {
+      "epoch": 0.24447717231222385,
+      "grad_norm": 0.5667039752006531,
+      "learning_rate": 8.137254901960784e-06,
+      "loss": 0.5037,
+      "step": 332
+    },
+    {
+      "epoch": 0.24521354933726067,
+      "grad_norm": 0.5189935564994812,
+      "learning_rate": 8.161764705882354e-06,
+      "loss": 0.4786,
+      "step": 333
+    },
+    {
+      "epoch": 0.24594992636229748,
+      "grad_norm": 0.5272833108901978,
+      "learning_rate": 8.186274509803922e-06,
+      "loss": 0.4986,
+      "step": 334
+    },
+    {
+      "epoch": 0.24668630338733433,
+      "grad_norm": 0.4853070080280304,
+      "learning_rate": 8.21078431372549e-06,
+      "loss": 0.4876,
+      "step": 335
+    },
+    {
+      "epoch": 0.24742268041237114,
+      "grad_norm": 0.6110916137695312,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.5359,
+      "step": 336
+    },
+    {
+      "epoch": 0.24815905743740796,
+      "grad_norm": 0.6302662491798401,
+      "learning_rate": 8.259803921568628e-06,
+      "loss": 0.5306,
+      "step": 337
+    },
+    {
+      "epoch": 0.24889543446244478,
+      "grad_norm": 0.5661644339561462,
+      "learning_rate": 8.284313725490198e-06,
+      "loss": 0.4894,
+      "step": 338
+    },
+    {
+      "epoch": 0.2496318114874816,
+      "grad_norm": 0.5667375326156616,
+      "learning_rate": 8.308823529411766e-06,
+      "loss": 0.4911,
+      "step": 339
+    },
+    {
+      "epoch": 0.2503681885125184,
+      "grad_norm": 0.5805349349975586,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.499,
+      "step": 340
+    },
+    {
+      "epoch": 0.2511045655375552,
+      "grad_norm": 0.5153322219848633,
+      "learning_rate": 8.357843137254903e-06,
+      "loss": 0.4693,
+      "step": 341
+    },
+    {
+      "epoch": 0.25184094256259204,
+      "grad_norm": 0.5583307147026062,
+      "learning_rate": 8.382352941176472e-06,
+      "loss": 0.4913,
+      "step": 342
+    },
+    {
+      "epoch": 0.25257731958762886,
+      "grad_norm": 0.5932314395904541,
+      "learning_rate": 8.40686274509804e-06,
+      "loss": 0.5091,
+      "step": 343
+    },
+    {
+      "epoch": 0.2533136966126657,
+      "grad_norm": 0.4746854603290558,
+      "learning_rate": 8.43137254901961e-06,
+      "loss": 0.4552,
+      "step": 344
+    },
+    {
+      "epoch": 0.2540500736377025,
+      "grad_norm": 0.6550336480140686,
+      "learning_rate": 8.455882352941177e-06,
+      "loss": 0.4583,
+      "step": 345
+    },
+    {
+      "epoch": 0.2547864506627393,
+      "grad_norm": 0.44437041878700256,
+      "learning_rate": 8.480392156862745e-06,
+      "loss": 0.4817,
+      "step": 346
+    },
+    {
+      "epoch": 0.2555228276877761,
+      "grad_norm": 0.6586700677871704,
+      "learning_rate": 8.504901960784314e-06,
+      "loss": 0.4843,
+      "step": 347
+    },
+    {
+      "epoch": 0.25625920471281294,
+      "grad_norm": 0.5380249619483948,
+      "learning_rate": 8.529411764705883e-06,
+      "loss": 0.4783,
+      "step": 348
+    },
+    {
+      "epoch": 0.25699558173784975,
+      "grad_norm": 0.5162436962127686,
+      "learning_rate": 8.553921568627451e-06,
+      "loss": 0.4796,
+      "step": 349
+    },
+    {
+      "epoch": 0.25773195876288657,
+      "grad_norm": 0.5282143354415894,
+      "learning_rate": 8.57843137254902e-06,
+      "loss": 0.4583,
+      "step": 350
+    },
+    {
+      "epoch": 0.25846833578792344,
+      "grad_norm": 0.5736419558525085,
+      "learning_rate": 8.60294117647059e-06,
+      "loss": 0.4989,
+      "step": 351
+    },
+    {
+      "epoch": 0.25920471281296026,
+      "grad_norm": 0.47756871581077576,
+      "learning_rate": 8.627450980392157e-06,
+      "loss": 0.4766,
+      "step": 352
+    },
+    {
+      "epoch": 0.2599410898379971,
+      "grad_norm": 0.5326557159423828,
+      "learning_rate": 8.651960784313727e-06,
+      "loss": 0.4931,
+      "step": 353
+    },
+    {
+      "epoch": 0.2606774668630339,
+      "grad_norm": 0.5259348750114441,
+      "learning_rate": 8.676470588235295e-06,
+      "loss": 0.4863,
+      "step": 354
+    },
+    {
+      "epoch": 0.2614138438880707,
+      "grad_norm": 0.5482640862464905,
+      "learning_rate": 8.700980392156863e-06,
+      "loss": 0.5174,
+      "step": 355
+    },
+    {
+      "epoch": 0.2621502209131075,
+      "grad_norm": 0.5132611989974976,
+      "learning_rate": 8.725490196078433e-06,
+      "loss": 0.4894,
+      "step": 356
+    },
+    {
+      "epoch": 0.26288659793814434,
+      "grad_norm": 0.5965814590454102,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.4922,
+      "step": 357
+    },
+    {
+      "epoch": 0.26362297496318116,
+      "grad_norm": 0.5167778730392456,
+      "learning_rate": 8.774509803921569e-06,
+      "loss": 0.4737,
+      "step": 358
+    },
+    {
+      "epoch": 0.26435935198821797,
+      "grad_norm": 0.510610818862915,
+      "learning_rate": 8.799019607843137e-06,
+      "loss": 0.4877,
+      "step": 359
+    },
+    {
+      "epoch": 0.2650957290132548,
+      "grad_norm": 0.5722100734710693,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.4597,
+      "step": 360
+    },
+    {
+      "epoch": 0.2658321060382916,
+      "grad_norm": 0.662177562713623,
+      "learning_rate": 8.848039215686275e-06,
+      "loss": 0.5086,
+      "step": 361
+    },
+    {
+      "epoch": 0.2665684830633284,
+      "grad_norm": 0.499646931886673,
+      "learning_rate": 8.872549019607843e-06,
+      "loss": 0.4513,
+      "step": 362
+    },
+    {
+      "epoch": 0.26730486008836524,
+      "grad_norm": 0.5374613404273987,
+      "learning_rate": 8.897058823529413e-06,
+      "loss": 0.4636,
+      "step": 363
+    },
+    {
+      "epoch": 0.26804123711340205,
+      "grad_norm": 0.6908944845199585,
+      "learning_rate": 8.921568627450982e-06,
+      "loss": 0.4969,
+      "step": 364
+    },
+    {
+      "epoch": 0.26877761413843887,
+      "grad_norm": 0.4661368131637573,
+      "learning_rate": 8.94607843137255e-06,
+      "loss": 0.4593,
+      "step": 365
+    },
+    {
+      "epoch": 0.2695139911634757,
+      "grad_norm": 0.5653083920478821,
+      "learning_rate": 8.970588235294119e-06,
+      "loss": 0.4714,
+      "step": 366
+    },
+    {
+      "epoch": 0.2702503681885125,
+      "grad_norm": 0.46112319827079773,
+      "learning_rate": 8.995098039215687e-06,
+      "loss": 0.4775,
+      "step": 367
+    },
+    {
+      "epoch": 0.2709867452135493,
+      "grad_norm": 0.5458109378814697,
+      "learning_rate": 9.019607843137256e-06,
+      "loss": 0.472,
+      "step": 368
+    },
+    {
+      "epoch": 0.27172312223858613,
+      "grad_norm": 0.5330044031143188,
+      "learning_rate": 9.044117647058824e-06,
+      "loss": 0.4798,
+      "step": 369
+    },
+    {
+      "epoch": 0.27245949926362295,
+      "grad_norm": 0.5421646237373352,
+      "learning_rate": 9.068627450980392e-06,
+      "loss": 0.5013,
+      "step": 370
+    },
+    {
+      "epoch": 0.27319587628865977,
+      "grad_norm": 0.575491726398468,
+      "learning_rate": 9.093137254901962e-06,
+      "loss": 0.4939,
+      "step": 371
+    },
+    {
+      "epoch": 0.27393225331369664,
+      "grad_norm": 0.5331813097000122,
+      "learning_rate": 9.11764705882353e-06,
+      "loss": 0.514,
+      "step": 372
+    },
+    {
+      "epoch": 0.27466863033873345,
+      "grad_norm": 0.5526221990585327,
+      "learning_rate": 9.142156862745098e-06,
+      "loss": 0.5139,
+      "step": 373
+    },
+    {
+      "epoch": 0.27540500736377027,
+      "grad_norm": 0.5274295806884766,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.4895,
+      "step": 374
+    },
+    {
+      "epoch": 0.2761413843888071,
+      "grad_norm": 0.5230691432952881,
+      "learning_rate": 9.191176470588236e-06,
+      "loss": 0.4956,
+      "step": 375
+    },
+    {
+      "epoch": 0.2768777614138439,
+      "grad_norm": 0.6479937434196472,
+      "learning_rate": 9.215686274509804e-06,
+      "loss": 0.4768,
+      "step": 376
+    },
+    {
+      "epoch": 0.2776141384388807,
+      "grad_norm": 0.5493687987327576,
+      "learning_rate": 9.240196078431374e-06,
+      "loss": 0.4733,
+      "step": 377
+    },
+    {
+      "epoch": 0.27835051546391754,
+      "grad_norm": 0.5068442225456238,
+      "learning_rate": 9.264705882352942e-06,
+      "loss": 0.4696,
+      "step": 378
+    },
+    {
+      "epoch": 0.27908689248895435,
+      "grad_norm": 0.49230822920799255,
+      "learning_rate": 9.28921568627451e-06,
+      "loss": 0.4747,
+      "step": 379
+    },
+    {
+      "epoch": 0.27982326951399117,
+      "grad_norm": 0.5067638158798218,
+      "learning_rate": 9.31372549019608e-06,
+      "loss": 0.4598,
+      "step": 380
+    },
+    {
+      "epoch": 0.280559646539028,
+      "grad_norm": 0.5510022640228271,
+      "learning_rate": 9.338235294117648e-06,
+      "loss": 0.4844,
+      "step": 381
+    },
+    {
+      "epoch": 0.2812960235640648,
+      "grad_norm": 0.5649453401565552,
+      "learning_rate": 9.362745098039216e-06,
+      "loss": 0.497,
+      "step": 382
+    },
+    {
+      "epoch": 0.2820324005891016,
+      "grad_norm": 0.549223780632019,
+      "learning_rate": 9.387254901960786e-06,
+      "loss": 0.4801,
+      "step": 383
+    },
+    {
+      "epoch": 0.28276877761413843,
+      "grad_norm": 0.46957284212112427,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.4752,
+      "step": 384
+    },
+    {
+      "epoch": 0.28350515463917525,
+      "grad_norm": 0.4900481104850769,
+      "learning_rate": 9.436274509803922e-06,
+      "loss": 0.4712,
+      "step": 385
+    },
+    {
+      "epoch": 0.28424153166421207,
+      "grad_norm": 0.49515727162361145,
+      "learning_rate": 9.46078431372549e-06,
+      "loss": 0.4827,
+      "step": 386
+    },
+    {
+      "epoch": 0.2849779086892489,
+      "grad_norm": 0.44085896015167236,
+      "learning_rate": 9.48529411764706e-06,
+      "loss": 0.4651,
+      "step": 387
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.5362505912780762,
+      "learning_rate": 9.509803921568628e-06,
+      "loss": 0.4951,
+      "step": 388
+    },
+    {
+      "epoch": 0.2864506627393225,
+      "grad_norm": 0.593925952911377,
+      "learning_rate": 9.534313725490198e-06,
+      "loss": 0.471,
+      "step": 389
+    },
+    {
+      "epoch": 0.28718703976435933,
+      "grad_norm": 0.6149821281433105,
+      "learning_rate": 9.558823529411766e-06,
+      "loss": 0.5252,
+      "step": 390
+    },
+    {
+      "epoch": 0.28792341678939615,
+      "grad_norm": 0.484279066324234,
+      "learning_rate": 9.583333333333335e-06,
+      "loss": 0.4968,
+      "step": 391
+    },
+    {
+      "epoch": 0.28865979381443296,
+      "grad_norm": 0.59568190574646,
+      "learning_rate": 9.607843137254903e-06,
+      "loss": 0.4778,
+      "step": 392
+    },
+    {
+      "epoch": 0.28939617083946984,
+      "grad_norm": 0.587205708026886,
+      "learning_rate": 9.632352941176471e-06,
+      "loss": 0.5125,
+      "step": 393
+    },
+    {
+      "epoch": 0.29013254786450665,
+      "grad_norm": 0.5671679377555847,
+      "learning_rate": 9.65686274509804e-06,
+      "loss": 0.5075,
+      "step": 394
+    },
+    {
+      "epoch": 0.29086892488954347,
+      "grad_norm": 0.5437478423118591,
+      "learning_rate": 9.68137254901961e-06,
+      "loss": 0.4632,
+      "step": 395
+    },
+    {
+      "epoch": 0.2916053019145803,
+      "grad_norm": 0.517405092716217,
+      "learning_rate": 9.705882352941177e-06,
+      "loss": 0.4851,
+      "step": 396
+    },
+    {
+      "epoch": 0.2923416789396171,
+      "grad_norm": 0.4883304238319397,
+      "learning_rate": 9.730392156862745e-06,
+      "loss": 0.4827,
+      "step": 397
+    },
+    {
+      "epoch": 0.2930780559646539,
+      "grad_norm": 0.5211227536201477,
+      "learning_rate": 9.754901960784315e-06,
+      "loss": 0.5063,
+      "step": 398
+    },
+    {
+      "epoch": 0.29381443298969073,
+      "grad_norm": 0.4950783848762512,
+      "learning_rate": 9.779411764705883e-06,
+      "loss": 0.4588,
+      "step": 399
+    },
+    {
+      "epoch": 0.29455081001472755,
+      "grad_norm": 0.5107828378677368,
+      "learning_rate": 9.803921568627451e-06,
+      "loss": 0.4775,
+      "step": 400
+    },
+    {
+      "epoch": 0.29528718703976436,
+      "grad_norm": 0.5515358448028564,
+      "learning_rate": 9.82843137254902e-06,
+      "loss": 0.4581,
+      "step": 401
+    },
+    {
+      "epoch": 0.2960235640648012,
+      "grad_norm": 0.6434028744697571,
+      "learning_rate": 9.852941176470589e-06,
+      "loss": 0.5061,
+      "step": 402
+    },
+    {
+      "epoch": 0.296759941089838,
+      "grad_norm": 0.6046220064163208,
+      "learning_rate": 9.877450980392159e-06,
+      "loss": 0.5053,
+      "step": 403
+    },
+    {
+      "epoch": 0.2974963181148748,
+      "grad_norm": 0.5827841758728027,
+      "learning_rate": 9.901960784313727e-06,
+      "loss": 0.4594,
+      "step": 404
+    },
+    {
+      "epoch": 0.29823269513991163,
+      "grad_norm": 0.5246471762657166,
+      "learning_rate": 9.926470588235295e-06,
+      "loss": 0.4909,
+      "step": 405
+    },
+    {
+      "epoch": 0.29896907216494845,
+      "grad_norm": 0.5674916505813599,
+      "learning_rate": 9.950980392156863e-06,
+      "loss": 0.4717,
+      "step": 406
+    },
+    {
+      "epoch": 0.29970544918998526,
+      "grad_norm": 0.5033355951309204,
+      "learning_rate": 9.975490196078433e-06,
+      "loss": 0.5008,
+      "step": 407
+    },
+    {
+      "epoch": 0.3004418262150221,
+      "grad_norm": 0.5390115976333618,
+      "learning_rate": 1e-05,
+      "loss": 0.5002,
+      "step": 408
+    },
+    {
+      "epoch": 0.3011782032400589,
+      "grad_norm": 0.5051551461219788,
+      "learning_rate": 9.999998164075549e-06,
+      "loss": 0.4609,
+      "step": 409
+    },
+    {
+      "epoch": 0.3019145802650957,
+      "grad_norm": 0.5789138078689575,
+      "learning_rate": 9.999992656303539e-06,
+      "loss": 0.5152,
+      "step": 410
+    },
+    {
+      "epoch": 0.3026509572901325,
+      "grad_norm": 0.4873914420604706,
+      "learning_rate": 9.999983476688016e-06,
+      "loss": 0.4836,
+      "step": 411
+    },
+    {
+      "epoch": 0.30338733431516934,
+      "grad_norm": 0.5575588345527649,
+      "learning_rate": 9.999970625235724e-06,
+      "loss": 0.472,
+      "step": 412
+    },
+    {
+      "epoch": 0.30412371134020616,
+      "grad_norm": 0.5004022121429443,
+      "learning_rate": 9.999954101956097e-06,
+      "loss": 0.5051,
+      "step": 413
+    },
+    {
+      "epoch": 0.30486008836524303,
+      "grad_norm": 0.5742695927619934,
+      "learning_rate": 9.999933906861272e-06,
+      "loss": 0.4613,
+      "step": 414
+    },
+    {
+      "epoch": 0.30559646539027985,
+      "grad_norm": 0.5978338122367859,
+      "learning_rate": 9.999910039966079e-06,
+      "loss": 0.4931,
+      "step": 415
+    },
+    {
+      "epoch": 0.30633284241531666,
+      "grad_norm": 0.5001004338264465,
+      "learning_rate": 9.999882501288043e-06,
+      "loss": 0.475,
+      "step": 416
+    },
+    {
+      "epoch": 0.3070692194403535,
+      "grad_norm": 0.608790397644043,
+      "learning_rate": 9.99985129084739e-06,
+      "loss": 0.4743,
+      "step": 417
+    },
+    {
+      "epoch": 0.3078055964653903,
+      "grad_norm": 0.5592244863510132,
+      "learning_rate": 9.99981640866704e-06,
+      "loss": 0.4767,
+      "step": 418
+    },
+    {
+      "epoch": 0.3085419734904271,
+      "grad_norm": 0.6118118166923523,
+      "learning_rate": 9.999777854772608e-06,
+      "loss": 0.5155,
+      "step": 419
+    },
+    {
+      "epoch": 0.30927835051546393,
+      "grad_norm": 0.691983699798584,
+      "learning_rate": 9.999735629192408e-06,
+      "loss": 0.4758,
+      "step": 420
+    },
+    {
+      "epoch": 0.31001472754050075,
+      "grad_norm": 0.5400111675262451,
+      "learning_rate": 9.99968973195745e-06,
+      "loss": 0.4831,
+      "step": 421
+    },
+    {
+      "epoch": 0.31075110456553756,
+      "grad_norm": 0.6896957159042358,
+      "learning_rate": 9.999640163101436e-06,
+      "loss": 0.4969,
+      "step": 422
+    },
+    {
+      "epoch": 0.3114874815905744,
+      "grad_norm": 0.5258437991142273,
+      "learning_rate": 9.99958692266077e-06,
+      "loss": 0.4701,
+      "step": 423
+    },
+    {
+      "epoch": 0.3122238586156112,
+      "grad_norm": 0.6300147175788879,
+      "learning_rate": 9.999530010674552e-06,
+      "loss": 0.5038,
+      "step": 424
+    },
+    {
+      "epoch": 0.312960235640648,
+      "grad_norm": 0.5327275991439819,
+      "learning_rate": 9.999469427184573e-06,
+      "loss": 0.4895,
+      "step": 425
+    },
+    {
+      "epoch": 0.3136966126656848,
+      "grad_norm": 0.5516902804374695,
+      "learning_rate": 9.999405172235325e-06,
+      "loss": 0.4983,
+      "step": 426
+    },
+    {
+      "epoch": 0.31443298969072164,
+      "grad_norm": 0.5630642771720886,
+      "learning_rate": 9.999337245873999e-06,
+      "loss": 0.4625,
+      "step": 427
+    },
+    {
+      "epoch": 0.31516936671575846,
+      "grad_norm": 0.5986700654029846,
+      "learning_rate": 9.999265648150472e-06,
+      "loss": 0.524,
+      "step": 428
+    },
+    {
+      "epoch": 0.3159057437407953,
+      "grad_norm": 0.5675637125968933,
+      "learning_rate": 9.999190379117324e-06,
+      "loss": 0.484,
+      "step": 429
+    },
+    {
+      "epoch": 0.3166421207658321,
+      "grad_norm": 0.6256847977638245,
+      "learning_rate": 9.999111438829834e-06,
+      "loss": 0.5181,
+      "step": 430
+    },
+    {
+      "epoch": 0.3173784977908689,
+      "grad_norm": 0.6200090050697327,
+      "learning_rate": 9.999028827345969e-06,
+      "loss": 0.4901,
+      "step": 431
+    },
+    {
+      "epoch": 0.3181148748159057,
+      "grad_norm": 0.5858426690101624,
+      "learning_rate": 9.9989425447264e-06,
+      "loss": 0.4901,
+      "step": 432
+    },
+    {
+      "epoch": 0.31885125184094254,
+      "grad_norm": 0.5074707269668579,
+      "learning_rate": 9.998852591034488e-06,
+      "loss": 0.4749,
+      "step": 433
+    },
+    {
+      "epoch": 0.31958762886597936,
+      "grad_norm": 0.5885580778121948,
+      "learning_rate": 9.998758966336296e-06,
+      "loss": 0.4836,
+      "step": 434
+    },
+    {
+      "epoch": 0.32032400589101623,
+      "grad_norm": 0.5307214856147766,
+      "learning_rate": 9.998661670700576e-06,
+      "loss": 0.475,
+      "step": 435
+    },
+    {
+      "epoch": 0.32106038291605304,
+      "grad_norm": 0.6851618885993958,
+      "learning_rate": 9.998560704198776e-06,
+      "loss": 0.4521,
+      "step": 436
+    },
+    {
+      "epoch": 0.32179675994108986,
+      "grad_norm": 0.6349673271179199,
+      "learning_rate": 9.99845606690505e-06,
+      "loss": 0.4875,
+      "step": 437
+    },
+    {
+      "epoch": 0.3225331369661267,
+      "grad_norm": 0.5425810217857361,
+      "learning_rate": 9.998347758896234e-06,
+      "loss": 0.5036,
+      "step": 438
+    },
+    {
+      "epoch": 0.3232695139911635,
+      "grad_norm": 0.6161746382713318,
+      "learning_rate": 9.99823578025187e-06,
+      "loss": 0.4791,
+      "step": 439
+    },
+    {
+      "epoch": 0.3240058910162003,
+      "grad_norm": 0.51716548204422,
+      "learning_rate": 9.99812013105419e-06,
+      "loss": 0.4881,
+      "step": 440
+    },
+    {
+      "epoch": 0.3247422680412371,
+      "grad_norm": 0.6496948003768921,
+      "learning_rate": 9.998000811388122e-06,
+      "loss": 0.4804,
+      "step": 441
+    },
+    {
+      "epoch": 0.32547864506627394,
+      "grad_norm": 0.6212929487228394,
+      "learning_rate": 9.997877821341294e-06,
+      "loss": 0.4871,
+      "step": 442
+    },
+    {
+      "epoch": 0.32621502209131076,
+      "grad_norm": 0.6089105010032654,
+      "learning_rate": 9.997751161004026e-06,
+      "loss": 0.4518,
+      "step": 443
+    },
+    {
+      "epoch": 0.3269513991163476,
+      "grad_norm": 0.6576497554779053,
+      "learning_rate": 9.99762083046933e-06,
+      "loss": 0.4673,
+      "step": 444
+    },
+    {
+      "epoch": 0.3276877761413844,
+      "grad_norm": 0.6152659058570862,
+      "learning_rate": 9.99748682983292e-06,
+      "loss": 0.4811,
+      "step": 445
+    },
+    {
+      "epoch": 0.3284241531664212,
+      "grad_norm": 0.6389955878257751,
+      "learning_rate": 9.9973491591932e-06,
+      "loss": 0.4737,
+      "step": 446
+    },
+    {
+      "epoch": 0.329160530191458,
+      "grad_norm": 0.645021915435791,
+      "learning_rate": 9.997207818651273e-06,
+      "loss": 0.491,
+      "step": 447
+    },
+    {
+      "epoch": 0.32989690721649484,
+      "grad_norm": 0.5410696864128113,
+      "learning_rate": 9.997062808310935e-06,
+      "loss": 0.4895,
+      "step": 448
+    },
+    {
+      "epoch": 0.33063328424153166,
+      "grad_norm": 0.693676769733429,
+      "learning_rate": 9.996914128278677e-06,
+      "loss": 0.5014,
+      "step": 449
+    },
+    {
+      "epoch": 0.33136966126656847,
+      "grad_norm": 0.5651475191116333,
+      "learning_rate": 9.996761778663682e-06,
+      "loss": 0.477,
+      "step": 450
+    },
+    {
+      "epoch": 0.3321060382916053,
+      "grad_norm": 0.6107382774353027,
+      "learning_rate": 9.996605759577836e-06,
+      "loss": 0.4553,
+      "step": 451
+    },
+    {
+      "epoch": 0.3328424153166421,
+      "grad_norm": 0.540756344795227,
+      "learning_rate": 9.996446071135711e-06,
+      "loss": 0.4961,
+      "step": 452
+    },
+    {
+      "epoch": 0.3335787923416789,
+      "grad_norm": 0.5446950793266296,
+      "learning_rate": 9.99628271345458e-06,
+      "loss": 0.4891,
+      "step": 453
+    },
+    {
+      "epoch": 0.33431516936671574,
+      "grad_norm": 0.4631234407424927,
+      "learning_rate": 9.996115686654406e-06,
+      "loss": 0.4734,
+      "step": 454
+    },
+    {
+      "epoch": 0.33505154639175255,
+      "grad_norm": 0.4855410158634186,
+      "learning_rate": 9.995944990857848e-06,
+      "loss": 0.4658,
+      "step": 455
+    },
+    {
+      "epoch": 0.33578792341678937,
+      "grad_norm": 0.5123491287231445,
+      "learning_rate": 9.995770626190263e-06,
+      "loss": 0.4627,
+      "step": 456
+    },
+    {
+      "epoch": 0.33652430044182624,
+      "grad_norm": 0.512749969959259,
+      "learning_rate": 9.995592592779695e-06,
+      "loss": 0.4798,
+      "step": 457
+    },
+    {
+      "epoch": 0.33726067746686306,
+      "grad_norm": 0.5249326229095459,
+      "learning_rate": 9.995410890756891e-06,
+      "loss": 0.5012,
+      "step": 458
+    },
+    {
+      "epoch": 0.3379970544918999,
+      "grad_norm": 0.47508466243743896,
+      "learning_rate": 9.995225520255282e-06,
+      "loss": 0.5039,
+      "step": 459
+    },
+    {
+      "epoch": 0.3387334315169367,
+      "grad_norm": 0.4680541455745697,
+      "learning_rate": 9.995036481411005e-06,
+      "loss": 0.4666,
+      "step": 460
+    },
+    {
+      "epoch": 0.3394698085419735,
+      "grad_norm": 0.4403444528579712,
+      "learning_rate": 9.994843774362878e-06,
+      "loss": 0.4613,
+      "step": 461
+    },
+    {
+      "epoch": 0.3402061855670103,
+      "grad_norm": 0.48063525557518005,
+      "learning_rate": 9.994647399252423e-06,
+      "loss": 0.4709,
+      "step": 462
+    },
+    {
+      "epoch": 0.34094256259204714,
+      "grad_norm": 0.4822491705417633,
+      "learning_rate": 9.99444735622385e-06,
+      "loss": 0.4947,
+      "step": 463
+    },
+    {
+      "epoch": 0.34167893961708395,
+      "grad_norm": 0.533244788646698,
+      "learning_rate": 9.994243645424067e-06,
+      "loss": 0.4927,
+      "step": 464
+    },
+    {
+      "epoch": 0.34241531664212077,
+      "grad_norm": 0.49757879972457886,
+      "learning_rate": 9.99403626700267e-06,
+      "loss": 0.4851,
+      "step": 465
+    },
+    {
+      "epoch": 0.3431516936671576,
+      "grad_norm": 0.583299994468689,
+      "learning_rate": 9.993825221111955e-06,
+      "loss": 0.5015,
+      "step": 466
+    },
+    {
+      "epoch": 0.3438880706921944,
+      "grad_norm": 0.46042028069496155,
+      "learning_rate": 9.993610507906904e-06,
+      "loss": 0.5083,
+      "step": 467
+    },
+    {
+      "epoch": 0.3446244477172312,
+      "grad_norm": 0.5615798830986023,
+      "learning_rate": 9.993392127545198e-06,
+      "loss": 0.4744,
+      "step": 468
+    },
+    {
+      "epoch": 0.34536082474226804,
+      "grad_norm": 0.4136447310447693,
+      "learning_rate": 9.99317008018721e-06,
+      "loss": 0.4552,
+      "step": 469
+    },
+    {
+      "epoch": 0.34609720176730485,
+      "grad_norm": 0.5388152599334717,
+      "learning_rate": 9.992944365996002e-06,
+      "loss": 0.4907,
+      "step": 470
+    },
+    {
+      "epoch": 0.34683357879234167,
+      "grad_norm": 0.5055995583534241,
+      "learning_rate": 9.992714985137336e-06,
+      "loss": 0.4649,
+      "step": 471
+    },
+    {
+      "epoch": 0.3475699558173785,
+      "grad_norm": 0.599743664264679,
+      "learning_rate": 9.992481937779655e-06,
+      "loss": 0.4747,
+      "step": 472
+    },
+    {
+      "epoch": 0.3483063328424153,
+      "grad_norm": 0.5145841836929321,
+      "learning_rate": 9.99224522409411e-06,
+      "loss": 0.5214,
+      "step": 473
+    },
+    {
+      "epoch": 0.3490427098674521,
+      "grad_norm": 0.5827757716178894,
+      "learning_rate": 9.99200484425453e-06,
+      "loss": 0.5151,
+      "step": 474
+    },
+    {
+      "epoch": 0.34977908689248893,
+      "grad_norm": 0.502983808517456,
+      "learning_rate": 9.991760798437448e-06,
+      "loss": 0.4808,
+      "step": 475
+    },
+    {
+      "epoch": 0.35051546391752575,
+      "grad_norm": 0.5331257581710815,
+      "learning_rate": 9.99151308682208e-06,
+      "loss": 0.4544,
+      "step": 476
+    },
+    {
+      "epoch": 0.35125184094256257,
+      "grad_norm": 0.5600625276565552,
+      "learning_rate": 9.99126170959034e-06,
+      "loss": 0.4814,
+      "step": 477
+    },
+    {
+      "epoch": 0.35198821796759944,
+      "grad_norm": 0.5749978423118591,
+      "learning_rate": 9.991006666926832e-06,
+      "loss": 0.4766,
+      "step": 478
+    },
+    {
+      "epoch": 0.35272459499263625,
+      "grad_norm": 0.5478549599647522,
+      "learning_rate": 9.990747959018849e-06,
+      "loss": 0.4763,
+      "step": 479
+    },
+    {
+      "epoch": 0.35346097201767307,
+      "grad_norm": 0.467970609664917,
+      "learning_rate": 9.990485586056381e-06,
+      "loss": 0.4753,
+      "step": 480
+    },
+    {
+      "epoch": 0.3541973490427099,
+      "grad_norm": 0.4842618405818939,
+      "learning_rate": 9.990219548232106e-06,
+      "loss": 0.4721,
+      "step": 481
+    },
+    {
+      "epoch": 0.3549337260677467,
+      "grad_norm": 0.4819396138191223,
+      "learning_rate": 9.989949845741393e-06,
+      "loss": 0.449,
+      "step": 482
+    },
+    {
+      "epoch": 0.3556701030927835,
+      "grad_norm": 0.5315747857093811,
+      "learning_rate": 9.989676478782305e-06,
+      "loss": 0.4866,
+      "step": 483
+    },
+    {
+      "epoch": 0.35640648011782033,
+      "grad_norm": 0.5199954509735107,
+      "learning_rate": 9.989399447555594e-06,
+      "loss": 0.4858,
+      "step": 484
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 0.6086426377296448,
+      "learning_rate": 9.989118752264704e-06,
+      "loss": 0.5055,
+      "step": 485
+    },
+    {
+      "epoch": 0.35787923416789397,
+      "grad_norm": 0.5935198664665222,
+      "learning_rate": 9.988834393115768e-06,
+      "loss": 0.4878,
+      "step": 486
+    },
+    {
+      "epoch": 0.3586156111929308,
+      "grad_norm": 0.5054832696914673,
+      "learning_rate": 9.988546370317609e-06,
+      "loss": 0.4357,
+      "step": 487
+    },
+    {
+      "epoch": 0.3593519882179676,
+      "grad_norm": 0.47322750091552734,
+      "learning_rate": 9.988254684081746e-06,
+      "loss": 0.4607,
+      "step": 488
+    },
+    {
+      "epoch": 0.3600883652430044,
+      "grad_norm": 0.4437432289123535,
+      "learning_rate": 9.987959334622381e-06,
+      "loss": 0.4772,
+      "step": 489
+    },
+    {
+      "epoch": 0.36082474226804123,
+      "grad_norm": 0.4980893135070801,
+      "learning_rate": 9.987660322156413e-06,
+      "loss": 0.4725,
+      "step": 490
+    },
+    {
+      "epoch": 0.36156111929307805,
+      "grad_norm": 0.5222845077514648,
+      "learning_rate": 9.987357646903427e-06,
+      "loss": 0.4915,
+      "step": 491
+    },
+    {
+      "epoch": 0.36229749631811486,
+      "grad_norm": 0.45025306940078735,
+      "learning_rate": 9.987051309085698e-06,
+      "loss": 0.4995,
+      "step": 492
+    },
+    {
+      "epoch": 0.3630338733431517,
+      "grad_norm": 0.4636288583278656,
+      "learning_rate": 9.986741308928189e-06,
+      "loss": 0.4748,
+      "step": 493
+    },
+    {
+      "epoch": 0.3637702503681885,
+      "grad_norm": 0.49730953574180603,
+      "learning_rate": 9.986427646658559e-06,
+      "loss": 0.4876,
+      "step": 494
+    },
+    {
+      "epoch": 0.3645066273932253,
+      "grad_norm": 0.5664610862731934,
+      "learning_rate": 9.986110322507149e-06,
+      "loss": 0.4727,
+      "step": 495
+    },
+    {
+      "epoch": 0.36524300441826213,
+      "grad_norm": 0.48833712935447693,
+      "learning_rate": 9.985789336706993e-06,
+      "loss": 0.5076,
+      "step": 496
+    },
+    {
+      "epoch": 0.36597938144329895,
+      "grad_norm": 0.5941821336746216,
+      "learning_rate": 9.985464689493814e-06,
+      "loss": 0.5057,
+      "step": 497
+    },
+    {
+      "epoch": 0.36671575846833576,
+      "grad_norm": 0.4826650023460388,
+      "learning_rate": 9.985136381106022e-06,
+      "loss": 0.5201,
+      "step": 498
+    },
+    {
+      "epoch": 0.36745213549337263,
+      "grad_norm": 0.49603334069252014,
+      "learning_rate": 9.984804411784717e-06,
+      "loss": 0.4921,
+      "step": 499
+    },
+    {
+      "epoch": 0.36818851251840945,
+      "grad_norm": 0.4792011082172394,
+      "learning_rate": 9.984468781773688e-06,
+      "loss": 0.459,
+      "step": 500
+    },
+    {
+      "epoch": 0.36892488954344627,
+      "grad_norm": 0.4913840591907501,
+      "learning_rate": 9.98412949131941e-06,
+      "loss": 0.4734,
+      "step": 501
+    },
+    {
+      "epoch": 0.3696612665684831,
+      "grad_norm": 0.5027479529380798,
+      "learning_rate": 9.983786540671052e-06,
+      "loss": 0.4432,
+      "step": 502
+    },
+    {
+      "epoch": 0.3703976435935199,
+      "grad_norm": 0.4886789321899414,
+      "learning_rate": 9.98343993008046e-06,
+      "loss": 0.4696,
+      "step": 503
+    },
+    {
+      "epoch": 0.3711340206185567,
+      "grad_norm": 0.4682316482067108,
+      "learning_rate": 9.983089659802178e-06,
+      "loss": 0.4516,
+      "step": 504
+    },
+    {
+      "epoch": 0.37187039764359353,
+      "grad_norm": 0.465826153755188,
+      "learning_rate": 9.982735730093436e-06,
+      "loss": 0.4827,
+      "step": 505
+    },
+    {
+      "epoch": 0.37260677466863035,
+      "grad_norm": 0.5015543699264526,
+      "learning_rate": 9.982378141214144e-06,
+      "loss": 0.4415,
+      "step": 506
+    },
+    {
+      "epoch": 0.37334315169366716,
+      "grad_norm": 0.5372326970100403,
+      "learning_rate": 9.98201689342691e-06,
+      "loss": 0.4782,
+      "step": 507
+    },
+    {
+      "epoch": 0.374079528718704,
+      "grad_norm": 0.4749949276447296,
+      "learning_rate": 9.98165198699702e-06,
+      "loss": 0.4636,
+      "step": 508
+    },
+    {
+      "epoch": 0.3748159057437408,
+      "grad_norm": 0.459778755903244,
+      "learning_rate": 9.98128342219245e-06,
+      "loss": 0.4608,
+      "step": 509
+    },
+    {
+      "epoch": 0.3755522827687776,
+      "grad_norm": 0.5350815653800964,
+      "learning_rate": 9.980911199283864e-06,
+      "loss": 0.4877,
+      "step": 510
+    },
+    {
+      "epoch": 0.37628865979381443,
+      "grad_norm": 0.47353699803352356,
+      "learning_rate": 9.98053531854461e-06,
+      "loss": 0.4867,
+      "step": 511
+    },
+    {
+      "epoch": 0.37702503681885124,
+      "grad_norm": 0.4845249652862549,
+      "learning_rate": 9.980155780250728e-06,
+      "loss": 0.4473,
+      "step": 512
+    },
+    {
+      "epoch": 0.37776141384388806,
+      "grad_norm": 0.47789323329925537,
+      "learning_rate": 9.979772584680933e-06,
+      "loss": 0.4566,
+      "step": 513
+    },
+    {
+      "epoch": 0.3784977908689249,
+      "grad_norm": 0.5107650756835938,
+      "learning_rate": 9.979385732116638e-06,
+      "loss": 0.4341,
+      "step": 514
+    },
+    {
+      "epoch": 0.3792341678939617,
+      "grad_norm": 0.5246970057487488,
+      "learning_rate": 9.978995222841932e-06,
+      "loss": 0.4741,
+      "step": 515
+    },
+    {
+      "epoch": 0.3799705449189985,
+      "grad_norm": 0.5087926387786865,
+      "learning_rate": 9.978601057143593e-06,
+      "loss": 0.4582,
+      "step": 516
+    },
+    {
+      "epoch": 0.3807069219440353,
+      "grad_norm": 0.5556607246398926,
+      "learning_rate": 9.978203235311088e-06,
+      "loss": 0.4683,
+      "step": 517
+    },
+    {
+      "epoch": 0.38144329896907214,
+      "grad_norm": 0.5219265818595886,
+      "learning_rate": 9.97780175763656e-06,
+      "loss": 0.4449,
+      "step": 518
+    },
+    {
+      "epoch": 0.38217967599410896,
+      "grad_norm": 0.5306047201156616,
+      "learning_rate": 9.977396624414848e-06,
+      "loss": 0.439,
+      "step": 519
+    },
+    {
+      "epoch": 0.38291605301914583,
+      "grad_norm": 0.638886570930481,
+      "learning_rate": 9.976987835943465e-06,
+      "loss": 0.4803,
+      "step": 520
+    },
+    {
+      "epoch": 0.38365243004418265,
+      "grad_norm": 0.513490617275238,
+      "learning_rate": 9.976575392522617e-06,
+      "loss": 0.4631,
+      "step": 521
+    },
+    {
+      "epoch": 0.38438880706921946,
+      "grad_norm": 0.5516048073768616,
+      "learning_rate": 9.976159294455186e-06,
+      "loss": 0.4958,
+      "step": 522
+    },
+    {
+      "epoch": 0.3851251840942563,
+      "grad_norm": 0.48548823595046997,
+      "learning_rate": 9.975739542046742e-06,
+      "loss": 0.5089,
+      "step": 523
+    },
+    {
+      "epoch": 0.3858615611192931,
+      "grad_norm": 0.5853198170661926,
+      "learning_rate": 9.975316135605543e-06,
+      "loss": 0.4525,
+      "step": 524
+    },
+    {
+      "epoch": 0.3865979381443299,
+      "grad_norm": 0.513939619064331,
+      "learning_rate": 9.97488907544252e-06,
+      "loss": 0.466,
+      "step": 525
+    },
+    {
+      "epoch": 0.3873343151693667,
+      "grad_norm": 0.5338413119316101,
+      "learning_rate": 9.974458361871299e-06,
+      "loss": 0.478,
+      "step": 526
+    },
+    {
+      "epoch": 0.38807069219440354,
+      "grad_norm": 0.4851873815059662,
+      "learning_rate": 9.974023995208177e-06,
+      "loss": 0.4578,
+      "step": 527
+    },
+    {
+      "epoch": 0.38880706921944036,
+      "grad_norm": 0.4788038730621338,
+      "learning_rate": 9.973585975772144e-06,
+      "loss": 0.4351,
+      "step": 528
+    },
+    {
+      "epoch": 0.3895434462444772,
+      "grad_norm": 0.5574104189872742,
+      "learning_rate": 9.973144303884867e-06,
+      "loss": 0.4512,
+      "step": 529
+    },
+    {
+      "epoch": 0.390279823269514,
+      "grad_norm": 0.5594552755355835,
+      "learning_rate": 9.972698979870698e-06,
+      "loss": 0.4996,
+      "step": 530
+    },
+    {
+      "epoch": 0.3910162002945508,
+      "grad_norm": 0.5972340703010559,
+      "learning_rate": 9.972250004056665e-06,
+      "loss": 0.4908,
+      "step": 531
+    },
+    {
+      "epoch": 0.3917525773195876,
+      "grad_norm": 0.5098032355308533,
+      "learning_rate": 9.971797376772488e-06,
+      "loss": 0.4873,
+      "step": 532
+    },
+    {
+      "epoch": 0.39248895434462444,
+      "grad_norm": 0.5212668776512146,
+      "learning_rate": 9.971341098350557e-06,
+      "loss": 0.4603,
+      "step": 533
+    },
+    {
+      "epoch": 0.39322533136966126,
+      "grad_norm": 0.49205219745635986,
+      "learning_rate": 9.970881169125955e-06,
+      "loss": 0.4892,
+      "step": 534
+    },
+    {
+      "epoch": 0.3939617083946981,
+      "grad_norm": 0.4715671241283417,
+      "learning_rate": 9.970417589436435e-06,
+      "loss": 0.4701,
+      "step": 535
+    },
+    {
+      "epoch": 0.3946980854197349,
+      "grad_norm": 0.49158528447151184,
+      "learning_rate": 9.969950359622438e-06,
+      "loss": 0.449,
+      "step": 536
+    },
+    {
+      "epoch": 0.3954344624447717,
+      "grad_norm": 0.5873647928237915,
+      "learning_rate": 9.969479480027086e-06,
+      "loss": 0.4757,
+      "step": 537
+    },
+    {
+      "epoch": 0.3961708394698085,
+      "grad_norm": 0.4452936053276062,
+      "learning_rate": 9.969004950996175e-06,
+      "loss": 0.4569,
+      "step": 538
+    },
+    {
+      "epoch": 0.39690721649484534,
+      "grad_norm": 0.5542396306991577,
+      "learning_rate": 9.968526772878185e-06,
+      "loss": 0.4877,
+      "step": 539
+    },
+    {
+      "epoch": 0.39764359351988215,
+      "grad_norm": 0.5141607522964478,
+      "learning_rate": 9.968044946024277e-06,
+      "loss": 0.4906,
+      "step": 540
+    },
+    {
+      "epoch": 0.39837997054491897,
+      "grad_norm": 0.47367358207702637,
+      "learning_rate": 9.967559470788292e-06,
+      "loss": 0.4683,
+      "step": 541
+    },
+    {
+      "epoch": 0.39911634756995584,
+      "grad_norm": 0.5380948781967163,
+      "learning_rate": 9.967070347526743e-06,
+      "loss": 0.4878,
+      "step": 542
+    },
+    {
+      "epoch": 0.39985272459499266,
+      "grad_norm": 0.46736347675323486,
+      "learning_rate": 9.966577576598833e-06,
+      "loss": 0.4865,
+      "step": 543
+    },
+    {
+      "epoch": 0.4005891016200295,
+      "grad_norm": 0.46713143587112427,
+      "learning_rate": 9.966081158366434e-06,
+      "loss": 0.4388,
+      "step": 544
+    },
+    {
+      "epoch": 0.4013254786450663,
+      "grad_norm": 0.5499324798583984,
+      "learning_rate": 9.965581093194103e-06,
+      "loss": 0.4916,
+      "step": 545
+    },
+    {
+      "epoch": 0.4020618556701031,
+      "grad_norm": 0.45428466796875,
+      "learning_rate": 9.965077381449073e-06,
+      "loss": 0.496,
+      "step": 546
+    },
+    {
+      "epoch": 0.4027982326951399,
+      "grad_norm": 0.58051598072052,
+      "learning_rate": 9.96457002350125e-06,
+      "loss": 0.4618,
+      "step": 547
+    },
+    {
+      "epoch": 0.40353460972017674,
+      "grad_norm": 0.4578842520713806,
+      "learning_rate": 9.96405901972323e-06,
+      "loss": 0.4808,
+      "step": 548
+    },
+    {
+      "epoch": 0.40427098674521356,
+      "grad_norm": 0.5422972440719604,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 0.4721,
+      "step": 549
+    },
+    {
+      "epoch": 0.4050073637702504,
+      "grad_norm": 0.5694018602371216,
+      "learning_rate": 9.96302607618032e-06,
+      "loss": 0.4825,
+      "step": 550
+    },
+    {
+      "epoch": 0.4057437407952872,
+      "grad_norm": 0.5665662884712219,
+      "learning_rate": 9.962504137173997e-06,
+      "loss": 0.4877,
+      "step": 551
+    },
+    {
+      "epoch": 0.406480117820324,
+      "grad_norm": 0.5937358140945435,
+      "learning_rate": 9.961978553854597e-06,
+      "loss": 0.454,
+      "step": 552
+    },
+    {
+      "epoch": 0.4072164948453608,
+      "grad_norm": 0.5692036747932434,
+      "learning_rate": 9.961449326608093e-06,
+      "loss": 0.45,
+      "step": 553
+    },
+    {
+      "epoch": 0.40795287187039764,
+      "grad_norm": 0.5986254811286926,
+      "learning_rate": 9.960916455823134e-06,
+      "loss": 0.4764,
+      "step": 554
+    },
+    {
+      "epoch": 0.40868924889543445,
+      "grad_norm": 0.540023922920227,
+      "learning_rate": 9.960379941891043e-06,
+      "loss": 0.4402,
+      "step": 555
+    },
+    {
+      "epoch": 0.40942562592047127,
+      "grad_norm": 0.620135486125946,
+      "learning_rate": 9.959839785205821e-06,
+      "loss": 0.4795,
+      "step": 556
+    },
+    {
+      "epoch": 0.4101620029455081,
+      "grad_norm": 0.5036307573318481,
+      "learning_rate": 9.959295986164139e-06,
+      "loss": 0.4725,
+      "step": 557
+    },
+    {
+      "epoch": 0.4108983799705449,
+      "grad_norm": 0.576543390750885,
+      "learning_rate": 9.958748545165353e-06,
+      "loss": 0.4593,
+      "step": 558
+    },
+    {
+      "epoch": 0.4116347569955817,
+      "grad_norm": 0.45533323287963867,
+      "learning_rate": 9.95819746261148e-06,
+      "loss": 0.4724,
+      "step": 559
+    },
+    {
+      "epoch": 0.41237113402061853,
+      "grad_norm": 0.5245439410209656,
+      "learning_rate": 9.957642738907226e-06,
+      "loss": 0.4688,
+      "step": 560
+    },
+    {
+      "epoch": 0.41310751104565535,
+      "grad_norm": 0.4937553107738495,
+      "learning_rate": 9.957084374459957e-06,
+      "loss": 0.48,
+      "step": 561
+    },
+    {
+      "epoch": 0.41384388807069217,
+      "grad_norm": 0.5153213739395142,
+      "learning_rate": 9.956522369679722e-06,
+      "loss": 0.4745,
+      "step": 562
+    },
+    {
+      "epoch": 0.41458026509572904,
+      "grad_norm": 0.5216047763824463,
+      "learning_rate": 9.955956724979239e-06,
+      "loss": 0.4751,
+      "step": 563
+    },
+    {
+      "epoch": 0.41531664212076586,
+      "grad_norm": 0.6182007789611816,
+      "learning_rate": 9.955387440773902e-06,
+      "loss": 0.4892,
+      "step": 564
+    },
+    {
+      "epoch": 0.41605301914580267,
+      "grad_norm": 0.5679191946983337,
+      "learning_rate": 9.954814517481774e-06,
+      "loss": 0.4579,
+      "step": 565
+    },
+    {
+      "epoch": 0.4167893961708395,
+      "grad_norm": 0.6047842502593994,
+      "learning_rate": 9.954237955523593e-06,
+      "loss": 0.5028,
+      "step": 566
+    },
+    {
+      "epoch": 0.4175257731958763,
+      "grad_norm": 0.6306226253509521,
+      "learning_rate": 9.953657755322772e-06,
+      "loss": 0.4831,
+      "step": 567
+    },
+    {
+      "epoch": 0.4182621502209131,
+      "grad_norm": 0.47558221220970154,
+      "learning_rate": 9.953073917305386e-06,
+      "loss": 0.4821,
+      "step": 568
+    },
+    {
+      "epoch": 0.41899852724594994,
+      "grad_norm": 0.725439190864563,
+      "learning_rate": 9.952486441900196e-06,
+      "loss": 0.4881,
+      "step": 569
+    },
+    {
+      "epoch": 0.41973490427098675,
+      "grad_norm": 0.5623874664306641,
+      "learning_rate": 9.95189532953862e-06,
+      "loss": 0.4738,
+      "step": 570
+    },
+    {
+      "epoch": 0.42047128129602357,
+      "grad_norm": 0.6907193064689636,
+      "learning_rate": 9.951300580654756e-06,
+      "loss": 0.4706,
+      "step": 571
+    },
+    {
+      "epoch": 0.4212076583210604,
+      "grad_norm": 0.6095152497291565,
+      "learning_rate": 9.950702195685366e-06,
+      "loss": 0.4737,
+      "step": 572
+    },
+    {
+      "epoch": 0.4219440353460972,
+      "grad_norm": 0.6079631447792053,
+      "learning_rate": 9.95010017506989e-06,
+      "loss": 0.4676,
+      "step": 573
+    },
+    {
+      "epoch": 0.422680412371134,
+      "grad_norm": 0.825157880783081,
+      "learning_rate": 9.949494519250433e-06,
+      "loss": 0.4791,
+      "step": 574
+    },
+    {
+      "epoch": 0.42341678939617083,
+      "grad_norm": 0.6472348570823669,
+      "learning_rate": 9.94888522867177e-06,
+      "loss": 0.4969,
+      "step": 575
+    },
+    {
+      "epoch": 0.42415316642120765,
+      "grad_norm": 0.6898149847984314,
+      "learning_rate": 9.948272303781346e-06,
+      "loss": 0.4648,
+      "step": 576
+    },
+    {
+      "epoch": 0.42488954344624447,
+      "grad_norm": 0.5716066360473633,
+      "learning_rate": 9.94765574502927e-06,
+      "loss": 0.4745,
+      "step": 577
+    },
+    {
+      "epoch": 0.4256259204712813,
+      "grad_norm": 0.6825960874557495,
+      "learning_rate": 9.94703555286833e-06,
+      "loss": 0.4888,
+      "step": 578
+    },
+    {
+      "epoch": 0.4263622974963181,
+      "grad_norm": 0.5997359156608582,
+      "learning_rate": 9.946411727753975e-06,
+      "loss": 0.4807,
+      "step": 579
+    },
+    {
+      "epoch": 0.4270986745213549,
+      "grad_norm": 0.5727107524871826,
+      "learning_rate": 9.945784270144321e-06,
+      "loss": 0.4724,
+      "step": 580
+    },
+    {
+      "epoch": 0.42783505154639173,
+      "grad_norm": 0.5531638264656067,
+      "learning_rate": 9.945153180500157e-06,
+      "loss": 0.4379,
+      "step": 581
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.5092889070510864,
+      "learning_rate": 9.944518459284934e-06,
+      "loss": 0.4745,
+      "step": 582
+    },
+    {
+      "epoch": 0.42930780559646536,
+      "grad_norm": 0.5107025504112244,
+      "learning_rate": 9.943880106964772e-06,
+      "loss": 0.494,
+      "step": 583
+    },
+    {
+      "epoch": 0.43004418262150224,
+      "grad_norm": 0.5222765803337097,
+      "learning_rate": 9.943238124008458e-06,
+      "loss": 0.4561,
+      "step": 584
+    },
+    {
+      "epoch": 0.43078055964653905,
+      "grad_norm": 0.550485372543335,
+      "learning_rate": 9.942592510887448e-06,
+      "loss": 0.4548,
+      "step": 585
+    },
+    {
+      "epoch": 0.43151693667157587,
+      "grad_norm": 0.5831854939460754,
+      "learning_rate": 9.941943268075855e-06,
+      "loss": 0.4955,
+      "step": 586
+    },
+    {
+      "epoch": 0.4322533136966127,
+      "grad_norm": 0.4990431070327759,
+      "learning_rate": 9.941290396050467e-06,
+      "loss": 0.49,
+      "step": 587
+    },
+    {
+      "epoch": 0.4329896907216495,
+      "grad_norm": 0.6468625664710999,
+      "learning_rate": 9.940633895290733e-06,
+      "loss": 0.4799,
+      "step": 588
+    },
+    {
+      "epoch": 0.4337260677466863,
+      "grad_norm": 0.5194735527038574,
+      "learning_rate": 9.939973766278768e-06,
+      "loss": 0.4453,
+      "step": 589
+    },
+    {
+      "epoch": 0.43446244477172313,
+      "grad_norm": 0.5722482204437256,
+      "learning_rate": 9.939310009499348e-06,
+      "loss": 0.4798,
+      "step": 590
+    },
+    {
+      "epoch": 0.43519882179675995,
+      "grad_norm": 0.5742457509040833,
+      "learning_rate": 9.938642625439918e-06,
+      "loss": 0.4668,
+      "step": 591
+    },
+    {
+      "epoch": 0.43593519882179677,
+      "grad_norm": 0.5003108382225037,
+      "learning_rate": 9.937971614590587e-06,
+      "loss": 0.4648,
+      "step": 592
+    },
+    {
+      "epoch": 0.4366715758468336,
+      "grad_norm": 0.5724197030067444,
+      "learning_rate": 9.93729697744412e-06,
+      "loss": 0.4841,
+      "step": 593
+    },
+    {
+      "epoch": 0.4374079528718704,
+      "grad_norm": 0.5018848180770874,
+      "learning_rate": 9.936618714495954e-06,
+      "loss": 0.4589,
+      "step": 594
+    },
+    {
+      "epoch": 0.4381443298969072,
+      "grad_norm": 0.5396544337272644,
+      "learning_rate": 9.935936826244183e-06,
+      "loss": 0.4753,
+      "step": 595
+    },
+    {
+      "epoch": 0.43888070692194403,
+      "grad_norm": 0.5358166694641113,
+      "learning_rate": 9.935251313189564e-06,
+      "loss": 0.4915,
+      "step": 596
+    },
+    {
+      "epoch": 0.43961708394698085,
+      "grad_norm": 0.4927530884742737,
+      "learning_rate": 9.934562175835521e-06,
+      "loss": 0.4586,
+      "step": 597
+    },
+    {
+      "epoch": 0.44035346097201766,
+      "grad_norm": 0.47724655270576477,
+      "learning_rate": 9.933869414688134e-06,
+      "loss": 0.4734,
+      "step": 598
+    },
+    {
+      "epoch": 0.4410898379970545,
+      "grad_norm": 0.46607303619384766,
+      "learning_rate": 9.933173030256142e-06,
+      "loss": 0.4718,
+      "step": 599
+    },
+    {
+      "epoch": 0.4418262150220913,
+      "grad_norm": 0.5244195461273193,
+      "learning_rate": 9.932473023050954e-06,
+      "loss": 0.4441,
+      "step": 600
+    },
+    {
+      "epoch": 0.4425625920471281,
+      "grad_norm": 0.5027062892913818,
+      "learning_rate": 9.931769393586633e-06,
+      "loss": 0.4456,
+      "step": 601
+    },
+    {
+      "epoch": 0.44329896907216493,
+      "grad_norm": 0.4999453127384186,
+      "learning_rate": 9.9310621423799e-06,
+      "loss": 0.4858,
+      "step": 602
+    },
+    {
+      "epoch": 0.44403534609720174,
+      "grad_norm": 0.5346736311912537,
+      "learning_rate": 9.930351269950144e-06,
+      "loss": 0.4985,
+      "step": 603
+    },
+    {
+      "epoch": 0.44477172312223856,
+      "grad_norm": 0.5561801791191101,
+      "learning_rate": 9.929636776819404e-06,
+      "loss": 0.4554,
+      "step": 604
+    },
+    {
+      "epoch": 0.44550810014727543,
+      "grad_norm": 0.5683428049087524,
+      "learning_rate": 9.928918663512382e-06,
+      "loss": 0.4764,
+      "step": 605
+    },
+    {
+      "epoch": 0.44624447717231225,
+      "grad_norm": 0.5989724397659302,
+      "learning_rate": 9.928196930556442e-06,
+      "loss": 0.473,
+      "step": 606
+    },
+    {
+      "epoch": 0.44698085419734906,
+      "grad_norm": 0.5784966349601746,
+      "learning_rate": 9.9274715784816e-06,
+      "loss": 0.4653,
+      "step": 607
+    },
+    {
+      "epoch": 0.4477172312223859,
+      "grad_norm": 0.5616557598114014,
+      "learning_rate": 9.926742607820535e-06,
+      "loss": 0.4722,
+      "step": 608
+    },
+    {
+      "epoch": 0.4484536082474227,
+      "grad_norm": 0.6208142638206482,
+      "learning_rate": 9.926010019108579e-06,
+      "loss": 0.4639,
+      "step": 609
+    },
+    {
+      "epoch": 0.4491899852724595,
+      "grad_norm": 0.5640535354614258,
+      "learning_rate": 9.925273812883724e-06,
+      "loss": 0.4703,
+      "step": 610
+    },
+    {
+      "epoch": 0.44992636229749633,
+      "grad_norm": 0.7158654928207397,
+      "learning_rate": 9.924533989686618e-06,
+      "loss": 0.4697,
+      "step": 611
+    },
+    {
+      "epoch": 0.45066273932253315,
+      "grad_norm": 0.6342428922653198,
+      "learning_rate": 9.923790550060564e-06,
+      "loss": 0.4662,
+      "step": 612
+    },
+    {
+      "epoch": 0.45139911634756996,
+      "grad_norm": 0.4764558970928192,
+      "learning_rate": 9.923043494551522e-06,
+      "loss": 0.4519,
+      "step": 613
+    },
+    {
+      "epoch": 0.4521354933726068,
+      "grad_norm": 0.6603193879127502,
+      "learning_rate": 9.922292823708106e-06,
+      "loss": 0.5068,
+      "step": 614
+    },
+    {
+      "epoch": 0.4528718703976436,
+      "grad_norm": 0.522085428237915,
+      "learning_rate": 9.921538538081588e-06,
+      "loss": 0.4602,
+      "step": 615
+    },
+    {
+      "epoch": 0.4536082474226804,
+      "grad_norm": 0.53127121925354,
+      "learning_rate": 9.92078063822589e-06,
+      "loss": 0.461,
+      "step": 616
+    },
+    {
+      "epoch": 0.4543446244477172,
+      "grad_norm": 0.6258383393287659,
+      "learning_rate": 9.920019124697594e-06,
+      "loss": 0.489,
+      "step": 617
+    },
+    {
+      "epoch": 0.45508100147275404,
+      "grad_norm": 0.49485811591148376,
+      "learning_rate": 9.919253998055928e-06,
+      "loss": 0.447,
+      "step": 618
+    },
+    {
+      "epoch": 0.45581737849779086,
+      "grad_norm": 0.5272852182388306,
+      "learning_rate": 9.918485258862781e-06,
+      "loss": 0.478,
+      "step": 619
+    },
+    {
+      "epoch": 0.4565537555228277,
+      "grad_norm": 0.4806794822216034,
+      "learning_rate": 9.917712907682694e-06,
+      "loss": 0.4353,
+      "step": 620
+    },
+    {
+      "epoch": 0.4572901325478645,
+      "grad_norm": 0.501160204410553,
+      "learning_rate": 9.916936945082854e-06,
+      "loss": 0.4682,
+      "step": 621
+    },
+    {
+      "epoch": 0.4580265095729013,
+      "grad_norm": 0.5172646641731262,
+      "learning_rate": 9.916157371633106e-06,
+      "loss": 0.4505,
+      "step": 622
+    },
+    {
+      "epoch": 0.4587628865979381,
+      "grad_norm": 0.5966675281524658,
+      "learning_rate": 9.915374187905945e-06,
+      "loss": 0.5016,
+      "step": 623
+    },
+    {
+      "epoch": 0.45949926362297494,
+      "grad_norm": 0.4914330840110779,
+      "learning_rate": 9.91458739447652e-06,
+      "loss": 0.4455,
+      "step": 624
+    },
+    {
+      "epoch": 0.46023564064801176,
+      "grad_norm": 0.6271175146102905,
+      "learning_rate": 9.913796991922624e-06,
+      "loss": 0.4604,
+      "step": 625
+    },
+    {
+      "epoch": 0.46097201767304863,
+      "grad_norm": 0.5928018689155579,
+      "learning_rate": 9.913002980824709e-06,
+      "loss": 0.4519,
+      "step": 626
+    },
+    {
+      "epoch": 0.46170839469808544,
+      "grad_norm": 0.4849533140659332,
+      "learning_rate": 9.912205361765868e-06,
+      "loss": 0.5011,
+      "step": 627
+    },
+    {
+      "epoch": 0.46244477172312226,
+      "grad_norm": 0.5498746633529663,
+      "learning_rate": 9.911404135331852e-06,
+      "loss": 0.4567,
+      "step": 628
+    },
+    {
+      "epoch": 0.4631811487481591,
+      "grad_norm": 0.4846838712692261,
+      "learning_rate": 9.910599302111057e-06,
+      "loss": 0.4607,
+      "step": 629
+    },
+    {
+      "epoch": 0.4639175257731959,
+      "grad_norm": 0.5435061454772949,
+      "learning_rate": 9.909790862694528e-06,
+      "loss": 0.4507,
+      "step": 630
+    },
+    {
+      "epoch": 0.4646539027982327,
+      "grad_norm": 0.5106296539306641,
+      "learning_rate": 9.908978817675959e-06,
+      "loss": 0.4366,
+      "step": 631
+    },
+    {
+      "epoch": 0.4653902798232695,
+      "grad_norm": 0.4310411512851715,
+      "learning_rate": 9.908163167651688e-06,
+      "loss": 0.4729,
+      "step": 632
+    },
+    {
+      "epoch": 0.46612665684830634,
+      "grad_norm": 0.5732407569885254,
+      "learning_rate": 9.907343913220707e-06,
+      "loss": 0.4729,
+      "step": 633
+    },
+    {
+      "epoch": 0.46686303387334316,
+      "grad_norm": 0.6053887009620667,
+      "learning_rate": 9.90652105498465e-06,
+      "loss": 0.4671,
+      "step": 634
+    },
+    {
+      "epoch": 0.46759941089838,
+      "grad_norm": 0.4543951153755188,
+      "learning_rate": 9.905694593547803e-06,
+      "loss": 0.4631,
+      "step": 635
+    },
+    {
+      "epoch": 0.4683357879234168,
+      "grad_norm": 0.5218227505683899,
+      "learning_rate": 9.90486452951709e-06,
+      "loss": 0.4638,
+      "step": 636
+    },
+    {
+      "epoch": 0.4690721649484536,
+      "grad_norm": 0.5472959876060486,
+      "learning_rate": 9.904030863502086e-06,
+      "loss": 0.4677,
+      "step": 637
+    },
+    {
+      "epoch": 0.4698085419734904,
+      "grad_norm": 0.5236535668373108,
+      "learning_rate": 9.903193596115011e-06,
+      "loss": 0.466,
+      "step": 638
+    },
+    {
+      "epoch": 0.47054491899852724,
+      "grad_norm": 0.5994406342506409,
+      "learning_rate": 9.902352727970729e-06,
+      "loss": 0.4811,
+      "step": 639
+    },
+    {
+      "epoch": 0.47128129602356406,
+      "grad_norm": 0.5658840537071228,
+      "learning_rate": 9.901508259686746e-06,
+      "loss": 0.4728,
+      "step": 640
+    },
+    {
+      "epoch": 0.47201767304860087,
+      "grad_norm": 0.5107308030128479,
+      "learning_rate": 9.900660191883217e-06,
+      "loss": 0.4834,
+      "step": 641
+    },
+    {
+      "epoch": 0.4727540500736377,
+      "grad_norm": 0.5555645227432251,
+      "learning_rate": 9.899808525182935e-06,
+      "loss": 0.471,
+      "step": 642
+    },
+    {
+      "epoch": 0.4734904270986745,
+      "grad_norm": 0.6290595531463623,
+      "learning_rate": 9.89895326021134e-06,
+      "loss": 0.4718,
+      "step": 643
+    },
+    {
+      "epoch": 0.4742268041237113,
+      "grad_norm": 0.5025970935821533,
+      "learning_rate": 9.898094397596511e-06,
+      "loss": 0.4752,
+      "step": 644
+    },
+    {
+      "epoch": 0.47496318114874814,
+      "grad_norm": 0.60362309217453,
+      "learning_rate": 9.897231937969172e-06,
+      "loss": 0.4636,
+      "step": 645
+    },
+    {
+      "epoch": 0.47569955817378495,
+      "grad_norm": 0.6537862420082092,
+      "learning_rate": 9.896365881962687e-06,
+      "loss": 0.4574,
+      "step": 646
+    },
+    {
+      "epoch": 0.47643593519882177,
+      "grad_norm": 0.4715963304042816,
+      "learning_rate": 9.895496230213061e-06,
+      "loss": 0.4596,
+      "step": 647
+    },
+    {
+      "epoch": 0.47717231222385864,
+      "grad_norm": 0.5870654582977295,
+      "learning_rate": 9.894622983358941e-06,
+      "loss": 0.4472,
+      "step": 648
+    },
+    {
+      "epoch": 0.47790868924889546,
+      "grad_norm": 0.6155905723571777,
+      "learning_rate": 9.893746142041612e-06,
+      "loss": 0.4841,
+      "step": 649
+    },
+    {
+      "epoch": 0.4786450662739323,
+      "grad_norm": 0.5059301853179932,
+      "learning_rate": 9.892865706905e-06,
+      "loss": 0.454,
+      "step": 650
+    },
+    {
+      "epoch": 0.4793814432989691,
+      "grad_norm": 0.6735835671424866,
+      "learning_rate": 9.891981678595671e-06,
+      "loss": 0.4788,
+      "step": 651
+    },
+    {
+      "epoch": 0.4801178203240059,
+      "grad_norm": 0.6185165643692017,
+      "learning_rate": 9.891094057762827e-06,
+      "loss": 0.4509,
+      "step": 652
+    },
+    {
+      "epoch": 0.4808541973490427,
+      "grad_norm": 0.5275318026542664,
+      "learning_rate": 9.89020284505831e-06,
+      "loss": 0.4064,
+      "step": 653
+    },
+    {
+      "epoch": 0.48159057437407954,
+      "grad_norm": 0.5368750095367432,
+      "learning_rate": 9.889308041136601e-06,
+      "loss": 0.4633,
+      "step": 654
+    },
+    {
+      "epoch": 0.48232695139911635,
+      "grad_norm": 0.6606622934341431,
+      "learning_rate": 9.888409646654818e-06,
+      "loss": 0.4619,
+      "step": 655
+    },
+    {
+      "epoch": 0.48306332842415317,
+      "grad_norm": 0.6033420562744141,
+      "learning_rate": 9.88750766227271e-06,
+      "loss": 0.4549,
+      "step": 656
+    },
+    {
+      "epoch": 0.48379970544919,
+      "grad_norm": 0.5704071521759033,
+      "learning_rate": 9.886602088652672e-06,
+      "loss": 0.4584,
+      "step": 657
+    },
+    {
+      "epoch": 0.4845360824742268,
+      "grad_norm": 0.6391714811325073,
+      "learning_rate": 9.885692926459729e-06,
+      "loss": 0.4612,
+      "step": 658
+    },
+    {
+      "epoch": 0.4852724594992636,
+      "grad_norm": 0.49151089787483215,
+      "learning_rate": 9.88478017636154e-06,
+      "loss": 0.4524,
+      "step": 659
+    },
+    {
+      "epoch": 0.48600883652430044,
+      "grad_norm": 0.5738299489021301,
+      "learning_rate": 9.883863839028402e-06,
+      "loss": 0.4854,
+      "step": 660
+    },
+    {
+      "epoch": 0.48674521354933725,
+      "grad_norm": 0.5694648027420044,
+      "learning_rate": 9.882943915133247e-06,
+      "loss": 0.4583,
+      "step": 661
+    },
+    {
+      "epoch": 0.48748159057437407,
+      "grad_norm": 0.5113805532455444,
+      "learning_rate": 9.88202040535164e-06,
+      "loss": 0.4714,
+      "step": 662
+    },
+    {
+      "epoch": 0.4882179675994109,
+      "grad_norm": 0.5874478220939636,
+      "learning_rate": 9.881093310361773e-06,
+      "loss": 0.4851,
+      "step": 663
+    },
+    {
+      "epoch": 0.4889543446244477,
+      "grad_norm": 0.5066226124763489,
+      "learning_rate": 9.880162630844483e-06,
+      "loss": 0.4644,
+      "step": 664
+    },
+    {
+      "epoch": 0.4896907216494845,
+      "grad_norm": 0.5173943042755127,
+      "learning_rate": 9.879228367483228e-06,
+      "loss": 0.4526,
+      "step": 665
+    },
+    {
+      "epoch": 0.49042709867452133,
+      "grad_norm": 0.531655490398407,
+      "learning_rate": 9.878290520964107e-06,
+      "loss": 0.4503,
+      "step": 666
+    },
+    {
+      "epoch": 0.49116347569955815,
+      "grad_norm": 0.5501636266708374,
+      "learning_rate": 9.877349091975844e-06,
+      "loss": 0.4694,
+      "step": 667
+    },
+    {
+      "epoch": 0.49189985272459497,
+      "grad_norm": 0.532542884349823,
+      "learning_rate": 9.876404081209796e-06,
+      "loss": 0.4588,
+      "step": 668
+    },
+    {
+      "epoch": 0.49263622974963184,
+      "grad_norm": 0.4820326864719391,
+      "learning_rate": 9.87545548935995e-06,
+      "loss": 0.4698,
+      "step": 669
+    },
+    {
+      "epoch": 0.49337260677466865,
+      "grad_norm": 0.4267001152038574,
+      "learning_rate": 9.874503317122925e-06,
+      "loss": 0.4621,
+      "step": 670
+    },
+    {
+      "epoch": 0.49410898379970547,
+      "grad_norm": 0.4171275198459625,
+      "learning_rate": 9.873547565197965e-06,
+      "loss": 0.4448,
+      "step": 671
+    },
+    {
+      "epoch": 0.4948453608247423,
+      "grad_norm": 0.4105445146560669,
+      "learning_rate": 9.872588234286946e-06,
+      "loss": 0.4507,
+      "step": 672
+    },
+    {
+      "epoch": 0.4955817378497791,
+      "grad_norm": 0.46572741866111755,
+      "learning_rate": 9.871625325094375e-06,
+      "loss": 0.4485,
+      "step": 673
+    },
+    {
+      "epoch": 0.4963181148748159,
+      "grad_norm": 0.44582173228263855,
+      "learning_rate": 9.870658838327378e-06,
+      "loss": 0.4712,
+      "step": 674
+    },
+    {
+      "epoch": 0.49705449189985274,
+      "grad_norm": 0.5541704893112183,
+      "learning_rate": 9.869688774695719e-06,
+      "loss": 0.4527,
+      "step": 675
+    },
+    {
+      "epoch": 0.49779086892488955,
+      "grad_norm": 0.45377492904663086,
+      "learning_rate": 9.86871513491178e-06,
+      "loss": 0.4764,
+      "step": 676
+    },
+    {
+      "epoch": 0.49852724594992637,
+      "grad_norm": 0.5734421610832214,
+      "learning_rate": 9.867737919690573e-06,
+      "loss": 0.4822,
+      "step": 677
+    },
+    {
+      "epoch": 0.4992636229749632,
+      "grad_norm": 0.49860337376594543,
+      "learning_rate": 9.866757129749733e-06,
+      "loss": 0.4502,
+      "step": 678
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4667682945728302,
+      "learning_rate": 9.865772765809528e-06,
+      "loss": 0.4951,
+      "step": 679
+    },
+    {
+      "epoch": 0.5007363770250368,
+      "grad_norm": 0.5397236347198486,
+      "learning_rate": 9.864784828592842e-06,
+      "loss": 0.4939,
+      "step": 680
+    },
+    {
+      "epoch": 0.5014727540500736,
+      "grad_norm": 0.5122597813606262,
+      "learning_rate": 9.863793318825186e-06,
+      "loss": 0.4452,
+      "step": 681
+    },
+    {
+      "epoch": 0.5022091310751104,
+      "grad_norm": 0.4826493263244629,
+      "learning_rate": 9.862798237234697e-06,
+      "loss": 0.4731,
+      "step": 682
+    },
+    {
+      "epoch": 0.5029455081001473,
+      "grad_norm": 0.4937388300895691,
+      "learning_rate": 9.86179958455213e-06,
+      "loss": 0.4707,
+      "step": 683
+    },
+    {
+      "epoch": 0.5036818851251841,
+      "grad_norm": 0.49966996908187866,
+      "learning_rate": 9.860797361510867e-06,
+      "loss": 0.4764,
+      "step": 684
+    },
+    {
+      "epoch": 0.5044182621502209,
+      "grad_norm": 0.41720449924468994,
+      "learning_rate": 9.859791568846908e-06,
+      "loss": 0.458,
+      "step": 685
+    },
+    {
+      "epoch": 0.5051546391752577,
+      "grad_norm": 0.5433468222618103,
+      "learning_rate": 9.858782207298881e-06,
+      "loss": 0.4631,
+      "step": 686
+    },
+    {
+      "epoch": 0.5058910162002945,
+      "grad_norm": 0.5088586807250977,
+      "learning_rate": 9.857769277608027e-06,
+      "loss": 0.4453,
+      "step": 687
+    },
+    {
+      "epoch": 0.5066273932253313,
+      "grad_norm": 0.5024508237838745,
+      "learning_rate": 9.856752780518214e-06,
+      "loss": 0.4519,
+      "step": 688
+    },
+    {
+      "epoch": 0.5073637702503682,
+      "grad_norm": 0.5998568534851074,
+      "learning_rate": 9.855732716775923e-06,
+      "loss": 0.463,
+      "step": 689
+    },
+    {
+      "epoch": 0.508100147275405,
+      "grad_norm": 0.5431100130081177,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.4632,
+      "step": 690
+    },
+    {
+      "epoch": 0.5088365243004418,
+      "grad_norm": 0.5772528648376465,
+      "learning_rate": 9.853681892332948e-06,
+      "loss": 0.4629,
+      "step": 691
+    },
+    {
+      "epoch": 0.5095729013254786,
+      "grad_norm": 0.4999123215675354,
+      "learning_rate": 9.852651133138328e-06,
+      "loss": 0.4269,
+      "step": 692
+    },
+    {
+      "epoch": 0.5103092783505154,
+      "grad_norm": 0.6174575686454773,
+      "learning_rate": 9.851616810303359e-06,
+      "loss": 0.4618,
+      "step": 693
+    },
+    {
+      "epoch": 0.5110456553755522,
+      "grad_norm": 0.5877047777175903,
+      "learning_rate": 9.850578924587614e-06,
+      "loss": 0.4857,
+      "step": 694
+    },
+    {
+      "epoch": 0.5117820324005891,
+      "grad_norm": 0.636028528213501,
+      "learning_rate": 9.849537476753286e-06,
+      "loss": 0.4832,
+      "step": 695
+    },
+    {
+      "epoch": 0.5125184094256259,
+      "grad_norm": 0.5791111588478088,
+      "learning_rate": 9.848492467565182e-06,
+      "loss": 0.493,
+      "step": 696
+    },
+    {
+      "epoch": 0.5132547864506627,
+      "grad_norm": 0.557636559009552,
+      "learning_rate": 9.847443897790728e-06,
+      "loss": 0.4739,
+      "step": 697
+    },
+    {
+      "epoch": 0.5139911634756995,
+      "grad_norm": 0.47123220562934875,
+      "learning_rate": 9.84639176819996e-06,
+      "loss": 0.463,
+      "step": 698
+    },
+    {
+      "epoch": 0.5147275405007363,
+      "grad_norm": 0.6015990376472473,
+      "learning_rate": 9.845336079565529e-06,
+      "loss": 0.4914,
+      "step": 699
+    },
+    {
+      "epoch": 0.5154639175257731,
+      "grad_norm": 0.4684229791164398,
+      "learning_rate": 9.844276832662704e-06,
+      "loss": 0.4573,
+      "step": 700
+    },
+    {
+      "epoch": 0.5162002945508101,
+      "grad_norm": 0.4829377233982086,
+      "learning_rate": 9.843214028269361e-06,
+      "loss": 0.4414,
+      "step": 701
+    },
+    {
+      "epoch": 0.5169366715758469,
+      "grad_norm": 0.5336502194404602,
+      "learning_rate": 9.842147667165993e-06,
+      "loss": 0.4746,
+      "step": 702
+    },
+    {
+      "epoch": 0.5176730486008837,
+      "grad_norm": 0.4682151675224304,
+      "learning_rate": 9.841077750135702e-06,
+      "loss": 0.4873,
+      "step": 703
+    },
+    {
+      "epoch": 0.5184094256259205,
+      "grad_norm": 0.5173882842063904,
+      "learning_rate": 9.840004277964204e-06,
+      "loss": 0.4497,
+      "step": 704
+    },
+    {
+      "epoch": 0.5191458026509573,
+      "grad_norm": 0.4670022130012512,
+      "learning_rate": 9.838927251439823e-06,
+      "loss": 0.4857,
+      "step": 705
+    },
+    {
+      "epoch": 0.5198821796759941,
+      "grad_norm": 0.5181158185005188,
+      "learning_rate": 9.837846671353498e-06,
+      "loss": 0.4822,
+      "step": 706
+    },
+    {
+      "epoch": 0.520618556701031,
+      "grad_norm": 0.4926871955394745,
+      "learning_rate": 9.83676253849877e-06,
+      "loss": 0.4677,
+      "step": 707
+    },
+    {
+      "epoch": 0.5213549337260678,
+      "grad_norm": 0.5156863331794739,
+      "learning_rate": 9.835674853671797e-06,
+      "loss": 0.4865,
+      "step": 708
+    },
+    {
+      "epoch": 0.5220913107511046,
+      "grad_norm": 0.5189216136932373,
+      "learning_rate": 9.83458361767134e-06,
+      "loss": 0.4697,
+      "step": 709
+    },
+    {
+      "epoch": 0.5228276877761414,
+      "grad_norm": 0.537659764289856,
+      "learning_rate": 9.83348883129877e-06,
+      "loss": 0.4682,
+      "step": 710
+    },
+    {
+      "epoch": 0.5235640648011782,
+      "grad_norm": 0.5328056812286377,
+      "learning_rate": 9.832390495358066e-06,
+      "loss": 0.464,
+      "step": 711
+    },
+    {
+      "epoch": 0.524300441826215,
+      "grad_norm": 0.5192391872406006,
+      "learning_rate": 9.831288610655812e-06,
+      "loss": 0.451,
+      "step": 712
+    },
+    {
+      "epoch": 0.5250368188512519,
+      "grad_norm": 0.48723873496055603,
+      "learning_rate": 9.830183178001199e-06,
+      "loss": 0.4747,
+      "step": 713
+    },
+    {
+      "epoch": 0.5257731958762887,
+      "grad_norm": 0.594840407371521,
+      "learning_rate": 9.829074198206024e-06,
+      "loss": 0.4407,
+      "step": 714
+    },
+    {
+      "epoch": 0.5265095729013255,
+      "grad_norm": 0.5472460389137268,
+      "learning_rate": 9.827961672084685e-06,
+      "loss": 0.4487,
+      "step": 715
+    },
+    {
+      "epoch": 0.5272459499263623,
+      "grad_norm": 0.5027500987052917,
+      "learning_rate": 9.82684560045419e-06,
+      "loss": 0.4904,
+      "step": 716
+    },
+    {
+      "epoch": 0.5279823269513991,
+      "grad_norm": 0.5732831358909607,
+      "learning_rate": 9.82572598413415e-06,
+      "loss": 0.4505,
+      "step": 717
+    },
+    {
+      "epoch": 0.5287187039764359,
+      "grad_norm": 0.5125916600227356,
+      "learning_rate": 9.824602823946776e-06,
+      "loss": 0.4428,
+      "step": 718
+    },
+    {
+      "epoch": 0.5294550810014728,
+      "grad_norm": 0.5110583901405334,
+      "learning_rate": 9.823476120716882e-06,
+      "loss": 0.4561,
+      "step": 719
+    },
+    {
+      "epoch": 0.5301914580265096,
+      "grad_norm": 0.5457249879837036,
+      "learning_rate": 9.822345875271884e-06,
+      "loss": 0.4834,
+      "step": 720
+    },
+    {
+      "epoch": 0.5309278350515464,
+      "grad_norm": 0.4954557418823242,
+      "learning_rate": 9.821212088441803e-06,
+      "loss": 0.4562,
+      "step": 721
+    },
+    {
+      "epoch": 0.5316642120765832,
+      "grad_norm": 0.6068676710128784,
+      "learning_rate": 9.820074761059255e-06,
+      "loss": 0.5063,
+      "step": 722
+    },
+    {
+      "epoch": 0.53240058910162,
+      "grad_norm": 0.49514177441596985,
+      "learning_rate": 9.81893389395946e-06,
+      "loss": 0.4364,
+      "step": 723
+    },
+    {
+      "epoch": 0.5331369661266568,
+      "grad_norm": 0.5643453598022461,
+      "learning_rate": 9.817789487980237e-06,
+      "loss": 0.4487,
+      "step": 724
+    },
+    {
+      "epoch": 0.5338733431516937,
+      "grad_norm": 0.4869299530982971,
+      "learning_rate": 9.816641543962001e-06,
+      "loss": 0.4697,
+      "step": 725
+    },
+    {
+      "epoch": 0.5346097201767305,
+      "grad_norm": 0.647627055644989,
+      "learning_rate": 9.815490062747773e-06,
+      "loss": 0.5138,
+      "step": 726
+    },
+    {
+      "epoch": 0.5353460972017673,
+      "grad_norm": 0.5390787124633789,
+      "learning_rate": 9.81433504518316e-06,
+      "loss": 0.4643,
+      "step": 727
+    },
+    {
+      "epoch": 0.5360824742268041,
+      "grad_norm": 0.5298830270767212,
+      "learning_rate": 9.813176492116372e-06,
+      "loss": 0.4745,
+      "step": 728
+    },
+    {
+      "epoch": 0.5368188512518409,
+      "grad_norm": 0.5326485633850098,
+      "learning_rate": 9.812014404398219e-06,
+      "loss": 0.4452,
+      "step": 729
+    },
+    {
+      "epoch": 0.5375552282768777,
+      "grad_norm": 0.47868359088897705,
+      "learning_rate": 9.810848782882101e-06,
+      "loss": 0.467,
+      "step": 730
+    },
+    {
+      "epoch": 0.5382916053019146,
+      "grad_norm": 0.5476110577583313,
+      "learning_rate": 9.809679628424016e-06,
+      "loss": 0.4368,
+      "step": 731
+    },
+    {
+      "epoch": 0.5390279823269514,
+      "grad_norm": 0.5308613181114197,
+      "learning_rate": 9.808506941882556e-06,
+      "loss": 0.4712,
+      "step": 732
+    },
+    {
+      "epoch": 0.5397643593519882,
+      "grad_norm": 0.4515834152698517,
+      "learning_rate": 9.807330724118906e-06,
+      "loss": 0.4381,
+      "step": 733
+    },
+    {
+      "epoch": 0.540500736377025,
+      "grad_norm": 0.5537406206130981,
+      "learning_rate": 9.806150975996843e-06,
+      "loss": 0.4615,
+      "step": 734
+    },
+    {
+      "epoch": 0.5412371134020618,
+      "grad_norm": 0.47350478172302246,
+      "learning_rate": 9.80496769838274e-06,
+      "loss": 0.4433,
+      "step": 735
+    },
+    {
+      "epoch": 0.5419734904270986,
+      "grad_norm": 0.4860948920249939,
+      "learning_rate": 9.803780892145562e-06,
+      "loss": 0.4558,
+      "step": 736
+    },
+    {
+      "epoch": 0.5427098674521355,
+      "grad_norm": 0.6019650101661682,
+      "learning_rate": 9.802590558156863e-06,
+      "loss": 0.4454,
+      "step": 737
+    },
+    {
+      "epoch": 0.5434462444771723,
+      "grad_norm": 0.467506468296051,
+      "learning_rate": 9.801396697290786e-06,
+      "loss": 0.4628,
+      "step": 738
+    },
+    {
+      "epoch": 0.5441826215022091,
+      "grad_norm": 0.4791230857372284,
+      "learning_rate": 9.800199310424067e-06,
+      "loss": 0.4563,
+      "step": 739
+    },
+    {
+      "epoch": 0.5449189985272459,
+      "grad_norm": 0.5122166275978088,
+      "learning_rate": 9.798998398436031e-06,
+      "loss": 0.4831,
+      "step": 740
+    },
+    {
+      "epoch": 0.5456553755522827,
+      "grad_norm": 0.47877249121665955,
+      "learning_rate": 9.797793962208593e-06,
+      "loss": 0.465,
+      "step": 741
+    },
+    {
+      "epoch": 0.5463917525773195,
+      "grad_norm": 0.49181488156318665,
+      "learning_rate": 9.796586002626253e-06,
+      "loss": 0.4793,
+      "step": 742
+    },
+    {
+      "epoch": 0.5471281296023565,
+      "grad_norm": 0.48321738839149475,
+      "learning_rate": 9.795374520576102e-06,
+      "loss": 0.4757,
+      "step": 743
+    },
+    {
+      "epoch": 0.5478645066273933,
+      "grad_norm": 0.5315781831741333,
+      "learning_rate": 9.794159516947812e-06,
+      "loss": 0.468,
+      "step": 744
+    },
+    {
+      "epoch": 0.5486008836524301,
+      "grad_norm": 0.4678581655025482,
+      "learning_rate": 9.792940992633649e-06,
+      "loss": 0.4663,
+      "step": 745
+    },
+    {
+      "epoch": 0.5493372606774669,
+      "grad_norm": 0.4767686724662781,
+      "learning_rate": 9.791718948528457e-06,
+      "loss": 0.4687,
+      "step": 746
+    },
+    {
+      "epoch": 0.5500736377025037,
+      "grad_norm": 0.5320640206336975,
+      "learning_rate": 9.790493385529671e-06,
+      "loss": 0.4672,
+      "step": 747
+    },
+    {
+      "epoch": 0.5508100147275405,
+      "grad_norm": 0.45434436202049255,
+      "learning_rate": 9.789264304537307e-06,
+      "loss": 0.442,
+      "step": 748
+    },
+    {
+      "epoch": 0.5515463917525774,
+      "grad_norm": 0.5533967614173889,
+      "learning_rate": 9.788031706453964e-06,
+      "loss": 0.4539,
+      "step": 749
+    },
+    {
+      "epoch": 0.5522827687776142,
+      "grad_norm": 0.5042517781257629,
+      "learning_rate": 9.786795592184824e-06,
+      "loss": 0.4619,
+      "step": 750
+    },
+    {
+      "epoch": 0.553019145802651,
+      "grad_norm": 0.4778529107570648,
+      "learning_rate": 9.785555962637654e-06,
+      "loss": 0.4744,
+      "step": 751
+    },
+    {
+      "epoch": 0.5537555228276878,
+      "grad_norm": 0.5131283402442932,
+      "learning_rate": 9.784312818722799e-06,
+      "loss": 0.4449,
+      "step": 752
+    },
+    {
+      "epoch": 0.5544918998527246,
+      "grad_norm": 0.48384329676628113,
+      "learning_rate": 9.783066161353188e-06,
+      "loss": 0.474,
+      "step": 753
+    },
+    {
+      "epoch": 0.5552282768777614,
+      "grad_norm": 0.49050474166870117,
+      "learning_rate": 9.781815991444326e-06,
+      "loss": 0.4445,
+      "step": 754
+    },
+    {
+      "epoch": 0.5559646539027983,
+      "grad_norm": 0.5421154499053955,
+      "learning_rate": 9.7805623099143e-06,
+      "loss": 0.4668,
+      "step": 755
+    },
+    {
+      "epoch": 0.5567010309278351,
+      "grad_norm": 0.5321326851844788,
+      "learning_rate": 9.779305117683781e-06,
+      "loss": 0.4762,
+      "step": 756
+    },
+    {
+      "epoch": 0.5574374079528719,
+      "grad_norm": 0.5349854230880737,
+      "learning_rate": 9.778044415676007e-06,
+      "loss": 0.4774,
+      "step": 757
+    },
+    {
+      "epoch": 0.5581737849779087,
+      "grad_norm": 0.5067935585975647,
+      "learning_rate": 9.776780204816801e-06,
+      "loss": 0.4862,
+      "step": 758
+    },
+    {
+      "epoch": 0.5589101620029455,
+      "grad_norm": 0.4486694037914276,
+      "learning_rate": 9.775512486034564e-06,
+      "loss": 0.4481,
+      "step": 759
+    },
+    {
+      "epoch": 0.5596465390279823,
+      "grad_norm": 0.5573259592056274,
+      "learning_rate": 9.774241260260266e-06,
+      "loss": 0.4695,
+      "step": 760
+    },
+    {
+      "epoch": 0.5603829160530192,
+      "grad_norm": 0.45694780349731445,
+      "learning_rate": 9.77296652842746e-06,
+      "loss": 0.4728,
+      "step": 761
+    },
+    {
+      "epoch": 0.561119293078056,
+      "grad_norm": 0.46451568603515625,
+      "learning_rate": 9.771688291472269e-06,
+      "loss": 0.458,
+      "step": 762
+    },
+    {
+      "epoch": 0.5618556701030928,
+      "grad_norm": 0.5605217218399048,
+      "learning_rate": 9.770406550333393e-06,
+      "loss": 0.4492,
+      "step": 763
+    },
+    {
+      "epoch": 0.5625920471281296,
+      "grad_norm": 0.4403326213359833,
+      "learning_rate": 9.769121305952102e-06,
+      "loss": 0.4721,
+      "step": 764
+    },
+    {
+      "epoch": 0.5633284241531664,
+      "grad_norm": 0.5193141102790833,
+      "learning_rate": 9.767832559272244e-06,
+      "loss": 0.461,
+      "step": 765
+    },
+    {
+      "epoch": 0.5640648011782032,
+      "grad_norm": 0.49891531467437744,
+      "learning_rate": 9.766540311240232e-06,
+      "loss": 0.451,
+      "step": 766
+    },
+    {
+      "epoch": 0.56480117820324,
+      "grad_norm": 0.5359540581703186,
+      "learning_rate": 9.765244562805055e-06,
+      "loss": 0.4595,
+      "step": 767
+    },
+    {
+      "epoch": 0.5655375552282769,
+      "grad_norm": 0.49058595299720764,
+      "learning_rate": 9.76394531491827e-06,
+      "loss": 0.4896,
+      "step": 768
+    },
+    {
+      "epoch": 0.5662739322533137,
+      "grad_norm": 0.4365098476409912,
+      "learning_rate": 9.762642568534012e-06,
+      "loss": 0.4575,
+      "step": 769
+    },
+    {
+      "epoch": 0.5670103092783505,
+      "grad_norm": 0.4465166926383972,
+      "learning_rate": 9.76133632460897e-06,
+      "loss": 0.4771,
+      "step": 770
+    },
+    {
+      "epoch": 0.5677466863033873,
+      "grad_norm": 0.4014427363872528,
+      "learning_rate": 9.760026584102414e-06,
+      "loss": 0.4658,
+      "step": 771
+    },
+    {
+      "epoch": 0.5684830633284241,
+      "grad_norm": 0.42262935638427734,
+      "learning_rate": 9.758713347976179e-06,
+      "loss": 0.4551,
+      "step": 772
+    },
+    {
+      "epoch": 0.569219440353461,
+      "grad_norm": 0.43725284934043884,
+      "learning_rate": 9.757396617194663e-06,
+      "loss": 0.4512,
+      "step": 773
+    },
+    {
+      "epoch": 0.5699558173784978,
+      "grad_norm": 0.4782891273498535,
+      "learning_rate": 9.756076392724836e-06,
+      "loss": 0.4364,
+      "step": 774
+    },
+    {
+      "epoch": 0.5706921944035346,
+      "grad_norm": 0.4645746350288391,
+      "learning_rate": 9.75475267553623e-06,
+      "loss": 0.4487,
+      "step": 775
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.4626657962799072,
+      "learning_rate": 9.75342546660094e-06,
+      "loss": 0.4305,
+      "step": 776
+    },
+    {
+      "epoch": 0.5721649484536082,
+      "grad_norm": 0.519809365272522,
+      "learning_rate": 9.752094766893635e-06,
+      "loss": 0.4818,
+      "step": 777
+    },
+    {
+      "epoch": 0.572901325478645,
+      "grad_norm": 0.48324060440063477,
+      "learning_rate": 9.750760577391535e-06,
+      "loss": 0.4035,
+      "step": 778
+    },
+    {
+      "epoch": 0.5736377025036818,
+      "grad_norm": 0.5473081469535828,
+      "learning_rate": 9.74942289907443e-06,
+      "loss": 0.459,
+      "step": 779
+    },
+    {
+      "epoch": 0.5743740795287187,
+      "grad_norm": 0.4750016927719116,
+      "learning_rate": 9.74808173292467e-06,
+      "loss": 0.4627,
+      "step": 780
+    },
+    {
+      "epoch": 0.5751104565537555,
+      "grad_norm": 0.5140059590339661,
+      "learning_rate": 9.746737079927166e-06,
+      "loss": 0.475,
+      "step": 781
+    },
+    {
+      "epoch": 0.5758468335787923,
+      "grad_norm": 0.6078734397888184,
+      "learning_rate": 9.745388941069395e-06,
+      "loss": 0.4972,
+      "step": 782
+    },
+    {
+      "epoch": 0.5765832106038291,
+      "grad_norm": 0.5263283848762512,
+      "learning_rate": 9.744037317341383e-06,
+      "loss": 0.4598,
+      "step": 783
+    },
+    {
+      "epoch": 0.5773195876288659,
+      "grad_norm": 0.47057968378067017,
+      "learning_rate": 9.742682209735727e-06,
+      "loss": 0.4601,
+      "step": 784
+    },
+    {
+      "epoch": 0.5780559646539027,
+      "grad_norm": 0.5394490361213684,
+      "learning_rate": 9.741323619247575e-06,
+      "loss": 0.4812,
+      "step": 785
+    },
+    {
+      "epoch": 0.5787923416789397,
+      "grad_norm": 0.43855324387550354,
+      "learning_rate": 9.739961546874637e-06,
+      "loss": 0.4499,
+      "step": 786
+    },
+    {
+      "epoch": 0.5795287187039765,
+      "grad_norm": 0.5713585019111633,
+      "learning_rate": 9.738595993617172e-06,
+      "loss": 0.471,
+      "step": 787
+    },
+    {
+      "epoch": 0.5802650957290133,
+      "grad_norm": 0.48629793524742126,
+      "learning_rate": 9.737226960478006e-06,
+      "loss": 0.456,
+      "step": 788
+    },
+    {
+      "epoch": 0.5810014727540501,
+      "grad_norm": 0.48818135261535645,
+      "learning_rate": 9.735854448462516e-06,
+      "loss": 0.4399,
+      "step": 789
+    },
+    {
+      "epoch": 0.5817378497790869,
+      "grad_norm": 0.5030027031898499,
+      "learning_rate": 9.73447845857863e-06,
+      "loss": 0.4885,
+      "step": 790
+    },
+    {
+      "epoch": 0.5824742268041238,
+      "grad_norm": 0.5217078328132629,
+      "learning_rate": 9.733098991836834e-06,
+      "loss": 0.4426,
+      "step": 791
+    },
+    {
+      "epoch": 0.5832106038291606,
+      "grad_norm": 0.5095168948173523,
+      "learning_rate": 9.731716049250169e-06,
+      "loss": 0.4469,
+      "step": 792
+    },
+    {
+      "epoch": 0.5839469808541974,
+      "grad_norm": 0.4390084743499756,
+      "learning_rate": 9.730329631834225e-06,
+      "loss": 0.4538,
+      "step": 793
+    },
+    {
+      "epoch": 0.5846833578792342,
+      "grad_norm": 0.5591133236885071,
+      "learning_rate": 9.728939740607145e-06,
+      "loss": 0.4437,
+      "step": 794
+    },
+    {
+      "epoch": 0.585419734904271,
+      "grad_norm": 0.48923259973526,
+      "learning_rate": 9.727546376589622e-06,
+      "loss": 0.4809,
+      "step": 795
+    },
+    {
+      "epoch": 0.5861561119293078,
+      "grad_norm": 0.5047422647476196,
+      "learning_rate": 9.726149540804901e-06,
+      "loss": 0.4616,
+      "step": 796
+    },
+    {
+      "epoch": 0.5868924889543446,
+      "grad_norm": 0.47945308685302734,
+      "learning_rate": 9.724749234278779e-06,
+      "loss": 0.4662,
+      "step": 797
+    },
+    {
+      "epoch": 0.5876288659793815,
+      "grad_norm": 0.45942962169647217,
+      "learning_rate": 9.723345458039595e-06,
+      "loss": 0.4572,
+      "step": 798
+    },
+    {
+      "epoch": 0.5883652430044183,
+      "grad_norm": 0.4904220998287201,
+      "learning_rate": 9.721938213118241e-06,
+      "loss": 0.4517,
+      "step": 799
+    },
+    {
+      "epoch": 0.5891016200294551,
+      "grad_norm": 0.4758506715297699,
+      "learning_rate": 9.720527500548155e-06,
+      "loss": 0.4392,
+      "step": 800
+    },
+    {
+      "epoch": 0.5898379970544919,
+      "grad_norm": 0.5180002450942993,
+      "learning_rate": 9.719113321365324e-06,
+      "loss": 0.4538,
+      "step": 801
+    },
+    {
+      "epoch": 0.5905743740795287,
+      "grad_norm": 0.49426352977752686,
+      "learning_rate": 9.717695676608275e-06,
+      "loss": 0.4865,
+      "step": 802
+    },
+    {
+      "epoch": 0.5913107511045655,
+      "grad_norm": 0.5237860083580017,
+      "learning_rate": 9.716274567318085e-06,
+      "loss": 0.4542,
+      "step": 803
+    },
+    {
+      "epoch": 0.5920471281296024,
+      "grad_norm": 0.48790404200553894,
+      "learning_rate": 9.714849994538373e-06,
+      "loss": 0.4931,
+      "step": 804
+    },
+    {
+      "epoch": 0.5927835051546392,
+      "grad_norm": 0.49453994631767273,
+      "learning_rate": 9.713421959315303e-06,
+      "loss": 0.4633,
+      "step": 805
+    },
+    {
+      "epoch": 0.593519882179676,
+      "grad_norm": 0.5625678896903992,
+      "learning_rate": 9.71199046269758e-06,
+      "loss": 0.4769,
+      "step": 806
+    },
+    {
+      "epoch": 0.5942562592047128,
+      "grad_norm": 0.5181511044502258,
+      "learning_rate": 9.710555505736456e-06,
+      "loss": 0.4895,
+      "step": 807
+    },
+    {
+      "epoch": 0.5949926362297496,
+      "grad_norm": 0.4666210114955902,
+      "learning_rate": 9.709117089485714e-06,
+      "loss": 0.4602,
+      "step": 808
+    },
+    {
+      "epoch": 0.5957290132547864,
+      "grad_norm": 0.42587047815322876,
+      "learning_rate": 9.707675215001685e-06,
+      "loss": 0.4632,
+      "step": 809
+    },
+    {
+      "epoch": 0.5964653902798233,
+      "grad_norm": 0.5286933183670044,
+      "learning_rate": 9.706229883343242e-06,
+      "loss": 0.4657,
+      "step": 810
+    },
+    {
+      "epoch": 0.5972017673048601,
+      "grad_norm": 0.4617254436016083,
+      "learning_rate": 9.704781095571788e-06,
+      "loss": 0.4559,
+      "step": 811
+    },
+    {
+      "epoch": 0.5979381443298969,
+      "grad_norm": 0.5746539235115051,
+      "learning_rate": 9.70332885275127e-06,
+      "loss": 0.4548,
+      "step": 812
+    },
+    {
+      "epoch": 0.5986745213549337,
+      "grad_norm": 0.4450950026512146,
+      "learning_rate": 9.701873155948177e-06,
+      "loss": 0.4597,
+      "step": 813
+    },
+    {
+      "epoch": 0.5994108983799705,
+      "grad_norm": 0.48545292019844055,
+      "learning_rate": 9.70041400623152e-06,
+      "loss": 0.4684,
+      "step": 814
+    },
+    {
+      "epoch": 0.6001472754050073,
+      "grad_norm": 0.45839008688926697,
+      "learning_rate": 9.698951404672858e-06,
+      "loss": 0.436,
+      "step": 815
+    },
+    {
+      "epoch": 0.6008836524300442,
+      "grad_norm": 0.5452147722244263,
+      "learning_rate": 9.697485352346282e-06,
+      "loss": 0.43,
+      "step": 816
+    },
+    {
+      "epoch": 0.601620029455081,
+      "grad_norm": 0.4968319237232208,
+      "learning_rate": 9.696015850328418e-06,
+      "loss": 0.4438,
+      "step": 817
+    },
+    {
+      "epoch": 0.6023564064801178,
+      "grad_norm": 0.5042880773544312,
+      "learning_rate": 9.694542899698422e-06,
+      "loss": 0.4463,
+      "step": 818
+    },
+    {
+      "epoch": 0.6030927835051546,
+      "grad_norm": 0.5175356864929199,
+      "learning_rate": 9.693066501537984e-06,
+      "loss": 0.4614,
+      "step": 819
+    },
+    {
+      "epoch": 0.6038291605301914,
+      "grad_norm": 0.5082677006721497,
+      "learning_rate": 9.691586656931326e-06,
+      "loss": 0.4635,
+      "step": 820
+    },
+    {
+      "epoch": 0.6045655375552282,
+      "grad_norm": 0.5881240367889404,
+      "learning_rate": 9.690103366965204e-06,
+      "loss": 0.4691,
+      "step": 821
+    },
+    {
+      "epoch": 0.605301914580265,
+      "grad_norm": 0.4930703341960907,
+      "learning_rate": 9.688616632728898e-06,
+      "loss": 0.439,
+      "step": 822
+    },
+    {
+      "epoch": 0.6060382916053019,
+      "grad_norm": 0.4951449930667877,
+      "learning_rate": 9.687126455314221e-06,
+      "loss": 0.4493,
+      "step": 823
+    },
+    {
+      "epoch": 0.6067746686303387,
+      "grad_norm": 0.4590352475643158,
+      "learning_rate": 9.685632835815519e-06,
+      "loss": 0.4778,
+      "step": 824
+    },
+    {
+      "epoch": 0.6075110456553755,
+      "grad_norm": 0.547599196434021,
+      "learning_rate": 9.684135775329653e-06,
+      "loss": 0.4312,
+      "step": 825
+    },
+    {
+      "epoch": 0.6082474226804123,
+      "grad_norm": 0.4426873028278351,
+      "learning_rate": 9.682635274956026e-06,
+      "loss": 0.4735,
+      "step": 826
+    },
+    {
+      "epoch": 0.6089837997054491,
+      "grad_norm": 0.41880539059638977,
+      "learning_rate": 9.681131335796557e-06,
+      "loss": 0.4593,
+      "step": 827
+    },
+    {
+      "epoch": 0.6097201767304861,
+      "grad_norm": 0.5596433877944946,
+      "learning_rate": 9.679623958955692e-06,
+      "loss": 0.4922,
+      "step": 828
+    },
+    {
+      "epoch": 0.6104565537555229,
+      "grad_norm": 0.45280906558036804,
+      "learning_rate": 9.678113145540406e-06,
+      "loss": 0.4737,
+      "step": 829
+    },
+    {
+      "epoch": 0.6111929307805597,
+      "grad_norm": 0.6396340727806091,
+      "learning_rate": 9.676598896660194e-06,
+      "loss": 0.4929,
+      "step": 830
+    },
+    {
+      "epoch": 0.6119293078055965,
+      "grad_norm": 0.4250088036060333,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 0.421,
+      "step": 831
+    },
+    {
+      "epoch": 0.6126656848306333,
+      "grad_norm": 0.5473159551620483,
+      "learning_rate": 9.673560096955588e-06,
+      "loss": 0.4563,
+      "step": 832
+    },
+    {
+      "epoch": 0.6134020618556701,
+      "grad_norm": 0.46349871158599854,
+      "learning_rate": 9.672035548362797e-06,
+      "loss": 0.4308,
+      "step": 833
+    },
+    {
+      "epoch": 0.614138438880707,
+      "grad_norm": 0.48488184809684753,
+      "learning_rate": 9.670507568768281e-06,
+      "loss": 0.4767,
+      "step": 834
+    },
+    {
+      "epoch": 0.6148748159057438,
+      "grad_norm": 0.5228757262229919,
+      "learning_rate": 9.668976159294145e-06,
+      "loss": 0.48,
+      "step": 835
+    },
+    {
+      "epoch": 0.6156111929307806,
+      "grad_norm": 0.5242437720298767,
+      "learning_rate": 9.66744132106501e-06,
+      "loss": 0.483,
+      "step": 836
+    },
+    {
+      "epoch": 0.6163475699558174,
+      "grad_norm": 0.47697046399116516,
+      "learning_rate": 9.665903055208013e-06,
+      "loss": 0.4704,
+      "step": 837
+    },
+    {
+      "epoch": 0.6170839469808542,
+      "grad_norm": 0.5204547047615051,
+      "learning_rate": 9.664361362852813e-06,
+      "loss": 0.4873,
+      "step": 838
+    },
+    {
+      "epoch": 0.617820324005891,
+      "grad_norm": 0.4981822967529297,
+      "learning_rate": 9.66281624513158e-06,
+      "loss": 0.4354,
+      "step": 839
+    },
+    {
+      "epoch": 0.6185567010309279,
+      "grad_norm": 0.4810592830181122,
+      "learning_rate": 9.661267703178999e-06,
+      "loss": 0.4763,
+      "step": 840
+    },
+    {
+      "epoch": 0.6192930780559647,
+      "grad_norm": 0.4789879024028778,
+      "learning_rate": 9.659715738132279e-06,
+      "loss": 0.4358,
+      "step": 841
+    },
+    {
+      "epoch": 0.6200294550810015,
+      "grad_norm": 0.48347872495651245,
+      "learning_rate": 9.658160351131129e-06,
+      "loss": 0.5049,
+      "step": 842
+    },
+    {
+      "epoch": 0.6207658321060383,
+      "grad_norm": 0.48118627071380615,
+      "learning_rate": 9.656601543317784e-06,
+      "loss": 0.49,
+      "step": 843
+    },
+    {
+      "epoch": 0.6215022091310751,
+      "grad_norm": 0.4531796872615814,
+      "learning_rate": 9.655039315836983e-06,
+      "loss": 0.4872,
+      "step": 844
+    },
+    {
+      "epoch": 0.6222385861561119,
+      "grad_norm": 0.4885636270046234,
+      "learning_rate": 9.653473669835978e-06,
+      "loss": 0.4479,
+      "step": 845
+    },
+    {
+      "epoch": 0.6229749631811488,
+      "grad_norm": 0.49448058009147644,
+      "learning_rate": 9.651904606464536e-06,
+      "loss": 0.4701,
+      "step": 846
+    },
+    {
+      "epoch": 0.6237113402061856,
+      "grad_norm": 0.47281432151794434,
+      "learning_rate": 9.650332126874924e-06,
+      "loss": 0.4437,
+      "step": 847
+    },
+    {
+      "epoch": 0.6244477172312224,
+      "grad_norm": 0.5343514084815979,
+      "learning_rate": 9.648756232221925e-06,
+      "loss": 0.4899,
+      "step": 848
+    },
+    {
+      "epoch": 0.6251840942562592,
+      "grad_norm": 0.42154327034950256,
+      "learning_rate": 9.647176923662833e-06,
+      "loss": 0.4486,
+      "step": 849
+    },
+    {
+      "epoch": 0.625920471281296,
+      "grad_norm": 0.4753609597682953,
+      "learning_rate": 9.645594202357438e-06,
+      "loss": 0.4371,
+      "step": 850
+    },
+    {
+      "epoch": 0.6266568483063328,
+      "grad_norm": 0.6016598343849182,
+      "learning_rate": 9.644008069468047e-06,
+      "loss": 0.4638,
+      "step": 851
+    },
+    {
+      "epoch": 0.6273932253313697,
+      "grad_norm": 0.46637803316116333,
+      "learning_rate": 9.642418526159467e-06,
+      "loss": 0.4541,
+      "step": 852
+    },
+    {
+      "epoch": 0.6281296023564065,
+      "grad_norm": 0.5309515595436096,
+      "learning_rate": 9.64082557359901e-06,
+      "loss": 0.4409,
+      "step": 853
+    },
+    {
+      "epoch": 0.6288659793814433,
+      "grad_norm": 0.6249402761459351,
+      "learning_rate": 9.639229212956494e-06,
+      "loss": 0.5012,
+      "step": 854
+    },
+    {
+      "epoch": 0.6296023564064801,
+      "grad_norm": 0.43967223167419434,
+      "learning_rate": 9.637629445404237e-06,
+      "loss": 0.4438,
+      "step": 855
+    },
+    {
+      "epoch": 0.6303387334315169,
+      "grad_norm": 0.45795974135398865,
+      "learning_rate": 9.636026272117058e-06,
+      "loss": 0.445,
+      "step": 856
+    },
+    {
+      "epoch": 0.6310751104565537,
+      "grad_norm": 0.5375895500183105,
+      "learning_rate": 9.63441969427228e-06,
+      "loss": 0.4441,
+      "step": 857
+    },
+    {
+      "epoch": 0.6318114874815906,
+      "grad_norm": 0.4524497985839844,
+      "learning_rate": 9.632809713049726e-06,
+      "loss": 0.4581,
+      "step": 858
+    },
+    {
+      "epoch": 0.6325478645066274,
+      "grad_norm": 0.4217367172241211,
+      "learning_rate": 9.631196329631719e-06,
+      "loss": 0.4681,
+      "step": 859
+    },
+    {
+      "epoch": 0.6332842415316642,
+      "grad_norm": 0.5063615441322327,
+      "learning_rate": 9.629579545203076e-06,
+      "loss": 0.4632,
+      "step": 860
+    },
+    {
+      "epoch": 0.634020618556701,
+      "grad_norm": 0.46327000856399536,
+      "learning_rate": 9.627959360951118e-06,
+      "loss": 0.437,
+      "step": 861
+    },
+    {
+      "epoch": 0.6347569955817378,
+      "grad_norm": 0.4578990042209625,
+      "learning_rate": 9.626335778065655e-06,
+      "loss": 0.4618,
+      "step": 862
+    },
+    {
+      "epoch": 0.6354933726067746,
+      "grad_norm": 0.47172173857688904,
+      "learning_rate": 9.624708797739002e-06,
+      "loss": 0.4533,
+      "step": 863
+    },
+    {
+      "epoch": 0.6362297496318114,
+      "grad_norm": 0.5421725511550903,
+      "learning_rate": 9.623078421165958e-06,
+      "loss": 0.4663,
+      "step": 864
+    },
+    {
+      "epoch": 0.6369661266568483,
+      "grad_norm": 0.4460313320159912,
+      "learning_rate": 9.62144464954383e-06,
+      "loss": 0.4486,
+      "step": 865
+    },
+    {
+      "epoch": 0.6377025036818851,
+      "grad_norm": 0.5286182165145874,
+      "learning_rate": 9.619807484072405e-06,
+      "loss": 0.4611,
+      "step": 866
+    },
+    {
+      "epoch": 0.6384388807069219,
+      "grad_norm": 0.5084875226020813,
+      "learning_rate": 9.618166925953969e-06,
+      "loss": 0.478,
+      "step": 867
+    },
+    {
+      "epoch": 0.6391752577319587,
+      "grad_norm": 0.5308467745780945,
+      "learning_rate": 9.6165229763933e-06,
+      "loss": 0.4512,
+      "step": 868
+    },
+    {
+      "epoch": 0.6399116347569955,
+      "grad_norm": 0.5583586096763611,
+      "learning_rate": 9.614875636597662e-06,
+      "loss": 0.4665,
+      "step": 869
+    },
+    {
+      "epoch": 0.6406480117820325,
+      "grad_norm": 0.4598456919193268,
+      "learning_rate": 9.613224907776814e-06,
+      "loss": 0.4138,
+      "step": 870
+    },
+    {
+      "epoch": 0.6413843888070693,
+      "grad_norm": 0.43438005447387695,
+      "learning_rate": 9.611570791143e-06,
+      "loss": 0.4584,
+      "step": 871
+    },
+    {
+      "epoch": 0.6421207658321061,
+      "grad_norm": 0.5358659029006958,
+      "learning_rate": 9.609913287910957e-06,
+      "loss": 0.4371,
+      "step": 872
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 0.4852744936943054,
+      "learning_rate": 9.608252399297899e-06,
+      "loss": 0.4621,
+      "step": 873
+    },
+    {
+      "epoch": 0.6435935198821797,
+      "grad_norm": 0.4329935610294342,
+      "learning_rate": 9.606588126523537e-06,
+      "loss": 0.4368,
+      "step": 874
+    },
+    {
+      "epoch": 0.6443298969072165,
+      "grad_norm": 0.5553391575813293,
+      "learning_rate": 9.60492047081006e-06,
+      "loss": 0.4636,
+      "step": 875
+    },
+    {
+      "epoch": 0.6450662739322534,
+      "grad_norm": 0.42962443828582764,
+      "learning_rate": 9.603249433382145e-06,
+      "loss": 0.4829,
+      "step": 876
+    },
+    {
+      "epoch": 0.6458026509572902,
+      "grad_norm": 0.47330227494239807,
+      "learning_rate": 9.60157501546695e-06,
+      "loss": 0.4412,
+      "step": 877
+    },
+    {
+      "epoch": 0.646539027982327,
+      "grad_norm": 0.444964200258255,
+      "learning_rate": 9.599897218294122e-06,
+      "loss": 0.4608,
+      "step": 878
+    },
+    {
+      "epoch": 0.6472754050073638,
+      "grad_norm": 0.5015583038330078,
+      "learning_rate": 9.598216043095779e-06,
+      "loss": 0.4652,
+      "step": 879
+    },
+    {
+      "epoch": 0.6480117820324006,
+      "grad_norm": 0.5097722411155701,
+      "learning_rate": 9.596531491106528e-06,
+      "loss": 0.4401,
+      "step": 880
+    },
+    {
+      "epoch": 0.6487481590574374,
+      "grad_norm": 0.5142177939414978,
+      "learning_rate": 9.594843563563452e-06,
+      "loss": 0.4416,
+      "step": 881
+    },
+    {
+      "epoch": 0.6494845360824743,
+      "grad_norm": 0.45564496517181396,
+      "learning_rate": 9.593152261706113e-06,
+      "loss": 0.4457,
+      "step": 882
+    },
+    {
+      "epoch": 0.6502209131075111,
+      "grad_norm": 0.4867049753665924,
+      "learning_rate": 9.591457586776555e-06,
+      "loss": 0.4653,
+      "step": 883
+    },
+    {
+      "epoch": 0.6509572901325479,
+      "grad_norm": 0.4787944257259369,
+      "learning_rate": 9.589759540019293e-06,
+      "loss": 0.4849,
+      "step": 884
+    },
+    {
+      "epoch": 0.6516936671575847,
+      "grad_norm": 0.4713253080844879,
+      "learning_rate": 9.588058122681324e-06,
+      "loss": 0.432,
+      "step": 885
+    },
+    {
+      "epoch": 0.6524300441826215,
+      "grad_norm": 0.48565471172332764,
+      "learning_rate": 9.586353336012115e-06,
+      "loss": 0.4606,
+      "step": 886
+    },
+    {
+      "epoch": 0.6531664212076583,
+      "grad_norm": 0.48033156991004944,
+      "learning_rate": 9.58464518126361e-06,
+      "loss": 0.4695,
+      "step": 887
+    },
+    {
+      "epoch": 0.6539027982326951,
+      "grad_norm": 0.5356956720352173,
+      "learning_rate": 9.582933659690228e-06,
+      "loss": 0.4774,
+      "step": 888
+    },
+    {
+      "epoch": 0.654639175257732,
+      "grad_norm": 0.5078469514846802,
+      "learning_rate": 9.58121877254886e-06,
+      "loss": 0.4458,
+      "step": 889
+    },
+    {
+      "epoch": 0.6553755522827688,
+      "grad_norm": 0.6260564923286438,
+      "learning_rate": 9.57950052109886e-06,
+      "loss": 0.446,
+      "step": 890
+    },
+    {
+      "epoch": 0.6561119293078056,
+      "grad_norm": 0.4698830246925354,
+      "learning_rate": 9.577778906602069e-06,
+      "loss": 0.4637,
+      "step": 891
+    },
+    {
+      "epoch": 0.6568483063328424,
+      "grad_norm": 0.5720396041870117,
+      "learning_rate": 9.576053930322784e-06,
+      "loss": 0.4609,
+      "step": 892
+    },
+    {
+      "epoch": 0.6575846833578792,
+      "grad_norm": 0.5099953413009644,
+      "learning_rate": 9.574325593527776e-06,
+      "loss": 0.4763,
+      "step": 893
+    },
+    {
+      "epoch": 0.658321060382916,
+      "grad_norm": 0.582254946231842,
+      "learning_rate": 9.572593897486283e-06,
+      "loss": 0.4653,
+      "step": 894
+    },
+    {
+      "epoch": 0.6590574374079529,
+      "grad_norm": 0.5127898454666138,
+      "learning_rate": 9.57085884347001e-06,
+      "loss": 0.47,
+      "step": 895
+    },
+    {
+      "epoch": 0.6597938144329897,
+      "grad_norm": 0.5360256433486938,
+      "learning_rate": 9.56912043275313e-06,
+      "loss": 0.4483,
+      "step": 896
+    },
+    {
+      "epoch": 0.6605301914580265,
+      "grad_norm": 0.5025308728218079,
+      "learning_rate": 9.567378666612279e-06,
+      "loss": 0.4378,
+      "step": 897
+    },
+    {
+      "epoch": 0.6612665684830633,
+      "grad_norm": 0.5077435374259949,
+      "learning_rate": 9.565633546326555e-06,
+      "loss": 0.4492,
+      "step": 898
+    },
+    {
+      "epoch": 0.6620029455081001,
+      "grad_norm": 0.580901026725769,
+      "learning_rate": 9.563885073177523e-06,
+      "loss": 0.4499,
+      "step": 899
+    },
+    {
+      "epoch": 0.6627393225331369,
+      "grad_norm": 0.4233223795890808,
+      "learning_rate": 9.56213324844921e-06,
+      "loss": 0.4618,
+      "step": 900
+    },
+    {
+      "epoch": 0.6634756995581738,
+      "grad_norm": 0.5685048699378967,
+      "learning_rate": 9.560378073428103e-06,
+      "loss": 0.4363,
+      "step": 901
+    },
+    {
+      "epoch": 0.6642120765832106,
+      "grad_norm": 0.5028063654899597,
+      "learning_rate": 9.558619549403148e-06,
+      "loss": 0.4684,
+      "step": 902
+    },
+    {
+      "epoch": 0.6649484536082474,
+      "grad_norm": 0.5596533417701721,
+      "learning_rate": 9.556857677665752e-06,
+      "loss": 0.4379,
+      "step": 903
+    },
+    {
+      "epoch": 0.6656848306332842,
+      "grad_norm": 0.391494482755661,
+      "learning_rate": 9.555092459509783e-06,
+      "loss": 0.4425,
+      "step": 904
+    },
+    {
+      "epoch": 0.666421207658321,
+      "grad_norm": 0.5344425439834595,
+      "learning_rate": 9.553323896231558e-06,
+      "loss": 0.4473,
+      "step": 905
+    },
+    {
+      "epoch": 0.6671575846833578,
+      "grad_norm": 0.4520135819911957,
+      "learning_rate": 9.551551989129864e-06,
+      "loss": 0.4698,
+      "step": 906
+    },
+    {
+      "epoch": 0.6678939617083947,
+      "grad_norm": 0.5516674518585205,
+      "learning_rate": 9.549776739505932e-06,
+      "loss": 0.4926,
+      "step": 907
+    },
+    {
+      "epoch": 0.6686303387334315,
+      "grad_norm": 0.4469415545463562,
+      "learning_rate": 9.547998148663449e-06,
+      "loss": 0.4509,
+      "step": 908
+    },
+    {
+      "epoch": 0.6693667157584683,
+      "grad_norm": 0.4923398494720459,
+      "learning_rate": 9.546216217908564e-06,
+      "loss": 0.4591,
+      "step": 909
+    },
+    {
+      "epoch": 0.6701030927835051,
+      "grad_norm": 0.5482118129730225,
+      "learning_rate": 9.54443094854987e-06,
+      "loss": 0.4566,
+      "step": 910
+    },
+    {
+      "epoch": 0.6708394698085419,
+      "grad_norm": 0.5054488778114319,
+      "learning_rate": 9.542642341898416e-06,
+      "loss": 0.4513,
+      "step": 911
+    },
+    {
+      "epoch": 0.6715758468335787,
+      "grad_norm": 0.6202924847602844,
+      "learning_rate": 9.540850399267698e-06,
+      "loss": 0.4827,
+      "step": 912
+    },
+    {
+      "epoch": 0.6723122238586157,
+      "grad_norm": 0.45753154158592224,
+      "learning_rate": 9.539055121973668e-06,
+      "loss": 0.4849,
+      "step": 913
+    },
+    {
+      "epoch": 0.6730486008836525,
+      "grad_norm": 0.7275657653808594,
+      "learning_rate": 9.537256511334722e-06,
+      "loss": 0.4659,
+      "step": 914
+    },
+    {
+      "epoch": 0.6737849779086893,
+      "grad_norm": 0.43268248438835144,
+      "learning_rate": 9.535454568671705e-06,
+      "loss": 0.4408,
+      "step": 915
+    },
+    {
+      "epoch": 0.6745213549337261,
+      "grad_norm": 0.6889506578445435,
+      "learning_rate": 9.53364929530791e-06,
+      "loss": 0.4649,
+      "step": 916
+    },
+    {
+      "epoch": 0.6752577319587629,
+      "grad_norm": 0.4373452365398407,
+      "learning_rate": 9.531840692569073e-06,
+      "loss": 0.4474,
+      "step": 917
+    },
+    {
+      "epoch": 0.6759941089837997,
+      "grad_norm": 0.6262958645820618,
+      "learning_rate": 9.530028761783379e-06,
+      "loss": 0.4589,
+      "step": 918
+    },
+    {
+      "epoch": 0.6767304860088366,
+      "grad_norm": 0.4766991436481476,
+      "learning_rate": 9.528213504281457e-06,
+      "loss": 0.4724,
+      "step": 919
+    },
+    {
+      "epoch": 0.6774668630338734,
+      "grad_norm": 0.43806901574134827,
+      "learning_rate": 9.526394921396373e-06,
+      "loss": 0.4478,
+      "step": 920
+    },
+    {
+      "epoch": 0.6782032400589102,
+      "grad_norm": 0.509490966796875,
+      "learning_rate": 9.524573014463643e-06,
+      "loss": 0.4739,
+      "step": 921
+    },
+    {
+      "epoch": 0.678939617083947,
+      "grad_norm": 0.4827909469604492,
+      "learning_rate": 9.52274778482122e-06,
+      "loss": 0.472,
+      "step": 922
+    },
+    {
+      "epoch": 0.6796759941089838,
+      "grad_norm": 0.46269240975379944,
+      "learning_rate": 9.520919233809494e-06,
+      "loss": 0.4498,
+      "step": 923
+    },
+    {
+      "epoch": 0.6804123711340206,
+      "grad_norm": 0.5521724224090576,
+      "learning_rate": 9.519087362771302e-06,
+      "loss": 0.4634,
+      "step": 924
+    },
+    {
+      "epoch": 0.6811487481590575,
+      "grad_norm": 0.5625545978546143,
+      "learning_rate": 9.517252173051912e-06,
+      "loss": 0.4552,
+      "step": 925
+    },
+    {
+      "epoch": 0.6818851251840943,
+      "grad_norm": 0.4687793552875519,
+      "learning_rate": 9.515413665999034e-06,
+      "loss": 0.4439,
+      "step": 926
+    },
+    {
+      "epoch": 0.6826215022091311,
+      "grad_norm": 0.6292194724082947,
+      "learning_rate": 9.51357184296281e-06,
+      "loss": 0.47,
+      "step": 927
+    },
+    {
+      "epoch": 0.6833578792341679,
+      "grad_norm": 0.5816941857337952,
+      "learning_rate": 9.51172670529582e-06,
+      "loss": 0.4765,
+      "step": 928
+    },
+    {
+      "epoch": 0.6840942562592047,
+      "grad_norm": 0.5308511853218079,
+      "learning_rate": 9.509878254353076e-06,
+      "loss": 0.4756,
+      "step": 929
+    },
+    {
+      "epoch": 0.6848306332842415,
+      "grad_norm": 0.6591737866401672,
+      "learning_rate": 9.508026491492027e-06,
+      "loss": 0.5005,
+      "step": 930
+    },
+    {
+      "epoch": 0.6855670103092784,
+      "grad_norm": 0.477727472782135,
+      "learning_rate": 9.50617141807255e-06,
+      "loss": 0.4663,
+      "step": 931
+    },
+    {
+      "epoch": 0.6863033873343152,
+      "grad_norm": 0.6698769927024841,
+      "learning_rate": 9.504313035456955e-06,
+      "loss": 0.4723,
+      "step": 932
+    },
+    {
+      "epoch": 0.687039764359352,
+      "grad_norm": 0.5323542952537537,
+      "learning_rate": 9.502451345009984e-06,
+      "loss": 0.4566,
+      "step": 933
+    },
+    {
+      "epoch": 0.6877761413843888,
+      "grad_norm": 0.5885183215141296,
+      "learning_rate": 9.500586348098803e-06,
+      "loss": 0.457,
+      "step": 934
+    },
+    {
+      "epoch": 0.6885125184094256,
+      "grad_norm": 0.6716952919960022,
+      "learning_rate": 9.498718046093013e-06,
+      "loss": 0.4514,
+      "step": 935
+    },
+    {
+      "epoch": 0.6892488954344624,
+      "grad_norm": 0.5240690112113953,
+      "learning_rate": 9.496846440364634e-06,
+      "loss": 0.4739,
+      "step": 936
+    },
+    {
+      "epoch": 0.6899852724594993,
+      "grad_norm": 0.7379181981086731,
+      "learning_rate": 9.49497153228812e-06,
+      "loss": 0.4672,
+      "step": 937
+    },
+    {
+      "epoch": 0.6907216494845361,
+      "grad_norm": 0.4711780250072479,
+      "learning_rate": 9.493093323240348e-06,
+      "loss": 0.4378,
+      "step": 938
+    },
+    {
+      "epoch": 0.6914580265095729,
+      "grad_norm": 0.661437451839447,
+      "learning_rate": 9.491211814600613e-06,
+      "loss": 0.4741,
+      "step": 939
+    },
+    {
+      "epoch": 0.6921944035346097,
+      "grad_norm": 0.5604520440101624,
+      "learning_rate": 9.489327007750644e-06,
+      "loss": 0.4706,
+      "step": 940
+    },
+    {
+      "epoch": 0.6929307805596465,
+      "grad_norm": 0.5789750814437866,
+      "learning_rate": 9.487438904074581e-06,
+      "loss": 0.4202,
+      "step": 941
+    },
+    {
+      "epoch": 0.6936671575846833,
+      "grad_norm": 0.6015188694000244,
+      "learning_rate": 9.485547504958993e-06,
+      "loss": 0.4321,
+      "step": 942
+    },
+    {
+      "epoch": 0.6944035346097202,
+      "grad_norm": 0.5518826246261597,
+      "learning_rate": 9.483652811792866e-06,
+      "loss": 0.4777,
+      "step": 943
+    },
+    {
+      "epoch": 0.695139911634757,
+      "grad_norm": 0.5786319375038147,
+      "learning_rate": 9.481754825967606e-06,
+      "loss": 0.4405,
+      "step": 944
+    },
+    {
+      "epoch": 0.6958762886597938,
+      "grad_norm": 0.5292356610298157,
+      "learning_rate": 9.479853548877033e-06,
+      "loss": 0.4367,
+      "step": 945
+    },
+    {
+      "epoch": 0.6966126656848306,
+      "grad_norm": 0.6690455675125122,
+      "learning_rate": 9.477948981917393e-06,
+      "loss": 0.4381,
+      "step": 946
+    },
+    {
+      "epoch": 0.6973490427098674,
+      "grad_norm": 0.4962507486343384,
+      "learning_rate": 9.476041126487341e-06,
+      "loss": 0.4595,
+      "step": 947
+    },
+    {
+      "epoch": 0.6980854197349042,
+      "grad_norm": 0.5820249915122986,
+      "learning_rate": 9.474129983987943e-06,
+      "loss": 0.4904,
+      "step": 948
+    },
+    {
+      "epoch": 0.698821796759941,
+      "grad_norm": 0.5604087710380554,
+      "learning_rate": 9.472215555822691e-06,
+      "loss": 0.4706,
+      "step": 949
+    },
+    {
+      "epoch": 0.6995581737849779,
+      "grad_norm": 0.4952147603034973,
+      "learning_rate": 9.47029784339748e-06,
+      "loss": 0.4787,
+      "step": 950
+    },
+    {
+      "epoch": 0.7002945508100147,
+      "grad_norm": 0.5814775228500366,
+      "learning_rate": 9.468376848120619e-06,
+      "loss": 0.4813,
+      "step": 951
+    },
+    {
+      "epoch": 0.7010309278350515,
+      "grad_norm": 0.4648406207561493,
+      "learning_rate": 9.466452571402833e-06,
+      "loss": 0.4607,
+      "step": 952
+    },
+    {
+      "epoch": 0.7017673048600883,
+      "grad_norm": 0.5979225039482117,
+      "learning_rate": 9.464525014657249e-06,
+      "loss": 0.4521,
+      "step": 953
+    },
+    {
+      "epoch": 0.7025036818851251,
+      "grad_norm": 0.4992789328098297,
+      "learning_rate": 9.462594179299408e-06,
+      "loss": 0.4686,
+      "step": 954
+    },
+    {
+      "epoch": 0.7032400589101621,
+      "grad_norm": 0.5263684391975403,
+      "learning_rate": 9.460660066747255e-06,
+      "loss": 0.4618,
+      "step": 955
+    },
+    {
+      "epoch": 0.7039764359351989,
+      "grad_norm": 0.5737230777740479,
+      "learning_rate": 9.458722678421146e-06,
+      "loss": 0.4294,
+      "step": 956
+    },
+    {
+      "epoch": 0.7047128129602357,
+      "grad_norm": 0.48960253596305847,
+      "learning_rate": 9.45678201574384e-06,
+      "loss": 0.4361,
+      "step": 957
+    },
+    {
+      "epoch": 0.7054491899852725,
+      "grad_norm": 0.6119993925094604,
+      "learning_rate": 9.454838080140501e-06,
+      "loss": 0.4593,
+      "step": 958
+    },
+    {
+      "epoch": 0.7061855670103093,
+      "grad_norm": 0.5781173706054688,
+      "learning_rate": 9.452890873038697e-06,
+      "loss": 0.4643,
+      "step": 959
+    },
+    {
+      "epoch": 0.7069219440353461,
+      "grad_norm": 0.7216101884841919,
+      "learning_rate": 9.450940395868397e-06,
+      "loss": 0.468,
+      "step": 960
+    },
+    {
+      "epoch": 0.707658321060383,
+      "grad_norm": 0.47883689403533936,
+      "learning_rate": 9.448986650061973e-06,
+      "loss": 0.4386,
+      "step": 961
+    },
+    {
+      "epoch": 0.7083946980854198,
+      "grad_norm": 0.5716425180435181,
+      "learning_rate": 9.447029637054198e-06,
+      "loss": 0.5083,
+      "step": 962
+    },
+    {
+      "epoch": 0.7091310751104566,
+      "grad_norm": 0.5576707720756531,
+      "learning_rate": 9.445069358282242e-06,
+      "loss": 0.4811,
+      "step": 963
+    },
+    {
+      "epoch": 0.7098674521354934,
+      "grad_norm": 0.4840952455997467,
+      "learning_rate": 9.443105815185674e-06,
+      "loss": 0.4485,
+      "step": 964
+    },
+    {
+      "epoch": 0.7106038291605302,
+      "grad_norm": 0.6055101156234741,
+      "learning_rate": 9.44113900920646e-06,
+      "loss": 0.4484,
+      "step": 965
+    },
+    {
+      "epoch": 0.711340206185567,
+      "grad_norm": 0.4516480267047882,
+      "learning_rate": 9.439168941788965e-06,
+      "loss": 0.4344,
+      "step": 966
+    },
+    {
+      "epoch": 0.7120765832106039,
+      "grad_norm": 0.5439156889915466,
+      "learning_rate": 9.437195614379947e-06,
+      "loss": 0.4537,
+      "step": 967
+    },
+    {
+      "epoch": 0.7128129602356407,
+      "grad_norm": 0.559701681137085,
+      "learning_rate": 9.435219028428558e-06,
+      "loss": 0.4527,
+      "step": 968
+    },
+    {
+      "epoch": 0.7135493372606775,
+      "grad_norm": 0.4636538326740265,
+      "learning_rate": 9.43323918538634e-06,
+      "loss": 0.47,
+      "step": 969
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.4914974272251129,
+      "learning_rate": 9.431256086707233e-06,
+      "loss": 0.4688,
+      "step": 970
+    },
+    {
+      "epoch": 0.7150220913107511,
+      "grad_norm": 0.5754413604736328,
+      "learning_rate": 9.429269733847563e-06,
+      "loss": 0.4754,
+      "step": 971
+    },
+    {
+      "epoch": 0.7157584683357879,
+      "grad_norm": 0.4941750168800354,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.4714,
+      "step": 972
+    },
+    {
+      "epoch": 0.7164948453608248,
+      "grad_norm": 0.5356074571609497,
+      "learning_rate": 9.425287271423797e-06,
+      "loss": 0.433,
+      "step": 973
+    },
+    {
+      "epoch": 0.7172312223858616,
+      "grad_norm": 0.5081741809844971,
+      "learning_rate": 9.4232911647843e-06,
+      "loss": 0.4652,
+      "step": 974
+    },
+    {
+      "epoch": 0.7179675994108984,
+      "grad_norm": 0.4957806169986725,
+      "learning_rate": 9.42129180981344e-06,
+      "loss": 0.4731,
+      "step": 975
+    },
+    {
+      "epoch": 0.7187039764359352,
+      "grad_norm": 0.5311683416366577,
+      "learning_rate": 9.41928920797948e-06,
+      "loss": 0.4396,
+      "step": 976
+    },
+    {
+      "epoch": 0.719440353460972,
+      "grad_norm": 0.44548124074935913,
+      "learning_rate": 9.417283360753073e-06,
+      "loss": 0.4488,
+      "step": 977
+    },
+    {
+      "epoch": 0.7201767304860088,
+      "grad_norm": 0.556957483291626,
+      "learning_rate": 9.415274269607253e-06,
+      "loss": 0.4302,
+      "step": 978
+    },
+    {
+      "epoch": 0.7209131075110456,
+      "grad_norm": 0.478667289018631,
+      "learning_rate": 9.413261936017433e-06,
+      "loss": 0.4154,
+      "step": 979
+    },
+    {
+      "epoch": 0.7216494845360825,
+      "grad_norm": 0.535645604133606,
+      "learning_rate": 9.41124636146141e-06,
+      "loss": 0.4511,
+      "step": 980
+    },
+    {
+      "epoch": 0.7223858615611193,
+      "grad_norm": 0.5650351643562317,
+      "learning_rate": 9.409227547419364e-06,
+      "loss": 0.4624,
+      "step": 981
+    },
+    {
+      "epoch": 0.7231222385861561,
+      "grad_norm": 0.4671155512332916,
+      "learning_rate": 9.407205495373849e-06,
+      "loss": 0.462,
+      "step": 982
+    },
+    {
+      "epoch": 0.7238586156111929,
+      "grad_norm": 0.4867047965526581,
+      "learning_rate": 9.405180206809799e-06,
+      "loss": 0.4013,
+      "step": 983
+    },
+    {
+      "epoch": 0.7245949926362297,
+      "grad_norm": 0.7005500197410583,
+      "learning_rate": 9.403151683214525e-06,
+      "loss": 0.4596,
+      "step": 984
+    },
+    {
+      "epoch": 0.7253313696612665,
+      "grad_norm": 0.47493258118629456,
+      "learning_rate": 9.401119926077714e-06,
+      "loss": 0.4439,
+      "step": 985
+    },
+    {
+      "epoch": 0.7260677466863034,
+      "grad_norm": 0.6544631123542786,
+      "learning_rate": 9.399084936891424e-06,
+      "loss": 0.4709,
+      "step": 986
+    },
+    {
+      "epoch": 0.7268041237113402,
+      "grad_norm": 0.4934577941894531,
+      "learning_rate": 9.397046717150095e-06,
+      "loss": 0.4537,
+      "step": 987
+    },
+    {
+      "epoch": 0.727540500736377,
+      "grad_norm": 0.5570245981216431,
+      "learning_rate": 9.39500526835053e-06,
+      "loss": 0.4462,
+      "step": 988
+    },
+    {
+      "epoch": 0.7282768777614138,
+      "grad_norm": 0.6459545493125916,
+      "learning_rate": 9.392960591991908e-06,
+      "loss": 0.4522,
+      "step": 989
+    },
+    {
+      "epoch": 0.7290132547864506,
+      "grad_norm": 0.5408301949501038,
+      "learning_rate": 9.39091268957578e-06,
+      "loss": 0.4684,
+      "step": 990
+    },
+    {
+      "epoch": 0.7297496318114874,
+      "grad_norm": 0.6333877444267273,
+      "learning_rate": 9.388861562606059e-06,
+      "loss": 0.4797,
+      "step": 991
+    },
+    {
+      "epoch": 0.7304860088365243,
+      "grad_norm": 0.4390884041786194,
+      "learning_rate": 9.386807212589036e-06,
+      "loss": 0.4377,
+      "step": 992
+    },
+    {
+      "epoch": 0.7312223858615611,
+      "grad_norm": 0.4864395260810852,
+      "learning_rate": 9.384749641033358e-06,
+      "loss": 0.478,
+      "step": 993
+    },
+    {
+      "epoch": 0.7319587628865979,
+      "grad_norm": 0.4967077076435089,
+      "learning_rate": 9.382688849450049e-06,
+      "loss": 0.4222,
+      "step": 994
+    },
+    {
+      "epoch": 0.7326951399116347,
+      "grad_norm": 0.45611369609832764,
+      "learning_rate": 9.380624839352486e-06,
+      "loss": 0.4561,
+      "step": 995
+    },
+    {
+      "epoch": 0.7334315169366715,
+      "grad_norm": 0.5901041626930237,
+      "learning_rate": 9.37855761225642e-06,
+      "loss": 0.4795,
+      "step": 996
+    },
+    {
+      "epoch": 0.7341678939617083,
+      "grad_norm": 0.4063359498977661,
+      "learning_rate": 9.37648716967996e-06,
+      "loss": 0.4684,
+      "step": 997
+    },
+    {
+      "epoch": 0.7349042709867453,
+      "grad_norm": 0.5191907286643982,
+      "learning_rate": 9.374413513143574e-06,
+      "loss": 0.438,
+      "step": 998
+    },
+    {
+      "epoch": 0.7356406480117821,
+      "grad_norm": 0.4642137885093689,
+      "learning_rate": 9.372336644170096e-06,
+      "loss": 0.432,
+      "step": 999
+    },
+    {
+      "epoch": 0.7363770250368189,
+      "grad_norm": 0.5185231566429138,
+      "learning_rate": 9.370256564284713e-06,
+      "loss": 0.4519,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7371134020618557,
+      "grad_norm": 0.4697466790676117,
+      "learning_rate": 9.368173275014973e-06,
+      "loss": 0.4226,
+      "step": 1001
+    },
+    {
+      "epoch": 0.7378497790868925,
+      "grad_norm": 0.49580705165863037,
+      "learning_rate": 9.366086777890785e-06,
+      "loss": 0.475,
+      "step": 1002
+    },
+    {
+      "epoch": 0.7385861561119293,
+      "grad_norm": 0.5398573875427246,
+      "learning_rate": 9.363997074444402e-06,
+      "loss": 0.4605,
+      "step": 1003
+    },
+    {
+      "epoch": 0.7393225331369662,
+      "grad_norm": 0.49332237243652344,
+      "learning_rate": 9.361904166210443e-06,
+      "loss": 0.4483,
+      "step": 1004
+    },
+    {
+      "epoch": 0.740058910162003,
+      "grad_norm": 0.6060042977333069,
+      "learning_rate": 9.359808054725877e-06,
+      "loss": 0.4546,
+      "step": 1005
+    },
+    {
+      "epoch": 0.7407952871870398,
+      "grad_norm": 0.5035642385482788,
+      "learning_rate": 9.357708741530025e-06,
+      "loss": 0.4692,
+      "step": 1006
+    },
+    {
+      "epoch": 0.7415316642120766,
+      "grad_norm": 0.5561889410018921,
+      "learning_rate": 9.355606228164559e-06,
+      "loss": 0.4416,
+      "step": 1007
+    },
+    {
+      "epoch": 0.7422680412371134,
+      "grad_norm": 0.5201199650764465,
+      "learning_rate": 9.3535005161735e-06,
+      "loss": 0.4581,
+      "step": 1008
+    },
+    {
+      "epoch": 0.7430044182621502,
+      "grad_norm": 0.5415538549423218,
+      "learning_rate": 9.351391607103222e-06,
+      "loss": 0.4417,
+      "step": 1009
+    },
+    {
+      "epoch": 0.7437407952871871,
+      "grad_norm": 0.5373315811157227,
+      "learning_rate": 9.34927950250244e-06,
+      "loss": 0.4449,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7444771723122239,
+      "grad_norm": 0.6178976893424988,
+      "learning_rate": 9.347164203922224e-06,
+      "loss": 0.4936,
+      "step": 1011
+    },
+    {
+      "epoch": 0.7452135493372607,
+      "grad_norm": 0.4898339807987213,
+      "learning_rate": 9.34504571291598e-06,
+      "loss": 0.4354,
+      "step": 1012
+    },
+    {
+      "epoch": 0.7459499263622975,
+      "grad_norm": 0.6177398562431335,
+      "learning_rate": 9.34292403103947e-06,
+      "loss": 0.4385,
+      "step": 1013
+    },
+    {
+      "epoch": 0.7466863033873343,
+      "grad_norm": 0.4545745551586151,
+      "learning_rate": 9.34079915985079e-06,
+      "loss": 0.4313,
+      "step": 1014
+    },
+    {
+      "epoch": 0.7474226804123711,
+      "grad_norm": 0.5530493259429932,
+      "learning_rate": 9.33867110091038e-06,
+      "loss": 0.4351,
+      "step": 1015
+    },
+    {
+      "epoch": 0.748159057437408,
+      "grad_norm": 0.562359094619751,
+      "learning_rate": 9.336539855781027e-06,
+      "loss": 0.4467,
+      "step": 1016
+    },
+    {
+      "epoch": 0.7488954344624448,
+      "grad_norm": 0.4444253742694855,
+      "learning_rate": 9.334405426027845e-06,
+      "loss": 0.4276,
+      "step": 1017
+    },
+    {
+      "epoch": 0.7496318114874816,
+      "grad_norm": 0.6157680749893188,
+      "learning_rate": 9.332267813218303e-06,
+      "loss": 0.4383,
+      "step": 1018
+    },
+    {
+      "epoch": 0.7503681885125184,
+      "grad_norm": 0.5156718492507935,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.4623,
+      "step": 1019
+    },
+    {
+      "epoch": 0.7511045655375552,
+      "grad_norm": 0.5296850204467773,
+      "learning_rate": 9.327983044711655e-06,
+      "loss": 0.4561,
+      "step": 1020
+    },
+    {
+      "epoch": 0.751840942562592,
+      "grad_norm": 0.5704274773597717,
+      "learning_rate": 9.325835892161156e-06,
+      "loss": 0.4747,
+      "step": 1021
+    },
+    {
+      "epoch": 0.7525773195876289,
+      "grad_norm": 0.5490828156471252,
+      "learning_rate": 9.323685562847497e-06,
+      "loss": 0.4119,
+      "step": 1022
+    },
+    {
+      "epoch": 0.7533136966126657,
+      "grad_norm": 0.4620678722858429,
+      "learning_rate": 9.321532058349817e-06,
+      "loss": 0.4218,
+      "step": 1023
+    },
+    {
+      "epoch": 0.7540500736377025,
+      "grad_norm": 0.507858157157898,
+      "learning_rate": 9.31937538024959e-06,
+      "loss": 0.4328,
+      "step": 1024
+    },
+    {
+      "epoch": 0.7547864506627393,
+      "grad_norm": 0.5123826861381531,
+      "learning_rate": 9.317215530130607e-06,
+      "loss": 0.4529,
+      "step": 1025
+    },
+    {
+      "epoch": 0.7555228276877761,
+      "grad_norm": 0.4884844124317169,
+      "learning_rate": 9.315052509579004e-06,
+      "loss": 0.4718,
+      "step": 1026
+    },
+    {
+      "epoch": 0.7562592047128129,
+      "grad_norm": 0.5704168677330017,
+      "learning_rate": 9.312886320183232e-06,
+      "loss": 0.4674,
+      "step": 1027
+    },
+    {
+      "epoch": 0.7569955817378498,
+      "grad_norm": 0.4477781057357788,
+      "learning_rate": 9.310716963534077e-06,
+      "loss": 0.443,
+      "step": 1028
+    },
+    {
+      "epoch": 0.7577319587628866,
+      "grad_norm": 0.5215590000152588,
+      "learning_rate": 9.30854444122465e-06,
+      "loss": 0.4639,
+      "step": 1029
+    },
+    {
+      "epoch": 0.7584683357879234,
+      "grad_norm": 0.5299932360649109,
+      "learning_rate": 9.306368754850386e-06,
+      "loss": 0.4503,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7592047128129602,
+      "grad_norm": 0.47621920704841614,
+      "learning_rate": 9.30418990600904e-06,
+      "loss": 0.4418,
+      "step": 1031
+    },
+    {
+      "epoch": 0.759941089837997,
+      "grad_norm": 0.5789692401885986,
+      "learning_rate": 9.302007896300697e-06,
+      "loss": 0.4675,
+      "step": 1032
+    },
+    {
+      "epoch": 0.7606774668630338,
+      "grad_norm": 0.4460338354110718,
+      "learning_rate": 9.299822727327758e-06,
+      "loss": 0.4253,
+      "step": 1033
+    },
+    {
+      "epoch": 0.7614138438880707,
+      "grad_norm": 0.5220574140548706,
+      "learning_rate": 9.297634400694943e-06,
+      "loss": 0.4548,
+      "step": 1034
+    },
+    {
+      "epoch": 0.7621502209131075,
+      "grad_norm": 0.5527031421661377,
+      "learning_rate": 9.295442918009295e-06,
+      "loss": 0.4268,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7628865979381443,
+      "grad_norm": 0.4812883734703064,
+      "learning_rate": 9.29324828088017e-06,
+      "loss": 0.4721,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7636229749631811,
+      "grad_norm": 0.5152131915092468,
+      "learning_rate": 9.291050490919244e-06,
+      "loss": 0.4758,
+      "step": 1037
+    },
+    {
+      "epoch": 0.7643593519882179,
+      "grad_norm": 0.5110999941825867,
+      "learning_rate": 9.288849549740513e-06,
+      "loss": 0.4721,
+      "step": 1038
+    },
+    {
+      "epoch": 0.7650957290132547,
+      "grad_norm": 0.4943920373916626,
+      "learning_rate": 9.286645458960272e-06,
+      "loss": 0.4636,
+      "step": 1039
+    },
+    {
+      "epoch": 0.7658321060382917,
+      "grad_norm": 0.4862247109413147,
+      "learning_rate": 9.28443822019715e-06,
+      "loss": 0.4717,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7665684830633285,
+      "grad_norm": 0.4800577163696289,
+      "learning_rate": 9.282227835072064e-06,
+      "loss": 0.4624,
+      "step": 1041
+    },
+    {
+      "epoch": 0.7673048600883653,
+      "grad_norm": 0.5938796401023865,
+      "learning_rate": 9.280014305208264e-06,
+      "loss": 0.4698,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7680412371134021,
+      "grad_norm": 0.45310524106025696,
+      "learning_rate": 9.277797632231295e-06,
+      "loss": 0.4681,
+      "step": 1043
+    },
+    {
+      "epoch": 0.7687776141384389,
+      "grad_norm": 0.6024627089500427,
+      "learning_rate": 9.275577817769015e-06,
+      "loss": 0.4819,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7695139911634757,
+      "grad_norm": 0.4621509611606598,
+      "learning_rate": 9.273354863451589e-06,
+      "loss": 0.4683,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7702503681885126,
+      "grad_norm": 0.5616011619567871,
+      "learning_rate": 9.271128770911489e-06,
+      "loss": 0.4924,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7709867452135494,
+      "grad_norm": 0.5046648979187012,
+      "learning_rate": 9.268899541783487e-06,
+      "loss": 0.4244,
+      "step": 1047
+    },
+    {
+      "epoch": 0.7717231222385862,
+      "grad_norm": 0.47864627838134766,
+      "learning_rate": 9.266667177704665e-06,
+      "loss": 0.4358,
+      "step": 1048
+    },
+    {
+      "epoch": 0.772459499263623,
+      "grad_norm": 0.45622187852859497,
+      "learning_rate": 9.2644316803144e-06,
+      "loss": 0.4492,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7731958762886598,
+      "grad_norm": 0.5317423343658447,
+      "learning_rate": 9.262193051254377e-06,
+      "loss": 0.4541,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7739322533136966,
+      "grad_norm": 0.5093345046043396,
+      "learning_rate": 9.259951292168576e-06,
+      "loss": 0.4874,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7746686303387335,
+      "grad_norm": 0.4720238745212555,
+      "learning_rate": 9.257706404703276e-06,
+      "loss": 0.4477,
+      "step": 1052
+    },
+    {
+      "epoch": 0.7754050073637703,
+      "grad_norm": 0.548051118850708,
+      "learning_rate": 9.255458390507059e-06,
+      "loss": 0.4761,
+      "step": 1053
+    },
+    {
+      "epoch": 0.7761413843888071,
+      "grad_norm": 0.4737755358219147,
+      "learning_rate": 9.253207251230793e-06,
+      "loss": 0.4517,
+      "step": 1054
+    },
+    {
+      "epoch": 0.7768777614138439,
+      "grad_norm": 0.40416955947875977,
+      "learning_rate": 9.250952988527648e-06,
+      "loss": 0.4095,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7776141384388807,
+      "grad_norm": 0.4669051468372345,
+      "learning_rate": 9.248695604053091e-06,
+      "loss": 0.4492,
+      "step": 1056
+    },
+    {
+      "epoch": 0.7783505154639175,
+      "grad_norm": 0.5601681470870972,
+      "learning_rate": 9.246435099464869e-06,
+      "loss": 0.4564,
+      "step": 1057
+    },
+    {
+      "epoch": 0.7790868924889544,
+      "grad_norm": 0.44851595163345337,
+      "learning_rate": 9.244171476423037e-06,
+      "loss": 0.4675,
+      "step": 1058
+    },
+    {
+      "epoch": 0.7798232695139912,
+      "grad_norm": 0.5892722606658936,
+      "learning_rate": 9.241904736589927e-06,
+      "loss": 0.4731,
+      "step": 1059
+    },
+    {
+      "epoch": 0.780559646539028,
+      "grad_norm": 0.47606587409973145,
+      "learning_rate": 9.239634881630162e-06,
+      "loss": 0.4443,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7812960235640648,
+      "grad_norm": 0.5107343196868896,
+      "learning_rate": 9.237361913210658e-06,
+      "loss": 0.4415,
+      "step": 1061
+    },
+    {
+      "epoch": 0.7820324005891016,
+      "grad_norm": 0.593949556350708,
+      "learning_rate": 9.235085833000613e-06,
+      "loss": 0.4537,
+      "step": 1062
+    },
+    {
+      "epoch": 0.7827687776141384,
+      "grad_norm": 0.4850008189678192,
+      "learning_rate": 9.232806642671513e-06,
+      "loss": 0.465,
+      "step": 1063
+    },
+    {
+      "epoch": 0.7835051546391752,
+      "grad_norm": 0.4913601577281952,
+      "learning_rate": 9.230524343897125e-06,
+      "loss": 0.4483,
+      "step": 1064
+    },
+    {
+      "epoch": 0.7842415316642121,
+      "grad_norm": 0.4335813820362091,
+      "learning_rate": 9.228238938353502e-06,
+      "loss": 0.4083,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7849779086892489,
+      "grad_norm": 0.5363730788230896,
+      "learning_rate": 9.225950427718974e-06,
+      "loss": 0.4521,
+      "step": 1066
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 0.5813281536102295,
+      "learning_rate": 9.223658813674157e-06,
+      "loss": 0.4493,
+      "step": 1067
+    },
+    {
+      "epoch": 0.7864506627393225,
+      "grad_norm": 0.466314435005188,
+      "learning_rate": 9.221364097901941e-06,
+      "loss": 0.446,
+      "step": 1068
+    },
+    {
+      "epoch": 0.7871870397643593,
+      "grad_norm": 0.5087846517562866,
+      "learning_rate": 9.219066282087497e-06,
+      "loss": 0.4284,
+      "step": 1069
+    },
+    {
+      "epoch": 0.7879234167893961,
+      "grad_norm": 0.5554249882698059,
+      "learning_rate": 9.216765367918272e-06,
+      "loss": 0.4804,
+      "step": 1070
+    },
+    {
+      "epoch": 0.788659793814433,
+      "grad_norm": 0.48049336671829224,
+      "learning_rate": 9.214461357083986e-06,
+      "loss": 0.4659,
+      "step": 1071
+    },
+    {
+      "epoch": 0.7893961708394698,
+      "grad_norm": 0.5227876901626587,
+      "learning_rate": 9.212154251276637e-06,
+      "loss": 0.4487,
+      "step": 1072
+    },
+    {
+      "epoch": 0.7901325478645066,
+      "grad_norm": 0.5827174186706543,
+      "learning_rate": 9.20984405219049e-06,
+      "loss": 0.4681,
+      "step": 1073
+    },
+    {
+      "epoch": 0.7908689248895434,
+      "grad_norm": 0.5332369804382324,
+      "learning_rate": 9.207530761522093e-06,
+      "loss": 0.4361,
+      "step": 1074
+    },
+    {
+      "epoch": 0.7916053019145802,
+      "grad_norm": 0.525380551815033,
+      "learning_rate": 9.205214380970247e-06,
+      "loss": 0.4415,
+      "step": 1075
+    },
+    {
+      "epoch": 0.792341678939617,
+      "grad_norm": 0.4778136610984802,
+      "learning_rate": 9.20289491223604e-06,
+      "loss": 0.4416,
+      "step": 1076
+    },
+    {
+      "epoch": 0.7930780559646539,
+      "grad_norm": 0.5096407532691956,
+      "learning_rate": 9.200572357022815e-06,
+      "loss": 0.4379,
+      "step": 1077
+    },
+    {
+      "epoch": 0.7938144329896907,
+      "grad_norm": 0.48358261585235596,
+      "learning_rate": 9.198246717036187e-06,
+      "loss": 0.4547,
+      "step": 1078
+    },
+    {
+      "epoch": 0.7945508100147275,
+      "grad_norm": 0.5392265319824219,
+      "learning_rate": 9.195917993984039e-06,
+      "loss": 0.4585,
+      "step": 1079
+    },
+    {
+      "epoch": 0.7952871870397643,
+      "grad_norm": 0.4468466341495514,
+      "learning_rate": 9.19358618957651e-06,
+      "loss": 0.4572,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7960235640648011,
+      "grad_norm": 0.48624467849731445,
+      "learning_rate": 9.191251305526013e-06,
+      "loss": 0.4327,
+      "step": 1081
+    },
+    {
+      "epoch": 0.7967599410898379,
+      "grad_norm": 0.49556195735931396,
+      "learning_rate": 9.18891334354721e-06,
+      "loss": 0.4285,
+      "step": 1082
+    },
+    {
+      "epoch": 0.7974963181148749,
+      "grad_norm": 0.427839994430542,
+      "learning_rate": 9.18657230535703e-06,
+      "loss": 0.4326,
+      "step": 1083
+    },
+    {
+      "epoch": 0.7982326951399117,
+      "grad_norm": 0.5406298041343689,
+      "learning_rate": 9.184228192674667e-06,
+      "loss": 0.455,
+      "step": 1084
+    },
+    {
+      "epoch": 0.7989690721649485,
+      "grad_norm": 0.4637623131275177,
+      "learning_rate": 9.18188100722156e-06,
+      "loss": 0.4596,
+      "step": 1085
+    },
+    {
+      "epoch": 0.7997054491899853,
+      "grad_norm": 0.4427599608898163,
+      "learning_rate": 9.179530750721413e-06,
+      "loss": 0.4371,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8004418262150221,
+      "grad_norm": 0.5007920861244202,
+      "learning_rate": 9.177177424900183e-06,
+      "loss": 0.4656,
+      "step": 1087
+    },
+    {
+      "epoch": 0.801178203240059,
+      "grad_norm": 0.5360900163650513,
+      "learning_rate": 9.174821031486083e-06,
+      "loss": 0.4214,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8019145802650958,
+      "grad_norm": 0.5234165191650391,
+      "learning_rate": 9.172461572209578e-06,
+      "loss": 0.4553,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8026509572901326,
+      "grad_norm": 0.5163874626159668,
+      "learning_rate": 9.17009904880338e-06,
+      "loss": 0.447,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8033873343151694,
+      "grad_norm": 0.6014302372932434,
+      "learning_rate": 9.167733463002457e-06,
+      "loss": 0.4715,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8041237113402062,
+      "grad_norm": 0.5282832384109497,
+      "learning_rate": 9.165364816544022e-06,
+      "loss": 0.4558,
+      "step": 1092
+    },
+    {
+      "epoch": 0.804860088365243,
+      "grad_norm": 0.547073483467102,
+      "learning_rate": 9.162993111167541e-06,
+      "loss": 0.4541,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8055964653902798,
+      "grad_norm": 0.5955381393432617,
+      "learning_rate": 9.160618348614718e-06,
+      "loss": 0.4368,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8063328424153167,
+      "grad_norm": 0.4820695221424103,
+      "learning_rate": 9.158240530629512e-06,
+      "loss": 0.4562,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8070692194403535,
+      "grad_norm": 0.4931407868862152,
+      "learning_rate": 9.155859658958117e-06,
+      "loss": 0.4747,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8078055964653903,
+      "grad_norm": 0.5072289705276489,
+      "learning_rate": 9.153475735348973e-06,
+      "loss": 0.4338,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8085419734904271,
+      "grad_norm": 0.4261424243450165,
+      "learning_rate": 9.151088761552764e-06,
+      "loss": 0.4415,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8092783505154639,
+      "grad_norm": 0.5440129637718201,
+      "learning_rate": 9.148698739322409e-06,
+      "loss": 0.4326,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8100147275405007,
+      "grad_norm": 0.4529050886631012,
+      "learning_rate": 9.146305670413069e-06,
+      "loss": 0.4579,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8107511045655376,
+      "grad_norm": 0.5012593865394592,
+      "learning_rate": 9.143909556582143e-06,
+      "loss": 0.4643,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8114874815905744,
+      "grad_norm": 0.5710225701332092,
+      "learning_rate": 9.141510399589261e-06,
+      "loss": 0.4484,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8122238586156112,
+      "grad_norm": 0.4354053735733032,
+      "learning_rate": 9.139108201196296e-06,
+      "loss": 0.4333,
+      "step": 1103
+    },
+    {
+      "epoch": 0.812960235640648,
+      "grad_norm": 0.49365589022636414,
+      "learning_rate": 9.136702963167348e-06,
+      "loss": 0.457,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8136966126656848,
+      "grad_norm": 0.46688565611839294,
+      "learning_rate": 9.134294687268749e-06,
+      "loss": 0.4454,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8144329896907216,
+      "grad_norm": 0.47631922364234924,
+      "learning_rate": 9.131883375269067e-06,
+      "loss": 0.4471,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8151693667157585,
+      "grad_norm": 0.54261714220047,
+      "learning_rate": 9.129469028939094e-06,
+      "loss": 0.4839,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8159057437407953,
+      "grad_norm": 0.5108979940414429,
+      "learning_rate": 9.127051650051854e-06,
+      "loss": 0.4306,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8166421207658321,
+      "grad_norm": 0.46782127022743225,
+      "learning_rate": 9.1246312403826e-06,
+      "loss": 0.4578,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8173784977908689,
+      "grad_norm": 0.5475788116455078,
+      "learning_rate": 9.122207801708802e-06,
+      "loss": 0.4498,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8181148748159057,
+      "grad_norm": 0.5272371172904968,
+      "learning_rate": 9.119781335810164e-06,
+      "loss": 0.4482,
+      "step": 1111
+    },
+    {
+      "epoch": 0.8188512518409425,
+      "grad_norm": 0.49965953826904297,
+      "learning_rate": 9.117351844468609e-06,
+      "loss": 0.4164,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8195876288659794,
+      "grad_norm": 0.5126081109046936,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 0.4744,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8203240058910162,
+      "grad_norm": 0.5301181674003601,
+      "learning_rate": 9.112483792595547e-06,
+      "loss": 0.4508,
+      "step": 1114
+    },
+    {
+      "epoch": 0.821060382916053,
+      "grad_norm": 0.49831271171569824,
+      "learning_rate": 9.110045235638991e-06,
+      "loss": 0.4435,
+      "step": 1115
+    },
+    {
+      "epoch": 0.8217967599410898,
+      "grad_norm": 0.48843833804130554,
+      "learning_rate": 9.107603660389414e-06,
+      "loss": 0.4381,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8225331369661266,
+      "grad_norm": 0.5341619253158569,
+      "learning_rate": 9.105159068639837e-06,
+      "loss": 0.4548,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8232695139911634,
+      "grad_norm": 0.5220252275466919,
+      "learning_rate": 9.102711462185495e-06,
+      "loss": 0.4459,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8240058910162003,
+      "grad_norm": 0.4931434392929077,
+      "learning_rate": 9.100260842823831e-06,
+      "loss": 0.4393,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8247422680412371,
+      "grad_norm": 0.5014846920967102,
+      "learning_rate": 9.097807212354513e-06,
+      "loss": 0.4551,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8254786450662739,
+      "grad_norm": 0.5581266283988953,
+      "learning_rate": 9.09535057257941e-06,
+      "loss": 0.4658,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8262150220913107,
+      "grad_norm": 0.46569985151290894,
+      "learning_rate": 9.092890925302601e-06,
+      "loss": 0.4414,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8269513991163475,
+      "grad_norm": 0.45610496401786804,
+      "learning_rate": 9.090428272330381e-06,
+      "loss": 0.4561,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8276877761413843,
+      "grad_norm": 0.4498854875564575,
+      "learning_rate": 9.087962615471246e-06,
+      "loss": 0.4501,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8284241531664213,
+      "grad_norm": 0.4839475452899933,
+      "learning_rate": 9.085493956535898e-06,
+      "loss": 0.4394,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8291605301914581,
+      "grad_norm": 0.44261297583580017,
+      "learning_rate": 9.083022297337251e-06,
+      "loss": 0.4284,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8298969072164949,
+      "grad_norm": 0.49953368306159973,
+      "learning_rate": 9.080547639690411e-06,
+      "loss": 0.4698,
+      "step": 1127
+    },
+    {
+      "epoch": 0.8306332842415317,
+      "grad_norm": 0.5280026197433472,
+      "learning_rate": 9.078069985412697e-06,
+      "loss": 0.435,
+      "step": 1128
+    },
+    {
+      "epoch": 0.8313696612665685,
+      "grad_norm": 0.46925580501556396,
+      "learning_rate": 9.075589336323619e-06,
+      "loss": 0.4702,
+      "step": 1129
+    },
+    {
+      "epoch": 0.8321060382916053,
+      "grad_norm": 0.49038681387901306,
+      "learning_rate": 9.073105694244892e-06,
+      "loss": 0.4367,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8328424153166422,
+      "grad_norm": 0.564074695110321,
+      "learning_rate": 9.070619061000429e-06,
+      "loss": 0.486,
+      "step": 1131
+    },
+    {
+      "epoch": 0.833578792341679,
+      "grad_norm": 0.45910879969596863,
+      "learning_rate": 9.068129438416337e-06,
+      "loss": 0.4243,
+      "step": 1132
+    },
+    {
+      "epoch": 0.8343151693667158,
+      "grad_norm": 0.5710626840591431,
+      "learning_rate": 9.065636828320919e-06,
+      "loss": 0.434,
+      "step": 1133
+    },
+    {
+      "epoch": 0.8350515463917526,
+      "grad_norm": 0.504901647567749,
+      "learning_rate": 9.063141232544676e-06,
+      "loss": 0.4441,
+      "step": 1134
+    },
+    {
+      "epoch": 0.8357879234167894,
+      "grad_norm": 0.5790115594863892,
+      "learning_rate": 9.060642652920295e-06,
+      "loss": 0.457,
+      "step": 1135
+    },
+    {
+      "epoch": 0.8365243004418262,
+      "grad_norm": 0.5210295915603638,
+      "learning_rate": 9.058141091282656e-06,
+      "loss": 0.4539,
+      "step": 1136
+    },
+    {
+      "epoch": 0.8372606774668631,
+      "grad_norm": 0.48215651512145996,
+      "learning_rate": 9.055636549468833e-06,
+      "loss": 0.4558,
+      "step": 1137
+    },
+    {
+      "epoch": 0.8379970544918999,
+      "grad_norm": 0.42006632685661316,
+      "learning_rate": 9.053129029318086e-06,
+      "loss": 0.424,
+      "step": 1138
+    },
+    {
+      "epoch": 0.8387334315169367,
+      "grad_norm": 0.45918166637420654,
+      "learning_rate": 9.050618532671862e-06,
+      "loss": 0.4302,
+      "step": 1139
+    },
+    {
+      "epoch": 0.8394698085419735,
+      "grad_norm": 0.5937727093696594,
+      "learning_rate": 9.048105061373793e-06,
+      "loss": 0.4467,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8402061855670103,
+      "grad_norm": 0.4482143521308899,
+      "learning_rate": 9.045588617269694e-06,
+      "loss": 0.4423,
+      "step": 1141
+    },
+    {
+      "epoch": 0.8409425625920471,
+      "grad_norm": 0.49220171570777893,
+      "learning_rate": 9.043069202207571e-06,
+      "loss": 0.4455,
+      "step": 1142
+    },
+    {
+      "epoch": 0.841678939617084,
+      "grad_norm": 0.4809843599796295,
+      "learning_rate": 9.040546818037602e-06,
+      "loss": 0.4628,
+      "step": 1143
+    },
+    {
+      "epoch": 0.8424153166421208,
+      "grad_norm": 0.49604564905166626,
+      "learning_rate": 9.038021466612151e-06,
+      "loss": 0.4283,
+      "step": 1144
+    },
+    {
+      "epoch": 0.8431516936671576,
+      "grad_norm": 0.4608924388885498,
+      "learning_rate": 9.035493149785758e-06,
+      "loss": 0.4359,
+      "step": 1145
+    },
+    {
+      "epoch": 0.8438880706921944,
+      "grad_norm": 0.5311315059661865,
+      "learning_rate": 9.032961869415147e-06,
+      "loss": 0.4655,
+      "step": 1146
+    },
+    {
+      "epoch": 0.8446244477172312,
+      "grad_norm": 0.491970032453537,
+      "learning_rate": 9.03042762735921e-06,
+      "loss": 0.4477,
+      "step": 1147
+    },
+    {
+      "epoch": 0.845360824742268,
+      "grad_norm": 0.5328875184059143,
+      "learning_rate": 9.027890425479016e-06,
+      "loss": 0.4565,
+      "step": 1148
+    },
+    {
+      "epoch": 0.8460972017673049,
+      "grad_norm": 0.46462082862854004,
+      "learning_rate": 9.025350265637816e-06,
+      "loss": 0.4307,
+      "step": 1149
+    },
+    {
+      "epoch": 0.8468335787923417,
+      "grad_norm": 0.568123996257782,
+      "learning_rate": 9.02280714970102e-06,
+      "loss": 0.4527,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8475699558173785,
+      "grad_norm": 0.39277321100234985,
+      "learning_rate": 9.02026107953622e-06,
+      "loss": 0.4382,
+      "step": 1151
+    },
+    {
+      "epoch": 0.8483063328424153,
+      "grad_norm": 0.46049660444259644,
+      "learning_rate": 9.01771205701317e-06,
+      "loss": 0.4496,
+      "step": 1152
+    },
+    {
+      "epoch": 0.8490427098674521,
+      "grad_norm": 0.5467735528945923,
+      "learning_rate": 9.015160084003798e-06,
+      "loss": 0.462,
+      "step": 1153
+    },
+    {
+      "epoch": 0.8497790868924889,
+      "grad_norm": 0.46097782254219055,
+      "learning_rate": 9.012605162382194e-06,
+      "loss": 0.4612,
+      "step": 1154
+    },
+    {
+      "epoch": 0.8505154639175257,
+      "grad_norm": 0.5207246541976929,
+      "learning_rate": 9.010047294024615e-06,
+      "loss": 0.4397,
+      "step": 1155
+    },
+    {
+      "epoch": 0.8512518409425626,
+      "grad_norm": 0.5254480838775635,
+      "learning_rate": 9.007486480809482e-06,
+      "loss": 0.4787,
+      "step": 1156
+    },
+    {
+      "epoch": 0.8519882179675994,
+      "grad_norm": 0.45604032278060913,
+      "learning_rate": 9.00492272461738e-06,
+      "loss": 0.4607,
+      "step": 1157
+    },
+    {
+      "epoch": 0.8527245949926362,
+      "grad_norm": 0.4617128372192383,
+      "learning_rate": 9.002356027331055e-06,
+      "loss": 0.4236,
+      "step": 1158
+    },
+    {
+      "epoch": 0.853460972017673,
+      "grad_norm": 0.5179611444473267,
+      "learning_rate": 8.999786390835408e-06,
+      "loss": 0.4647,
+      "step": 1159
+    },
+    {
+      "epoch": 0.8541973490427098,
+      "grad_norm": 0.4486055374145508,
+      "learning_rate": 8.997213817017508e-06,
+      "loss": 0.4423,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8549337260677466,
+      "grad_norm": 0.5427169799804688,
+      "learning_rate": 8.99463830776657e-06,
+      "loss": 0.4648,
+      "step": 1161
+    },
+    {
+      "epoch": 0.8556701030927835,
+      "grad_norm": 0.5229467749595642,
+      "learning_rate": 8.992059864973972e-06,
+      "loss": 0.4559,
+      "step": 1162
+    },
+    {
+      "epoch": 0.8564064801178203,
+      "grad_norm": 0.5158513784408569,
+      "learning_rate": 8.989478490533247e-06,
+      "loss": 0.4168,
+      "step": 1163
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.47246885299682617,
+      "learning_rate": 8.986894186340075e-06,
+      "loss": 0.4156,
+      "step": 1164
+    },
+    {
+      "epoch": 0.8578792341678939,
+      "grad_norm": 0.5099225044250488,
+      "learning_rate": 8.984306954292293e-06,
+      "loss": 0.4577,
+      "step": 1165
+    },
+    {
+      "epoch": 0.8586156111929307,
+      "grad_norm": 0.5285944938659668,
+      "learning_rate": 8.981716796289886e-06,
+      "loss": 0.4423,
+      "step": 1166
+    },
+    {
+      "epoch": 0.8593519882179675,
+      "grad_norm": 0.43009987473487854,
+      "learning_rate": 8.979123714234986e-06,
+      "loss": 0.4311,
+      "step": 1167
+    },
+    {
+      "epoch": 0.8600883652430045,
+      "grad_norm": 0.5090437531471252,
+      "learning_rate": 8.976527710031875e-06,
+      "loss": 0.4636,
+      "step": 1168
+    },
+    {
+      "epoch": 0.8608247422680413,
+      "grad_norm": 0.44846466183662415,
+      "learning_rate": 8.97392878558698e-06,
+      "loss": 0.4243,
+      "step": 1169
+    },
+    {
+      "epoch": 0.8615611192930781,
+      "grad_norm": 0.5008236765861511,
+      "learning_rate": 8.971326942808876e-06,
+      "loss": 0.4539,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8622974963181149,
+      "grad_norm": 0.5144447088241577,
+      "learning_rate": 8.968722183608272e-06,
+      "loss": 0.4791,
+      "step": 1171
+    },
+    {
+      "epoch": 0.8630338733431517,
+      "grad_norm": 0.4842917025089264,
+      "learning_rate": 8.966114509898026e-06,
+      "loss": 0.4568,
+      "step": 1172
+    },
+    {
+      "epoch": 0.8637702503681886,
+      "grad_norm": 0.4209856688976288,
+      "learning_rate": 8.963503923593138e-06,
+      "loss": 0.4375,
+      "step": 1173
+    },
+    {
+      "epoch": 0.8645066273932254,
+      "grad_norm": 0.49867910146713257,
+      "learning_rate": 8.96089042661074e-06,
+      "loss": 0.4882,
+      "step": 1174
+    },
+    {
+      "epoch": 0.8652430044182622,
+      "grad_norm": 0.43283653259277344,
+      "learning_rate": 8.958274020870107e-06,
+      "loss": 0.437,
+      "step": 1175
+    },
+    {
+      "epoch": 0.865979381443299,
+      "grad_norm": 0.4610222578048706,
+      "learning_rate": 8.955654708292647e-06,
+      "loss": 0.4323,
+      "step": 1176
+    },
+    {
+      "epoch": 0.8667157584683358,
+      "grad_norm": 0.5031896829605103,
+      "learning_rate": 8.953032490801908e-06,
+      "loss": 0.4147,
+      "step": 1177
+    },
+    {
+      "epoch": 0.8674521354933726,
+      "grad_norm": 0.45876145362854004,
+      "learning_rate": 8.950407370323563e-06,
+      "loss": 0.4496,
+      "step": 1178
+    },
+    {
+      "epoch": 0.8681885125184094,
+      "grad_norm": 0.4437890648841858,
+      "learning_rate": 8.94777934878542e-06,
+      "loss": 0.4182,
+      "step": 1179
+    },
+    {
+      "epoch": 0.8689248895434463,
+      "grad_norm": 0.49486371874809265,
+      "learning_rate": 8.945148428117423e-06,
+      "loss": 0.4497,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8696612665684831,
+      "grad_norm": 0.554208517074585,
+      "learning_rate": 8.942514610251639e-06,
+      "loss": 0.4753,
+      "step": 1181
+    },
+    {
+      "epoch": 0.8703976435935199,
+      "grad_norm": 0.4523577094078064,
+      "learning_rate": 8.939877897122262e-06,
+      "loss": 0.4476,
+      "step": 1182
+    },
+    {
+      "epoch": 0.8711340206185567,
+      "grad_norm": 0.43716442584991455,
+      "learning_rate": 8.937238290665617e-06,
+      "loss": 0.43,
+      "step": 1183
+    },
+    {
+      "epoch": 0.8718703976435935,
+      "grad_norm": 0.4811909794807434,
+      "learning_rate": 8.934595792820152e-06,
+      "loss": 0.4329,
+      "step": 1184
+    },
+    {
+      "epoch": 0.8726067746686303,
+      "grad_norm": 0.5379281044006348,
+      "learning_rate": 8.931950405526436e-06,
+      "loss": 0.4529,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8733431516936672,
+      "grad_norm": 0.4634367525577545,
+      "learning_rate": 8.92930213072716e-06,
+      "loss": 0.4269,
+      "step": 1186
+    },
+    {
+      "epoch": 0.874079528718704,
+      "grad_norm": 0.46587860584259033,
+      "learning_rate": 8.926650970367138e-06,
+      "loss": 0.4399,
+      "step": 1187
+    },
+    {
+      "epoch": 0.8748159057437408,
+      "grad_norm": 0.47557711601257324,
+      "learning_rate": 8.923996926393306e-06,
+      "loss": 0.4365,
+      "step": 1188
+    },
+    {
+      "epoch": 0.8755522827687776,
+      "grad_norm": 0.48660027980804443,
+      "learning_rate": 8.921340000754708e-06,
+      "loss": 0.4772,
+      "step": 1189
+    },
+    {
+      "epoch": 0.8762886597938144,
+      "grad_norm": 0.4610178470611572,
+      "learning_rate": 8.918680195402512e-06,
+      "loss": 0.4276,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8770250368188512,
+      "grad_norm": 0.41225728392601013,
+      "learning_rate": 8.916017512290001e-06,
+      "loss": 0.4269,
+      "step": 1191
+    },
+    {
+      "epoch": 0.8777614138438881,
+      "grad_norm": 0.4658215641975403,
+      "learning_rate": 8.913351953372565e-06,
+      "loss": 0.4116,
+      "step": 1192
+    },
+    {
+      "epoch": 0.8784977908689249,
+      "grad_norm": 0.4793921709060669,
+      "learning_rate": 8.910683520607713e-06,
+      "loss": 0.4811,
+      "step": 1193
+    },
+    {
+      "epoch": 0.8792341678939617,
+      "grad_norm": 0.4011600613594055,
+      "learning_rate": 8.90801221595506e-06,
+      "loss": 0.4319,
+      "step": 1194
+    },
+    {
+      "epoch": 0.8799705449189985,
+      "grad_norm": 0.4331344962120056,
+      "learning_rate": 8.90533804137633e-06,
+      "loss": 0.4457,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8807069219440353,
+      "grad_norm": 0.5520123839378357,
+      "learning_rate": 8.902660998835359e-06,
+      "loss": 0.4549,
+      "step": 1196
+    },
+    {
+      "epoch": 0.8814432989690721,
+      "grad_norm": 0.4395567774772644,
+      "learning_rate": 8.899981090298084e-06,
+      "loss": 0.4334,
+      "step": 1197
+    },
+    {
+      "epoch": 0.882179675994109,
+      "grad_norm": 0.4339301288127899,
+      "learning_rate": 8.89729831773255e-06,
+      "loss": 0.4518,
+      "step": 1198
+    },
+    {
+      "epoch": 0.8829160530191458,
+      "grad_norm": 0.4404997229576111,
+      "learning_rate": 8.894612683108905e-06,
+      "loss": 0.4428,
+      "step": 1199
+    },
+    {
+      "epoch": 0.8836524300441826,
+      "grad_norm": 0.47511598467826843,
+      "learning_rate": 8.891924188399395e-06,
+      "loss": 0.4455,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8843888070692194,
+      "grad_norm": 0.5081692934036255,
+      "learning_rate": 8.889232835578372e-06,
+      "loss": 0.5261,
+      "step": 1201
+    },
+    {
+      "epoch": 0.8851251840942562,
+      "grad_norm": 0.5229424834251404,
+      "learning_rate": 8.886538626622282e-06,
+      "loss": 0.4371,
+      "step": 1202
+    },
+    {
+      "epoch": 0.885861561119293,
+      "grad_norm": 0.45727550983428955,
+      "learning_rate": 8.883841563509671e-06,
+      "loss": 0.414,
+      "step": 1203
+    },
+    {
+      "epoch": 0.8865979381443299,
+      "grad_norm": 0.4745890498161316,
+      "learning_rate": 8.881141648221185e-06,
+      "loss": 0.472,
+      "step": 1204
+    },
+    {
+      "epoch": 0.8873343151693667,
+      "grad_norm": 0.48544570803642273,
+      "learning_rate": 8.878438882739554e-06,
+      "loss": 0.4719,
+      "step": 1205
+    },
+    {
+      "epoch": 0.8880706921944035,
+      "grad_norm": 0.45275935530662537,
+      "learning_rate": 8.87573326904961e-06,
+      "loss": 0.4395,
+      "step": 1206
+    },
+    {
+      "epoch": 0.8888070692194403,
+      "grad_norm": 0.527344822883606,
+      "learning_rate": 8.873024809138272e-06,
+      "loss": 0.4796,
+      "step": 1207
+    },
+    {
+      "epoch": 0.8895434462444771,
+      "grad_norm": 0.44065412878990173,
+      "learning_rate": 8.870313504994556e-06,
+      "loss": 0.4587,
+      "step": 1208
+    },
+    {
+      "epoch": 0.8902798232695139,
+      "grad_norm": 0.5130428671836853,
+      "learning_rate": 8.867599358609557e-06,
+      "loss": 0.4672,
+      "step": 1209
+    },
+    {
+      "epoch": 0.8910162002945509,
+      "grad_norm": 0.4739178419113159,
+      "learning_rate": 8.864882371976466e-06,
+      "loss": 0.4384,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8917525773195877,
+      "grad_norm": 0.4686276614665985,
+      "learning_rate": 8.862162547090551e-06,
+      "loss": 0.4473,
+      "step": 1211
+    },
+    {
+      "epoch": 0.8924889543446245,
+      "grad_norm": 0.43980199098587036,
+      "learning_rate": 8.859439885949175e-06,
+      "loss": 0.4568,
+      "step": 1212
+    },
+    {
+      "epoch": 0.8932253313696613,
+      "grad_norm": 0.43443796038627625,
+      "learning_rate": 8.856714390551774e-06,
+      "loss": 0.4529,
+      "step": 1213
+    },
+    {
+      "epoch": 0.8939617083946981,
+      "grad_norm": 0.514228343963623,
+      "learning_rate": 8.853986062899869e-06,
+      "loss": 0.4468,
+      "step": 1214
+    },
+    {
+      "epoch": 0.894698085419735,
+      "grad_norm": 0.4369834363460541,
+      "learning_rate": 8.851254904997062e-06,
+      "loss": 0.4721,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8954344624447718,
+      "grad_norm": 0.5049661993980408,
+      "learning_rate": 8.848520918849035e-06,
+      "loss": 0.4094,
+      "step": 1216
+    },
+    {
+      "epoch": 0.8961708394698086,
+      "grad_norm": 0.4989210069179535,
+      "learning_rate": 8.845784106463545e-06,
+      "loss": 0.4519,
+      "step": 1217
+    },
+    {
+      "epoch": 0.8969072164948454,
+      "grad_norm": 0.4148315489292145,
+      "learning_rate": 8.84304446985042e-06,
+      "loss": 0.4464,
+      "step": 1218
+    },
+    {
+      "epoch": 0.8976435935198822,
+      "grad_norm": 0.43474555015563965,
+      "learning_rate": 8.84030201102157e-06,
+      "loss": 0.4372,
+      "step": 1219
+    },
+    {
+      "epoch": 0.898379970544919,
+      "grad_norm": 0.49237704277038574,
+      "learning_rate": 8.837556731990973e-06,
+      "loss": 0.4582,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8991163475699558,
+      "grad_norm": 0.46191561222076416,
+      "learning_rate": 8.83480863477468e-06,
+      "loss": 0.4557,
+      "step": 1221
+    },
+    {
+      "epoch": 0.8998527245949927,
+      "grad_norm": 0.41119834780693054,
+      "learning_rate": 8.832057721390809e-06,
+      "loss": 0.4343,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9005891016200295,
+      "grad_norm": 0.4598017632961273,
+      "learning_rate": 8.829303993859548e-06,
+      "loss": 0.4636,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9013254786450663,
+      "grad_norm": 0.4588959515094757,
+      "learning_rate": 8.826547454203152e-06,
+      "loss": 0.4331,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9020618556701031,
+      "grad_norm": 0.44713202118873596,
+      "learning_rate": 8.823788104445941e-06,
+      "loss": 0.4329,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9027982326951399,
+      "grad_norm": 0.5540334582328796,
+      "learning_rate": 8.821025946614295e-06,
+      "loss": 0.4508,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9035346097201767,
+      "grad_norm": 0.42183777689933777,
+      "learning_rate": 8.818260982736662e-06,
+      "loss": 0.4633,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9042709867452136,
+      "grad_norm": 0.5915690064430237,
+      "learning_rate": 8.815493214843546e-06,
+      "loss": 0.4515,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9050073637702504,
+      "grad_norm": 0.5356271862983704,
+      "learning_rate": 8.812722644967515e-06,
+      "loss": 0.4205,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9057437407952872,
+      "grad_norm": 0.4718015789985657,
+      "learning_rate": 8.809949275143189e-06,
+      "loss": 0.4572,
+      "step": 1230
+    },
+    {
+      "epoch": 0.906480117820324,
+      "grad_norm": 0.5468553900718689,
+      "learning_rate": 8.807173107407248e-06,
+      "loss": 0.4489,
+      "step": 1231
+    },
+    {
+      "epoch": 0.9072164948453608,
+      "grad_norm": 0.559502899646759,
+      "learning_rate": 8.804394143798426e-06,
+      "loss": 0.4629,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9079528718703976,
+      "grad_norm": 0.47452765703201294,
+      "learning_rate": 8.801612386357508e-06,
+      "loss": 0.4457,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9086892488954345,
+      "grad_norm": 0.5304099917411804,
+      "learning_rate": 8.798827837127336e-06,
+      "loss": 0.4492,
+      "step": 1234
+    },
+    {
+      "epoch": 0.9094256259204713,
+      "grad_norm": 0.5169060826301575,
+      "learning_rate": 8.796040498152797e-06,
+      "loss": 0.4748,
+      "step": 1235
+    },
+    {
+      "epoch": 0.9101620029455081,
+      "grad_norm": 0.49267488718032837,
+      "learning_rate": 8.793250371480827e-06,
+      "loss": 0.4579,
+      "step": 1236
+    },
+    {
+      "epoch": 0.9108983799705449,
+      "grad_norm": 0.4913206100463867,
+      "learning_rate": 8.790457459160414e-06,
+      "loss": 0.4255,
+      "step": 1237
+    },
+    {
+      "epoch": 0.9116347569955817,
+      "grad_norm": 0.531548798084259,
+      "learning_rate": 8.787661763242585e-06,
+      "loss": 0.4728,
+      "step": 1238
+    },
+    {
+      "epoch": 0.9123711340206185,
+      "grad_norm": 0.52623450756073,
+      "learning_rate": 8.784863285780419e-06,
+      "loss": 0.4661,
+      "step": 1239
+    },
+    {
+      "epoch": 0.9131075110456554,
+      "grad_norm": 0.5517471432685852,
+      "learning_rate": 8.782062028829028e-06,
+      "loss": 0.4553,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9138438880706922,
+      "grad_norm": 0.472700297832489,
+      "learning_rate": 8.779257994445574e-06,
+      "loss": 0.4422,
+      "step": 1241
+    },
+    {
+      "epoch": 0.914580265095729,
+      "grad_norm": 0.44898468255996704,
+      "learning_rate": 8.776451184689253e-06,
+      "loss": 0.4124,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9153166421207658,
+      "grad_norm": 0.48845183849334717,
+      "learning_rate": 8.773641601621303e-06,
+      "loss": 0.4469,
+      "step": 1243
+    },
+    {
+      "epoch": 0.9160530191458026,
+      "grad_norm": 0.5202846527099609,
+      "learning_rate": 8.770829247304998e-06,
+      "loss": 0.4723,
+      "step": 1244
+    },
+    {
+      "epoch": 0.9167893961708394,
+      "grad_norm": 0.44977980852127075,
+      "learning_rate": 8.768014123805642e-06,
+      "loss": 0.4413,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9175257731958762,
+      "grad_norm": 0.46154212951660156,
+      "learning_rate": 8.765196233190579e-06,
+      "loss": 0.4411,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9182621502209131,
+      "grad_norm": 0.5938853621482849,
+      "learning_rate": 8.762375577529184e-06,
+      "loss": 0.4399,
+      "step": 1247
+    },
+    {
+      "epoch": 0.9189985272459499,
+      "grad_norm": 0.43907907605171204,
+      "learning_rate": 8.75955215889286e-06,
+      "loss": 0.4402,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9197349042709867,
+      "grad_norm": 0.4637261629104614,
+      "learning_rate": 8.756725979355039e-06,
+      "loss": 0.4587,
+      "step": 1249
+    },
+    {
+      "epoch": 0.9204712812960235,
+      "grad_norm": 0.45832645893096924,
+      "learning_rate": 8.753897040991183e-06,
+      "loss": 0.4579,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9212076583210603,
+      "grad_norm": 0.44324028491973877,
+      "learning_rate": 8.751065345878778e-06,
+      "loss": 0.453,
+      "step": 1251
+    },
+    {
+      "epoch": 0.9219440353460973,
+      "grad_norm": 0.5125420689582825,
+      "learning_rate": 8.748230896097338e-06,
+      "loss": 0.4422,
+      "step": 1252
+    },
+    {
+      "epoch": 0.9226804123711341,
+      "grad_norm": 0.5083498954772949,
+      "learning_rate": 8.745393693728395e-06,
+      "loss": 0.4661,
+      "step": 1253
+    },
+    {
+      "epoch": 0.9234167893961709,
+      "grad_norm": 0.4523887038230896,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.4469,
+      "step": 1254
+    },
+    {
+      "epoch": 0.9241531664212077,
+      "grad_norm": 0.5064306259155273,
+      "learning_rate": 8.739711039564245e-06,
+      "loss": 0.4462,
+      "step": 1255
+    },
+    {
+      "epoch": 0.9248895434462445,
+      "grad_norm": 0.48103612661361694,
+      "learning_rate": 8.736865591942208e-06,
+      "loss": 0.4334,
+      "step": 1256
+    },
+    {
+      "epoch": 0.9256259204712813,
+      "grad_norm": 0.43229183554649353,
+      "learning_rate": 8.734017400079002e-06,
+      "loss": 0.4749,
+      "step": 1257
+    },
+    {
+      "epoch": 0.9263622974963182,
+      "grad_norm": 0.47184067964553833,
+      "learning_rate": 8.731166466066258e-06,
+      "loss": 0.4692,
+      "step": 1258
+    },
+    {
+      "epoch": 0.927098674521355,
+      "grad_norm": 0.49023813009262085,
+      "learning_rate": 8.728312791997612e-06,
+      "loss": 0.4593,
+      "step": 1259
+    },
+    {
+      "epoch": 0.9278350515463918,
+      "grad_norm": 0.5471561551094055,
+      "learning_rate": 8.725456379968717e-06,
+      "loss": 0.465,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 0.41933873295783997,
+      "learning_rate": 8.722597232077236e-06,
+      "loss": 0.4507,
+      "step": 1261
+    },
+    {
+      "epoch": 0.9293078055964654,
+      "grad_norm": 0.4452081322669983,
+      "learning_rate": 8.71973535042284e-06,
+      "loss": 0.4604,
+      "step": 1262
+    },
+    {
+      "epoch": 0.9300441826215022,
+      "grad_norm": 0.40033042430877686,
+      "learning_rate": 8.716870737107211e-06,
+      "loss": 0.4578,
+      "step": 1263
+    },
+    {
+      "epoch": 0.930780559646539,
+      "grad_norm": 0.4695224463939667,
+      "learning_rate": 8.714003394234031e-06,
+      "loss": 0.4582,
+      "step": 1264
+    },
+    {
+      "epoch": 0.9315169366715759,
+      "grad_norm": 0.40169644355773926,
+      "learning_rate": 8.711133323908993e-06,
+      "loss": 0.4523,
+      "step": 1265
+    },
+    {
+      "epoch": 0.9322533136966127,
+      "grad_norm": 0.43690577149391174,
+      "learning_rate": 8.708260528239788e-06,
+      "loss": 0.4569,
+      "step": 1266
+    },
+    {
+      "epoch": 0.9329896907216495,
+      "grad_norm": 0.44741305708885193,
+      "learning_rate": 8.705385009336111e-06,
+      "loss": 0.4363,
+      "step": 1267
+    },
+    {
+      "epoch": 0.9337260677466863,
+      "grad_norm": 0.6049298644065857,
+      "learning_rate": 8.702506769309656e-06,
+      "loss": 0.4569,
+      "step": 1268
+    },
+    {
+      "epoch": 0.9344624447717231,
+      "grad_norm": 0.43075326085090637,
+      "learning_rate": 8.699625810274115e-06,
+      "loss": 0.4377,
+      "step": 1269
+    },
+    {
+      "epoch": 0.93519882179676,
+      "grad_norm": 0.5428841710090637,
+      "learning_rate": 8.696742134345178e-06,
+      "loss": 0.437,
+      "step": 1270
+    },
+    {
+      "epoch": 0.9359351988217968,
+      "grad_norm": 0.5976313948631287,
+      "learning_rate": 8.69385574364053e-06,
+      "loss": 0.4581,
+      "step": 1271
+    },
+    {
+      "epoch": 0.9366715758468336,
+      "grad_norm": 0.48488497734069824,
+      "learning_rate": 8.690966640279846e-06,
+      "loss": 0.448,
+      "step": 1272
+    },
+    {
+      "epoch": 0.9374079528718704,
+      "grad_norm": 0.5577574372291565,
+      "learning_rate": 8.688074826384801e-06,
+      "loss": 0.4572,
+      "step": 1273
+    },
+    {
+      "epoch": 0.9381443298969072,
+      "grad_norm": 0.459681898355484,
+      "learning_rate": 8.685180304079051e-06,
+      "loss": 0.4488,
+      "step": 1274
+    },
+    {
+      "epoch": 0.938880706921944,
+      "grad_norm": 0.5070257782936096,
+      "learning_rate": 8.682283075488249e-06,
+      "loss": 0.4495,
+      "step": 1275
+    },
+    {
+      "epoch": 0.9396170839469808,
+      "grad_norm": 0.486531525850296,
+      "learning_rate": 8.679383142740033e-06,
+      "loss": 0.4402,
+      "step": 1276
+    },
+    {
+      "epoch": 0.9403534609720177,
+      "grad_norm": 0.5276297330856323,
+      "learning_rate": 8.676480507964021e-06,
+      "loss": 0.4487,
+      "step": 1277
+    },
+    {
+      "epoch": 0.9410898379970545,
+      "grad_norm": 0.47423624992370605,
+      "learning_rate": 8.673575173291826e-06,
+      "loss": 0.449,
+      "step": 1278
+    },
+    {
+      "epoch": 0.9418262150220913,
+      "grad_norm": 0.5563042163848877,
+      "learning_rate": 8.670667140857034e-06,
+      "loss": 0.4449,
+      "step": 1279
+    },
+    {
+      "epoch": 0.9425625920471281,
+      "grad_norm": 0.47434091567993164,
+      "learning_rate": 8.667756412795217e-06,
+      "loss": 0.4455,
+      "step": 1280
+    },
+    {
+      "epoch": 0.9432989690721649,
+      "grad_norm": 0.6123890280723572,
+      "learning_rate": 8.664842991243927e-06,
+      "loss": 0.4782,
+      "step": 1281
+    },
+    {
+      "epoch": 0.9440353460972017,
+      "grad_norm": 0.42824527621269226,
+      "learning_rate": 8.661926878342692e-06,
+      "loss": 0.4512,
+      "step": 1282
+    },
+    {
+      "epoch": 0.9447717231222386,
+      "grad_norm": 0.5722915530204773,
+      "learning_rate": 8.659008076233016e-06,
+      "loss": 0.4485,
+      "step": 1283
+    },
+    {
+      "epoch": 0.9455081001472754,
+      "grad_norm": 0.5747035145759583,
+      "learning_rate": 8.656086587058381e-06,
+      "loss": 0.4503,
+      "step": 1284
+    },
+    {
+      "epoch": 0.9462444771723122,
+      "grad_norm": 0.4897293150424957,
+      "learning_rate": 8.65316241296424e-06,
+      "loss": 0.4605,
+      "step": 1285
+    },
+    {
+      "epoch": 0.946980854197349,
+      "grad_norm": 0.5299695134162903,
+      "learning_rate": 8.650235556098017e-06,
+      "loss": 0.4241,
+      "step": 1286
+    },
+    {
+      "epoch": 0.9477172312223858,
+      "grad_norm": 0.5436539053916931,
+      "learning_rate": 8.647306018609108e-06,
+      "loss": 0.4211,
+      "step": 1287
+    },
+    {
+      "epoch": 0.9484536082474226,
+      "grad_norm": 0.49147042632102966,
+      "learning_rate": 8.644373802648877e-06,
+      "loss": 0.4528,
+      "step": 1288
+    },
+    {
+      "epoch": 0.9491899852724595,
+      "grad_norm": 0.5229616165161133,
+      "learning_rate": 8.641438910370655e-06,
+      "loss": 0.4492,
+      "step": 1289
+    },
+    {
+      "epoch": 0.9499263622974963,
+      "grad_norm": 0.49656403064727783,
+      "learning_rate": 8.638501343929735e-06,
+      "loss": 0.4382,
+      "step": 1290
+    },
+    {
+      "epoch": 0.9506627393225331,
+      "grad_norm": 0.47837382555007935,
+      "learning_rate": 8.635561105483384e-06,
+      "loss": 0.469,
+      "step": 1291
+    },
+    {
+      "epoch": 0.9513991163475699,
+      "grad_norm": 0.48511695861816406,
+      "learning_rate": 8.632618197190817e-06,
+      "loss": 0.4372,
+      "step": 1292
+    },
+    {
+      "epoch": 0.9521354933726067,
+      "grad_norm": 0.5691758394241333,
+      "learning_rate": 8.629672621213221e-06,
+      "loss": 0.4331,
+      "step": 1293
+    },
+    {
+      "epoch": 0.9528718703976435,
+      "grad_norm": 0.46881407499313354,
+      "learning_rate": 8.626724379713736e-06,
+      "loss": 0.4669,
+      "step": 1294
+    },
+    {
+      "epoch": 0.9536082474226805,
+      "grad_norm": 0.5768502354621887,
+      "learning_rate": 8.623773474857461e-06,
+      "loss": 0.4613,
+      "step": 1295
+    },
+    {
+      "epoch": 0.9543446244477173,
+      "grad_norm": 0.49087420105934143,
+      "learning_rate": 8.620819908811455e-06,
+      "loss": 0.4296,
+      "step": 1296
+    },
+    {
+      "epoch": 0.9550810014727541,
+      "grad_norm": 0.45804309844970703,
+      "learning_rate": 8.617863683744726e-06,
+      "loss": 0.4376,
+      "step": 1297
+    },
+    {
+      "epoch": 0.9558173784977909,
+      "grad_norm": 0.47427794337272644,
+      "learning_rate": 8.614904801828234e-06,
+      "loss": 0.4583,
+      "step": 1298
+    },
+    {
+      "epoch": 0.9565537555228277,
+      "grad_norm": 0.44913750886917114,
+      "learning_rate": 8.611943265234895e-06,
+      "loss": 0.4409,
+      "step": 1299
+    },
+    {
+      "epoch": 0.9572901325478645,
+      "grad_norm": 0.491207480430603,
+      "learning_rate": 8.608979076139572e-06,
+      "loss": 0.449,
+      "step": 1300
+    },
+    {
+      "epoch": 0.9580265095729014,
+      "grad_norm": 0.43761712312698364,
+      "learning_rate": 8.606012236719073e-06,
+      "loss": 0.442,
+      "step": 1301
+    },
+    {
+      "epoch": 0.9587628865979382,
+      "grad_norm": 0.39326536655426025,
+      "learning_rate": 8.60304274915216e-06,
+      "loss": 0.4258,
+      "step": 1302
+    },
+    {
+      "epoch": 0.959499263622975,
+      "grad_norm": 0.4258727431297302,
+      "learning_rate": 8.600070615619528e-06,
+      "loss": 0.4392,
+      "step": 1303
+    },
+    {
+      "epoch": 0.9602356406480118,
+      "grad_norm": 0.4401472210884094,
+      "learning_rate": 8.597095838303831e-06,
+      "loss": 0.4483,
+      "step": 1304
+    },
+    {
+      "epoch": 0.9609720176730486,
+      "grad_norm": 0.4828743040561676,
+      "learning_rate": 8.594118419389648e-06,
+      "loss": 0.4307,
+      "step": 1305
+    },
+    {
+      "epoch": 0.9617083946980854,
+      "grad_norm": 0.4741956889629364,
+      "learning_rate": 8.591138361063508e-06,
+      "loss": 0.4331,
+      "step": 1306
+    },
+    {
+      "epoch": 0.9624447717231223,
+      "grad_norm": 0.4706476926803589,
+      "learning_rate": 8.588155665513877e-06,
+      "loss": 0.4425,
+      "step": 1307
+    },
+    {
+      "epoch": 0.9631811487481591,
+      "grad_norm": 0.4412619471549988,
+      "learning_rate": 8.585170334931156e-06,
+      "loss": 0.4312,
+      "step": 1308
+    },
+    {
+      "epoch": 0.9639175257731959,
+      "grad_norm": 0.4449722170829773,
+      "learning_rate": 8.58218237150768e-06,
+      "loss": 0.4158,
+      "step": 1309
+    },
+    {
+      "epoch": 0.9646539027982327,
+      "grad_norm": 0.45096832513809204,
+      "learning_rate": 8.579191777437721e-06,
+      "loss": 0.4532,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9653902798232695,
+      "grad_norm": 0.4758065938949585,
+      "learning_rate": 8.57619855491748e-06,
+      "loss": 0.4518,
+      "step": 1311
+    },
+    {
+      "epoch": 0.9661266568483063,
+      "grad_norm": 0.4714369773864746,
+      "learning_rate": 8.57320270614509e-06,
+      "loss": 0.4579,
+      "step": 1312
+    },
+    {
+      "epoch": 0.9668630338733432,
+      "grad_norm": 0.44195324182510376,
+      "learning_rate": 8.57020423332061e-06,
+      "loss": 0.4256,
+      "step": 1313
+    },
+    {
+      "epoch": 0.96759941089838,
+      "grad_norm": 0.49739697575569153,
+      "learning_rate": 8.567203138646027e-06,
+      "loss": 0.4411,
+      "step": 1314
+    },
+    {
+      "epoch": 0.9683357879234168,
+      "grad_norm": 0.48561814427375793,
+      "learning_rate": 8.564199424325259e-06,
+      "loss": 0.439,
+      "step": 1315
+    },
+    {
+      "epoch": 0.9690721649484536,
+      "grad_norm": 0.44757330417633057,
+      "learning_rate": 8.56119309256414e-06,
+      "loss": 0.468,
+      "step": 1316
+    },
+    {
+      "epoch": 0.9698085419734904,
+      "grad_norm": 0.5125359296798706,
+      "learning_rate": 8.558184145570427e-06,
+      "loss": 0.4643,
+      "step": 1317
+    },
+    {
+      "epoch": 0.9705449189985272,
+      "grad_norm": 0.6268301606178284,
+      "learning_rate": 8.555172585553804e-06,
+      "loss": 0.4455,
+      "step": 1318
+    },
+    {
+      "epoch": 0.9712812960235641,
+      "grad_norm": 0.4266813099384308,
+      "learning_rate": 8.552158414725868e-06,
+      "loss": 0.4357,
+      "step": 1319
+    },
+    {
+      "epoch": 0.9720176730486009,
+      "grad_norm": 0.49251487851142883,
+      "learning_rate": 8.549141635300135e-06,
+      "loss": 0.4315,
+      "step": 1320
+    },
+    {
+      "epoch": 0.9727540500736377,
+      "grad_norm": 0.4862709939479828,
+      "learning_rate": 8.546122249492035e-06,
+      "loss": 0.4352,
+      "step": 1321
+    },
+    {
+      "epoch": 0.9734904270986745,
+      "grad_norm": 0.4898791015148163,
+      "learning_rate": 8.543100259518916e-06,
+      "loss": 0.4408,
+      "step": 1322
+    },
+    {
+      "epoch": 0.9742268041237113,
+      "grad_norm": 0.45198628306388855,
+      "learning_rate": 8.540075667600034e-06,
+      "loss": 0.4316,
+      "step": 1323
+    },
+    {
+      "epoch": 0.9749631811487481,
+      "grad_norm": 0.5984836220741272,
+      "learning_rate": 8.53704847595656e-06,
+      "loss": 0.4535,
+      "step": 1324
+    },
+    {
+      "epoch": 0.975699558173785,
+      "grad_norm": 0.5049937963485718,
+      "learning_rate": 8.534018686811572e-06,
+      "loss": 0.4115,
+      "step": 1325
+    },
+    {
+      "epoch": 0.9764359351988218,
+      "grad_norm": 0.41473087668418884,
+      "learning_rate": 8.530986302390053e-06,
+      "loss": 0.463,
+      "step": 1326
+    },
+    {
+      "epoch": 0.9771723122238586,
+      "grad_norm": 0.4933287501335144,
+      "learning_rate": 8.527951324918897e-06,
+      "loss": 0.4182,
+      "step": 1327
+    },
+    {
+      "epoch": 0.9779086892488954,
+      "grad_norm": 0.6397290229797363,
+      "learning_rate": 8.5249137566269e-06,
+      "loss": 0.4784,
+      "step": 1328
+    },
+    {
+      "epoch": 0.9786450662739322,
+      "grad_norm": 0.45069581270217896,
+      "learning_rate": 8.521873599744758e-06,
+      "loss": 0.4373,
+      "step": 1329
+    },
+    {
+      "epoch": 0.979381443298969,
+      "grad_norm": 0.5195262432098389,
+      "learning_rate": 8.518830856505072e-06,
+      "loss": 0.4582,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9801178203240059,
+      "grad_norm": 0.6199840307235718,
+      "learning_rate": 8.515785529142339e-06,
+      "loss": 0.4265,
+      "step": 1331
+    },
+    {
+      "epoch": 0.9808541973490427,
+      "grad_norm": 0.5363901257514954,
+      "learning_rate": 8.512737619892958e-06,
+      "loss": 0.4388,
+      "step": 1332
+    },
+    {
+      "epoch": 0.9815905743740795,
+      "grad_norm": 0.576992392539978,
+      "learning_rate": 8.509687130995223e-06,
+      "loss": 0.4475,
+      "step": 1333
+    },
+    {
+      "epoch": 0.9823269513991163,
+      "grad_norm": 0.4745303690433502,
+      "learning_rate": 8.506634064689314e-06,
+      "loss": 0.432,
+      "step": 1334
+    },
+    {
+      "epoch": 0.9830633284241531,
+      "grad_norm": 0.4878941476345062,
+      "learning_rate": 8.503578423217316e-06,
+      "loss": 0.433,
+      "step": 1335
+    },
+    {
+      "epoch": 0.9837997054491899,
+      "grad_norm": 0.5367373824119568,
+      "learning_rate": 8.500520208823199e-06,
+      "loss": 0.4369,
+      "step": 1336
+    },
+    {
+      "epoch": 0.9845360824742269,
+      "grad_norm": 0.5170448422431946,
+      "learning_rate": 8.497459423752824e-06,
+      "loss": 0.4418,
+      "step": 1337
+    },
+    {
+      "epoch": 0.9852724594992637,
+      "grad_norm": 0.45060834288597107,
+      "learning_rate": 8.494396070253934e-06,
+      "loss": 0.4293,
+      "step": 1338
+    },
+    {
+      "epoch": 0.9860088365243005,
+      "grad_norm": 0.43550148606300354,
+      "learning_rate": 8.49133015057617e-06,
+      "loss": 0.437,
+      "step": 1339
+    },
+    {
+      "epoch": 0.9867452135493373,
+      "grad_norm": 0.4647277891635895,
+      "learning_rate": 8.488261666971047e-06,
+      "loss": 0.4149,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9874815905743741,
+      "grad_norm": 0.5019136667251587,
+      "learning_rate": 8.485190621691967e-06,
+      "loss": 0.4437,
+      "step": 1341
+    },
+    {
+      "epoch": 0.9882179675994109,
+      "grad_norm": 0.41769254207611084,
+      "learning_rate": 8.482117016994213e-06,
+      "loss": 0.4457,
+      "step": 1342
+    },
+    {
+      "epoch": 0.9889543446244478,
+      "grad_norm": 0.4599890112876892,
+      "learning_rate": 8.479040855134949e-06,
+      "loss": 0.4314,
+      "step": 1343
+    },
+    {
+      "epoch": 0.9896907216494846,
+      "grad_norm": 0.495074063539505,
+      "learning_rate": 8.475962138373212e-06,
+      "loss": 0.4334,
+      "step": 1344
+    },
+    {
+      "epoch": 0.9904270986745214,
+      "grad_norm": 0.41426074504852295,
+      "learning_rate": 8.472880868969922e-06,
+      "loss": 0.4461,
+      "step": 1345
+    },
+    {
+      "epoch": 0.9911634756995582,
+      "grad_norm": 0.4830911457538605,
+      "learning_rate": 8.469797049187867e-06,
+      "loss": 0.4623,
+      "step": 1346
+    },
+    {
+      "epoch": 0.991899852724595,
+      "grad_norm": 0.5529137849807739,
+      "learning_rate": 8.466710681291714e-06,
+      "loss": 0.4372,
+      "step": 1347
+    },
+    {
+      "epoch": 0.9926362297496318,
+      "grad_norm": 0.46961885690689087,
+      "learning_rate": 8.463621767547998e-06,
+      "loss": 0.4219,
+      "step": 1348
+    },
+    {
+      "epoch": 0.9933726067746687,
+      "grad_norm": 0.4638763666152954,
+      "learning_rate": 8.46053031022512e-06,
+      "loss": 0.4388,
+      "step": 1349
+    },
+    {
+      "epoch": 0.9941089837997055,
+      "grad_norm": 0.44377005100250244,
+      "learning_rate": 8.457436311593358e-06,
+      "loss": 0.4169,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9948453608247423,
+      "grad_norm": 0.5292149782180786,
+      "learning_rate": 8.454339773924849e-06,
+      "loss": 0.4521,
+      "step": 1351
+    },
+    {
+      "epoch": 0.9955817378497791,
+      "grad_norm": 0.5035756230354309,
+      "learning_rate": 8.451240699493597e-06,
+      "loss": 0.4451,
+      "step": 1352
+    },
+    {
+      "epoch": 0.9963181148748159,
+      "grad_norm": 0.42530128359794617,
+      "learning_rate": 8.448139090575467e-06,
+      "loss": 0.434,
+      "step": 1353
+    },
+    {
+      "epoch": 0.9970544918998527,
+      "grad_norm": 0.5264377593994141,
+      "learning_rate": 8.445034949448188e-06,
+      "loss": 0.4654,
+      "step": 1354
+    },
+    {
+      "epoch": 0.9977908689248896,
+      "grad_norm": 0.5930060148239136,
+      "learning_rate": 8.441928278391349e-06,
+      "loss": 0.4539,
+      "step": 1355
+    },
+    {
+      "epoch": 0.9985272459499264,
+      "grad_norm": 0.43737247586250305,
+      "learning_rate": 8.438819079686391e-06,
+      "loss": 0.4495,
+      "step": 1356
+    },
+    {
+      "epoch": 0.9992636229749632,
+      "grad_norm": 0.46241891384124756,
+      "learning_rate": 8.43570735561662e-06,
+      "loss": 0.4729,
+      "step": 1357
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5586955547332764,
+      "learning_rate": 8.43259310846719e-06,
+      "loss": 0.4411,
+      "step": 1358
+    },
+    {
+      "epoch": 1.0007363770250368,
+      "grad_norm": 0.5139601230621338,
+      "learning_rate": 8.429476340525111e-06,
+      "loss": 0.4186,
+      "step": 1359
+    },
+    {
+      "epoch": 1.0014727540500736,
+      "grad_norm": 0.42976438999176025,
+      "learning_rate": 8.426357054079244e-06,
+      "loss": 0.4069,
+      "step": 1360
+    },
+    {
+      "epoch": 1.0022091310751104,
+      "grad_norm": 0.537762463092804,
+      "learning_rate": 8.423235251420297e-06,
+      "loss": 0.3928,
+      "step": 1361
+    },
+    {
+      "epoch": 1.0029455081001473,
+      "grad_norm": 0.4901559054851532,
+      "learning_rate": 8.420110934840826e-06,
+      "loss": 0.4392,
+      "step": 1362
+    },
+    {
+      "epoch": 1.003681885125184,
+      "grad_norm": 0.44034528732299805,
+      "learning_rate": 8.416984106635238e-06,
+      "loss": 0.4228,
+      "step": 1363
+    },
+    {
+      "epoch": 1.004418262150221,
+      "grad_norm": 0.4630548059940338,
+      "learning_rate": 8.413854769099779e-06,
+      "loss": 0.4053,
+      "step": 1364
+    },
+    {
+      "epoch": 1.0051546391752577,
+      "grad_norm": 0.5146474242210388,
+      "learning_rate": 8.410722924532541e-06,
+      "loss": 0.3992,
+      "step": 1365
+    },
+    {
+      "epoch": 1.0058910162002945,
+      "grad_norm": 0.46642136573791504,
+      "learning_rate": 8.407588575233457e-06,
+      "loss": 0.4437,
+      "step": 1366
+    },
+    {
+      "epoch": 1.0066273932253313,
+      "grad_norm": 0.4860120415687561,
+      "learning_rate": 8.404451723504295e-06,
+      "loss": 0.421,
+      "step": 1367
+    },
+    {
+      "epoch": 1.0073637702503682,
+      "grad_norm": 0.5072521567344666,
+      "learning_rate": 8.401312371648667e-06,
+      "loss": 0.4082,
+      "step": 1368
+    },
+    {
+      "epoch": 1.008100147275405,
+      "grad_norm": 0.4237157702445984,
+      "learning_rate": 8.398170521972017e-06,
+      "loss": 0.4189,
+      "step": 1369
+    },
+    {
+      "epoch": 1.0088365243004418,
+      "grad_norm": 0.4586620628833771,
+      "learning_rate": 8.395026176781627e-06,
+      "loss": 0.3847,
+      "step": 1370
+    },
+    {
+      "epoch": 1.0095729013254786,
+      "grad_norm": 0.4629966616630554,
+      "learning_rate": 8.391879338386604e-06,
+      "loss": 0.4173,
+      "step": 1371
+    },
+    {
+      "epoch": 1.0103092783505154,
+      "grad_norm": 0.4874536097049713,
+      "learning_rate": 8.388730009097895e-06,
+      "loss": 0.4035,
+      "step": 1372
+    },
+    {
+      "epoch": 1.0110456553755522,
+      "grad_norm": 0.4574902355670929,
+      "learning_rate": 8.385578191228272e-06,
+      "loss": 0.425,
+      "step": 1373
+    },
+    {
+      "epoch": 1.011782032400589,
+      "grad_norm": 0.3942805230617523,
+      "learning_rate": 8.382423887092333e-06,
+      "loss": 0.4186,
+      "step": 1374
+    },
+    {
+      "epoch": 1.0125184094256259,
+      "grad_norm": 0.4842493236064911,
+      "learning_rate": 8.379267099006506e-06,
+      "loss": 0.4149,
+      "step": 1375
+    },
+    {
+      "epoch": 1.0132547864506627,
+      "grad_norm": 0.46591871976852417,
+      "learning_rate": 8.376107829289037e-06,
+      "loss": 0.4112,
+      "step": 1376
+    },
+    {
+      "epoch": 1.0139911634756995,
+      "grad_norm": 0.4761437177658081,
+      "learning_rate": 8.372946080260002e-06,
+      "loss": 0.4109,
+      "step": 1377
+    },
+    {
+      "epoch": 1.0147275405007363,
+      "grad_norm": 0.4643495976924896,
+      "learning_rate": 8.369781854241293e-06,
+      "loss": 0.4222,
+      "step": 1378
+    },
+    {
+      "epoch": 1.0154639175257731,
+      "grad_norm": 0.4508487284183502,
+      "learning_rate": 8.36661515355662e-06,
+      "loss": 0.4215,
+      "step": 1379
+    },
+    {
+      "epoch": 1.01620029455081,
+      "grad_norm": 0.5168802738189697,
+      "learning_rate": 8.363445980531515e-06,
+      "loss": 0.4213,
+      "step": 1380
+    },
+    {
+      "epoch": 1.0169366715758468,
+      "grad_norm": 0.46994489431381226,
+      "learning_rate": 8.360274337493321e-06,
+      "loss": 0.4292,
+      "step": 1381
+    },
+    {
+      "epoch": 1.0176730486008836,
+      "grad_norm": 0.44330844283103943,
+      "learning_rate": 8.3571002267712e-06,
+      "loss": 0.4107,
+      "step": 1382
+    },
+    {
+      "epoch": 1.0184094256259204,
+      "grad_norm": 0.4858071506023407,
+      "learning_rate": 8.353923650696119e-06,
+      "loss": 0.3989,
+      "step": 1383
+    },
+    {
+      "epoch": 1.0191458026509572,
+      "grad_norm": 0.5103859305381775,
+      "learning_rate": 8.35074461160086e-06,
+      "loss": 0.4168,
+      "step": 1384
+    },
+    {
+      "epoch": 1.019882179675994,
+      "grad_norm": 0.4278160631656647,
+      "learning_rate": 8.347563111820014e-06,
+      "loss": 0.4058,
+      "step": 1385
+    },
+    {
+      "epoch": 1.0206185567010309,
+      "grad_norm": 0.436535507440567,
+      "learning_rate": 8.34437915368998e-06,
+      "loss": 0.3877,
+      "step": 1386
+    },
+    {
+      "epoch": 1.0213549337260677,
+      "grad_norm": 0.47809407114982605,
+      "learning_rate": 8.341192739548958e-06,
+      "loss": 0.3958,
+      "step": 1387
+    },
+    {
+      "epoch": 1.0220913107511045,
+      "grad_norm": 0.4272827208042145,
+      "learning_rate": 8.338003871736957e-06,
+      "loss": 0.4139,
+      "step": 1388
+    },
+    {
+      "epoch": 1.0228276877761413,
+      "grad_norm": 0.4393133819103241,
+      "learning_rate": 8.334812552595782e-06,
+      "loss": 0.4266,
+      "step": 1389
+    },
+    {
+      "epoch": 1.0235640648011781,
+      "grad_norm": 0.45444509387016296,
+      "learning_rate": 8.331618784469043e-06,
+      "loss": 0.4055,
+      "step": 1390
+    },
+    {
+      "epoch": 1.024300441826215,
+      "grad_norm": 0.46266138553619385,
+      "learning_rate": 8.328422569702148e-06,
+      "loss": 0.4283,
+      "step": 1391
+    },
+    {
+      "epoch": 1.0250368188512518,
+      "grad_norm": 0.43913209438323975,
+      "learning_rate": 8.325223910642297e-06,
+      "loss": 0.4317,
+      "step": 1392
+    },
+    {
+      "epoch": 1.0257731958762886,
+      "grad_norm": 0.4571021795272827,
+      "learning_rate": 8.322022809638492e-06,
+      "loss": 0.4255,
+      "step": 1393
+    },
+    {
+      "epoch": 1.0265095729013254,
+      "grad_norm": 0.4817020893096924,
+      "learning_rate": 8.318819269041524e-06,
+      "loss": 0.4324,
+      "step": 1394
+    },
+    {
+      "epoch": 1.0272459499263622,
+      "grad_norm": 0.4041571021080017,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 0.417,
+      "step": 1395
+    },
+    {
+      "epoch": 1.027982326951399,
+      "grad_norm": 0.4440697729587555,
+      "learning_rate": 8.312404878480222e-06,
+      "loss": 0.4043,
+      "step": 1396
+    },
+    {
+      "epoch": 1.0287187039764358,
+      "grad_norm": 0.47406235337257385,
+      "learning_rate": 8.309194033226423e-06,
+      "loss": 0.4089,
+      "step": 1397
+    },
+    {
+      "epoch": 1.0294550810014726,
+      "grad_norm": 0.4187905192375183,
+      "learning_rate": 8.305980757800525e-06,
+      "loss": 0.4108,
+      "step": 1398
+    },
+    {
+      "epoch": 1.0301914580265095,
+      "grad_norm": 0.45452675223350525,
+      "learning_rate": 8.302765054562261e-06,
+      "loss": 0.4159,
+      "step": 1399
+    },
+    {
+      "epoch": 1.0309278350515463,
+      "grad_norm": 0.4963740110397339,
+      "learning_rate": 8.299546925873148e-06,
+      "loss": 0.411,
+      "step": 1400
+    },
+    {
+      "epoch": 1.031664212076583,
+      "grad_norm": 0.48701077699661255,
+      "learning_rate": 8.296326374096482e-06,
+      "loss": 0.409,
+      "step": 1401
+    },
+    {
+      "epoch": 1.0324005891016201,
+      "grad_norm": 0.4532162547111511,
+      "learning_rate": 8.293103401597338e-06,
+      "loss": 0.3964,
+      "step": 1402
+    },
+    {
+      "epoch": 1.033136966126657,
+      "grad_norm": 0.4221610128879547,
+      "learning_rate": 8.28987801074257e-06,
+      "loss": 0.4084,
+      "step": 1403
+    },
+    {
+      "epoch": 1.0338733431516938,
+      "grad_norm": 0.5044291615486145,
+      "learning_rate": 8.286650203900808e-06,
+      "loss": 0.4088,
+      "step": 1404
+    },
+    {
+      "epoch": 1.0346097201767306,
+      "grad_norm": 0.4653428792953491,
+      "learning_rate": 8.283419983442453e-06,
+      "loss": 0.3964,
+      "step": 1405
+    },
+    {
+      "epoch": 1.0353460972017674,
+      "grad_norm": 0.45748192071914673,
+      "learning_rate": 8.280187351739686e-06,
+      "loss": 0.4184,
+      "step": 1406
+    },
+    {
+      "epoch": 1.0360824742268042,
+      "grad_norm": 0.4691620171070099,
+      "learning_rate": 8.276952311166451e-06,
+      "loss": 0.3849,
+      "step": 1407
+    },
+    {
+      "epoch": 1.036818851251841,
+      "grad_norm": 0.46832898259162903,
+      "learning_rate": 8.273714864098466e-06,
+      "loss": 0.4052,
+      "step": 1408
+    },
+    {
+      "epoch": 1.0375552282768779,
+      "grad_norm": 0.3961854875087738,
+      "learning_rate": 8.270475012913212e-06,
+      "loss": 0.3872,
+      "step": 1409
+    },
+    {
+      "epoch": 1.0382916053019147,
+      "grad_norm": 0.4253145158290863,
+      "learning_rate": 8.267232759989938e-06,
+      "loss": 0.3978,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0390279823269515,
+      "grad_norm": 0.43363869190216064,
+      "learning_rate": 8.26398810770966e-06,
+      "loss": 0.4387,
+      "step": 1411
+    },
+    {
+      "epoch": 1.0397643593519883,
+      "grad_norm": 0.43698909878730774,
+      "learning_rate": 8.260741058455147e-06,
+      "loss": 0.4094,
+      "step": 1412
+    },
+    {
+      "epoch": 1.0405007363770251,
+      "grad_norm": 0.3974965214729309,
+      "learning_rate": 8.257491614610939e-06,
+      "loss": 0.407,
+      "step": 1413
+    },
+    {
+      "epoch": 1.041237113402062,
+      "grad_norm": 0.4418043792247772,
+      "learning_rate": 8.254239778563325e-06,
+      "loss": 0.4174,
+      "step": 1414
+    },
+    {
+      "epoch": 1.0419734904270987,
+      "grad_norm": 0.43214893341064453,
+      "learning_rate": 8.250985552700359e-06,
+      "loss": 0.4399,
+      "step": 1415
+    },
+    {
+      "epoch": 1.0427098674521356,
+      "grad_norm": 0.41666722297668457,
+      "learning_rate": 8.247728939411845e-06,
+      "loss": 0.4377,
+      "step": 1416
+    },
+    {
+      "epoch": 1.0434462444771724,
+      "grad_norm": 0.45747169852256775,
+      "learning_rate": 8.24446994108934e-06,
+      "loss": 0.4204,
+      "step": 1417
+    },
+    {
+      "epoch": 1.0441826215022092,
+      "grad_norm": 0.4472166895866394,
+      "learning_rate": 8.241208560126154e-06,
+      "loss": 0.4299,
+      "step": 1418
+    },
+    {
+      "epoch": 1.044918998527246,
+      "grad_norm": 0.42849603295326233,
+      "learning_rate": 8.237944798917347e-06,
+      "loss": 0.419,
+      "step": 1419
+    },
+    {
+      "epoch": 1.0456553755522828,
+      "grad_norm": 0.46297961473464966,
+      "learning_rate": 8.234678659859729e-06,
+      "loss": 0.4288,
+      "step": 1420
+    },
+    {
+      "epoch": 1.0463917525773196,
+      "grad_norm": 0.5132995247840881,
+      "learning_rate": 8.231410145351853e-06,
+      "loss": 0.4088,
+      "step": 1421
+    },
+    {
+      "epoch": 1.0471281296023565,
+      "grad_norm": 0.45529741048812866,
+      "learning_rate": 8.228139257794012e-06,
+      "loss": 0.4097,
+      "step": 1422
+    },
+    {
+      "epoch": 1.0478645066273933,
+      "grad_norm": 0.4541131854057312,
+      "learning_rate": 8.224865999588254e-06,
+      "loss": 0.3927,
+      "step": 1423
+    },
+    {
+      "epoch": 1.04860088365243,
+      "grad_norm": 0.46962690353393555,
+      "learning_rate": 8.221590373138358e-06,
+      "loss": 0.4196,
+      "step": 1424
+    },
+    {
+      "epoch": 1.049337260677467,
+      "grad_norm": 0.39268723130226135,
+      "learning_rate": 8.218312380849844e-06,
+      "loss": 0.4267,
+      "step": 1425
+    },
+    {
+      "epoch": 1.0500736377025037,
+      "grad_norm": 0.38890156149864197,
+      "learning_rate": 8.21503202512997e-06,
+      "loss": 0.4095,
+      "step": 1426
+    },
+    {
+      "epoch": 1.0508100147275405,
+      "grad_norm": 0.4562273621559143,
+      "learning_rate": 8.211749308387734e-06,
+      "loss": 0.4437,
+      "step": 1427
+    },
+    {
+      "epoch": 1.0515463917525774,
+      "grad_norm": 0.43432724475860596,
+      "learning_rate": 8.208464233033862e-06,
+      "loss": 0.4181,
+      "step": 1428
+    },
+    {
+      "epoch": 1.0522827687776142,
+      "grad_norm": 0.4545213580131531,
+      "learning_rate": 8.205176801480811e-06,
+      "loss": 0.4148,
+      "step": 1429
+    },
+    {
+      "epoch": 1.053019145802651,
+      "grad_norm": 0.5053450465202332,
+      "learning_rate": 8.201887016142776e-06,
+      "loss": 0.4027,
+      "step": 1430
+    },
+    {
+      "epoch": 1.0537555228276878,
+      "grad_norm": 0.4846493601799011,
+      "learning_rate": 8.198594879435673e-06,
+      "loss": 0.4324,
+      "step": 1431
+    },
+    {
+      "epoch": 1.0544918998527246,
+      "grad_norm": 0.4266805648803711,
+      "learning_rate": 8.19530039377715e-06,
+      "loss": 0.4072,
+      "step": 1432
+    },
+    {
+      "epoch": 1.0552282768777614,
+      "grad_norm": 0.5038962364196777,
+      "learning_rate": 8.192003561586576e-06,
+      "loss": 0.4021,
+      "step": 1433
+    },
+    {
+      "epoch": 1.0559646539027983,
+      "grad_norm": 0.5069649815559387,
+      "learning_rate": 8.188704385285046e-06,
+      "loss": 0.4373,
+      "step": 1434
+    },
+    {
+      "epoch": 1.056701030927835,
+      "grad_norm": 0.40627750754356384,
+      "learning_rate": 8.185402867295373e-06,
+      "loss": 0.4307,
+      "step": 1435
+    },
+    {
+      "epoch": 1.0574374079528719,
+      "grad_norm": 0.5366407632827759,
+      "learning_rate": 8.182099010042095e-06,
+      "loss": 0.4244,
+      "step": 1436
+    },
+    {
+      "epoch": 1.0581737849779087,
+      "grad_norm": 0.4338124096393585,
+      "learning_rate": 8.178792815951465e-06,
+      "loss": 0.4091,
+      "step": 1437
+    },
+    {
+      "epoch": 1.0589101620029455,
+      "grad_norm": 0.4336191415786743,
+      "learning_rate": 8.175484287451448e-06,
+      "loss": 0.4111,
+      "step": 1438
+    },
+    {
+      "epoch": 1.0596465390279823,
+      "grad_norm": 0.4909955561161041,
+      "learning_rate": 8.172173426971732e-06,
+      "loss": 0.4237,
+      "step": 1439
+    },
+    {
+      "epoch": 1.0603829160530192,
+      "grad_norm": 0.6215251088142395,
+      "learning_rate": 8.168860236943709e-06,
+      "loss": 0.4154,
+      "step": 1440
+    },
+    {
+      "epoch": 1.061119293078056,
+      "grad_norm": 0.4355964958667755,
+      "learning_rate": 8.16554471980049e-06,
+      "loss": 0.3951,
+      "step": 1441
+    },
+    {
+      "epoch": 1.0618556701030928,
+      "grad_norm": 0.5015247464179993,
+      "learning_rate": 8.162226877976886e-06,
+      "loss": 0.4373,
+      "step": 1442
+    },
+    {
+      "epoch": 1.0625920471281296,
+      "grad_norm": 0.49945569038391113,
+      "learning_rate": 8.158906713909425e-06,
+      "loss": 0.3987,
+      "step": 1443
+    },
+    {
+      "epoch": 1.0633284241531664,
+      "grad_norm": 0.492780864238739,
+      "learning_rate": 8.155584230036328e-06,
+      "loss": 0.424,
+      "step": 1444
+    },
+    {
+      "epoch": 1.0640648011782032,
+      "grad_norm": 0.5154081583023071,
+      "learning_rate": 8.152259428797535e-06,
+      "loss": 0.4135,
+      "step": 1445
+    },
+    {
+      "epoch": 1.06480117820324,
+      "grad_norm": 0.4992247521877289,
+      "learning_rate": 8.148932312634674e-06,
+      "loss": 0.4126,
+      "step": 1446
+    },
+    {
+      "epoch": 1.0655375552282769,
+      "grad_norm": 0.579197883605957,
+      "learning_rate": 8.14560288399108e-06,
+      "loss": 0.4069,
+      "step": 1447
+    },
+    {
+      "epoch": 1.0662739322533137,
+      "grad_norm": 0.5475592613220215,
+      "learning_rate": 8.142271145311784e-06,
+      "loss": 0.4106,
+      "step": 1448
+    },
+    {
+      "epoch": 1.0670103092783505,
+      "grad_norm": 0.5077992081642151,
+      "learning_rate": 8.138937099043516e-06,
+      "loss": 0.4135,
+      "step": 1449
+    },
+    {
+      "epoch": 1.0677466863033873,
+      "grad_norm": 0.5219954252243042,
+      "learning_rate": 8.135600747634697e-06,
+      "loss": 0.4654,
+      "step": 1450
+    },
+    {
+      "epoch": 1.0684830633284241,
+      "grad_norm": 0.5722024440765381,
+      "learning_rate": 8.132262093535444e-06,
+      "loss": 0.4448,
+      "step": 1451
+    },
+    {
+      "epoch": 1.069219440353461,
+      "grad_norm": 0.4120500683784485,
+      "learning_rate": 8.128921139197563e-06,
+      "loss": 0.3952,
+      "step": 1452
+    },
+    {
+      "epoch": 1.0699558173784978,
+      "grad_norm": 0.46952569484710693,
+      "learning_rate": 8.125577887074552e-06,
+      "loss": 0.4353,
+      "step": 1453
+    },
+    {
+      "epoch": 1.0706921944035346,
+      "grad_norm": 0.5226830244064331,
+      "learning_rate": 8.12223233962159e-06,
+      "loss": 0.409,
+      "step": 1454
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 0.5324341058731079,
+      "learning_rate": 8.118884499295549e-06,
+      "loss": 0.4086,
+      "step": 1455
+    },
+    {
+      "epoch": 1.0721649484536082,
+      "grad_norm": 0.4469471573829651,
+      "learning_rate": 8.115534368554981e-06,
+      "loss": 0.4064,
+      "step": 1456
+    },
+    {
+      "epoch": 1.072901325478645,
+      "grad_norm": 0.5327169895172119,
+      "learning_rate": 8.112181949860121e-06,
+      "loss": 0.4437,
+      "step": 1457
+    },
+    {
+      "epoch": 1.0736377025036818,
+      "grad_norm": 0.5192805528640747,
+      "learning_rate": 8.108827245672884e-06,
+      "loss": 0.4017,
+      "step": 1458
+    },
+    {
+      "epoch": 1.0743740795287187,
+      "grad_norm": 0.3944444954395294,
+      "learning_rate": 8.105470258456863e-06,
+      "loss": 0.3981,
+      "step": 1459
+    },
+    {
+      "epoch": 1.0751104565537555,
+      "grad_norm": 0.4764431416988373,
+      "learning_rate": 8.102110990677328e-06,
+      "loss": 0.4081,
+      "step": 1460
+    },
+    {
+      "epoch": 1.0758468335787923,
+      "grad_norm": 0.49513620138168335,
+      "learning_rate": 8.098749444801226e-06,
+      "loss": 0.4406,
+      "step": 1461
+    },
+    {
+      "epoch": 1.076583210603829,
+      "grad_norm": 0.4354887306690216,
+      "learning_rate": 8.095385623297171e-06,
+      "loss": 0.4379,
+      "step": 1462
+    },
+    {
+      "epoch": 1.077319587628866,
+      "grad_norm": 0.541980504989624,
+      "learning_rate": 8.092019528635453e-06,
+      "loss": 0.43,
+      "step": 1463
+    },
+    {
+      "epoch": 1.0780559646539027,
+      "grad_norm": 0.47262078523635864,
+      "learning_rate": 8.088651163288032e-06,
+      "loss": 0.4201,
+      "step": 1464
+    },
+    {
+      "epoch": 1.0787923416789396,
+      "grad_norm": 0.4839096665382385,
+      "learning_rate": 8.085280529728533e-06,
+      "loss": 0.4137,
+      "step": 1465
+    },
+    {
+      "epoch": 1.0795287187039764,
+      "grad_norm": 0.4608899652957916,
+      "learning_rate": 8.081907630432246e-06,
+      "loss": 0.4211,
+      "step": 1466
+    },
+    {
+      "epoch": 1.0802650957290132,
+      "grad_norm": 0.4837184250354767,
+      "learning_rate": 8.078532467876126e-06,
+      "loss": 0.4036,
+      "step": 1467
+    },
+    {
+      "epoch": 1.08100147275405,
+      "grad_norm": 0.518398642539978,
+      "learning_rate": 8.075155044538792e-06,
+      "loss": 0.4052,
+      "step": 1468
+    },
+    {
+      "epoch": 1.0817378497790868,
+      "grad_norm": 0.5079612731933594,
+      "learning_rate": 8.071775362900522e-06,
+      "loss": 0.4255,
+      "step": 1469
+    },
+    {
+      "epoch": 1.0824742268041236,
+      "grad_norm": 0.4972943365573883,
+      "learning_rate": 8.068393425443253e-06,
+      "loss": 0.3831,
+      "step": 1470
+    },
+    {
+      "epoch": 1.0832106038291605,
+      "grad_norm": 0.4453538656234741,
+      "learning_rate": 8.065009234650574e-06,
+      "loss": 0.4174,
+      "step": 1471
+    },
+    {
+      "epoch": 1.0839469808541973,
+      "grad_norm": 0.46142661571502686,
+      "learning_rate": 8.061622793007735e-06,
+      "loss": 0.4331,
+      "step": 1472
+    },
+    {
+      "epoch": 1.084683357879234,
+      "grad_norm": 0.4330517053604126,
+      "learning_rate": 8.058234103001634e-06,
+      "loss": 0.4285,
+      "step": 1473
+    },
+    {
+      "epoch": 1.085419734904271,
+      "grad_norm": 0.4767371118068695,
+      "learning_rate": 8.054843167120827e-06,
+      "loss": 0.4198,
+      "step": 1474
+    },
+    {
+      "epoch": 1.0861561119293077,
+      "grad_norm": 0.45682060718536377,
+      "learning_rate": 8.051449987855512e-06,
+      "loss": 0.4344,
+      "step": 1475
+    },
+    {
+      "epoch": 1.0868924889543445,
+      "grad_norm": 0.44379401206970215,
+      "learning_rate": 8.048054567697537e-06,
+      "loss": 0.4296,
+      "step": 1476
+    },
+    {
+      "epoch": 1.0876288659793814,
+      "grad_norm": 0.4649368226528168,
+      "learning_rate": 8.044656909140395e-06,
+      "loss": 0.4213,
+      "step": 1477
+    },
+    {
+      "epoch": 1.0883652430044182,
+      "grad_norm": 0.4579814672470093,
+      "learning_rate": 8.041257014679228e-06,
+      "loss": 0.4117,
+      "step": 1478
+    },
+    {
+      "epoch": 1.089101620029455,
+      "grad_norm": 0.4541153013706207,
+      "learning_rate": 8.037854886810813e-06,
+      "loss": 0.4197,
+      "step": 1479
+    },
+    {
+      "epoch": 1.0898379970544918,
+      "grad_norm": 0.39233219623565674,
+      "learning_rate": 8.034450528033565e-06,
+      "loss": 0.3982,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0905743740795286,
+      "grad_norm": 0.40796148777008057,
+      "learning_rate": 8.031043940847551e-06,
+      "loss": 0.4377,
+      "step": 1481
+    },
+    {
+      "epoch": 1.0913107511045654,
+      "grad_norm": 0.47203153371810913,
+      "learning_rate": 8.027635127754462e-06,
+      "loss": 0.4251,
+      "step": 1482
+    },
+    {
+      "epoch": 1.0920471281296025,
+      "grad_norm": 0.4739782512187958,
+      "learning_rate": 8.024224091257628e-06,
+      "loss": 0.4257,
+      "step": 1483
+    },
+    {
+      "epoch": 1.0927835051546393,
+      "grad_norm": 0.3920868933200836,
+      "learning_rate": 8.020810833862009e-06,
+      "loss": 0.406,
+      "step": 1484
+    },
+    {
+      "epoch": 1.093519882179676,
+      "grad_norm": 0.5111119151115417,
+      "learning_rate": 8.017395358074198e-06,
+      "loss": 0.4216,
+      "step": 1485
+    },
+    {
+      "epoch": 1.094256259204713,
+      "grad_norm": 0.46156689524650574,
+      "learning_rate": 8.013977666402421e-06,
+      "loss": 0.4185,
+      "step": 1486
+    },
+    {
+      "epoch": 1.0949926362297497,
+      "grad_norm": 0.42204272747039795,
+      "learning_rate": 8.010557761356523e-06,
+      "loss": 0.417,
+      "step": 1487
+    },
+    {
+      "epoch": 1.0957290132547866,
+      "grad_norm": 0.5379983186721802,
+      "learning_rate": 8.007135645447982e-06,
+      "loss": 0.4235,
+      "step": 1488
+    },
+    {
+      "epoch": 1.0964653902798234,
+      "grad_norm": 0.4320377707481384,
+      "learning_rate": 8.003711321189895e-06,
+      "loss": 0.4081,
+      "step": 1489
+    },
+    {
+      "epoch": 1.0972017673048602,
+      "grad_norm": 0.4671606421470642,
+      "learning_rate": 8.000284791096983e-06,
+      "loss": 0.4235,
+      "step": 1490
+    },
+    {
+      "epoch": 1.097938144329897,
+      "grad_norm": 0.47538307309150696,
+      "learning_rate": 7.996856057685587e-06,
+      "loss": 0.45,
+      "step": 1491
+    },
+    {
+      "epoch": 1.0986745213549338,
+      "grad_norm": 0.45516252517700195,
+      "learning_rate": 7.993425123473662e-06,
+      "loss": 0.4002,
+      "step": 1492
+    },
+    {
+      "epoch": 1.0994108983799706,
+      "grad_norm": 0.4619099199771881,
+      "learning_rate": 7.989991990980786e-06,
+      "loss": 0.4214,
+      "step": 1493
+    },
+    {
+      "epoch": 1.1001472754050075,
+      "grad_norm": 0.36979731917381287,
+      "learning_rate": 7.986556662728145e-06,
+      "loss": 0.4225,
+      "step": 1494
+    },
+    {
+      "epoch": 1.1008836524300443,
+      "grad_norm": 0.43840059638023376,
+      "learning_rate": 7.983119141238543e-06,
+      "loss": 0.4197,
+      "step": 1495
+    },
+    {
+      "epoch": 1.101620029455081,
+      "grad_norm": 0.4606439769268036,
+      "learning_rate": 7.97967942903639e-06,
+      "loss": 0.4014,
+      "step": 1496
+    },
+    {
+      "epoch": 1.102356406480118,
+      "grad_norm": 0.38191360235214233,
+      "learning_rate": 7.976237528647705e-06,
+      "loss": 0.4149,
+      "step": 1497
+    },
+    {
+      "epoch": 1.1030927835051547,
+      "grad_norm": 0.4278278648853302,
+      "learning_rate": 7.97279344260012e-06,
+      "loss": 0.4283,
+      "step": 1498
+    },
+    {
+      "epoch": 1.1038291605301915,
+      "grad_norm": 0.4436081051826477,
+      "learning_rate": 7.969347173422866e-06,
+      "loss": 0.4274,
+      "step": 1499
+    },
+    {
+      "epoch": 1.1045655375552283,
+      "grad_norm": 0.40234655141830444,
+      "learning_rate": 7.965898723646777e-06,
+      "loss": 0.4044,
+      "step": 1500
+    },
+    {
+      "epoch": 1.1053019145802652,
+      "grad_norm": 0.38506656885147095,
+      "learning_rate": 7.962448095804292e-06,
+      "loss": 0.4066,
+      "step": 1501
+    },
+    {
+      "epoch": 1.106038291605302,
+      "grad_norm": 0.4169124364852905,
+      "learning_rate": 7.958995292429447e-06,
+      "loss": 0.423,
+      "step": 1502
+    },
+    {
+      "epoch": 1.1067746686303388,
+      "grad_norm": 0.4359695613384247,
+      "learning_rate": 7.955540316057877e-06,
+      "loss": 0.4202,
+      "step": 1503
+    },
+    {
+      "epoch": 1.1075110456553756,
+      "grad_norm": 0.4167039394378662,
+      "learning_rate": 7.952083169226813e-06,
+      "loss": 0.3861,
+      "step": 1504
+    },
+    {
+      "epoch": 1.1082474226804124,
+      "grad_norm": 0.43994608521461487,
+      "learning_rate": 7.948623854475079e-06,
+      "loss": 0.4144,
+      "step": 1505
+    },
+    {
+      "epoch": 1.1089837997054492,
+      "grad_norm": 0.4638996422290802,
+      "learning_rate": 7.94516237434309e-06,
+      "loss": 0.4085,
+      "step": 1506
+    },
+    {
+      "epoch": 1.109720176730486,
+      "grad_norm": 0.4726799726486206,
+      "learning_rate": 7.941698731372851e-06,
+      "loss": 0.3952,
+      "step": 1507
+    },
+    {
+      "epoch": 1.1104565537555229,
+      "grad_norm": 0.5001874566078186,
+      "learning_rate": 7.938232928107963e-06,
+      "loss": 0.4176,
+      "step": 1508
+    },
+    {
+      "epoch": 1.1111929307805597,
+      "grad_norm": 0.39368122816085815,
+      "learning_rate": 7.9347649670936e-06,
+      "loss": 0.4093,
+      "step": 1509
+    },
+    {
+      "epoch": 1.1119293078055965,
+      "grad_norm": 0.4525279998779297,
+      "learning_rate": 7.93129485087653e-06,
+      "loss": 0.4261,
+      "step": 1510
+    },
+    {
+      "epoch": 1.1126656848306333,
+      "grad_norm": 0.43985629081726074,
+      "learning_rate": 7.927822582005104e-06,
+      "loss": 0.4309,
+      "step": 1511
+    },
+    {
+      "epoch": 1.1134020618556701,
+      "grad_norm": 0.4164014160633087,
+      "learning_rate": 7.924348163029249e-06,
+      "loss": 0.4273,
+      "step": 1512
+    },
+    {
+      "epoch": 1.114138438880707,
+      "grad_norm": 0.5301308035850525,
+      "learning_rate": 7.920871596500473e-06,
+      "loss": 0.4277,
+      "step": 1513
+    },
+    {
+      "epoch": 1.1148748159057438,
+      "grad_norm": 0.4031001925468445,
+      "learning_rate": 7.917392884971863e-06,
+      "loss": 0.4058,
+      "step": 1514
+    },
+    {
+      "epoch": 1.1156111929307806,
+      "grad_norm": 0.564858078956604,
+      "learning_rate": 7.913912030998079e-06,
+      "loss": 0.4363,
+      "step": 1515
+    },
+    {
+      "epoch": 1.1163475699558174,
+      "grad_norm": 0.45828789472579956,
+      "learning_rate": 7.910429037135355e-06,
+      "loss": 0.4088,
+      "step": 1516
+    },
+    {
+      "epoch": 1.1170839469808542,
+      "grad_norm": 0.5494905114173889,
+      "learning_rate": 7.906943905941495e-06,
+      "loss": 0.4241,
+      "step": 1517
+    },
+    {
+      "epoch": 1.117820324005891,
+      "grad_norm": 0.5447287559509277,
+      "learning_rate": 7.903456639975875e-06,
+      "loss": 0.4177,
+      "step": 1518
+    },
+    {
+      "epoch": 1.1185567010309279,
+      "grad_norm": 0.4168299734592438,
+      "learning_rate": 7.89996724179944e-06,
+      "loss": 0.4134,
+      "step": 1519
+    },
+    {
+      "epoch": 1.1192930780559647,
+      "grad_norm": 0.5956445336341858,
+      "learning_rate": 7.896475713974696e-06,
+      "loss": 0.4133,
+      "step": 1520
+    },
+    {
+      "epoch": 1.1200294550810015,
+      "grad_norm": 0.5098801255226135,
+      "learning_rate": 7.892982059065714e-06,
+      "loss": 0.4245,
+      "step": 1521
+    },
+    {
+      "epoch": 1.1207658321060383,
+      "grad_norm": 0.4896615743637085,
+      "learning_rate": 7.889486279638134e-06,
+      "loss": 0.4074,
+      "step": 1522
+    },
+    {
+      "epoch": 1.1215022091310751,
+      "grad_norm": 0.5458217263221741,
+      "learning_rate": 7.885988378259145e-06,
+      "loss": 0.4018,
+      "step": 1523
+    },
+    {
+      "epoch": 1.122238586156112,
+      "grad_norm": 0.49367034435272217,
+      "learning_rate": 7.882488357497504e-06,
+      "loss": 0.4176,
+      "step": 1524
+    },
+    {
+      "epoch": 1.1229749631811488,
+      "grad_norm": 0.5893204212188721,
+      "learning_rate": 7.87898621992352e-06,
+      "loss": 0.4162,
+      "step": 1525
+    },
+    {
+      "epoch": 1.1237113402061856,
+      "grad_norm": 0.5065829753875732,
+      "learning_rate": 7.875481968109052e-06,
+      "loss": 0.4279,
+      "step": 1526
+    },
+    {
+      "epoch": 1.1244477172312224,
+      "grad_norm": 0.5080648064613342,
+      "learning_rate": 7.871975604627524e-06,
+      "loss": 0.3956,
+      "step": 1527
+    },
+    {
+      "epoch": 1.1251840942562592,
+      "grad_norm": 0.6169611811637878,
+      "learning_rate": 7.8684671320539e-06,
+      "loss": 0.3985,
+      "step": 1528
+    },
+    {
+      "epoch": 1.125920471281296,
+      "grad_norm": 0.4453141987323761,
+      "learning_rate": 7.864956552964695e-06,
+      "loss": 0.4053,
+      "step": 1529
+    },
+    {
+      "epoch": 1.1266568483063328,
+      "grad_norm": 0.5291115641593933,
+      "learning_rate": 7.861443869937973e-06,
+      "loss": 0.414,
+      "step": 1530
+    },
+    {
+      "epoch": 1.1273932253313697,
+      "grad_norm": 0.5258786678314209,
+      "learning_rate": 7.857929085553344e-06,
+      "loss": 0.4303,
+      "step": 1531
+    },
+    {
+      "epoch": 1.1281296023564065,
+      "grad_norm": 0.4741217792034149,
+      "learning_rate": 7.854412202391958e-06,
+      "loss": 0.4174,
+      "step": 1532
+    },
+    {
+      "epoch": 1.1288659793814433,
+      "grad_norm": 0.5719811320304871,
+      "learning_rate": 7.850893223036508e-06,
+      "loss": 0.4043,
+      "step": 1533
+    },
+    {
+      "epoch": 1.12960235640648,
+      "grad_norm": 0.46365654468536377,
+      "learning_rate": 7.847372150071227e-06,
+      "loss": 0.4338,
+      "step": 1534
+    },
+    {
+      "epoch": 1.130338733431517,
+      "grad_norm": 0.4973761737346649,
+      "learning_rate": 7.843848986081882e-06,
+      "loss": 0.3943,
+      "step": 1535
+    },
+    {
+      "epoch": 1.1310751104565537,
+      "grad_norm": 0.5263963937759399,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.3909,
+      "step": 1536
+    },
+    {
+      "epoch": 1.1318114874815906,
+      "grad_norm": 0.452614426612854,
+      "learning_rate": 7.836796395381761e-06,
+      "loss": 0.415,
+      "step": 1537
+    },
+    {
+      "epoch": 1.1325478645066274,
+      "grad_norm": 0.49693626165390015,
+      "learning_rate": 7.833266973850192e-06,
+      "loss": 0.4189,
+      "step": 1538
+    },
+    {
+      "epoch": 1.1332842415316642,
+      "grad_norm": 0.5099273920059204,
+      "learning_rate": 7.829735471652978e-06,
+      "loss": 0.4125,
+      "step": 1539
+    },
+    {
+      "epoch": 1.134020618556701,
+      "grad_norm": 0.40911126136779785,
+      "learning_rate": 7.826201891383542e-06,
+      "loss": 0.4016,
+      "step": 1540
+    },
+    {
+      "epoch": 1.1347569955817378,
+      "grad_norm": 0.525810182094574,
+      "learning_rate": 7.822666235636844e-06,
+      "loss": 0.4202,
+      "step": 1541
+    },
+    {
+      "epoch": 1.1354933726067746,
+      "grad_norm": 0.40537041425704956,
+      "learning_rate": 7.819128507009361e-06,
+      "loss": 0.4088,
+      "step": 1542
+    },
+    {
+      "epoch": 1.1362297496318114,
+      "grad_norm": 0.42212677001953125,
+      "learning_rate": 7.815588708099094e-06,
+      "loss": 0.4222,
+      "step": 1543
+    },
+    {
+      "epoch": 1.1369661266568483,
+      "grad_norm": 0.5106800198554993,
+      "learning_rate": 7.812046841505563e-06,
+      "loss": 0.4592,
+      "step": 1544
+    },
+    {
+      "epoch": 1.137702503681885,
+      "grad_norm": 0.419968843460083,
+      "learning_rate": 7.808502909829807e-06,
+      "loss": 0.4089,
+      "step": 1545
+    },
+    {
+      "epoch": 1.138438880706922,
+      "grad_norm": 0.4281230568885803,
+      "learning_rate": 7.804956915674387e-06,
+      "loss": 0.3874,
+      "step": 1546
+    },
+    {
+      "epoch": 1.1391752577319587,
+      "grad_norm": 0.502241313457489,
+      "learning_rate": 7.80140886164337e-06,
+      "loss": 0.4284,
+      "step": 1547
+    },
+    {
+      "epoch": 1.1399116347569955,
+      "grad_norm": 0.39548102021217346,
+      "learning_rate": 7.79785875034234e-06,
+      "loss": 0.408,
+      "step": 1548
+    },
+    {
+      "epoch": 1.1406480117820323,
+      "grad_norm": 0.4249773919582367,
+      "learning_rate": 7.794306584378392e-06,
+      "loss": 0.4015,
+      "step": 1549
+    },
+    {
+      "epoch": 1.1413843888070692,
+      "grad_norm": 0.4784921109676361,
+      "learning_rate": 7.79075236636013e-06,
+      "loss": 0.4199,
+      "step": 1550
+    },
+    {
+      "epoch": 1.142120765832106,
+      "grad_norm": 0.42127084732055664,
+      "learning_rate": 7.787196098897664e-06,
+      "loss": 0.3893,
+      "step": 1551
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.4333012104034424,
+      "learning_rate": 7.783637784602608e-06,
+      "loss": 0.3984,
+      "step": 1552
+    },
+    {
+      "epoch": 1.1435935198821796,
+      "grad_norm": 0.47243404388427734,
+      "learning_rate": 7.780077426088083e-06,
+      "loss": 0.4212,
+      "step": 1553
+    },
+    {
+      "epoch": 1.1443298969072164,
+      "grad_norm": 0.47604048252105713,
+      "learning_rate": 7.776515025968707e-06,
+      "loss": 0.3994,
+      "step": 1554
+    },
+    {
+      "epoch": 1.1450662739322532,
+      "grad_norm": 0.5063879489898682,
+      "learning_rate": 7.772950586860599e-06,
+      "loss": 0.4086,
+      "step": 1555
+    },
+    {
+      "epoch": 1.14580265095729,
+      "grad_norm": 0.43437930941581726,
+      "learning_rate": 7.769384111381375e-06,
+      "loss": 0.4358,
+      "step": 1556
+    },
+    {
+      "epoch": 1.1465390279823269,
+      "grad_norm": 0.46195188164711,
+      "learning_rate": 7.76581560215015e-06,
+      "loss": 0.4083,
+      "step": 1557
+    },
+    {
+      "epoch": 1.1472754050073637,
+      "grad_norm": 0.4431914985179901,
+      "learning_rate": 7.762245061787525e-06,
+      "loss": 0.4024,
+      "step": 1558
+    },
+    {
+      "epoch": 1.1480117820324005,
+      "grad_norm": 0.42202213406562805,
+      "learning_rate": 7.758672492915598e-06,
+      "loss": 0.3988,
+      "step": 1559
+    },
+    {
+      "epoch": 1.1487481590574373,
+      "grad_norm": 0.4758620858192444,
+      "learning_rate": 7.755097898157957e-06,
+      "loss": 0.4307,
+      "step": 1560
+    },
+    {
+      "epoch": 1.1494845360824741,
+      "grad_norm": 0.4043138027191162,
+      "learning_rate": 7.751521280139675e-06,
+      "loss": 0.3967,
+      "step": 1561
+    },
+    {
+      "epoch": 1.150220913107511,
+      "grad_norm": 0.41827574372291565,
+      "learning_rate": 7.747942641487313e-06,
+      "loss": 0.4107,
+      "step": 1562
+    },
+    {
+      "epoch": 1.1509572901325478,
+      "grad_norm": 0.45348092913627625,
+      "learning_rate": 7.74436198482892e-06,
+      "loss": 0.4278,
+      "step": 1563
+    },
+    {
+      "epoch": 1.1516936671575846,
+      "grad_norm": 0.43559592962265015,
+      "learning_rate": 7.74077931279401e-06,
+      "loss": 0.4146,
+      "step": 1564
+    },
+    {
+      "epoch": 1.1524300441826214,
+      "grad_norm": 0.3940446078777313,
+      "learning_rate": 7.7371946280136e-06,
+      "loss": 0.404,
+      "step": 1565
+    },
+    {
+      "epoch": 1.1531664212076582,
+      "grad_norm": 0.4286969006061554,
+      "learning_rate": 7.73360793312017e-06,
+      "loss": 0.4078,
+      "step": 1566
+    },
+    {
+      "epoch": 1.153902798232695,
+      "grad_norm": 0.4312724471092224,
+      "learning_rate": 7.730019230747681e-06,
+      "loss": 0.4088,
+      "step": 1567
+    },
+    {
+      "epoch": 1.1546391752577319,
+      "grad_norm": 0.41734251379966736,
+      "learning_rate": 7.726428523531565e-06,
+      "loss": 0.4228,
+      "step": 1568
+    },
+    {
+      "epoch": 1.1553755522827687,
+      "grad_norm": 0.435044527053833,
+      "learning_rate": 7.722835814108733e-06,
+      "loss": 0.4059,
+      "step": 1569
+    },
+    {
+      "epoch": 1.1561119293078055,
+      "grad_norm": 0.443087637424469,
+      "learning_rate": 7.719241105117559e-06,
+      "loss": 0.3977,
+      "step": 1570
+    },
+    {
+      "epoch": 1.1568483063328423,
+      "grad_norm": 0.3917474150657654,
+      "learning_rate": 7.715644399197893e-06,
+      "loss": 0.3999,
+      "step": 1571
+    },
+    {
+      "epoch": 1.1575846833578791,
+      "grad_norm": 0.4369930922985077,
+      "learning_rate": 7.712045698991043e-06,
+      "loss": 0.4112,
+      "step": 1572
+    },
+    {
+      "epoch": 1.158321060382916,
+      "grad_norm": 0.42742857336997986,
+      "learning_rate": 7.708445007139785e-06,
+      "loss": 0.4092,
+      "step": 1573
+    },
+    {
+      "epoch": 1.1590574374079528,
+      "grad_norm": 0.41142895817756653,
+      "learning_rate": 7.70484232628836e-06,
+      "loss": 0.4413,
+      "step": 1574
+    },
+    {
+      "epoch": 1.1597938144329896,
+      "grad_norm": 0.4239861071109772,
+      "learning_rate": 7.70123765908247e-06,
+      "loss": 0.4047,
+      "step": 1575
+    },
+    {
+      "epoch": 1.1605301914580266,
+      "grad_norm": 0.41946011781692505,
+      "learning_rate": 7.69763100816927e-06,
+      "loss": 0.3995,
+      "step": 1576
+    },
+    {
+      "epoch": 1.1612665684830634,
+      "grad_norm": 0.41380631923675537,
+      "learning_rate": 7.69402237619738e-06,
+      "loss": 0.4251,
+      "step": 1577
+    },
+    {
+      "epoch": 1.1620029455081002,
+      "grad_norm": 0.4237883985042572,
+      "learning_rate": 7.690411765816864e-06,
+      "loss": 0.4196,
+      "step": 1578
+    },
+    {
+      "epoch": 1.162739322533137,
+      "grad_norm": 0.4093558192253113,
+      "learning_rate": 7.68679917967925e-06,
+      "loss": 0.412,
+      "step": 1579
+    },
+    {
+      "epoch": 1.1634756995581739,
+      "grad_norm": 0.41598761081695557,
+      "learning_rate": 7.683184620437511e-06,
+      "loss": 0.4031,
+      "step": 1580
+    },
+    {
+      "epoch": 1.1642120765832107,
+      "grad_norm": 0.43564292788505554,
+      "learning_rate": 7.67956809074607e-06,
+      "loss": 0.397,
+      "step": 1581
+    },
+    {
+      "epoch": 1.1649484536082475,
+      "grad_norm": 0.40577253699302673,
+      "learning_rate": 7.675949593260797e-06,
+      "loss": 0.3964,
+      "step": 1582
+    },
+    {
+      "epoch": 1.1656848306332843,
+      "grad_norm": 0.43315377831459045,
+      "learning_rate": 7.672329130639007e-06,
+      "loss": 0.4423,
+      "step": 1583
+    },
+    {
+      "epoch": 1.1664212076583211,
+      "grad_norm": 0.5157422423362732,
+      "learning_rate": 7.668706705539458e-06,
+      "loss": 0.4398,
+      "step": 1584
+    },
+    {
+      "epoch": 1.167157584683358,
+      "grad_norm": 0.4576352834701538,
+      "learning_rate": 7.66508232062235e-06,
+      "loss": 0.4075,
+      "step": 1585
+    },
+    {
+      "epoch": 1.1678939617083948,
+      "grad_norm": 0.3952282965183258,
+      "learning_rate": 7.661455978549322e-06,
+      "loss": 0.4418,
+      "step": 1586
+    },
+    {
+      "epoch": 1.1686303387334316,
+      "grad_norm": 0.40662047266960144,
+      "learning_rate": 7.657827681983448e-06,
+      "loss": 0.3875,
+      "step": 1587
+    },
+    {
+      "epoch": 1.1693667157584684,
+      "grad_norm": 0.46831056475639343,
+      "learning_rate": 7.654197433589243e-06,
+      "loss": 0.4206,
+      "step": 1588
+    },
+    {
+      "epoch": 1.1701030927835052,
+      "grad_norm": 0.44352883100509644,
+      "learning_rate": 7.65056523603265e-06,
+      "loss": 0.3882,
+      "step": 1589
+    },
+    {
+      "epoch": 1.170839469808542,
+      "grad_norm": 0.4131197929382324,
+      "learning_rate": 7.646931091981045e-06,
+      "loss": 0.4203,
+      "step": 1590
+    },
+    {
+      "epoch": 1.1715758468335788,
+      "grad_norm": 0.543073296546936,
+      "learning_rate": 7.643295004103232e-06,
+      "loss": 0.4035,
+      "step": 1591
+    },
+    {
+      "epoch": 1.1723122238586157,
+      "grad_norm": 0.42148828506469727,
+      "learning_rate": 7.639656975069447e-06,
+      "loss": 0.4257,
+      "step": 1592
+    },
+    {
+      "epoch": 1.1730486008836525,
+      "grad_norm": 0.45353934168815613,
+      "learning_rate": 7.636017007551349e-06,
+      "loss": 0.4306,
+      "step": 1593
+    },
+    {
+      "epoch": 1.1737849779086893,
+      "grad_norm": 0.42976027727127075,
+      "learning_rate": 7.632375104222014e-06,
+      "loss": 0.3962,
+      "step": 1594
+    },
+    {
+      "epoch": 1.1745213549337261,
+      "grad_norm": 0.4995291829109192,
+      "learning_rate": 7.628731267755952e-06,
+      "loss": 0.3974,
+      "step": 1595
+    },
+    {
+      "epoch": 1.175257731958763,
+      "grad_norm": 0.4624010920524597,
+      "learning_rate": 7.6250855008290856e-06,
+      "loss": 0.4175,
+      "step": 1596
+    },
+    {
+      "epoch": 1.1759941089837997,
+      "grad_norm": 0.46264103055000305,
+      "learning_rate": 7.6214378061187546e-06,
+      "loss": 0.3981,
+      "step": 1597
+    },
+    {
+      "epoch": 1.1767304860088366,
+      "grad_norm": 0.48678502440452576,
+      "learning_rate": 7.617788186303714e-06,
+      "loss": 0.4185,
+      "step": 1598
+    },
+    {
+      "epoch": 1.1774668630338734,
+      "grad_norm": 0.4638168513774872,
+      "learning_rate": 7.6141366440641365e-06,
+      "loss": 0.4137,
+      "step": 1599
+    },
+    {
+      "epoch": 1.1782032400589102,
+      "grad_norm": 0.4336228668689728,
+      "learning_rate": 7.610483182081607e-06,
+      "loss": 0.4418,
+      "step": 1600
+    },
+    {
+      "epoch": 1.178939617083947,
+      "grad_norm": 0.47200411558151245,
+      "learning_rate": 7.606827803039112e-06,
+      "loss": 0.4105,
+      "step": 1601
+    },
+    {
+      "epoch": 1.1796759941089838,
+      "grad_norm": 0.5027320981025696,
+      "learning_rate": 7.603170509621054e-06,
+      "loss": 0.4476,
+      "step": 1602
+    },
+    {
+      "epoch": 1.1804123711340206,
+      "grad_norm": 0.4112722873687744,
+      "learning_rate": 7.5995113045132395e-06,
+      "loss": 0.4147,
+      "step": 1603
+    },
+    {
+      "epoch": 1.1811487481590575,
+      "grad_norm": 0.4506515562534332,
+      "learning_rate": 7.595850190402877e-06,
+      "loss": 0.4221,
+      "step": 1604
+    },
+    {
+      "epoch": 1.1818851251840943,
+      "grad_norm": 0.4421546757221222,
+      "learning_rate": 7.59218716997858e-06,
+      "loss": 0.4313,
+      "step": 1605
+    },
+    {
+      "epoch": 1.182621502209131,
+      "grad_norm": 0.4537235200405121,
+      "learning_rate": 7.588522245930357e-06,
+      "loss": 0.4175,
+      "step": 1606
+    },
+    {
+      "epoch": 1.183357879234168,
+      "grad_norm": 0.4480210244655609,
+      "learning_rate": 7.584855420949619e-06,
+      "loss": 0.4213,
+      "step": 1607
+    },
+    {
+      "epoch": 1.1840942562592047,
+      "grad_norm": 0.408025324344635,
+      "learning_rate": 7.581186697729172e-06,
+      "loss": 0.4099,
+      "step": 1608
+    },
+    {
+      "epoch": 1.1848306332842415,
+      "grad_norm": 0.45869341492652893,
+      "learning_rate": 7.577516078963215e-06,
+      "loss": 0.4143,
+      "step": 1609
+    },
+    {
+      "epoch": 1.1855670103092784,
+      "grad_norm": 0.48638081550598145,
+      "learning_rate": 7.573843567347339e-06,
+      "loss": 0.4361,
+      "step": 1610
+    },
+    {
+      "epoch": 1.1863033873343152,
+      "grad_norm": 0.39438486099243164,
+      "learning_rate": 7.570169165578527e-06,
+      "loss": 0.4394,
+      "step": 1611
+    },
+    {
+      "epoch": 1.187039764359352,
+      "grad_norm": 0.49038755893707275,
+      "learning_rate": 7.566492876355147e-06,
+      "loss": 0.416,
+      "step": 1612
+    },
+    {
+      "epoch": 1.1877761413843888,
+      "grad_norm": 0.4872884452342987,
+      "learning_rate": 7.562814702376955e-06,
+      "loss": 0.4071,
+      "step": 1613
+    },
+    {
+      "epoch": 1.1885125184094256,
+      "grad_norm": 0.4326101541519165,
+      "learning_rate": 7.559134646345092e-06,
+      "loss": 0.4061,
+      "step": 1614
+    },
+    {
+      "epoch": 1.1892488954344624,
+      "grad_norm": 0.48473793268203735,
+      "learning_rate": 7.5554527109620775e-06,
+      "loss": 0.4206,
+      "step": 1615
+    },
+    {
+      "epoch": 1.1899852724594993,
+      "grad_norm": 0.43738964200019836,
+      "learning_rate": 7.551768898931816e-06,
+      "loss": 0.4113,
+      "step": 1616
+    },
+    {
+      "epoch": 1.190721649484536,
+      "grad_norm": 0.41719716787338257,
+      "learning_rate": 7.548083212959588e-06,
+      "loss": 0.4126,
+      "step": 1617
+    },
+    {
+      "epoch": 1.1914580265095729,
+      "grad_norm": 0.4838084280490875,
+      "learning_rate": 7.5443956557520485e-06,
+      "loss": 0.4494,
+      "step": 1618
+    },
+    {
+      "epoch": 1.1921944035346097,
+      "grad_norm": 0.4023309051990509,
+      "learning_rate": 7.540706230017227e-06,
+      "loss": 0.4212,
+      "step": 1619
+    },
+    {
+      "epoch": 1.1929307805596465,
+      "grad_norm": 0.42563098669052124,
+      "learning_rate": 7.537014938464529e-06,
+      "loss": 0.378,
+      "step": 1620
+    },
+    {
+      "epoch": 1.1936671575846833,
+      "grad_norm": 0.4844202697277069,
+      "learning_rate": 7.533321783804726e-06,
+      "loss": 0.4024,
+      "step": 1621
+    },
+    {
+      "epoch": 1.1944035346097202,
+      "grad_norm": 0.39639124274253845,
+      "learning_rate": 7.529626768749958e-06,
+      "loss": 0.4082,
+      "step": 1622
+    },
+    {
+      "epoch": 1.195139911634757,
+      "grad_norm": 0.45201244950294495,
+      "learning_rate": 7.525929896013735e-06,
+      "loss": 0.4252,
+      "step": 1623
+    },
+    {
+      "epoch": 1.1958762886597938,
+      "grad_norm": 0.463316947221756,
+      "learning_rate": 7.5222311683109265e-06,
+      "loss": 0.417,
+      "step": 1624
+    },
+    {
+      "epoch": 1.1966126656848306,
+      "grad_norm": 0.42931288480758667,
+      "learning_rate": 7.518530588357769e-06,
+      "loss": 0.4369,
+      "step": 1625
+    },
+    {
+      "epoch": 1.1973490427098674,
+      "grad_norm": 0.44762808084487915,
+      "learning_rate": 7.514828158871852e-06,
+      "loss": 0.3949,
+      "step": 1626
+    },
+    {
+      "epoch": 1.1980854197349042,
+      "grad_norm": 0.43206796050071716,
+      "learning_rate": 7.511123882572133e-06,
+      "loss": 0.4018,
+      "step": 1627
+    },
+    {
+      "epoch": 1.198821796759941,
+      "grad_norm": 0.41934624314308167,
+      "learning_rate": 7.5074177621789155e-06,
+      "loss": 0.4152,
+      "step": 1628
+    },
+    {
+      "epoch": 1.1995581737849779,
+      "grad_norm": 0.41448119282722473,
+      "learning_rate": 7.503709800413868e-06,
+      "loss": 0.4424,
+      "step": 1629
+    },
+    {
+      "epoch": 1.2002945508100147,
+      "grad_norm": 0.4471980333328247,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.417,
+      "step": 1630
+    },
+    {
+      "epoch": 1.2010309278350515,
+      "grad_norm": 0.4638178050518036,
+      "learning_rate": 7.496288363661681e-06,
+      "loss": 0.431,
+      "step": 1631
+    },
+    {
+      "epoch": 1.2017673048600883,
+      "grad_norm": 0.4889911115169525,
+      "learning_rate": 7.492574894124624e-06,
+      "loss": 0.4159,
+      "step": 1632
+    },
+    {
+      "epoch": 1.2025036818851251,
+      "grad_norm": 0.4356183707714081,
+      "learning_rate": 7.4888595941158844e-06,
+      "loss": 0.4258,
+      "step": 1633
+    },
+    {
+      "epoch": 1.203240058910162,
+      "grad_norm": 0.4645734429359436,
+      "learning_rate": 7.485142466363873e-06,
+      "loss": 0.3947,
+      "step": 1634
+    },
+    {
+      "epoch": 1.2039764359351988,
+      "grad_norm": 0.44492483139038086,
+      "learning_rate": 7.481423513598331e-06,
+      "loss": 0.4154,
+      "step": 1635
+    },
+    {
+      "epoch": 1.2047128129602356,
+      "grad_norm": 0.42611250281333923,
+      "learning_rate": 7.477702738550346e-06,
+      "loss": 0.4105,
+      "step": 1636
+    },
+    {
+      "epoch": 1.2054491899852724,
+      "grad_norm": 0.4502268135547638,
+      "learning_rate": 7.473980143952344e-06,
+      "loss": 0.3943,
+      "step": 1637
+    },
+    {
+      "epoch": 1.2061855670103092,
+      "grad_norm": 0.4995657801628113,
+      "learning_rate": 7.470255732538086e-06,
+      "loss": 0.401,
+      "step": 1638
+    },
+    {
+      "epoch": 1.206921944035346,
+      "grad_norm": 0.399395227432251,
+      "learning_rate": 7.466529507042666e-06,
+      "loss": 0.3953,
+      "step": 1639
+    },
+    {
+      "epoch": 1.2076583210603828,
+      "grad_norm": 0.472028523683548,
+      "learning_rate": 7.462801470202513e-06,
+      "loss": 0.412,
+      "step": 1640
+    },
+    {
+      "epoch": 1.2083946980854197,
+      "grad_norm": 0.44773492217063904,
+      "learning_rate": 7.459071624755382e-06,
+      "loss": 0.4121,
+      "step": 1641
+    },
+    {
+      "epoch": 1.2091310751104565,
+      "grad_norm": 0.4189164340496063,
+      "learning_rate": 7.455339973440361e-06,
+      "loss": 0.398,
+      "step": 1642
+    },
+    {
+      "epoch": 1.2098674521354933,
+      "grad_norm": 0.40162959694862366,
+      "learning_rate": 7.4516065189978625e-06,
+      "loss": 0.4068,
+      "step": 1643
+    },
+    {
+      "epoch": 1.21060382916053,
+      "grad_norm": 0.4244077503681183,
+      "learning_rate": 7.4478712641696194e-06,
+      "loss": 0.4342,
+      "step": 1644
+    },
+    {
+      "epoch": 1.211340206185567,
+      "grad_norm": 0.4497550427913666,
+      "learning_rate": 7.444134211698692e-06,
+      "loss": 0.4067,
+      "step": 1645
+    },
+    {
+      "epoch": 1.2120765832106037,
+      "grad_norm": 0.4071500301361084,
+      "learning_rate": 7.44039536432946e-06,
+      "loss": 0.4084,
+      "step": 1646
+    },
+    {
+      "epoch": 1.2128129602356406,
+      "grad_norm": 0.413670152425766,
+      "learning_rate": 7.43665472480762e-06,
+      "loss": 0.439,
+      "step": 1647
+    },
+    {
+      "epoch": 1.2135493372606774,
+      "grad_norm": 0.4371226727962494,
+      "learning_rate": 7.4329122958801806e-06,
+      "loss": 0.3991,
+      "step": 1648
+    },
+    {
+      "epoch": 1.2142857142857142,
+      "grad_norm": 0.4222114384174347,
+      "learning_rate": 7.4291680802954716e-06,
+      "loss": 0.4208,
+      "step": 1649
+    },
+    {
+      "epoch": 1.2150220913107512,
+      "grad_norm": 0.4362742304801941,
+      "learning_rate": 7.425422080803132e-06,
+      "loss": 0.3812,
+      "step": 1650
+    },
+    {
+      "epoch": 1.215758468335788,
+      "grad_norm": 0.4180275499820709,
+      "learning_rate": 7.42167430015411e-06,
+      "loss": 0.4073,
+      "step": 1651
+    },
+    {
+      "epoch": 1.2164948453608249,
+      "grad_norm": 0.42266976833343506,
+      "learning_rate": 7.417924741100662e-06,
+      "loss": 0.4104,
+      "step": 1652
+    },
+    {
+      "epoch": 1.2172312223858617,
+      "grad_norm": 0.44959259033203125,
+      "learning_rate": 7.414173406396351e-06,
+      "loss": 0.4324,
+      "step": 1653
+    },
+    {
+      "epoch": 1.2179675994108985,
+      "grad_norm": 0.42745327949523926,
+      "learning_rate": 7.410420298796045e-06,
+      "loss": 0.413,
+      "step": 1654
+    },
+    {
+      "epoch": 1.2187039764359353,
+      "grad_norm": 0.3998710811138153,
+      "learning_rate": 7.406665421055912e-06,
+      "loss": 0.4059,
+      "step": 1655
+    },
+    {
+      "epoch": 1.2194403534609721,
+      "grad_norm": 0.4011445641517639,
+      "learning_rate": 7.402908775933419e-06,
+      "loss": 0.4191,
+      "step": 1656
+    },
+    {
+      "epoch": 1.220176730486009,
+      "grad_norm": 0.44070056080818176,
+      "learning_rate": 7.399150366187336e-06,
+      "loss": 0.4215,
+      "step": 1657
+    },
+    {
+      "epoch": 1.2209131075110458,
+      "grad_norm": 0.4228590726852417,
+      "learning_rate": 7.395390194577724e-06,
+      "loss": 0.4107,
+      "step": 1658
+    },
+    {
+      "epoch": 1.2216494845360826,
+      "grad_norm": 0.3656286299228668,
+      "learning_rate": 7.391628263865939e-06,
+      "loss": 0.3822,
+      "step": 1659
+    },
+    {
+      "epoch": 1.2223858615611194,
+      "grad_norm": 0.4292581081390381,
+      "learning_rate": 7.387864576814628e-06,
+      "loss": 0.4188,
+      "step": 1660
+    },
+    {
+      "epoch": 1.2231222385861562,
+      "grad_norm": 0.39621493220329285,
+      "learning_rate": 7.3840991361877326e-06,
+      "loss": 0.3921,
+      "step": 1661
+    },
+    {
+      "epoch": 1.223858615611193,
+      "grad_norm": 0.4454437792301178,
+      "learning_rate": 7.380331944750476e-06,
+      "loss": 0.4097,
+      "step": 1662
+    },
+    {
+      "epoch": 1.2245949926362298,
+      "grad_norm": 0.39751824736595154,
+      "learning_rate": 7.37656300526937e-06,
+      "loss": 0.4232,
+      "step": 1663
+    },
+    {
+      "epoch": 1.2253313696612667,
+      "grad_norm": 0.4180833399295807,
+      "learning_rate": 7.37279232051221e-06,
+      "loss": 0.4162,
+      "step": 1664
+    },
+    {
+      "epoch": 1.2260677466863035,
+      "grad_norm": 0.41712233424186707,
+      "learning_rate": 7.369019893248074e-06,
+      "loss": 0.4304,
+      "step": 1665
+    },
+    {
+      "epoch": 1.2268041237113403,
+      "grad_norm": 0.38904157280921936,
+      "learning_rate": 7.365245726247316e-06,
+      "loss": 0.427,
+      "step": 1666
+    },
+    {
+      "epoch": 1.227540500736377,
+      "grad_norm": 0.41479575634002686,
+      "learning_rate": 7.361469822281573e-06,
+      "loss": 0.4401,
+      "step": 1667
+    },
+    {
+      "epoch": 1.228276877761414,
+      "grad_norm": 0.3810448944568634,
+      "learning_rate": 7.3576921841237535e-06,
+      "loss": 0.4167,
+      "step": 1668
+    },
+    {
+      "epoch": 1.2290132547864507,
+      "grad_norm": 0.42002904415130615,
+      "learning_rate": 7.353912814548042e-06,
+      "loss": 0.4009,
+      "step": 1669
+    },
+    {
+      "epoch": 1.2297496318114876,
+      "grad_norm": 0.3993929326534271,
+      "learning_rate": 7.350131716329891e-06,
+      "loss": 0.4106,
+      "step": 1670
+    },
+    {
+      "epoch": 1.2304860088365244,
+      "grad_norm": 0.4273282587528229,
+      "learning_rate": 7.346348892246026e-06,
+      "loss": 0.4382,
+      "step": 1671
+    },
+    {
+      "epoch": 1.2312223858615612,
+      "grad_norm": 0.43270593881607056,
+      "learning_rate": 7.342564345074441e-06,
+      "loss": 0.4107,
+      "step": 1672
+    },
+    {
+      "epoch": 1.231958762886598,
+      "grad_norm": 0.40336766839027405,
+      "learning_rate": 7.338778077594388e-06,
+      "loss": 0.4238,
+      "step": 1673
+    },
+    {
+      "epoch": 1.2326951399116348,
+      "grad_norm": 0.39280226826667786,
+      "learning_rate": 7.33499009258639e-06,
+      "loss": 0.3768,
+      "step": 1674
+    },
+    {
+      "epoch": 1.2334315169366716,
+      "grad_norm": 0.4784250557422638,
+      "learning_rate": 7.331200392832232e-06,
+      "loss": 0.4272,
+      "step": 1675
+    },
+    {
+      "epoch": 1.2341678939617085,
+      "grad_norm": 0.3877667188644409,
+      "learning_rate": 7.32740898111495e-06,
+      "loss": 0.3862,
+      "step": 1676
+    },
+    {
+      "epoch": 1.2349042709867453,
+      "grad_norm": 0.4045974016189575,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 0.4099,
+      "step": 1677
+    },
+    {
+      "epoch": 1.235640648011782,
+      "grad_norm": 0.4475846588611603,
+      "learning_rate": 7.319821032929467e-06,
+      "loss": 0.4142,
+      "step": 1678
+    },
+    {
+      "epoch": 1.236377025036819,
+      "grad_norm": 0.4150718152523041,
+      "learning_rate": 7.316024502033627e-06,
+      "loss": 0.428,
+      "step": 1679
+    },
+    {
+      "epoch": 1.2371134020618557,
+      "grad_norm": 0.4547623097896576,
+      "learning_rate": 7.31222627031938e-06,
+      "loss": 0.4304,
+      "step": 1680
+    },
+    {
+      "epoch": 1.2378497790868925,
+      "grad_norm": 0.4688529968261719,
+      "learning_rate": 7.308426340576034e-06,
+      "loss": 0.3903,
+      "step": 1681
+    },
+    {
+      "epoch": 1.2385861561119293,
+      "grad_norm": 0.5233607888221741,
+      "learning_rate": 7.30462471559414e-06,
+      "loss": 0.3849,
+      "step": 1682
+    },
+    {
+      "epoch": 1.2393225331369662,
+      "grad_norm": 0.43747377395629883,
+      "learning_rate": 7.3008213981655005e-06,
+      "loss": 0.3894,
+      "step": 1683
+    },
+    {
+      "epoch": 1.240058910162003,
+      "grad_norm": 0.40247631072998047,
+      "learning_rate": 7.297016391083154e-06,
+      "loss": 0.376,
+      "step": 1684
+    },
+    {
+      "epoch": 1.2407952871870398,
+      "grad_norm": 0.49333998560905457,
+      "learning_rate": 7.2932096971413815e-06,
+      "loss": 0.4105,
+      "step": 1685
+    },
+    {
+      "epoch": 1.2415316642120766,
+      "grad_norm": 0.4000438153743744,
+      "learning_rate": 7.289401319135707e-06,
+      "loss": 0.3952,
+      "step": 1686
+    },
+    {
+      "epoch": 1.2422680412371134,
+      "grad_norm": 0.44068193435668945,
+      "learning_rate": 7.285591259862888e-06,
+      "loss": 0.379,
+      "step": 1687
+    },
+    {
+      "epoch": 1.2430044182621502,
+      "grad_norm": 0.4409189820289612,
+      "learning_rate": 7.281779522120914e-06,
+      "loss": 0.4173,
+      "step": 1688
+    },
+    {
+      "epoch": 1.243740795287187,
+      "grad_norm": 0.4082186222076416,
+      "learning_rate": 7.277966108709014e-06,
+      "loss": 0.407,
+      "step": 1689
+    },
+    {
+      "epoch": 1.2444771723122239,
+      "grad_norm": 0.40812796354293823,
+      "learning_rate": 7.27415102242764e-06,
+      "loss": 0.4243,
+      "step": 1690
+    },
+    {
+      "epoch": 1.2452135493372607,
+      "grad_norm": 0.41487666964530945,
+      "learning_rate": 7.2703342660784785e-06,
+      "loss": 0.4032,
+      "step": 1691
+    },
+    {
+      "epoch": 1.2459499263622975,
+      "grad_norm": 0.44201499223709106,
+      "learning_rate": 7.266515842464438e-06,
+      "loss": 0.448,
+      "step": 1692
+    },
+    {
+      "epoch": 1.2466863033873343,
+      "grad_norm": 0.4068033695220947,
+      "learning_rate": 7.262695754389655e-06,
+      "loss": 0.4022,
+      "step": 1693
+    },
+    {
+      "epoch": 1.2474226804123711,
+      "grad_norm": 0.3832601010799408,
+      "learning_rate": 7.258874004659487e-06,
+      "loss": 0.4315,
+      "step": 1694
+    },
+    {
+      "epoch": 1.248159057437408,
+      "grad_norm": 0.4322513937950134,
+      "learning_rate": 7.25505059608051e-06,
+      "loss": 0.4123,
+      "step": 1695
+    },
+    {
+      "epoch": 1.2488954344624448,
+      "grad_norm": 0.3983667194843292,
+      "learning_rate": 7.25122553146052e-06,
+      "loss": 0.3817,
+      "step": 1696
+    },
+    {
+      "epoch": 1.2496318114874816,
+      "grad_norm": 0.44032666087150574,
+      "learning_rate": 7.247398813608531e-06,
+      "loss": 0.3921,
+      "step": 1697
+    },
+    {
+      "epoch": 1.2503681885125184,
+      "grad_norm": 0.451885461807251,
+      "learning_rate": 7.243570445334767e-06,
+      "loss": 0.4332,
+      "step": 1698
+    },
+    {
+      "epoch": 1.2511045655375552,
+      "grad_norm": 0.46099600195884705,
+      "learning_rate": 7.239740429450664e-06,
+      "loss": 0.4037,
+      "step": 1699
+    },
+    {
+      "epoch": 1.251840942562592,
+      "grad_norm": 0.4341306984424591,
+      "learning_rate": 7.235908768768875e-06,
+      "loss": 0.3931,
+      "step": 1700
+    },
+    {
+      "epoch": 1.2525773195876289,
+      "grad_norm": 0.4982093274593353,
+      "learning_rate": 7.232075466103254e-06,
+      "loss": 0.4291,
+      "step": 1701
+    },
+    {
+      "epoch": 1.2533136966126657,
+      "grad_norm": 0.4553844928741455,
+      "learning_rate": 7.228240524268858e-06,
+      "loss": 0.4011,
+      "step": 1702
+    },
+    {
+      "epoch": 1.2540500736377025,
+      "grad_norm": 0.43320947885513306,
+      "learning_rate": 7.224403946081957e-06,
+      "loss": 0.4023,
+      "step": 1703
+    },
+    {
+      "epoch": 1.2547864506627393,
+      "grad_norm": 0.46215543150901794,
+      "learning_rate": 7.220565734360019e-06,
+      "loss": 0.4213,
+      "step": 1704
+    },
+    {
+      "epoch": 1.2555228276877761,
+      "grad_norm": 0.4323497712612152,
+      "learning_rate": 7.216725891921707e-06,
+      "loss": 0.4226,
+      "step": 1705
+    },
+    {
+      "epoch": 1.256259204712813,
+      "grad_norm": 0.47635823488235474,
+      "learning_rate": 7.212884421586889e-06,
+      "loss": 0.399,
+      "step": 1706
+    },
+    {
+      "epoch": 1.2569955817378498,
+      "grad_norm": 0.40797707438468933,
+      "learning_rate": 7.20904132617662e-06,
+      "loss": 0.3783,
+      "step": 1707
+    },
+    {
+      "epoch": 1.2577319587628866,
+      "grad_norm": 0.5090800523757935,
+      "learning_rate": 7.2051966085131584e-06,
+      "loss": 0.4148,
+      "step": 1708
+    },
+    {
+      "epoch": 1.2584683357879234,
+      "grad_norm": 0.5189807415008545,
+      "learning_rate": 7.201350271419945e-06,
+      "loss": 0.4367,
+      "step": 1709
+    },
+    {
+      "epoch": 1.2592047128129602,
+      "grad_norm": 0.40211692452430725,
+      "learning_rate": 7.197502317721616e-06,
+      "loss": 0.3966,
+      "step": 1710
+    },
+    {
+      "epoch": 1.259941089837997,
+      "grad_norm": 0.45923155546188354,
+      "learning_rate": 7.19365275024399e-06,
+      "loss": 0.4394,
+      "step": 1711
+    },
+    {
+      "epoch": 1.2606774668630338,
+      "grad_norm": 0.48686525225639343,
+      "learning_rate": 7.189801571814075e-06,
+      "loss": 0.4252,
+      "step": 1712
+    },
+    {
+      "epoch": 1.2614138438880707,
+      "grad_norm": 0.43244504928588867,
+      "learning_rate": 7.185948785260058e-06,
+      "loss": 0.3866,
+      "step": 1713
+    },
+    {
+      "epoch": 1.2621502209131075,
+      "grad_norm": 0.4492274820804596,
+      "learning_rate": 7.182094393411312e-06,
+      "loss": 0.423,
+      "step": 1714
+    },
+    {
+      "epoch": 1.2628865979381443,
+      "grad_norm": 0.404514878988266,
+      "learning_rate": 7.178238399098381e-06,
+      "loss": 0.3901,
+      "step": 1715
+    },
+    {
+      "epoch": 1.263622974963181,
+      "grad_norm": 0.41889336705207825,
+      "learning_rate": 7.174380805152997e-06,
+      "loss": 0.402,
+      "step": 1716
+    },
+    {
+      "epoch": 1.264359351988218,
+      "grad_norm": 0.4595990777015686,
+      "learning_rate": 7.170521614408056e-06,
+      "loss": 0.3852,
+      "step": 1717
+    },
+    {
+      "epoch": 1.2650957290132547,
+      "grad_norm": 0.422296941280365,
+      "learning_rate": 7.166660829697633e-06,
+      "loss": 0.4092,
+      "step": 1718
+    },
+    {
+      "epoch": 1.2658321060382915,
+      "grad_norm": 0.49134594202041626,
+      "learning_rate": 7.162798453856969e-06,
+      "loss": 0.4246,
+      "step": 1719
+    },
+    {
+      "epoch": 1.2665684830633284,
+      "grad_norm": 0.4637044072151184,
+      "learning_rate": 7.1589344897224795e-06,
+      "loss": 0.3807,
+      "step": 1720
+    },
+    {
+      "epoch": 1.2673048600883652,
+      "grad_norm": 0.49215713143348694,
+      "learning_rate": 7.155068940131741e-06,
+      "loss": 0.4121,
+      "step": 1721
+    },
+    {
+      "epoch": 1.268041237113402,
+      "grad_norm": 0.46819546818733215,
+      "learning_rate": 7.151201807923497e-06,
+      "loss": 0.4045,
+      "step": 1722
+    },
+    {
+      "epoch": 1.2687776141384388,
+      "grad_norm": 0.4482595920562744,
+      "learning_rate": 7.1473330959376515e-06,
+      "loss": 0.3915,
+      "step": 1723
+    },
+    {
+      "epoch": 1.2695139911634756,
+      "grad_norm": 0.5111578702926636,
+      "learning_rate": 7.143462807015271e-06,
+      "loss": 0.4096,
+      "step": 1724
+    },
+    {
+      "epoch": 1.2702503681885124,
+      "grad_norm": 0.3854685425758362,
+      "learning_rate": 7.139590943998579e-06,
+      "loss": 0.4059,
+      "step": 1725
+    },
+    {
+      "epoch": 1.2709867452135493,
+      "grad_norm": 0.542492151260376,
+      "learning_rate": 7.135717509730953e-06,
+      "loss": 0.4306,
+      "step": 1726
+    },
+    {
+      "epoch": 1.271723122238586,
+      "grad_norm": 0.4321967661380768,
+      "learning_rate": 7.1318425070569284e-06,
+      "loss": 0.3901,
+      "step": 1727
+    },
+    {
+      "epoch": 1.272459499263623,
+      "grad_norm": 0.41840946674346924,
+      "learning_rate": 7.127965938822187e-06,
+      "loss": 0.4079,
+      "step": 1728
+    },
+    {
+      "epoch": 1.2731958762886597,
+      "grad_norm": 0.3965054750442505,
+      "learning_rate": 7.124087807873565e-06,
+      "loss": 0.4081,
+      "step": 1729
+    },
+    {
+      "epoch": 1.2739322533136965,
+      "grad_norm": 0.4440041184425354,
+      "learning_rate": 7.1202081170590455e-06,
+      "loss": 0.4106,
+      "step": 1730
+    },
+    {
+      "epoch": 1.2746686303387333,
+      "grad_norm": 0.4273045063018799,
+      "learning_rate": 7.116326869227755e-06,
+      "loss": 0.3929,
+      "step": 1731
+    },
+    {
+      "epoch": 1.2754050073637702,
+      "grad_norm": 0.3961130976676941,
+      "learning_rate": 7.112444067229966e-06,
+      "loss": 0.4022,
+      "step": 1732
+    },
+    {
+      "epoch": 1.276141384388807,
+      "grad_norm": 0.3908890187740326,
+      "learning_rate": 7.108559713917089e-06,
+      "loss": 0.3974,
+      "step": 1733
+    },
+    {
+      "epoch": 1.2768777614138438,
+      "grad_norm": 0.4090268909931183,
+      "learning_rate": 7.104673812141676e-06,
+      "loss": 0.4551,
+      "step": 1734
+    },
+    {
+      "epoch": 1.2776141384388806,
+      "grad_norm": 0.38256412744522095,
+      "learning_rate": 7.100786364757417e-06,
+      "loss": 0.4439,
+      "step": 1735
+    },
+    {
+      "epoch": 1.2783505154639174,
+      "grad_norm": 0.37507322430610657,
+      "learning_rate": 7.096897374619134e-06,
+      "loss": 0.4087,
+      "step": 1736
+    },
+    {
+      "epoch": 1.2790868924889542,
+      "grad_norm": 0.39156273007392883,
+      "learning_rate": 7.093006844582787e-06,
+      "loss": 0.3903,
+      "step": 1737
+    },
+    {
+      "epoch": 1.279823269513991,
+      "grad_norm": 0.37378448247909546,
+      "learning_rate": 7.08911477750546e-06,
+      "loss": 0.4178,
+      "step": 1738
+    },
+    {
+      "epoch": 1.2805596465390279,
+      "grad_norm": 0.4035623371601105,
+      "learning_rate": 7.085221176245374e-06,
+      "loss": 0.4079,
+      "step": 1739
+    },
+    {
+      "epoch": 1.2812960235640647,
+      "grad_norm": 0.4087090492248535,
+      "learning_rate": 7.081326043661867e-06,
+      "loss": 0.39,
+      "step": 1740
+    },
+    {
+      "epoch": 1.2820324005891015,
+      "grad_norm": 0.43084970116615295,
+      "learning_rate": 7.0774293826154095e-06,
+      "loss": 0.4189,
+      "step": 1741
+    },
+    {
+      "epoch": 1.2827687776141383,
+      "grad_norm": 0.40026846528053284,
+      "learning_rate": 7.073531195967591e-06,
+      "loss": 0.4454,
+      "step": 1742
+    },
+    {
+      "epoch": 1.2835051546391751,
+      "grad_norm": 0.39230623841285706,
+      "learning_rate": 7.069631486581123e-06,
+      "loss": 0.4106,
+      "step": 1743
+    },
+    {
+      "epoch": 1.284241531664212,
+      "grad_norm": 0.39439305663108826,
+      "learning_rate": 7.065730257319832e-06,
+      "loss": 0.4,
+      "step": 1744
+    },
+    {
+      "epoch": 1.2849779086892488,
+      "grad_norm": 0.37384194135665894,
+      "learning_rate": 7.061827511048666e-06,
+      "loss": 0.3976,
+      "step": 1745
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.4040299355983734,
+      "learning_rate": 7.057923250633681e-06,
+      "loss": 0.4322,
+      "step": 1746
+    },
+    {
+      "epoch": 1.2864506627393224,
+      "grad_norm": 0.42770475149154663,
+      "learning_rate": 7.054017478942048e-06,
+      "loss": 0.4132,
+      "step": 1747
+    },
+    {
+      "epoch": 1.2871870397643592,
+      "grad_norm": 0.43052366375923157,
+      "learning_rate": 7.050110198842052e-06,
+      "loss": 0.4313,
+      "step": 1748
+    },
+    {
+      "epoch": 1.287923416789396,
+      "grad_norm": 0.3632519841194153,
+      "learning_rate": 7.046201413203076e-06,
+      "loss": 0.4088,
+      "step": 1749
+    },
+    {
+      "epoch": 1.2886597938144329,
+      "grad_norm": 0.3969815671443939,
+      "learning_rate": 7.042291124895615e-06,
+      "loss": 0.4149,
+      "step": 1750
+    },
+    {
+      "epoch": 1.28939617083947,
+      "grad_norm": 0.45275336503982544,
+      "learning_rate": 7.038379336791269e-06,
+      "loss": 0.4252,
+      "step": 1751
+    },
+    {
+      "epoch": 1.2901325478645067,
+      "grad_norm": 0.4112440049648285,
+      "learning_rate": 7.034466051762736e-06,
+      "loss": 0.4302,
+      "step": 1752
+    },
+    {
+      "epoch": 1.2908689248895435,
+      "grad_norm": 0.4475271999835968,
+      "learning_rate": 7.030551272683814e-06,
+      "loss": 0.4077,
+      "step": 1753
+    },
+    {
+      "epoch": 1.2916053019145803,
+      "grad_norm": 0.4204852283000946,
+      "learning_rate": 7.026635002429399e-06,
+      "loss": 0.4264,
+      "step": 1754
+    },
+    {
+      "epoch": 1.2923416789396172,
+      "grad_norm": 0.44256094098091125,
+      "learning_rate": 7.02271724387548e-06,
+      "loss": 0.4211,
+      "step": 1755
+    },
+    {
+      "epoch": 1.293078055964654,
+      "grad_norm": 0.4253244698047638,
+      "learning_rate": 7.018797999899142e-06,
+      "loss": 0.4134,
+      "step": 1756
+    },
+    {
+      "epoch": 1.2938144329896908,
+      "grad_norm": 0.42249786853790283,
+      "learning_rate": 7.014877273378557e-06,
+      "loss": 0.4062,
+      "step": 1757
+    },
+    {
+      "epoch": 1.2945508100147276,
+      "grad_norm": 0.42850828170776367,
+      "learning_rate": 7.010955067192991e-06,
+      "loss": 0.4063,
+      "step": 1758
+    },
+    {
+      "epoch": 1.2952871870397644,
+      "grad_norm": 0.4714277386665344,
+      "learning_rate": 7.0070313842227946e-06,
+      "loss": 0.409,
+      "step": 1759
+    },
+    {
+      "epoch": 1.2960235640648012,
+      "grad_norm": 0.3643467426300049,
+      "learning_rate": 7.003106227349399e-06,
+      "loss": 0.4297,
+      "step": 1760
+    },
+    {
+      "epoch": 1.296759941089838,
+      "grad_norm": 0.4631706178188324,
+      "learning_rate": 6.999179599455321e-06,
+      "loss": 0.4398,
+      "step": 1761
+    },
+    {
+      "epoch": 1.2974963181148749,
+      "grad_norm": 0.37905317544937134,
+      "learning_rate": 6.995251503424158e-06,
+      "loss": 0.3931,
+      "step": 1762
+    },
+    {
+      "epoch": 1.2982326951399117,
+      "grad_norm": 0.4009278118610382,
+      "learning_rate": 6.991321942140587e-06,
+      "loss": 0.4023,
+      "step": 1763
+    },
+    {
+      "epoch": 1.2989690721649485,
+      "grad_norm": 0.4024566113948822,
+      "learning_rate": 6.987390918490356e-06,
+      "loss": 0.4188,
+      "step": 1764
+    },
+    {
+      "epoch": 1.2997054491899853,
+      "grad_norm": 0.4234688878059387,
+      "learning_rate": 6.983458435360291e-06,
+      "loss": 0.4308,
+      "step": 1765
+    },
+    {
+      "epoch": 1.3004418262150221,
+      "grad_norm": 0.4323263466358185,
+      "learning_rate": 6.9795244956382904e-06,
+      "loss": 0.3954,
+      "step": 1766
+    },
+    {
+      "epoch": 1.301178203240059,
+      "grad_norm": 0.4084095358848572,
+      "learning_rate": 6.975589102213318e-06,
+      "loss": 0.3908,
+      "step": 1767
+    },
+    {
+      "epoch": 1.3019145802650958,
+      "grad_norm": 0.3725346624851227,
+      "learning_rate": 6.97165225797541e-06,
+      "loss": 0.4085,
+      "step": 1768
+    },
+    {
+      "epoch": 1.3026509572901326,
+      "grad_norm": 0.4671725928783417,
+      "learning_rate": 6.9677139658156656e-06,
+      "loss": 0.4093,
+      "step": 1769
+    },
+    {
+      "epoch": 1.3033873343151694,
+      "grad_norm": 0.42323094606399536,
+      "learning_rate": 6.963774228626246e-06,
+      "loss": 0.371,
+      "step": 1770
+    },
+    {
+      "epoch": 1.3041237113402062,
+      "grad_norm": 0.37016546726226807,
+      "learning_rate": 6.959833049300376e-06,
+      "loss": 0.4009,
+      "step": 1771
+    },
+    {
+      "epoch": 1.304860088365243,
+      "grad_norm": 0.4504646360874176,
+      "learning_rate": 6.955890430732338e-06,
+      "loss": 0.3951,
+      "step": 1772
+    },
+    {
+      "epoch": 1.3055964653902798,
+      "grad_norm": 0.41494229435920715,
+      "learning_rate": 6.9519463758174745e-06,
+      "loss": 0.4112,
+      "step": 1773
+    },
+    {
+      "epoch": 1.3063328424153167,
+      "grad_norm": 0.3828340768814087,
+      "learning_rate": 6.948000887452177e-06,
+      "loss": 0.4081,
+      "step": 1774
+    },
+    {
+      "epoch": 1.3070692194403535,
+      "grad_norm": 0.4123505651950836,
+      "learning_rate": 6.944053968533895e-06,
+      "loss": 0.4267,
+      "step": 1775
+    },
+    {
+      "epoch": 1.3078055964653903,
+      "grad_norm": 0.42603743076324463,
+      "learning_rate": 6.9401056219611255e-06,
+      "loss": 0.3866,
+      "step": 1776
+    },
+    {
+      "epoch": 1.3085419734904271,
+      "grad_norm": 0.48614487051963806,
+      "learning_rate": 6.936155850633417e-06,
+      "loss": 0.4409,
+      "step": 1777
+    },
+    {
+      "epoch": 1.309278350515464,
+      "grad_norm": 0.37237557768821716,
+      "learning_rate": 6.932204657451358e-06,
+      "loss": 0.4096,
+      "step": 1778
+    },
+    {
+      "epoch": 1.3100147275405007,
+      "grad_norm": 0.45322996377944946,
+      "learning_rate": 6.928252045316588e-06,
+      "loss": 0.4214,
+      "step": 1779
+    },
+    {
+      "epoch": 1.3107511045655376,
+      "grad_norm": 0.3440614640712738,
+      "learning_rate": 6.924298017131786e-06,
+      "loss": 0.3917,
+      "step": 1780
+    },
+    {
+      "epoch": 1.3114874815905744,
+      "grad_norm": 0.38658320903778076,
+      "learning_rate": 6.920342575800672e-06,
+      "loss": 0.419,
+      "step": 1781
+    },
+    {
+      "epoch": 1.3122238586156112,
+      "grad_norm": 0.4042324721813202,
+      "learning_rate": 6.916385724227998e-06,
+      "loss": 0.4234,
+      "step": 1782
+    },
+    {
+      "epoch": 1.312960235640648,
+      "grad_norm": 0.41090258955955505,
+      "learning_rate": 6.912427465319561e-06,
+      "loss": 0.4129,
+      "step": 1783
+    },
+    {
+      "epoch": 1.3136966126656848,
+      "grad_norm": 0.378579705953598,
+      "learning_rate": 6.908467801982186e-06,
+      "loss": 0.3884,
+      "step": 1784
+    },
+    {
+      "epoch": 1.3144329896907216,
+      "grad_norm": 0.43072155117988586,
+      "learning_rate": 6.9045067371237285e-06,
+      "loss": 0.4319,
+      "step": 1785
+    },
+    {
+      "epoch": 1.3151693667157585,
+      "grad_norm": 0.42852944135665894,
+      "learning_rate": 6.9005442736530745e-06,
+      "loss": 0.4116,
+      "step": 1786
+    },
+    {
+      "epoch": 1.3159057437407953,
+      "grad_norm": 0.40988779067993164,
+      "learning_rate": 6.8965804144801386e-06,
+      "loss": 0.425,
+      "step": 1787
+    },
+    {
+      "epoch": 1.316642120765832,
+      "grad_norm": 0.4157440960407257,
+      "learning_rate": 6.89261516251586e-06,
+      "loss": 0.4266,
+      "step": 1788
+    },
+    {
+      "epoch": 1.317378497790869,
+      "grad_norm": 0.46512264013290405,
+      "learning_rate": 6.888648520672198e-06,
+      "loss": 0.4201,
+      "step": 1789
+    },
+    {
+      "epoch": 1.3181148748159057,
+      "grad_norm": 0.496389776468277,
+      "learning_rate": 6.8846804918621355e-06,
+      "loss": 0.4487,
+      "step": 1790
+    },
+    {
+      "epoch": 1.3188512518409425,
+      "grad_norm": 0.3850422501564026,
+      "learning_rate": 6.880711078999673e-06,
+      "loss": 0.4016,
+      "step": 1791
+    },
+    {
+      "epoch": 1.3195876288659794,
+      "grad_norm": 0.44369614124298096,
+      "learning_rate": 6.876740284999828e-06,
+      "loss": 0.4188,
+      "step": 1792
+    },
+    {
+      "epoch": 1.3203240058910162,
+      "grad_norm": 0.43615850806236267,
+      "learning_rate": 6.872768112778629e-06,
+      "loss": 0.4164,
+      "step": 1793
+    },
+    {
+      "epoch": 1.321060382916053,
+      "grad_norm": 0.46753209829330444,
+      "learning_rate": 6.868794565253123e-06,
+      "loss": 0.4041,
+      "step": 1794
+    },
+    {
+      "epoch": 1.3217967599410898,
+      "grad_norm": 0.4193538725376129,
+      "learning_rate": 6.864819645341361e-06,
+      "loss": 0.3932,
+      "step": 1795
+    },
+    {
+      "epoch": 1.3225331369661266,
+      "grad_norm": 0.437418669462204,
+      "learning_rate": 6.860843355962403e-06,
+      "loss": 0.3834,
+      "step": 1796
+    },
+    {
+      "epoch": 1.3232695139911634,
+      "grad_norm": 0.4375527799129486,
+      "learning_rate": 6.856865700036317e-06,
+      "loss": 0.4254,
+      "step": 1797
+    },
+    {
+      "epoch": 1.3240058910162003,
+      "grad_norm": 0.4403863251209259,
+      "learning_rate": 6.852886680484175e-06,
+      "loss": 0.4204,
+      "step": 1798
+    },
+    {
+      "epoch": 1.324742268041237,
+      "grad_norm": 0.39266374707221985,
+      "learning_rate": 6.848906300228047e-06,
+      "loss": 0.3803,
+      "step": 1799
+    },
+    {
+      "epoch": 1.3254786450662739,
+      "grad_norm": 0.41610684990882874,
+      "learning_rate": 6.844924562191003e-06,
+      "loss": 0.3783,
+      "step": 1800
+    },
+    {
+      "epoch": 1.3262150220913107,
+      "grad_norm": 0.39313656091690063,
+      "learning_rate": 6.8409414692971125e-06,
+      "loss": 0.3979,
+      "step": 1801
+    },
+    {
+      "epoch": 1.3269513991163475,
+      "grad_norm": 0.3985959589481354,
+      "learning_rate": 6.836957024471439e-06,
+      "loss": 0.4142,
+      "step": 1802
+    },
+    {
+      "epoch": 1.3276877761413843,
+      "grad_norm": 0.38051387667655945,
+      "learning_rate": 6.832971230640037e-06,
+      "loss": 0.4066,
+      "step": 1803
+    },
+    {
+      "epoch": 1.3284241531664212,
+      "grad_norm": 0.4025249481201172,
+      "learning_rate": 6.828984090729954e-06,
+      "loss": 0.3993,
+      "step": 1804
+    },
+    {
+      "epoch": 1.329160530191458,
+      "grad_norm": 0.39893102645874023,
+      "learning_rate": 6.8249956076692235e-06,
+      "loss": 0.4082,
+      "step": 1805
+    },
+    {
+      "epoch": 1.3298969072164948,
+      "grad_norm": 0.3757968246936798,
+      "learning_rate": 6.8210057843868715e-06,
+      "loss": 0.4043,
+      "step": 1806
+    },
+    {
+      "epoch": 1.3306332842415316,
+      "grad_norm": 0.377785861492157,
+      "learning_rate": 6.817014623812898e-06,
+      "loss": 0.3817,
+      "step": 1807
+    },
+    {
+      "epoch": 1.3313696612665684,
+      "grad_norm": 0.39483869075775146,
+      "learning_rate": 6.813022128878292e-06,
+      "loss": 0.4089,
+      "step": 1808
+    },
+    {
+      "epoch": 1.3321060382916052,
+      "grad_norm": 0.4306437373161316,
+      "learning_rate": 6.809028302515024e-06,
+      "loss": 0.3828,
+      "step": 1809
+    },
+    {
+      "epoch": 1.332842415316642,
+      "grad_norm": 0.38789859414100647,
+      "learning_rate": 6.805033147656037e-06,
+      "loss": 0.4043,
+      "step": 1810
+    },
+    {
+      "epoch": 1.3335787923416789,
+      "grad_norm": 0.3801842927932739,
+      "learning_rate": 6.801036667235252e-06,
+      "loss": 0.4094,
+      "step": 1811
+    },
+    {
+      "epoch": 1.3343151693667157,
+      "grad_norm": 0.3741806447505951,
+      "learning_rate": 6.797038864187564e-06,
+      "loss": 0.4316,
+      "step": 1812
+    },
+    {
+      "epoch": 1.3350515463917525,
+      "grad_norm": 0.39872273802757263,
+      "learning_rate": 6.79303974144884e-06,
+      "loss": 0.4268,
+      "step": 1813
+    },
+    {
+      "epoch": 1.3357879234167893,
+      "grad_norm": 0.4184839427471161,
+      "learning_rate": 6.789039301955913e-06,
+      "loss": 0.4332,
+      "step": 1814
+    },
+    {
+      "epoch": 1.3365243004418264,
+      "grad_norm": 0.3965596854686737,
+      "learning_rate": 6.785037548646586e-06,
+      "loss": 0.4486,
+      "step": 1815
+    },
+    {
+      "epoch": 1.3372606774668632,
+      "grad_norm": 0.4066341817378998,
+      "learning_rate": 6.781034484459624e-06,
+      "loss": 0.3996,
+      "step": 1816
+    },
+    {
+      "epoch": 1.3379970544919,
+      "grad_norm": 0.47760289907455444,
+      "learning_rate": 6.777030112334759e-06,
+      "loss": 0.4175,
+      "step": 1817
+    },
+    {
+      "epoch": 1.3387334315169368,
+      "grad_norm": 0.39844220876693726,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.4186,
+      "step": 1818
+    },
+    {
+      "epoch": 1.3394698085419736,
+      "grad_norm": 0.419331431388855,
+      "learning_rate": 6.769017456035033e-06,
+      "loss": 0.4229,
+      "step": 1819
+    },
+    {
+      "epoch": 1.3402061855670104,
+      "grad_norm": 0.41925719380378723,
+      "learning_rate": 6.765009177744425e-06,
+      "loss": 0.4078,
+      "step": 1820
+    },
+    {
+      "epoch": 1.3409425625920472,
+      "grad_norm": 0.39159253239631653,
+      "learning_rate": 6.760999603284413e-06,
+      "loss": 0.394,
+      "step": 1821
+    },
+    {
+      "epoch": 1.341678939617084,
+      "grad_norm": 0.38817527890205383,
+      "learning_rate": 6.756988735599508e-06,
+      "loss": 0.3846,
+      "step": 1822
+    },
+    {
+      "epoch": 1.3424153166421209,
+      "grad_norm": 0.4380631148815155,
+      "learning_rate": 6.752976577635169e-06,
+      "loss": 0.4231,
+      "step": 1823
+    },
+    {
+      "epoch": 1.3431516936671577,
+      "grad_norm": 0.4459487795829773,
+      "learning_rate": 6.748963132337807e-06,
+      "loss": 0.4201,
+      "step": 1824
+    },
+    {
+      "epoch": 1.3438880706921945,
+      "grad_norm": 0.3807379901409149,
+      "learning_rate": 6.7449484026547705e-06,
+      "loss": 0.4175,
+      "step": 1825
+    },
+    {
+      "epoch": 1.3446244477172313,
+      "grad_norm": 0.3830302655696869,
+      "learning_rate": 6.740932391534358e-06,
+      "loss": 0.4129,
+      "step": 1826
+    },
+    {
+      "epoch": 1.3453608247422681,
+      "grad_norm": 0.4267512559890747,
+      "learning_rate": 6.736915101925807e-06,
+      "loss": 0.4108,
+      "step": 1827
+    },
+    {
+      "epoch": 1.346097201767305,
+      "grad_norm": 0.3901468515396118,
+      "learning_rate": 6.732896536779293e-06,
+      "loss": 0.3991,
+      "step": 1828
+    },
+    {
+      "epoch": 1.3468335787923418,
+      "grad_norm": 0.41451314091682434,
+      "learning_rate": 6.728876699045927e-06,
+      "loss": 0.4133,
+      "step": 1829
+    },
+    {
+      "epoch": 1.3475699558173786,
+      "grad_norm": 0.4063207507133484,
+      "learning_rate": 6.7248555916777595e-06,
+      "loss": 0.4133,
+      "step": 1830
+    },
+    {
+      "epoch": 1.3483063328424154,
+      "grad_norm": 0.4275752902030945,
+      "learning_rate": 6.720833217627769e-06,
+      "loss": 0.4337,
+      "step": 1831
+    },
+    {
+      "epoch": 1.3490427098674522,
+      "grad_norm": 0.451399028301239,
+      "learning_rate": 6.716809579849865e-06,
+      "loss": 0.3908,
+      "step": 1832
+    },
+    {
+      "epoch": 1.349779086892489,
+      "grad_norm": 0.43301716446876526,
+      "learning_rate": 6.712784681298885e-06,
+      "loss": 0.4326,
+      "step": 1833
+    },
+    {
+      "epoch": 1.3505154639175259,
+      "grad_norm": 0.4233209192752838,
+      "learning_rate": 6.708758524930594e-06,
+      "loss": 0.4022,
+      "step": 1834
+    },
+    {
+      "epoch": 1.3512518409425627,
+      "grad_norm": 0.43682563304901123,
+      "learning_rate": 6.704731113701679e-06,
+      "loss": 0.4212,
+      "step": 1835
+    },
+    {
+      "epoch": 1.3519882179675995,
+      "grad_norm": 0.4808812439441681,
+      "learning_rate": 6.70070245056975e-06,
+      "loss": 0.4156,
+      "step": 1836
+    },
+    {
+      "epoch": 1.3527245949926363,
+      "grad_norm": 0.40924352407455444,
+      "learning_rate": 6.696672538493334e-06,
+      "loss": 0.3947,
+      "step": 1837
+    },
+    {
+      "epoch": 1.3534609720176731,
+      "grad_norm": 0.3757144808769226,
+      "learning_rate": 6.692641380431879e-06,
+      "loss": 0.4008,
+      "step": 1838
+    },
+    {
+      "epoch": 1.35419734904271,
+      "grad_norm": 0.40984469652175903,
+      "learning_rate": 6.688608979345742e-06,
+      "loss": 0.4306,
+      "step": 1839
+    },
+    {
+      "epoch": 1.3549337260677468,
+      "grad_norm": 0.39980268478393555,
+      "learning_rate": 6.6845753381961995e-06,
+      "loss": 0.4105,
+      "step": 1840
+    },
+    {
+      "epoch": 1.3556701030927836,
+      "grad_norm": 0.38676726818084717,
+      "learning_rate": 6.680540459945435e-06,
+      "loss": 0.3773,
+      "step": 1841
+    },
+    {
+      "epoch": 1.3564064801178204,
+      "grad_norm": 0.4054774045944214,
+      "learning_rate": 6.676504347556541e-06,
+      "loss": 0.4035,
+      "step": 1842
+    },
+    {
+      "epoch": 1.3571428571428572,
+      "grad_norm": 0.42290937900543213,
+      "learning_rate": 6.6724670039935145e-06,
+      "loss": 0.4226,
+      "step": 1843
+    },
+    {
+      "epoch": 1.357879234167894,
+      "grad_norm": 0.41523557901382446,
+      "learning_rate": 6.668428432221262e-06,
+      "loss": 0.4143,
+      "step": 1844
+    },
+    {
+      "epoch": 1.3586156111929308,
+      "grad_norm": 0.40057843923568726,
+      "learning_rate": 6.664388635205587e-06,
+      "loss": 0.4092,
+      "step": 1845
+    },
+    {
+      "epoch": 1.3593519882179677,
+      "grad_norm": 0.43508538603782654,
+      "learning_rate": 6.660347615913194e-06,
+      "loss": 0.3645,
+      "step": 1846
+    },
+    {
+      "epoch": 1.3600883652430045,
+      "grad_norm": 0.4160832464694977,
+      "learning_rate": 6.656305377311686e-06,
+      "loss": 0.3927,
+      "step": 1847
+    },
+    {
+      "epoch": 1.3608247422680413,
+      "grad_norm": 0.40695255994796753,
+      "learning_rate": 6.652261922369562e-06,
+      "loss": 0.4079,
+      "step": 1848
+    },
+    {
+      "epoch": 1.361561119293078,
+      "grad_norm": 0.4502415657043457,
+      "learning_rate": 6.648217254056211e-06,
+      "loss": 0.41,
+      "step": 1849
+    },
+    {
+      "epoch": 1.362297496318115,
+      "grad_norm": 0.4875112473964691,
+      "learning_rate": 6.644171375341915e-06,
+      "loss": 0.4101,
+      "step": 1850
+    },
+    {
+      "epoch": 1.3630338733431517,
+      "grad_norm": 0.4880272150039673,
+      "learning_rate": 6.640124289197845e-06,
+      "loss": 0.4239,
+      "step": 1851
+    },
+    {
+      "epoch": 1.3637702503681886,
+      "grad_norm": 0.44013795256614685,
+      "learning_rate": 6.636075998596063e-06,
+      "loss": 0.3872,
+      "step": 1852
+    },
+    {
+      "epoch": 1.3645066273932254,
+      "grad_norm": 0.43611839413642883,
+      "learning_rate": 6.632026506509507e-06,
+      "loss": 0.4029,
+      "step": 1853
+    },
+    {
+      "epoch": 1.3652430044182622,
+      "grad_norm": 0.4688226878643036,
+      "learning_rate": 6.627975815912002e-06,
+      "loss": 0.4,
+      "step": 1854
+    },
+    {
+      "epoch": 1.365979381443299,
+      "grad_norm": 0.4303816258907318,
+      "learning_rate": 6.623923929778253e-06,
+      "loss": 0.4301,
+      "step": 1855
+    },
+    {
+      "epoch": 1.3667157584683358,
+      "grad_norm": 0.516590416431427,
+      "learning_rate": 6.6198708510838446e-06,
+      "loss": 0.4398,
+      "step": 1856
+    },
+    {
+      "epoch": 1.3674521354933726,
+      "grad_norm": 0.4422401785850525,
+      "learning_rate": 6.615816582805235e-06,
+      "loss": 0.3925,
+      "step": 1857
+    },
+    {
+      "epoch": 1.3681885125184094,
+      "grad_norm": 0.39282315969467163,
+      "learning_rate": 6.611761127919753e-06,
+      "loss": 0.3969,
+      "step": 1858
+    },
+    {
+      "epoch": 1.3689248895434463,
+      "grad_norm": 0.4454099237918854,
+      "learning_rate": 6.607704489405605e-06,
+      "loss": 0.4306,
+      "step": 1859
+    },
+    {
+      "epoch": 1.369661266568483,
+      "grad_norm": 0.45609748363494873,
+      "learning_rate": 6.603646670241863e-06,
+      "loss": 0.3761,
+      "step": 1860
+    },
+    {
+      "epoch": 1.37039764359352,
+      "grad_norm": 0.40115681290626526,
+      "learning_rate": 6.599587673408469e-06,
+      "loss": 0.4091,
+      "step": 1861
+    },
+    {
+      "epoch": 1.3711340206185567,
+      "grad_norm": 0.4406599998474121,
+      "learning_rate": 6.595527501886223e-06,
+      "loss": 0.4157,
+      "step": 1862
+    },
+    {
+      "epoch": 1.3718703976435935,
+      "grad_norm": 0.3757615089416504,
+      "learning_rate": 6.591466158656795e-06,
+      "loss": 0.3968,
+      "step": 1863
+    },
+    {
+      "epoch": 1.3726067746686303,
+      "grad_norm": 0.4482031464576721,
+      "learning_rate": 6.5874036467027135e-06,
+      "loss": 0.4094,
+      "step": 1864
+    },
+    {
+      "epoch": 1.3733431516936672,
+      "grad_norm": 0.4158189594745636,
+      "learning_rate": 6.583339969007364e-06,
+      "loss": 0.4047,
+      "step": 1865
+    },
+    {
+      "epoch": 1.374079528718704,
+      "grad_norm": 0.45515987277030945,
+      "learning_rate": 6.579275128554986e-06,
+      "loss": 0.4326,
+      "step": 1866
+    },
+    {
+      "epoch": 1.3748159057437408,
+      "grad_norm": 0.49017441272735596,
+      "learning_rate": 6.575209128330679e-06,
+      "loss": 0.4162,
+      "step": 1867
+    },
+    {
+      "epoch": 1.3755522827687776,
+      "grad_norm": 0.5305281281471252,
+      "learning_rate": 6.57114197132039e-06,
+      "loss": 0.402,
+      "step": 1868
+    },
+    {
+      "epoch": 1.3762886597938144,
+      "grad_norm": 0.42159077525138855,
+      "learning_rate": 6.567073660510914e-06,
+      "loss": 0.4187,
+      "step": 1869
+    },
+    {
+      "epoch": 1.3770250368188512,
+      "grad_norm": 0.4378039538860321,
+      "learning_rate": 6.563004198889899e-06,
+      "loss": 0.4001,
+      "step": 1870
+    },
+    {
+      "epoch": 1.377761413843888,
+      "grad_norm": 0.45202401280403137,
+      "learning_rate": 6.5589335894458305e-06,
+      "loss": 0.4204,
+      "step": 1871
+    },
+    {
+      "epoch": 1.3784977908689249,
+      "grad_norm": 0.4391757547855377,
+      "learning_rate": 6.554861835168045e-06,
+      "loss": 0.4141,
+      "step": 1872
+    },
+    {
+      "epoch": 1.3792341678939617,
+      "grad_norm": 0.4077272415161133,
+      "learning_rate": 6.550788939046713e-06,
+      "loss": 0.3808,
+      "step": 1873
+    },
+    {
+      "epoch": 1.3799705449189985,
+      "grad_norm": 0.488126665353775,
+      "learning_rate": 6.546714904072848e-06,
+      "loss": 0.4128,
+      "step": 1874
+    },
+    {
+      "epoch": 1.3807069219440353,
+      "grad_norm": 0.43175673484802246,
+      "learning_rate": 6.542639733238297e-06,
+      "loss": 0.416,
+      "step": 1875
+    },
+    {
+      "epoch": 1.3814432989690721,
+      "grad_norm": 0.39616307616233826,
+      "learning_rate": 6.538563429535742e-06,
+      "loss": 0.4148,
+      "step": 1876
+    },
+    {
+      "epoch": 1.382179675994109,
+      "grad_norm": 0.436383455991745,
+      "learning_rate": 6.534485995958699e-06,
+      "loss": 0.4205,
+      "step": 1877
+    },
+    {
+      "epoch": 1.3829160530191458,
+      "grad_norm": 0.5000892281532288,
+      "learning_rate": 6.530407435501513e-06,
+      "loss": 0.3923,
+      "step": 1878
+    },
+    {
+      "epoch": 1.3836524300441826,
+      "grad_norm": 0.4896223545074463,
+      "learning_rate": 6.5263277511593515e-06,
+      "loss": 0.4397,
+      "step": 1879
+    },
+    {
+      "epoch": 1.3843888070692194,
+      "grad_norm": 0.3927074372768402,
+      "learning_rate": 6.522246945928214e-06,
+      "loss": 0.422,
+      "step": 1880
+    },
+    {
+      "epoch": 1.3851251840942562,
+      "grad_norm": 0.4764426350593567,
+      "learning_rate": 6.518165022804921e-06,
+      "loss": 0.4127,
+      "step": 1881
+    },
+    {
+      "epoch": 1.385861561119293,
+      "grad_norm": 0.4523772895336151,
+      "learning_rate": 6.514081984787112e-06,
+      "loss": 0.4211,
+      "step": 1882
+    },
+    {
+      "epoch": 1.3865979381443299,
+      "grad_norm": 0.3522360324859619,
+      "learning_rate": 6.509997834873246e-06,
+      "loss": 0.4101,
+      "step": 1883
+    },
+    {
+      "epoch": 1.3873343151693667,
+      "grad_norm": 0.4114728271961212,
+      "learning_rate": 6.505912576062602e-06,
+      "loss": 0.4017,
+      "step": 1884
+    },
+    {
+      "epoch": 1.3880706921944035,
+      "grad_norm": 0.4329925775527954,
+      "learning_rate": 6.501826211355269e-06,
+      "loss": 0.4095,
+      "step": 1885
+    },
+    {
+      "epoch": 1.3888070692194403,
+      "grad_norm": 0.4298880398273468,
+      "learning_rate": 6.497738743752151e-06,
+      "loss": 0.4246,
+      "step": 1886
+    },
+    {
+      "epoch": 1.3895434462444771,
+      "grad_norm": 0.4396612048149109,
+      "learning_rate": 6.493650176254958e-06,
+      "loss": 0.4231,
+      "step": 1887
+    },
+    {
+      "epoch": 1.390279823269514,
+      "grad_norm": 0.3601161241531372,
+      "learning_rate": 6.4895605118662116e-06,
+      "loss": 0.3881,
+      "step": 1888
+    },
+    {
+      "epoch": 1.3910162002945508,
+      "grad_norm": 0.42812642455101013,
+      "learning_rate": 6.485469753589241e-06,
+      "loss": 0.4094,
+      "step": 1889
+    },
+    {
+      "epoch": 1.3917525773195876,
+      "grad_norm": 0.41203832626342773,
+      "learning_rate": 6.481377904428171e-06,
+      "loss": 0.4305,
+      "step": 1890
+    },
+    {
+      "epoch": 1.3924889543446244,
+      "grad_norm": 0.4372691214084625,
+      "learning_rate": 6.4772849673879335e-06,
+      "loss": 0.4255,
+      "step": 1891
+    },
+    {
+      "epoch": 1.3932253313696612,
+      "grad_norm": 0.47426924109458923,
+      "learning_rate": 6.473190945474258e-06,
+      "loss": 0.4377,
+      "step": 1892
+    },
+    {
+      "epoch": 1.393961708394698,
+      "grad_norm": 0.4277667999267578,
+      "learning_rate": 6.469095841693671e-06,
+      "loss": 0.3954,
+      "step": 1893
+    },
+    {
+      "epoch": 1.3946980854197348,
+      "grad_norm": 0.3962320387363434,
+      "learning_rate": 6.4649996590534915e-06,
+      "loss": 0.3877,
+      "step": 1894
+    },
+    {
+      "epoch": 1.3954344624447717,
+      "grad_norm": 0.4663618206977844,
+      "learning_rate": 6.460902400561835e-06,
+      "loss": 0.4046,
+      "step": 1895
+    },
+    {
+      "epoch": 1.3961708394698085,
+      "grad_norm": 0.4559689164161682,
+      "learning_rate": 6.456804069227601e-06,
+      "loss": 0.4166,
+      "step": 1896
+    },
+    {
+      "epoch": 1.3969072164948453,
+      "grad_norm": 0.3791521489620209,
+      "learning_rate": 6.452704668060481e-06,
+      "loss": 0.4049,
+      "step": 1897
+    },
+    {
+      "epoch": 1.397643593519882,
+      "grad_norm": 0.49199846386909485,
+      "learning_rate": 6.448604200070953e-06,
+      "loss": 0.4207,
+      "step": 1898
+    },
+    {
+      "epoch": 1.398379970544919,
+      "grad_norm": 0.4463452994823456,
+      "learning_rate": 6.444502668270276e-06,
+      "loss": 0.4088,
+      "step": 1899
+    },
+    {
+      "epoch": 1.3991163475699557,
+      "grad_norm": 0.47948887944221497,
+      "learning_rate": 6.440400075670491e-06,
+      "loss": 0.3906,
+      "step": 1900
+    },
+    {
+      "epoch": 1.3998527245949925,
+      "grad_norm": 0.42631468176841736,
+      "learning_rate": 6.4362964252844165e-06,
+      "loss": 0.4024,
+      "step": 1901
+    },
+    {
+      "epoch": 1.4005891016200294,
+      "grad_norm": 0.39506101608276367,
+      "learning_rate": 6.432191720125651e-06,
+      "loss": 0.4013,
+      "step": 1902
+    },
+    {
+      "epoch": 1.4013254786450662,
+      "grad_norm": 0.4285804331302643,
+      "learning_rate": 6.428085963208567e-06,
+      "loss": 0.4148,
+      "step": 1903
+    },
+    {
+      "epoch": 1.402061855670103,
+      "grad_norm": 0.4068101942539215,
+      "learning_rate": 6.423979157548306e-06,
+      "loss": 0.4113,
+      "step": 1904
+    },
+    {
+      "epoch": 1.4027982326951398,
+      "grad_norm": 0.39695197343826294,
+      "learning_rate": 6.419871306160782e-06,
+      "loss": 0.3925,
+      "step": 1905
+    },
+    {
+      "epoch": 1.4035346097201766,
+      "grad_norm": 0.38716959953308105,
+      "learning_rate": 6.415762412062678e-06,
+      "loss": 0.38,
+      "step": 1906
+    },
+    {
+      "epoch": 1.4042709867452134,
+      "grad_norm": 0.41700947284698486,
+      "learning_rate": 6.411652478271444e-06,
+      "loss": 0.3994,
+      "step": 1907
+    },
+    {
+      "epoch": 1.4050073637702503,
+      "grad_norm": 0.4346196949481964,
+      "learning_rate": 6.407541507805286e-06,
+      "loss": 0.4043,
+      "step": 1908
+    },
+    {
+      "epoch": 1.405743740795287,
+      "grad_norm": 0.433353990316391,
+      "learning_rate": 6.403429503683178e-06,
+      "loss": 0.4124,
+      "step": 1909
+    },
+    {
+      "epoch": 1.406480117820324,
+      "grad_norm": 0.4605219066143036,
+      "learning_rate": 6.399316468924856e-06,
+      "loss": 0.4196,
+      "step": 1910
+    },
+    {
+      "epoch": 1.4072164948453607,
+      "grad_norm": 0.4418054223060608,
+      "learning_rate": 6.395202406550803e-06,
+      "loss": 0.4054,
+      "step": 1911
+    },
+    {
+      "epoch": 1.4079528718703975,
+      "grad_norm": 0.4400234818458557,
+      "learning_rate": 6.391087319582264e-06,
+      "loss": 0.4206,
+      "step": 1912
+    },
+    {
+      "epoch": 1.4086892488954343,
+      "grad_norm": 0.40220165252685547,
+      "learning_rate": 6.386971211041235e-06,
+      "loss": 0.4301,
+      "step": 1913
+    },
+    {
+      "epoch": 1.4094256259204712,
+      "grad_norm": 0.41866376996040344,
+      "learning_rate": 6.382854083950462e-06,
+      "loss": 0.3882,
+      "step": 1914
+    },
+    {
+      "epoch": 1.410162002945508,
+      "grad_norm": 0.4194769561290741,
+      "learning_rate": 6.378735941333437e-06,
+      "loss": 0.4037,
+      "step": 1915
+    },
+    {
+      "epoch": 1.4108983799705448,
+      "grad_norm": 0.41453391313552856,
+      "learning_rate": 6.374616786214402e-06,
+      "loss": 0.4195,
+      "step": 1916
+    },
+    {
+      "epoch": 1.4116347569955816,
+      "grad_norm": 0.4037197232246399,
+      "learning_rate": 6.370496621618338e-06,
+      "loss": 0.3969,
+      "step": 1917
+    },
+    {
+      "epoch": 1.4123711340206184,
+      "grad_norm": 0.4220995008945465,
+      "learning_rate": 6.366375450570971e-06,
+      "loss": 0.4141,
+      "step": 1918
+    },
+    {
+      "epoch": 1.4131075110456552,
+      "grad_norm": 0.45782655477523804,
+      "learning_rate": 6.362253276098762e-06,
+      "loss": 0.3984,
+      "step": 1919
+    },
+    {
+      "epoch": 1.413843888070692,
+      "grad_norm": 0.40615370869636536,
+      "learning_rate": 6.358130101228914e-06,
+      "loss": 0.4279,
+      "step": 1920
+    },
+    {
+      "epoch": 1.414580265095729,
+      "grad_norm": 0.42293697595596313,
+      "learning_rate": 6.35400592898936e-06,
+      "loss": 0.4001,
+      "step": 1921
+    },
+    {
+      "epoch": 1.415316642120766,
+      "grad_norm": 0.4643053710460663,
+      "learning_rate": 6.34988076240877e-06,
+      "loss": 0.4087,
+      "step": 1922
+    },
+    {
+      "epoch": 1.4160530191458027,
+      "grad_norm": 0.42279914021492004,
+      "learning_rate": 6.345754604516539e-06,
+      "loss": 0.3953,
+      "step": 1923
+    },
+    {
+      "epoch": 1.4167893961708395,
+      "grad_norm": 0.4467233121395111,
+      "learning_rate": 6.341627458342794e-06,
+      "loss": 0.4182,
+      "step": 1924
+    },
+    {
+      "epoch": 1.4175257731958764,
+      "grad_norm": 0.4454042911529541,
+      "learning_rate": 6.337499326918386e-06,
+      "loss": 0.4092,
+      "step": 1925
+    },
+    {
+      "epoch": 1.4182621502209132,
+      "grad_norm": 0.414069265127182,
+      "learning_rate": 6.33337021327489e-06,
+      "loss": 0.4254,
+      "step": 1926
+    },
+    {
+      "epoch": 1.41899852724595,
+      "grad_norm": 0.4927258789539337,
+      "learning_rate": 6.329240120444602e-06,
+      "loss": 0.4215,
+      "step": 1927
+    },
+    {
+      "epoch": 1.4197349042709868,
+      "grad_norm": 0.4353216588497162,
+      "learning_rate": 6.325109051460538e-06,
+      "loss": 0.4077,
+      "step": 1928
+    },
+    {
+      "epoch": 1.4204712812960236,
+      "grad_norm": 0.4361936151981354,
+      "learning_rate": 6.3209770093564315e-06,
+      "loss": 0.439,
+      "step": 1929
+    },
+    {
+      "epoch": 1.4212076583210604,
+      "grad_norm": 0.4463266432285309,
+      "learning_rate": 6.316843997166726e-06,
+      "loss": 0.4217,
+      "step": 1930
+    },
+    {
+      "epoch": 1.4219440353460973,
+      "grad_norm": 0.4338516294956207,
+      "learning_rate": 6.312710017926582e-06,
+      "loss": 0.4278,
+      "step": 1931
+    },
+    {
+      "epoch": 1.422680412371134,
+      "grad_norm": 0.48313939571380615,
+      "learning_rate": 6.3085750746718725e-06,
+      "loss": 0.3838,
+      "step": 1932
+    },
+    {
+      "epoch": 1.423416789396171,
+      "grad_norm": 0.5277740955352783,
+      "learning_rate": 6.30443917043917e-06,
+      "loss": 0.4097,
+      "step": 1933
+    },
+    {
+      "epoch": 1.4241531664212077,
+      "grad_norm": 0.47533461451530457,
+      "learning_rate": 6.30030230826576e-06,
+      "loss": 0.3995,
+      "step": 1934
+    },
+    {
+      "epoch": 1.4248895434462445,
+      "grad_norm": 0.4214441776275635,
+      "learning_rate": 6.296164491189628e-06,
+      "loss": 0.4226,
+      "step": 1935
+    },
+    {
+      "epoch": 1.4256259204712813,
+      "grad_norm": 0.4258248507976532,
+      "learning_rate": 6.292025722249463e-06,
+      "loss": 0.3888,
+      "step": 1936
+    },
+    {
+      "epoch": 1.4263622974963182,
+      "grad_norm": 0.4134536683559418,
+      "learning_rate": 6.287886004484651e-06,
+      "loss": 0.3852,
+      "step": 1937
+    },
+    {
+      "epoch": 1.427098674521355,
+      "grad_norm": 0.47609132528305054,
+      "learning_rate": 6.283745340935277e-06,
+      "loss": 0.3968,
+      "step": 1938
+    },
+    {
+      "epoch": 1.4278350515463918,
+      "grad_norm": 0.44796431064605713,
+      "learning_rate": 6.279603734642117e-06,
+      "loss": 0.4421,
+      "step": 1939
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.404325395822525,
+      "learning_rate": 6.275461188646641e-06,
+      "loss": 0.4123,
+      "step": 1940
+    },
+    {
+      "epoch": 1.4293078055964654,
+      "grad_norm": 0.41975611448287964,
+      "learning_rate": 6.271317705991014e-06,
+      "loss": 0.432,
+      "step": 1941
+    },
+    {
+      "epoch": 1.4300441826215022,
+      "grad_norm": 0.46170738339424133,
+      "learning_rate": 6.267173289718079e-06,
+      "loss": 0.4272,
+      "step": 1942
+    },
+    {
+      "epoch": 1.430780559646539,
+      "grad_norm": 0.388175904750824,
+      "learning_rate": 6.263027942871375e-06,
+      "loss": 0.3877,
+      "step": 1943
+    },
+    {
+      "epoch": 1.4315169366715759,
+      "grad_norm": 0.459796667098999,
+      "learning_rate": 6.258881668495116e-06,
+      "loss": 0.3874,
+      "step": 1944
+    },
+    {
+      "epoch": 1.4322533136966127,
+      "grad_norm": 0.4751020073890686,
+      "learning_rate": 6.2547344696342015e-06,
+      "loss": 0.393,
+      "step": 1945
+    },
+    {
+      "epoch": 1.4329896907216495,
+      "grad_norm": 0.3899977207183838,
+      "learning_rate": 6.250586349334209e-06,
+      "loss": 0.3918,
+      "step": 1946
+    },
+    {
+      "epoch": 1.4337260677466863,
+      "grad_norm": 0.43887919187545776,
+      "learning_rate": 6.246437310641395e-06,
+      "loss": 0.43,
+      "step": 1947
+    },
+    {
+      "epoch": 1.4344624447717231,
+      "grad_norm": 0.4462708830833435,
+      "learning_rate": 6.242287356602684e-06,
+      "loss": 0.3989,
+      "step": 1948
+    },
+    {
+      "epoch": 1.43519882179676,
+      "grad_norm": 0.4143645167350769,
+      "learning_rate": 6.238136490265681e-06,
+      "loss": 0.4257,
+      "step": 1949
+    },
+    {
+      "epoch": 1.4359351988217968,
+      "grad_norm": 0.4817536175251007,
+      "learning_rate": 6.2339847146786515e-06,
+      "loss": 0.3969,
+      "step": 1950
+    },
+    {
+      "epoch": 1.4366715758468336,
+      "grad_norm": 0.40567225217819214,
+      "learning_rate": 6.22983203289054e-06,
+      "loss": 0.4277,
+      "step": 1951
+    },
+    {
+      "epoch": 1.4374079528718704,
+      "grad_norm": 0.4788074791431427,
+      "learning_rate": 6.225678447950947e-06,
+      "loss": 0.4243,
+      "step": 1952
+    },
+    {
+      "epoch": 1.4381443298969072,
+      "grad_norm": 0.38317757844924927,
+      "learning_rate": 6.2215239629101385e-06,
+      "loss": 0.4197,
+      "step": 1953
+    },
+    {
+      "epoch": 1.438880706921944,
+      "grad_norm": 0.40520721673965454,
+      "learning_rate": 6.217368580819049e-06,
+      "loss": 0.4112,
+      "step": 1954
+    },
+    {
+      "epoch": 1.4396170839469808,
+      "grad_norm": 0.4457045793533325,
+      "learning_rate": 6.213212304729259e-06,
+      "loss": 0.4104,
+      "step": 1955
+    },
+    {
+      "epoch": 1.4403534609720177,
+      "grad_norm": 0.42782163619995117,
+      "learning_rate": 6.209055137693014e-06,
+      "loss": 0.4247,
+      "step": 1956
+    },
+    {
+      "epoch": 1.4410898379970545,
+      "grad_norm": 0.4254503548145294,
+      "learning_rate": 6.204897082763213e-06,
+      "loss": 0.3672,
+      "step": 1957
+    },
+    {
+      "epoch": 1.4418262150220913,
+      "grad_norm": 0.3664149343967438,
+      "learning_rate": 6.200738142993406e-06,
+      "loss": 0.3824,
+      "step": 1958
+    },
+    {
+      "epoch": 1.4425625920471281,
+      "grad_norm": 0.41897517442703247,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 0.3947,
+      "step": 1959
+    },
+    {
+      "epoch": 1.443298969072165,
+      "grad_norm": 0.3825157582759857,
+      "learning_rate": 6.1924176211512145e-06,
+      "loss": 0.3827,
+      "step": 1960
+    },
+    {
+      "epoch": 1.4440353460972017,
+      "grad_norm": 0.42175137996673584,
+      "learning_rate": 6.1882560451891715e-06,
+      "loss": 0.4241,
+      "step": 1961
+    },
+    {
+      "epoch": 1.4447717231222386,
+      "grad_norm": 0.40888282656669617,
+      "learning_rate": 6.1840935966077985e-06,
+      "loss": 0.4041,
+      "step": 1962
+    },
+    {
+      "epoch": 1.4455081001472754,
+      "grad_norm": 0.42432424426078796,
+      "learning_rate": 6.179930278463868e-06,
+      "loss": 0.4155,
+      "step": 1963
+    },
+    {
+      "epoch": 1.4462444771723122,
+      "grad_norm": 0.4558820128440857,
+      "learning_rate": 6.175766093814798e-06,
+      "loss": 0.4044,
+      "step": 1964
+    },
+    {
+      "epoch": 1.446980854197349,
+      "grad_norm": 0.506184458732605,
+      "learning_rate": 6.1716010457186395e-06,
+      "loss": 0.4198,
+      "step": 1965
+    },
+    {
+      "epoch": 1.4477172312223858,
+      "grad_norm": 0.38299667835235596,
+      "learning_rate": 6.167435137234078e-06,
+      "loss": 0.4086,
+      "step": 1966
+    },
+    {
+      "epoch": 1.4484536082474226,
+      "grad_norm": 0.43511974811553955,
+      "learning_rate": 6.16326837142043e-06,
+      "loss": 0.4063,
+      "step": 1967
+    },
+    {
+      "epoch": 1.4491899852724595,
+      "grad_norm": 0.48710158467292786,
+      "learning_rate": 6.1591007513376425e-06,
+      "loss": 0.4157,
+      "step": 1968
+    },
+    {
+      "epoch": 1.4499263622974963,
+      "grad_norm": 0.41424429416656494,
+      "learning_rate": 6.15493228004629e-06,
+      "loss": 0.397,
+      "step": 1969
+    },
+    {
+      "epoch": 1.450662739322533,
+      "grad_norm": 0.4269460439682007,
+      "learning_rate": 6.1507629606075724e-06,
+      "loss": 0.4171,
+      "step": 1970
+    },
+    {
+      "epoch": 1.45139911634757,
+      "grad_norm": 0.4956137239933014,
+      "learning_rate": 6.14659279608331e-06,
+      "loss": 0.4108,
+      "step": 1971
+    },
+    {
+      "epoch": 1.4521354933726067,
+      "grad_norm": 0.4469659626483917,
+      "learning_rate": 6.142421789535948e-06,
+      "loss": 0.4267,
+      "step": 1972
+    },
+    {
+      "epoch": 1.4528718703976435,
+      "grad_norm": 0.3990239202976227,
+      "learning_rate": 6.138249944028547e-06,
+      "loss": 0.4151,
+      "step": 1973
+    },
+    {
+      "epoch": 1.4536082474226804,
+      "grad_norm": 0.4268157184123993,
+      "learning_rate": 6.134077262624783e-06,
+      "loss": 0.42,
+      "step": 1974
+    },
+    {
+      "epoch": 1.4543446244477172,
+      "grad_norm": 0.45455217361450195,
+      "learning_rate": 6.129903748388948e-06,
+      "loss": 0.3977,
+      "step": 1975
+    },
+    {
+      "epoch": 1.455081001472754,
+      "grad_norm": 0.3931489884853363,
+      "learning_rate": 6.125729404385946e-06,
+      "loss": 0.4218,
+      "step": 1976
+    },
+    {
+      "epoch": 1.4558173784977908,
+      "grad_norm": 0.4898168444633484,
+      "learning_rate": 6.121554233681286e-06,
+      "loss": 0.3984,
+      "step": 1977
+    },
+    {
+      "epoch": 1.4565537555228276,
+      "grad_norm": 0.44921234250068665,
+      "learning_rate": 6.11737823934109e-06,
+      "loss": 0.4049,
+      "step": 1978
+    },
+    {
+      "epoch": 1.4572901325478644,
+      "grad_norm": 0.4154300093650818,
+      "learning_rate": 6.11320142443208e-06,
+      "loss": 0.4296,
+      "step": 1979
+    },
+    {
+      "epoch": 1.4580265095729013,
+      "grad_norm": 0.46535080671310425,
+      "learning_rate": 6.109023792021586e-06,
+      "loss": 0.4384,
+      "step": 1980
+    },
+    {
+      "epoch": 1.458762886597938,
+      "grad_norm": 0.42891719937324524,
+      "learning_rate": 6.1048453451775305e-06,
+      "loss": 0.4186,
+      "step": 1981
+    },
+    {
+      "epoch": 1.4594992636229749,
+      "grad_norm": 0.4581204056739807,
+      "learning_rate": 6.100666086968441e-06,
+      "loss": 0.3773,
+      "step": 1982
+    },
+    {
+      "epoch": 1.4602356406480117,
+      "grad_norm": 0.3952482342720032,
+      "learning_rate": 6.09648602046344e-06,
+      "loss": 0.4141,
+      "step": 1983
+    },
+    {
+      "epoch": 1.4609720176730487,
+      "grad_norm": 0.4057900905609131,
+      "learning_rate": 6.0923051487322385e-06,
+      "loss": 0.4384,
+      "step": 1984
+    },
+    {
+      "epoch": 1.4617083946980856,
+      "grad_norm": 0.4145631194114685,
+      "learning_rate": 6.088123474845144e-06,
+      "loss": 0.3874,
+      "step": 1985
+    },
+    {
+      "epoch": 1.4624447717231224,
+      "grad_norm": 0.3992542624473572,
+      "learning_rate": 6.0839410018730515e-06,
+      "loss": 0.414,
+      "step": 1986
+    },
+    {
+      "epoch": 1.4631811487481592,
+      "grad_norm": 0.43685290217399597,
+      "learning_rate": 6.079757732887444e-06,
+      "loss": 0.3838,
+      "step": 1987
+    },
+    {
+      "epoch": 1.463917525773196,
+      "grad_norm": 0.3933916389942169,
+      "learning_rate": 6.075573670960385e-06,
+      "loss": 0.3845,
+      "step": 1988
+    },
+    {
+      "epoch": 1.4646539027982328,
+      "grad_norm": 0.4441211223602295,
+      "learning_rate": 6.071388819164525e-06,
+      "loss": 0.4222,
+      "step": 1989
+    },
+    {
+      "epoch": 1.4653902798232696,
+      "grad_norm": 0.4381551444530487,
+      "learning_rate": 6.067203180573094e-06,
+      "loss": 0.3978,
+      "step": 1990
+    },
+    {
+      "epoch": 1.4661266568483065,
+      "grad_norm": 0.40585675835609436,
+      "learning_rate": 6.063016758259896e-06,
+      "loss": 0.4072,
+      "step": 1991
+    },
+    {
+      "epoch": 1.4668630338733433,
+      "grad_norm": 0.358671098947525,
+      "learning_rate": 6.058829555299314e-06,
+      "loss": 0.4078,
+      "step": 1992
+    },
+    {
+      "epoch": 1.46759941089838,
+      "grad_norm": 0.39564085006713867,
+      "learning_rate": 6.054641574766304e-06,
+      "loss": 0.4084,
+      "step": 1993
+    },
+    {
+      "epoch": 1.468335787923417,
+      "grad_norm": 0.40318918228149414,
+      "learning_rate": 6.05045281973639e-06,
+      "loss": 0.3892,
+      "step": 1994
+    },
+    {
+      "epoch": 1.4690721649484537,
+      "grad_norm": 0.4056055247783661,
+      "learning_rate": 6.04626329328567e-06,
+      "loss": 0.4119,
+      "step": 1995
+    },
+    {
+      "epoch": 1.4698085419734905,
+      "grad_norm": 0.40646982192993164,
+      "learning_rate": 6.042072998490805e-06,
+      "loss": 0.4069,
+      "step": 1996
+    },
+    {
+      "epoch": 1.4705449189985274,
+      "grad_norm": 0.396869033575058,
+      "learning_rate": 6.0378819384290185e-06,
+      "loss": 0.4212,
+      "step": 1997
+    },
+    {
+      "epoch": 1.4712812960235642,
+      "grad_norm": 0.46342048048973083,
+      "learning_rate": 6.033690116178101e-06,
+      "loss": 0.401,
+      "step": 1998
+    },
+    {
+      "epoch": 1.472017673048601,
+      "grad_norm": 0.34812402725219727,
+      "learning_rate": 6.0294975348163985e-06,
+      "loss": 0.3684,
+      "step": 1999
+    },
+    {
+      "epoch": 1.4727540500736378,
+      "grad_norm": 0.4260614812374115,
+      "learning_rate": 6.025304197422819e-06,
+      "loss": 0.3867,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4734904270986746,
+      "grad_norm": 0.4525951147079468,
+      "learning_rate": 6.0211101070768184e-06,
+      "loss": 0.3997,
+      "step": 2001
+    },
+    {
+      "epoch": 1.4742268041237114,
+      "grad_norm": 0.4634397625923157,
+      "learning_rate": 6.016915266858413e-06,
+      "loss": 0.4346,
+      "step": 2002
+    },
+    {
+      "epoch": 1.4749631811487482,
+      "grad_norm": 0.40612179040908813,
+      "learning_rate": 6.0127196798481645e-06,
+      "loss": 0.3975,
+      "step": 2003
+    },
+    {
+      "epoch": 1.475699558173785,
+      "grad_norm": 0.4513106048107147,
+      "learning_rate": 6.008523349127188e-06,
+      "loss": 0.3806,
+      "step": 2004
+    },
+    {
+      "epoch": 1.4764359351988219,
+      "grad_norm": 0.5352683067321777,
+      "learning_rate": 6.004326277777141e-06,
+      "loss": 0.4328,
+      "step": 2005
+    },
+    {
+      "epoch": 1.4771723122238587,
+      "grad_norm": 0.4118804931640625,
+      "learning_rate": 6.000128468880223e-06,
+      "loss": 0.4485,
+      "step": 2006
+    },
+    {
+      "epoch": 1.4779086892488955,
+      "grad_norm": 0.4650607109069824,
+      "learning_rate": 5.995929925519181e-06,
+      "loss": 0.4315,
+      "step": 2007
+    },
+    {
+      "epoch": 1.4786450662739323,
+      "grad_norm": 0.5012993216514587,
+      "learning_rate": 5.991730650777297e-06,
+      "loss": 0.3924,
+      "step": 2008
+    },
+    {
+      "epoch": 1.4793814432989691,
+      "grad_norm": 0.4178714156150818,
+      "learning_rate": 5.987530647738394e-06,
+      "loss": 0.42,
+      "step": 2009
+    },
+    {
+      "epoch": 1.480117820324006,
+      "grad_norm": 0.4390731453895569,
+      "learning_rate": 5.983329919486824e-06,
+      "loss": 0.4017,
+      "step": 2010
+    },
+    {
+      "epoch": 1.4808541973490428,
+      "grad_norm": 0.4657975137233734,
+      "learning_rate": 5.9791284691074765e-06,
+      "loss": 0.3949,
+      "step": 2011
+    },
+    {
+      "epoch": 1.4815905743740796,
+      "grad_norm": 0.4687153100967407,
+      "learning_rate": 5.974926299685772e-06,
+      "loss": 0.3857,
+      "step": 2012
+    },
+    {
+      "epoch": 1.4823269513991164,
+      "grad_norm": 0.41643956303596497,
+      "learning_rate": 5.970723414307652e-06,
+      "loss": 0.4026,
+      "step": 2013
+    },
+    {
+      "epoch": 1.4830633284241532,
+      "grad_norm": 0.4140578806400299,
+      "learning_rate": 5.966519816059591e-06,
+      "loss": 0.3904,
+      "step": 2014
+    },
+    {
+      "epoch": 1.48379970544919,
+      "grad_norm": 0.42964425683021545,
+      "learning_rate": 5.962315508028584e-06,
+      "loss": 0.3904,
+      "step": 2015
+    },
+    {
+      "epoch": 1.4845360824742269,
+      "grad_norm": 0.4009730815887451,
+      "learning_rate": 5.958110493302148e-06,
+      "loss": 0.4047,
+      "step": 2016
+    },
+    {
+      "epoch": 1.4852724594992637,
+      "grad_norm": 0.41644203662872314,
+      "learning_rate": 5.95390477496832e-06,
+      "loss": 0.4019,
+      "step": 2017
+    },
+    {
+      "epoch": 1.4860088365243005,
+      "grad_norm": 0.43529626727104187,
+      "learning_rate": 5.94969835611565e-06,
+      "loss": 0.4447,
+      "step": 2018
+    },
+    {
+      "epoch": 1.4867452135493373,
+      "grad_norm": 0.4235115051269531,
+      "learning_rate": 5.945491239833206e-06,
+      "loss": 0.4281,
+      "step": 2019
+    },
+    {
+      "epoch": 1.4874815905743741,
+      "grad_norm": 0.441995769739151,
+      "learning_rate": 5.941283429210568e-06,
+      "loss": 0.3974,
+      "step": 2020
+    },
+    {
+      "epoch": 1.488217967599411,
+      "grad_norm": 0.4208744764328003,
+      "learning_rate": 5.937074927337824e-06,
+      "loss": 0.3994,
+      "step": 2021
+    },
+    {
+      "epoch": 1.4889543446244478,
+      "grad_norm": 0.3799182176589966,
+      "learning_rate": 5.932865737305571e-06,
+      "loss": 0.4188,
+      "step": 2022
+    },
+    {
+      "epoch": 1.4896907216494846,
+      "grad_norm": 0.4479272961616516,
+      "learning_rate": 5.928655862204911e-06,
+      "loss": 0.4072,
+      "step": 2023
+    },
+    {
+      "epoch": 1.4904270986745214,
+      "grad_norm": 0.4461783766746521,
+      "learning_rate": 5.924445305127448e-06,
+      "loss": 0.4119,
+      "step": 2024
+    },
+    {
+      "epoch": 1.4911634756995582,
+      "grad_norm": 0.40956565737724304,
+      "learning_rate": 5.9202340691652895e-06,
+      "loss": 0.4073,
+      "step": 2025
+    },
+    {
+      "epoch": 1.491899852724595,
+      "grad_norm": 0.3731072247028351,
+      "learning_rate": 5.916022157411038e-06,
+      "loss": 0.4237,
+      "step": 2026
+    },
+    {
+      "epoch": 1.4926362297496318,
+      "grad_norm": 0.42282289266586304,
+      "learning_rate": 5.911809572957796e-06,
+      "loss": 0.415,
+      "step": 2027
+    },
+    {
+      "epoch": 1.4933726067746687,
+      "grad_norm": 0.41434556245803833,
+      "learning_rate": 5.907596318899157e-06,
+      "loss": 0.4238,
+      "step": 2028
+    },
+    {
+      "epoch": 1.4941089837997055,
+      "grad_norm": 0.4469594657421112,
+      "learning_rate": 5.9033823983292095e-06,
+      "loss": 0.4217,
+      "step": 2029
+    },
+    {
+      "epoch": 1.4948453608247423,
+      "grad_norm": 0.4168426990509033,
+      "learning_rate": 5.899167814342527e-06,
+      "loss": 0.4015,
+      "step": 2030
+    },
+    {
+      "epoch": 1.495581737849779,
+      "grad_norm": 0.3973442018032074,
+      "learning_rate": 5.8949525700341735e-06,
+      "loss": 0.3986,
+      "step": 2031
+    },
+    {
+      "epoch": 1.496318114874816,
+      "grad_norm": 0.5031341314315796,
+      "learning_rate": 5.890736668499696e-06,
+      "loss": 0.4217,
+      "step": 2032
+    },
+    {
+      "epoch": 1.4970544918998527,
+      "grad_norm": 0.42431867122650146,
+      "learning_rate": 5.886520112835128e-06,
+      "loss": 0.3919,
+      "step": 2033
+    },
+    {
+      "epoch": 1.4977908689248896,
+      "grad_norm": 0.4433683454990387,
+      "learning_rate": 5.8823029061369785e-06,
+      "loss": 0.393,
+      "step": 2034
+    },
+    {
+      "epoch": 1.4985272459499264,
+      "grad_norm": 0.3885188400745392,
+      "learning_rate": 5.878085051502236e-06,
+      "loss": 0.3966,
+      "step": 2035
+    },
+    {
+      "epoch": 1.4992636229749632,
+      "grad_norm": 0.4179115295410156,
+      "learning_rate": 5.873866552028367e-06,
+      "loss": 0.39,
+      "step": 2036
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.409787118434906,
+      "learning_rate": 5.86964741081331e-06,
+      "loss": 0.4102,
+      "step": 2037
+    },
+    {
+      "epoch": 1.5007363770250368,
+      "grad_norm": 0.4352511465549469,
+      "learning_rate": 5.865427630955475e-06,
+      "loss": 0.4002,
+      "step": 2038
+    },
+    {
+      "epoch": 1.5014727540500736,
+      "grad_norm": 0.47778865694999695,
+      "learning_rate": 5.861207215553739e-06,
+      "loss": 0.3996,
+      "step": 2039
+    },
+    {
+      "epoch": 1.5022091310751104,
+      "grad_norm": 0.3906075954437256,
+      "learning_rate": 5.856986167707448e-06,
+      "loss": 0.3844,
+      "step": 2040
+    },
+    {
+      "epoch": 1.5029455081001473,
+      "grad_norm": 0.4253922402858734,
+      "learning_rate": 5.852764490516414e-06,
+      "loss": 0.4071,
+      "step": 2041
+    },
+    {
+      "epoch": 1.503681885125184,
+      "grad_norm": 0.39142486453056335,
+      "learning_rate": 5.8485421870809076e-06,
+      "loss": 0.3863,
+      "step": 2042
+    },
+    {
+      "epoch": 1.504418262150221,
+      "grad_norm": 0.4321049153804779,
+      "learning_rate": 5.8443192605016604e-06,
+      "loss": 0.4135,
+      "step": 2043
+    },
+    {
+      "epoch": 1.5051546391752577,
+      "grad_norm": 0.4090045988559723,
+      "learning_rate": 5.840095713879864e-06,
+      "loss": 0.4048,
+      "step": 2044
+    },
+    {
+      "epoch": 1.5058910162002945,
+      "grad_norm": 0.4378100037574768,
+      "learning_rate": 5.83587155031716e-06,
+      "loss": 0.4017,
+      "step": 2045
+    },
+    {
+      "epoch": 1.5066273932253313,
+      "grad_norm": 0.42427340149879456,
+      "learning_rate": 5.831646772915651e-06,
+      "loss": 0.4201,
+      "step": 2046
+    },
+    {
+      "epoch": 1.5073637702503682,
+      "grad_norm": 0.4345465898513794,
+      "learning_rate": 5.827421384777883e-06,
+      "loss": 0.3987,
+      "step": 2047
+    },
+    {
+      "epoch": 1.508100147275405,
+      "grad_norm": 0.4228680729866028,
+      "learning_rate": 5.823195389006853e-06,
+      "loss": 0.4231,
+      "step": 2048
+    },
+    {
+      "epoch": 1.5088365243004418,
+      "grad_norm": 0.3961130976676941,
+      "learning_rate": 5.818968788706006e-06,
+      "loss": 0.4138,
+      "step": 2049
+    },
+    {
+      "epoch": 1.5095729013254786,
+      "grad_norm": 0.3891131579875946,
+      "learning_rate": 5.814741586979228e-06,
+      "loss": 0.3979,
+      "step": 2050
+    },
+    {
+      "epoch": 1.5103092783505154,
+      "grad_norm": 0.44312554597854614,
+      "learning_rate": 5.810513786930849e-06,
+      "loss": 0.4171,
+      "step": 2051
+    },
+    {
+      "epoch": 1.5110456553755522,
+      "grad_norm": 0.36081069707870483,
+      "learning_rate": 5.806285391665639e-06,
+      "loss": 0.3745,
+      "step": 2052
+    },
+    {
+      "epoch": 1.511782032400589,
+      "grad_norm": 0.43281090259552,
+      "learning_rate": 5.8020564042888015e-06,
+      "loss": 0.408,
+      "step": 2053
+    },
+    {
+      "epoch": 1.5125184094256259,
+      "grad_norm": 0.4164915680885315,
+      "learning_rate": 5.7978268279059795e-06,
+      "loss": 0.3897,
+      "step": 2054
+    },
+    {
+      "epoch": 1.5132547864506627,
+      "grad_norm": 0.41813069581985474,
+      "learning_rate": 5.7935966656232434e-06,
+      "loss": 0.4268,
+      "step": 2055
+    },
+    {
+      "epoch": 1.5139911634756995,
+      "grad_norm": 0.4125601053237915,
+      "learning_rate": 5.789365920547098e-06,
+      "loss": 0.3895,
+      "step": 2056
+    },
+    {
+      "epoch": 1.5147275405007363,
+      "grad_norm": 0.4070327579975128,
+      "learning_rate": 5.785134595784473e-06,
+      "loss": 0.3862,
+      "step": 2057
+    },
+    {
+      "epoch": 1.5154639175257731,
+      "grad_norm": 0.40182122588157654,
+      "learning_rate": 5.780902694442727e-06,
+      "loss": 0.4227,
+      "step": 2058
+    },
+    {
+      "epoch": 1.51620029455081,
+      "grad_norm": 0.40783995389938354,
+      "learning_rate": 5.776670219629643e-06,
+      "loss": 0.3972,
+      "step": 2059
+    },
+    {
+      "epoch": 1.5169366715758468,
+      "grad_norm": 0.4468885064125061,
+      "learning_rate": 5.772437174453418e-06,
+      "loss": 0.384,
+      "step": 2060
+    },
+    {
+      "epoch": 1.5176730486008836,
+      "grad_norm": 0.4095504581928253,
+      "learning_rate": 5.768203562022674e-06,
+      "loss": 0.4271,
+      "step": 2061
+    },
+    {
+      "epoch": 1.5184094256259204,
+      "grad_norm": 0.4125244915485382,
+      "learning_rate": 5.7639693854464495e-06,
+      "loss": 0.429,
+      "step": 2062
+    },
+    {
+      "epoch": 1.5191458026509572,
+      "grad_norm": 0.3764050602912903,
+      "learning_rate": 5.7597346478341946e-06,
+      "loss": 0.4135,
+      "step": 2063
+    },
+    {
+      "epoch": 1.519882179675994,
+      "grad_norm": 0.3860231637954712,
+      "learning_rate": 5.755499352295772e-06,
+      "loss": 0.387,
+      "step": 2064
+    },
+    {
+      "epoch": 1.5206185567010309,
+      "grad_norm": 0.42220139503479004,
+      "learning_rate": 5.751263501941454e-06,
+      "loss": 0.385,
+      "step": 2065
+    },
+    {
+      "epoch": 1.5213549337260677,
+      "grad_norm": 0.4375130236148834,
+      "learning_rate": 5.747027099881925e-06,
+      "loss": 0.4225,
+      "step": 2066
+    },
+    {
+      "epoch": 1.5220913107511045,
+      "grad_norm": 0.3937590420246124,
+      "learning_rate": 5.742790149228268e-06,
+      "loss": 0.4172,
+      "step": 2067
+    },
+    {
+      "epoch": 1.5228276877761413,
+      "grad_norm": 0.49965283274650574,
+      "learning_rate": 5.738552653091971e-06,
+      "loss": 0.3976,
+      "step": 2068
+    },
+    {
+      "epoch": 1.5235640648011781,
+      "grad_norm": 0.4149169623851776,
+      "learning_rate": 5.734314614584924e-06,
+      "loss": 0.4299,
+      "step": 2069
+    },
+    {
+      "epoch": 1.524300441826215,
+      "grad_norm": 0.41810843348503113,
+      "learning_rate": 5.730076036819414e-06,
+      "loss": 0.3996,
+      "step": 2070
+    },
+    {
+      "epoch": 1.5250368188512518,
+      "grad_norm": 0.42622384428977966,
+      "learning_rate": 5.725836922908125e-06,
+      "loss": 0.387,
+      "step": 2071
+    },
+    {
+      "epoch": 1.5257731958762886,
+      "grad_norm": 0.40818580985069275,
+      "learning_rate": 5.7215972759641335e-06,
+      "loss": 0.4264,
+      "step": 2072
+    },
+    {
+      "epoch": 1.5265095729013254,
+      "grad_norm": 0.4584078788757324,
+      "learning_rate": 5.71735709910091e-06,
+      "loss": 0.4179,
+      "step": 2073
+    },
+    {
+      "epoch": 1.5272459499263622,
+      "grad_norm": 0.44075214862823486,
+      "learning_rate": 5.7131163954323085e-06,
+      "loss": 0.4033,
+      "step": 2074
+    },
+    {
+      "epoch": 1.527982326951399,
+      "grad_norm": 0.4163413643836975,
+      "learning_rate": 5.708875168072577e-06,
+      "loss": 0.4279,
+      "step": 2075
+    },
+    {
+      "epoch": 1.5287187039764358,
+      "grad_norm": 0.40220654010772705,
+      "learning_rate": 5.704633420136343e-06,
+      "loss": 0.3998,
+      "step": 2076
+    },
+    {
+      "epoch": 1.5294550810014726,
+      "grad_norm": 0.45772379636764526,
+      "learning_rate": 5.700391154738619e-06,
+      "loss": 0.3933,
+      "step": 2077
+    },
+    {
+      "epoch": 1.5301914580265095,
+      "grad_norm": 0.38176655769348145,
+      "learning_rate": 5.696148374994795e-06,
+      "loss": 0.4204,
+      "step": 2078
+    },
+    {
+      "epoch": 1.5309278350515463,
+      "grad_norm": 0.4314991533756256,
+      "learning_rate": 5.691905084020642e-06,
+      "loss": 0.3963,
+      "step": 2079
+    },
+    {
+      "epoch": 1.531664212076583,
+      "grad_norm": 0.42802858352661133,
+      "learning_rate": 5.687661284932306e-06,
+      "loss": 0.4353,
+      "step": 2080
+    },
+    {
+      "epoch": 1.53240058910162,
+      "grad_norm": 0.4672757089138031,
+      "learning_rate": 5.6834169808463e-06,
+      "loss": 0.439,
+      "step": 2081
+    },
+    {
+      "epoch": 1.5331369661266567,
+      "grad_norm": 0.42437198758125305,
+      "learning_rate": 5.679172174879516e-06,
+      "loss": 0.3863,
+      "step": 2082
+    },
+    {
+      "epoch": 1.5338733431516935,
+      "grad_norm": 0.404613196849823,
+      "learning_rate": 5.67492687014921e-06,
+      "loss": 0.4005,
+      "step": 2083
+    },
+    {
+      "epoch": 1.5346097201767304,
+      "grad_norm": 0.41044384241104126,
+      "learning_rate": 5.6706810697730095e-06,
+      "loss": 0.4065,
+      "step": 2084
+    },
+    {
+      "epoch": 1.5353460972017672,
+      "grad_norm": 0.4328327178955078,
+      "learning_rate": 5.666434776868895e-06,
+      "loss": 0.4264,
+      "step": 2085
+    },
+    {
+      "epoch": 1.536082474226804,
+      "grad_norm": 0.44456425309181213,
+      "learning_rate": 5.662187994555221e-06,
+      "loss": 0.3825,
+      "step": 2086
+    },
+    {
+      "epoch": 1.5368188512518408,
+      "grad_norm": 0.4055061638355255,
+      "learning_rate": 5.657940725950693e-06,
+      "loss": 0.3757,
+      "step": 2087
+    },
+    {
+      "epoch": 1.5375552282768776,
+      "grad_norm": 0.4078313112258911,
+      "learning_rate": 5.65369297417438e-06,
+      "loss": 0.414,
+      "step": 2088
+    },
+    {
+      "epoch": 1.5382916053019144,
+      "grad_norm": 0.4202408194541931,
+      "learning_rate": 5.6494447423457e-06,
+      "loss": 0.406,
+      "step": 2089
+    },
+    {
+      "epoch": 1.5390279823269513,
+      "grad_norm": 0.4479289650917053,
+      "learning_rate": 5.645196033584426e-06,
+      "loss": 0.384,
+      "step": 2090
+    },
+    {
+      "epoch": 1.539764359351988,
+      "grad_norm": 0.40371423959732056,
+      "learning_rate": 5.640946851010682e-06,
+      "loss": 0.3914,
+      "step": 2091
+    },
+    {
+      "epoch": 1.540500736377025,
+      "grad_norm": 0.3889717161655426,
+      "learning_rate": 5.636697197744941e-06,
+      "loss": 0.4069,
+      "step": 2092
+    },
+    {
+      "epoch": 1.5412371134020617,
+      "grad_norm": 0.4180935323238373,
+      "learning_rate": 5.6324470769080165e-06,
+      "loss": 0.4073,
+      "step": 2093
+    },
+    {
+      "epoch": 1.5419734904270985,
+      "grad_norm": 0.46457764506340027,
+      "learning_rate": 5.6281964916210715e-06,
+      "loss": 0.3946,
+      "step": 2094
+    },
+    {
+      "epoch": 1.5427098674521353,
+      "grad_norm": 0.41641589999198914,
+      "learning_rate": 5.6239454450056066e-06,
+      "loss": 0.4287,
+      "step": 2095
+    },
+    {
+      "epoch": 1.5434462444771722,
+      "grad_norm": 0.40457090735435486,
+      "learning_rate": 5.6196939401834625e-06,
+      "loss": 0.4061,
+      "step": 2096
+    },
+    {
+      "epoch": 1.544182621502209,
+      "grad_norm": 0.42012038826942444,
+      "learning_rate": 5.615441980276814e-06,
+      "loss": 0.4113,
+      "step": 2097
+    },
+    {
+      "epoch": 1.5449189985272458,
+      "grad_norm": 0.4231841564178467,
+      "learning_rate": 5.611189568408173e-06,
+      "loss": 0.4352,
+      "step": 2098
+    },
+    {
+      "epoch": 1.5456553755522826,
+      "grad_norm": 0.3875311017036438,
+      "learning_rate": 5.6069367077003835e-06,
+      "loss": 0.4069,
+      "step": 2099
+    },
+    {
+      "epoch": 1.5463917525773194,
+      "grad_norm": 0.38472917675971985,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.3963,
+      "step": 2100
+    },
+    {
+      "epoch": 1.5471281296023565,
+      "grad_norm": 0.48356834053993225,
+      "learning_rate": 5.598429652260371e-06,
+      "loss": 0.4407,
+      "step": 2101
+    },
+    {
+      "epoch": 1.5478645066273933,
+      "grad_norm": 0.3837643563747406,
+      "learning_rate": 5.594175463775475e-06,
+      "loss": 0.3983,
+      "step": 2102
+    },
+    {
+      "epoch": 1.54860088365243,
+      "grad_norm": 0.4311898648738861,
+      "learning_rate": 5.5899208389460715e-06,
+      "loss": 0.4,
+      "step": 2103
+    },
+    {
+      "epoch": 1.549337260677467,
+      "grad_norm": 0.38184744119644165,
+      "learning_rate": 5.5856657808966315e-06,
+      "loss": 0.3998,
+      "step": 2104
+    },
+    {
+      "epoch": 1.5500736377025037,
+      "grad_norm": 0.4166722297668457,
+      "learning_rate": 5.581410292751941e-06,
+      "loss": 0.3881,
+      "step": 2105
+    },
+    {
+      "epoch": 1.5508100147275405,
+      "grad_norm": 0.44956570863723755,
+      "learning_rate": 5.577154377637101e-06,
+      "loss": 0.4095,
+      "step": 2106
+    },
+    {
+      "epoch": 1.5515463917525774,
+      "grad_norm": 0.3997005522251129,
+      "learning_rate": 5.572898038677526e-06,
+      "loss": 0.4035,
+      "step": 2107
+    },
+    {
+      "epoch": 1.5522827687776142,
+      "grad_norm": 0.4489591121673584,
+      "learning_rate": 5.5686412789989444e-06,
+      "loss": 0.4001,
+      "step": 2108
+    },
+    {
+      "epoch": 1.553019145802651,
+      "grad_norm": 0.40616336464881897,
+      "learning_rate": 5.5643841017273915e-06,
+      "loss": 0.3858,
+      "step": 2109
+    },
+    {
+      "epoch": 1.5537555228276878,
+      "grad_norm": 0.46725019812583923,
+      "learning_rate": 5.560126509989209e-06,
+      "loss": 0.3954,
+      "step": 2110
+    },
+    {
+      "epoch": 1.5544918998527246,
+      "grad_norm": 0.42120271921157837,
+      "learning_rate": 5.5558685069110444e-06,
+      "loss": 0.4105,
+      "step": 2111
+    },
+    {
+      "epoch": 1.5552282768777614,
+      "grad_norm": 0.361070454120636,
+      "learning_rate": 5.5516100956198445e-06,
+      "loss": 0.3762,
+      "step": 2112
+    },
+    {
+      "epoch": 1.5559646539027983,
+      "grad_norm": 0.4115205705165863,
+      "learning_rate": 5.547351279242861e-06,
+      "loss": 0.4182,
+      "step": 2113
+    },
+    {
+      "epoch": 1.556701030927835,
+      "grad_norm": 0.40333011746406555,
+      "learning_rate": 5.543092060907639e-06,
+      "loss": 0.3663,
+      "step": 2114
+    },
+    {
+      "epoch": 1.5574374079528719,
+      "grad_norm": 0.4221840500831604,
+      "learning_rate": 5.538832443742018e-06,
+      "loss": 0.4132,
+      "step": 2115
+    },
+    {
+      "epoch": 1.5581737849779087,
+      "grad_norm": 0.40356510877609253,
+      "learning_rate": 5.5345724308741326e-06,
+      "loss": 0.3685,
+      "step": 2116
+    },
+    {
+      "epoch": 1.5589101620029455,
+      "grad_norm": 0.47017496824264526,
+      "learning_rate": 5.5303120254324104e-06,
+      "loss": 0.4117,
+      "step": 2117
+    },
+    {
+      "epoch": 1.5596465390279823,
+      "grad_norm": 0.3889496624469757,
+      "learning_rate": 5.52605123054556e-06,
+      "loss": 0.4128,
+      "step": 2118
+    },
+    {
+      "epoch": 1.5603829160530192,
+      "grad_norm": 0.44326820969581604,
+      "learning_rate": 5.521790049342583e-06,
+      "loss": 0.4213,
+      "step": 2119
+    },
+    {
+      "epoch": 1.561119293078056,
+      "grad_norm": 0.4094356298446655,
+      "learning_rate": 5.5175284849527635e-06,
+      "loss": 0.4006,
+      "step": 2120
+    },
+    {
+      "epoch": 1.5618556701030928,
+      "grad_norm": 0.46180570125579834,
+      "learning_rate": 5.513266540505662e-06,
+      "loss": 0.4114,
+      "step": 2121
+    },
+    {
+      "epoch": 1.5625920471281296,
+      "grad_norm": 0.38660722970962524,
+      "learning_rate": 5.509004219131124e-06,
+      "loss": 0.3904,
+      "step": 2122
+    },
+    {
+      "epoch": 1.5633284241531664,
+      "grad_norm": 0.42590051889419556,
+      "learning_rate": 5.504741523959269e-06,
+      "loss": 0.4155,
+      "step": 2123
+    },
+    {
+      "epoch": 1.5640648011782032,
+      "grad_norm": 0.4073706269264221,
+      "learning_rate": 5.500478458120493e-06,
+      "loss": 0.4318,
+      "step": 2124
+    },
+    {
+      "epoch": 1.56480117820324,
+      "grad_norm": 0.39798110723495483,
+      "learning_rate": 5.49621502474546e-06,
+      "loss": 0.416,
+      "step": 2125
+    },
+    {
+      "epoch": 1.5655375552282769,
+      "grad_norm": 0.4096386432647705,
+      "learning_rate": 5.491951226965108e-06,
+      "loss": 0.4398,
+      "step": 2126
+    },
+    {
+      "epoch": 1.5662739322533137,
+      "grad_norm": 0.4136483669281006,
+      "learning_rate": 5.48768706791064e-06,
+      "loss": 0.4249,
+      "step": 2127
+    },
+    {
+      "epoch": 1.5670103092783505,
+      "grad_norm": 0.435358464717865,
+      "learning_rate": 5.4834225507135284e-06,
+      "loss": 0.3931,
+      "step": 2128
+    },
+    {
+      "epoch": 1.5677466863033873,
+      "grad_norm": 0.39365580677986145,
+      "learning_rate": 5.479157678505503e-06,
+      "loss": 0.419,
+      "step": 2129
+    },
+    {
+      "epoch": 1.5684830633284241,
+      "grad_norm": 0.4026069939136505,
+      "learning_rate": 5.474892454418559e-06,
+      "loss": 0.393,
+      "step": 2130
+    },
+    {
+      "epoch": 1.569219440353461,
+      "grad_norm": 0.44560348987579346,
+      "learning_rate": 5.470626881584948e-06,
+      "loss": 0.4176,
+      "step": 2131
+    },
+    {
+      "epoch": 1.5699558173784978,
+      "grad_norm": 0.4150668978691101,
+      "learning_rate": 5.466360963137175e-06,
+      "loss": 0.4245,
+      "step": 2132
+    },
+    {
+      "epoch": 1.5706921944035346,
+      "grad_norm": 0.3756403923034668,
+      "learning_rate": 5.462094702208004e-06,
+      "loss": 0.4057,
+      "step": 2133
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.45393243432044983,
+      "learning_rate": 5.4578281019304494e-06,
+      "loss": 0.411,
+      "step": 2134
+    },
+    {
+      "epoch": 1.5721649484536082,
+      "grad_norm": 0.4269341826438904,
+      "learning_rate": 5.453561165437771e-06,
+      "loss": 0.3946,
+      "step": 2135
+    },
+    {
+      "epoch": 1.572901325478645,
+      "grad_norm": 0.39662691950798035,
+      "learning_rate": 5.449293895863478e-06,
+      "loss": 0.3958,
+      "step": 2136
+    },
+    {
+      "epoch": 1.5736377025036818,
+      "grad_norm": 0.5086665153503418,
+      "learning_rate": 5.445026296341325e-06,
+      "loss": 0.4376,
+      "step": 2137
+    },
+    {
+      "epoch": 1.5743740795287187,
+      "grad_norm": 0.43581289052963257,
+      "learning_rate": 5.440758370005309e-06,
+      "loss": 0.4384,
+      "step": 2138
+    },
+    {
+      "epoch": 1.5751104565537555,
+      "grad_norm": 0.39093270897865295,
+      "learning_rate": 5.4364901199896655e-06,
+      "loss": 0.4093,
+      "step": 2139
+    },
+    {
+      "epoch": 1.5758468335787923,
+      "grad_norm": 0.4165237545967102,
+      "learning_rate": 5.432221549428867e-06,
+      "loss": 0.3921,
+      "step": 2140
+    },
+    {
+      "epoch": 1.576583210603829,
+      "grad_norm": 0.4502768814563751,
+      "learning_rate": 5.427952661457624e-06,
+      "loss": 0.4246,
+      "step": 2141
+    },
+    {
+      "epoch": 1.577319587628866,
+      "grad_norm": 0.42256587743759155,
+      "learning_rate": 5.42368345921088e-06,
+      "loss": 0.3885,
+      "step": 2142
+    },
+    {
+      "epoch": 1.5780559646539027,
+      "grad_norm": 0.3865189552307129,
+      "learning_rate": 5.419413945823806e-06,
+      "loss": 0.4251,
+      "step": 2143
+    },
+    {
+      "epoch": 1.5787923416789398,
+      "grad_norm": 0.405563086271286,
+      "learning_rate": 5.415144124431805e-06,
+      "loss": 0.4107,
+      "step": 2144
+    },
+    {
+      "epoch": 1.5795287187039766,
+      "grad_norm": 0.4749312102794647,
+      "learning_rate": 5.410873998170503e-06,
+      "loss": 0.4022,
+      "step": 2145
+    },
+    {
+      "epoch": 1.5802650957290134,
+      "grad_norm": 0.41955462098121643,
+      "learning_rate": 5.4066035701757535e-06,
+      "loss": 0.4113,
+      "step": 2146
+    },
+    {
+      "epoch": 1.5810014727540502,
+      "grad_norm": 0.34526562690734863,
+      "learning_rate": 5.402332843583631e-06,
+      "loss": 0.4176,
+      "step": 2147
+    },
+    {
+      "epoch": 1.581737849779087,
+      "grad_norm": 0.4131619334220886,
+      "learning_rate": 5.398061821530423e-06,
+      "loss": 0.398,
+      "step": 2148
+    },
+    {
+      "epoch": 1.5824742268041239,
+      "grad_norm": 0.38839566707611084,
+      "learning_rate": 5.393790507152645e-06,
+      "loss": 0.4031,
+      "step": 2149
+    },
+    {
+      "epoch": 1.5832106038291607,
+      "grad_norm": 0.35662680864334106,
+      "learning_rate": 5.389518903587016e-06,
+      "loss": 0.3957,
+      "step": 2150
+    },
+    {
+      "epoch": 1.5839469808541975,
+      "grad_norm": 0.417187362909317,
+      "learning_rate": 5.3852470139704786e-06,
+      "loss": 0.4268,
+      "step": 2151
+    },
+    {
+      "epoch": 1.5846833578792343,
+      "grad_norm": 0.4425226151943207,
+      "learning_rate": 5.380974841440173e-06,
+      "loss": 0.4115,
+      "step": 2152
+    },
+    {
+      "epoch": 1.5854197349042711,
+      "grad_norm": 0.41201913356781006,
+      "learning_rate": 5.376702389133458e-06,
+      "loss": 0.4056,
+      "step": 2153
+    },
+    {
+      "epoch": 1.586156111929308,
+      "grad_norm": 0.39344313740730286,
+      "learning_rate": 5.37242966018789e-06,
+      "loss": 0.4178,
+      "step": 2154
+    },
+    {
+      "epoch": 1.5868924889543448,
+      "grad_norm": 0.41077542304992676,
+      "learning_rate": 5.3681566577412355e-06,
+      "loss": 0.4046,
+      "step": 2155
+    },
+    {
+      "epoch": 1.5876288659793816,
+      "grad_norm": 0.40460658073425293,
+      "learning_rate": 5.363883384931456e-06,
+      "loss": 0.4302,
+      "step": 2156
+    },
+    {
+      "epoch": 1.5883652430044184,
+      "grad_norm": 0.43949198722839355,
+      "learning_rate": 5.359609844896717e-06,
+      "loss": 0.4148,
+      "step": 2157
+    },
+    {
+      "epoch": 1.5891016200294552,
+      "grad_norm": 0.43986040353775024,
+      "learning_rate": 5.355336040775373e-06,
+      "loss": 0.4092,
+      "step": 2158
+    },
+    {
+      "epoch": 1.589837997054492,
+      "grad_norm": 0.41094833612442017,
+      "learning_rate": 5.3510619757059775e-06,
+      "loss": 0.3936,
+      "step": 2159
+    },
+    {
+      "epoch": 1.5905743740795288,
+      "grad_norm": 0.456833153963089,
+      "learning_rate": 5.346787652827279e-06,
+      "loss": 0.416,
+      "step": 2160
+    },
+    {
+      "epoch": 1.5913107511045657,
+      "grad_norm": 0.43459081649780273,
+      "learning_rate": 5.3425130752782065e-06,
+      "loss": 0.4134,
+      "step": 2161
+    },
+    {
+      "epoch": 1.5920471281296025,
+      "grad_norm": 0.41692793369293213,
+      "learning_rate": 5.33823824619788e-06,
+      "loss": 0.3983,
+      "step": 2162
+    },
+    {
+      "epoch": 1.5927835051546393,
+      "grad_norm": 0.37748250365257263,
+      "learning_rate": 5.3339631687256085e-06,
+      "loss": 0.399,
+      "step": 2163
+    },
+    {
+      "epoch": 1.593519882179676,
+      "grad_norm": 0.40077048540115356,
+      "learning_rate": 5.3296878460008785e-06,
+      "loss": 0.3902,
+      "step": 2164
+    },
+    {
+      "epoch": 1.594256259204713,
+      "grad_norm": 0.4162245988845825,
+      "learning_rate": 5.325412281163356e-06,
+      "loss": 0.3972,
+      "step": 2165
+    },
+    {
+      "epoch": 1.5949926362297497,
+      "grad_norm": 0.41424018144607544,
+      "learning_rate": 5.321136477352887e-06,
+      "loss": 0.396,
+      "step": 2166
+    },
+    {
+      "epoch": 1.5957290132547866,
+      "grad_norm": 0.4058535099029541,
+      "learning_rate": 5.3168604377094945e-06,
+      "loss": 0.3772,
+      "step": 2167
+    },
+    {
+      "epoch": 1.5964653902798234,
+      "grad_norm": 0.44171908497810364,
+      "learning_rate": 5.312584165373372e-06,
+      "loss": 0.4135,
+      "step": 2168
+    },
+    {
+      "epoch": 1.5972017673048602,
+      "grad_norm": 0.46828219294548035,
+      "learning_rate": 5.308307663484884e-06,
+      "loss": 0.4229,
+      "step": 2169
+    },
+    {
+      "epoch": 1.597938144329897,
+      "grad_norm": 0.39885610342025757,
+      "learning_rate": 5.304030935184564e-06,
+      "loss": 0.4146,
+      "step": 2170
+    },
+    {
+      "epoch": 1.5986745213549338,
+      "grad_norm": 0.4212936460971832,
+      "learning_rate": 5.299753983613114e-06,
+      "loss": 0.4033,
+      "step": 2171
+    },
+    {
+      "epoch": 1.5994108983799706,
+      "grad_norm": 0.4296552836894989,
+      "learning_rate": 5.2954768119113975e-06,
+      "loss": 0.4278,
+      "step": 2172
+    },
+    {
+      "epoch": 1.6001472754050075,
+      "grad_norm": 0.4296092987060547,
+      "learning_rate": 5.291199423220438e-06,
+      "loss": 0.4141,
+      "step": 2173
+    },
+    {
+      "epoch": 1.6008836524300443,
+      "grad_norm": 0.3816363215446472,
+      "learning_rate": 5.286921820681421e-06,
+      "loss": 0.4128,
+      "step": 2174
+    },
+    {
+      "epoch": 1.601620029455081,
+      "grad_norm": 0.415924608707428,
+      "learning_rate": 5.28264400743569e-06,
+      "loss": 0.4133,
+      "step": 2175
+    },
+    {
+      "epoch": 1.602356406480118,
+      "grad_norm": 0.47628721594810486,
+      "learning_rate": 5.278365986624743e-06,
+      "loss": 0.4195,
+      "step": 2176
+    },
+    {
+      "epoch": 1.6030927835051547,
+      "grad_norm": 0.477327823638916,
+      "learning_rate": 5.274087761390224e-06,
+      "loss": 0.4116,
+      "step": 2177
+    },
+    {
+      "epoch": 1.6038291605301915,
+      "grad_norm": 0.42795825004577637,
+      "learning_rate": 5.269809334873939e-06,
+      "loss": 0.4033,
+      "step": 2178
+    },
+    {
+      "epoch": 1.6045655375552283,
+      "grad_norm": 0.4993734657764435,
+      "learning_rate": 5.2655307102178285e-06,
+      "loss": 0.4451,
+      "step": 2179
+    },
+    {
+      "epoch": 1.6053019145802652,
+      "grad_norm": 0.49656498432159424,
+      "learning_rate": 5.26125189056399e-06,
+      "loss": 0.4144,
+      "step": 2180
+    },
+    {
+      "epoch": 1.606038291605302,
+      "grad_norm": 0.4029616713523865,
+      "learning_rate": 5.256972879054659e-06,
+      "loss": 0.4113,
+      "step": 2181
+    },
+    {
+      "epoch": 1.6067746686303388,
+      "grad_norm": 0.46401447057724,
+      "learning_rate": 5.2526936788322106e-06,
+      "loss": 0.4219,
+      "step": 2182
+    },
+    {
+      "epoch": 1.6075110456553756,
+      "grad_norm": 0.49594005942344666,
+      "learning_rate": 5.248414293039159e-06,
+      "loss": 0.4095,
+      "step": 2183
+    },
+    {
+      "epoch": 1.6082474226804124,
+      "grad_norm": 0.5078505873680115,
+      "learning_rate": 5.244134724818158e-06,
+      "loss": 0.4391,
+      "step": 2184
+    },
+    {
+      "epoch": 1.6089837997054492,
+      "grad_norm": 0.3824053108692169,
+      "learning_rate": 5.2398549773119945e-06,
+      "loss": 0.391,
+      "step": 2185
+    },
+    {
+      "epoch": 1.609720176730486,
+      "grad_norm": 0.44562169909477234,
+      "learning_rate": 5.235575053663582e-06,
+      "loss": 0.4052,
+      "step": 2186
+    },
+    {
+      "epoch": 1.6104565537555229,
+      "grad_norm": 0.44519391655921936,
+      "learning_rate": 5.231294957015969e-06,
+      "loss": 0.4049,
+      "step": 2187
+    },
+    {
+      "epoch": 1.6111929307805597,
+      "grad_norm": 0.38943272829055786,
+      "learning_rate": 5.2270146905123285e-06,
+      "loss": 0.3984,
+      "step": 2188
+    },
+    {
+      "epoch": 1.6119293078055965,
+      "grad_norm": 0.4190487265586853,
+      "learning_rate": 5.222734257295963e-06,
+      "loss": 0.3971,
+      "step": 2189
+    },
+    {
+      "epoch": 1.6126656848306333,
+      "grad_norm": 0.5010129809379578,
+      "learning_rate": 5.218453660510287e-06,
+      "loss": 0.3993,
+      "step": 2190
+    },
+    {
+      "epoch": 1.6134020618556701,
+      "grad_norm": 0.4155033826828003,
+      "learning_rate": 5.214172903298843e-06,
+      "loss": 0.4258,
+      "step": 2191
+    },
+    {
+      "epoch": 1.614138438880707,
+      "grad_norm": 0.441825270652771,
+      "learning_rate": 5.209891988805292e-06,
+      "loss": 0.4368,
+      "step": 2192
+    },
+    {
+      "epoch": 1.6148748159057438,
+      "grad_norm": 0.49732720851898193,
+      "learning_rate": 5.205610920173408e-06,
+      "loss": 0.3995,
+      "step": 2193
+    },
+    {
+      "epoch": 1.6156111929307806,
+      "grad_norm": 0.39198291301727295,
+      "learning_rate": 5.201329700547077e-06,
+      "loss": 0.412,
+      "step": 2194
+    },
+    {
+      "epoch": 1.6163475699558174,
+      "grad_norm": 0.37358447909355164,
+      "learning_rate": 5.197048333070297e-06,
+      "loss": 0.4017,
+      "step": 2195
+    },
+    {
+      "epoch": 1.6170839469808542,
+      "grad_norm": 0.49914005398750305,
+      "learning_rate": 5.192766820887177e-06,
+      "loss": 0.4346,
+      "step": 2196
+    },
+    {
+      "epoch": 1.617820324005891,
+      "grad_norm": 0.45152711868286133,
+      "learning_rate": 5.188485167141929e-06,
+      "loss": 0.3845,
+      "step": 2197
+    },
+    {
+      "epoch": 1.6185567010309279,
+      "grad_norm": 0.3699303865432739,
+      "learning_rate": 5.1842033749788686e-06,
+      "loss": 0.3937,
+      "step": 2198
+    },
+    {
+      "epoch": 1.6192930780559647,
+      "grad_norm": 0.578906238079071,
+      "learning_rate": 5.179921447542417e-06,
+      "loss": 0.4365,
+      "step": 2199
+    },
+    {
+      "epoch": 1.6200294550810015,
+      "grad_norm": 0.3776850998401642,
+      "learning_rate": 5.175639387977091e-06,
+      "loss": 0.4229,
+      "step": 2200
+    },
+    {
+      "epoch": 1.6207658321060383,
+      "grad_norm": 0.5020015239715576,
+      "learning_rate": 5.171357199427507e-06,
+      "loss": 0.4087,
+      "step": 2201
+    },
+    {
+      "epoch": 1.6215022091310751,
+      "grad_norm": 0.447489857673645,
+      "learning_rate": 5.1670748850383734e-06,
+      "loss": 0.4046,
+      "step": 2202
+    },
+    {
+      "epoch": 1.622238586156112,
+      "grad_norm": 0.4035995602607727,
+      "learning_rate": 5.162792447954494e-06,
+      "loss": 0.3878,
+      "step": 2203
+    },
+    {
+      "epoch": 1.6229749631811488,
+      "grad_norm": 0.43608716130256653,
+      "learning_rate": 5.158509891320759e-06,
+      "loss": 0.4031,
+      "step": 2204
+    },
+    {
+      "epoch": 1.6237113402061856,
+      "grad_norm": 0.5111272931098938,
+      "learning_rate": 5.154227218282149e-06,
+      "loss": 0.3981,
+      "step": 2205
+    },
+    {
+      "epoch": 1.6244477172312224,
+      "grad_norm": 0.40405896306037903,
+      "learning_rate": 5.1499444319837326e-06,
+      "loss": 0.3946,
+      "step": 2206
+    },
+    {
+      "epoch": 1.6251840942562592,
+      "grad_norm": 0.4375455975532532,
+      "learning_rate": 5.145661535570656e-06,
+      "loss": 0.4045,
+      "step": 2207
+    },
+    {
+      "epoch": 1.625920471281296,
+      "grad_norm": 0.4746772348880768,
+      "learning_rate": 5.141378532188148e-06,
+      "loss": 0.4113,
+      "step": 2208
+    },
+    {
+      "epoch": 1.6266568483063328,
+      "grad_norm": 0.4946158826351166,
+      "learning_rate": 5.137095424981519e-06,
+      "loss": 0.4086,
+      "step": 2209
+    },
+    {
+      "epoch": 1.6273932253313697,
+      "grad_norm": 0.4054524600505829,
+      "learning_rate": 5.1328122170961534e-06,
+      "loss": 0.4218,
+      "step": 2210
+    },
+    {
+      "epoch": 1.6281296023564065,
+      "grad_norm": 0.5024380683898926,
+      "learning_rate": 5.128528911677509e-06,
+      "loss": 0.3965,
+      "step": 2211
+    },
+    {
+      "epoch": 1.6288659793814433,
+      "grad_norm": 0.409605473279953,
+      "learning_rate": 5.124245511871115e-06,
+      "loss": 0.4167,
+      "step": 2212
+    },
+    {
+      "epoch": 1.62960235640648,
+      "grad_norm": 0.44012802839279175,
+      "learning_rate": 5.119962020822572e-06,
+      "loss": 0.431,
+      "step": 2213
+    },
+    {
+      "epoch": 1.630338733431517,
+      "grad_norm": 0.41772133111953735,
+      "learning_rate": 5.115678441677546e-06,
+      "loss": 0.3957,
+      "step": 2214
+    },
+    {
+      "epoch": 1.6310751104565537,
+      "grad_norm": 0.3946409523487091,
+      "learning_rate": 5.111394777581769e-06,
+      "loss": 0.422,
+      "step": 2215
+    },
+    {
+      "epoch": 1.6318114874815906,
+      "grad_norm": 0.4510660767555237,
+      "learning_rate": 5.107111031681034e-06,
+      "loss": 0.4063,
+      "step": 2216
+    },
+    {
+      "epoch": 1.6325478645066274,
+      "grad_norm": 0.4569968581199646,
+      "learning_rate": 5.1028272071211916e-06,
+      "loss": 0.4193,
+      "step": 2217
+    },
+    {
+      "epoch": 1.6332842415316642,
+      "grad_norm": 0.45792651176452637,
+      "learning_rate": 5.098543307048158e-06,
+      "loss": 0.4073,
+      "step": 2218
+    },
+    {
+      "epoch": 1.634020618556701,
+      "grad_norm": 0.42194387316703796,
+      "learning_rate": 5.094259334607896e-06,
+      "loss": 0.3922,
+      "step": 2219
+    },
+    {
+      "epoch": 1.6347569955817378,
+      "grad_norm": 0.48982447385787964,
+      "learning_rate": 5.089975292946427e-06,
+      "loss": 0.4192,
+      "step": 2220
+    },
+    {
+      "epoch": 1.6354933726067746,
+      "grad_norm": 0.4341419041156769,
+      "learning_rate": 5.085691185209824e-06,
+      "loss": 0.428,
+      "step": 2221
+    },
+    {
+      "epoch": 1.6362297496318114,
+      "grad_norm": 0.4825488030910492,
+      "learning_rate": 5.081407014544202e-06,
+      "loss": 0.4034,
+      "step": 2222
+    },
+    {
+      "epoch": 1.6369661266568483,
+      "grad_norm": 0.4492826759815216,
+      "learning_rate": 5.07712278409573e-06,
+      "loss": 0.4292,
+      "step": 2223
+    },
+    {
+      "epoch": 1.637702503681885,
+      "grad_norm": 0.44333913922309875,
+      "learning_rate": 5.0728384970106135e-06,
+      "loss": 0.3939,
+      "step": 2224
+    },
+    {
+      "epoch": 1.638438880706922,
+      "grad_norm": 0.3804095387458801,
+      "learning_rate": 5.068554156435108e-06,
+      "loss": 0.3795,
+      "step": 2225
+    },
+    {
+      "epoch": 1.6391752577319587,
+      "grad_norm": 0.3924141228199005,
+      "learning_rate": 5.0642697655155e-06,
+      "loss": 0.3771,
+      "step": 2226
+    },
+    {
+      "epoch": 1.6399116347569955,
+      "grad_norm": 0.4596264660358429,
+      "learning_rate": 5.059985327398121e-06,
+      "loss": 0.4098,
+      "step": 2227
+    },
+    {
+      "epoch": 1.6406480117820323,
+      "grad_norm": 0.4128621220588684,
+      "learning_rate": 5.0557008452293275e-06,
+      "loss": 0.4235,
+      "step": 2228
+    },
+    {
+      "epoch": 1.6413843888070692,
+      "grad_norm": 0.44195055961608887,
+      "learning_rate": 5.051416322155519e-06,
+      "loss": 0.397,
+      "step": 2229
+    },
+    {
+      "epoch": 1.642120765832106,
+      "grad_norm": 0.4374493956565857,
+      "learning_rate": 5.047131761323115e-06,
+      "loss": 0.3965,
+      "step": 2230
+    },
+    {
+      "epoch": 1.6428571428571428,
+      "grad_norm": 0.39338651299476624,
+      "learning_rate": 5.0428471658785715e-06,
+      "loss": 0.3714,
+      "step": 2231
+    },
+    {
+      "epoch": 1.6435935198821796,
+      "grad_norm": 0.4788668155670166,
+      "learning_rate": 5.038562538968363e-06,
+      "loss": 0.4252,
+      "step": 2232
+    },
+    {
+      "epoch": 1.6443298969072164,
+      "grad_norm": 0.4350014328956604,
+      "learning_rate": 5.034277883738992e-06,
+      "loss": 0.3947,
+      "step": 2233
+    },
+    {
+      "epoch": 1.6450662739322532,
+      "grad_norm": 0.3968852460384369,
+      "learning_rate": 5.029993203336978e-06,
+      "loss": 0.4066,
+      "step": 2234
+    },
+    {
+      "epoch": 1.64580265095729,
+      "grad_norm": 0.3912464678287506,
+      "learning_rate": 5.025708500908864e-06,
+      "loss": 0.4011,
+      "step": 2235
+    },
+    {
+      "epoch": 1.6465390279823269,
+      "grad_norm": 0.4370138943195343,
+      "learning_rate": 5.021423779601202e-06,
+      "loss": 0.3789,
+      "step": 2236
+    },
+    {
+      "epoch": 1.6472754050073637,
+      "grad_norm": 0.39066141843795776,
+      "learning_rate": 5.017139042560564e-06,
+      "loss": 0.3597,
+      "step": 2237
+    },
+    {
+      "epoch": 1.6480117820324005,
+      "grad_norm": 0.45047131180763245,
+      "learning_rate": 5.01285429293353e-06,
+      "loss": 0.4255,
+      "step": 2238
+    },
+    {
+      "epoch": 1.6487481590574373,
+      "grad_norm": 0.4357801377773285,
+      "learning_rate": 5.008569533866693e-06,
+      "loss": 0.4201,
+      "step": 2239
+    },
+    {
+      "epoch": 1.6494845360824741,
+      "grad_norm": 0.3798350393772125,
+      "learning_rate": 5.00428476850665e-06,
+      "loss": 0.3872,
+      "step": 2240
+    },
+    {
+      "epoch": 1.650220913107511,
+      "grad_norm": 0.4238215386867523,
+      "learning_rate": 5e-06,
+      "loss": 0.4003,
+      "step": 2241
+    },
+    {
+      "epoch": 1.6509572901325478,
+      "grad_norm": 0.4555947482585907,
+      "learning_rate": 4.995715231493352e-06,
+      "loss": 0.4197,
+      "step": 2242
+    },
+    {
+      "epoch": 1.6516936671575846,
+      "grad_norm": 0.3676789700984955,
+      "learning_rate": 4.991430466133308e-06,
+      "loss": 0.3704,
+      "step": 2243
+    },
+    {
+      "epoch": 1.6524300441826214,
+      "grad_norm": 0.43429771065711975,
+      "learning_rate": 4.98714570706647e-06,
+      "loss": 0.4075,
+      "step": 2244
+    },
+    {
+      "epoch": 1.6531664212076582,
+      "grad_norm": 0.4155738055706024,
+      "learning_rate": 4.982860957439437e-06,
+      "loss": 0.4313,
+      "step": 2245
+    },
+    {
+      "epoch": 1.653902798232695,
+      "grad_norm": 0.38362860679626465,
+      "learning_rate": 4.978576220398801e-06,
+      "loss": 0.4021,
+      "step": 2246
+    },
+    {
+      "epoch": 1.6546391752577319,
+      "grad_norm": 0.39530959725379944,
+      "learning_rate": 4.97429149909114e-06,
+      "loss": 0.3926,
+      "step": 2247
+    },
+    {
+      "epoch": 1.6553755522827687,
+      "grad_norm": 0.36841920018196106,
+      "learning_rate": 4.970006796663023e-06,
+      "loss": 0.4198,
+      "step": 2248
+    },
+    {
+      "epoch": 1.6561119293078055,
+      "grad_norm": 0.34673774242401123,
+      "learning_rate": 4.965722116261009e-06,
+      "loss": 0.413,
+      "step": 2249
+    },
+    {
+      "epoch": 1.6568483063328423,
+      "grad_norm": 0.35781288146972656,
+      "learning_rate": 4.961437461031638e-06,
+      "loss": 0.4232,
+      "step": 2250
+    },
+    {
+      "epoch": 1.6575846833578791,
+      "grad_norm": 0.4127185046672821,
+      "learning_rate": 4.95715283412143e-06,
+      "loss": 0.4162,
+      "step": 2251
+    },
+    {
+      "epoch": 1.658321060382916,
+      "grad_norm": 0.3733428120613098,
+      "learning_rate": 4.952868238676885e-06,
+      "loss": 0.3892,
+      "step": 2252
+    },
+    {
+      "epoch": 1.6590574374079528,
+      "grad_norm": 0.46769607067108154,
+      "learning_rate": 4.948583677844482e-06,
+      "loss": 0.3946,
+      "step": 2253
+    },
+    {
+      "epoch": 1.6597938144329896,
+      "grad_norm": 0.39663955569267273,
+      "learning_rate": 4.944299154770673e-06,
+      "loss": 0.3957,
+      "step": 2254
+    },
+    {
+      "epoch": 1.6605301914580264,
+      "grad_norm": 0.3862704932689667,
+      "learning_rate": 4.940014672601881e-06,
+      "loss": 0.4042,
+      "step": 2255
+    },
+    {
+      "epoch": 1.6612665684830632,
+      "grad_norm": 0.36831381916999817,
+      "learning_rate": 4.9357302344845005e-06,
+      "loss": 0.388,
+      "step": 2256
+    },
+    {
+      "epoch": 1.6620029455081,
+      "grad_norm": 0.4497531056404114,
+      "learning_rate": 4.931445843564893e-06,
+      "loss": 0.4393,
+      "step": 2257
+    },
+    {
+      "epoch": 1.6627393225331368,
+      "grad_norm": 0.4758419692516327,
+      "learning_rate": 4.927161502989387e-06,
+      "loss": 0.3868,
+      "step": 2258
+    },
+    {
+      "epoch": 1.6634756995581736,
+      "grad_norm": 0.3863910734653473,
+      "learning_rate": 4.922877215904272e-06,
+      "loss": 0.4029,
+      "step": 2259
+    },
+    {
+      "epoch": 1.6642120765832105,
+      "grad_norm": 0.4318389892578125,
+      "learning_rate": 4.918592985455799e-06,
+      "loss": 0.4101,
+      "step": 2260
+    },
+    {
+      "epoch": 1.6649484536082473,
+      "grad_norm": 0.44303011894226074,
+      "learning_rate": 4.914308814790178e-06,
+      "loss": 0.3769,
+      "step": 2261
+    },
+    {
+      "epoch": 1.665684830633284,
+      "grad_norm": 0.4603565037250519,
+      "learning_rate": 4.910024707053573e-06,
+      "loss": 0.417,
+      "step": 2262
+    },
+    {
+      "epoch": 1.666421207658321,
+      "grad_norm": 0.3724450170993805,
+      "learning_rate": 4.905740665392106e-06,
+      "loss": 0.4129,
+      "step": 2263
+    },
+    {
+      "epoch": 1.6671575846833577,
+      "grad_norm": 0.4588872790336609,
+      "learning_rate": 4.901456692951844e-06,
+      "loss": 0.4192,
+      "step": 2264
+    },
+    {
+      "epoch": 1.6678939617083945,
+      "grad_norm": 0.36367759108543396,
+      "learning_rate": 4.89717279287881e-06,
+      "loss": 0.4206,
+      "step": 2265
+    },
+    {
+      "epoch": 1.6686303387334314,
+      "grad_norm": 0.3877205550670624,
+      "learning_rate": 4.892888968318968e-06,
+      "loss": 0.4204,
+      "step": 2266
+    },
+    {
+      "epoch": 1.6693667157584682,
+      "grad_norm": 0.4392288625240326,
+      "learning_rate": 4.888605222418232e-06,
+      "loss": 0.4133,
+      "step": 2267
+    },
+    {
+      "epoch": 1.670103092783505,
+      "grad_norm": 0.35620570182800293,
+      "learning_rate": 4.884321558322455e-06,
+      "loss": 0.3808,
+      "step": 2268
+    },
+    {
+      "epoch": 1.6708394698085418,
+      "grad_norm": 0.39953136444091797,
+      "learning_rate": 4.8800379791774285e-06,
+      "loss": 0.4151,
+      "step": 2269
+    },
+    {
+      "epoch": 1.6715758468335786,
+      "grad_norm": 0.46267053484916687,
+      "learning_rate": 4.875754488128885e-06,
+      "loss": 0.3956,
+      "step": 2270
+    },
+    {
+      "epoch": 1.6723122238586157,
+      "grad_norm": 0.401598185300827,
+      "learning_rate": 4.871471088322493e-06,
+      "loss": 0.4193,
+      "step": 2271
+    },
+    {
+      "epoch": 1.6730486008836525,
+      "grad_norm": 0.41850388050079346,
+      "learning_rate": 4.867187782903847e-06,
+      "loss": 0.3883,
+      "step": 2272
+    },
+    {
+      "epoch": 1.6737849779086893,
+      "grad_norm": 0.47216498851776123,
+      "learning_rate": 4.862904575018482e-06,
+      "loss": 0.4087,
+      "step": 2273
+    },
+    {
+      "epoch": 1.6745213549337261,
+      "grad_norm": 0.416972279548645,
+      "learning_rate": 4.8586214678118536e-06,
+      "loss": 0.3824,
+      "step": 2274
+    },
+    {
+      "epoch": 1.675257731958763,
+      "grad_norm": 0.41057777404785156,
+      "learning_rate": 4.854338464429346e-06,
+      "loss": 0.4095,
+      "step": 2275
+    },
+    {
+      "epoch": 1.6759941089837997,
+      "grad_norm": 0.41428229212760925,
+      "learning_rate": 4.850055568016268e-06,
+      "loss": 0.4096,
+      "step": 2276
+    },
+    {
+      "epoch": 1.6767304860088366,
+      "grad_norm": 0.5290152430534363,
+      "learning_rate": 4.845772781717852e-06,
+      "loss": 0.3914,
+      "step": 2277
+    },
+    {
+      "epoch": 1.6774668630338734,
+      "grad_norm": 0.41980114579200745,
+      "learning_rate": 4.841490108679242e-06,
+      "loss": 0.3946,
+      "step": 2278
+    },
+    {
+      "epoch": 1.6782032400589102,
+      "grad_norm": 0.3944283127784729,
+      "learning_rate": 4.837207552045509e-06,
+      "loss": 0.3936,
+      "step": 2279
+    },
+    {
+      "epoch": 1.678939617083947,
+      "grad_norm": 0.46467381715774536,
+      "learning_rate": 4.832925114961629e-06,
+      "loss": 0.4159,
+      "step": 2280
+    },
+    {
+      "epoch": 1.6796759941089838,
+      "grad_norm": 0.4479691982269287,
+      "learning_rate": 4.828642800572495e-06,
+      "loss": 0.3904,
+      "step": 2281
+    },
+    {
+      "epoch": 1.6804123711340206,
+      "grad_norm": 0.39899787306785583,
+      "learning_rate": 4.8243606120229095e-06,
+      "loss": 0.4172,
+      "step": 2282
+    },
+    {
+      "epoch": 1.6811487481590575,
+      "grad_norm": 0.4519674479961395,
+      "learning_rate": 4.820078552457584e-06,
+      "loss": 0.401,
+      "step": 2283
+    },
+    {
+      "epoch": 1.6818851251840943,
+      "grad_norm": 0.405024915933609,
+      "learning_rate": 4.815796625021132e-06,
+      "loss": 0.4126,
+      "step": 2284
+    },
+    {
+      "epoch": 1.682621502209131,
+      "grad_norm": 0.42666709423065186,
+      "learning_rate": 4.811514832858072e-06,
+      "loss": 0.4016,
+      "step": 2285
+    },
+    {
+      "epoch": 1.683357879234168,
+      "grad_norm": 0.49751371145248413,
+      "learning_rate": 4.8072331791128244e-06,
+      "loss": 0.4145,
+      "step": 2286
+    },
+    {
+      "epoch": 1.6840942562592047,
+      "grad_norm": 0.48239147663116455,
+      "learning_rate": 4.802951666929704e-06,
+      "loss": 0.4144,
+      "step": 2287
+    },
+    {
+      "epoch": 1.6848306332842415,
+      "grad_norm": 0.41547465324401855,
+      "learning_rate": 4.798670299452926e-06,
+      "loss": 0.4038,
+      "step": 2288
+    },
+    {
+      "epoch": 1.6855670103092784,
+      "grad_norm": 0.4863029420375824,
+      "learning_rate": 4.794389079826594e-06,
+      "loss": 0.3982,
+      "step": 2289
+    },
+    {
+      "epoch": 1.6863033873343152,
+      "grad_norm": 0.48596158623695374,
+      "learning_rate": 4.790108011194709e-06,
+      "loss": 0.4637,
+      "step": 2290
+    },
+    {
+      "epoch": 1.687039764359352,
+      "grad_norm": 0.3679700791835785,
+      "learning_rate": 4.785827096701159e-06,
+      "loss": 0.3796,
+      "step": 2291
+    },
+    {
+      "epoch": 1.6877761413843888,
+      "grad_norm": 0.39074385166168213,
+      "learning_rate": 4.781546339489716e-06,
+      "loss": 0.4275,
+      "step": 2292
+    },
+    {
+      "epoch": 1.6885125184094256,
+      "grad_norm": 0.47816234827041626,
+      "learning_rate": 4.777265742704039e-06,
+      "loss": 0.4048,
+      "step": 2293
+    },
+    {
+      "epoch": 1.6892488954344624,
+      "grad_norm": 0.4297253489494324,
+      "learning_rate": 4.7729853094876714e-06,
+      "loss": 0.3735,
+      "step": 2294
+    },
+    {
+      "epoch": 1.6899852724594993,
+      "grad_norm": 0.3920181095600128,
+      "learning_rate": 4.768705042984031e-06,
+      "loss": 0.4444,
+      "step": 2295
+    },
+    {
+      "epoch": 1.690721649484536,
+      "grad_norm": 0.35152071714401245,
+      "learning_rate": 4.7644249463364205e-06,
+      "loss": 0.4234,
+      "step": 2296
+    },
+    {
+      "epoch": 1.6914580265095729,
+      "grad_norm": 0.4087672233581543,
+      "learning_rate": 4.760145022688007e-06,
+      "loss": 0.4192,
+      "step": 2297
+    },
+    {
+      "epoch": 1.6921944035346097,
+      "grad_norm": 0.40535783767700195,
+      "learning_rate": 4.755865275181843e-06,
+      "loss": 0.4036,
+      "step": 2298
+    },
+    {
+      "epoch": 1.6929307805596465,
+      "grad_norm": 0.4112594425678253,
+      "learning_rate": 4.751585706960842e-06,
+      "loss": 0.4382,
+      "step": 2299
+    },
+    {
+      "epoch": 1.6936671575846833,
+      "grad_norm": 0.40678101778030396,
+      "learning_rate": 4.747306321167791e-06,
+      "loss": 0.4198,
+      "step": 2300
+    },
+    {
+      "epoch": 1.6944035346097202,
+      "grad_norm": 0.38269445300102234,
+      "learning_rate": 4.743027120945342e-06,
+      "loss": 0.4061,
+      "step": 2301
+    },
+    {
+      "epoch": 1.695139911634757,
+      "grad_norm": 0.47511664032936096,
+      "learning_rate": 4.73874810943601e-06,
+      "loss": 0.4329,
+      "step": 2302
+    },
+    {
+      "epoch": 1.6958762886597938,
+      "grad_norm": 0.3984917998313904,
+      "learning_rate": 4.7344692897821714e-06,
+      "loss": 0.4191,
+      "step": 2303
+    },
+    {
+      "epoch": 1.6966126656848306,
+      "grad_norm": 0.37280645966529846,
+      "learning_rate": 4.7301906651260634e-06,
+      "loss": 0.4067,
+      "step": 2304
+    },
+    {
+      "epoch": 1.6973490427098674,
+      "grad_norm": 0.3684954047203064,
+      "learning_rate": 4.725912238609779e-06,
+      "loss": 0.4016,
+      "step": 2305
+    },
+    {
+      "epoch": 1.6980854197349042,
+      "grad_norm": 0.3745935559272766,
+      "learning_rate": 4.7216340133752604e-06,
+      "loss": 0.3998,
+      "step": 2306
+    },
+    {
+      "epoch": 1.698821796759941,
+      "grad_norm": 0.360403448343277,
+      "learning_rate": 4.717355992564311e-06,
+      "loss": 0.3958,
+      "step": 2307
+    },
+    {
+      "epoch": 1.6995581737849779,
+      "grad_norm": 0.3953647017478943,
+      "learning_rate": 4.7130781793185805e-06,
+      "loss": 0.404,
+      "step": 2308
+    },
+    {
+      "epoch": 1.7002945508100147,
+      "grad_norm": 0.40129730105400085,
+      "learning_rate": 4.708800576779564e-06,
+      "loss": 0.4205,
+      "step": 2309
+    },
+    {
+      "epoch": 1.7010309278350515,
+      "grad_norm": 0.3984296917915344,
+      "learning_rate": 4.704523188088604e-06,
+      "loss": 0.3888,
+      "step": 2310
+    },
+    {
+      "epoch": 1.7017673048600883,
+      "grad_norm": 0.37107083201408386,
+      "learning_rate": 4.700246016386887e-06,
+      "loss": 0.3904,
+      "step": 2311
+    },
+    {
+      "epoch": 1.7025036818851251,
+      "grad_norm": 0.38005733489990234,
+      "learning_rate": 4.695969064815436e-06,
+      "loss": 0.4087,
+      "step": 2312
+    },
+    {
+      "epoch": 1.7032400589101622,
+      "grad_norm": 0.37871333956718445,
+      "learning_rate": 4.6916923365151185e-06,
+      "loss": 0.4028,
+      "step": 2313
+    },
+    {
+      "epoch": 1.703976435935199,
+      "grad_norm": 0.3813686668872833,
+      "learning_rate": 4.68741583462663e-06,
+      "loss": 0.3869,
+      "step": 2314
+    },
+    {
+      "epoch": 1.7047128129602358,
+      "grad_norm": 0.37767162919044495,
+      "learning_rate": 4.683139562290506e-06,
+      "loss": 0.4025,
+      "step": 2315
+    },
+    {
+      "epoch": 1.7054491899852726,
+      "grad_norm": 0.38888660073280334,
+      "learning_rate": 4.678863522647114e-06,
+      "loss": 0.3863,
+      "step": 2316
+    },
+    {
+      "epoch": 1.7061855670103094,
+      "grad_norm": 0.4239829182624817,
+      "learning_rate": 4.6745877188366464e-06,
+      "loss": 0.3987,
+      "step": 2317
+    },
+    {
+      "epoch": 1.7069219440353463,
+      "grad_norm": 0.38922736048698425,
+      "learning_rate": 4.670312153999123e-06,
+      "loss": 0.3987,
+      "step": 2318
+    },
+    {
+      "epoch": 1.707658321060383,
+      "grad_norm": 0.42677411437034607,
+      "learning_rate": 4.666036831274392e-06,
+      "loss": 0.3937,
+      "step": 2319
+    },
+    {
+      "epoch": 1.7083946980854199,
+      "grad_norm": 0.43811002373695374,
+      "learning_rate": 4.66176175380212e-06,
+      "loss": 0.4189,
+      "step": 2320
+    },
+    {
+      "epoch": 1.7091310751104567,
+      "grad_norm": 0.41332265734672546,
+      "learning_rate": 4.657486924721797e-06,
+      "loss": 0.4231,
+      "step": 2321
+    },
+    {
+      "epoch": 1.7098674521354935,
+      "grad_norm": 0.41495949029922485,
+      "learning_rate": 4.653212347172723e-06,
+      "loss": 0.3893,
+      "step": 2322
+    },
+    {
+      "epoch": 1.7106038291605303,
+      "grad_norm": 0.42686596512794495,
+      "learning_rate": 4.648938024294023e-06,
+      "loss": 0.4398,
+      "step": 2323
+    },
+    {
+      "epoch": 1.7113402061855671,
+      "grad_norm": 0.38408711552619934,
+      "learning_rate": 4.644663959224629e-06,
+      "loss": 0.4055,
+      "step": 2324
+    },
+    {
+      "epoch": 1.712076583210604,
+      "grad_norm": 0.4538039565086365,
+      "learning_rate": 4.640390155103285e-06,
+      "loss": 0.4302,
+      "step": 2325
+    },
+    {
+      "epoch": 1.7128129602356408,
+      "grad_norm": 0.40232333540916443,
+      "learning_rate": 4.636116615068545e-06,
+      "loss": 0.4076,
+      "step": 2326
+    },
+    {
+      "epoch": 1.7135493372606776,
+      "grad_norm": 0.45923569798469543,
+      "learning_rate": 4.631843342258765e-06,
+      "loss": 0.4284,
+      "step": 2327
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.390129953622818,
+      "learning_rate": 4.627570339812109e-06,
+      "loss": 0.4171,
+      "step": 2328
+    },
+    {
+      "epoch": 1.7150220913107512,
+      "grad_norm": 0.44369444251060486,
+      "learning_rate": 4.623297610866544e-06,
+      "loss": 0.4287,
+      "step": 2329
+    },
+    {
+      "epoch": 1.715758468335788,
+      "grad_norm": 0.42557501792907715,
+      "learning_rate": 4.619025158559829e-06,
+      "loss": 0.4052,
+      "step": 2330
+    },
+    {
+      "epoch": 1.7164948453608249,
+      "grad_norm": 0.3788171708583832,
+      "learning_rate": 4.614752986029524e-06,
+      "loss": 0.3884,
+      "step": 2331
+    },
+    {
+      "epoch": 1.7172312223858617,
+      "grad_norm": 0.3878194987773895,
+      "learning_rate": 4.610481096412985e-06,
+      "loss": 0.4091,
+      "step": 2332
+    },
+    {
+      "epoch": 1.7179675994108985,
+      "grad_norm": 0.40998604893684387,
+      "learning_rate": 4.606209492847356e-06,
+      "loss": 0.379,
+      "step": 2333
+    },
+    {
+      "epoch": 1.7187039764359353,
+      "grad_norm": 0.3821749985218048,
+      "learning_rate": 4.6019381784695774e-06,
+      "loss": 0.4005,
+      "step": 2334
+    },
+    {
+      "epoch": 1.7194403534609721,
+      "grad_norm": 0.35601040720939636,
+      "learning_rate": 4.597667156416371e-06,
+      "loss": 0.3966,
+      "step": 2335
+    },
+    {
+      "epoch": 1.720176730486009,
+      "grad_norm": 0.40410375595092773,
+      "learning_rate": 4.5933964298242465e-06,
+      "loss": 0.3946,
+      "step": 2336
+    },
+    {
+      "epoch": 1.7209131075110458,
+      "grad_norm": 0.37720221281051636,
+      "learning_rate": 4.589126001829497e-06,
+      "loss": 0.3987,
+      "step": 2337
+    },
+    {
+      "epoch": 1.7216494845360826,
+      "grad_norm": 0.3496411144733429,
+      "learning_rate": 4.584855875568198e-06,
+      "loss": 0.4229,
+      "step": 2338
+    },
+    {
+      "epoch": 1.7223858615611194,
+      "grad_norm": 0.3647906482219696,
+      "learning_rate": 4.580586054176196e-06,
+      "loss": 0.4183,
+      "step": 2339
+    },
+    {
+      "epoch": 1.7231222385861562,
+      "grad_norm": 0.36781540513038635,
+      "learning_rate": 4.576316540789122e-06,
+      "loss": 0.3892,
+      "step": 2340
+    },
+    {
+      "epoch": 1.723858615611193,
+      "grad_norm": 0.36236733198165894,
+      "learning_rate": 4.572047338542377e-06,
+      "loss": 0.4132,
+      "step": 2341
+    },
+    {
+      "epoch": 1.7245949926362298,
+      "grad_norm": 0.3683478832244873,
+      "learning_rate": 4.567778450571135e-06,
+      "loss": 0.4394,
+      "step": 2342
+    },
+    {
+      "epoch": 1.7253313696612667,
+      "grad_norm": 0.378169447183609,
+      "learning_rate": 4.563509880010336e-06,
+      "loss": 0.3853,
+      "step": 2343
+    },
+    {
+      "epoch": 1.7260677466863035,
+      "grad_norm": 0.3431897759437561,
+      "learning_rate": 4.559241629994693e-06,
+      "loss": 0.4018,
+      "step": 2344
+    },
+    {
+      "epoch": 1.7268041237113403,
+      "grad_norm": 0.3948942720890045,
+      "learning_rate": 4.554973703658676e-06,
+      "loss": 0.3987,
+      "step": 2345
+    },
+    {
+      "epoch": 1.727540500736377,
+      "grad_norm": 0.36369240283966064,
+      "learning_rate": 4.550706104136523e-06,
+      "loss": 0.3903,
+      "step": 2346
+    },
+    {
+      "epoch": 1.728276877761414,
+      "grad_norm": 0.40486857295036316,
+      "learning_rate": 4.546438834562232e-06,
+      "loss": 0.4185,
+      "step": 2347
+    },
+    {
+      "epoch": 1.7290132547864507,
+      "grad_norm": 0.3831326365470886,
+      "learning_rate": 4.542171898069553e-06,
+      "loss": 0.3756,
+      "step": 2348
+    },
+    {
+      "epoch": 1.7297496318114876,
+      "grad_norm": 0.40144047141075134,
+      "learning_rate": 4.537905297791997e-06,
+      "loss": 0.3718,
+      "step": 2349
+    },
+    {
+      "epoch": 1.7304860088365244,
+      "grad_norm": 0.38407352566719055,
+      "learning_rate": 4.5336390368628265e-06,
+      "loss": 0.4093,
+      "step": 2350
+    },
+    {
+      "epoch": 1.7312223858615612,
+      "grad_norm": 0.37621355056762695,
+      "learning_rate": 4.529373118415053e-06,
+      "loss": 0.4144,
+      "step": 2351
+    },
+    {
+      "epoch": 1.731958762886598,
+      "grad_norm": 0.34284254908561707,
+      "learning_rate": 4.525107545581442e-06,
+      "loss": 0.4234,
+      "step": 2352
+    },
+    {
+      "epoch": 1.7326951399116348,
+      "grad_norm": 0.42767757177352905,
+      "learning_rate": 4.5208423214944975e-06,
+      "loss": 0.4072,
+      "step": 2353
+    },
+    {
+      "epoch": 1.7334315169366716,
+      "grad_norm": 0.3716091215610504,
+      "learning_rate": 4.5165774492864715e-06,
+      "loss": 0.3886,
+      "step": 2354
+    },
+    {
+      "epoch": 1.7341678939617085,
+      "grad_norm": 0.35501131415367126,
+      "learning_rate": 4.512312932089361e-06,
+      "loss": 0.405,
+      "step": 2355
+    },
+    {
+      "epoch": 1.7349042709867453,
+      "grad_norm": 0.4200528562068939,
+      "learning_rate": 4.508048773034895e-06,
+      "loss": 0.4098,
+      "step": 2356
+    },
+    {
+      "epoch": 1.735640648011782,
+      "grad_norm": 0.3673626780509949,
+      "learning_rate": 4.503784975254543e-06,
+      "loss": 0.3913,
+      "step": 2357
+    },
+    {
+      "epoch": 1.736377025036819,
+      "grad_norm": 0.39019590616226196,
+      "learning_rate": 4.499521541879508e-06,
+      "loss": 0.4088,
+      "step": 2358
+    },
+    {
+      "epoch": 1.7371134020618557,
+      "grad_norm": 0.3743307888507843,
+      "learning_rate": 4.495258476040732e-06,
+      "loss": 0.4009,
+      "step": 2359
+    },
+    {
+      "epoch": 1.7378497790868925,
+      "grad_norm": 0.39364027976989746,
+      "learning_rate": 4.4909957808688765e-06,
+      "loss": 0.4161,
+      "step": 2360
+    },
+    {
+      "epoch": 1.7385861561119293,
+      "grad_norm": 0.3424508571624756,
+      "learning_rate": 4.486733459494338e-06,
+      "loss": 0.4315,
+      "step": 2361
+    },
+    {
+      "epoch": 1.7393225331369662,
+      "grad_norm": 0.3977791368961334,
+      "learning_rate": 4.482471515047237e-06,
+      "loss": 0.3828,
+      "step": 2362
+    },
+    {
+      "epoch": 1.740058910162003,
+      "grad_norm": 0.36530202627182007,
+      "learning_rate": 4.478209950657418e-06,
+      "loss": 0.399,
+      "step": 2363
+    },
+    {
+      "epoch": 1.7407952871870398,
+      "grad_norm": 0.40208595991134644,
+      "learning_rate": 4.4739487694544415e-06,
+      "loss": 0.4193,
+      "step": 2364
+    },
+    {
+      "epoch": 1.7415316642120766,
+      "grad_norm": 0.36320358514785767,
+      "learning_rate": 4.469687974567591e-06,
+      "loss": 0.3844,
+      "step": 2365
+    },
+    {
+      "epoch": 1.7422680412371134,
+      "grad_norm": 0.3930068016052246,
+      "learning_rate": 4.465427569125868e-06,
+      "loss": 0.4007,
+      "step": 2366
+    },
+    {
+      "epoch": 1.7430044182621502,
+      "grad_norm": 0.3515762984752655,
+      "learning_rate": 4.461167556257984e-06,
+      "loss": 0.422,
+      "step": 2367
+    },
+    {
+      "epoch": 1.743740795287187,
+      "grad_norm": 0.39239639043807983,
+      "learning_rate": 4.456907939092363e-06,
+      "loss": 0.4105,
+      "step": 2368
+    },
+    {
+      "epoch": 1.7444771723122239,
+      "grad_norm": 0.3900611102581024,
+      "learning_rate": 4.45264872075714e-06,
+      "loss": 0.3909,
+      "step": 2369
+    },
+    {
+      "epoch": 1.7452135493372607,
+      "grad_norm": 0.38840755820274353,
+      "learning_rate": 4.448389904380156e-06,
+      "loss": 0.4077,
+      "step": 2370
+    },
+    {
+      "epoch": 1.7459499263622975,
+      "grad_norm": 0.3633229732513428,
+      "learning_rate": 4.444131493088956e-06,
+      "loss": 0.4119,
+      "step": 2371
+    },
+    {
+      "epoch": 1.7466863033873343,
+      "grad_norm": 0.3752785325050354,
+      "learning_rate": 4.4398734900107935e-06,
+      "loss": 0.3841,
+      "step": 2372
+    },
+    {
+      "epoch": 1.7474226804123711,
+      "grad_norm": 0.42519140243530273,
+      "learning_rate": 4.43561589827261e-06,
+      "loss": 0.4221,
+      "step": 2373
+    },
+    {
+      "epoch": 1.748159057437408,
+      "grad_norm": 0.40567177534103394,
+      "learning_rate": 4.431358721001058e-06,
+      "loss": 0.401,
+      "step": 2374
+    },
+    {
+      "epoch": 1.7488954344624448,
+      "grad_norm": 0.335718035697937,
+      "learning_rate": 4.427101961322475e-06,
+      "loss": 0.3703,
+      "step": 2375
+    },
+    {
+      "epoch": 1.7496318114874816,
+      "grad_norm": 0.39199185371398926,
+      "learning_rate": 4.422845622362901e-06,
+      "loss": 0.3967,
+      "step": 2376
+    },
+    {
+      "epoch": 1.7503681885125184,
+      "grad_norm": 0.3723236918449402,
+      "learning_rate": 4.418589707248061e-06,
+      "loss": 0.3873,
+      "step": 2377
+    },
+    {
+      "epoch": 1.7511045655375552,
+      "grad_norm": 0.41921404004096985,
+      "learning_rate": 4.414334219103369e-06,
+      "loss": 0.3941,
+      "step": 2378
+    },
+    {
+      "epoch": 1.751840942562592,
+      "grad_norm": 0.37618371844291687,
+      "learning_rate": 4.4100791610539285e-06,
+      "loss": 0.3763,
+      "step": 2379
+    },
+    {
+      "epoch": 1.7525773195876289,
+      "grad_norm": 0.4034104645252228,
+      "learning_rate": 4.4058245362245276e-06,
+      "loss": 0.3917,
+      "step": 2380
+    },
+    {
+      "epoch": 1.7533136966126657,
+      "grad_norm": 0.43745455145835876,
+      "learning_rate": 4.401570347739631e-06,
+      "loss": 0.4249,
+      "step": 2381
+    },
+    {
+      "epoch": 1.7540500736377025,
+      "grad_norm": 0.3758779466152191,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.413,
+      "step": 2382
+    },
+    {
+      "epoch": 1.7547864506627393,
+      "grad_norm": 0.3742757737636566,
+      "learning_rate": 4.393063292299618e-06,
+      "loss": 0.4179,
+      "step": 2383
+    },
+    {
+      "epoch": 1.7555228276877761,
+      "grad_norm": 0.4073251783847809,
+      "learning_rate": 4.388810431591829e-06,
+      "loss": 0.4235,
+      "step": 2384
+    },
+    {
+      "epoch": 1.756259204712813,
+      "grad_norm": 0.4208831787109375,
+      "learning_rate": 4.384558019723188e-06,
+      "loss": 0.4119,
+      "step": 2385
+    },
+    {
+      "epoch": 1.7569955817378498,
+      "grad_norm": 0.41460785269737244,
+      "learning_rate": 4.380306059816539e-06,
+      "loss": 0.3966,
+      "step": 2386
+    },
+    {
+      "epoch": 1.7577319587628866,
+      "grad_norm": 0.4289226830005646,
+      "learning_rate": 4.376054554994394e-06,
+      "loss": 0.4164,
+      "step": 2387
+    },
+    {
+      "epoch": 1.7584683357879234,
+      "grad_norm": 0.41257333755493164,
+      "learning_rate": 4.371803508378929e-06,
+      "loss": 0.3822,
+      "step": 2388
+    },
+    {
+      "epoch": 1.7592047128129602,
+      "grad_norm": 0.44865962862968445,
+      "learning_rate": 4.367552923091985e-06,
+      "loss": 0.395,
+      "step": 2389
+    },
+    {
+      "epoch": 1.759941089837997,
+      "grad_norm": 0.37914708256721497,
+      "learning_rate": 4.363302802255062e-06,
+      "loss": 0.4037,
+      "step": 2390
+    },
+    {
+      "epoch": 1.7606774668630338,
+      "grad_norm": 0.3959915339946747,
+      "learning_rate": 4.359053148989319e-06,
+      "loss": 0.3997,
+      "step": 2391
+    },
+    {
+      "epoch": 1.7614138438880707,
+      "grad_norm": 0.5142814517021179,
+      "learning_rate": 4.354803966415576e-06,
+      "loss": 0.4012,
+      "step": 2392
+    },
+    {
+      "epoch": 1.7621502209131075,
+      "grad_norm": 0.4417358338832855,
+      "learning_rate": 4.350555257654302e-06,
+      "loss": 0.3962,
+      "step": 2393
+    },
+    {
+      "epoch": 1.7628865979381443,
+      "grad_norm": 0.41511955857276917,
+      "learning_rate": 4.346307025825621e-06,
+      "loss": 0.407,
+      "step": 2394
+    },
+    {
+      "epoch": 1.763622974963181,
+      "grad_norm": 0.3971770405769348,
+      "learning_rate": 4.342059274049308e-06,
+      "loss": 0.3975,
+      "step": 2395
+    },
+    {
+      "epoch": 1.764359351988218,
+      "grad_norm": 0.43736016750335693,
+      "learning_rate": 4.33781200544478e-06,
+      "loss": 0.3925,
+      "step": 2396
+    },
+    {
+      "epoch": 1.7650957290132547,
+      "grad_norm": 0.3964605927467346,
+      "learning_rate": 4.333565223131107e-06,
+      "loss": 0.3938,
+      "step": 2397
+    },
+    {
+      "epoch": 1.7658321060382915,
+      "grad_norm": 0.4375699758529663,
+      "learning_rate": 4.329318930226993e-06,
+      "loss": 0.4229,
+      "step": 2398
+    },
+    {
+      "epoch": 1.7665684830633284,
+      "grad_norm": 0.34781819581985474,
+      "learning_rate": 4.325073129850791e-06,
+      "loss": 0.3878,
+      "step": 2399
+    },
+    {
+      "epoch": 1.7673048600883652,
+      "grad_norm": 0.39587196707725525,
+      "learning_rate": 4.320827825120485e-06,
+      "loss": 0.4248,
+      "step": 2400
+    },
+    {
+      "epoch": 1.768041237113402,
+      "grad_norm": 0.47028040885925293,
+      "learning_rate": 4.3165830191537016e-06,
+      "loss": 0.3747,
+      "step": 2401
+    },
+    {
+      "epoch": 1.7687776141384388,
+      "grad_norm": 0.3784812390804291,
+      "learning_rate": 4.312338715067697e-06,
+      "loss": 0.3972,
+      "step": 2402
+    },
+    {
+      "epoch": 1.7695139911634756,
+      "grad_norm": 0.45428574085235596,
+      "learning_rate": 4.308094915979359e-06,
+      "loss": 0.4051,
+      "step": 2403
+    },
+    {
+      "epoch": 1.7702503681885124,
+      "grad_norm": 0.4495220184326172,
+      "learning_rate": 4.303851625005205e-06,
+      "loss": 0.4132,
+      "step": 2404
+    },
+    {
+      "epoch": 1.7709867452135493,
+      "grad_norm": 0.39823389053344727,
+      "learning_rate": 4.2996088452613835e-06,
+      "loss": 0.4036,
+      "step": 2405
+    },
+    {
+      "epoch": 1.771723122238586,
+      "grad_norm": 0.4087408483028412,
+      "learning_rate": 4.295366579863658e-06,
+      "loss": 0.3894,
+      "step": 2406
+    },
+    {
+      "epoch": 1.772459499263623,
+      "grad_norm": 0.5459381937980652,
+      "learning_rate": 4.291124831927425e-06,
+      "loss": 0.4391,
+      "step": 2407
+    },
+    {
+      "epoch": 1.7731958762886597,
+      "grad_norm": 0.4444357752799988,
+      "learning_rate": 4.286883604567693e-06,
+      "loss": 0.3851,
+      "step": 2408
+    },
+    {
+      "epoch": 1.7739322533136965,
+      "grad_norm": 0.4461228549480438,
+      "learning_rate": 4.282642900899092e-06,
+      "loss": 0.4251,
+      "step": 2409
+    },
+    {
+      "epoch": 1.7746686303387333,
+      "grad_norm": 0.41404518485069275,
+      "learning_rate": 4.278402724035868e-06,
+      "loss": 0.4339,
+      "step": 2410
+    },
+    {
+      "epoch": 1.7754050073637702,
+      "grad_norm": 0.4226773977279663,
+      "learning_rate": 4.274163077091876e-06,
+      "loss": 0.42,
+      "step": 2411
+    },
+    {
+      "epoch": 1.776141384388807,
+      "grad_norm": 0.44457805156707764,
+      "learning_rate": 4.269923963180587e-06,
+      "loss": 0.4236,
+      "step": 2412
+    },
+    {
+      "epoch": 1.7768777614138438,
+      "grad_norm": 0.4070783853530884,
+      "learning_rate": 4.265685385415077e-06,
+      "loss": 0.3761,
+      "step": 2413
+    },
+    {
+      "epoch": 1.7776141384388806,
+      "grad_norm": 0.3834463953971863,
+      "learning_rate": 4.261447346908032e-06,
+      "loss": 0.4151,
+      "step": 2414
+    },
+    {
+      "epoch": 1.7783505154639174,
+      "grad_norm": 0.41372165083885193,
+      "learning_rate": 4.257209850771734e-06,
+      "loss": 0.3919,
+      "step": 2415
+    },
+    {
+      "epoch": 1.7790868924889542,
+      "grad_norm": 0.4155184328556061,
+      "learning_rate": 4.2529729001180765e-06,
+      "loss": 0.4034,
+      "step": 2416
+    },
+    {
+      "epoch": 1.779823269513991,
+      "grad_norm": 0.3758399784564972,
+      "learning_rate": 4.248736498058547e-06,
+      "loss": 0.4108,
+      "step": 2417
+    },
+    {
+      "epoch": 1.7805596465390279,
+      "grad_norm": 0.38244372606277466,
+      "learning_rate": 4.24450064770423e-06,
+      "loss": 0.3716,
+      "step": 2418
+    },
+    {
+      "epoch": 1.7812960235640647,
+      "grad_norm": 0.4086960256099701,
+      "learning_rate": 4.240265352165806e-06,
+      "loss": 0.3858,
+      "step": 2419
+    },
+    {
+      "epoch": 1.7820324005891015,
+      "grad_norm": 0.37736374139785767,
+      "learning_rate": 4.236030614553552e-06,
+      "loss": 0.4,
+      "step": 2420
+    },
+    {
+      "epoch": 1.7827687776141383,
+      "grad_norm": 0.4141034185886383,
+      "learning_rate": 4.2317964379773265e-06,
+      "loss": 0.4311,
+      "step": 2421
+    },
+    {
+      "epoch": 1.7835051546391751,
+      "grad_norm": 0.37181559205055237,
+      "learning_rate": 4.2275628255465846e-06,
+      "loss": 0.3799,
+      "step": 2422
+    },
+    {
+      "epoch": 1.784241531664212,
+      "grad_norm": 0.3771205246448517,
+      "learning_rate": 4.223329780370359e-06,
+      "loss": 0.4282,
+      "step": 2423
+    },
+    {
+      "epoch": 1.7849779086892488,
+      "grad_norm": 0.4032301604747772,
+      "learning_rate": 4.219097305557274e-06,
+      "loss": 0.4305,
+      "step": 2424
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 0.39853084087371826,
+      "learning_rate": 4.214865404215528e-06,
+      "loss": 0.3973,
+      "step": 2425
+    },
+    {
+      "epoch": 1.7864506627393224,
+      "grad_norm": 0.35678479075431824,
+      "learning_rate": 4.210634079452904e-06,
+      "loss": 0.368,
+      "step": 2426
+    },
+    {
+      "epoch": 1.7871870397643592,
+      "grad_norm": 0.3927466869354248,
+      "learning_rate": 4.206403334376757e-06,
+      "loss": 0.4107,
+      "step": 2427
+    },
+    {
+      "epoch": 1.787923416789396,
+      "grad_norm": 0.3804020285606384,
+      "learning_rate": 4.202173172094022e-06,
+      "loss": 0.3749,
+      "step": 2428
+    },
+    {
+      "epoch": 1.7886597938144329,
+      "grad_norm": 0.3756907284259796,
+      "learning_rate": 4.1979435957111984e-06,
+      "loss": 0.4019,
+      "step": 2429
+    },
+    {
+      "epoch": 1.7893961708394697,
+      "grad_norm": 0.3835069239139557,
+      "learning_rate": 4.193714608334361e-06,
+      "loss": 0.383,
+      "step": 2430
+    },
+    {
+      "epoch": 1.7901325478645065,
+      "grad_norm": 0.41579073667526245,
+      "learning_rate": 4.189486213069152e-06,
+      "loss": 0.4195,
+      "step": 2431
+    },
+    {
+      "epoch": 1.7908689248895433,
+      "grad_norm": 0.39560478925704956,
+      "learning_rate": 4.1852584130207745e-06,
+      "loss": 0.4095,
+      "step": 2432
+    },
+    {
+      "epoch": 1.7916053019145801,
+      "grad_norm": 0.4340824782848358,
+      "learning_rate": 4.181031211293997e-06,
+      "loss": 0.4334,
+      "step": 2433
+    },
+    {
+      "epoch": 1.792341678939617,
+      "grad_norm": 0.41419485211372375,
+      "learning_rate": 4.176804610993149e-06,
+      "loss": 0.4022,
+      "step": 2434
+    },
+    {
+      "epoch": 1.7930780559646537,
+      "grad_norm": 0.40266767144203186,
+      "learning_rate": 4.17257861522212e-06,
+      "loss": 0.4214,
+      "step": 2435
+    },
+    {
+      "epoch": 1.7938144329896906,
+      "grad_norm": 0.35882508754730225,
+      "learning_rate": 4.1683532270843505e-06,
+      "loss": 0.3855,
+      "step": 2436
+    },
+    {
+      "epoch": 1.7945508100147274,
+      "grad_norm": 0.37283796072006226,
+      "learning_rate": 4.16412844968284e-06,
+      "loss": 0.3659,
+      "step": 2437
+    },
+    {
+      "epoch": 1.7952871870397642,
+      "grad_norm": 0.3643818199634552,
+      "learning_rate": 4.1599042861201375e-06,
+      "loss": 0.4015,
+      "step": 2438
+    },
+    {
+      "epoch": 1.796023564064801,
+      "grad_norm": 0.38770896196365356,
+      "learning_rate": 4.155680739498342e-06,
+      "loss": 0.379,
+      "step": 2439
+    },
+    {
+      "epoch": 1.7967599410898378,
+      "grad_norm": 0.3598364293575287,
+      "learning_rate": 4.151457812919094e-06,
+      "loss": 0.4056,
+      "step": 2440
+    },
+    {
+      "epoch": 1.7974963181148749,
+      "grad_norm": 0.39613017439842224,
+      "learning_rate": 4.147235509483587e-06,
+      "loss": 0.3952,
+      "step": 2441
+    },
+    {
+      "epoch": 1.7982326951399117,
+      "grad_norm": 0.4370924234390259,
+      "learning_rate": 4.1430138322925535e-06,
+      "loss": 0.4109,
+      "step": 2442
+    },
+    {
+      "epoch": 1.7989690721649485,
+      "grad_norm": 0.3646906912326813,
+      "learning_rate": 4.138792784446263e-06,
+      "loss": 0.3945,
+      "step": 2443
+    },
+    {
+      "epoch": 1.7997054491899853,
+      "grad_norm": 0.34723278880119324,
+      "learning_rate": 4.134572369044526e-06,
+      "loss": 0.3952,
+      "step": 2444
+    },
+    {
+      "epoch": 1.8004418262150221,
+      "grad_norm": 0.40729156136512756,
+      "learning_rate": 4.1303525891866905e-06,
+      "loss": 0.4143,
+      "step": 2445
+    },
+    {
+      "epoch": 1.801178203240059,
+      "grad_norm": 0.4514944851398468,
+      "learning_rate": 4.126133447971633e-06,
+      "loss": 0.392,
+      "step": 2446
+    },
+    {
+      "epoch": 1.8019145802650958,
+      "grad_norm": 0.3769078254699707,
+      "learning_rate": 4.121914948497764e-06,
+      "loss": 0.3999,
+      "step": 2447
+    },
+    {
+      "epoch": 1.8026509572901326,
+      "grad_norm": 0.39932695031166077,
+      "learning_rate": 4.117697093863023e-06,
+      "loss": 0.4274,
+      "step": 2448
+    },
+    {
+      "epoch": 1.8033873343151694,
+      "grad_norm": 0.3611906170845032,
+      "learning_rate": 4.113479887164873e-06,
+      "loss": 0.3788,
+      "step": 2449
+    },
+    {
+      "epoch": 1.8041237113402062,
+      "grad_norm": 0.38393762707710266,
+      "learning_rate": 4.109263331500305e-06,
+      "loss": 0.4218,
+      "step": 2450
+    },
+    {
+      "epoch": 1.804860088365243,
+      "grad_norm": 0.42800167202949524,
+      "learning_rate": 4.105047429965828e-06,
+      "loss": 0.3984,
+      "step": 2451
+    },
+    {
+      "epoch": 1.8055964653902798,
+      "grad_norm": 0.38685426115989685,
+      "learning_rate": 4.1008321856574745e-06,
+      "loss": 0.4097,
+      "step": 2452
+    },
+    {
+      "epoch": 1.8063328424153167,
+      "grad_norm": 0.3722650408744812,
+      "learning_rate": 4.096617601670793e-06,
+      "loss": 0.4073,
+      "step": 2453
+    },
+    {
+      "epoch": 1.8070692194403535,
+      "grad_norm": 0.3909144997596741,
+      "learning_rate": 4.092403681100844e-06,
+      "loss": 0.403,
+      "step": 2454
+    },
+    {
+      "epoch": 1.8078055964653903,
+      "grad_norm": 0.42803841829299927,
+      "learning_rate": 4.0881904270422045e-06,
+      "loss": 0.3873,
+      "step": 2455
+    },
+    {
+      "epoch": 1.8085419734904271,
+      "grad_norm": 0.42633092403411865,
+      "learning_rate": 4.083977842588963e-06,
+      "loss": 0.3925,
+      "step": 2456
+    },
+    {
+      "epoch": 1.809278350515464,
+      "grad_norm": 0.3902944028377533,
+      "learning_rate": 4.079765930834714e-06,
+      "loss": 0.4063,
+      "step": 2457
+    },
+    {
+      "epoch": 1.8100147275405007,
+      "grad_norm": 0.3900681436061859,
+      "learning_rate": 4.075554694872554e-06,
+      "loss": 0.406,
+      "step": 2458
+    },
+    {
+      "epoch": 1.8107511045655376,
+      "grad_norm": 0.412332147359848,
+      "learning_rate": 4.071344137795091e-06,
+      "loss": 0.4169,
+      "step": 2459
+    },
+    {
+      "epoch": 1.8114874815905744,
+      "grad_norm": 0.39725714921951294,
+      "learning_rate": 4.067134262694431e-06,
+      "loss": 0.4163,
+      "step": 2460
+    },
+    {
+      "epoch": 1.8122238586156112,
+      "grad_norm": 0.36094245314598083,
+      "learning_rate": 4.062925072662177e-06,
+      "loss": 0.3998,
+      "step": 2461
+    },
+    {
+      "epoch": 1.812960235640648,
+      "grad_norm": 0.35459452867507935,
+      "learning_rate": 4.0587165707894326e-06,
+      "loss": 0.4008,
+      "step": 2462
+    },
+    {
+      "epoch": 1.8136966126656848,
+      "grad_norm": 0.3944514989852905,
+      "learning_rate": 4.054508760166795e-06,
+      "loss": 0.4151,
+      "step": 2463
+    },
+    {
+      "epoch": 1.8144329896907216,
+      "grad_norm": 0.41395866870880127,
+      "learning_rate": 4.050301643884352e-06,
+      "loss": 0.4049,
+      "step": 2464
+    },
+    {
+      "epoch": 1.8151693667157585,
+      "grad_norm": 0.3743531405925751,
+      "learning_rate": 4.046095225031683e-06,
+      "loss": 0.4226,
+      "step": 2465
+    },
+    {
+      "epoch": 1.8159057437407953,
+      "grad_norm": 0.4096841812133789,
+      "learning_rate": 4.0418895066978536e-06,
+      "loss": 0.3879,
+      "step": 2466
+    },
+    {
+      "epoch": 1.816642120765832,
+      "grad_norm": 0.41024044156074524,
+      "learning_rate": 4.037684491971417e-06,
+      "loss": 0.4129,
+      "step": 2467
+    },
+    {
+      "epoch": 1.817378497790869,
+      "grad_norm": 0.36897313594818115,
+      "learning_rate": 4.033480183940412e-06,
+      "loss": 0.391,
+      "step": 2468
+    },
+    {
+      "epoch": 1.8181148748159057,
+      "grad_norm": 0.3991081416606903,
+      "learning_rate": 4.029276585692349e-06,
+      "loss": 0.4263,
+      "step": 2469
+    },
+    {
+      "epoch": 1.8188512518409425,
+      "grad_norm": 0.43920043110847473,
+      "learning_rate": 4.0250737003142294e-06,
+      "loss": 0.433,
+      "step": 2470
+    },
+    {
+      "epoch": 1.8195876288659794,
+      "grad_norm": 0.47898420691490173,
+      "learning_rate": 4.0208715308925235e-06,
+      "loss": 0.4085,
+      "step": 2471
+    },
+    {
+      "epoch": 1.8203240058910162,
+      "grad_norm": 0.37598884105682373,
+      "learning_rate": 4.016670080513176e-06,
+      "loss": 0.4126,
+      "step": 2472
+    },
+    {
+      "epoch": 1.821060382916053,
+      "grad_norm": 0.36710143089294434,
+      "learning_rate": 4.012469352261608e-06,
+      "loss": 0.3943,
+      "step": 2473
+    },
+    {
+      "epoch": 1.8217967599410898,
+      "grad_norm": 0.42525747418403625,
+      "learning_rate": 4.0082693492227035e-06,
+      "loss": 0.3982,
+      "step": 2474
+    },
+    {
+      "epoch": 1.8225331369661266,
+      "grad_norm": 0.39572814106941223,
+      "learning_rate": 4.004070074480821e-06,
+      "loss": 0.3859,
+      "step": 2475
+    },
+    {
+      "epoch": 1.8232695139911634,
+      "grad_norm": 0.38209137320518494,
+      "learning_rate": 3.999871531119779e-06,
+      "loss": 0.3998,
+      "step": 2476
+    },
+    {
+      "epoch": 1.8240058910162003,
+      "grad_norm": 0.36567169427871704,
+      "learning_rate": 3.995673722222861e-06,
+      "loss": 0.397,
+      "step": 2477
+    },
+    {
+      "epoch": 1.824742268041237,
+      "grad_norm": 0.4170894920825958,
+      "learning_rate": 3.991476650872813e-06,
+      "loss": 0.4181,
+      "step": 2478
+    },
+    {
+      "epoch": 1.8254786450662739,
+      "grad_norm": 0.43204647302627563,
+      "learning_rate": 3.987280320151835e-06,
+      "loss": 0.4415,
+      "step": 2479
+    },
+    {
+      "epoch": 1.8262150220913107,
+      "grad_norm": 0.39173027873039246,
+      "learning_rate": 3.983084733141588e-06,
+      "loss": 0.3853,
+      "step": 2480
+    },
+    {
+      "epoch": 1.8269513991163475,
+      "grad_norm": 0.4281696081161499,
+      "learning_rate": 3.978889892923183e-06,
+      "loss": 0.4305,
+      "step": 2481
+    },
+    {
+      "epoch": 1.8276877761413843,
+      "grad_norm": 0.44608938694000244,
+      "learning_rate": 3.974695802577184e-06,
+      "loss": 0.4322,
+      "step": 2482
+    },
+    {
+      "epoch": 1.8284241531664214,
+      "grad_norm": 0.4017941951751709,
+      "learning_rate": 3.970502465183602e-06,
+      "loss": 0.4158,
+      "step": 2483
+    },
+    {
+      "epoch": 1.8291605301914582,
+      "grad_norm": 0.4041925072669983,
+      "learning_rate": 3.966309883821901e-06,
+      "loss": 0.4031,
+      "step": 2484
+    },
+    {
+      "epoch": 1.829896907216495,
+      "grad_norm": 0.4443046748638153,
+      "learning_rate": 3.962118061570982e-06,
+      "loss": 0.3905,
+      "step": 2485
+    },
+    {
+      "epoch": 1.8306332842415318,
+      "grad_norm": 0.45764559507369995,
+      "learning_rate": 3.957927001509197e-06,
+      "loss": 0.399,
+      "step": 2486
+    },
+    {
+      "epoch": 1.8313696612665686,
+      "grad_norm": 0.4235036373138428,
+      "learning_rate": 3.953736706714331e-06,
+      "loss": 0.396,
+      "step": 2487
+    },
+    {
+      "epoch": 1.8321060382916055,
+      "grad_norm": 0.46383094787597656,
+      "learning_rate": 3.94954718026361e-06,
+      "loss": 0.4079,
+      "step": 2488
+    },
+    {
+      "epoch": 1.8328424153166423,
+      "grad_norm": 0.38650867342948914,
+      "learning_rate": 3.945358425233697e-06,
+      "loss": 0.4062,
+      "step": 2489
+    },
+    {
+      "epoch": 1.833578792341679,
+      "grad_norm": 0.40243032574653625,
+      "learning_rate": 3.941170444700688e-06,
+      "loss": 0.3963,
+      "step": 2490
+    },
+    {
+      "epoch": 1.834315169366716,
+      "grad_norm": 0.4477875530719757,
+      "learning_rate": 3.9369832417401055e-06,
+      "loss": 0.3711,
+      "step": 2491
+    },
+    {
+      "epoch": 1.8350515463917527,
+      "grad_norm": 0.4357861578464508,
+      "learning_rate": 3.9327968194269074e-06,
+      "loss": 0.4004,
+      "step": 2492
+    },
+    {
+      "epoch": 1.8357879234167895,
+      "grad_norm": 0.34190165996551514,
+      "learning_rate": 3.928611180835476e-06,
+      "loss": 0.4045,
+      "step": 2493
+    },
+    {
+      "epoch": 1.8365243004418264,
+      "grad_norm": 0.40084031224250793,
+      "learning_rate": 3.924426329039616e-06,
+      "loss": 0.3943,
+      "step": 2494
+    },
+    {
+      "epoch": 1.8372606774668632,
+      "grad_norm": 0.43317151069641113,
+      "learning_rate": 3.920242267112557e-06,
+      "loss": 0.4437,
+      "step": 2495
+    },
+    {
+      "epoch": 1.8379970544919,
+      "grad_norm": 0.42097267508506775,
+      "learning_rate": 3.916058998126949e-06,
+      "loss": 0.4135,
+      "step": 2496
+    },
+    {
+      "epoch": 1.8387334315169368,
+      "grad_norm": 0.43369096517562866,
+      "learning_rate": 3.911876525154857e-06,
+      "loss": 0.4216,
+      "step": 2497
+    },
+    {
+      "epoch": 1.8394698085419736,
+      "grad_norm": 0.4318563640117645,
+      "learning_rate": 3.907694851267764e-06,
+      "loss": 0.396,
+      "step": 2498
+    },
+    {
+      "epoch": 1.8402061855670104,
+      "grad_norm": 0.39135801792144775,
+      "learning_rate": 3.903513979536563e-06,
+      "loss": 0.4167,
+      "step": 2499
+    },
+    {
+      "epoch": 1.8409425625920472,
+      "grad_norm": 0.4389367997646332,
+      "learning_rate": 3.899333913031561e-06,
+      "loss": 0.412,
+      "step": 2500
+    },
+    {
+      "epoch": 1.841678939617084,
+      "grad_norm": 0.4623863995075226,
+      "learning_rate": 3.895154654822471e-06,
+      "loss": 0.3903,
+      "step": 2501
+    },
+    {
+      "epoch": 1.8424153166421209,
+      "grad_norm": 0.40219375491142273,
+      "learning_rate": 3.890976207978416e-06,
+      "loss": 0.4075,
+      "step": 2502
+    },
+    {
+      "epoch": 1.8431516936671577,
+      "grad_norm": 0.36119547486305237,
+      "learning_rate": 3.8867985755679206e-06,
+      "loss": 0.4064,
+      "step": 2503
+    },
+    {
+      "epoch": 1.8438880706921945,
+      "grad_norm": 0.4446427524089813,
+      "learning_rate": 3.882621760658911e-06,
+      "loss": 0.4002,
+      "step": 2504
+    },
+    {
+      "epoch": 1.8446244477172313,
+      "grad_norm": 0.38301101326942444,
+      "learning_rate": 3.878445766318714e-06,
+      "loss": 0.3946,
+      "step": 2505
+    },
+    {
+      "epoch": 1.8453608247422681,
+      "grad_norm": 0.39597806334495544,
+      "learning_rate": 3.874270595614057e-06,
+      "loss": 0.3955,
+      "step": 2506
+    },
+    {
+      "epoch": 1.846097201767305,
+      "grad_norm": 0.36354929208755493,
+      "learning_rate": 3.870096251611053e-06,
+      "loss": 0.415,
+      "step": 2507
+    },
+    {
+      "epoch": 1.8468335787923418,
+      "grad_norm": 0.4049472510814667,
+      "learning_rate": 3.865922737375219e-06,
+      "loss": 0.4134,
+      "step": 2508
+    },
+    {
+      "epoch": 1.8475699558173786,
+      "grad_norm": 0.42379382252693176,
+      "learning_rate": 3.861750055971455e-06,
+      "loss": 0.436,
+      "step": 2509
+    },
+    {
+      "epoch": 1.8483063328424154,
+      "grad_norm": 0.3723948001861572,
+      "learning_rate": 3.857578210464053e-06,
+      "loss": 0.3914,
+      "step": 2510
+    },
+    {
+      "epoch": 1.8490427098674522,
+      "grad_norm": 0.4040578007698059,
+      "learning_rate": 3.8534072039166915e-06,
+      "loss": 0.3967,
+      "step": 2511
+    },
+    {
+      "epoch": 1.849779086892489,
+      "grad_norm": 0.3998672664165497,
+      "learning_rate": 3.849237039392429e-06,
+      "loss": 0.3973,
+      "step": 2512
+    },
+    {
+      "epoch": 1.8505154639175259,
+      "grad_norm": 0.3948405086994171,
+      "learning_rate": 3.845067719953711e-06,
+      "loss": 0.4459,
+      "step": 2513
+    },
+    {
+      "epoch": 1.8512518409425627,
+      "grad_norm": 0.38502955436706543,
+      "learning_rate": 3.840899248662358e-06,
+      "loss": 0.4169,
+      "step": 2514
+    },
+    {
+      "epoch": 1.8519882179675995,
+      "grad_norm": 0.34823498129844666,
+      "learning_rate": 3.836731628579573e-06,
+      "loss": 0.4141,
+      "step": 2515
+    },
+    {
+      "epoch": 1.8527245949926363,
+      "grad_norm": 0.4114006459712982,
+      "learning_rate": 3.832564862765924e-06,
+      "loss": 0.4317,
+      "step": 2516
+    },
+    {
+      "epoch": 1.8534609720176731,
+      "grad_norm": 0.34312447905540466,
+      "learning_rate": 3.828398954281361e-06,
+      "loss": 0.3961,
+      "step": 2517
+    },
+    {
+      "epoch": 1.85419734904271,
+      "grad_norm": 0.3569032549858093,
+      "learning_rate": 3.8242339061852035e-06,
+      "loss": 0.4025,
+      "step": 2518
+    },
+    {
+      "epoch": 1.8549337260677468,
+      "grad_norm": 0.37142735719680786,
+      "learning_rate": 3.8200697215361336e-06,
+      "loss": 0.3778,
+      "step": 2519
+    },
+    {
+      "epoch": 1.8556701030927836,
+      "grad_norm": 0.39146125316619873,
+      "learning_rate": 3.815906403392203e-06,
+      "loss": 0.4004,
+      "step": 2520
+    },
+    {
+      "epoch": 1.8564064801178204,
+      "grad_norm": 0.3406389057636261,
+      "learning_rate": 3.8117439548108293e-06,
+      "loss": 0.4247,
+      "step": 2521
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.3523464500904083,
+      "learning_rate": 3.8075823788487863e-06,
+      "loss": 0.4202,
+      "step": 2522
+    },
+    {
+      "epoch": 1.857879234167894,
+      "grad_norm": 0.38543546199798584,
+      "learning_rate": 3.803421678562213e-06,
+      "loss": 0.4254,
+      "step": 2523
+    },
+    {
+      "epoch": 1.8586156111929308,
+      "grad_norm": 0.3914850056171417,
+      "learning_rate": 3.799261857006597e-06,
+      "loss": 0.402,
+      "step": 2524
+    },
+    {
+      "epoch": 1.8593519882179677,
+      "grad_norm": 0.3607421815395355,
+      "learning_rate": 3.7951029172367883e-06,
+      "loss": 0.396,
+      "step": 2525
+    },
+    {
+      "epoch": 1.8600883652430045,
+      "grad_norm": 0.3871685862541199,
+      "learning_rate": 3.790944862306988e-06,
+      "loss": 0.4304,
+      "step": 2526
+    },
+    {
+      "epoch": 1.8608247422680413,
+      "grad_norm": 0.4246087074279785,
+      "learning_rate": 3.786787695270743e-06,
+      "loss": 0.4147,
+      "step": 2527
+    },
+    {
+      "epoch": 1.861561119293078,
+      "grad_norm": 0.35566386580467224,
+      "learning_rate": 3.7826314191809522e-06,
+      "loss": 0.4074,
+      "step": 2528
+    },
+    {
+      "epoch": 1.862297496318115,
+      "grad_norm": 0.3580811321735382,
+      "learning_rate": 3.778476037089861e-06,
+      "loss": 0.4046,
+      "step": 2529
+    },
+    {
+      "epoch": 1.8630338733431517,
+      "grad_norm": 0.35247814655303955,
+      "learning_rate": 3.774321552049054e-06,
+      "loss": 0.3959,
+      "step": 2530
+    },
+    {
+      "epoch": 1.8637702503681886,
+      "grad_norm": 0.37611591815948486,
+      "learning_rate": 3.7701679671094602e-06,
+      "loss": 0.4283,
+      "step": 2531
+    },
+    {
+      "epoch": 1.8645066273932254,
+      "grad_norm": 0.3860566020011902,
+      "learning_rate": 3.7660152853213494e-06,
+      "loss": 0.4123,
+      "step": 2532
+    },
+    {
+      "epoch": 1.8652430044182622,
+      "grad_norm": 0.37459200620651245,
+      "learning_rate": 3.7618635097343225e-06,
+      "loss": 0.4377,
+      "step": 2533
+    },
+    {
+      "epoch": 1.865979381443299,
+      "grad_norm": 0.37669306993484497,
+      "learning_rate": 3.7577126433973176e-06,
+      "loss": 0.4092,
+      "step": 2534
+    },
+    {
+      "epoch": 1.8667157584683358,
+      "grad_norm": 0.3490535020828247,
+      "learning_rate": 3.7535626893586062e-06,
+      "loss": 0.4069,
+      "step": 2535
+    },
+    {
+      "epoch": 1.8674521354933726,
+      "grad_norm": 0.3674035370349884,
+      "learning_rate": 3.749413650665792e-06,
+      "loss": 0.4336,
+      "step": 2536
+    },
+    {
+      "epoch": 1.8681885125184094,
+      "grad_norm": 0.3696489930152893,
+      "learning_rate": 3.7452655303657993e-06,
+      "loss": 0.4025,
+      "step": 2537
+    },
+    {
+      "epoch": 1.8689248895434463,
+      "grad_norm": 0.3966639041900635,
+      "learning_rate": 3.7411183315048847e-06,
+      "loss": 0.4173,
+      "step": 2538
+    },
+    {
+      "epoch": 1.869661266568483,
+      "grad_norm": 0.3363037705421448,
+      "learning_rate": 3.736972057128626e-06,
+      "loss": 0.4027,
+      "step": 2539
+    },
+    {
+      "epoch": 1.87039764359352,
+      "grad_norm": 0.35173100233078003,
+      "learning_rate": 3.732826710281923e-06,
+      "loss": 0.3908,
+      "step": 2540
+    },
+    {
+      "epoch": 1.8711340206185567,
+      "grad_norm": 0.3510563373565674,
+      "learning_rate": 3.728682294008988e-06,
+      "loss": 0.3961,
+      "step": 2541
+    },
+    {
+      "epoch": 1.8718703976435935,
+      "grad_norm": 0.3618358373641968,
+      "learning_rate": 3.7245388113533596e-06,
+      "loss": 0.3936,
+      "step": 2542
+    },
+    {
+      "epoch": 1.8726067746686303,
+      "grad_norm": 0.33672353625297546,
+      "learning_rate": 3.7203962653578853e-06,
+      "loss": 0.4081,
+      "step": 2543
+    },
+    {
+      "epoch": 1.8733431516936672,
+      "grad_norm": 0.36822593212127686,
+      "learning_rate": 3.7162546590647254e-06,
+      "loss": 0.4108,
+      "step": 2544
+    },
+    {
+      "epoch": 1.874079528718704,
+      "grad_norm": 0.3691384196281433,
+      "learning_rate": 3.7121139955153497e-06,
+      "loss": 0.379,
+      "step": 2545
+    },
+    {
+      "epoch": 1.8748159057437408,
+      "grad_norm": 0.4163464307785034,
+      "learning_rate": 3.7079742777505373e-06,
+      "loss": 0.4063,
+      "step": 2546
+    },
+    {
+      "epoch": 1.8755522827687776,
+      "grad_norm": 0.3840535879135132,
+      "learning_rate": 3.7038355088103726e-06,
+      "loss": 0.403,
+      "step": 2547
+    },
+    {
+      "epoch": 1.8762886597938144,
+      "grad_norm": 0.3876000642776489,
+      "learning_rate": 3.699697691734243e-06,
+      "loss": 0.4056,
+      "step": 2548
+    },
+    {
+      "epoch": 1.8770250368188512,
+      "grad_norm": 0.3980104625225067,
+      "learning_rate": 3.695560829560832e-06,
+      "loss": 0.3986,
+      "step": 2549
+    },
+    {
+      "epoch": 1.877761413843888,
+      "grad_norm": 0.3940967917442322,
+      "learning_rate": 3.691424925328129e-06,
+      "loss": 0.3895,
+      "step": 2550
+    },
+    {
+      "epoch": 1.8784977908689249,
+      "grad_norm": 0.37271252274513245,
+      "learning_rate": 3.687289982073419e-06,
+      "loss": 0.4393,
+      "step": 2551
+    },
+    {
+      "epoch": 1.8792341678939617,
+      "grad_norm": 0.37549808621406555,
+      "learning_rate": 3.683156002833276e-06,
+      "loss": 0.4248,
+      "step": 2552
+    },
+    {
+      "epoch": 1.8799705449189985,
+      "grad_norm": 0.3900524079799652,
+      "learning_rate": 3.6790229906435706e-06,
+      "loss": 0.4255,
+      "step": 2553
+    },
+    {
+      "epoch": 1.8807069219440353,
+      "grad_norm": 0.36957311630249023,
+      "learning_rate": 3.674890948539463e-06,
+      "loss": 0.404,
+      "step": 2554
+    },
+    {
+      "epoch": 1.8814432989690721,
+      "grad_norm": 0.37756621837615967,
+      "learning_rate": 3.670759879555399e-06,
+      "loss": 0.4144,
+      "step": 2555
+    },
+    {
+      "epoch": 1.882179675994109,
+      "grad_norm": 0.4113319516181946,
+      "learning_rate": 3.666629786725111e-06,
+      "loss": 0.3993,
+      "step": 2556
+    },
+    {
+      "epoch": 1.8829160530191458,
+      "grad_norm": 0.4135149419307709,
+      "learning_rate": 3.6625006730816157e-06,
+      "loss": 0.4202,
+      "step": 2557
+    },
+    {
+      "epoch": 1.8836524300441826,
+      "grad_norm": 0.3912547826766968,
+      "learning_rate": 3.6583725416572093e-06,
+      "loss": 0.4029,
+      "step": 2558
+    },
+    {
+      "epoch": 1.8843888070692194,
+      "grad_norm": 0.36167773604393005,
+      "learning_rate": 3.6542453954834632e-06,
+      "loss": 0.4039,
+      "step": 2559
+    },
+    {
+      "epoch": 1.8851251840942562,
+      "grad_norm": 0.3650279641151428,
+      "learning_rate": 3.650119237591232e-06,
+      "loss": 0.4051,
+      "step": 2560
+    },
+    {
+      "epoch": 1.885861561119293,
+      "grad_norm": 0.358697772026062,
+      "learning_rate": 3.6459940710106414e-06,
+      "loss": 0.3967,
+      "step": 2561
+    },
+    {
+      "epoch": 1.8865979381443299,
+      "grad_norm": 0.41603848338127136,
+      "learning_rate": 3.6418698987710872e-06,
+      "loss": 0.3998,
+      "step": 2562
+    },
+    {
+      "epoch": 1.8873343151693667,
+      "grad_norm": 0.3741896450519562,
+      "learning_rate": 3.637746723901238e-06,
+      "loss": 0.4212,
+      "step": 2563
+    },
+    {
+      "epoch": 1.8880706921944035,
+      "grad_norm": 0.39252546429634094,
+      "learning_rate": 3.6336245494290305e-06,
+      "loss": 0.387,
+      "step": 2564
+    },
+    {
+      "epoch": 1.8888070692194403,
+      "grad_norm": 0.35629981756210327,
+      "learning_rate": 3.6295033783816636e-06,
+      "loss": 0.3794,
+      "step": 2565
+    },
+    {
+      "epoch": 1.8895434462444771,
+      "grad_norm": 0.3638868033885956,
+      "learning_rate": 3.6253832137856e-06,
+      "loss": 0.4164,
+      "step": 2566
+    },
+    {
+      "epoch": 1.890279823269514,
+      "grad_norm": 0.380978524684906,
+      "learning_rate": 3.621264058666564e-06,
+      "loss": 0.3887,
+      "step": 2567
+    },
+    {
+      "epoch": 1.8910162002945508,
+      "grad_norm": 0.3995063304901123,
+      "learning_rate": 3.6171459160495393e-06,
+      "loss": 0.408,
+      "step": 2568
+    },
+    {
+      "epoch": 1.8917525773195876,
+      "grad_norm": 0.4066588878631592,
+      "learning_rate": 3.6130287889587665e-06,
+      "loss": 0.4023,
+      "step": 2569
+    },
+    {
+      "epoch": 1.8924889543446244,
+      "grad_norm": 0.4209347069263458,
+      "learning_rate": 3.6089126804177373e-06,
+      "loss": 0.3898,
+      "step": 2570
+    },
+    {
+      "epoch": 1.8932253313696612,
+      "grad_norm": 0.41782402992248535,
+      "learning_rate": 3.6047975934491983e-06,
+      "loss": 0.4186,
+      "step": 2571
+    },
+    {
+      "epoch": 1.893961708394698,
+      "grad_norm": 0.41970953345298767,
+      "learning_rate": 3.6006835310751464e-06,
+      "loss": 0.4078,
+      "step": 2572
+    },
+    {
+      "epoch": 1.8946980854197348,
+      "grad_norm": 0.3769712746143341,
+      "learning_rate": 3.596570496316822e-06,
+      "loss": 0.4206,
+      "step": 2573
+    },
+    {
+      "epoch": 1.8954344624447717,
+      "grad_norm": 0.38179120421409607,
+      "learning_rate": 3.592458492194717e-06,
+      "loss": 0.3897,
+      "step": 2574
+    },
+    {
+      "epoch": 1.8961708394698085,
+      "grad_norm": 0.42509886622428894,
+      "learning_rate": 3.5883475217285592e-06,
+      "loss": 0.4142,
+      "step": 2575
+    },
+    {
+      "epoch": 1.8969072164948453,
+      "grad_norm": 0.38956883549690247,
+      "learning_rate": 3.5842375879373237e-06,
+      "loss": 0.3984,
+      "step": 2576
+    },
+    {
+      "epoch": 1.897643593519882,
+      "grad_norm": 0.34472212195396423,
+      "learning_rate": 3.5801286938392195e-06,
+      "loss": 0.4137,
+      "step": 2577
+    },
+    {
+      "epoch": 1.898379970544919,
+      "grad_norm": 0.39913445711135864,
+      "learning_rate": 3.5760208424516957e-06,
+      "loss": 0.3989,
+      "step": 2578
+    },
+    {
+      "epoch": 1.8991163475699557,
+      "grad_norm": 0.38616591691970825,
+      "learning_rate": 3.571914036791435e-06,
+      "loss": 0.384,
+      "step": 2579
+    },
+    {
+      "epoch": 1.8998527245949925,
+      "grad_norm": 0.35932883620262146,
+      "learning_rate": 3.5678082798743498e-06,
+      "loss": 0.3966,
+      "step": 2580
+    },
+    {
+      "epoch": 1.9005891016200294,
+      "grad_norm": 0.40238866209983826,
+      "learning_rate": 3.5637035747155835e-06,
+      "loss": 0.4105,
+      "step": 2581
+    },
+    {
+      "epoch": 1.9013254786450662,
+      "grad_norm": 0.39923685789108276,
+      "learning_rate": 3.5595999243295114e-06,
+      "loss": 0.4016,
+      "step": 2582
+    },
+    {
+      "epoch": 1.902061855670103,
+      "grad_norm": 0.428325891494751,
+      "learning_rate": 3.5554973317297255e-06,
+      "loss": 0.3744,
+      "step": 2583
+    },
+    {
+      "epoch": 1.9027982326951398,
+      "grad_norm": 0.3603360652923584,
+      "learning_rate": 3.5513957999290483e-06,
+      "loss": 0.3955,
+      "step": 2584
+    },
+    {
+      "epoch": 1.9035346097201766,
+      "grad_norm": 0.42727532982826233,
+      "learning_rate": 3.5472953319395196e-06,
+      "loss": 0.4234,
+      "step": 2585
+    },
+    {
+      "epoch": 1.9042709867452134,
+      "grad_norm": 0.39020195603370667,
+      "learning_rate": 3.5431959307724e-06,
+      "loss": 0.4095,
+      "step": 2586
+    },
+    {
+      "epoch": 1.9050073637702503,
+      "grad_norm": 0.3618599474430084,
+      "learning_rate": 3.539097599438167e-06,
+      "loss": 0.3776,
+      "step": 2587
+    },
+    {
+      "epoch": 1.905743740795287,
+      "grad_norm": 0.37120410799980164,
+      "learning_rate": 3.5350003409465085e-06,
+      "loss": 0.4031,
+      "step": 2588
+    },
+    {
+      "epoch": 1.906480117820324,
+      "grad_norm": 0.4274289309978485,
+      "learning_rate": 3.530904158306329e-06,
+      "loss": 0.3971,
+      "step": 2589
+    },
+    {
+      "epoch": 1.9072164948453607,
+      "grad_norm": 0.36874106526374817,
+      "learning_rate": 3.526809054525744e-06,
+      "loss": 0.3911,
+      "step": 2590
+    },
+    {
+      "epoch": 1.9079528718703975,
+      "grad_norm": 0.3638487756252289,
+      "learning_rate": 3.522715032612069e-06,
+      "loss": 0.405,
+      "step": 2591
+    },
+    {
+      "epoch": 1.9086892488954343,
+      "grad_norm": 0.3634064495563507,
+      "learning_rate": 3.518622095571831e-06,
+      "loss": 0.4145,
+      "step": 2592
+    },
+    {
+      "epoch": 1.9094256259204712,
+      "grad_norm": 0.3739914894104004,
+      "learning_rate": 3.5145302464107612e-06,
+      "loss": 0.3862,
+      "step": 2593
+    },
+    {
+      "epoch": 1.910162002945508,
+      "grad_norm": 0.39826780557632446,
+      "learning_rate": 3.510439488133789e-06,
+      "loss": 0.4123,
+      "step": 2594
+    },
+    {
+      "epoch": 1.9108983799705448,
+      "grad_norm": 0.37097451090812683,
+      "learning_rate": 3.506349823745043e-06,
+      "loss": 0.3913,
+      "step": 2595
+    },
+    {
+      "epoch": 1.9116347569955816,
+      "grad_norm": 0.3629193603992462,
+      "learning_rate": 3.5022612562478507e-06,
+      "loss": 0.4055,
+      "step": 2596
+    },
+    {
+      "epoch": 1.9123711340206184,
+      "grad_norm": 0.368109792470932,
+      "learning_rate": 3.498173788644732e-06,
+      "loss": 0.3793,
+      "step": 2597
+    },
+    {
+      "epoch": 1.9131075110456552,
+      "grad_norm": 0.418338418006897,
+      "learning_rate": 3.494087423937399e-06,
+      "loss": 0.3781,
+      "step": 2598
+    },
+    {
+      "epoch": 1.913843888070692,
+      "grad_norm": 0.35174235701560974,
+      "learning_rate": 3.4900021651267557e-06,
+      "loss": 0.405,
+      "step": 2599
+    },
+    {
+      "epoch": 1.9145802650957289,
+      "grad_norm": 0.3879396915435791,
+      "learning_rate": 3.485918015212891e-06,
+      "loss": 0.383,
+      "step": 2600
+    },
+    {
+      "epoch": 1.9153166421207657,
+      "grad_norm": 0.34615907073020935,
+      "learning_rate": 3.481834977195081e-06,
+      "loss": 0.3978,
+      "step": 2601
+    },
+    {
+      "epoch": 1.9160530191458025,
+      "grad_norm": 0.36179783940315247,
+      "learning_rate": 3.4777530540717875e-06,
+      "loss": 0.4361,
+      "step": 2602
+    },
+    {
+      "epoch": 1.9167893961708393,
+      "grad_norm": 0.37738144397735596,
+      "learning_rate": 3.4736722488406493e-06,
+      "loss": 0.409,
+      "step": 2603
+    },
+    {
+      "epoch": 1.9175257731958761,
+      "grad_norm": 0.37794360518455505,
+      "learning_rate": 3.4695925644984885e-06,
+      "loss": 0.4211,
+      "step": 2604
+    },
+    {
+      "epoch": 1.918262150220913,
+      "grad_norm": 0.40854412317276,
+      "learning_rate": 3.465514004041301e-06,
+      "loss": 0.4055,
+      "step": 2605
+    },
+    {
+      "epoch": 1.9189985272459498,
+      "grad_norm": 0.3971635103225708,
+      "learning_rate": 3.461436570464258e-06,
+      "loss": 0.3971,
+      "step": 2606
+    },
+    {
+      "epoch": 1.9197349042709866,
+      "grad_norm": 0.3623662292957306,
+      "learning_rate": 3.4573602667617056e-06,
+      "loss": 0.4185,
+      "step": 2607
+    },
+    {
+      "epoch": 1.9204712812960234,
+      "grad_norm": 0.4009900391101837,
+      "learning_rate": 3.453285095927154e-06,
+      "loss": 0.4067,
+      "step": 2608
+    },
+    {
+      "epoch": 1.9212076583210602,
+      "grad_norm": 0.34463950991630554,
+      "learning_rate": 3.4492110609532892e-06,
+      "loss": 0.4056,
+      "step": 2609
+    },
+    {
+      "epoch": 1.9219440353460973,
+      "grad_norm": 0.3612362742424011,
+      "learning_rate": 3.4451381648319573e-06,
+      "loss": 0.3852,
+      "step": 2610
+    },
+    {
+      "epoch": 1.922680412371134,
+      "grad_norm": 0.3724992275238037,
+      "learning_rate": 3.4410664105541703e-06,
+      "loss": 0.4193,
+      "step": 2611
+    },
+    {
+      "epoch": 1.923416789396171,
+      "grad_norm": 0.3696814775466919,
+      "learning_rate": 3.4369958011101035e-06,
+      "loss": 0.4328,
+      "step": 2612
+    },
+    {
+      "epoch": 1.9241531664212077,
+      "grad_norm": 0.3654807507991791,
+      "learning_rate": 3.4329263394890867e-06,
+      "loss": 0.3699,
+      "step": 2613
+    },
+    {
+      "epoch": 1.9248895434462445,
+      "grad_norm": 0.3380682170391083,
+      "learning_rate": 3.4288580286796106e-06,
+      "loss": 0.3954,
+      "step": 2614
+    },
+    {
+      "epoch": 1.9256259204712813,
+      "grad_norm": 0.3763212263584137,
+      "learning_rate": 3.424790871669321e-06,
+      "loss": 0.395,
+      "step": 2615
+    },
+    {
+      "epoch": 1.9263622974963182,
+      "grad_norm": 0.37345150113105774,
+      "learning_rate": 3.4207248714450157e-06,
+      "loss": 0.4032,
+      "step": 2616
+    },
+    {
+      "epoch": 1.927098674521355,
+      "grad_norm": 0.3797363340854645,
+      "learning_rate": 3.416660030992639e-06,
+      "loss": 0.4183,
+      "step": 2617
+    },
+    {
+      "epoch": 1.9278350515463918,
+      "grad_norm": 0.39499834179878235,
+      "learning_rate": 3.4125963532972878e-06,
+      "loss": 0.4318,
+      "step": 2618
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 0.41415366530418396,
+      "learning_rate": 3.4085338413432066e-06,
+      "loss": 0.3798,
+      "step": 2619
+    },
+    {
+      "epoch": 1.9293078055964654,
+      "grad_norm": 0.38226163387298584,
+      "learning_rate": 3.4044724981137787e-06,
+      "loss": 0.4323,
+      "step": 2620
+    },
+    {
+      "epoch": 1.9300441826215022,
+      "grad_norm": 0.41554006934165955,
+      "learning_rate": 3.4004123265915328e-06,
+      "loss": 0.3985,
+      "step": 2621
+    },
+    {
+      "epoch": 1.930780559646539,
+      "grad_norm": 0.40494853258132935,
+      "learning_rate": 3.3963533297581375e-06,
+      "loss": 0.3894,
+      "step": 2622
+    },
+    {
+      "epoch": 1.9315169366715759,
+      "grad_norm": 0.42371952533721924,
+      "learning_rate": 3.3922955105943953e-06,
+      "loss": 0.4125,
+      "step": 2623
+    },
+    {
+      "epoch": 1.9322533136966127,
+      "grad_norm": 0.4033835530281067,
+      "learning_rate": 3.3882388720802496e-06,
+      "loss": 0.4221,
+      "step": 2624
+    },
+    {
+      "epoch": 1.9329896907216495,
+      "grad_norm": 0.42939725518226624,
+      "learning_rate": 3.384183417194767e-06,
+      "loss": 0.4095,
+      "step": 2625
+    },
+    {
+      "epoch": 1.9337260677466863,
+      "grad_norm": 0.41357123851776123,
+      "learning_rate": 3.380129148916156e-06,
+      "loss": 0.3923,
+      "step": 2626
+    },
+    {
+      "epoch": 1.9344624447717231,
+      "grad_norm": 0.41408517956733704,
+      "learning_rate": 3.3760760702217477e-06,
+      "loss": 0.3849,
+      "step": 2627
+    },
+    {
+      "epoch": 1.93519882179676,
+      "grad_norm": 0.3655821681022644,
+      "learning_rate": 3.3720241840879992e-06,
+      "loss": 0.4033,
+      "step": 2628
+    },
+    {
+      "epoch": 1.9359351988217968,
+      "grad_norm": 0.40222102403640747,
+      "learning_rate": 3.367973493490494e-06,
+      "loss": 0.4061,
+      "step": 2629
+    },
+    {
+      "epoch": 1.9366715758468336,
+      "grad_norm": 0.37809598445892334,
+      "learning_rate": 3.363924001403939e-06,
+      "loss": 0.3973,
+      "step": 2630
+    },
+    {
+      "epoch": 1.9374079528718704,
+      "grad_norm": 0.3718864321708679,
+      "learning_rate": 3.3598757108021546e-06,
+      "loss": 0.4259,
+      "step": 2631
+    },
+    {
+      "epoch": 1.9381443298969072,
+      "grad_norm": 0.409039169549942,
+      "learning_rate": 3.355828624658087e-06,
+      "loss": 0.4057,
+      "step": 2632
+    },
+    {
+      "epoch": 1.938880706921944,
+      "grad_norm": 0.34290969371795654,
+      "learning_rate": 3.351782745943792e-06,
+      "loss": 0.3939,
+      "step": 2633
+    },
+    {
+      "epoch": 1.9396170839469808,
+      "grad_norm": 0.378534734249115,
+      "learning_rate": 3.3477380776304412e-06,
+      "loss": 0.3941,
+      "step": 2634
+    },
+    {
+      "epoch": 1.9403534609720177,
+      "grad_norm": 0.3930615782737732,
+      "learning_rate": 3.343694622688315e-06,
+      "loss": 0.4266,
+      "step": 2635
+    },
+    {
+      "epoch": 1.9410898379970545,
+      "grad_norm": 0.3948494791984558,
+      "learning_rate": 3.3396523840868065e-06,
+      "loss": 0.4136,
+      "step": 2636
+    },
+    {
+      "epoch": 1.9418262150220913,
+      "grad_norm": 0.3820838928222656,
+      "learning_rate": 3.3356113647944144e-06,
+      "loss": 0.403,
+      "step": 2637
+    },
+    {
+      "epoch": 1.9425625920471281,
+      "grad_norm": 0.3924919664859772,
+      "learning_rate": 3.3315715677787387e-06,
+      "loss": 0.4077,
+      "step": 2638
+    },
+    {
+      "epoch": 1.943298969072165,
+      "grad_norm": 0.3638979196548462,
+      "learning_rate": 3.3275329960064855e-06,
+      "loss": 0.4086,
+      "step": 2639
+    },
+    {
+      "epoch": 1.9440353460972017,
+      "grad_norm": 0.3670336902141571,
+      "learning_rate": 3.3234956524434615e-06,
+      "loss": 0.4018,
+      "step": 2640
+    },
+    {
+      "epoch": 1.9447717231222386,
+      "grad_norm": 0.38897061347961426,
+      "learning_rate": 3.319459540054567e-06,
+      "loss": 0.4143,
+      "step": 2641
+    },
+    {
+      "epoch": 1.9455081001472754,
+      "grad_norm": 0.403804749250412,
+      "learning_rate": 3.315424661803802e-06,
+      "loss": 0.3792,
+      "step": 2642
+    },
+    {
+      "epoch": 1.9462444771723122,
+      "grad_norm": 0.3691318929195404,
+      "learning_rate": 3.3113910206542595e-06,
+      "loss": 0.4074,
+      "step": 2643
+    },
+    {
+      "epoch": 1.946980854197349,
+      "grad_norm": 0.41701748967170715,
+      "learning_rate": 3.307358619568123e-06,
+      "loss": 0.4037,
+      "step": 2644
+    },
+    {
+      "epoch": 1.9477172312223858,
+      "grad_norm": 0.3514380156993866,
+      "learning_rate": 3.303327461506667e-06,
+      "loss": 0.4278,
+      "step": 2645
+    },
+    {
+      "epoch": 1.9484536082474226,
+      "grad_norm": 0.3677442669868469,
+      "learning_rate": 3.29929754943025e-06,
+      "loss": 0.4029,
+      "step": 2646
+    },
+    {
+      "epoch": 1.9491899852724595,
+      "grad_norm": 0.38052093982696533,
+      "learning_rate": 3.295268886298321e-06,
+      "loss": 0.3938,
+      "step": 2647
+    },
+    {
+      "epoch": 1.9499263622974963,
+      "grad_norm": 0.4505428671836853,
+      "learning_rate": 3.2912414750694064e-06,
+      "loss": 0.3936,
+      "step": 2648
+    },
+    {
+      "epoch": 1.950662739322533,
+      "grad_norm": 0.35540086030960083,
+      "learning_rate": 3.2872153187011175e-06,
+      "loss": 0.4038,
+      "step": 2649
+    },
+    {
+      "epoch": 1.95139911634757,
+      "grad_norm": 0.38356146216392517,
+      "learning_rate": 3.2831904201501376e-06,
+      "loss": 0.4276,
+      "step": 2650
+    },
+    {
+      "epoch": 1.9521354933726067,
+      "grad_norm": 0.38261398673057556,
+      "learning_rate": 3.2791667823722327e-06,
+      "loss": 0.3967,
+      "step": 2651
+    },
+    {
+      "epoch": 1.9528718703976435,
+      "grad_norm": 0.3730379343032837,
+      "learning_rate": 3.2751444083222418e-06,
+      "loss": 0.3942,
+      "step": 2652
+    },
+    {
+      "epoch": 1.9536082474226806,
+      "grad_norm": 0.41036689281463623,
+      "learning_rate": 3.271123300954074e-06,
+      "loss": 0.4031,
+      "step": 2653
+    },
+    {
+      "epoch": 1.9543446244477174,
+      "grad_norm": 0.37355658411979675,
+      "learning_rate": 3.2671034632207084e-06,
+      "loss": 0.4094,
+      "step": 2654
+    },
+    {
+      "epoch": 1.9550810014727542,
+      "grad_norm": 0.3889113962650299,
+      "learning_rate": 3.263084898074194e-06,
+      "loss": 0.4123,
+      "step": 2655
+    },
+    {
+      "epoch": 1.955817378497791,
+      "grad_norm": 0.3863618075847626,
+      "learning_rate": 3.2590676084656425e-06,
+      "loss": 0.3765,
+      "step": 2656
+    },
+    {
+      "epoch": 1.9565537555228278,
+      "grad_norm": 0.3829159438610077,
+      "learning_rate": 3.2550515973452295e-06,
+      "loss": 0.3963,
+      "step": 2657
+    },
+    {
+      "epoch": 1.9572901325478647,
+      "grad_norm": 0.3452022671699524,
+      "learning_rate": 3.251036867662195e-06,
+      "loss": 0.4252,
+      "step": 2658
+    },
+    {
+      "epoch": 1.9580265095729015,
+      "grad_norm": 0.3539165258407593,
+      "learning_rate": 3.247023422364831e-06,
+      "loss": 0.3867,
+      "step": 2659
+    },
+    {
+      "epoch": 1.9587628865979383,
+      "grad_norm": 0.3868674039840698,
+      "learning_rate": 3.243011264400494e-06,
+      "loss": 0.421,
+      "step": 2660
+    },
+    {
+      "epoch": 1.959499263622975,
+      "grad_norm": 0.39568665623664856,
+      "learning_rate": 3.2390003967155887e-06,
+      "loss": 0.4064,
+      "step": 2661
+    },
+    {
+      "epoch": 1.960235640648012,
+      "grad_norm": 0.3922138214111328,
+      "learning_rate": 3.2349908222555764e-06,
+      "loss": 0.3997,
+      "step": 2662
+    },
+    {
+      "epoch": 1.9609720176730487,
+      "grad_norm": 0.3632813096046448,
+      "learning_rate": 3.230982543964969e-06,
+      "loss": 0.3785,
+      "step": 2663
+    },
+    {
+      "epoch": 1.9617083946980856,
+      "grad_norm": 0.38700589537620544,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.4231,
+      "step": 2664
+    },
+    {
+      "epoch": 1.9624447717231224,
+      "grad_norm": 0.37696290016174316,
+      "learning_rate": 3.2229698876652415e-06,
+      "loss": 0.3819,
+      "step": 2665
+    },
+    {
+      "epoch": 1.9631811487481592,
+      "grad_norm": 0.3456031084060669,
+      "learning_rate": 3.218965515540377e-06,
+      "loss": 0.4044,
+      "step": 2666
+    },
+    {
+      "epoch": 1.963917525773196,
+      "grad_norm": 0.39847907423973083,
+      "learning_rate": 3.214962451353416e-06,
+      "loss": 0.4108,
+      "step": 2667
+    },
+    {
+      "epoch": 1.9646539027982328,
+      "grad_norm": 0.3617440462112427,
+      "learning_rate": 3.2109606980440887e-06,
+      "loss": 0.4052,
+      "step": 2668
+    },
+    {
+      "epoch": 1.9653902798232696,
+      "grad_norm": 0.37616562843322754,
+      "learning_rate": 3.2069602585511605e-06,
+      "loss": 0.4069,
+      "step": 2669
+    },
+    {
+      "epoch": 1.9661266568483065,
+      "grad_norm": 0.39469170570373535,
+      "learning_rate": 3.202961135812437e-06,
+      "loss": 0.3785,
+      "step": 2670
+    },
+    {
+      "epoch": 1.9668630338733433,
+      "grad_norm": 0.4196336269378662,
+      "learning_rate": 3.1989633327647485e-06,
+      "loss": 0.4297,
+      "step": 2671
+    },
+    {
+      "epoch": 1.96759941089838,
+      "grad_norm": 0.380405068397522,
+      "learning_rate": 3.1949668523439635e-06,
+      "loss": 0.3883,
+      "step": 2672
+    },
+    {
+      "epoch": 1.968335787923417,
+      "grad_norm": 0.37013182044029236,
+      "learning_rate": 3.190971697484977e-06,
+      "loss": 0.3756,
+      "step": 2673
+    },
+    {
+      "epoch": 1.9690721649484537,
+      "grad_norm": 0.34765076637268066,
+      "learning_rate": 3.186977871121708e-06,
+      "loss": 0.3944,
+      "step": 2674
+    },
+    {
+      "epoch": 1.9698085419734905,
+      "grad_norm": 0.3879580497741699,
+      "learning_rate": 3.182985376187105e-06,
+      "loss": 0.4227,
+      "step": 2675
+    },
+    {
+      "epoch": 1.9705449189985274,
+      "grad_norm": 0.3524974584579468,
+      "learning_rate": 3.178994215613131e-06,
+      "loss": 0.3805,
+      "step": 2676
+    },
+    {
+      "epoch": 1.9712812960235642,
+      "grad_norm": 0.3391323983669281,
+      "learning_rate": 3.1750043923307773e-06,
+      "loss": 0.4046,
+      "step": 2677
+    },
+    {
+      "epoch": 1.972017673048601,
+      "grad_norm": 0.3604549169540405,
+      "learning_rate": 3.1710159092700475e-06,
+      "loss": 0.3858,
+      "step": 2678
+    },
+    {
+      "epoch": 1.9727540500736378,
+      "grad_norm": 0.36499884724617004,
+      "learning_rate": 3.167028769359964e-06,
+      "loss": 0.4425,
+      "step": 2679
+    },
+    {
+      "epoch": 1.9734904270986746,
+      "grad_norm": 0.31244251132011414,
+      "learning_rate": 3.1630429755285623e-06,
+      "loss": 0.4021,
+      "step": 2680
+    },
+    {
+      "epoch": 1.9742268041237114,
+      "grad_norm": 0.3878641128540039,
+      "learning_rate": 3.1590585307028884e-06,
+      "loss": 0.4057,
+      "step": 2681
+    },
+    {
+      "epoch": 1.9749631811487482,
+      "grad_norm": 0.3576429784297943,
+      "learning_rate": 3.1550754378089976e-06,
+      "loss": 0.3675,
+      "step": 2682
+    },
+    {
+      "epoch": 1.975699558173785,
+      "grad_norm": 0.40466341376304626,
+      "learning_rate": 3.1510936997719557e-06,
+      "loss": 0.3959,
+      "step": 2683
+    },
+    {
+      "epoch": 1.9764359351988219,
+      "grad_norm": 0.3640960156917572,
+      "learning_rate": 3.1471133195158266e-06,
+      "loss": 0.3905,
+      "step": 2684
+    },
+    {
+      "epoch": 1.9771723122238587,
+      "grad_norm": 0.3552342653274536,
+      "learning_rate": 3.143134299963684e-06,
+      "loss": 0.4215,
+      "step": 2685
+    },
+    {
+      "epoch": 1.9779086892488955,
+      "grad_norm": 0.3452814817428589,
+      "learning_rate": 3.1391566440375987e-06,
+      "loss": 0.3995,
+      "step": 2686
+    },
+    {
+      "epoch": 1.9786450662739323,
+      "grad_norm": 0.3464387357234955,
+      "learning_rate": 3.1351803546586407e-06,
+      "loss": 0.4106,
+      "step": 2687
+    },
+    {
+      "epoch": 1.9793814432989691,
+      "grad_norm": 0.3663497567176819,
+      "learning_rate": 3.131205434746879e-06,
+      "loss": 0.3728,
+      "step": 2688
+    },
+    {
+      "epoch": 1.980117820324006,
+      "grad_norm": 0.3594132661819458,
+      "learning_rate": 3.1272318872213713e-06,
+      "loss": 0.4075,
+      "step": 2689
+    },
+    {
+      "epoch": 1.9808541973490428,
+      "grad_norm": 0.35411345958709717,
+      "learning_rate": 3.123259715000173e-06,
+      "loss": 0.3754,
+      "step": 2690
+    },
+    {
+      "epoch": 1.9815905743740796,
+      "grad_norm": 0.33865928649902344,
+      "learning_rate": 3.1192889210003285e-06,
+      "loss": 0.3723,
+      "step": 2691
+    },
+    {
+      "epoch": 1.9823269513991164,
+      "grad_norm": 0.3422465920448303,
+      "learning_rate": 3.115319508137866e-06,
+      "loss": 0.387,
+      "step": 2692
+    },
+    {
+      "epoch": 1.9830633284241532,
+      "grad_norm": 0.3784734010696411,
+      "learning_rate": 3.1113514793278037e-06,
+      "loss": 0.397,
+      "step": 2693
+    },
+    {
+      "epoch": 1.98379970544919,
+      "grad_norm": 0.41923490166664124,
+      "learning_rate": 3.1073848374841416e-06,
+      "loss": 0.4199,
+      "step": 2694
+    },
+    {
+      "epoch": 1.9845360824742269,
+      "grad_norm": 0.37350550293922424,
+      "learning_rate": 3.1034195855198622e-06,
+      "loss": 0.4023,
+      "step": 2695
+    },
+    {
+      "epoch": 1.9852724594992637,
+      "grad_norm": 0.3785882294178009,
+      "learning_rate": 3.0994557263469267e-06,
+      "loss": 0.4117,
+      "step": 2696
+    },
+    {
+      "epoch": 1.9860088365243005,
+      "grad_norm": 0.36809927225112915,
+      "learning_rate": 3.0954932628762723e-06,
+      "loss": 0.3928,
+      "step": 2697
+    },
+    {
+      "epoch": 1.9867452135493373,
+      "grad_norm": 0.39854827523231506,
+      "learning_rate": 3.0915321980178153e-06,
+      "loss": 0.3881,
+      "step": 2698
+    },
+    {
+      "epoch": 1.9874815905743741,
+      "grad_norm": 0.37375113368034363,
+      "learning_rate": 3.0875725346804385e-06,
+      "loss": 0.4052,
+      "step": 2699
+    },
+    {
+      "epoch": 1.988217967599411,
+      "grad_norm": 0.4068243205547333,
+      "learning_rate": 3.0836142757720034e-06,
+      "loss": 0.3912,
+      "step": 2700
+    },
+    {
+      "epoch": 1.9889543446244478,
+      "grad_norm": 0.3465851843357086,
+      "learning_rate": 3.0796574241993306e-06,
+      "loss": 0.4136,
+      "step": 2701
+    },
+    {
+      "epoch": 1.9896907216494846,
+      "grad_norm": 0.39400115609169006,
+      "learning_rate": 3.0757019828682145e-06,
+      "loss": 0.3915,
+      "step": 2702
+    },
+    {
+      "epoch": 1.9904270986745214,
+      "grad_norm": 0.33876892924308777,
+      "learning_rate": 3.0717479546834136e-06,
+      "loss": 0.3967,
+      "step": 2703
+    },
+    {
+      "epoch": 1.9911634756995582,
+      "grad_norm": 0.3738095462322235,
+      "learning_rate": 3.0677953425486435e-06,
+      "loss": 0.3866,
+      "step": 2704
+    },
+    {
+      "epoch": 1.991899852724595,
+      "grad_norm": 0.350265771150589,
+      "learning_rate": 3.063844149366585e-06,
+      "loss": 0.3789,
+      "step": 2705
+    },
+    {
+      "epoch": 1.9926362297496318,
+      "grad_norm": 0.3564341068267822,
+      "learning_rate": 3.0598943780388744e-06,
+      "loss": 0.3948,
+      "step": 2706
+    },
+    {
+      "epoch": 1.9933726067746687,
+      "grad_norm": 0.3786764442920685,
+      "learning_rate": 3.055946031466105e-06,
+      "loss": 0.4204,
+      "step": 2707
+    },
+    {
+      "epoch": 1.9941089837997055,
+      "grad_norm": 0.3561117649078369,
+      "learning_rate": 3.0519991125478244e-06,
+      "loss": 0.3828,
+      "step": 2708
+    },
+    {
+      "epoch": 1.9948453608247423,
+      "grad_norm": 0.40353891253471375,
+      "learning_rate": 3.0480536241825263e-06,
+      "loss": 0.3999,
+      "step": 2709
+    },
+    {
+      "epoch": 1.995581737849779,
+      "grad_norm": 0.3841717839241028,
+      "learning_rate": 3.0441095692676625e-06,
+      "loss": 0.4079,
+      "step": 2710
+    },
+    {
+      "epoch": 1.996318114874816,
+      "grad_norm": 0.4203062057495117,
+      "learning_rate": 3.040166950699626e-06,
+      "loss": 0.395,
+      "step": 2711
+    },
+    {
+      "epoch": 1.9970544918998527,
+      "grad_norm": 0.4298144280910492,
+      "learning_rate": 3.0362257713737552e-06,
+      "loss": 0.466,
+      "step": 2712
+    },
+    {
+      "epoch": 1.9977908689248896,
+      "grad_norm": 0.41384029388427734,
+      "learning_rate": 3.0322860341843365e-06,
+      "loss": 0.3908,
+      "step": 2713
+    },
+    {
+      "epoch": 1.9985272459499264,
+      "grad_norm": 0.35945290327072144,
+      "learning_rate": 3.028347742024591e-06,
+      "loss": 0.3889,
+      "step": 2714
+    },
+    {
+      "epoch": 1.9992636229749632,
+      "grad_norm": 0.3887770473957062,
+      "learning_rate": 3.024410897786682e-06,
+      "loss": 0.4116,
+      "step": 2715
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.4592723250389099,
+      "learning_rate": 3.020475504361711e-06,
+      "loss": 0.4304,
+      "step": 2716
+    },
+    {
+      "epoch": 2.000736377025037,
+      "grad_norm": 0.4854089915752411,
+      "learning_rate": 3.01654156463971e-06,
+      "loss": 0.3914,
+      "step": 2717
+    },
+    {
+      "epoch": 2.0014727540500736,
+      "grad_norm": 0.40188172459602356,
+      "learning_rate": 3.0126090815096466e-06,
+      "loss": 0.3823,
+      "step": 2718
+    },
+    {
+      "epoch": 2.0022091310751104,
+      "grad_norm": 0.3685808777809143,
+      "learning_rate": 3.008678057859415e-06,
+      "loss": 0.3927,
+      "step": 2719
+    },
+    {
+      "epoch": 2.0029455081001473,
+      "grad_norm": 0.41367441415786743,
+      "learning_rate": 3.004748496575842e-06,
+      "loss": 0.3738,
+      "step": 2720
+    },
+    {
+      "epoch": 2.003681885125184,
+      "grad_norm": 0.3970171809196472,
+      "learning_rate": 3.0008204005446807e-06,
+      "loss": 0.3864,
+      "step": 2721
+    },
+    {
+      "epoch": 2.004418262150221,
+      "grad_norm": 0.38890549540519714,
+      "learning_rate": 2.996893772650602e-06,
+      "loss": 0.3532,
+      "step": 2722
+    },
+    {
+      "epoch": 2.0051546391752577,
+      "grad_norm": 0.39022096991539,
+      "learning_rate": 2.992968615777206e-06,
+      "loss": 0.3809,
+      "step": 2723
+    },
+    {
+      "epoch": 2.0058910162002945,
+      "grad_norm": 0.3781045973300934,
+      "learning_rate": 2.989044932807008e-06,
+      "loss": 0.3753,
+      "step": 2724
+    },
+    {
+      "epoch": 2.0066273932253313,
+      "grad_norm": 0.40674319863319397,
+      "learning_rate": 2.9851227266214444e-06,
+      "loss": 0.3748,
+      "step": 2725
+    },
+    {
+      "epoch": 2.007363770250368,
+      "grad_norm": 0.42673784494400024,
+      "learning_rate": 2.981202000100861e-06,
+      "loss": 0.371,
+      "step": 2726
+    },
+    {
+      "epoch": 2.008100147275405,
+      "grad_norm": 0.3645564615726471,
+      "learning_rate": 2.9772827561245223e-06,
+      "loss": 0.3759,
+      "step": 2727
+    },
+    {
+      "epoch": 2.008836524300442,
+      "grad_norm": 0.3563685715198517,
+      "learning_rate": 2.9733649975706035e-06,
+      "loss": 0.3681,
+      "step": 2728
+    },
+    {
+      "epoch": 2.0095729013254786,
+      "grad_norm": 0.35634204745292664,
+      "learning_rate": 2.969448727316188e-06,
+      "loss": 0.3784,
+      "step": 2729
+    },
+    {
+      "epoch": 2.0103092783505154,
+      "grad_norm": 0.3639596998691559,
+      "learning_rate": 2.9655339482372647e-06,
+      "loss": 0.3652,
+      "step": 2730
+    },
+    {
+      "epoch": 2.0110456553755522,
+      "grad_norm": 0.3867950141429901,
+      "learning_rate": 2.961620663208732e-06,
+      "loss": 0.3561,
+      "step": 2731
+    },
+    {
+      "epoch": 2.011782032400589,
+      "grad_norm": 0.3902873396873474,
+      "learning_rate": 2.957708875104386e-06,
+      "loss": 0.3777,
+      "step": 2732
+    },
+    {
+      "epoch": 2.012518409425626,
+      "grad_norm": 0.39293259382247925,
+      "learning_rate": 2.9537985867969277e-06,
+      "loss": 0.358,
+      "step": 2733
+    },
+    {
+      "epoch": 2.0132547864506627,
+      "grad_norm": 0.37629497051239014,
+      "learning_rate": 2.9498898011579514e-06,
+      "loss": 0.3884,
+      "step": 2734
+    },
+    {
+      "epoch": 2.0139911634756995,
+      "grad_norm": 0.3420172929763794,
+      "learning_rate": 2.9459825210579534e-06,
+      "loss": 0.3858,
+      "step": 2735
+    },
+    {
+      "epoch": 2.0147275405007363,
+      "grad_norm": 0.3435642123222351,
+      "learning_rate": 2.942076749366321e-06,
+      "loss": 0.3672,
+      "step": 2736
+    },
+    {
+      "epoch": 2.015463917525773,
+      "grad_norm": 0.36072492599487305,
+      "learning_rate": 2.938172488951336e-06,
+      "loss": 0.3591,
+      "step": 2737
+    },
+    {
+      "epoch": 2.01620029455081,
+      "grad_norm": 0.34583789110183716,
+      "learning_rate": 2.9342697426801693e-06,
+      "loss": 0.3348,
+      "step": 2738
+    },
+    {
+      "epoch": 2.0169366715758468,
+      "grad_norm": 0.3557235598564148,
+      "learning_rate": 2.9303685134188785e-06,
+      "loss": 0.3624,
+      "step": 2739
+    },
+    {
+      "epoch": 2.0176730486008836,
+      "grad_norm": 0.3426262140274048,
+      "learning_rate": 2.9264688040324098e-06,
+      "loss": 0.3786,
+      "step": 2740
+    },
+    {
+      "epoch": 2.0184094256259204,
+      "grad_norm": 0.33903318643569946,
+      "learning_rate": 2.922570617384591e-06,
+      "loss": 0.3707,
+      "step": 2741
+    },
+    {
+      "epoch": 2.0191458026509572,
+      "grad_norm": 0.32387295365333557,
+      "learning_rate": 2.918673956338136e-06,
+      "loss": 0.3661,
+      "step": 2742
+    },
+    {
+      "epoch": 2.019882179675994,
+      "grad_norm": 0.34796783328056335,
+      "learning_rate": 2.914778823754628e-06,
+      "loss": 0.3941,
+      "step": 2743
+    },
+    {
+      "epoch": 2.020618556701031,
+      "grad_norm": 0.3746124804019928,
+      "learning_rate": 2.9108852224945405e-06,
+      "loss": 0.3809,
+      "step": 2744
+    },
+    {
+      "epoch": 2.0213549337260677,
+      "grad_norm": 0.32552647590637207,
+      "learning_rate": 2.9069931554172155e-06,
+      "loss": 0.3749,
+      "step": 2745
+    },
+    {
+      "epoch": 2.0220913107511045,
+      "grad_norm": 0.36219868063926697,
+      "learning_rate": 2.9031026253808657e-06,
+      "loss": 0.3647,
+      "step": 2746
+    },
+    {
+      "epoch": 2.0228276877761413,
+      "grad_norm": 0.311967134475708,
+      "learning_rate": 2.899213635242585e-06,
+      "loss": 0.3714,
+      "step": 2747
+    },
+    {
+      "epoch": 2.023564064801178,
+      "grad_norm": 0.382050484418869,
+      "learning_rate": 2.8953261878583263e-06,
+      "loss": 0.3629,
+      "step": 2748
+    },
+    {
+      "epoch": 2.024300441826215,
+      "grad_norm": 0.3814060688018799,
+      "learning_rate": 2.8914402860829116e-06,
+      "loss": 0.3756,
+      "step": 2749
+    },
+    {
+      "epoch": 2.0250368188512518,
+      "grad_norm": 0.3468574285507202,
+      "learning_rate": 2.8875559327700376e-06,
+      "loss": 0.3673,
+      "step": 2750
+    },
+    {
+      "epoch": 2.0257731958762886,
+      "grad_norm": 0.3503856658935547,
+      "learning_rate": 2.8836731307722456e-06,
+      "loss": 0.3999,
+      "step": 2751
+    },
+    {
+      "epoch": 2.0265095729013254,
+      "grad_norm": 0.34949731826782227,
+      "learning_rate": 2.8797918829409553e-06,
+      "loss": 0.371,
+      "step": 2752
+    },
+    {
+      "epoch": 2.027245949926362,
+      "grad_norm": 0.4045935273170471,
+      "learning_rate": 2.8759121921264366e-06,
+      "loss": 0.3934,
+      "step": 2753
+    },
+    {
+      "epoch": 2.027982326951399,
+      "grad_norm": 0.3662734031677246,
+      "learning_rate": 2.8720340611778134e-06,
+      "loss": 0.3674,
+      "step": 2754
+    },
+    {
+      "epoch": 2.028718703976436,
+      "grad_norm": 0.3531613051891327,
+      "learning_rate": 2.8681574929430732e-06,
+      "loss": 0.3619,
+      "step": 2755
+    },
+    {
+      "epoch": 2.0294550810014726,
+      "grad_norm": 0.35535216331481934,
+      "learning_rate": 2.8642824902690482e-06,
+      "loss": 0.3672,
+      "step": 2756
+    },
+    {
+      "epoch": 2.0301914580265095,
+      "grad_norm": 0.39943090081214905,
+      "learning_rate": 2.860409056001421e-06,
+      "loss": 0.3925,
+      "step": 2757
+    },
+    {
+      "epoch": 2.0309278350515463,
+      "grad_norm": 0.39442721009254456,
+      "learning_rate": 2.8565371929847286e-06,
+      "loss": 0.3568,
+      "step": 2758
+    },
+    {
+      "epoch": 2.031664212076583,
+      "grad_norm": 0.35805344581604004,
+      "learning_rate": 2.852666904062351e-06,
+      "loss": 0.3424,
+      "step": 2759
+    },
+    {
+      "epoch": 2.03240058910162,
+      "grad_norm": 0.3514004051685333,
+      "learning_rate": 2.8487981920765044e-06,
+      "loss": 0.3695,
+      "step": 2760
+    },
+    {
+      "epoch": 2.0331369661266567,
+      "grad_norm": 0.3559187054634094,
+      "learning_rate": 2.844931059868261e-06,
+      "loss": 0.4159,
+      "step": 2761
+    },
+    {
+      "epoch": 2.0338733431516935,
+      "grad_norm": 0.3435509502887726,
+      "learning_rate": 2.841065510277523e-06,
+      "loss": 0.3477,
+      "step": 2762
+    },
+    {
+      "epoch": 2.0346097201767304,
+      "grad_norm": 0.3598922789096832,
+      "learning_rate": 2.8372015461430313e-06,
+      "loss": 0.3543,
+      "step": 2763
+    },
+    {
+      "epoch": 2.035346097201767,
+      "grad_norm": 0.4100363552570343,
+      "learning_rate": 2.833339170302369e-06,
+      "loss": 0.3485,
+      "step": 2764
+    },
+    {
+      "epoch": 2.036082474226804,
+      "grad_norm": 0.3745668828487396,
+      "learning_rate": 2.829478385591946e-06,
+      "loss": 0.3597,
+      "step": 2765
+    },
+    {
+      "epoch": 2.036818851251841,
+      "grad_norm": 0.32710880041122437,
+      "learning_rate": 2.8256191948470034e-06,
+      "loss": 0.3525,
+      "step": 2766
+    },
+    {
+      "epoch": 2.0375552282768776,
+      "grad_norm": 0.4090001583099365,
+      "learning_rate": 2.8217616009016203e-06,
+      "loss": 0.3994,
+      "step": 2767
+    },
+    {
+      "epoch": 2.0382916053019144,
+      "grad_norm": 0.35250866413116455,
+      "learning_rate": 2.81790560658869e-06,
+      "loss": 0.3575,
+      "step": 2768
+    },
+    {
+      "epoch": 2.0390279823269513,
+      "grad_norm": 0.3470795452594757,
+      "learning_rate": 2.8140512147399436e-06,
+      "loss": 0.3801,
+      "step": 2769
+    },
+    {
+      "epoch": 2.039764359351988,
+      "grad_norm": 0.39775145053863525,
+      "learning_rate": 2.8101984281859276e-06,
+      "loss": 0.3803,
+      "step": 2770
+    },
+    {
+      "epoch": 2.040500736377025,
+      "grad_norm": 0.38484060764312744,
+      "learning_rate": 2.8063472497560107e-06,
+      "loss": 0.3948,
+      "step": 2771
+    },
+    {
+      "epoch": 2.0412371134020617,
+      "grad_norm": 0.35866910219192505,
+      "learning_rate": 2.802497682278385e-06,
+      "loss": 0.3644,
+      "step": 2772
+    },
+    {
+      "epoch": 2.0419734904270985,
+      "grad_norm": 0.3320389688014984,
+      "learning_rate": 2.7986497285800564e-06,
+      "loss": 0.3525,
+      "step": 2773
+    },
+    {
+      "epoch": 2.0427098674521353,
+      "grad_norm": 0.3605724573135376,
+      "learning_rate": 2.7948033914868415e-06,
+      "loss": 0.3861,
+      "step": 2774
+    },
+    {
+      "epoch": 2.043446244477172,
+      "grad_norm": 0.35822010040283203,
+      "learning_rate": 2.7909586738233816e-06,
+      "loss": 0.3836,
+      "step": 2775
+    },
+    {
+      "epoch": 2.044182621502209,
+      "grad_norm": 0.38547834753990173,
+      "learning_rate": 2.787115578413113e-06,
+      "loss": 0.3731,
+      "step": 2776
+    },
+    {
+      "epoch": 2.044918998527246,
+      "grad_norm": 0.36046162247657776,
+      "learning_rate": 2.7832741080782944e-06,
+      "loss": 0.3705,
+      "step": 2777
+    },
+    {
+      "epoch": 2.0456553755522826,
+      "grad_norm": 0.36100590229034424,
+      "learning_rate": 2.7794342656399835e-06,
+      "loss": 0.3573,
+      "step": 2778
+    },
+    {
+      "epoch": 2.0463917525773194,
+      "grad_norm": 0.37001779675483704,
+      "learning_rate": 2.775596053918043e-06,
+      "loss": 0.3919,
+      "step": 2779
+    },
+    {
+      "epoch": 2.0471281296023562,
+      "grad_norm": 0.37465184926986694,
+      "learning_rate": 2.7717594757311435e-06,
+      "loss": 0.4072,
+      "step": 2780
+    },
+    {
+      "epoch": 2.047864506627393,
+      "grad_norm": 0.35224416851997375,
+      "learning_rate": 2.7679245338967497e-06,
+      "loss": 0.393,
+      "step": 2781
+    },
+    {
+      "epoch": 2.04860088365243,
+      "grad_norm": 0.38229209184646606,
+      "learning_rate": 2.764091231231125e-06,
+      "loss": 0.391,
+      "step": 2782
+    },
+    {
+      "epoch": 2.0493372606774667,
+      "grad_norm": 0.3451941907405853,
+      "learning_rate": 2.7602595705493353e-06,
+      "loss": 0.367,
+      "step": 2783
+    },
+    {
+      "epoch": 2.0500736377025035,
+      "grad_norm": 0.33509618043899536,
+      "learning_rate": 2.7564295546652366e-06,
+      "loss": 0.3684,
+      "step": 2784
+    },
+    {
+      "epoch": 2.0508100147275403,
+      "grad_norm": 0.38282108306884766,
+      "learning_rate": 2.7526011863914702e-06,
+      "loss": 0.3657,
+      "step": 2785
+    },
+    {
+      "epoch": 2.051546391752577,
+      "grad_norm": 0.38125553727149963,
+      "learning_rate": 2.748774468539481e-06,
+      "loss": 0.3667,
+      "step": 2786
+    },
+    {
+      "epoch": 2.052282768777614,
+      "grad_norm": 0.34986355900764465,
+      "learning_rate": 2.74494940391949e-06,
+      "loss": 0.4248,
+      "step": 2787
+    },
+    {
+      "epoch": 2.0530191458026508,
+      "grad_norm": 0.38790905475616455,
+      "learning_rate": 2.7411259953405143e-06,
+      "loss": 0.3802,
+      "step": 2788
+    },
+    {
+      "epoch": 2.0537555228276876,
+      "grad_norm": 0.3548150956630707,
+      "learning_rate": 2.737304245610346e-06,
+      "loss": 0.3833,
+      "step": 2789
+    },
+    {
+      "epoch": 2.0544918998527244,
+      "grad_norm": 0.3673829436302185,
+      "learning_rate": 2.7334841575355618e-06,
+      "loss": 0.3856,
+      "step": 2790
+    },
+    {
+      "epoch": 2.055228276877761,
+      "grad_norm": 0.363076388835907,
+      "learning_rate": 2.7296657339215227e-06,
+      "loss": 0.3783,
+      "step": 2791
+    },
+    {
+      "epoch": 2.055964653902798,
+      "grad_norm": 0.3587580621242523,
+      "learning_rate": 2.725848977572363e-06,
+      "loss": 0.386,
+      "step": 2792
+    },
+    {
+      "epoch": 2.056701030927835,
+      "grad_norm": 0.3971223831176758,
+      "learning_rate": 2.722033891290988e-06,
+      "loss": 0.3445,
+      "step": 2793
+    },
+    {
+      "epoch": 2.0574374079528717,
+      "grad_norm": 0.38008826971054077,
+      "learning_rate": 2.7182204778790878e-06,
+      "loss": 0.3613,
+      "step": 2794
+    },
+    {
+      "epoch": 2.0581737849779085,
+      "grad_norm": 0.3624361753463745,
+      "learning_rate": 2.714408740137115e-06,
+      "loss": 0.3675,
+      "step": 2795
+    },
+    {
+      "epoch": 2.0589101620029453,
+      "grad_norm": 0.33494141697883606,
+      "learning_rate": 2.7105986808642936e-06,
+      "loss": 0.3914,
+      "step": 2796
+    },
+    {
+      "epoch": 2.059646539027982,
+      "grad_norm": 0.36965957283973694,
+      "learning_rate": 2.7067903028586193e-06,
+      "loss": 0.3601,
+      "step": 2797
+    },
+    {
+      "epoch": 2.060382916053019,
+      "grad_norm": 0.3501898944377899,
+      "learning_rate": 2.702983608916849e-06,
+      "loss": 0.3996,
+      "step": 2798
+    },
+    {
+      "epoch": 2.0611192930780557,
+      "grad_norm": 0.41025105118751526,
+      "learning_rate": 2.6991786018345e-06,
+      "loss": 0.3741,
+      "step": 2799
+    },
+    {
+      "epoch": 2.0618556701030926,
+      "grad_norm": 0.3536095917224884,
+      "learning_rate": 2.69537528440586e-06,
+      "loss": 0.3831,
+      "step": 2800
+    },
+    {
+      "epoch": 2.0625920471281294,
+      "grad_norm": 0.369924396276474,
+      "learning_rate": 2.6915736594239676e-06,
+      "loss": 0.3879,
+      "step": 2801
+    },
+    {
+      "epoch": 2.063328424153166,
+      "grad_norm": 0.3758985996246338,
+      "learning_rate": 2.6877737296806217e-06,
+      "loss": 0.3922,
+      "step": 2802
+    },
+    {
+      "epoch": 2.064064801178203,
+      "grad_norm": 0.4005085527896881,
+      "learning_rate": 2.6839754979663752e-06,
+      "loss": 0.3852,
+      "step": 2803
+    },
+    {
+      "epoch": 2.0648011782032403,
+      "grad_norm": 0.3522018790245056,
+      "learning_rate": 2.6801789670705335e-06,
+      "loss": 0.3979,
+      "step": 2804
+    },
+    {
+      "epoch": 2.065537555228277,
+      "grad_norm": 0.36220279335975647,
+      "learning_rate": 2.6763841397811576e-06,
+      "loss": 0.3662,
+      "step": 2805
+    },
+    {
+      "epoch": 2.066273932253314,
+      "grad_norm": 0.37038183212280273,
+      "learning_rate": 2.6725910188850523e-06,
+      "loss": 0.364,
+      "step": 2806
+    },
+    {
+      "epoch": 2.0670103092783507,
+      "grad_norm": 0.3805944323539734,
+      "learning_rate": 2.668799607167769e-06,
+      "loss": 0.389,
+      "step": 2807
+    },
+    {
+      "epoch": 2.0677466863033875,
+      "grad_norm": 0.33866146206855774,
+      "learning_rate": 2.6650099074136095e-06,
+      "loss": 0.348,
+      "step": 2808
+    },
+    {
+      "epoch": 2.0684830633284244,
+      "grad_norm": 0.37582066655158997,
+      "learning_rate": 2.6612219224056133e-06,
+      "loss": 0.3865,
+      "step": 2809
+    },
+    {
+      "epoch": 2.069219440353461,
+      "grad_norm": 0.39235755801200867,
+      "learning_rate": 2.657435654925562e-06,
+      "loss": 0.3762,
+      "step": 2810
+    },
+    {
+      "epoch": 2.069955817378498,
+      "grad_norm": 0.3813096284866333,
+      "learning_rate": 2.6536511077539757e-06,
+      "loss": 0.3941,
+      "step": 2811
+    },
+    {
+      "epoch": 2.070692194403535,
+      "grad_norm": 0.35484588146209717,
+      "learning_rate": 2.6498682836701094e-06,
+      "loss": 0.388,
+      "step": 2812
+    },
+    {
+      "epoch": 2.0714285714285716,
+      "grad_norm": 0.37006160616874695,
+      "learning_rate": 2.6460871854519594e-06,
+      "loss": 0.4055,
+      "step": 2813
+    },
+    {
+      "epoch": 2.0721649484536084,
+      "grad_norm": 0.35195285081863403,
+      "learning_rate": 2.6423078158762473e-06,
+      "loss": 0.3766,
+      "step": 2814
+    },
+    {
+      "epoch": 2.0729013254786453,
+      "grad_norm": 0.3882370591163635,
+      "learning_rate": 2.638530177718427e-06,
+      "loss": 0.388,
+      "step": 2815
+    },
+    {
+      "epoch": 2.073637702503682,
+      "grad_norm": 0.33952176570892334,
+      "learning_rate": 2.6347542737526843e-06,
+      "loss": 0.3726,
+      "step": 2816
+    },
+    {
+      "epoch": 2.074374079528719,
+      "grad_norm": 0.37573882937431335,
+      "learning_rate": 2.6309801067519293e-06,
+      "loss": 0.3818,
+      "step": 2817
+    },
+    {
+      "epoch": 2.0751104565537557,
+      "grad_norm": 0.3656890094280243,
+      "learning_rate": 2.6272076794877915e-06,
+      "loss": 0.397,
+      "step": 2818
+    },
+    {
+      "epoch": 2.0758468335787925,
+      "grad_norm": 0.39007052779197693,
+      "learning_rate": 2.623436994730632e-06,
+      "loss": 0.3641,
+      "step": 2819
+    },
+    {
+      "epoch": 2.0765832106038293,
+      "grad_norm": 0.36530953645706177,
+      "learning_rate": 2.619668055249527e-06,
+      "loss": 0.3543,
+      "step": 2820
+    },
+    {
+      "epoch": 2.077319587628866,
+      "grad_norm": 0.3379687964916229,
+      "learning_rate": 2.6159008638122687e-06,
+      "loss": 0.3812,
+      "step": 2821
+    },
+    {
+      "epoch": 2.078055964653903,
+      "grad_norm": 0.3618925213813782,
+      "learning_rate": 2.6121354231853725e-06,
+      "loss": 0.4019,
+      "step": 2822
+    },
+    {
+      "epoch": 2.07879234167894,
+      "grad_norm": 0.39091405272483826,
+      "learning_rate": 2.608371736134063e-06,
+      "loss": 0.3696,
+      "step": 2823
+    },
+    {
+      "epoch": 2.0795287187039766,
+      "grad_norm": 0.3801884055137634,
+      "learning_rate": 2.6046098054222767e-06,
+      "loss": 0.3822,
+      "step": 2824
+    },
+    {
+      "epoch": 2.0802650957290134,
+      "grad_norm": 0.34748730063438416,
+      "learning_rate": 2.6008496338126643e-06,
+      "loss": 0.3625,
+      "step": 2825
+    },
+    {
+      "epoch": 2.0810014727540502,
+      "grad_norm": 0.3482074737548828,
+      "learning_rate": 2.5970912240665815e-06,
+      "loss": 0.3685,
+      "step": 2826
+    },
+    {
+      "epoch": 2.081737849779087,
+      "grad_norm": 0.39526116847991943,
+      "learning_rate": 2.59333457894409e-06,
+      "loss": 0.3852,
+      "step": 2827
+    },
+    {
+      "epoch": 2.082474226804124,
+      "grad_norm": 0.39386582374572754,
+      "learning_rate": 2.5895797012039576e-06,
+      "loss": 0.3473,
+      "step": 2828
+    },
+    {
+      "epoch": 2.0832106038291607,
+      "grad_norm": 0.40795037150382996,
+      "learning_rate": 2.5858265936036496e-06,
+      "loss": 0.394,
+      "step": 2829
+    },
+    {
+      "epoch": 2.0839469808541975,
+      "grad_norm": 0.40214234590530396,
+      "learning_rate": 2.582075258899339e-06,
+      "loss": 0.3657,
+      "step": 2830
+    },
+    {
+      "epoch": 2.0846833578792343,
+      "grad_norm": 0.35180026292800903,
+      "learning_rate": 2.578325699845892e-06,
+      "loss": 0.3508,
+      "step": 2831
+    },
+    {
+      "epoch": 2.085419734904271,
+      "grad_norm": 0.3568512201309204,
+      "learning_rate": 2.5745779191968686e-06,
+      "loss": 0.3434,
+      "step": 2832
+    },
+    {
+      "epoch": 2.086156111929308,
+      "grad_norm": 0.3792460262775421,
+      "learning_rate": 2.5708319197045297e-06,
+      "loss": 0.3691,
+      "step": 2833
+    },
+    {
+      "epoch": 2.0868924889543448,
+      "grad_norm": 0.36163848638534546,
+      "learning_rate": 2.567087704119821e-06,
+      "loss": 0.3726,
+      "step": 2834
+    },
+    {
+      "epoch": 2.0876288659793816,
+      "grad_norm": 0.3781639039516449,
+      "learning_rate": 2.5633452751923825e-06,
+      "loss": 0.3749,
+      "step": 2835
+    },
+    {
+      "epoch": 2.0883652430044184,
+      "grad_norm": 0.3682844340801239,
+      "learning_rate": 2.5596046356705418e-06,
+      "loss": 0.3525,
+      "step": 2836
+    },
+    {
+      "epoch": 2.089101620029455,
+      "grad_norm": 0.3800516128540039,
+      "learning_rate": 2.5558657883013078e-06,
+      "loss": 0.3867,
+      "step": 2837
+    },
+    {
+      "epoch": 2.089837997054492,
+      "grad_norm": 0.34438201785087585,
+      "learning_rate": 2.5521287358303814e-06,
+      "loss": 0.4029,
+      "step": 2838
+    },
+    {
+      "epoch": 2.090574374079529,
+      "grad_norm": 0.3582353889942169,
+      "learning_rate": 2.54839348100214e-06,
+      "loss": 0.35,
+      "step": 2839
+    },
+    {
+      "epoch": 2.0913107511045657,
+      "grad_norm": 0.35245490074157715,
+      "learning_rate": 2.544660026559639e-06,
+      "loss": 0.3818,
+      "step": 2840
+    },
+    {
+      "epoch": 2.0920471281296025,
+      "grad_norm": 0.3359908163547516,
+      "learning_rate": 2.5409283752446183e-06,
+      "loss": 0.3835,
+      "step": 2841
+    },
+    {
+      "epoch": 2.0927835051546393,
+      "grad_norm": 0.3465385437011719,
+      "learning_rate": 2.537198529797489e-06,
+      "loss": 0.3917,
+      "step": 2842
+    },
+    {
+      "epoch": 2.093519882179676,
+      "grad_norm": 0.3186095952987671,
+      "learning_rate": 2.533470492957335e-06,
+      "loss": 0.3766,
+      "step": 2843
+    },
+    {
+      "epoch": 2.094256259204713,
+      "grad_norm": 0.34522688388824463,
+      "learning_rate": 2.5297442674619153e-06,
+      "loss": 0.3691,
+      "step": 2844
+    },
+    {
+      "epoch": 2.0949926362297497,
+      "grad_norm": 0.3759070932865143,
+      "learning_rate": 2.526019856047656e-06,
+      "loss": 0.3514,
+      "step": 2845
+    },
+    {
+      "epoch": 2.0957290132547866,
+      "grad_norm": 0.37958410382270813,
+      "learning_rate": 2.5222972614496543e-06,
+      "loss": 0.3999,
+      "step": 2846
+    },
+    {
+      "epoch": 2.0964653902798234,
+      "grad_norm": 0.33293333649635315,
+      "learning_rate": 2.518576486401671e-06,
+      "loss": 0.3622,
+      "step": 2847
+    },
+    {
+      "epoch": 2.09720176730486,
+      "grad_norm": 0.39282292127609253,
+      "learning_rate": 2.514857533636128e-06,
+      "loss": 0.3799,
+      "step": 2848
+    },
+    {
+      "epoch": 2.097938144329897,
+      "grad_norm": 0.3422982394695282,
+      "learning_rate": 2.5111404058841155e-06,
+      "loss": 0.3917,
+      "step": 2849
+    },
+    {
+      "epoch": 2.098674521354934,
+      "grad_norm": 0.3647412359714508,
+      "learning_rate": 2.5074251058753783e-06,
+      "loss": 0.3606,
+      "step": 2850
+    },
+    {
+      "epoch": 2.0994108983799706,
+      "grad_norm": 0.38597914576530457,
+      "learning_rate": 2.5037116363383203e-06,
+      "loss": 0.3649,
+      "step": 2851
+    },
+    {
+      "epoch": 2.1001472754050075,
+      "grad_norm": 0.33769500255584717,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.37,
+      "step": 2852
+    },
+    {
+      "epoch": 2.1008836524300443,
+      "grad_norm": 0.3958123028278351,
+      "learning_rate": 2.4962901995861348e-06,
+      "loss": 0.3439,
+      "step": 2853
+    },
+    {
+      "epoch": 2.101620029455081,
+      "grad_norm": 0.34865254163742065,
+      "learning_rate": 2.4925822378210844e-06,
+      "loss": 0.3636,
+      "step": 2854
+    },
+    {
+      "epoch": 2.102356406480118,
+      "grad_norm": 0.3949892222881317,
+      "learning_rate": 2.488876117427869e-06,
+      "loss": 0.3538,
+      "step": 2855
+    },
+    {
+      "epoch": 2.1030927835051547,
+      "grad_norm": 0.39913707971572876,
+      "learning_rate": 2.4851718411281495e-06,
+      "loss": 0.368,
+      "step": 2856
+    },
+    {
+      "epoch": 2.1038291605301915,
+      "grad_norm": 0.3734777569770813,
+      "learning_rate": 2.4814694116422326e-06,
+      "loss": 0.3516,
+      "step": 2857
+    },
+    {
+      "epoch": 2.1045655375552283,
+      "grad_norm": 0.414492130279541,
+      "learning_rate": 2.477768831689074e-06,
+      "loss": 0.3833,
+      "step": 2858
+    },
+    {
+      "epoch": 2.105301914580265,
+      "grad_norm": 0.3294821083545685,
+      "learning_rate": 2.4740701039862663e-06,
+      "loss": 0.3678,
+      "step": 2859
+    },
+    {
+      "epoch": 2.106038291605302,
+      "grad_norm": 0.38018667697906494,
+      "learning_rate": 2.4703732312500438e-06,
+      "loss": 0.3991,
+      "step": 2860
+    },
+    {
+      "epoch": 2.106774668630339,
+      "grad_norm": 0.35948431491851807,
+      "learning_rate": 2.466678216195277e-06,
+      "loss": 0.361,
+      "step": 2861
+    },
+    {
+      "epoch": 2.1075110456553756,
+      "grad_norm": 0.3856082856655121,
+      "learning_rate": 2.462985061535472e-06,
+      "loss": 0.369,
+      "step": 2862
+    },
+    {
+      "epoch": 2.1082474226804124,
+      "grad_norm": 0.39372214674949646,
+      "learning_rate": 2.459293769982774e-06,
+      "loss": 0.3786,
+      "step": 2863
+    },
+    {
+      "epoch": 2.1089837997054492,
+      "grad_norm": 0.3526458144187927,
+      "learning_rate": 2.455604344247954e-06,
+      "loss": 0.3703,
+      "step": 2864
+    },
+    {
+      "epoch": 2.109720176730486,
+      "grad_norm": 0.3489939570426941,
+      "learning_rate": 2.4519167870404126e-06,
+      "loss": 0.3932,
+      "step": 2865
+    },
+    {
+      "epoch": 2.110456553755523,
+      "grad_norm": 0.37256160378456116,
+      "learning_rate": 2.4482311010681842e-06,
+      "loss": 0.3982,
+      "step": 2866
+    },
+    {
+      "epoch": 2.1111929307805597,
+      "grad_norm": 0.3323703110218048,
+      "learning_rate": 2.4445472890379233e-06,
+      "loss": 0.3791,
+      "step": 2867
+    },
+    {
+      "epoch": 2.1119293078055965,
+      "grad_norm": 0.36218979954719543,
+      "learning_rate": 2.4408653536549104e-06,
+      "loss": 0.4038,
+      "step": 2868
+    },
+    {
+      "epoch": 2.1126656848306333,
+      "grad_norm": 0.34477296471595764,
+      "learning_rate": 2.437185297623047e-06,
+      "loss": 0.391,
+      "step": 2869
+    },
+    {
+      "epoch": 2.11340206185567,
+      "grad_norm": 0.35292166471481323,
+      "learning_rate": 2.4335071236448536e-06,
+      "loss": 0.3793,
+      "step": 2870
+    },
+    {
+      "epoch": 2.114138438880707,
+      "grad_norm": 0.35486897826194763,
+      "learning_rate": 2.4298308344214745e-06,
+      "loss": 0.3849,
+      "step": 2871
+    },
+    {
+      "epoch": 2.1148748159057438,
+      "grad_norm": 0.38103657960891724,
+      "learning_rate": 2.4261564326526623e-06,
+      "loss": 0.3523,
+      "step": 2872
+    },
+    {
+      "epoch": 2.1156111929307806,
+      "grad_norm": 0.37378567457199097,
+      "learning_rate": 2.422483921036785e-06,
+      "loss": 0.3847,
+      "step": 2873
+    },
+    {
+      "epoch": 2.1163475699558174,
+      "grad_norm": 0.38318321108818054,
+      "learning_rate": 2.418813302270829e-06,
+      "loss": 0.3825,
+      "step": 2874
+    },
+    {
+      "epoch": 2.1170839469808542,
+      "grad_norm": 0.34228307008743286,
+      "learning_rate": 2.415144579050382e-06,
+      "loss": 0.3661,
+      "step": 2875
+    },
+    {
+      "epoch": 2.117820324005891,
+      "grad_norm": 0.3472125828266144,
+      "learning_rate": 2.411477754069645e-06,
+      "loss": 0.3491,
+      "step": 2876
+    },
+    {
+      "epoch": 2.118556701030928,
+      "grad_norm": 0.355693519115448,
+      "learning_rate": 2.4078128300214225e-06,
+      "loss": 0.3481,
+      "step": 2877
+    },
+    {
+      "epoch": 2.1192930780559647,
+      "grad_norm": 0.40808922052383423,
+      "learning_rate": 2.4041498095971253e-06,
+      "loss": 0.4022,
+      "step": 2878
+    },
+    {
+      "epoch": 2.1200294550810015,
+      "grad_norm": 0.3449529707431793,
+      "learning_rate": 2.4004886954867618e-06,
+      "loss": 0.361,
+      "step": 2879
+    },
+    {
+      "epoch": 2.1207658321060383,
+      "grad_norm": 0.35305774211883545,
+      "learning_rate": 2.3968294903789474e-06,
+      "loss": 0.3913,
+      "step": 2880
+    },
+    {
+      "epoch": 2.121502209131075,
+      "grad_norm": 0.3608260750770569,
+      "learning_rate": 2.393172196960891e-06,
+      "loss": 0.3521,
+      "step": 2881
+    },
+    {
+      "epoch": 2.122238586156112,
+      "grad_norm": 0.3713111877441406,
+      "learning_rate": 2.3895168179183947e-06,
+      "loss": 0.389,
+      "step": 2882
+    },
+    {
+      "epoch": 2.1229749631811488,
+      "grad_norm": 0.3446366488933563,
+      "learning_rate": 2.3858633559358635e-06,
+      "loss": 0.3765,
+      "step": 2883
+    },
+    {
+      "epoch": 2.1237113402061856,
+      "grad_norm": 0.36745747923851013,
+      "learning_rate": 2.3822118136962876e-06,
+      "loss": 0.38,
+      "step": 2884
+    },
+    {
+      "epoch": 2.1244477172312224,
+      "grad_norm": 0.3752921223640442,
+      "learning_rate": 2.378562193881248e-06,
+      "loss": 0.3714,
+      "step": 2885
+    },
+    {
+      "epoch": 2.125184094256259,
+      "grad_norm": 0.33874091506004333,
+      "learning_rate": 2.3749144991709174e-06,
+      "loss": 0.3857,
+      "step": 2886
+    },
+    {
+      "epoch": 2.125920471281296,
+      "grad_norm": 0.3628979027271271,
+      "learning_rate": 2.371268732244048e-06,
+      "loss": 0.3781,
+      "step": 2887
+    },
+    {
+      "epoch": 2.126656848306333,
+      "grad_norm": 0.36751553416252136,
+      "learning_rate": 2.367624895777987e-06,
+      "loss": 0.3642,
+      "step": 2888
+    },
+    {
+      "epoch": 2.1273932253313697,
+      "grad_norm": 0.3799552023410797,
+      "learning_rate": 2.3639829924486546e-06,
+      "loss": 0.3399,
+      "step": 2889
+    },
+    {
+      "epoch": 2.1281296023564065,
+      "grad_norm": 0.34371480345726013,
+      "learning_rate": 2.3603430249305532e-06,
+      "loss": 0.3956,
+      "step": 2890
+    },
+    {
+      "epoch": 2.1288659793814433,
+      "grad_norm": 0.3942258954048157,
+      "learning_rate": 2.356704995896768e-06,
+      "loss": 0.3714,
+      "step": 2891
+    },
+    {
+      "epoch": 2.12960235640648,
+      "grad_norm": 0.34906941652297974,
+      "learning_rate": 2.353068908018957e-06,
+      "loss": 0.3743,
+      "step": 2892
+    },
+    {
+      "epoch": 2.130338733431517,
+      "grad_norm": 0.3614822328090668,
+      "learning_rate": 2.3494347639673513e-06,
+      "loss": 0.3632,
+      "step": 2893
+    },
+    {
+      "epoch": 2.1310751104565537,
+      "grad_norm": 0.3540303111076355,
+      "learning_rate": 2.3458025664107587e-06,
+      "loss": 0.3442,
+      "step": 2894
+    },
+    {
+      "epoch": 2.1318114874815906,
+      "grad_norm": 0.3677811622619629,
+      "learning_rate": 2.342172318016552e-06,
+      "loss": 0.3421,
+      "step": 2895
+    },
+    {
+      "epoch": 2.1325478645066274,
+      "grad_norm": 0.3830873668193817,
+      "learning_rate": 2.33854402145068e-06,
+      "loss": 0.378,
+      "step": 2896
+    },
+    {
+      "epoch": 2.133284241531664,
+      "grad_norm": 0.405880868434906,
+      "learning_rate": 2.3349176793776523e-06,
+      "loss": 0.3765,
+      "step": 2897
+    },
+    {
+      "epoch": 2.134020618556701,
+      "grad_norm": 0.37011435627937317,
+      "learning_rate": 2.3312932944605433e-06,
+      "loss": 0.3852,
+      "step": 2898
+    },
+    {
+      "epoch": 2.134756995581738,
+      "grad_norm": 0.3674939274787903,
+      "learning_rate": 2.3276708693609947e-06,
+      "loss": 0.3663,
+      "step": 2899
+    },
+    {
+      "epoch": 2.1354933726067746,
+      "grad_norm": 0.3470740020275116,
+      "learning_rate": 2.324050406739205e-06,
+      "loss": 0.3652,
+      "step": 2900
+    },
+    {
+      "epoch": 2.1362297496318114,
+      "grad_norm": 0.3683933913707733,
+      "learning_rate": 2.32043190925393e-06,
+      "loss": 0.3959,
+      "step": 2901
+    },
+    {
+      "epoch": 2.1369661266568483,
+      "grad_norm": 0.3439938426017761,
+      "learning_rate": 2.316815379562491e-06,
+      "loss": 0.3843,
+      "step": 2902
+    },
+    {
+      "epoch": 2.137702503681885,
+      "grad_norm": 0.3798215389251709,
+      "learning_rate": 2.3132008203207508e-06,
+      "loss": 0.3494,
+      "step": 2903
+    },
+    {
+      "epoch": 2.138438880706922,
+      "grad_norm": 0.3750232458114624,
+      "learning_rate": 2.309588234183137e-06,
+      "loss": 0.388,
+      "step": 2904
+    },
+    {
+      "epoch": 2.1391752577319587,
+      "grad_norm": 0.34800633788108826,
+      "learning_rate": 2.3059776238026233e-06,
+      "loss": 0.3883,
+      "step": 2905
+    },
+    {
+      "epoch": 2.1399116347569955,
+      "grad_norm": 0.3639180660247803,
+      "learning_rate": 2.30236899183073e-06,
+      "loss": 0.3853,
+      "step": 2906
+    },
+    {
+      "epoch": 2.1406480117820323,
+      "grad_norm": 0.3714924156665802,
+      "learning_rate": 2.298762340917531e-06,
+      "loss": 0.415,
+      "step": 2907
+    },
+    {
+      "epoch": 2.141384388807069,
+      "grad_norm": 0.4161515235900879,
+      "learning_rate": 2.295157673711641e-06,
+      "loss": 0.3851,
+      "step": 2908
+    },
+    {
+      "epoch": 2.142120765832106,
+      "grad_norm": 0.35217177867889404,
+      "learning_rate": 2.2915549928602153e-06,
+      "loss": 0.3988,
+      "step": 2909
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.3613126575946808,
+      "learning_rate": 2.2879543010089613e-06,
+      "loss": 0.3599,
+      "step": 2910
+    },
+    {
+      "epoch": 2.1435935198821796,
+      "grad_norm": 0.34729936718940735,
+      "learning_rate": 2.2843556008021105e-06,
+      "loss": 0.3698,
+      "step": 2911
+    },
+    {
+      "epoch": 2.1443298969072164,
+      "grad_norm": 0.37736669182777405,
+      "learning_rate": 2.280758894882441e-06,
+      "loss": 0.3627,
+      "step": 2912
+    },
+    {
+      "epoch": 2.1450662739322532,
+      "grad_norm": 0.36949247121810913,
+      "learning_rate": 2.2771641858912684e-06,
+      "loss": 0.3889,
+      "step": 2913
+    },
+    {
+      "epoch": 2.14580265095729,
+      "grad_norm": 0.3440004587173462,
+      "learning_rate": 2.2735714764684368e-06,
+      "loss": 0.3574,
+      "step": 2914
+    },
+    {
+      "epoch": 2.146539027982327,
+      "grad_norm": 0.3486247658729553,
+      "learning_rate": 2.269980769252321e-06,
+      "loss": 0.3554,
+      "step": 2915
+    },
+    {
+      "epoch": 2.1472754050073637,
+      "grad_norm": 0.35198965668678284,
+      "learning_rate": 2.2663920668798316e-06,
+      "loss": 0.3627,
+      "step": 2916
+    },
+    {
+      "epoch": 2.1480117820324005,
+      "grad_norm": 0.37342336773872375,
+      "learning_rate": 2.262805371986402e-06,
+      "loss": 0.375,
+      "step": 2917
+    },
+    {
+      "epoch": 2.1487481590574373,
+      "grad_norm": 0.36178645491600037,
+      "learning_rate": 2.2592206872059913e-06,
+      "loss": 0.4036,
+      "step": 2918
+    },
+    {
+      "epoch": 2.149484536082474,
+      "grad_norm": 0.37667861580848694,
+      "learning_rate": 2.255638015171085e-06,
+      "loss": 0.3922,
+      "step": 2919
+    },
+    {
+      "epoch": 2.150220913107511,
+      "grad_norm": 0.3384239077568054,
+      "learning_rate": 2.2520573585126863e-06,
+      "loss": 0.3869,
+      "step": 2920
+    },
+    {
+      "epoch": 2.1509572901325478,
+      "grad_norm": 0.3617817759513855,
+      "learning_rate": 2.248478719860326e-06,
+      "loss": 0.3777,
+      "step": 2921
+    },
+    {
+      "epoch": 2.1516936671575846,
+      "grad_norm": 0.34681081771850586,
+      "learning_rate": 2.2449021018420454e-06,
+      "loss": 0.355,
+      "step": 2922
+    },
+    {
+      "epoch": 2.1524300441826214,
+      "grad_norm": 0.3603213131427765,
+      "learning_rate": 2.2413275070844026e-06,
+      "loss": 0.4205,
+      "step": 2923
+    },
+    {
+      "epoch": 2.153166421207658,
+      "grad_norm": 0.33105942606925964,
+      "learning_rate": 2.2377549382124767e-06,
+      "loss": 0.3259,
+      "step": 2924
+    },
+    {
+      "epoch": 2.153902798232695,
+      "grad_norm": 0.36263999342918396,
+      "learning_rate": 2.2341843978498525e-06,
+      "loss": 0.383,
+      "step": 2925
+    },
+    {
+      "epoch": 2.154639175257732,
+      "grad_norm": 0.34897348284721375,
+      "learning_rate": 2.230615888618624e-06,
+      "loss": 0.3873,
+      "step": 2926
+    },
+    {
+      "epoch": 2.1553755522827687,
+      "grad_norm": 0.3384336829185486,
+      "learning_rate": 2.2270494131394034e-06,
+      "loss": 0.3643,
+      "step": 2927
+    },
+    {
+      "epoch": 2.1561119293078055,
+      "grad_norm": 0.3632971942424774,
+      "learning_rate": 2.223484974031294e-06,
+      "loss": 0.3668,
+      "step": 2928
+    },
+    {
+      "epoch": 2.1568483063328423,
+      "grad_norm": 0.3339279592037201,
+      "learning_rate": 2.2199225739119184e-06,
+      "loss": 0.3488,
+      "step": 2929
+    },
+    {
+      "epoch": 2.157584683357879,
+      "grad_norm": 0.35420891642570496,
+      "learning_rate": 2.216362215397393e-06,
+      "loss": 0.3732,
+      "step": 2930
+    },
+    {
+      "epoch": 2.158321060382916,
+      "grad_norm": 0.35036933422088623,
+      "learning_rate": 2.2128039011023367e-06,
+      "loss": 0.3618,
+      "step": 2931
+    },
+    {
+      "epoch": 2.1590574374079528,
+      "grad_norm": 0.40435558557510376,
+      "learning_rate": 2.2092476336398706e-06,
+      "loss": 0.4052,
+      "step": 2932
+    },
+    {
+      "epoch": 2.1597938144329896,
+      "grad_norm": 0.32747867703437805,
+      "learning_rate": 2.2056934156216094e-06,
+      "loss": 0.3612,
+      "step": 2933
+    },
+    {
+      "epoch": 2.1605301914580264,
+      "grad_norm": 0.3270984888076782,
+      "learning_rate": 2.2021412496576598e-06,
+      "loss": 0.3827,
+      "step": 2934
+    },
+    {
+      "epoch": 2.161266568483063,
+      "grad_norm": 0.3557640314102173,
+      "learning_rate": 2.198591138356633e-06,
+      "loss": 0.3625,
+      "step": 2935
+    },
+    {
+      "epoch": 2.1620029455081,
+      "grad_norm": 0.3336459994316101,
+      "learning_rate": 2.195043084325616e-06,
+      "loss": 0.35,
+      "step": 2936
+    },
+    {
+      "epoch": 2.162739322533137,
+      "grad_norm": 0.3505702316761017,
+      "learning_rate": 2.191497090170193e-06,
+      "loss": 0.3816,
+      "step": 2937
+    },
+    {
+      "epoch": 2.1634756995581736,
+      "grad_norm": 0.37018442153930664,
+      "learning_rate": 2.1879531584944396e-06,
+      "loss": 0.3812,
+      "step": 2938
+    },
+    {
+      "epoch": 2.1642120765832105,
+      "grad_norm": 0.33809441328048706,
+      "learning_rate": 2.1844112919009087e-06,
+      "loss": 0.39,
+      "step": 2939
+    },
+    {
+      "epoch": 2.1649484536082473,
+      "grad_norm": 0.32393452525138855,
+      "learning_rate": 2.1808714929906394e-06,
+      "loss": 0.3462,
+      "step": 2940
+    },
+    {
+      "epoch": 2.165684830633284,
+      "grad_norm": 0.33284613490104675,
+      "learning_rate": 2.1773337643631565e-06,
+      "loss": 0.3845,
+      "step": 2941
+    },
+    {
+      "epoch": 2.166421207658321,
+      "grad_norm": 0.36973825097084045,
+      "learning_rate": 2.173798108616459e-06,
+      "loss": 0.3597,
+      "step": 2942
+    },
+    {
+      "epoch": 2.1671575846833577,
+      "grad_norm": 0.3738713264465332,
+      "learning_rate": 2.1702645283470238e-06,
+      "loss": 0.381,
+      "step": 2943
+    },
+    {
+      "epoch": 2.1678939617083945,
+      "grad_norm": 0.3623276352882385,
+      "learning_rate": 2.166733026149811e-06,
+      "loss": 0.3546,
+      "step": 2944
+    },
+    {
+      "epoch": 2.1686303387334314,
+      "grad_norm": 0.4014292061328888,
+      "learning_rate": 2.1632036046182416e-06,
+      "loss": 0.3645,
+      "step": 2945
+    },
+    {
+      "epoch": 2.169366715758468,
+      "grad_norm": 0.37799927592277527,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.3611,
+      "step": 2946
+    },
+    {
+      "epoch": 2.170103092783505,
+      "grad_norm": 0.379955917596817,
+      "learning_rate": 2.15615101391812e-06,
+      "loss": 0.3564,
+      "step": 2947
+    },
+    {
+      "epoch": 2.170839469808542,
+      "grad_norm": 0.36554086208343506,
+      "learning_rate": 2.1526278499287746e-06,
+      "loss": 0.3748,
+      "step": 2948
+    },
+    {
+      "epoch": 2.1715758468335786,
+      "grad_norm": 0.3485044538974762,
+      "learning_rate": 2.1491067769634927e-06,
+      "loss": 0.3738,
+      "step": 2949
+    },
+    {
+      "epoch": 2.1723122238586154,
+      "grad_norm": 0.4192858934402466,
+      "learning_rate": 2.145587797608043e-06,
+      "loss": 0.3907,
+      "step": 2950
+    },
+    {
+      "epoch": 2.1730486008836523,
+      "grad_norm": 0.3680093288421631,
+      "learning_rate": 2.1420709144466557e-06,
+      "loss": 0.3673,
+      "step": 2951
+    },
+    {
+      "epoch": 2.173784977908689,
+      "grad_norm": 0.3767206370830536,
+      "learning_rate": 2.1385561300620287e-06,
+      "loss": 0.3542,
+      "step": 2952
+    },
+    {
+      "epoch": 2.174521354933726,
+      "grad_norm": 0.3542313575744629,
+      "learning_rate": 2.1350434470353065e-06,
+      "loss": 0.3815,
+      "step": 2953
+    },
+    {
+      "epoch": 2.1752577319587627,
+      "grad_norm": 0.3587740957736969,
+      "learning_rate": 2.131532867946102e-06,
+      "loss": 0.3831,
+      "step": 2954
+    },
+    {
+      "epoch": 2.1759941089837995,
+      "grad_norm": 0.35472214221954346,
+      "learning_rate": 2.1280243953724784e-06,
+      "loss": 0.3649,
+      "step": 2955
+    },
+    {
+      "epoch": 2.1767304860088363,
+      "grad_norm": 0.37398767471313477,
+      "learning_rate": 2.1245180318909482e-06,
+      "loss": 0.3801,
+      "step": 2956
+    },
+    {
+      "epoch": 2.177466863033873,
+      "grad_norm": 0.3110312521457672,
+      "learning_rate": 2.121013780076483e-06,
+      "loss": 0.3539,
+      "step": 2957
+    },
+    {
+      "epoch": 2.17820324005891,
+      "grad_norm": 0.34271514415740967,
+      "learning_rate": 2.1175116425024978e-06,
+      "loss": 0.3796,
+      "step": 2958
+    },
+    {
+      "epoch": 2.178939617083947,
+      "grad_norm": 0.34827741980552673,
+      "learning_rate": 2.1140116217408554e-06,
+      "loss": 0.3739,
+      "step": 2959
+    },
+    {
+      "epoch": 2.1796759941089836,
+      "grad_norm": 0.327886700630188,
+      "learning_rate": 2.110513720361869e-06,
+      "loss": 0.3734,
+      "step": 2960
+    },
+    {
+      "epoch": 2.1804123711340204,
+      "grad_norm": 0.3339656889438629,
+      "learning_rate": 2.107017940934286e-06,
+      "loss": 0.3918,
+      "step": 2961
+    },
+    {
+      "epoch": 2.1811487481590572,
+      "grad_norm": 0.380188912153244,
+      "learning_rate": 2.1035242860253064e-06,
+      "loss": 0.3954,
+      "step": 2962
+    },
+    {
+      "epoch": 2.181885125184094,
+      "grad_norm": 0.33889806270599365,
+      "learning_rate": 2.100032758200562e-06,
+      "loss": 0.3591,
+      "step": 2963
+    },
+    {
+      "epoch": 2.182621502209131,
+      "grad_norm": 0.3376297056674957,
+      "learning_rate": 2.0965433600241247e-06,
+      "loss": 0.3968,
+      "step": 2964
+    },
+    {
+      "epoch": 2.1833578792341677,
+      "grad_norm": 0.3318977355957031,
+      "learning_rate": 2.093056094058506e-06,
+      "loss": 0.3602,
+      "step": 2965
+    },
+    {
+      "epoch": 2.184094256259205,
+      "grad_norm": 0.3732307255268097,
+      "learning_rate": 2.089570962864647e-06,
+      "loss": 0.3863,
+      "step": 2966
+    },
+    {
+      "epoch": 2.1848306332842418,
+      "grad_norm": 0.33816441893577576,
+      "learning_rate": 2.0860879690019216e-06,
+      "loss": 0.3826,
+      "step": 2967
+    },
+    {
+      "epoch": 2.1855670103092786,
+      "grad_norm": 0.38040101528167725,
+      "learning_rate": 2.0826071150281374e-06,
+      "loss": 0.3581,
+      "step": 2968
+    },
+    {
+      "epoch": 2.1863033873343154,
+      "grad_norm": 0.3592909276485443,
+      "learning_rate": 2.0791284034995296e-06,
+      "loss": 0.4081,
+      "step": 2969
+    },
+    {
+      "epoch": 2.187039764359352,
+      "grad_norm": 0.330105185508728,
+      "learning_rate": 2.0756518369707528e-06,
+      "loss": 0.3798,
+      "step": 2970
+    },
+    {
+      "epoch": 2.187776141384389,
+      "grad_norm": 0.34812307357788086,
+      "learning_rate": 2.0721774179948978e-06,
+      "loss": 0.4039,
+      "step": 2971
+    },
+    {
+      "epoch": 2.188512518409426,
+      "grad_norm": 0.39521628618240356,
+      "learning_rate": 2.0687051491234717e-06,
+      "loss": 0.3804,
+      "step": 2972
+    },
+    {
+      "epoch": 2.1892488954344627,
+      "grad_norm": 0.3390514552593231,
+      "learning_rate": 2.0652350329064012e-06,
+      "loss": 0.3691,
+      "step": 2973
+    },
+    {
+      "epoch": 2.1899852724594995,
+      "grad_norm": 0.38476595282554626,
+      "learning_rate": 2.061767071892039e-06,
+      "loss": 0.3884,
+      "step": 2974
+    },
+    {
+      "epoch": 2.1907216494845363,
+      "grad_norm": 0.35555726289749146,
+      "learning_rate": 2.0583012686271493e-06,
+      "loss": 0.3852,
+      "step": 2975
+    },
+    {
+      "epoch": 2.191458026509573,
+      "grad_norm": 0.3225259482860565,
+      "learning_rate": 2.0548376256569107e-06,
+      "loss": 0.3712,
+      "step": 2976
+    },
+    {
+      "epoch": 2.19219440353461,
+      "grad_norm": 0.3521542251110077,
+      "learning_rate": 2.051376145524924e-06,
+      "loss": 0.389,
+      "step": 2977
+    },
+    {
+      "epoch": 2.1929307805596467,
+      "grad_norm": 0.3622699975967407,
+      "learning_rate": 2.047916830773187e-06,
+      "loss": 0.3854,
+      "step": 2978
+    },
+    {
+      "epoch": 2.1936671575846836,
+      "grad_norm": 0.36842185258865356,
+      "learning_rate": 2.044459683942124e-06,
+      "loss": 0.3563,
+      "step": 2979
+    },
+    {
+      "epoch": 2.1944035346097204,
+      "grad_norm": 0.4047960042953491,
+      "learning_rate": 2.041004707570555e-06,
+      "loss": 0.3704,
+      "step": 2980
+    },
+    {
+      "epoch": 2.195139911634757,
+      "grad_norm": 0.3790772259235382,
+      "learning_rate": 2.037551904195709e-06,
+      "loss": 0.3918,
+      "step": 2981
+    },
+    {
+      "epoch": 2.195876288659794,
+      "grad_norm": 0.31942737102508545,
+      "learning_rate": 2.0341012763532243e-06,
+      "loss": 0.3562,
+      "step": 2982
+    },
+    {
+      "epoch": 2.196612665684831,
+      "grad_norm": 0.36221376061439514,
+      "learning_rate": 2.0306528265771357e-06,
+      "loss": 0.3856,
+      "step": 2983
+    },
+    {
+      "epoch": 2.1973490427098676,
+      "grad_norm": 0.37802380323410034,
+      "learning_rate": 2.0272065573998794e-06,
+      "loss": 0.3645,
+      "step": 2984
+    },
+    {
+      "epoch": 2.1980854197349045,
+      "grad_norm": 0.34147152304649353,
+      "learning_rate": 2.0237624713522945e-06,
+      "loss": 0.3898,
+      "step": 2985
+    },
+    {
+      "epoch": 2.1988217967599413,
+      "grad_norm": 0.3583637475967407,
+      "learning_rate": 2.020320570963612e-06,
+      "loss": 0.3768,
+      "step": 2986
+    },
+    {
+      "epoch": 2.199558173784978,
+      "grad_norm": 0.3441274166107178,
+      "learning_rate": 2.0168808587614584e-06,
+      "loss": 0.3712,
+      "step": 2987
+    },
+    {
+      "epoch": 2.200294550810015,
+      "grad_norm": 0.3442659080028534,
+      "learning_rate": 2.0134433372718565e-06,
+      "loss": 0.3423,
+      "step": 2988
+    },
+    {
+      "epoch": 2.2010309278350517,
+      "grad_norm": 0.34774288535118103,
+      "learning_rate": 2.010008009019215e-06,
+      "loss": 0.3854,
+      "step": 2989
+    },
+    {
+      "epoch": 2.2017673048600885,
+      "grad_norm": 0.3395615518093109,
+      "learning_rate": 2.0065748765263386e-06,
+      "loss": 0.3523,
+      "step": 2990
+    },
+    {
+      "epoch": 2.2025036818851254,
+      "grad_norm": 0.3748759627342224,
+      "learning_rate": 2.003143942314415e-06,
+      "loss": 0.3714,
+      "step": 2991
+    },
+    {
+      "epoch": 2.203240058910162,
+      "grad_norm": 0.34823179244995117,
+      "learning_rate": 1.999715208903017e-06,
+      "loss": 0.3958,
+      "step": 2992
+    },
+    {
+      "epoch": 2.203976435935199,
+      "grad_norm": 0.32306548953056335,
+      "learning_rate": 1.996288678810105e-06,
+      "loss": 0.3813,
+      "step": 2993
+    },
+    {
+      "epoch": 2.204712812960236,
+      "grad_norm": 0.3721083700656891,
+      "learning_rate": 1.9928643545520204e-06,
+      "loss": 0.3786,
+      "step": 2994
+    },
+    {
+      "epoch": 2.2054491899852726,
+      "grad_norm": 0.34985750913619995,
+      "learning_rate": 1.989442238643478e-06,
+      "loss": 0.3816,
+      "step": 2995
+    },
+    {
+      "epoch": 2.2061855670103094,
+      "grad_norm": 0.3577035367488861,
+      "learning_rate": 1.9860223335975815e-06,
+      "loss": 0.355,
+      "step": 2996
+    },
+    {
+      "epoch": 2.2069219440353463,
+      "grad_norm": 0.3292270302772522,
+      "learning_rate": 1.9826046419258037e-06,
+      "loss": 0.3322,
+      "step": 2997
+    },
+    {
+      "epoch": 2.207658321060383,
+      "grad_norm": 0.3422560393810272,
+      "learning_rate": 1.9791891661379926e-06,
+      "loss": 0.3514,
+      "step": 2998
+    },
+    {
+      "epoch": 2.20839469808542,
+      "grad_norm": 0.3499544560909271,
+      "learning_rate": 1.975775908742374e-06,
+      "loss": 0.3692,
+      "step": 2999
+    },
+    {
+      "epoch": 2.2091310751104567,
+      "grad_norm": 0.35650497674942017,
+      "learning_rate": 1.972364872245539e-06,
+      "loss": 0.382,
+      "step": 3000
+    },
+    {
+      "epoch": 2.2098674521354935,
+      "grad_norm": 0.3358071446418762,
+      "learning_rate": 1.9689560591524482e-06,
+      "loss": 0.3655,
+      "step": 3001
+    },
+    {
+      "epoch": 2.2106038291605303,
+      "grad_norm": 0.3230755925178528,
+      "learning_rate": 1.965549471966436e-06,
+      "loss": 0.3629,
+      "step": 3002
+    },
+    {
+      "epoch": 2.211340206185567,
+      "grad_norm": 0.33048614859580994,
+      "learning_rate": 1.96214511318919e-06,
+      "loss": 0.3856,
+      "step": 3003
+    },
+    {
+      "epoch": 2.212076583210604,
+      "grad_norm": 0.36040183901786804,
+      "learning_rate": 1.958742985320774e-06,
+      "loss": 0.3601,
+      "step": 3004
+    },
+    {
+      "epoch": 2.212812960235641,
+      "grad_norm": 0.3354277014732361,
+      "learning_rate": 1.955343090859606e-06,
+      "loss": 0.3607,
+      "step": 3005
+    },
+    {
+      "epoch": 2.2135493372606776,
+      "grad_norm": 0.3312050700187683,
+      "learning_rate": 1.9519454323024644e-06,
+      "loss": 0.3716,
+      "step": 3006
+    },
+    {
+      "epoch": 2.2142857142857144,
+      "grad_norm": 0.34683001041412354,
+      "learning_rate": 1.9485500121444896e-06,
+      "loss": 0.3692,
+      "step": 3007
+    },
+    {
+      "epoch": 2.2150220913107512,
+      "grad_norm": 0.3769376575946808,
+      "learning_rate": 1.945156832879174e-06,
+      "loss": 0.3664,
+      "step": 3008
+    },
+    {
+      "epoch": 2.215758468335788,
+      "grad_norm": 0.33873477578163147,
+      "learning_rate": 1.941765896998365e-06,
+      "loss": 0.3909,
+      "step": 3009
+    },
+    {
+      "epoch": 2.216494845360825,
+      "grad_norm": 0.3695991635322571,
+      "learning_rate": 1.938377206992266e-06,
+      "loss": 0.3717,
+      "step": 3010
+    },
+    {
+      "epoch": 2.2172312223858617,
+      "grad_norm": 0.3528919816017151,
+      "learning_rate": 1.934990765349427e-06,
+      "loss": 0.3708,
+      "step": 3011
+    },
+    {
+      "epoch": 2.2179675994108985,
+      "grad_norm": 0.3522799015045166,
+      "learning_rate": 1.931606574556749e-06,
+      "loss": 0.373,
+      "step": 3012
+    },
+    {
+      "epoch": 2.2187039764359353,
+      "grad_norm": 0.3132975101470947,
+      "learning_rate": 1.928224637099479e-06,
+      "loss": 0.35,
+      "step": 3013
+    },
+    {
+      "epoch": 2.219440353460972,
+      "grad_norm": 0.35459184646606445,
+      "learning_rate": 1.9248449554612076e-06,
+      "loss": 0.3724,
+      "step": 3014
+    },
+    {
+      "epoch": 2.220176730486009,
+      "grad_norm": 0.367097944021225,
+      "learning_rate": 1.9214675321238753e-06,
+      "loss": 0.3754,
+      "step": 3015
+    },
+    {
+      "epoch": 2.2209131075110458,
+      "grad_norm": 0.38313212990760803,
+      "learning_rate": 1.9180923695677565e-06,
+      "loss": 0.3632,
+      "step": 3016
+    },
+    {
+      "epoch": 2.2216494845360826,
+      "grad_norm": 0.3429620563983917,
+      "learning_rate": 1.9147194702714683e-06,
+      "loss": 0.3543,
+      "step": 3017
+    },
+    {
+      "epoch": 2.2223858615611194,
+      "grad_norm": 0.31693652272224426,
+      "learning_rate": 1.911348836711969e-06,
+      "loss": 0.3526,
+      "step": 3018
+    },
+    {
+      "epoch": 2.223122238586156,
+      "grad_norm": 0.33426809310913086,
+      "learning_rate": 1.907980471364548e-06,
+      "loss": 0.3795,
+      "step": 3019
+    },
+    {
+      "epoch": 2.223858615611193,
+      "grad_norm": 0.3561391532421112,
+      "learning_rate": 1.9046143767028309e-06,
+      "loss": 0.3824,
+      "step": 3020
+    },
+    {
+      "epoch": 2.22459499263623,
+      "grad_norm": 0.32854175567626953,
+      "learning_rate": 1.9012505551987764e-06,
+      "loss": 0.3357,
+      "step": 3021
+    },
+    {
+      "epoch": 2.2253313696612667,
+      "grad_norm": 0.34215104579925537,
+      "learning_rate": 1.897889009322672e-06,
+      "loss": 0.3826,
+      "step": 3022
+    },
+    {
+      "epoch": 2.2260677466863035,
+      "grad_norm": 0.3710014522075653,
+      "learning_rate": 1.8945297415431379e-06,
+      "loss": 0.3799,
+      "step": 3023
+    },
+    {
+      "epoch": 2.2268041237113403,
+      "grad_norm": 0.35444051027297974,
+      "learning_rate": 1.8911727543271174e-06,
+      "loss": 0.396,
+      "step": 3024
+    },
+    {
+      "epoch": 2.227540500736377,
+      "grad_norm": 0.3570059835910797,
+      "learning_rate": 1.8878180501398796e-06,
+      "loss": 0.3755,
+      "step": 3025
+    },
+    {
+      "epoch": 2.228276877761414,
+      "grad_norm": 0.34448522329330444,
+      "learning_rate": 1.88446563144502e-06,
+      "loss": 0.4058,
+      "step": 3026
+    },
+    {
+      "epoch": 2.2290132547864507,
+      "grad_norm": 0.3360097408294678,
+      "learning_rate": 1.8811155007044523e-06,
+      "loss": 0.3788,
+      "step": 3027
+    },
+    {
+      "epoch": 2.2297496318114876,
+      "grad_norm": 0.35772281885147095,
+      "learning_rate": 1.8777676603784122e-06,
+      "loss": 0.4028,
+      "step": 3028
+    },
+    {
+      "epoch": 2.2304860088365244,
+      "grad_norm": 0.31866273283958435,
+      "learning_rate": 1.8744221129254514e-06,
+      "loss": 0.3906,
+      "step": 3029
+    },
+    {
+      "epoch": 2.231222385861561,
+      "grad_norm": 0.3577582538127899,
+      "learning_rate": 1.871078860802439e-06,
+      "loss": 0.3542,
+      "step": 3030
+    },
+    {
+      "epoch": 2.231958762886598,
+      "grad_norm": 0.34226885437965393,
+      "learning_rate": 1.8677379064645567e-06,
+      "loss": 0.3399,
+      "step": 3031
+    },
+    {
+      "epoch": 2.232695139911635,
+      "grad_norm": 0.35378527641296387,
+      "learning_rate": 1.8643992523653043e-06,
+      "loss": 0.3938,
+      "step": 3032
+    },
+    {
+      "epoch": 2.2334315169366716,
+      "grad_norm": 0.35914602875709534,
+      "learning_rate": 1.8610629009564863e-06,
+      "loss": 0.3751,
+      "step": 3033
+    },
+    {
+      "epoch": 2.2341678939617085,
+      "grad_norm": 0.35784912109375,
+      "learning_rate": 1.8577288546882167e-06,
+      "loss": 0.3771,
+      "step": 3034
+    },
+    {
+      "epoch": 2.2349042709867453,
+      "grad_norm": 0.3862041234970093,
+      "learning_rate": 1.8543971160089213e-06,
+      "loss": 0.3834,
+      "step": 3035
+    },
+    {
+      "epoch": 2.235640648011782,
+      "grad_norm": 0.38658207654953003,
+      "learning_rate": 1.8510676873653278e-06,
+      "loss": 0.3619,
+      "step": 3036
+    },
+    {
+      "epoch": 2.236377025036819,
+      "grad_norm": 0.3452564775943756,
+      "learning_rate": 1.8477405712024671e-06,
+      "loss": 0.3496,
+      "step": 3037
+    },
+    {
+      "epoch": 2.2371134020618557,
+      "grad_norm": 0.37417715787887573,
+      "learning_rate": 1.8444157699636728e-06,
+      "loss": 0.3811,
+      "step": 3038
+    },
+    {
+      "epoch": 2.2378497790868925,
+      "grad_norm": 0.3347567319869995,
+      "learning_rate": 1.8410932860905767e-06,
+      "loss": 0.3752,
+      "step": 3039
+    },
+    {
+      "epoch": 2.2385861561119293,
+      "grad_norm": 0.3563539981842041,
+      "learning_rate": 1.8377731220231144e-06,
+      "loss": 0.3723,
+      "step": 3040
+    },
+    {
+      "epoch": 2.239322533136966,
+      "grad_norm": 0.33468982577323914,
+      "learning_rate": 1.834455280199512e-06,
+      "loss": 0.3565,
+      "step": 3041
+    },
+    {
+      "epoch": 2.240058910162003,
+      "grad_norm": 0.37427690625190735,
+      "learning_rate": 1.8311397630562905e-06,
+      "loss": 0.3703,
+      "step": 3042
+    },
+    {
+      "epoch": 2.24079528718704,
+      "grad_norm": 0.3921281099319458,
+      "learning_rate": 1.8278265730282696e-06,
+      "loss": 0.3775,
+      "step": 3043
+    },
+    {
+      "epoch": 2.2415316642120766,
+      "grad_norm": 0.3752121925354004,
+      "learning_rate": 1.824515712548553e-06,
+      "loss": 0.3651,
+      "step": 3044
+    },
+    {
+      "epoch": 2.2422680412371134,
+      "grad_norm": 0.34277763962745667,
+      "learning_rate": 1.821207184048538e-06,
+      "loss": 0.3717,
+      "step": 3045
+    },
+    {
+      "epoch": 2.2430044182621502,
+      "grad_norm": 0.36394500732421875,
+      "learning_rate": 1.8179009899579069e-06,
+      "loss": 0.3698,
+      "step": 3046
+    },
+    {
+      "epoch": 2.243740795287187,
+      "grad_norm": 0.3427956700325012,
+      "learning_rate": 1.8145971327046274e-06,
+      "loss": 0.377,
+      "step": 3047
+    },
+    {
+      "epoch": 2.244477172312224,
+      "grad_norm": 0.3212227523326874,
+      "learning_rate": 1.8112956147149558e-06,
+      "loss": 0.3747,
+      "step": 3048
+    },
+    {
+      "epoch": 2.2452135493372607,
+      "grad_norm": 0.3556148409843445,
+      "learning_rate": 1.8079964384134252e-06,
+      "loss": 0.3795,
+      "step": 3049
+    },
+    {
+      "epoch": 2.2459499263622975,
+      "grad_norm": 0.3533985912799835,
+      "learning_rate": 1.80469960622285e-06,
+      "loss": 0.3946,
+      "step": 3050
+    },
+    {
+      "epoch": 2.2466863033873343,
+      "grad_norm": 0.32733723521232605,
+      "learning_rate": 1.8014051205643268e-06,
+      "loss": 0.3854,
+      "step": 3051
+    },
+    {
+      "epoch": 2.247422680412371,
+      "grad_norm": 0.35768672823905945,
+      "learning_rate": 1.7981129838572248e-06,
+      "loss": 0.3638,
+      "step": 3052
+    },
+    {
+      "epoch": 2.248159057437408,
+      "grad_norm": 0.36508840322494507,
+      "learning_rate": 1.79482319851919e-06,
+      "loss": 0.3859,
+      "step": 3053
+    },
+    {
+      "epoch": 2.2488954344624448,
+      "grad_norm": 0.33305230736732483,
+      "learning_rate": 1.7915357669661409e-06,
+      "loss": 0.3635,
+      "step": 3054
+    },
+    {
+      "epoch": 2.2496318114874816,
+      "grad_norm": 0.324288934469223,
+      "learning_rate": 1.7882506916122683e-06,
+      "loss": 0.3887,
+      "step": 3055
+    },
+    {
+      "epoch": 2.2503681885125184,
+      "grad_norm": 0.3586910665035248,
+      "learning_rate": 1.7849679748700305e-06,
+      "loss": 0.3818,
+      "step": 3056
+    },
+    {
+      "epoch": 2.2511045655375552,
+      "grad_norm": 0.3524543344974518,
+      "learning_rate": 1.7816876191501587e-06,
+      "loss": 0.4012,
+      "step": 3057
+    },
+    {
+      "epoch": 2.251840942562592,
+      "grad_norm": 0.3519776463508606,
+      "learning_rate": 1.7784096268616453e-06,
+      "loss": 0.3727,
+      "step": 3058
+    },
+    {
+      "epoch": 2.252577319587629,
+      "grad_norm": 0.36303767561912537,
+      "learning_rate": 1.7751340004117468e-06,
+      "loss": 0.349,
+      "step": 3059
+    },
+    {
+      "epoch": 2.2533136966126657,
+      "grad_norm": 0.3571464717388153,
+      "learning_rate": 1.771860742205988e-06,
+      "loss": 0.3939,
+      "step": 3060
+    },
+    {
+      "epoch": 2.2540500736377025,
+      "grad_norm": 0.33723917603492737,
+      "learning_rate": 1.7685898546481495e-06,
+      "loss": 0.3779,
+      "step": 3061
+    },
+    {
+      "epoch": 2.2547864506627393,
+      "grad_norm": 0.3464319109916687,
+      "learning_rate": 1.7653213401402718e-06,
+      "loss": 0.36,
+      "step": 3062
+    },
+    {
+      "epoch": 2.255522827687776,
+      "grad_norm": 0.3434007167816162,
+      "learning_rate": 1.7620552010826535e-06,
+      "loss": 0.3766,
+      "step": 3063
+    },
+    {
+      "epoch": 2.256259204712813,
+      "grad_norm": 0.3645339608192444,
+      "learning_rate": 1.7587914398738466e-06,
+      "loss": 0.3689,
+      "step": 3064
+    },
+    {
+      "epoch": 2.2569955817378498,
+      "grad_norm": 0.33665645122528076,
+      "learning_rate": 1.7555300589106616e-06,
+      "loss": 0.397,
+      "step": 3065
+    },
+    {
+      "epoch": 2.2577319587628866,
+      "grad_norm": 0.312653124332428,
+      "learning_rate": 1.752271060588157e-06,
+      "loss": 0.3567,
+      "step": 3066
+    },
+    {
+      "epoch": 2.2584683357879234,
+      "grad_norm": 0.34965354204177856,
+      "learning_rate": 1.7490144472996412e-06,
+      "loss": 0.3739,
+      "step": 3067
+    },
+    {
+      "epoch": 2.25920471281296,
+      "grad_norm": 0.30937907099723816,
+      "learning_rate": 1.7457602214366754e-06,
+      "loss": 0.3815,
+      "step": 3068
+    },
+    {
+      "epoch": 2.259941089837997,
+      "grad_norm": 0.3516156077384949,
+      "learning_rate": 1.7425083853890628e-06,
+      "loss": 0.3779,
+      "step": 3069
+    },
+    {
+      "epoch": 2.260677466863034,
+      "grad_norm": 0.36380839347839355,
+      "learning_rate": 1.7392589415448546e-06,
+      "loss": 0.3853,
+      "step": 3070
+    },
+    {
+      "epoch": 2.2614138438880707,
+      "grad_norm": 0.3635159134864807,
+      "learning_rate": 1.736011892290343e-06,
+      "loss": 0.3666,
+      "step": 3071
+    },
+    {
+      "epoch": 2.2621502209131075,
+      "grad_norm": 0.3385698199272156,
+      "learning_rate": 1.732767240010062e-06,
+      "loss": 0.3674,
+      "step": 3072
+    },
+    {
+      "epoch": 2.2628865979381443,
+      "grad_norm": 0.3308032155036926,
+      "learning_rate": 1.7295249870867898e-06,
+      "loss": 0.3906,
+      "step": 3073
+    },
+    {
+      "epoch": 2.263622974963181,
+      "grad_norm": 0.3098700940608978,
+      "learning_rate": 1.726285135901536e-06,
+      "loss": 0.3532,
+      "step": 3074
+    },
+    {
+      "epoch": 2.264359351988218,
+      "grad_norm": 0.33167368173599243,
+      "learning_rate": 1.7230476888335484e-06,
+      "loss": 0.3623,
+      "step": 3075
+    },
+    {
+      "epoch": 2.2650957290132547,
+      "grad_norm": 0.3469424247741699,
+      "learning_rate": 1.7198126482603144e-06,
+      "loss": 0.3977,
+      "step": 3076
+    },
+    {
+      "epoch": 2.2658321060382915,
+      "grad_norm": 0.3413713574409485,
+      "learning_rate": 1.7165800165575475e-06,
+      "loss": 0.3831,
+      "step": 3077
+    },
+    {
+      "epoch": 2.2665684830633284,
+      "grad_norm": 0.3475739061832428,
+      "learning_rate": 1.7133497960991945e-06,
+      "loss": 0.3655,
+      "step": 3078
+    },
+    {
+      "epoch": 2.267304860088365,
+      "grad_norm": 0.3709220886230469,
+      "learning_rate": 1.7101219892574321e-06,
+      "loss": 0.3622,
+      "step": 3079
+    },
+    {
+      "epoch": 2.268041237113402,
+      "grad_norm": 0.3361673951148987,
+      "learning_rate": 1.706896598402663e-06,
+      "loss": 0.3978,
+      "step": 3080
+    },
+    {
+      "epoch": 2.268777614138439,
+      "grad_norm": 0.3467946946620941,
+      "learning_rate": 1.7036736259035197e-06,
+      "loss": 0.3752,
+      "step": 3081
+    },
+    {
+      "epoch": 2.2695139911634756,
+      "grad_norm": 0.35126227140426636,
+      "learning_rate": 1.7004530741268532e-06,
+      "loss": 0.394,
+      "step": 3082
+    },
+    {
+      "epoch": 2.2702503681885124,
+      "grad_norm": 0.36080968379974365,
+      "learning_rate": 1.697234945437739e-06,
+      "loss": 0.3753,
+      "step": 3083
+    },
+    {
+      "epoch": 2.2709867452135493,
+      "grad_norm": 0.35190704464912415,
+      "learning_rate": 1.6940192421994766e-06,
+      "loss": 0.3586,
+      "step": 3084
+    },
+    {
+      "epoch": 2.271723122238586,
+      "grad_norm": 0.31936824321746826,
+      "learning_rate": 1.6908059667735793e-06,
+      "loss": 0.3687,
+      "step": 3085
+    },
+    {
+      "epoch": 2.272459499263623,
+      "grad_norm": 0.33945798873901367,
+      "learning_rate": 1.6875951215197779e-06,
+      "loss": 0.3583,
+      "step": 3086
+    },
+    {
+      "epoch": 2.2731958762886597,
+      "grad_norm": 0.3654190003871918,
+      "learning_rate": 1.6843867087960252e-06,
+      "loss": 0.3991,
+      "step": 3087
+    },
+    {
+      "epoch": 2.2739322533136965,
+      "grad_norm": 0.33993858098983765,
+      "learning_rate": 1.6811807309584776e-06,
+      "loss": 0.3679,
+      "step": 3088
+    },
+    {
+      "epoch": 2.2746686303387333,
+      "grad_norm": 0.35775047540664673,
+      "learning_rate": 1.6779771903615083e-06,
+      "loss": 0.3821,
+      "step": 3089
+    },
+    {
+      "epoch": 2.27540500736377,
+      "grad_norm": 0.33738553524017334,
+      "learning_rate": 1.6747760893577037e-06,
+      "loss": 0.3891,
+      "step": 3090
+    },
+    {
+      "epoch": 2.276141384388807,
+      "grad_norm": 0.39476099610328674,
+      "learning_rate": 1.6715774302978544e-06,
+      "loss": 0.3794,
+      "step": 3091
+    },
+    {
+      "epoch": 2.276877761413844,
+      "grad_norm": 0.3633752167224884,
+      "learning_rate": 1.6683812155309577e-06,
+      "loss": 0.3904,
+      "step": 3092
+    },
+    {
+      "epoch": 2.2776141384388806,
+      "grad_norm": 0.31952276825904846,
+      "learning_rate": 1.665187447404219e-06,
+      "loss": 0.3837,
+      "step": 3093
+    },
+    {
+      "epoch": 2.2783505154639174,
+      "grad_norm": 0.33152905106544495,
+      "learning_rate": 1.6619961282630453e-06,
+      "loss": 0.3555,
+      "step": 3094
+    },
+    {
+      "epoch": 2.2790868924889542,
+      "grad_norm": 0.3404163420200348,
+      "learning_rate": 1.6588072604510435e-06,
+      "loss": 0.3725,
+      "step": 3095
+    },
+    {
+      "epoch": 2.279823269513991,
+      "grad_norm": 0.36314231157302856,
+      "learning_rate": 1.6556208463100226e-06,
+      "loss": 0.3619,
+      "step": 3096
+    },
+    {
+      "epoch": 2.280559646539028,
+      "grad_norm": 0.38192078471183777,
+      "learning_rate": 1.6524368881799863e-06,
+      "loss": 0.3697,
+      "step": 3097
+    },
+    {
+      "epoch": 2.2812960235640647,
+      "grad_norm": 0.3242843747138977,
+      "learning_rate": 1.6492553883991418e-06,
+      "loss": 0.3806,
+      "step": 3098
+    },
+    {
+      "epoch": 2.2820324005891015,
+      "grad_norm": 0.33095407485961914,
+      "learning_rate": 1.646076349303884e-06,
+      "loss": 0.382,
+      "step": 3099
+    },
+    {
+      "epoch": 2.2827687776141383,
+      "grad_norm": 0.371073454618454,
+      "learning_rate": 1.642899773228801e-06,
+      "loss": 0.4286,
+      "step": 3100
+    },
+    {
+      "epoch": 2.283505154639175,
+      "grad_norm": 0.35626521706581116,
+      "learning_rate": 1.6397256625066787e-06,
+      "loss": 0.3677,
+      "step": 3101
+    },
+    {
+      "epoch": 2.284241531664212,
+      "grad_norm": 0.3136880695819855,
+      "learning_rate": 1.6365540194684853e-06,
+      "loss": 0.3528,
+      "step": 3102
+    },
+    {
+      "epoch": 2.2849779086892488,
+      "grad_norm": 0.35781604051589966,
+      "learning_rate": 1.633384846443381e-06,
+      "loss": 0.3813,
+      "step": 3103
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.3532848358154297,
+      "learning_rate": 1.6302181457587092e-06,
+      "loss": 0.368,
+      "step": 3104
+    },
+    {
+      "epoch": 2.2864506627393224,
+      "grad_norm": 0.3363364040851593,
+      "learning_rate": 1.6270539197399988e-06,
+      "loss": 0.3679,
+      "step": 3105
+    },
+    {
+      "epoch": 2.287187039764359,
+      "grad_norm": 0.347859263420105,
+      "learning_rate": 1.6238921707109639e-06,
+      "loss": 0.371,
+      "step": 3106
+    },
+    {
+      "epoch": 2.287923416789396,
+      "grad_norm": 0.3195333480834961,
+      "learning_rate": 1.620732900993497e-06,
+      "loss": 0.3878,
+      "step": 3107
+    },
+    {
+      "epoch": 2.288659793814433,
+      "grad_norm": 0.35446897149086,
+      "learning_rate": 1.6175761129076673e-06,
+      "loss": 0.4139,
+      "step": 3108
+    },
+    {
+      "epoch": 2.2893961708394697,
+      "grad_norm": 0.35336872935295105,
+      "learning_rate": 1.614421808771729e-06,
+      "loss": 0.3528,
+      "step": 3109
+    },
+    {
+      "epoch": 2.2901325478645065,
+      "grad_norm": 0.3762224018573761,
+      "learning_rate": 1.6112699909021057e-06,
+      "loss": 0.3475,
+      "step": 3110
+    },
+    {
+      "epoch": 2.2908689248895433,
+      "grad_norm": 0.39090195298194885,
+      "learning_rate": 1.608120661613396e-06,
+      "loss": 0.3518,
+      "step": 3111
+    },
+    {
+      "epoch": 2.29160530191458,
+      "grad_norm": 0.33557429909706116,
+      "learning_rate": 1.604973823218376e-06,
+      "loss": 0.3848,
+      "step": 3112
+    },
+    {
+      "epoch": 2.292341678939617,
+      "grad_norm": 0.3590867519378662,
+      "learning_rate": 1.6018294780279848e-06,
+      "loss": 0.3553,
+      "step": 3113
+    },
+    {
+      "epoch": 2.2930780559646537,
+      "grad_norm": 0.3268950581550598,
+      "learning_rate": 1.598687628351334e-06,
+      "loss": 0.3818,
+      "step": 3114
+    },
+    {
+      "epoch": 2.2938144329896906,
+      "grad_norm": 0.3529122769832611,
+      "learning_rate": 1.5955482764957063e-06,
+      "loss": 0.3755,
+      "step": 3115
+    },
+    {
+      "epoch": 2.2945508100147274,
+      "grad_norm": 0.32063764333724976,
+      "learning_rate": 1.5924114247665457e-06,
+      "loss": 0.3589,
+      "step": 3116
+    },
+    {
+      "epoch": 2.295287187039764,
+      "grad_norm": 0.3151834309101105,
+      "learning_rate": 1.5892770754674596e-06,
+      "loss": 0.3749,
+      "step": 3117
+    },
+    {
+      "epoch": 2.296023564064801,
+      "grad_norm": 0.3227955996990204,
+      "learning_rate": 1.5861452309002219e-06,
+      "loss": 0.3449,
+      "step": 3118
+    },
+    {
+      "epoch": 2.296759941089838,
+      "grad_norm": 0.3374328315258026,
+      "learning_rate": 1.5830158933647638e-06,
+      "loss": 0.3702,
+      "step": 3119
+    },
+    {
+      "epoch": 2.2974963181148746,
+      "grad_norm": 0.34967485070228577,
+      "learning_rate": 1.5798890651591759e-06,
+      "loss": 0.3723,
+      "step": 3120
+    },
+    {
+      "epoch": 2.2982326951399115,
+      "grad_norm": 0.32275229692459106,
+      "learning_rate": 1.576764748579706e-06,
+      "loss": 0.3772,
+      "step": 3121
+    },
+    {
+      "epoch": 2.2989690721649483,
+      "grad_norm": 0.3539294898509979,
+      "learning_rate": 1.5736429459207569e-06,
+      "loss": 0.3473,
+      "step": 3122
+    },
+    {
+      "epoch": 2.299705449189985,
+      "grad_norm": 0.35446813702583313,
+      "learning_rate": 1.570523659474889e-06,
+      "loss": 0.3826,
+      "step": 3123
+    },
+    {
+      "epoch": 2.300441826215022,
+      "grad_norm": 0.3261672258377075,
+      "learning_rate": 1.5674068915328105e-06,
+      "loss": 0.3832,
+      "step": 3124
+    },
+    {
+      "epoch": 2.3011782032400587,
+      "grad_norm": 0.32315734028816223,
+      "learning_rate": 1.56429264438338e-06,
+      "loss": 0.3884,
+      "step": 3125
+    },
+    {
+      "epoch": 2.3019145802650955,
+      "grad_norm": 0.32102611660957336,
+      "learning_rate": 1.561180920313609e-06,
+      "loss": 0.401,
+      "step": 3126
+    },
+    {
+      "epoch": 2.3026509572901324,
+      "grad_norm": 0.3752593398094177,
+      "learning_rate": 1.5580717216086533e-06,
+      "loss": 0.3684,
+      "step": 3127
+    },
+    {
+      "epoch": 2.303387334315169,
+      "grad_norm": 0.331100732088089,
+      "learning_rate": 1.5549650505518115e-06,
+      "loss": 0.3856,
+      "step": 3128
+    },
+    {
+      "epoch": 2.304123711340206,
+      "grad_norm": 0.3341582417488098,
+      "learning_rate": 1.5518609094245351e-06,
+      "loss": 0.3876,
+      "step": 3129
+    },
+    {
+      "epoch": 2.304860088365243,
+      "grad_norm": 0.3470800817012787,
+      "learning_rate": 1.5487593005064038e-06,
+      "loss": 0.3762,
+      "step": 3130
+    },
+    {
+      "epoch": 2.3055964653902796,
+      "grad_norm": 0.3255835175514221,
+      "learning_rate": 1.5456602260751513e-06,
+      "loss": 0.3874,
+      "step": 3131
+    },
+    {
+      "epoch": 2.3063328424153164,
+      "grad_norm": 0.317374587059021,
+      "learning_rate": 1.5425636884066426e-06,
+      "loss": 0.4035,
+      "step": 3132
+    },
+    {
+      "epoch": 2.3070692194403533,
+      "grad_norm": 0.3365512788295746,
+      "learning_rate": 1.539469689774879e-06,
+      "loss": 0.3671,
+      "step": 3133
+    },
+    {
+      "epoch": 2.30780559646539,
+      "grad_norm": 0.34714415669441223,
+      "learning_rate": 1.5363782324520033e-06,
+      "loss": 0.3717,
+      "step": 3134
+    },
+    {
+      "epoch": 2.308541973490427,
+      "grad_norm": 0.3375949263572693,
+      "learning_rate": 1.5332893187082864e-06,
+      "loss": 0.3973,
+      "step": 3135
+    },
+    {
+      "epoch": 2.3092783505154637,
+      "grad_norm": 0.305399626493454,
+      "learning_rate": 1.5302029508121325e-06,
+      "loss": 0.38,
+      "step": 3136
+    },
+    {
+      "epoch": 2.3100147275405005,
+      "grad_norm": 0.3382294774055481,
+      "learning_rate": 1.5271191310300803e-06,
+      "loss": 0.3695,
+      "step": 3137
+    },
+    {
+      "epoch": 2.3107511045655373,
+      "grad_norm": 0.35040998458862305,
+      "learning_rate": 1.5240378616267887e-06,
+      "loss": 0.3784,
+      "step": 3138
+    },
+    {
+      "epoch": 2.311487481590574,
+      "grad_norm": 0.3383118510246277,
+      "learning_rate": 1.5209591448650535e-06,
+      "loss": 0.3911,
+      "step": 3139
+    },
+    {
+      "epoch": 2.312223858615611,
+      "grad_norm": 0.32253560423851013,
+      "learning_rate": 1.5178829830057883e-06,
+      "loss": 0.3547,
+      "step": 3140
+    },
+    {
+      "epoch": 2.312960235640648,
+      "grad_norm": 0.37971949577331543,
+      "learning_rate": 1.5148093783080337e-06,
+      "loss": 0.3687,
+      "step": 3141
+    },
+    {
+      "epoch": 2.3136966126656846,
+      "grad_norm": 0.34527388215065,
+      "learning_rate": 1.5117383330289542e-06,
+      "loss": 0.4067,
+      "step": 3142
+    },
+    {
+      "epoch": 2.3144329896907214,
+      "grad_norm": 0.3223811686038971,
+      "learning_rate": 1.5086698494238316e-06,
+      "loss": 0.3777,
+      "step": 3143
+    },
+    {
+      "epoch": 2.3151693667157582,
+      "grad_norm": 0.3294554054737091,
+      "learning_rate": 1.5056039297460656e-06,
+      "loss": 0.3806,
+      "step": 3144
+    },
+    {
+      "epoch": 2.315905743740795,
+      "grad_norm": 0.4138847291469574,
+      "learning_rate": 1.5025405762471795e-06,
+      "loss": 0.366,
+      "step": 3145
+    },
+    {
+      "epoch": 2.316642120765832,
+      "grad_norm": 0.3485645353794098,
+      "learning_rate": 1.4994797911768034e-06,
+      "loss": 0.4099,
+      "step": 3146
+    },
+    {
+      "epoch": 2.3173784977908687,
+      "grad_norm": 0.3449089527130127,
+      "learning_rate": 1.4964215767826846e-06,
+      "loss": 0.3593,
+      "step": 3147
+    },
+    {
+      "epoch": 2.3181148748159055,
+      "grad_norm": 0.3338415324687958,
+      "learning_rate": 1.4933659353106872e-06,
+      "loss": 0.363,
+      "step": 3148
+    },
+    {
+      "epoch": 2.3188512518409423,
+      "grad_norm": 0.33928152918815613,
+      "learning_rate": 1.4903128690047802e-06,
+      "loss": 0.3571,
+      "step": 3149
+    },
+    {
+      "epoch": 2.319587628865979,
+      "grad_norm": 0.3429686725139618,
+      "learning_rate": 1.4872623801070413e-06,
+      "loss": 0.3747,
+      "step": 3150
+    },
+    {
+      "epoch": 2.3203240058910164,
+      "grad_norm": 0.32575222849845886,
+      "learning_rate": 1.4842144708576606e-06,
+      "loss": 0.3772,
+      "step": 3151
+    },
+    {
+      "epoch": 2.321060382916053,
+      "grad_norm": 0.34492257237434387,
+      "learning_rate": 1.4811691434949293e-06,
+      "loss": 0.3755,
+      "step": 3152
+    },
+    {
+      "epoch": 2.32179675994109,
+      "grad_norm": 0.3562105596065521,
+      "learning_rate": 1.4781264002552425e-06,
+      "loss": 0.3717,
+      "step": 3153
+    },
+    {
+      "epoch": 2.322533136966127,
+      "grad_norm": 0.3609287440776825,
+      "learning_rate": 1.4750862433731028e-06,
+      "loss": 0.357,
+      "step": 3154
+    },
+    {
+      "epoch": 2.3232695139911637,
+      "grad_norm": 0.33014407753944397,
+      "learning_rate": 1.4720486750811035e-06,
+      "loss": 0.3536,
+      "step": 3155
+    },
+    {
+      "epoch": 2.3240058910162005,
+      "grad_norm": 0.3423437476158142,
+      "learning_rate": 1.4690136976099479e-06,
+      "loss": 0.3642,
+      "step": 3156
+    },
+    {
+      "epoch": 2.3247422680412373,
+      "grad_norm": 0.35252103209495544,
+      "learning_rate": 1.4659813131884304e-06,
+      "loss": 0.3835,
+      "step": 3157
+    },
+    {
+      "epoch": 2.325478645066274,
+      "grad_norm": 0.3483889400959015,
+      "learning_rate": 1.46295152404344e-06,
+      "loss": 0.3714,
+      "step": 3158
+    },
+    {
+      "epoch": 2.326215022091311,
+      "grad_norm": 0.3345074951648712,
+      "learning_rate": 1.4599243323999668e-06,
+      "loss": 0.365,
+      "step": 3159
+    },
+    {
+      "epoch": 2.3269513991163477,
+      "grad_norm": 0.334255188703537,
+      "learning_rate": 1.4568997404810858e-06,
+      "loss": 0.3851,
+      "step": 3160
+    },
+    {
+      "epoch": 2.3276877761413846,
+      "grad_norm": 0.33251067996025085,
+      "learning_rate": 1.4538777505079654e-06,
+      "loss": 0.3928,
+      "step": 3161
+    },
+    {
+      "epoch": 2.3284241531664214,
+      "grad_norm": 0.3211064040660858,
+      "learning_rate": 1.4508583646998674e-06,
+      "loss": 0.4159,
+      "step": 3162
+    },
+    {
+      "epoch": 2.329160530191458,
+      "grad_norm": 0.33041635155677795,
+      "learning_rate": 1.4478415852741328e-06,
+      "loss": 0.349,
+      "step": 3163
+    },
+    {
+      "epoch": 2.329896907216495,
+      "grad_norm": 0.31442853808403015,
+      "learning_rate": 1.4448274144461965e-06,
+      "loss": 0.3824,
+      "step": 3164
+    },
+    {
+      "epoch": 2.330633284241532,
+      "grad_norm": 0.3113677501678467,
+      "learning_rate": 1.4418158544295734e-06,
+      "loss": 0.3695,
+      "step": 3165
+    },
+    {
+      "epoch": 2.3313696612665686,
+      "grad_norm": 0.33691975474357605,
+      "learning_rate": 1.4388069074358612e-06,
+      "loss": 0.3742,
+      "step": 3166
+    },
+    {
+      "epoch": 2.3321060382916055,
+      "grad_norm": 0.33256834745407104,
+      "learning_rate": 1.4358005756747417e-06,
+      "loss": 0.341,
+      "step": 3167
+    },
+    {
+      "epoch": 2.3328424153166423,
+      "grad_norm": 0.33882421255111694,
+      "learning_rate": 1.4327968613539734e-06,
+      "loss": 0.3785,
+      "step": 3168
+    },
+    {
+      "epoch": 2.333578792341679,
+      "grad_norm": 0.33136385679244995,
+      "learning_rate": 1.429795766679391e-06,
+      "loss": 0.3777,
+      "step": 3169
+    },
+    {
+      "epoch": 2.334315169366716,
+      "grad_norm": 0.31647056341171265,
+      "learning_rate": 1.426797293854912e-06,
+      "loss": 0.3867,
+      "step": 3170
+    },
+    {
+      "epoch": 2.3350515463917527,
+      "grad_norm": 0.3487012982368469,
+      "learning_rate": 1.4238014450825227e-06,
+      "loss": 0.3682,
+      "step": 3171
+    },
+    {
+      "epoch": 2.3357879234167895,
+      "grad_norm": 0.3635872006416321,
+      "learning_rate": 1.4208082225622804e-06,
+      "loss": 0.385,
+      "step": 3172
+    },
+    {
+      "epoch": 2.3365243004418264,
+      "grad_norm": 0.35022327303886414,
+      "learning_rate": 1.4178176284923212e-06,
+      "loss": 0.4066,
+      "step": 3173
+    },
+    {
+      "epoch": 2.337260677466863,
+      "grad_norm": 0.35386213660240173,
+      "learning_rate": 1.4148296650688465e-06,
+      "loss": 0.3982,
+      "step": 3174
+    },
+    {
+      "epoch": 2.3379970544919,
+      "grad_norm": 0.3819175660610199,
+      "learning_rate": 1.4118443344861237e-06,
+      "loss": 0.3898,
+      "step": 3175
+    },
+    {
+      "epoch": 2.338733431516937,
+      "grad_norm": 0.36631712317466736,
+      "learning_rate": 1.408861638936493e-06,
+      "loss": 0.3976,
+      "step": 3176
+    },
+    {
+      "epoch": 2.3394698085419736,
+      "grad_norm": 0.3479391634464264,
+      "learning_rate": 1.4058815806103542e-06,
+      "loss": 0.3825,
+      "step": 3177
+    },
+    {
+      "epoch": 2.3402061855670104,
+      "grad_norm": 0.35590627789497375,
+      "learning_rate": 1.4029041616961703e-06,
+      "loss": 0.3747,
+      "step": 3178
+    },
+    {
+      "epoch": 2.3409425625920472,
+      "grad_norm": 0.3458181321620941,
+      "learning_rate": 1.3999293843804728e-06,
+      "loss": 0.3736,
+      "step": 3179
+    },
+    {
+      "epoch": 2.341678939617084,
+      "grad_norm": 0.33873867988586426,
+      "learning_rate": 1.3969572508478424e-06,
+      "loss": 0.373,
+      "step": 3180
+    },
+    {
+      "epoch": 2.342415316642121,
+      "grad_norm": 0.33975568413734436,
+      "learning_rate": 1.3939877632809279e-06,
+      "loss": 0.3431,
+      "step": 3181
+    },
+    {
+      "epoch": 2.3431516936671577,
+      "grad_norm": 0.3956592082977295,
+      "learning_rate": 1.3910209238604306e-06,
+      "loss": 0.3857,
+      "step": 3182
+    },
+    {
+      "epoch": 2.3438880706921945,
+      "grad_norm": 0.35657671093940735,
+      "learning_rate": 1.3880567347651052e-06,
+      "loss": 0.3717,
+      "step": 3183
+    },
+    {
+      "epoch": 2.3446244477172313,
+      "grad_norm": 0.34503045678138733,
+      "learning_rate": 1.3850951981717665e-06,
+      "loss": 0.3607,
+      "step": 3184
+    },
+    {
+      "epoch": 2.345360824742268,
+      "grad_norm": 0.31892770528793335,
+      "learning_rate": 1.3821363162552753e-06,
+      "loss": 0.3666,
+      "step": 3185
+    },
+    {
+      "epoch": 2.346097201767305,
+      "grad_norm": 0.3312123715877533,
+      "learning_rate": 1.3791800911885444e-06,
+      "loss": 0.3279,
+      "step": 3186
+    },
+    {
+      "epoch": 2.346833578792342,
+      "grad_norm": 0.37443795800209045,
+      "learning_rate": 1.3762265251425394e-06,
+      "loss": 0.3779,
+      "step": 3187
+    },
+    {
+      "epoch": 2.3475699558173786,
+      "grad_norm": 0.3352743685245514,
+      "learning_rate": 1.373275620286265e-06,
+      "loss": 0.3839,
+      "step": 3188
+    },
+    {
+      "epoch": 2.3483063328424154,
+      "grad_norm": 0.3558715879917145,
+      "learning_rate": 1.370327378786781e-06,
+      "loss": 0.3788,
+      "step": 3189
+    },
+    {
+      "epoch": 2.3490427098674522,
+      "grad_norm": 0.38027918338775635,
+      "learning_rate": 1.367381802809185e-06,
+      "loss": 0.4096,
+      "step": 3190
+    },
+    {
+      "epoch": 2.349779086892489,
+      "grad_norm": 0.31835508346557617,
+      "learning_rate": 1.3644388945166175e-06,
+      "loss": 0.4012,
+      "step": 3191
+    },
+    {
+      "epoch": 2.350515463917526,
+      "grad_norm": 0.31716689467430115,
+      "learning_rate": 1.3614986560702648e-06,
+      "loss": 0.3676,
+      "step": 3192
+    },
+    {
+      "epoch": 2.3512518409425627,
+      "grad_norm": 0.3417465090751648,
+      "learning_rate": 1.3585610896293472e-06,
+      "loss": 0.369,
+      "step": 3193
+    },
+    {
+      "epoch": 2.3519882179675995,
+      "grad_norm": 0.3349650800228119,
+      "learning_rate": 1.3556261973511236e-06,
+      "loss": 0.3702,
+      "step": 3194
+    },
+    {
+      "epoch": 2.3527245949926363,
+      "grad_norm": 0.38491392135620117,
+      "learning_rate": 1.3526939813908929e-06,
+      "loss": 0.3594,
+      "step": 3195
+    },
+    {
+      "epoch": 2.353460972017673,
+      "grad_norm": 0.3387065529823303,
+      "learning_rate": 1.349764443901984e-06,
+      "loss": 0.3557,
+      "step": 3196
+    },
+    {
+      "epoch": 2.35419734904271,
+      "grad_norm": 0.33061766624450684,
+      "learning_rate": 1.346837587035762e-06,
+      "loss": 0.3777,
+      "step": 3197
+    },
+    {
+      "epoch": 2.3549337260677468,
+      "grad_norm": 0.34983956813812256,
+      "learning_rate": 1.343913412941621e-06,
+      "loss": 0.392,
+      "step": 3198
+    },
+    {
+      "epoch": 2.3556701030927836,
+      "grad_norm": 0.3425159454345703,
+      "learning_rate": 1.3409919237669843e-06,
+      "loss": 0.3779,
+      "step": 3199
+    },
+    {
+      "epoch": 2.3564064801178204,
+      "grad_norm": 0.3214556574821472,
+      "learning_rate": 1.33807312165731e-06,
+      "loss": 0.3965,
+      "step": 3200
+    },
+    {
+      "epoch": 2.357142857142857,
+      "grad_norm": 0.32517537474632263,
+      "learning_rate": 1.335157008756075e-06,
+      "loss": 0.3657,
+      "step": 3201
+    },
+    {
+      "epoch": 2.357879234167894,
+      "grad_norm": 0.3494378626346588,
+      "learning_rate": 1.3322435872047835e-06,
+      "loss": 0.3905,
+      "step": 3202
+    },
+    {
+      "epoch": 2.358615611192931,
+      "grad_norm": 0.39801037311553955,
+      "learning_rate": 1.329332859142967e-06,
+      "loss": 0.3801,
+      "step": 3203
+    },
+    {
+      "epoch": 2.3593519882179677,
+      "grad_norm": 0.34806138277053833,
+      "learning_rate": 1.326424826708177e-06,
+      "loss": 0.3711,
+      "step": 3204
+    },
+    {
+      "epoch": 2.3600883652430045,
+      "grad_norm": 0.37914684414863586,
+      "learning_rate": 1.3235194920359795e-06,
+      "loss": 0.3568,
+      "step": 3205
+    },
+    {
+      "epoch": 2.3608247422680413,
+      "grad_norm": 0.3150370419025421,
+      "learning_rate": 1.3206168572599692e-06,
+      "loss": 0.3577,
+      "step": 3206
+    },
+    {
+      "epoch": 2.361561119293078,
+      "grad_norm": 0.3247841000556946,
+      "learning_rate": 1.3177169245117522e-06,
+      "loss": 0.3998,
+      "step": 3207
+    },
+    {
+      "epoch": 2.362297496318115,
+      "grad_norm": 0.3235202729701996,
+      "learning_rate": 1.3148196959209491e-06,
+      "loss": 0.3604,
+      "step": 3208
+    },
+    {
+      "epoch": 2.3630338733431517,
+      "grad_norm": 0.3443073034286499,
+      "learning_rate": 1.3119251736152005e-06,
+      "loss": 0.3561,
+      "step": 3209
+    },
+    {
+      "epoch": 2.3637702503681886,
+      "grad_norm": 0.3384738266468048,
+      "learning_rate": 1.309033359720155e-06,
+      "loss": 0.3632,
+      "step": 3210
+    },
+    {
+      "epoch": 2.3645066273932254,
+      "grad_norm": 0.3304847478866577,
+      "learning_rate": 1.3061442563594718e-06,
+      "loss": 0.3745,
+      "step": 3211
+    },
+    {
+      "epoch": 2.365243004418262,
+      "grad_norm": 0.3340161144733429,
+      "learning_rate": 1.3032578656548228e-06,
+      "loss": 0.3665,
+      "step": 3212
+    },
+    {
+      "epoch": 2.365979381443299,
+      "grad_norm": 0.3465518355369568,
+      "learning_rate": 1.3003741897258864e-06,
+      "loss": 0.3678,
+      "step": 3213
+    },
+    {
+      "epoch": 2.366715758468336,
+      "grad_norm": 0.31364724040031433,
+      "learning_rate": 1.297493230690346e-06,
+      "loss": 0.3602,
+      "step": 3214
+    },
+    {
+      "epoch": 2.3674521354933726,
+      "grad_norm": 0.3540259599685669,
+      "learning_rate": 1.2946149906638905e-06,
+      "loss": 0.3733,
+      "step": 3215
+    },
+    {
+      "epoch": 2.3681885125184094,
+      "grad_norm": 0.34416845440864563,
+      "learning_rate": 1.2917394717602123e-06,
+      "loss": 0.3512,
+      "step": 3216
+    },
+    {
+      "epoch": 2.3689248895434463,
+      "grad_norm": 0.37289807200431824,
+      "learning_rate": 1.2888666760910074e-06,
+      "loss": 0.3755,
+      "step": 3217
+    },
+    {
+      "epoch": 2.369661266568483,
+      "grad_norm": 0.3461972773075104,
+      "learning_rate": 1.285996605765969e-06,
+      "loss": 0.4092,
+      "step": 3218
+    },
+    {
+      "epoch": 2.37039764359352,
+      "grad_norm": 0.36387526988983154,
+      "learning_rate": 1.283129262892789e-06,
+      "loss": 0.3788,
+      "step": 3219
+    },
+    {
+      "epoch": 2.3711340206185567,
+      "grad_norm": 0.32595691084861755,
+      "learning_rate": 1.2802646495771592e-06,
+      "loss": 0.3588,
+      "step": 3220
+    },
+    {
+      "epoch": 2.3718703976435935,
+      "grad_norm": 0.332273006439209,
+      "learning_rate": 1.2774027679227647e-06,
+      "loss": 0.3772,
+      "step": 3221
+    },
+    {
+      "epoch": 2.3726067746686303,
+      "grad_norm": 0.3550896942615509,
+      "learning_rate": 1.2745436200312844e-06,
+      "loss": 0.375,
+      "step": 3222
+    },
+    {
+      "epoch": 2.373343151693667,
+      "grad_norm": 0.3492611348628998,
+      "learning_rate": 1.2716872080023901e-06,
+      "loss": 0.3792,
+      "step": 3223
+    },
+    {
+      "epoch": 2.374079528718704,
+      "grad_norm": 0.33285483717918396,
+      "learning_rate": 1.2688335339337433e-06,
+      "loss": 0.3837,
+      "step": 3224
+    },
+    {
+      "epoch": 2.374815905743741,
+      "grad_norm": 0.3760431110858917,
+      "learning_rate": 1.2659825999209985e-06,
+      "loss": 0.3811,
+      "step": 3225
+    },
+    {
+      "epoch": 2.3755522827687776,
+      "grad_norm": 0.33756300806999207,
+      "learning_rate": 1.263134408057794e-06,
+      "loss": 0.3708,
+      "step": 3226
+    },
+    {
+      "epoch": 2.3762886597938144,
+      "grad_norm": 0.36119866371154785,
+      "learning_rate": 1.2602889604357548e-06,
+      "loss": 0.3681,
+      "step": 3227
+    },
+    {
+      "epoch": 2.3770250368188512,
+      "grad_norm": 0.33655065298080444,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.385,
+      "step": 3228
+    },
+    {
+      "epoch": 2.377761413843888,
+      "grad_norm": 0.3267086446285248,
+      "learning_rate": 1.2546063062716069e-06,
+      "loss": 0.3656,
+      "step": 3229
+    },
+    {
+      "epoch": 2.378497790868925,
+      "grad_norm": 0.3475969731807709,
+      "learning_rate": 1.2517691039026625e-06,
+      "loss": 0.3735,
+      "step": 3230
+    },
+    {
+      "epoch": 2.3792341678939617,
+      "grad_norm": 0.3716298043727875,
+      "learning_rate": 1.2489346541212226e-06,
+      "loss": 0.4026,
+      "step": 3231
+    },
+    {
+      "epoch": 2.3799705449189985,
+      "grad_norm": 0.39196276664733887,
+      "learning_rate": 1.2461029590088198e-06,
+      "loss": 0.3645,
+      "step": 3232
+    },
+    {
+      "epoch": 2.3807069219440353,
+      "grad_norm": 0.34049636125564575,
+      "learning_rate": 1.2432740206449629e-06,
+      "loss": 0.3597,
+      "step": 3233
+    },
+    {
+      "epoch": 2.381443298969072,
+      "grad_norm": 0.3242121636867523,
+      "learning_rate": 1.240447841107143e-06,
+      "loss": 0.397,
+      "step": 3234
+    },
+    {
+      "epoch": 2.382179675994109,
+      "grad_norm": 0.3249550759792328,
+      "learning_rate": 1.2376244224708183e-06,
+      "loss": 0.3732,
+      "step": 3235
+    },
+    {
+      "epoch": 2.3829160530191458,
+      "grad_norm": 0.3562273681163788,
+      "learning_rate": 1.2348037668094214e-06,
+      "loss": 0.3755,
+      "step": 3236
+    },
+    {
+      "epoch": 2.3836524300441826,
+      "grad_norm": 0.3847813904285431,
+      "learning_rate": 1.2319858761943598e-06,
+      "loss": 0.3614,
+      "step": 3237
+    },
+    {
+      "epoch": 2.3843888070692194,
+      "grad_norm": 0.32016924023628235,
+      "learning_rate": 1.2291707526950047e-06,
+      "loss": 0.3499,
+      "step": 3238
+    },
+    {
+      "epoch": 2.3851251840942562,
+      "grad_norm": 0.38904672861099243,
+      "learning_rate": 1.2263583983786986e-06,
+      "loss": 0.3994,
+      "step": 3239
+    },
+    {
+      "epoch": 2.385861561119293,
+      "grad_norm": 0.34415403008461,
+      "learning_rate": 1.2235488153107488e-06,
+      "loss": 0.3723,
+      "step": 3240
+    },
+    {
+      "epoch": 2.38659793814433,
+      "grad_norm": 0.3804604709148407,
+      "learning_rate": 1.2207420055544278e-06,
+      "loss": 0.3707,
+      "step": 3241
+    },
+    {
+      "epoch": 2.3873343151693667,
+      "grad_norm": 0.34501025080680847,
+      "learning_rate": 1.2179379711709738e-06,
+      "loss": 0.4066,
+      "step": 3242
+    },
+    {
+      "epoch": 2.3880706921944035,
+      "grad_norm": 0.35405653715133667,
+      "learning_rate": 1.2151367142195842e-06,
+      "loss": 0.376,
+      "step": 3243
+    },
+    {
+      "epoch": 2.3888070692194403,
+      "grad_norm": 0.3774116039276123,
+      "learning_rate": 1.212338236757415e-06,
+      "loss": 0.3763,
+      "step": 3244
+    },
+    {
+      "epoch": 2.389543446244477,
+      "grad_norm": 0.35209256410598755,
+      "learning_rate": 1.2095425408395873e-06,
+      "loss": 0.3543,
+      "step": 3245
+    },
+    {
+      "epoch": 2.390279823269514,
+      "grad_norm": 0.36879295110702515,
+      "learning_rate": 1.2067496285191743e-06,
+      "loss": 0.3765,
+      "step": 3246
+    },
+    {
+      "epoch": 2.3910162002945508,
+      "grad_norm": 0.3600074052810669,
+      "learning_rate": 1.2039595018472055e-06,
+      "loss": 0.377,
+      "step": 3247
+    },
+    {
+      "epoch": 2.3917525773195876,
+      "grad_norm": 0.3357848525047302,
+      "learning_rate": 1.2011721628726663e-06,
+      "loss": 0.3659,
+      "step": 3248
+    },
+    {
+      "epoch": 2.3924889543446244,
+      "grad_norm": 0.3500244617462158,
+      "learning_rate": 1.1983876136424926e-06,
+      "loss": 0.3847,
+      "step": 3249
+    },
+    {
+      "epoch": 2.393225331369661,
+      "grad_norm": 0.3420659303665161,
+      "learning_rate": 1.1956058562015766e-06,
+      "loss": 0.392,
+      "step": 3250
+    },
+    {
+      "epoch": 2.393961708394698,
+      "grad_norm": 0.33460474014282227,
+      "learning_rate": 1.192826892592755e-06,
+      "loss": 0.3708,
+      "step": 3251
+    },
+    {
+      "epoch": 2.394698085419735,
+      "grad_norm": 0.34583914279937744,
+      "learning_rate": 1.1900507248568128e-06,
+      "loss": 0.349,
+      "step": 3252
+    },
+    {
+      "epoch": 2.3954344624447717,
+      "grad_norm": 0.36606159806251526,
+      "learning_rate": 1.1872773550324873e-06,
+      "loss": 0.3739,
+      "step": 3253
+    },
+    {
+      "epoch": 2.3961708394698085,
+      "grad_norm": 0.34434834122657776,
+      "learning_rate": 1.1845067851564557e-06,
+      "loss": 0.3459,
+      "step": 3254
+    },
+    {
+      "epoch": 2.3969072164948453,
+      "grad_norm": 0.3151196539402008,
+      "learning_rate": 1.1817390172633402e-06,
+      "loss": 0.3635,
+      "step": 3255
+    },
+    {
+      "epoch": 2.397643593519882,
+      "grad_norm": 0.32296085357666016,
+      "learning_rate": 1.1789740533857075e-06,
+      "loss": 0.3863,
+      "step": 3256
+    },
+    {
+      "epoch": 2.398379970544919,
+      "grad_norm": 0.3849099576473236,
+      "learning_rate": 1.1762118955540609e-06,
+      "loss": 0.3943,
+      "step": 3257
+    },
+    {
+      "epoch": 2.3991163475699557,
+      "grad_norm": 0.3584287762641907,
+      "learning_rate": 1.1734525457968488e-06,
+      "loss": 0.3823,
+      "step": 3258
+    },
+    {
+      "epoch": 2.3998527245949925,
+      "grad_norm": 0.33477911353111267,
+      "learning_rate": 1.1706960061404527e-06,
+      "loss": 0.37,
+      "step": 3259
+    },
+    {
+      "epoch": 2.4005891016200294,
+      "grad_norm": 0.3298894762992859,
+      "learning_rate": 1.1679422786091909e-06,
+      "loss": 0.3887,
+      "step": 3260
+    },
+    {
+      "epoch": 2.401325478645066,
+      "grad_norm": 0.3459241986274719,
+      "learning_rate": 1.1651913652253199e-06,
+      "loss": 0.3546,
+      "step": 3261
+    },
+    {
+      "epoch": 2.402061855670103,
+      "grad_norm": 0.3287370502948761,
+      "learning_rate": 1.162443268009027e-06,
+      "loss": 0.3604,
+      "step": 3262
+    },
+    {
+      "epoch": 2.40279823269514,
+      "grad_norm": 0.33065786957740784,
+      "learning_rate": 1.1596979889784304e-06,
+      "loss": 0.384,
+      "step": 3263
+    },
+    {
+      "epoch": 2.4035346097201766,
+      "grad_norm": 0.3277767598628998,
+      "learning_rate": 1.1569555301495817e-06,
+      "loss": 0.3492,
+      "step": 3264
+    },
+    {
+      "epoch": 2.4042709867452134,
+      "grad_norm": 0.3225124478340149,
+      "learning_rate": 1.1542158935364584e-06,
+      "loss": 0.3909,
+      "step": 3265
+    },
+    {
+      "epoch": 2.4050073637702503,
+      "grad_norm": 0.30637234449386597,
+      "learning_rate": 1.1514790811509658e-06,
+      "loss": 0.4044,
+      "step": 3266
+    },
+    {
+      "epoch": 2.405743740795287,
+      "grad_norm": 0.34995532035827637,
+      "learning_rate": 1.148745095002939e-06,
+      "loss": 0.3824,
+      "step": 3267
+    },
+    {
+      "epoch": 2.406480117820324,
+      "grad_norm": 0.33643367886543274,
+      "learning_rate": 1.1460139371001339e-06,
+      "loss": 0.3694,
+      "step": 3268
+    },
+    {
+      "epoch": 2.4072164948453607,
+      "grad_norm": 0.3490047752857208,
+      "learning_rate": 1.1432856094482282e-06,
+      "loss": 0.384,
+      "step": 3269
+    },
+    {
+      "epoch": 2.4079528718703975,
+      "grad_norm": 0.3088352382183075,
+      "learning_rate": 1.1405601140508265e-06,
+      "loss": 0.3662,
+      "step": 3270
+    },
+    {
+      "epoch": 2.4086892488954343,
+      "grad_norm": 0.3435070514678955,
+      "learning_rate": 1.1378374529094494e-06,
+      "loss": 0.3965,
+      "step": 3271
+    },
+    {
+      "epoch": 2.409425625920471,
+      "grad_norm": 0.33832401037216187,
+      "learning_rate": 1.135117628023536e-06,
+      "loss": 0.3821,
+      "step": 3272
+    },
+    {
+      "epoch": 2.410162002945508,
+      "grad_norm": 0.335674524307251,
+      "learning_rate": 1.1324006413904437e-06,
+      "loss": 0.4041,
+      "step": 3273
+    },
+    {
+      "epoch": 2.410898379970545,
+      "grad_norm": 0.37129271030426025,
+      "learning_rate": 1.1296864950054443e-06,
+      "loss": 0.364,
+      "step": 3274
+    },
+    {
+      "epoch": 2.4116347569955816,
+      "grad_norm": 0.35193946957588196,
+      "learning_rate": 1.1269751908617277e-06,
+      "loss": 0.404,
+      "step": 3275
+    },
+    {
+      "epoch": 2.4123711340206184,
+      "grad_norm": 0.34441494941711426,
+      "learning_rate": 1.124266730950392e-06,
+      "loss": 0.3571,
+      "step": 3276
+    },
+    {
+      "epoch": 2.4131075110456552,
+      "grad_norm": 0.3434298634529114,
+      "learning_rate": 1.1215611172604468e-06,
+      "loss": 0.3831,
+      "step": 3277
+    },
+    {
+      "epoch": 2.413843888070692,
+      "grad_norm": 0.3040885925292969,
+      "learning_rate": 1.1188583517788165e-06,
+      "loss": 0.3555,
+      "step": 3278
+    },
+    {
+      "epoch": 2.414580265095729,
+      "grad_norm": 0.3662799596786499,
+      "learning_rate": 1.1161584364903287e-06,
+      "loss": 0.3932,
+      "step": 3279
+    },
+    {
+      "epoch": 2.4153166421207657,
+      "grad_norm": 0.3392142653465271,
+      "learning_rate": 1.1134613733777195e-06,
+      "loss": 0.3553,
+      "step": 3280
+    },
+    {
+      "epoch": 2.4160530191458025,
+      "grad_norm": 0.35965248942375183,
+      "learning_rate": 1.1107671644216305e-06,
+      "loss": 0.3529,
+      "step": 3281
+    },
+    {
+      "epoch": 2.4167893961708393,
+      "grad_norm": 0.37585723400115967,
+      "learning_rate": 1.1080758116006057e-06,
+      "loss": 0.3913,
+      "step": 3282
+    },
+    {
+      "epoch": 2.417525773195876,
+      "grad_norm": 0.3380414545536041,
+      "learning_rate": 1.1053873168910966e-06,
+      "loss": 0.3933,
+      "step": 3283
+    },
+    {
+      "epoch": 2.418262150220913,
+      "grad_norm": 0.33005619049072266,
+      "learning_rate": 1.1027016822674509e-06,
+      "loss": 0.3837,
+      "step": 3284
+    },
+    {
+      "epoch": 2.4189985272459498,
+      "grad_norm": 0.3498542904853821,
+      "learning_rate": 1.1000189097019164e-06,
+      "loss": 0.3881,
+      "step": 3285
+    },
+    {
+      "epoch": 2.4197349042709866,
+      "grad_norm": 0.37784454226493835,
+      "learning_rate": 1.0973390011646422e-06,
+      "loss": 0.3898,
+      "step": 3286
+    },
+    {
+      "epoch": 2.4204712812960234,
+      "grad_norm": 0.3332594335079193,
+      "learning_rate": 1.0946619586236711e-06,
+      "loss": 0.3571,
+      "step": 3287
+    },
+    {
+      "epoch": 2.42120765832106,
+      "grad_norm": 0.32110390067100525,
+      "learning_rate": 1.0919877840449428e-06,
+      "loss": 0.3904,
+      "step": 3288
+    },
+    {
+      "epoch": 2.421944035346097,
+      "grad_norm": 0.3465670645236969,
+      "learning_rate": 1.0893164793922894e-06,
+      "loss": 0.3595,
+      "step": 3289
+    },
+    {
+      "epoch": 2.422680412371134,
+      "grad_norm": 0.33734622597694397,
+      "learning_rate": 1.0866480466274377e-06,
+      "loss": 0.3817,
+      "step": 3290
+    },
+    {
+      "epoch": 2.4234167893961707,
+      "grad_norm": 0.3218252956867218,
+      "learning_rate": 1.0839824877100008e-06,
+      "loss": 0.3472,
+      "step": 3291
+    },
+    {
+      "epoch": 2.4241531664212075,
+      "grad_norm": 0.3475230634212494,
+      "learning_rate": 1.0813198045974888e-06,
+      "loss": 0.3629,
+      "step": 3292
+    },
+    {
+      "epoch": 2.4248895434462443,
+      "grad_norm": 0.32285812497138977,
+      "learning_rate": 1.0786599992452933e-06,
+      "loss": 0.3819,
+      "step": 3293
+    },
+    {
+      "epoch": 2.425625920471281,
+      "grad_norm": 0.34469369053840637,
+      "learning_rate": 1.0760030736066952e-06,
+      "loss": 0.3668,
+      "step": 3294
+    },
+    {
+      "epoch": 2.426362297496318,
+      "grad_norm": 0.3158765435218811,
+      "learning_rate": 1.0733490296328613e-06,
+      "loss": 0.3542,
+      "step": 3295
+    },
+    {
+      "epoch": 2.4270986745213547,
+      "grad_norm": 0.3300797641277313,
+      "learning_rate": 1.0706978692728416e-06,
+      "loss": 0.4024,
+      "step": 3296
+    },
+    {
+      "epoch": 2.4278350515463916,
+      "grad_norm": 0.31223219633102417,
+      "learning_rate": 1.0680495944735665e-06,
+      "loss": 0.3649,
+      "step": 3297
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.35305702686309814,
+      "learning_rate": 1.0654042071798498e-06,
+      "loss": 0.3598,
+      "step": 3298
+    },
+    {
+      "epoch": 2.429307805596465,
+      "grad_norm": 0.354264497756958,
+      "learning_rate": 1.0627617093343833e-06,
+      "loss": 0.3895,
+      "step": 3299
+    },
+    {
+      "epoch": 2.4300441826215025,
+      "grad_norm": 0.34128403663635254,
+      "learning_rate": 1.060122102877739e-06,
+      "loss": 0.3702,
+      "step": 3300
+    },
+    {
+      "epoch": 2.4307805596465393,
+      "grad_norm": 0.36130642890930176,
+      "learning_rate": 1.0574853897483634e-06,
+      "loss": 0.4276,
+      "step": 3301
+    },
+    {
+      "epoch": 2.431516936671576,
+      "grad_norm": 0.3270364999771118,
+      "learning_rate": 1.054851571882578e-06,
+      "loss": 0.3915,
+      "step": 3302
+    },
+    {
+      "epoch": 2.432253313696613,
+      "grad_norm": 0.33622244000434875,
+      "learning_rate": 1.052220651214581e-06,
+      "loss": 0.4009,
+      "step": 3303
+    },
+    {
+      "epoch": 2.4329896907216497,
+      "grad_norm": 0.3255276381969452,
+      "learning_rate": 1.0495926296764398e-06,
+      "loss": 0.3953,
+      "step": 3304
+    },
+    {
+      "epoch": 2.4337260677466865,
+      "grad_norm": 0.32380256056785583,
+      "learning_rate": 1.0469675091980946e-06,
+      "loss": 0.3766,
+      "step": 3305
+    },
+    {
+      "epoch": 2.4344624447717234,
+      "grad_norm": 0.32424572110176086,
+      "learning_rate": 1.0443452917073538e-06,
+      "loss": 0.3914,
+      "step": 3306
+    },
+    {
+      "epoch": 2.43519882179676,
+      "grad_norm": 0.3544393479824066,
+      "learning_rate": 1.041725979129894e-06,
+      "loss": 0.3624,
+      "step": 3307
+    },
+    {
+      "epoch": 2.435935198821797,
+      "grad_norm": 0.3281401991844177,
+      "learning_rate": 1.0391095733892614e-06,
+      "loss": 0.3437,
+      "step": 3308
+    },
+    {
+      "epoch": 2.436671575846834,
+      "grad_norm": 0.32801955938339233,
+      "learning_rate": 1.0364960764068643e-06,
+      "loss": 0.3675,
+      "step": 3309
+    },
+    {
+      "epoch": 2.4374079528718706,
+      "grad_norm": 0.3724784553050995,
+      "learning_rate": 1.033885490101974e-06,
+      "loss": 0.3646,
+      "step": 3310
+    },
+    {
+      "epoch": 2.4381443298969074,
+      "grad_norm": 0.35734328627586365,
+      "learning_rate": 1.0312778163917298e-06,
+      "loss": 0.3746,
+      "step": 3311
+    },
+    {
+      "epoch": 2.4388807069219443,
+      "grad_norm": 0.33092063665390015,
+      "learning_rate": 1.0286730571911264e-06,
+      "loss": 0.3573,
+      "step": 3312
+    },
+    {
+      "epoch": 2.439617083946981,
+      "grad_norm": 0.32385748624801636,
+      "learning_rate": 1.0260712144130192e-06,
+      "loss": 0.3724,
+      "step": 3313
+    },
+    {
+      "epoch": 2.440353460972018,
+      "grad_norm": 0.3588411808013916,
+      "learning_rate": 1.0234722899681265e-06,
+      "loss": 0.3727,
+      "step": 3314
+    },
+    {
+      "epoch": 2.4410898379970547,
+      "grad_norm": 0.32793402671813965,
+      "learning_rate": 1.020876285765015e-06,
+      "loss": 0.3588,
+      "step": 3315
+    },
+    {
+      "epoch": 2.4418262150220915,
+      "grad_norm": 0.37099236249923706,
+      "learning_rate": 1.018283203710116e-06,
+      "loss": 0.3759,
+      "step": 3316
+    },
+    {
+      "epoch": 2.4425625920471283,
+      "grad_norm": 0.35512179136276245,
+      "learning_rate": 1.0156930457077085e-06,
+      "loss": 0.3909,
+      "step": 3317
+    },
+    {
+      "epoch": 2.443298969072165,
+      "grad_norm": 0.32418885827064514,
+      "learning_rate": 1.0131058136599254e-06,
+      "loss": 0.3577,
+      "step": 3318
+    },
+    {
+      "epoch": 2.444035346097202,
+      "grad_norm": 0.33136194944381714,
+      "learning_rate": 1.0105215094667542e-06,
+      "loss": 0.3603,
+      "step": 3319
+    },
+    {
+      "epoch": 2.444771723122239,
+      "grad_norm": 0.35086795687675476,
+      "learning_rate": 1.0079401350260288e-06,
+      "loss": 0.3666,
+      "step": 3320
+    },
+    {
+      "epoch": 2.4455081001472756,
+      "grad_norm": 0.3403591215610504,
+      "learning_rate": 1.0053616922334307e-06,
+      "loss": 0.3901,
+      "step": 3321
+    },
+    {
+      "epoch": 2.4462444771723124,
+      "grad_norm": 0.3403468132019043,
+      "learning_rate": 1.0027861829824953e-06,
+      "loss": 0.3854,
+      "step": 3322
+    },
+    {
+      "epoch": 2.4469808541973492,
+      "grad_norm": 0.3459208011627197,
+      "learning_rate": 1.0002136091645936e-06,
+      "loss": 0.3169,
+      "step": 3323
+    },
+    {
+      "epoch": 2.447717231222386,
+      "grad_norm": 0.3217117190361023,
+      "learning_rate": 9.976439726689469e-07,
+      "loss": 0.3866,
+      "step": 3324
+    },
+    {
+      "epoch": 2.448453608247423,
+      "grad_norm": 0.37422022223472595,
+      "learning_rate": 9.95077275382621e-07,
+      "loss": 0.3865,
+      "step": 3325
+    },
+    {
+      "epoch": 2.4491899852724597,
+      "grad_norm": 0.37057486176490784,
+      "learning_rate": 9.925135191905194e-07,
+      "loss": 0.3944,
+      "step": 3326
+    },
+    {
+      "epoch": 2.4499263622974965,
+      "grad_norm": 0.33397674560546875,
+      "learning_rate": 9.89952705975386e-07,
+      "loss": 0.3958,
+      "step": 3327
+    },
+    {
+      "epoch": 2.4506627393225333,
+      "grad_norm": 0.3693501353263855,
+      "learning_rate": 9.873948376178073e-07,
+      "loss": 0.3158,
+      "step": 3328
+    },
+    {
+      "epoch": 2.45139911634757,
+      "grad_norm": 0.3479422330856323,
+      "learning_rate": 9.84839915996203e-07,
+      "loss": 0.3486,
+      "step": 3329
+    },
+    {
+      "epoch": 2.452135493372607,
+      "grad_norm": 0.3450046479701996,
+      "learning_rate": 9.822879429868304e-07,
+      "loss": 0.3718,
+      "step": 3330
+    },
+    {
+      "epoch": 2.4528718703976438,
+      "grad_norm": 0.33801180124282837,
+      "learning_rate": 9.79738920463782e-07,
+      "loss": 0.3553,
+      "step": 3331
+    },
+    {
+      "epoch": 2.4536082474226806,
+      "grad_norm": 0.3484099507331848,
+      "learning_rate": 9.771928502989802e-07,
+      "loss": 0.3978,
+      "step": 3332
+    },
+    {
+      "epoch": 2.4543446244477174,
+      "grad_norm": 0.34799516201019287,
+      "learning_rate": 9.746497343621857e-07,
+      "loss": 0.393,
+      "step": 3333
+    },
+    {
+      "epoch": 2.455081001472754,
+      "grad_norm": 0.3594471216201782,
+      "learning_rate": 9.721095745209847e-07,
+      "loss": 0.3844,
+      "step": 3334
+    },
+    {
+      "epoch": 2.455817378497791,
+      "grad_norm": 0.35299617052078247,
+      "learning_rate": 9.695723726407918e-07,
+      "loss": 0.3746,
+      "step": 3335
+    },
+    {
+      "epoch": 2.456553755522828,
+      "grad_norm": 0.309285432100296,
+      "learning_rate": 9.670381305848547e-07,
+      "loss": 0.3796,
+      "step": 3336
+    },
+    {
+      "epoch": 2.4572901325478647,
+      "grad_norm": 0.3416523039340973,
+      "learning_rate": 9.64506850214243e-07,
+      "loss": 0.3789,
+      "step": 3337
+    },
+    {
+      "epoch": 2.4580265095729015,
+      "grad_norm": 0.3610781729221344,
+      "learning_rate": 9.6197853338785e-07,
+      "loss": 0.3822,
+      "step": 3338
+    },
+    {
+      "epoch": 2.4587628865979383,
+      "grad_norm": 0.36001071333885193,
+      "learning_rate": 9.594531819624003e-07,
+      "loss": 0.4023,
+      "step": 3339
+    },
+    {
+      "epoch": 2.459499263622975,
+      "grad_norm": 0.3570241630077362,
+      "learning_rate": 9.569307977924304e-07,
+      "loss": 0.3697,
+      "step": 3340
+    },
+    {
+      "epoch": 2.460235640648012,
+      "grad_norm": 0.34070777893066406,
+      "learning_rate": 9.544113827303064e-07,
+      "loss": 0.3789,
+      "step": 3341
+    },
+    {
+      "epoch": 2.4609720176730487,
+      "grad_norm": 0.31999388337135315,
+      "learning_rate": 9.518949386262088e-07,
+      "loss": 0.3746,
+      "step": 3342
+    },
+    {
+      "epoch": 2.4617083946980856,
+      "grad_norm": 0.3107585608959198,
+      "learning_rate": 9.493814673281382e-07,
+      "loss": 0.3745,
+      "step": 3343
+    },
+    {
+      "epoch": 2.4624447717231224,
+      "grad_norm": 0.33291155099868774,
+      "learning_rate": 9.468709706819141e-07,
+      "loss": 0.3693,
+      "step": 3344
+    },
+    {
+      "epoch": 2.463181148748159,
+      "grad_norm": 0.35321640968322754,
+      "learning_rate": 9.443634505311671e-07,
+      "loss": 0.3677,
+      "step": 3345
+    },
+    {
+      "epoch": 2.463917525773196,
+      "grad_norm": 0.31461623311042786,
+      "learning_rate": 9.418589087173441e-07,
+      "loss": 0.4084,
+      "step": 3346
+    },
+    {
+      "epoch": 2.464653902798233,
+      "grad_norm": 0.311948299407959,
+      "learning_rate": 9.393573470797079e-07,
+      "loss": 0.3756,
+      "step": 3347
+    },
+    {
+      "epoch": 2.4653902798232696,
+      "grad_norm": 0.34136369824409485,
+      "learning_rate": 9.368587674553265e-07,
+      "loss": 0.3904,
+      "step": 3348
+    },
+    {
+      "epoch": 2.4661266568483065,
+      "grad_norm": 0.3355327546596527,
+      "learning_rate": 9.343631716790813e-07,
+      "loss": 0.3842,
+      "step": 3349
+    },
+    {
+      "epoch": 2.4668630338733433,
+      "grad_norm": 0.32075121998786926,
+      "learning_rate": 9.318705615836648e-07,
+      "loss": 0.3478,
+      "step": 3350
+    },
+    {
+      "epoch": 2.46759941089838,
+      "grad_norm": 0.31247884035110474,
+      "learning_rate": 9.293809389995734e-07,
+      "loss": 0.3843,
+      "step": 3351
+    },
+    {
+      "epoch": 2.468335787923417,
+      "grad_norm": 0.33083632588386536,
+      "learning_rate": 9.268943057551089e-07,
+      "loss": 0.3705,
+      "step": 3352
+    },
+    {
+      "epoch": 2.4690721649484537,
+      "grad_norm": 0.34602615237236023,
+      "learning_rate": 9.244106636763827e-07,
+      "loss": 0.3816,
+      "step": 3353
+    },
+    {
+      "epoch": 2.4698085419734905,
+      "grad_norm": 0.3348204493522644,
+      "learning_rate": 9.219300145873051e-07,
+      "loss": 0.4182,
+      "step": 3354
+    },
+    {
+      "epoch": 2.4705449189985274,
+      "grad_norm": 0.30488619208335876,
+      "learning_rate": 9.19452360309589e-07,
+      "loss": 0.3337,
+      "step": 3355
+    },
+    {
+      "epoch": 2.471281296023564,
+      "grad_norm": 0.34817051887512207,
+      "learning_rate": 9.169777026627514e-07,
+      "loss": 0.3839,
+      "step": 3356
+    },
+    {
+      "epoch": 2.472017673048601,
+      "grad_norm": 0.328900545835495,
+      "learning_rate": 9.145060434641017e-07,
+      "loss": 0.3649,
+      "step": 3357
+    },
+    {
+      "epoch": 2.472754050073638,
+      "grad_norm": 0.36652466654777527,
+      "learning_rate": 9.120373845287561e-07,
+      "loss": 0.3494,
+      "step": 3358
+    },
+    {
+      "epoch": 2.4734904270986746,
+      "grad_norm": 0.3472610116004944,
+      "learning_rate": 9.095717276696214e-07,
+      "loss": 0.3894,
+      "step": 3359
+    },
+    {
+      "epoch": 2.4742268041237114,
+      "grad_norm": 0.3122188448905945,
+      "learning_rate": 9.071090746973999e-07,
+      "loss": 0.3546,
+      "step": 3360
+    },
+    {
+      "epoch": 2.4749631811487482,
+      "grad_norm": 0.338001549243927,
+      "learning_rate": 9.046494274205924e-07,
+      "loss": 0.363,
+      "step": 3361
+    },
+    {
+      "epoch": 2.475699558173785,
+      "grad_norm": 0.32908204197883606,
+      "learning_rate": 9.021927876454883e-07,
+      "loss": 0.3786,
+      "step": 3362
+    },
+    {
+      "epoch": 2.476435935198822,
+      "grad_norm": 0.3352196216583252,
+      "learning_rate": 8.997391571761682e-07,
+      "loss": 0.3681,
+      "step": 3363
+    },
+    {
+      "epoch": 2.4771723122238587,
+      "grad_norm": 0.3196353316307068,
+      "learning_rate": 8.972885378145079e-07,
+      "loss": 0.3735,
+      "step": 3364
+    },
+    {
+      "epoch": 2.4779086892488955,
+      "grad_norm": 0.3419160544872284,
+      "learning_rate": 8.94840931360163e-07,
+      "loss": 0.3501,
+      "step": 3365
+    },
+    {
+      "epoch": 2.4786450662739323,
+      "grad_norm": 0.3332716226577759,
+      "learning_rate": 8.923963396105861e-07,
+      "loss": 0.3606,
+      "step": 3366
+    },
+    {
+      "epoch": 2.479381443298969,
+      "grad_norm": 0.3240850269794464,
+      "learning_rate": 8.899547643610102e-07,
+      "loss": 0.3774,
+      "step": 3367
+    },
+    {
+      "epoch": 2.480117820324006,
+      "grad_norm": 0.3325689136981964,
+      "learning_rate": 8.875162074044524e-07,
+      "loss": 0.3788,
+      "step": 3368
+    },
+    {
+      "epoch": 2.4808541973490428,
+      "grad_norm": 0.3567977547645569,
+      "learning_rate": 8.850806705317183e-07,
+      "loss": 0.3516,
+      "step": 3369
+    },
+    {
+      "epoch": 2.4815905743740796,
+      "grad_norm": 0.33295658230781555,
+      "learning_rate": 8.826481555313909e-07,
+      "loss": 0.3668,
+      "step": 3370
+    },
+    {
+      "epoch": 2.4823269513991164,
+      "grad_norm": 0.3270358443260193,
+      "learning_rate": 8.802186641898352e-07,
+      "loss": 0.3957,
+      "step": 3371
+    },
+    {
+      "epoch": 2.4830633284241532,
+      "grad_norm": 0.3151583671569824,
+      "learning_rate": 8.777921982911996e-07,
+      "loss": 0.4113,
+      "step": 3372
+    },
+    {
+      "epoch": 2.48379970544919,
+      "grad_norm": 0.3360562026500702,
+      "learning_rate": 8.753687596174021e-07,
+      "loss": 0.4078,
+      "step": 3373
+    },
+    {
+      "epoch": 2.484536082474227,
+      "grad_norm": 0.3369217813014984,
+      "learning_rate": 8.729483499481467e-07,
+      "loss": 0.383,
+      "step": 3374
+    },
+    {
+      "epoch": 2.4852724594992637,
+      "grad_norm": 0.3392618000507355,
+      "learning_rate": 8.705309710609078e-07,
+      "loss": 0.3923,
+      "step": 3375
+    },
+    {
+      "epoch": 2.4860088365243005,
+      "grad_norm": 0.3362715244293213,
+      "learning_rate": 8.681166247309348e-07,
+      "loss": 0.3598,
+      "step": 3376
+    },
+    {
+      "epoch": 2.4867452135493373,
+      "grad_norm": 0.34049803018569946,
+      "learning_rate": 8.65705312731252e-07,
+      "loss": 0.3656,
+      "step": 3377
+    },
+    {
+      "epoch": 2.487481590574374,
+      "grad_norm": 0.328102171421051,
+      "learning_rate": 8.632970368326537e-07,
+      "loss": 0.3924,
+      "step": 3378
+    },
+    {
+      "epoch": 2.488217967599411,
+      "grad_norm": 0.3466600775718689,
+      "learning_rate": 8.608917988037036e-07,
+      "loss": 0.3835,
+      "step": 3379
+    },
+    {
+      "epoch": 2.4889543446244478,
+      "grad_norm": 0.3212660551071167,
+      "learning_rate": 8.584896004107379e-07,
+      "loss": 0.382,
+      "step": 3380
+    },
+    {
+      "epoch": 2.4896907216494846,
+      "grad_norm": 0.3672233819961548,
+      "learning_rate": 8.56090443417859e-07,
+      "loss": 0.3818,
+      "step": 3381
+    },
+    {
+      "epoch": 2.4904270986745214,
+      "grad_norm": 0.3334003984928131,
+      "learning_rate": 8.536943295869315e-07,
+      "loss": 0.3788,
+      "step": 3382
+    },
+    {
+      "epoch": 2.491163475699558,
+      "grad_norm": 0.35568714141845703,
+      "learning_rate": 8.513012606775928e-07,
+      "loss": 0.3687,
+      "step": 3383
+    },
+    {
+      "epoch": 2.491899852724595,
+      "grad_norm": 0.34635692834854126,
+      "learning_rate": 8.489112384472386e-07,
+      "loss": 0.3541,
+      "step": 3384
+    },
+    {
+      "epoch": 2.492636229749632,
+      "grad_norm": 0.36689338088035583,
+      "learning_rate": 8.46524264651028e-07,
+      "loss": 0.3686,
+      "step": 3385
+    },
+    {
+      "epoch": 2.4933726067746687,
+      "grad_norm": 0.36320188641548157,
+      "learning_rate": 8.441403410418853e-07,
+      "loss": 0.3795,
+      "step": 3386
+    },
+    {
+      "epoch": 2.4941089837997055,
+      "grad_norm": 0.3351554274559021,
+      "learning_rate": 8.417594693704901e-07,
+      "loss": 0.365,
+      "step": 3387
+    },
+    {
+      "epoch": 2.4948453608247423,
+      "grad_norm": 0.3230125606060028,
+      "learning_rate": 8.393816513852815e-07,
+      "loss": 0.3867,
+      "step": 3388
+    },
+    {
+      "epoch": 2.495581737849779,
+      "grad_norm": 0.3341889977455139,
+      "learning_rate": 8.370068888324612e-07,
+      "loss": 0.3554,
+      "step": 3389
+    },
+    {
+      "epoch": 2.496318114874816,
+      "grad_norm": 0.3187844753265381,
+      "learning_rate": 8.346351834559784e-07,
+      "loss": 0.388,
+      "step": 3390
+    },
+    {
+      "epoch": 2.4970544918998527,
+      "grad_norm": 0.3333573639392853,
+      "learning_rate": 8.322665369975447e-07,
+      "loss": 0.3775,
+      "step": 3391
+    },
+    {
+      "epoch": 2.4977908689248896,
+      "grad_norm": 0.32489851117134094,
+      "learning_rate": 8.299009511966221e-07,
+      "loss": 0.3604,
+      "step": 3392
+    },
+    {
+      "epoch": 2.4985272459499264,
+      "grad_norm": 0.3213289976119995,
+      "learning_rate": 8.275384277904231e-07,
+      "loss": 0.3635,
+      "step": 3393
+    },
+    {
+      "epoch": 2.499263622974963,
+      "grad_norm": 0.341377854347229,
+      "learning_rate": 8.251789685139172e-07,
+      "loss": 0.3761,
+      "step": 3394
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.34892818331718445,
+      "learning_rate": 8.228225750998176e-07,
+      "loss": 0.3652,
+      "step": 3395
+    },
+    {
+      "epoch": 2.500736377025037,
+      "grad_norm": 0.31448036432266235,
+      "learning_rate": 8.204692492785876e-07,
+      "loss": 0.3864,
+      "step": 3396
+    },
+    {
+      "epoch": 2.5014727540500736,
+      "grad_norm": 0.31953367590904236,
+      "learning_rate": 8.181189927784416e-07,
+      "loss": 0.364,
+      "step": 3397
+    },
+    {
+      "epoch": 2.5022091310751104,
+      "grad_norm": 0.33577442169189453,
+      "learning_rate": 8.157718073253351e-07,
+      "loss": 0.3808,
+      "step": 3398
+    },
+    {
+      "epoch": 2.5029455081001473,
+      "grad_norm": 0.3457181453704834,
+      "learning_rate": 8.134276946429703e-07,
+      "loss": 0.3665,
+      "step": 3399
+    },
+    {
+      "epoch": 2.503681885125184,
+      "grad_norm": 0.34147438406944275,
+      "learning_rate": 8.110866564527925e-07,
+      "loss": 0.3755,
+      "step": 3400
+    },
+    {
+      "epoch": 2.504418262150221,
+      "grad_norm": 0.3595358729362488,
+      "learning_rate": 8.087486944739886e-07,
+      "loss": 0.4067,
+      "step": 3401
+    },
+    {
+      "epoch": 2.5051546391752577,
+      "grad_norm": 0.32780689001083374,
+      "learning_rate": 8.064138104234897e-07,
+      "loss": 0.3573,
+      "step": 3402
+    },
+    {
+      "epoch": 2.5058910162002945,
+      "grad_norm": 0.3280833065509796,
+      "learning_rate": 8.040820060159621e-07,
+      "loss": 0.3645,
+      "step": 3403
+    },
+    {
+      "epoch": 2.5066273932253313,
+      "grad_norm": 0.3275584280490875,
+      "learning_rate": 8.017532829638119e-07,
+      "loss": 0.3812,
+      "step": 3404
+    },
+    {
+      "epoch": 2.507363770250368,
+      "grad_norm": 0.363805890083313,
+      "learning_rate": 7.994276429771857e-07,
+      "loss": 0.3802,
+      "step": 3405
+    },
+    {
+      "epoch": 2.508100147275405,
+      "grad_norm": 0.317568302154541,
+      "learning_rate": 7.971050877639624e-07,
+      "loss": 0.348,
+      "step": 3406
+    },
+    {
+      "epoch": 2.508836524300442,
+      "grad_norm": 0.3445456922054291,
+      "learning_rate": 7.947856190297538e-07,
+      "loss": 0.3498,
+      "step": 3407
+    },
+    {
+      "epoch": 2.5095729013254786,
+      "grad_norm": 0.3341994285583496,
+      "learning_rate": 7.924692384779098e-07,
+      "loss": 0.3574,
+      "step": 3408
+    },
+    {
+      "epoch": 2.5103092783505154,
+      "grad_norm": 0.368876188993454,
+      "learning_rate": 7.901559478095106e-07,
+      "loss": 0.4066,
+      "step": 3409
+    },
+    {
+      "epoch": 2.5110456553755522,
+      "grad_norm": 0.333023339509964,
+      "learning_rate": 7.878457487233643e-07,
+      "loss": 0.3624,
+      "step": 3410
+    },
+    {
+      "epoch": 2.511782032400589,
+      "grad_norm": 0.33965203166007996,
+      "learning_rate": 7.85538642916015e-07,
+      "loss": 0.379,
+      "step": 3411
+    },
+    {
+      "epoch": 2.512518409425626,
+      "grad_norm": 0.3300594091415405,
+      "learning_rate": 7.832346320817297e-07,
+      "loss": 0.3837,
+      "step": 3412
+    },
+    {
+      "epoch": 2.5132547864506627,
+      "grad_norm": 0.3793734312057495,
+      "learning_rate": 7.809337179125031e-07,
+      "loss": 0.3478,
+      "step": 3413
+    },
+    {
+      "epoch": 2.5139911634756995,
+      "grad_norm": 0.3749076724052429,
+      "learning_rate": 7.786359020980605e-07,
+      "loss": 0.3635,
+      "step": 3414
+    },
+    {
+      "epoch": 2.5147275405007363,
+      "grad_norm": 0.3277282118797302,
+      "learning_rate": 7.763411863258441e-07,
+      "loss": 0.3736,
+      "step": 3415
+    },
+    {
+      "epoch": 2.515463917525773,
+      "grad_norm": 0.3639467656612396,
+      "learning_rate": 7.740495722810271e-07,
+      "loss": 0.3579,
+      "step": 3416
+    },
+    {
+      "epoch": 2.51620029455081,
+      "grad_norm": 0.36798664927482605,
+      "learning_rate": 7.717610616464999e-07,
+      "loss": 0.3939,
+      "step": 3417
+    },
+    {
+      "epoch": 2.5169366715758468,
+      "grad_norm": 0.3840385675430298,
+      "learning_rate": 7.694756561028754e-07,
+      "loss": 0.3643,
+      "step": 3418
+    },
+    {
+      "epoch": 2.5176730486008836,
+      "grad_norm": 0.3123991787433624,
+      "learning_rate": 7.671933573284878e-07,
+      "loss": 0.3923,
+      "step": 3419
+    },
+    {
+      "epoch": 2.5184094256259204,
+      "grad_norm": 0.37913936376571655,
+      "learning_rate": 7.649141669993881e-07,
+      "loss": 0.3921,
+      "step": 3420
+    },
+    {
+      "epoch": 2.5191458026509572,
+      "grad_norm": 0.30917784571647644,
+      "learning_rate": 7.626380867893429e-07,
+      "loss": 0.3538,
+      "step": 3421
+    },
+    {
+      "epoch": 2.519882179675994,
+      "grad_norm": 0.3407180607318878,
+      "learning_rate": 7.603651183698396e-07,
+      "loss": 0.4078,
+      "step": 3422
+    },
+    {
+      "epoch": 2.520618556701031,
+      "grad_norm": 0.34745657444000244,
+      "learning_rate": 7.580952634100758e-07,
+      "loss": 0.387,
+      "step": 3423
+    },
+    {
+      "epoch": 2.5213549337260677,
+      "grad_norm": 0.33488914370536804,
+      "learning_rate": 7.558285235769647e-07,
+      "loss": 0.3858,
+      "step": 3424
+    },
+    {
+      "epoch": 2.5220913107511045,
+      "grad_norm": 0.3565431237220764,
+      "learning_rate": 7.535649005351309e-07,
+      "loss": 0.3824,
+      "step": 3425
+    },
+    {
+      "epoch": 2.5228276877761413,
+      "grad_norm": 0.3106916546821594,
+      "learning_rate": 7.513043959469107e-07,
+      "loss": 0.3919,
+      "step": 3426
+    },
+    {
+      "epoch": 2.523564064801178,
+      "grad_norm": 0.34295180439949036,
+      "learning_rate": 7.49047011472352e-07,
+      "loss": 0.398,
+      "step": 3427
+    },
+    {
+      "epoch": 2.524300441826215,
+      "grad_norm": 0.316560834646225,
+      "learning_rate": 7.467927487692089e-07,
+      "loss": 0.3448,
+      "step": 3428
+    },
+    {
+      "epoch": 2.5250368188512518,
+      "grad_norm": 0.33750441670417786,
+      "learning_rate": 7.445416094929426e-07,
+      "loss": 0.3661,
+      "step": 3429
+    },
+    {
+      "epoch": 2.5257731958762886,
+      "grad_norm": 0.34076401591300964,
+      "learning_rate": 7.422935952967236e-07,
+      "loss": 0.3744,
+      "step": 3430
+    },
+    {
+      "epoch": 2.5265095729013254,
+      "grad_norm": 0.3119591772556305,
+      "learning_rate": 7.40048707831425e-07,
+      "loss": 0.3562,
+      "step": 3431
+    },
+    {
+      "epoch": 2.527245949926362,
+      "grad_norm": 0.33838844299316406,
+      "learning_rate": 7.378069487456241e-07,
+      "loss": 0.3647,
+      "step": 3432
+    },
+    {
+      "epoch": 2.527982326951399,
+      "grad_norm": 0.3333272337913513,
+      "learning_rate": 7.355683196856006e-07,
+      "loss": 0.3804,
+      "step": 3433
+    },
+    {
+      "epoch": 2.528718703976436,
+      "grad_norm": 0.31554344296455383,
+      "learning_rate": 7.333328222953356e-07,
+      "loss": 0.3709,
+      "step": 3434
+    },
+    {
+      "epoch": 2.5294550810014726,
+      "grad_norm": 0.3351151645183563,
+      "learning_rate": 7.311004582165132e-07,
+      "loss": 0.3745,
+      "step": 3435
+    },
+    {
+      "epoch": 2.5301914580265095,
+      "grad_norm": 0.3532677888870239,
+      "learning_rate": 7.288712290885119e-07,
+      "loss": 0.3772,
+      "step": 3436
+    },
+    {
+      "epoch": 2.5309278350515463,
+      "grad_norm": 0.33610615134239197,
+      "learning_rate": 7.266451365484106e-07,
+      "loss": 0.37,
+      "step": 3437
+    },
+    {
+      "epoch": 2.531664212076583,
+      "grad_norm": 0.33613523840904236,
+      "learning_rate": 7.244221822309855e-07,
+      "loss": 0.3705,
+      "step": 3438
+    },
+    {
+      "epoch": 2.53240058910162,
+      "grad_norm": 0.3339458107948303,
+      "learning_rate": 7.222023677687062e-07,
+      "loss": 0.382,
+      "step": 3439
+    },
+    {
+      "epoch": 2.5331369661266567,
+      "grad_norm": 0.3765323758125305,
+      "learning_rate": 7.199856947917372e-07,
+      "loss": 0.3996,
+      "step": 3440
+    },
+    {
+      "epoch": 2.5338733431516935,
+      "grad_norm": 0.3355490565299988,
+      "learning_rate": 7.177721649279367e-07,
+      "loss": 0.3722,
+      "step": 3441
+    },
+    {
+      "epoch": 2.5346097201767304,
+      "grad_norm": 0.33720365166664124,
+      "learning_rate": 7.155617798028542e-07,
+      "loss": 0.3867,
+      "step": 3442
+    },
+    {
+      "epoch": 2.535346097201767,
+      "grad_norm": 0.3521978259086609,
+      "learning_rate": 7.133545410397274e-07,
+      "loss": 0.3626,
+      "step": 3443
+    },
+    {
+      "epoch": 2.536082474226804,
+      "grad_norm": 0.38076502084732056,
+      "learning_rate": 7.111504502594896e-07,
+      "loss": 0.3998,
+      "step": 3444
+    },
+    {
+      "epoch": 2.536818851251841,
+      "grad_norm": 0.3390524983406067,
+      "learning_rate": 7.089495090807564e-07,
+      "loss": 0.4084,
+      "step": 3445
+    },
+    {
+      "epoch": 2.5375552282768776,
+      "grad_norm": 0.3228042423725128,
+      "learning_rate": 7.067517191198314e-07,
+      "loss": 0.3477,
+      "step": 3446
+    },
+    {
+      "epoch": 2.5382916053019144,
+      "grad_norm": 0.373737633228302,
+      "learning_rate": 7.045570819907072e-07,
+      "loss": 0.3804,
+      "step": 3447
+    },
+    {
+      "epoch": 2.5390279823269513,
+      "grad_norm": 0.33321553468704224,
+      "learning_rate": 7.023655993050588e-07,
+      "loss": 0.3792,
+      "step": 3448
+    },
+    {
+      "epoch": 2.539764359351988,
+      "grad_norm": 0.35181403160095215,
+      "learning_rate": 7.001772726722439e-07,
+      "loss": 0.3664,
+      "step": 3449
+    },
+    {
+      "epoch": 2.540500736377025,
+      "grad_norm": 0.36001092195510864,
+      "learning_rate": 6.979921036993042e-07,
+      "loss": 0.4203,
+      "step": 3450
+    },
+    {
+      "epoch": 2.5412371134020617,
+      "grad_norm": 0.32719293236732483,
+      "learning_rate": 6.958100939909601e-07,
+      "loss": 0.3662,
+      "step": 3451
+    },
+    {
+      "epoch": 2.5419734904270985,
+      "grad_norm": 0.33018410205841064,
+      "learning_rate": 6.936312451496157e-07,
+      "loss": 0.3636,
+      "step": 3452
+    },
+    {
+      "epoch": 2.5427098674521353,
+      "grad_norm": 0.32865098118782043,
+      "learning_rate": 6.914555587753508e-07,
+      "loss": 0.3847,
+      "step": 3453
+    },
+    {
+      "epoch": 2.543446244477172,
+      "grad_norm": 0.324142187833786,
+      "learning_rate": 6.892830364659231e-07,
+      "loss": 0.3712,
+      "step": 3454
+    },
+    {
+      "epoch": 2.544182621502209,
+      "grad_norm": 0.3413415849208832,
+      "learning_rate": 6.871136798167693e-07,
+      "loss": 0.3993,
+      "step": 3455
+    },
+    {
+      "epoch": 2.544918998527246,
+      "grad_norm": 0.32384759187698364,
+      "learning_rate": 6.849474904209979e-07,
+      "loss": 0.3785,
+      "step": 3456
+    },
+    {
+      "epoch": 2.5456553755522826,
+      "grad_norm": 0.32682374119758606,
+      "learning_rate": 6.827844698693931e-07,
+      "loss": 0.3639,
+      "step": 3457
+    },
+    {
+      "epoch": 2.5463917525773194,
+      "grad_norm": 0.3174643814563751,
+      "learning_rate": 6.806246197504118e-07,
+      "loss": 0.3797,
+      "step": 3458
+    },
+    {
+      "epoch": 2.5471281296023562,
+      "grad_norm": 0.3432275056838989,
+      "learning_rate": 6.784679416501822e-07,
+      "loss": 0.3791,
+      "step": 3459
+    },
+    {
+      "epoch": 2.547864506627393,
+      "grad_norm": 0.35215064883232117,
+      "learning_rate": 6.763144371525048e-07,
+      "loss": 0.3873,
+      "step": 3460
+    },
+    {
+      "epoch": 2.54860088365243,
+      "grad_norm": 0.36379846930503845,
+      "learning_rate": 6.741641078388472e-07,
+      "loss": 0.3611,
+      "step": 3461
+    },
+    {
+      "epoch": 2.5493372606774667,
+      "grad_norm": 0.3444019854068756,
+      "learning_rate": 6.72016955288346e-07,
+      "loss": 0.3479,
+      "step": 3462
+    },
+    {
+      "epoch": 2.5500736377025035,
+      "grad_norm": 0.3353855311870575,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.3835,
+      "step": 3463
+    },
+    {
+      "epoch": 2.5508100147275403,
+      "grad_norm": 0.3298718333244324,
+      "learning_rate": 6.677321867816983e-07,
+      "loss": 0.4032,
+      "step": 3464
+    },
+    {
+      "epoch": 2.551546391752577,
+      "grad_norm": 0.3183014392852783,
+      "learning_rate": 6.655945739721548e-07,
+      "loss": 0.3547,
+      "step": 3465
+    },
+    {
+      "epoch": 2.552282768777614,
+      "grad_norm": 0.33290722966194153,
+      "learning_rate": 6.634601442189753e-07,
+      "loss": 0.3874,
+      "step": 3466
+    },
+    {
+      "epoch": 2.5530191458026508,
+      "grad_norm": 0.3307499289512634,
+      "learning_rate": 6.613288990896205e-07,
+      "loss": 0.3761,
+      "step": 3467
+    },
+    {
+      "epoch": 2.5537555228276876,
+      "grad_norm": 0.3292166590690613,
+      "learning_rate": 6.592008401492106e-07,
+      "loss": 0.3838,
+      "step": 3468
+    },
+    {
+      "epoch": 2.5544918998527244,
+      "grad_norm": 0.32803595066070557,
+      "learning_rate": 6.570759689605305e-07,
+      "loss": 0.3759,
+      "step": 3469
+    },
+    {
+      "epoch": 2.555228276877761,
+      "grad_norm": 0.36094194650650024,
+      "learning_rate": 6.549542870840203e-07,
+      "loss": 0.346,
+      "step": 3470
+    },
+    {
+      "epoch": 2.555964653902798,
+      "grad_norm": 0.35194647312164307,
+      "learning_rate": 6.528357960777776e-07,
+      "loss": 0.3795,
+      "step": 3471
+    },
+    {
+      "epoch": 2.556701030927835,
+      "grad_norm": 0.3414342701435089,
+      "learning_rate": 6.507204974975611e-07,
+      "loss": 0.3704,
+      "step": 3472
+    },
+    {
+      "epoch": 2.5574374079528717,
+      "grad_norm": 0.3302089273929596,
+      "learning_rate": 6.486083928967801e-07,
+      "loss": 0.3581,
+      "step": 3473
+    },
+    {
+      "epoch": 2.5581737849779085,
+      "grad_norm": 0.3138299286365509,
+      "learning_rate": 6.46499483826501e-07,
+      "loss": 0.3649,
+      "step": 3474
+    },
+    {
+      "epoch": 2.5589101620029453,
+      "grad_norm": 0.34113311767578125,
+      "learning_rate": 6.443937718354426e-07,
+      "loss": 0.3937,
+      "step": 3475
+    },
+    {
+      "epoch": 2.559646539027982,
+      "grad_norm": 0.33165475726127625,
+      "learning_rate": 6.422912584699753e-07,
+      "loss": 0.3952,
+      "step": 3476
+    },
+    {
+      "epoch": 2.560382916053019,
+      "grad_norm": 0.35434436798095703,
+      "learning_rate": 6.401919452741234e-07,
+      "loss": 0.3813,
+      "step": 3477
+    },
+    {
+      "epoch": 2.5611192930780557,
+      "grad_norm": 0.3147828280925751,
+      "learning_rate": 6.380958337895582e-07,
+      "loss": 0.378,
+      "step": 3478
+    },
+    {
+      "epoch": 2.5618556701030926,
+      "grad_norm": 0.34505346417427063,
+      "learning_rate": 6.360029255555994e-07,
+      "loss": 0.3648,
+      "step": 3479
+    },
+    {
+      "epoch": 2.5625920471281294,
+      "grad_norm": 0.3115389049053192,
+      "learning_rate": 6.339132221092181e-07,
+      "loss": 0.3581,
+      "step": 3480
+    },
+    {
+      "epoch": 2.563328424153166,
+      "grad_norm": 0.3120824694633484,
+      "learning_rate": 6.318267249850274e-07,
+      "loss": 0.3574,
+      "step": 3481
+    },
+    {
+      "epoch": 2.564064801178203,
+      "grad_norm": 0.3659776747226715,
+      "learning_rate": 6.297434357152882e-07,
+      "loss": 0.3897,
+      "step": 3482
+    },
+    {
+      "epoch": 2.56480117820324,
+      "grad_norm": 0.33508267998695374,
+      "learning_rate": 6.276633558299056e-07,
+      "loss": 0.3856,
+      "step": 3483
+    },
+    {
+      "epoch": 2.5655375552282766,
+      "grad_norm": 0.3201551139354706,
+      "learning_rate": 6.25586486856426e-07,
+      "loss": 0.3766,
+      "step": 3484
+    },
+    {
+      "epoch": 2.5662739322533135,
+      "grad_norm": 0.332343727350235,
+      "learning_rate": 6.23512830320041e-07,
+      "loss": 0.3716,
+      "step": 3485
+    },
+    {
+      "epoch": 2.5670103092783503,
+      "grad_norm": 0.346399188041687,
+      "learning_rate": 6.214423877435805e-07,
+      "loss": 0.3845,
+      "step": 3486
+    },
+    {
+      "epoch": 2.567746686303387,
+      "grad_norm": 0.3369497060775757,
+      "learning_rate": 6.193751606475141e-07,
+      "loss": 0.3658,
+      "step": 3487
+    },
+    {
+      "epoch": 2.568483063328424,
+      "grad_norm": 0.3280666768550873,
+      "learning_rate": 6.17311150549953e-07,
+      "loss": 0.3755,
+      "step": 3488
+    },
+    {
+      "epoch": 2.5692194403534607,
+      "grad_norm": 0.30463147163391113,
+      "learning_rate": 6.152503589666426e-07,
+      "loss": 0.3658,
+      "step": 3489
+    },
+    {
+      "epoch": 2.5699558173784975,
+      "grad_norm": 0.3326549828052521,
+      "learning_rate": 6.131927874109661e-07,
+      "loss": 0.3723,
+      "step": 3490
+    },
+    {
+      "epoch": 2.5706921944035344,
+      "grad_norm": 0.338368684053421,
+      "learning_rate": 6.111384373939416e-07,
+      "loss": 0.366,
+      "step": 3491
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.3462516963481903,
+      "learning_rate": 6.090873104242213e-07,
+      "loss": 0.3906,
+      "step": 3492
+    },
+    {
+      "epoch": 2.572164948453608,
+      "grad_norm": 0.33473119139671326,
+      "learning_rate": 6.070394080080921e-07,
+      "loss": 0.3634,
+      "step": 3493
+    },
+    {
+      "epoch": 2.572901325478645,
+      "grad_norm": 0.3650546073913574,
+      "learning_rate": 6.049947316494709e-07,
+      "loss": 0.3895,
+      "step": 3494
+    },
+    {
+      "epoch": 2.5736377025036816,
+      "grad_norm": 0.3254354000091553,
+      "learning_rate": 6.029532828499052e-07,
+      "loss": 0.3694,
+      "step": 3495
+    },
+    {
+      "epoch": 2.5743740795287184,
+      "grad_norm": 0.3232567012310028,
+      "learning_rate": 6.009150631085758e-07,
+      "loss": 0.3675,
+      "step": 3496
+    },
+    {
+      "epoch": 2.5751104565537553,
+      "grad_norm": 0.3334692120552063,
+      "learning_rate": 5.988800739222884e-07,
+      "loss": 0.3539,
+      "step": 3497
+    },
+    {
+      "epoch": 2.575846833578792,
+      "grad_norm": 0.32919347286224365,
+      "learning_rate": 5.968483167854761e-07,
+      "loss": 0.3744,
+      "step": 3498
+    },
+    {
+      "epoch": 2.576583210603829,
+      "grad_norm": 0.3350676894187927,
+      "learning_rate": 5.948197931902034e-07,
+      "loss": 0.3817,
+      "step": 3499
+    },
+    {
+      "epoch": 2.5773195876288657,
+      "grad_norm": 0.32519227266311646,
+      "learning_rate": 5.927945046261541e-07,
+      "loss": 0.3669,
+      "step": 3500
+    },
+    {
+      "epoch": 2.5780559646539025,
+      "grad_norm": 0.31917646527290344,
+      "learning_rate": 5.90772452580638e-07,
+      "loss": 0.3524,
+      "step": 3501
+    },
+    {
+      "epoch": 2.57879234167894,
+      "grad_norm": 0.3143576979637146,
+      "learning_rate": 5.887536385385917e-07,
+      "loss": 0.3696,
+      "step": 3502
+    },
+    {
+      "epoch": 2.5795287187039766,
+      "grad_norm": 0.32816874980926514,
+      "learning_rate": 5.867380639825698e-07,
+      "loss": 0.3613,
+      "step": 3503
+    },
+    {
+      "epoch": 2.5802650957290134,
+      "grad_norm": 0.3384314179420471,
+      "learning_rate": 5.847257303927484e-07,
+      "loss": 0.3812,
+      "step": 3504
+    },
+    {
+      "epoch": 2.5810014727540502,
+      "grad_norm": 0.34819257259368896,
+      "learning_rate": 5.827166392469269e-07,
+      "loss": 0.3749,
+      "step": 3505
+    },
+    {
+      "epoch": 2.581737849779087,
+      "grad_norm": 0.31272298097610474,
+      "learning_rate": 5.807107920205202e-07,
+      "loss": 0.3579,
+      "step": 3506
+    },
+    {
+      "epoch": 2.582474226804124,
+      "grad_norm": 0.3496764004230499,
+      "learning_rate": 5.78708190186561e-07,
+      "loss": 0.3594,
+      "step": 3507
+    },
+    {
+      "epoch": 2.5832106038291607,
+      "grad_norm": 0.3062533438205719,
+      "learning_rate": 5.767088352157002e-07,
+      "loss": 0.3901,
+      "step": 3508
+    },
+    {
+      "epoch": 2.5839469808541975,
+      "grad_norm": 0.4096584618091583,
+      "learning_rate": 5.747127285762027e-07,
+      "loss": 0.372,
+      "step": 3509
+    },
+    {
+      "epoch": 2.5846833578792343,
+      "grad_norm": 0.32007288932800293,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.366,
+      "step": 3510
+    },
+    {
+      "epoch": 2.585419734904271,
+      "grad_norm": 0.32174500823020935,
+      "learning_rate": 5.707302661524372e-07,
+      "loss": 0.3938,
+      "step": 3511
+    },
+    {
+      "epoch": 2.586156111929308,
+      "grad_norm": 0.3417740762233734,
+      "learning_rate": 5.687439132927674e-07,
+      "loss": 0.3665,
+      "step": 3512
+    },
+    {
+      "epoch": 2.5868924889543448,
+      "grad_norm": 0.32081058621406555,
+      "learning_rate": 5.66760814613661e-07,
+      "loss": 0.3919,
+      "step": 3513
+    },
+    {
+      "epoch": 2.5876288659793816,
+      "grad_norm": 0.32988831400871277,
+      "learning_rate": 5.647809715714442e-07,
+      "loss": 0.3802,
+      "step": 3514
+    },
+    {
+      "epoch": 2.5883652430044184,
+      "grad_norm": 0.3118469715118408,
+      "learning_rate": 5.628043856200543e-07,
+      "loss": 0.3748,
+      "step": 3515
+    },
+    {
+      "epoch": 2.589101620029455,
+      "grad_norm": 0.33467116951942444,
+      "learning_rate": 5.60831058211036e-07,
+      "loss": 0.4001,
+      "step": 3516
+    },
+    {
+      "epoch": 2.589837997054492,
+      "grad_norm": 0.3081044852733612,
+      "learning_rate": 5.588609907935405e-07,
+      "loss": 0.3654,
+      "step": 3517
+    },
+    {
+      "epoch": 2.590574374079529,
+      "grad_norm": 0.3295198976993561,
+      "learning_rate": 5.568941848143284e-07,
+      "loss": 0.3907,
+      "step": 3518
+    },
+    {
+      "epoch": 2.5913107511045657,
+      "grad_norm": 0.32055968046188354,
+      "learning_rate": 5.549306417177602e-07,
+      "loss": 0.3872,
+      "step": 3519
+    },
+    {
+      "epoch": 2.5920471281296025,
+      "grad_norm": 0.29698818922042847,
+      "learning_rate": 5.529703629458027e-07,
+      "loss": 0.3768,
+      "step": 3520
+    },
+    {
+      "epoch": 2.5927835051546393,
+      "grad_norm": 0.3199003040790558,
+      "learning_rate": 5.510133499380271e-07,
+      "loss": 0.359,
+      "step": 3521
+    },
+    {
+      "epoch": 2.593519882179676,
+      "grad_norm": 0.3205602169036865,
+      "learning_rate": 5.490596041316038e-07,
+      "loss": 0.3838,
+      "step": 3522
+    },
+    {
+      "epoch": 2.594256259204713,
+      "grad_norm": 0.32970884442329407,
+      "learning_rate": 5.471091269613033e-07,
+      "loss": 0.3974,
+      "step": 3523
+    },
+    {
+      "epoch": 2.5949926362297497,
+      "grad_norm": 0.3231904208660126,
+      "learning_rate": 5.451619198594998e-07,
+      "loss": 0.4017,
+      "step": 3524
+    },
+    {
+      "epoch": 2.5957290132547866,
+      "grad_norm": 0.3256881535053253,
+      "learning_rate": 5.432179842561614e-07,
+      "loss": 0.3776,
+      "step": 3525
+    },
+    {
+      "epoch": 2.5964653902798234,
+      "grad_norm": 0.3105742633342743,
+      "learning_rate": 5.412773215788547e-07,
+      "loss": 0.4146,
+      "step": 3526
+    },
+    {
+      "epoch": 2.59720176730486,
+      "grad_norm": 0.36772292852401733,
+      "learning_rate": 5.393399332527466e-07,
+      "loss": 0.3675,
+      "step": 3527
+    },
+    {
+      "epoch": 2.597938144329897,
+      "grad_norm": 0.34043705463409424,
+      "learning_rate": 5.374058207005945e-07,
+      "loss": 0.3676,
+      "step": 3528
+    },
+    {
+      "epoch": 2.598674521354934,
+      "grad_norm": 0.31147414445877075,
+      "learning_rate": 5.354749853427521e-07,
+      "loss": 0.3456,
+      "step": 3529
+    },
+    {
+      "epoch": 2.5994108983799706,
+      "grad_norm": 0.3361349105834961,
+      "learning_rate": 5.335474285971681e-07,
+      "loss": 0.3442,
+      "step": 3530
+    },
+    {
+      "epoch": 2.6001472754050075,
+      "grad_norm": 0.3142765164375305,
+      "learning_rate": 5.316231518793802e-07,
+      "loss": 0.3597,
+      "step": 3531
+    },
+    {
+      "epoch": 2.6008836524300443,
+      "grad_norm": 0.3216022253036499,
+      "learning_rate": 5.297021566025212e-07,
+      "loss": 0.3665,
+      "step": 3532
+    },
+    {
+      "epoch": 2.601620029455081,
+      "grad_norm": 0.32801806926727295,
+      "learning_rate": 5.277844441773105e-07,
+      "loss": 0.392,
+      "step": 3533
+    },
+    {
+      "epoch": 2.602356406480118,
+      "grad_norm": 0.3432222902774811,
+      "learning_rate": 5.258700160120567e-07,
+      "loss": 0.3727,
+      "step": 3534
+    },
+    {
+      "epoch": 2.6030927835051547,
+      "grad_norm": 0.3140788972377777,
+      "learning_rate": 5.239588735126611e-07,
+      "loss": 0.3747,
+      "step": 3535
+    },
+    {
+      "epoch": 2.6038291605301915,
+      "grad_norm": 0.323508083820343,
+      "learning_rate": 5.220510180826071e-07,
+      "loss": 0.416,
+      "step": 3536
+    },
+    {
+      "epoch": 2.6045655375552283,
+      "grad_norm": 0.3127088248729706,
+      "learning_rate": 5.201464511229659e-07,
+      "loss": 0.3921,
+      "step": 3537
+    },
+    {
+      "epoch": 2.605301914580265,
+      "grad_norm": 0.3234788179397583,
+      "learning_rate": 5.182451740323957e-07,
+      "loss": 0.3813,
+      "step": 3538
+    },
+    {
+      "epoch": 2.606038291605302,
+      "grad_norm": 0.30984804034233093,
+      "learning_rate": 5.163471882071352e-07,
+      "loss": 0.3887,
+      "step": 3539
+    },
+    {
+      "epoch": 2.606774668630339,
+      "grad_norm": 0.32947883009910583,
+      "learning_rate": 5.144524950410074e-07,
+      "loss": 0.3717,
+      "step": 3540
+    },
+    {
+      "epoch": 2.6075110456553756,
+      "grad_norm": 0.3143480718135834,
+      "learning_rate": 5.125610959254213e-07,
+      "loss": 0.3805,
+      "step": 3541
+    },
+    {
+      "epoch": 2.6082474226804124,
+      "grad_norm": 0.32195669412612915,
+      "learning_rate": 5.10672992249358e-07,
+      "loss": 0.3862,
+      "step": 3542
+    },
+    {
+      "epoch": 2.6089837997054492,
+      "grad_norm": 0.3320513367652893,
+      "learning_rate": 5.087881853993876e-07,
+      "loss": 0.3418,
+      "step": 3543
+    },
+    {
+      "epoch": 2.609720176730486,
+      "grad_norm": 0.3392578959465027,
+      "learning_rate": 5.069066767596542e-07,
+      "loss": 0.3469,
+      "step": 3544
+    },
+    {
+      "epoch": 2.610456553755523,
+      "grad_norm": 0.3689731955528259,
+      "learning_rate": 5.0502846771188e-07,
+      "loss": 0.3818,
+      "step": 3545
+    },
+    {
+      "epoch": 2.6111929307805597,
+      "grad_norm": 0.33059579133987427,
+      "learning_rate": 5.031535596353665e-07,
+      "loss": 0.3901,
+      "step": 3546
+    },
+    {
+      "epoch": 2.6119293078055965,
+      "grad_norm": 0.3137247860431671,
+      "learning_rate": 5.012819539069885e-07,
+      "loss": 0.3799,
+      "step": 3547
+    },
+    {
+      "epoch": 2.6126656848306333,
+      "grad_norm": 0.3432486057281494,
+      "learning_rate": 4.994136519011966e-07,
+      "loss": 0.3573,
+      "step": 3548
+    },
+    {
+      "epoch": 2.61340206185567,
+      "grad_norm": 0.33022403717041016,
+      "learning_rate": 4.975486549900177e-07,
+      "loss": 0.369,
+      "step": 3549
+    },
+    {
+      "epoch": 2.614138438880707,
+      "grad_norm": 0.33566609025001526,
+      "learning_rate": 4.956869645430451e-07,
+      "loss": 0.3732,
+      "step": 3550
+    },
+    {
+      "epoch": 2.6148748159057438,
+      "grad_norm": 0.3244359791278839,
+      "learning_rate": 4.938285819274507e-07,
+      "loss": 0.3276,
+      "step": 3551
+    },
+    {
+      "epoch": 2.6156111929307806,
+      "grad_norm": 0.31504350900650024,
+      "learning_rate": 4.919735085079746e-07,
+      "loss": 0.3572,
+      "step": 3552
+    },
+    {
+      "epoch": 2.6163475699558174,
+      "grad_norm": 0.32959744334220886,
+      "learning_rate": 4.901217456469248e-07,
+      "loss": 0.4114,
+      "step": 3553
+    },
+    {
+      "epoch": 2.6170839469808542,
+      "grad_norm": 0.30462411046028137,
+      "learning_rate": 4.882732947041818e-07,
+      "loss": 0.3974,
+      "step": 3554
+    },
+    {
+      "epoch": 2.617820324005891,
+      "grad_norm": 0.33037570118904114,
+      "learning_rate": 4.86428157037192e-07,
+      "loss": 0.3725,
+      "step": 3555
+    },
+    {
+      "epoch": 2.618556701030928,
+      "grad_norm": 0.3179605007171631,
+      "learning_rate": 4.845863340009671e-07,
+      "loss": 0.4015,
+      "step": 3556
+    },
+    {
+      "epoch": 2.6192930780559647,
+      "grad_norm": 0.332682341337204,
+      "learning_rate": 4.827478269480895e-07,
+      "loss": 0.3676,
+      "step": 3557
+    },
+    {
+      "epoch": 2.6200294550810015,
+      "grad_norm": 0.33066973090171814,
+      "learning_rate": 4.809126372286999e-07,
+      "loss": 0.3644,
+      "step": 3558
+    },
+    {
+      "epoch": 2.6207658321060383,
+      "grad_norm": 0.34845325350761414,
+      "learning_rate": 4.790807661905067e-07,
+      "loss": 0.3876,
+      "step": 3559
+    },
+    {
+      "epoch": 2.621502209131075,
+      "grad_norm": 0.3088925778865814,
+      "learning_rate": 4.772522151787822e-07,
+      "loss": 0.3701,
+      "step": 3560
+    },
+    {
+      "epoch": 2.622238586156112,
+      "grad_norm": 0.3307948708534241,
+      "learning_rate": 4.7542698553635856e-07,
+      "loss": 0.3932,
+      "step": 3561
+    },
+    {
+      "epoch": 2.6229749631811488,
+      "grad_norm": 0.33199846744537354,
+      "learning_rate": 4.7360507860362723e-07,
+      "loss": 0.362,
+      "step": 3562
+    },
+    {
+      "epoch": 2.6237113402061856,
+      "grad_norm": 0.3131435215473175,
+      "learning_rate": 4.7178649571854473e-07,
+      "loss": 0.3647,
+      "step": 3563
+    },
+    {
+      "epoch": 2.6244477172312224,
+      "grad_norm": 0.3339531719684601,
+      "learning_rate": 4.699712382166216e-07,
+      "loss": 0.3888,
+      "step": 3564
+    },
+    {
+      "epoch": 2.625184094256259,
+      "grad_norm": 0.3675123155117035,
+      "learning_rate": 4.6815930743092765e-07,
+      "loss": 0.3772,
+      "step": 3565
+    },
+    {
+      "epoch": 2.625920471281296,
+      "grad_norm": 0.34725135564804077,
+      "learning_rate": 4.663507046920929e-07,
+      "loss": 0.349,
+      "step": 3566
+    },
+    {
+      "epoch": 2.626656848306333,
+      "grad_norm": 0.31065019965171814,
+      "learning_rate": 4.6454543132829653e-07,
+      "loss": 0.3587,
+      "step": 3567
+    },
+    {
+      "epoch": 2.6273932253313697,
+      "grad_norm": 0.32565373182296753,
+      "learning_rate": 4.627434886652793e-07,
+      "loss": 0.3668,
+      "step": 3568
+    },
+    {
+      "epoch": 2.6281296023564065,
+      "grad_norm": 0.33230826258659363,
+      "learning_rate": 4.6094487802633315e-07,
+      "loss": 0.3763,
+      "step": 3569
+    },
+    {
+      "epoch": 2.6288659793814433,
+      "grad_norm": 0.31987035274505615,
+      "learning_rate": 4.591496007323021e-07,
+      "loss": 0.38,
+      "step": 3570
+    },
+    {
+      "epoch": 2.62960235640648,
+      "grad_norm": 0.31742241978645325,
+      "learning_rate": 4.573576581015854e-07,
+      "loss": 0.3474,
+      "step": 3571
+    },
+    {
+      "epoch": 2.630338733431517,
+      "grad_norm": 0.31509506702423096,
+      "learning_rate": 4.55569051450131e-07,
+      "loss": 0.3646,
+      "step": 3572
+    },
+    {
+      "epoch": 2.6310751104565537,
+      "grad_norm": 0.3257356584072113,
+      "learning_rate": 4.537837820914359e-07,
+      "loss": 0.3668,
+      "step": 3573
+    },
+    {
+      "epoch": 2.6318114874815906,
+      "grad_norm": 0.32111799716949463,
+      "learning_rate": 4.520018513365515e-07,
+      "loss": 0.3933,
+      "step": 3574
+    },
+    {
+      "epoch": 2.6325478645066274,
+      "grad_norm": 0.3151605725288391,
+      "learning_rate": 4.5022326049406986e-07,
+      "loss": 0.3676,
+      "step": 3575
+    },
+    {
+      "epoch": 2.633284241531664,
+      "grad_norm": 0.3598038852214813,
+      "learning_rate": 4.484480108701372e-07,
+      "loss": 0.4136,
+      "step": 3576
+    },
+    {
+      "epoch": 2.634020618556701,
+      "grad_norm": 0.2969145178794861,
+      "learning_rate": 4.4667610376844197e-07,
+      "loss": 0.3821,
+      "step": 3577
+    },
+    {
+      "epoch": 2.634756995581738,
+      "grad_norm": 0.33178281784057617,
+      "learning_rate": 4.449075404902187e-07,
+      "loss": 0.3559,
+      "step": 3578
+    },
+    {
+      "epoch": 2.6354933726067746,
+      "grad_norm": 0.3466475307941437,
+      "learning_rate": 4.4314232233424845e-07,
+      "loss": 0.3827,
+      "step": 3579
+    },
+    {
+      "epoch": 2.6362297496318114,
+      "grad_norm": 0.3220832645893097,
+      "learning_rate": 4.413804505968533e-07,
+      "loss": 0.367,
+      "step": 3580
+    },
+    {
+      "epoch": 2.6369661266568483,
+      "grad_norm": 0.32306790351867676,
+      "learning_rate": 4.3962192657189707e-07,
+      "loss": 0.3961,
+      "step": 3581
+    },
+    {
+      "epoch": 2.637702503681885,
+      "grad_norm": 0.32913732528686523,
+      "learning_rate": 4.378667515507895e-07,
+      "loss": 0.3701,
+      "step": 3582
+    },
+    {
+      "epoch": 2.638438880706922,
+      "grad_norm": 0.3247510492801666,
+      "learning_rate": 4.361149268224779e-07,
+      "loss": 0.3603,
+      "step": 3583
+    },
+    {
+      "epoch": 2.6391752577319587,
+      "grad_norm": 0.36141979694366455,
+      "learning_rate": 4.34366453673446e-07,
+      "loss": 0.3905,
+      "step": 3584
+    },
+    {
+      "epoch": 2.6399116347569955,
+      "grad_norm": 0.31596043705940247,
+      "learning_rate": 4.326213333877227e-07,
+      "loss": 0.3527,
+      "step": 3585
+    },
+    {
+      "epoch": 2.6406480117820323,
+      "grad_norm": 0.3106141984462738,
+      "learning_rate": 4.308795672468713e-07,
+      "loss": 0.3539,
+      "step": 3586
+    },
+    {
+      "epoch": 2.641384388807069,
+      "grad_norm": 0.30769434571266174,
+      "learning_rate": 4.291411565299902e-07,
+      "loss": 0.3495,
+      "step": 3587
+    },
+    {
+      "epoch": 2.642120765832106,
+      "grad_norm": 0.30311083793640137,
+      "learning_rate": 4.2740610251371826e-07,
+      "loss": 0.3988,
+      "step": 3588
+    },
+    {
+      "epoch": 2.642857142857143,
+      "grad_norm": 0.3294108510017395,
+      "learning_rate": 4.256744064722246e-07,
+      "loss": 0.3609,
+      "step": 3589
+    },
+    {
+      "epoch": 2.6435935198821796,
+      "grad_norm": 0.3281387686729431,
+      "learning_rate": 4.2394606967721683e-07,
+      "loss": 0.3651,
+      "step": 3590
+    },
+    {
+      "epoch": 2.6443298969072164,
+      "grad_norm": 0.33501261472702026,
+      "learning_rate": 4.222210933979326e-07,
+      "loss": 0.3573,
+      "step": 3591
+    },
+    {
+      "epoch": 2.6450662739322532,
+      "grad_norm": 0.35645949840545654,
+      "learning_rate": 4.204994789011396e-07,
+      "loss": 0.4035,
+      "step": 3592
+    },
+    {
+      "epoch": 2.64580265095729,
+      "grad_norm": 0.3307779133319855,
+      "learning_rate": 4.187812274511427e-07,
+      "loss": 0.3763,
+      "step": 3593
+    },
+    {
+      "epoch": 2.646539027982327,
+      "grad_norm": 0.3367134630680084,
+      "learning_rate": 4.17066340309773e-07,
+      "loss": 0.3359,
+      "step": 3594
+    },
+    {
+      "epoch": 2.6472754050073637,
+      "grad_norm": 0.33751773834228516,
+      "learning_rate": 4.153548187363904e-07,
+      "loss": 0.3989,
+      "step": 3595
+    },
+    {
+      "epoch": 2.6480117820324005,
+      "grad_norm": 0.29782766103744507,
+      "learning_rate": 4.1364666398788613e-07,
+      "loss": 0.3574,
+      "step": 3596
+    },
+    {
+      "epoch": 2.6487481590574373,
+      "grad_norm": 0.3283192217350006,
+      "learning_rate": 4.1194187731867783e-07,
+      "loss": 0.3527,
+      "step": 3597
+    },
+    {
+      "epoch": 2.649484536082474,
+      "grad_norm": 0.32778191566467285,
+      "learning_rate": 4.102404599807075e-07,
+      "loss": 0.4059,
+      "step": 3598
+    },
+    {
+      "epoch": 2.650220913107511,
+      "grad_norm": 0.35166746377944946,
+      "learning_rate": 4.0854241322344665e-07,
+      "loss": 0.3882,
+      "step": 3599
+    },
+    {
+      "epoch": 2.6509572901325478,
+      "grad_norm": 0.34499865770339966,
+      "learning_rate": 4.0684773829388737e-07,
+      "loss": 0.3763,
+      "step": 3600
+    },
+    {
+      "epoch": 2.6516936671575846,
+      "grad_norm": 0.3195962607860565,
+      "learning_rate": 4.0515643643655014e-07,
+      "loss": 0.3616,
+      "step": 3601
+    },
+    {
+      "epoch": 2.6524300441826214,
+      "grad_norm": 0.34196287393569946,
+      "learning_rate": 4.034685088934737e-07,
+      "loss": 0.3798,
+      "step": 3602
+    },
+    {
+      "epoch": 2.653166421207658,
+      "grad_norm": 0.32536935806274414,
+      "learning_rate": 4.0178395690422143e-07,
+      "loss": 0.3623,
+      "step": 3603
+    },
+    {
+      "epoch": 2.653902798232695,
+      "grad_norm": 0.3281365633010864,
+      "learning_rate": 4.001027817058789e-07,
+      "loss": 0.3523,
+      "step": 3604
+    },
+    {
+      "epoch": 2.654639175257732,
+      "grad_norm": 0.32271289825439453,
+      "learning_rate": 3.9842498453304955e-07,
+      "loss": 0.3578,
+      "step": 3605
+    },
+    {
+      "epoch": 2.6553755522827687,
+      "grad_norm": 0.3313661515712738,
+      "learning_rate": 3.9675056661785563e-07,
+      "loss": 0.3415,
+      "step": 3606
+    },
+    {
+      "epoch": 2.6561119293078055,
+      "grad_norm": 0.31950780749320984,
+      "learning_rate": 3.950795291899412e-07,
+      "loss": 0.3595,
+      "step": 3607
+    },
+    {
+      "epoch": 2.6568483063328423,
+      "grad_norm": 0.34532180428504944,
+      "learning_rate": 3.934118734764647e-07,
+      "loss": 0.3879,
+      "step": 3608
+    },
+    {
+      "epoch": 2.657584683357879,
+      "grad_norm": 0.3165544271469116,
+      "learning_rate": 3.9174760070210204e-07,
+      "loss": 0.3682,
+      "step": 3609
+    },
+    {
+      "epoch": 2.658321060382916,
+      "grad_norm": 0.33208510279655457,
+      "learning_rate": 3.9008671208904503e-07,
+      "loss": 0.3667,
+      "step": 3610
+    },
+    {
+      "epoch": 2.6590574374079528,
+      "grad_norm": 0.31021976470947266,
+      "learning_rate": 3.8842920885699906e-07,
+      "loss": 0.3645,
+      "step": 3611
+    },
+    {
+      "epoch": 2.6597938144329896,
+      "grad_norm": 0.32414910197257996,
+      "learning_rate": 3.8677509222318557e-07,
+      "loss": 0.3607,
+      "step": 3612
+    },
+    {
+      "epoch": 2.6605301914580264,
+      "grad_norm": 0.33595502376556396,
+      "learning_rate": 3.8512436340233826e-07,
+      "loss": 0.3728,
+      "step": 3613
+    },
+    {
+      "epoch": 2.661266568483063,
+      "grad_norm": 0.31958842277526855,
+      "learning_rate": 3.8347702360670036e-07,
+      "loss": 0.3722,
+      "step": 3614
+    },
+    {
+      "epoch": 2.6620029455081,
+      "grad_norm": 0.3532390594482422,
+      "learning_rate": 3.8183307404603074e-07,
+      "loss": 0.3657,
+      "step": 3615
+    },
+    {
+      "epoch": 2.662739322533137,
+      "grad_norm": 0.3612724840641022,
+      "learning_rate": 3.8019251592759656e-07,
+      "loss": 0.388,
+      "step": 3616
+    },
+    {
+      "epoch": 2.6634756995581736,
+      "grad_norm": 0.3217339515686035,
+      "learning_rate": 3.785553504561712e-07,
+      "loss": 0.3428,
+      "step": 3617
+    },
+    {
+      "epoch": 2.6642120765832105,
+      "grad_norm": 0.2959875762462616,
+      "learning_rate": 3.769215788340419e-07,
+      "loss": 0.3665,
+      "step": 3618
+    },
+    {
+      "epoch": 2.6649484536082473,
+      "grad_norm": 0.3248257637023926,
+      "learning_rate": 3.752912022610006e-07,
+      "loss": 0.3725,
+      "step": 3619
+    },
+    {
+      "epoch": 2.665684830633284,
+      "grad_norm": 0.3298156261444092,
+      "learning_rate": 3.736642219343456e-07,
+      "loss": 0.3673,
+      "step": 3620
+    },
+    {
+      "epoch": 2.666421207658321,
+      "grad_norm": 0.32262980937957764,
+      "learning_rate": 3.720406390488834e-07,
+      "loss": 0.3928,
+      "step": 3621
+    },
+    {
+      "epoch": 2.6671575846833577,
+      "grad_norm": 0.30602526664733887,
+      "learning_rate": 3.7042045479692424e-07,
+      "loss": 0.3647,
+      "step": 3622
+    },
+    {
+      "epoch": 2.6678939617083945,
+      "grad_norm": 0.3311411142349243,
+      "learning_rate": 3.6880367036828124e-07,
+      "loss": 0.3779,
+      "step": 3623
+    },
+    {
+      "epoch": 2.6686303387334314,
+      "grad_norm": 0.32879799604415894,
+      "learning_rate": 3.671902869502736e-07,
+      "loss": 0.3925,
+      "step": 3624
+    },
+    {
+      "epoch": 2.669366715758468,
+      "grad_norm": 0.3125240206718445,
+      "learning_rate": 3.6558030572772075e-07,
+      "loss": 0.3984,
+      "step": 3625
+    },
+    {
+      "epoch": 2.670103092783505,
+      "grad_norm": 0.3394649624824524,
+      "learning_rate": 3.639737278829436e-07,
+      "loss": 0.3657,
+      "step": 3626
+    },
+    {
+      "epoch": 2.670839469808542,
+      "grad_norm": 0.34995949268341064,
+      "learning_rate": 3.623705545957651e-07,
+      "loss": 0.4049,
+      "step": 3627
+    },
+    {
+      "epoch": 2.6715758468335786,
+      "grad_norm": 0.33862966299057007,
+      "learning_rate": 3.607707870435062e-07,
+      "loss": 0.3941,
+      "step": 3628
+    },
+    {
+      "epoch": 2.672312223858616,
+      "grad_norm": 0.3330628573894501,
+      "learning_rate": 3.5917442640098997e-07,
+      "loss": 0.3516,
+      "step": 3629
+    },
+    {
+      "epoch": 2.6730486008836527,
+      "grad_norm": 0.30992215871810913,
+      "learning_rate": 3.575814738405331e-07,
+      "loss": 0.3642,
+      "step": 3630
+    },
+    {
+      "epoch": 2.6737849779086895,
+      "grad_norm": 0.3173280954360962,
+      "learning_rate": 3.559919305319526e-07,
+      "loss": 0.3858,
+      "step": 3631
+    },
+    {
+      "epoch": 2.6745213549337263,
+      "grad_norm": 0.32444649934768677,
+      "learning_rate": 3.544057976425619e-07,
+      "loss": 0.3647,
+      "step": 3632
+    },
+    {
+      "epoch": 2.675257731958763,
+      "grad_norm": 0.2939029932022095,
+      "learning_rate": 3.528230763371687e-07,
+      "loss": 0.359,
+      "step": 3633
+    },
+    {
+      "epoch": 2.6759941089838,
+      "grad_norm": 0.3513110876083374,
+      "learning_rate": 3.51243767778075e-07,
+      "loss": 0.3474,
+      "step": 3634
+    },
+    {
+      "epoch": 2.676730486008837,
+      "grad_norm": 0.3469097912311554,
+      "learning_rate": 3.49667873125078e-07,
+      "loss": 0.3416,
+      "step": 3635
+    },
+    {
+      "epoch": 2.6774668630338736,
+      "grad_norm": 0.3251095116138458,
+      "learning_rate": 3.480953935354658e-07,
+      "loss": 0.3585,
+      "step": 3636
+    },
+    {
+      "epoch": 2.6782032400589104,
+      "grad_norm": 0.31582942605018616,
+      "learning_rate": 3.4652633016402205e-07,
+      "loss": 0.3648,
+      "step": 3637
+    },
+    {
+      "epoch": 2.6789396170839472,
+      "grad_norm": 0.31770941615104675,
+      "learning_rate": 3.449606841630182e-07,
+      "loss": 0.3846,
+      "step": 3638
+    },
+    {
+      "epoch": 2.679675994108984,
+      "grad_norm": 0.3245222866535187,
+      "learning_rate": 3.433984566822163e-07,
+      "loss": 0.3573,
+      "step": 3639
+    },
+    {
+      "epoch": 2.680412371134021,
+      "grad_norm": 0.33965256810188293,
+      "learning_rate": 3.4183964886887135e-07,
+      "loss": 0.3547,
+      "step": 3640
+    },
+    {
+      "epoch": 2.6811487481590577,
+      "grad_norm": 0.34220483899116516,
+      "learning_rate": 3.4028426186772435e-07,
+      "loss": 0.3624,
+      "step": 3641
+    },
+    {
+      "epoch": 2.6818851251840945,
+      "grad_norm": 0.32621511816978455,
+      "learning_rate": 3.387322968210022e-07,
+      "loss": 0.3795,
+      "step": 3642
+    },
+    {
+      "epoch": 2.6826215022091313,
+      "grad_norm": 0.3251863121986389,
+      "learning_rate": 3.3718375486842314e-07,
+      "loss": 0.3535,
+      "step": 3643
+    },
+    {
+      "epoch": 2.683357879234168,
+      "grad_norm": 0.32561933994293213,
+      "learning_rate": 3.3563863714718927e-07,
+      "loss": 0.3637,
+      "step": 3644
+    },
+    {
+      "epoch": 2.684094256259205,
+      "grad_norm": 0.3464414179325104,
+      "learning_rate": 3.340969447919873e-07,
+      "loss": 0.3708,
+      "step": 3645
+    },
+    {
+      "epoch": 2.6848306332842418,
+      "grad_norm": 0.3460819721221924,
+      "learning_rate": 3.3255867893499105e-07,
+      "loss": 0.366,
+      "step": 3646
+    },
+    {
+      "epoch": 2.6855670103092786,
+      "grad_norm": 0.36896535754203796,
+      "learning_rate": 3.3102384070585523e-07,
+      "loss": 0.3482,
+      "step": 3647
+    },
+    {
+      "epoch": 2.6863033873343154,
+      "grad_norm": 0.3020630478858948,
+      "learning_rate": 3.2949243123171994e-07,
+      "loss": 0.3855,
+      "step": 3648
+    },
+    {
+      "epoch": 2.687039764359352,
+      "grad_norm": 0.3055335283279419,
+      "learning_rate": 3.279644516372049e-07,
+      "loss": 0.3841,
+      "step": 3649
+    },
+    {
+      "epoch": 2.687776141384389,
+      "grad_norm": 0.3210819363594055,
+      "learning_rate": 3.264399030444132e-07,
+      "loss": 0.3661,
+      "step": 3650
+    },
+    {
+      "epoch": 2.688512518409426,
+      "grad_norm": 0.3282409608364105,
+      "learning_rate": 3.2491878657292643e-07,
+      "loss": 0.3595,
+      "step": 3651
+    },
+    {
+      "epoch": 2.6892488954344627,
+      "grad_norm": 0.3092752695083618,
+      "learning_rate": 3.2340110333980656e-07,
+      "loss": 0.3499,
+      "step": 3652
+    },
+    {
+      "epoch": 2.6899852724594995,
+      "grad_norm": 0.31985411047935486,
+      "learning_rate": 3.218868544595938e-07,
+      "loss": 0.3618,
+      "step": 3653
+    },
+    {
+      "epoch": 2.6907216494845363,
+      "grad_norm": 0.33014747500419617,
+      "learning_rate": 3.20376041044308e-07,
+      "loss": 0.3551,
+      "step": 3654
+    },
+    {
+      "epoch": 2.691458026509573,
+      "grad_norm": 0.307456910610199,
+      "learning_rate": 3.18868664203445e-07,
+      "loss": 0.3722,
+      "step": 3655
+    },
+    {
+      "epoch": 2.69219440353461,
+      "grad_norm": 0.35932809114456177,
+      "learning_rate": 3.1736472504397485e-07,
+      "loss": 0.3935,
+      "step": 3656
+    },
+    {
+      "epoch": 2.6929307805596467,
+      "grad_norm": 0.3306938409805298,
+      "learning_rate": 3.1586422467034695e-07,
+      "loss": 0.379,
+      "step": 3657
+    },
+    {
+      "epoch": 2.6936671575846836,
+      "grad_norm": 0.3393386900424957,
+      "learning_rate": 3.143671641844831e-07,
+      "loss": 0.4004,
+      "step": 3658
+    },
+    {
+      "epoch": 2.6944035346097204,
+      "grad_norm": 0.3137757480144501,
+      "learning_rate": 3.128735446857784e-07,
+      "loss": 0.3594,
+      "step": 3659
+    },
+    {
+      "epoch": 2.695139911634757,
+      "grad_norm": 0.33764535188674927,
+      "learning_rate": 3.1138336727110307e-07,
+      "loss": 0.3994,
+      "step": 3660
+    },
+    {
+      "epoch": 2.695876288659794,
+      "grad_norm": 0.32632213830947876,
+      "learning_rate": 3.098966330347969e-07,
+      "loss": 0.3772,
+      "step": 3661
+    },
+    {
+      "epoch": 2.696612665684831,
+      "grad_norm": 0.3276118040084839,
+      "learning_rate": 3.0841334306867367e-07,
+      "loss": 0.3561,
+      "step": 3662
+    },
+    {
+      "epoch": 2.6973490427098676,
+      "grad_norm": 0.3348173201084137,
+      "learning_rate": 3.06933498462017e-07,
+      "loss": 0.357,
+      "step": 3663
+    },
+    {
+      "epoch": 2.6980854197349045,
+      "grad_norm": 0.372728168964386,
+      "learning_rate": 3.0545710030157824e-07,
+      "loss": 0.3482,
+      "step": 3664
+    },
+    {
+      "epoch": 2.6988217967599413,
+      "grad_norm": 0.34344926476478577,
+      "learning_rate": 3.039841496715823e-07,
+      "loss": 0.39,
+      "step": 3665
+    },
+    {
+      "epoch": 2.699558173784978,
+      "grad_norm": 0.3021678924560547,
+      "learning_rate": 3.0251464765371774e-07,
+      "loss": 0.3888,
+      "step": 3666
+    },
+    {
+      "epoch": 2.700294550810015,
+      "grad_norm": 0.3373548984527588,
+      "learning_rate": 3.010485953271425e-07,
+      "loss": 0.3803,
+      "step": 3667
+    },
+    {
+      "epoch": 2.7010309278350517,
+      "grad_norm": 0.31935036182403564,
+      "learning_rate": 2.9958599376848194e-07,
+      "loss": 0.3878,
+      "step": 3668
+    },
+    {
+      "epoch": 2.7017673048600885,
+      "grad_norm": 0.3091143071651459,
+      "learning_rate": 2.9812684405182536e-07,
+      "loss": 0.336,
+      "step": 3669
+    },
+    {
+      "epoch": 2.7025036818851254,
+      "grad_norm": 0.325328528881073,
+      "learning_rate": 2.9667114724872937e-07,
+      "loss": 0.4175,
+      "step": 3670
+    },
+    {
+      "epoch": 2.703240058910162,
+      "grad_norm": 0.31764811277389526,
+      "learning_rate": 2.9521890442821276e-07,
+      "loss": 0.3726,
+      "step": 3671
+    },
+    {
+      "epoch": 2.703976435935199,
+      "grad_norm": 0.3514196276664734,
+      "learning_rate": 2.9377011665675913e-07,
+      "loss": 0.3768,
+      "step": 3672
+    },
+    {
+      "epoch": 2.704712812960236,
+      "grad_norm": 0.3268696069717407,
+      "learning_rate": 2.923247849983146e-07,
+      "loss": 0.3878,
+      "step": 3673
+    },
+    {
+      "epoch": 2.7054491899852726,
+      "grad_norm": 0.3018622100353241,
+      "learning_rate": 2.908829105142874e-07,
+      "loss": 0.3717,
+      "step": 3674
+    },
+    {
+      "epoch": 2.7061855670103094,
+      "grad_norm": 0.28456223011016846,
+      "learning_rate": 2.89444494263546e-07,
+      "loss": 0.3918,
+      "step": 3675
+    },
+    {
+      "epoch": 2.7069219440353463,
+      "grad_norm": 0.35831522941589355,
+      "learning_rate": 2.8800953730242e-07,
+      "loss": 0.3505,
+      "step": 3676
+    },
+    {
+      "epoch": 2.707658321060383,
+      "grad_norm": 0.33213359117507935,
+      "learning_rate": 2.865780406846985e-07,
+      "loss": 0.3614,
+      "step": 3677
+    },
+    {
+      "epoch": 2.70839469808542,
+      "grad_norm": 0.3209593594074249,
+      "learning_rate": 2.85150005461628e-07,
+      "loss": 0.3597,
+      "step": 3678
+    },
+    {
+      "epoch": 2.7091310751104567,
+      "grad_norm": 0.3153224587440491,
+      "learning_rate": 2.8372543268191723e-07,
+      "loss": 0.3796,
+      "step": 3679
+    },
+    {
+      "epoch": 2.7098674521354935,
+      "grad_norm": 0.3322901129722595,
+      "learning_rate": 2.823043233917272e-07,
+      "loss": 0.3776,
+      "step": 3680
+    },
+    {
+      "epoch": 2.7106038291605303,
+      "grad_norm": 0.35406193137168884,
+      "learning_rate": 2.8088667863467754e-07,
+      "loss": 0.3886,
+      "step": 3681
+    },
+    {
+      "epoch": 2.711340206185567,
+      "grad_norm": 0.3339800536632538,
+      "learning_rate": 2.794724994518455e-07,
+      "loss": 0.3846,
+      "step": 3682
+    },
+    {
+      "epoch": 2.712076583210604,
+      "grad_norm": 0.33495986461639404,
+      "learning_rate": 2.7806178688175977e-07,
+      "loss": 0.3788,
+      "step": 3683
+    },
+    {
+      "epoch": 2.712812960235641,
+      "grad_norm": 0.3515438437461853,
+      "learning_rate": 2.7665454196040665e-07,
+      "loss": 0.3865,
+      "step": 3684
+    },
+    {
+      "epoch": 2.7135493372606776,
+      "grad_norm": 0.320209801197052,
+      "learning_rate": 2.752507657212228e-07,
+      "loss": 0.3725,
+      "step": 3685
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.3267118036746979,
+      "learning_rate": 2.738504591950991e-07,
+      "loss": 0.3785,
+      "step": 3686
+    },
+    {
+      "epoch": 2.7150220913107512,
+      "grad_norm": 0.34039753675460815,
+      "learning_rate": 2.724536234103792e-07,
+      "loss": 0.3622,
+      "step": 3687
+    },
+    {
+      "epoch": 2.715758468335788,
+      "grad_norm": 0.33974501490592957,
+      "learning_rate": 2.710602593928574e-07,
+      "loss": 0.3561,
+      "step": 3688
+    },
+    {
+      "epoch": 2.716494845360825,
+      "grad_norm": 0.3247467279434204,
+      "learning_rate": 2.6967036816577643e-07,
+      "loss": 0.3628,
+      "step": 3689
+    },
+    {
+      "epoch": 2.7172312223858617,
+      "grad_norm": 0.3339967131614685,
+      "learning_rate": 2.6828395074983195e-07,
+      "loss": 0.3635,
+      "step": 3690
+    },
+    {
+      "epoch": 2.7179675994108985,
+      "grad_norm": 0.31294745206832886,
+      "learning_rate": 2.6690100816316675e-07,
+      "loss": 0.3706,
+      "step": 3691
+    },
+    {
+      "epoch": 2.7187039764359353,
+      "grad_norm": 0.3254539370536804,
+      "learning_rate": 2.655215414213719e-07,
+      "loss": 0.3796,
+      "step": 3692
+    },
+    {
+      "epoch": 2.719440353460972,
+      "grad_norm": 0.32704704999923706,
+      "learning_rate": 2.6414555153748635e-07,
+      "loss": 0.3642,
+      "step": 3693
+    },
+    {
+      "epoch": 2.720176730486009,
+      "grad_norm": 0.31729063391685486,
+      "learning_rate": 2.627730395219941e-07,
+      "loss": 0.3979,
+      "step": 3694
+    },
+    {
+      "epoch": 2.7209131075110458,
+      "grad_norm": 0.3445037603378296,
+      "learning_rate": 2.6140400638282826e-07,
+      "loss": 0.3648,
+      "step": 3695
+    },
+    {
+      "epoch": 2.7216494845360826,
+      "grad_norm": 0.3214196562767029,
+      "learning_rate": 2.6003845312536526e-07,
+      "loss": 0.397,
+      "step": 3696
+    },
+    {
+      "epoch": 2.7223858615611194,
+      "grad_norm": 0.3424557149410248,
+      "learning_rate": 2.5867638075242454e-07,
+      "loss": 0.3801,
+      "step": 3697
+    },
+    {
+      "epoch": 2.723122238586156,
+      "grad_norm": 0.36900395154953003,
+      "learning_rate": 2.573177902642726e-07,
+      "loss": 0.3874,
+      "step": 3698
+    },
+    {
+      "epoch": 2.723858615611193,
+      "grad_norm": 0.33150961995124817,
+      "learning_rate": 2.5596268265861646e-07,
+      "loss": 0.3587,
+      "step": 3699
+    },
+    {
+      "epoch": 2.72459499263623,
+      "grad_norm": 0.35658320784568787,
+      "learning_rate": 2.5461105893060667e-07,
+      "loss": 0.3807,
+      "step": 3700
+    },
+    {
+      "epoch": 2.7253313696612667,
+      "grad_norm": 0.3134189248085022,
+      "learning_rate": 2.532629200728343e-07,
+      "loss": 0.3832,
+      "step": 3701
+    },
+    {
+      "epoch": 2.7260677466863035,
+      "grad_norm": 0.3331991136074066,
+      "learning_rate": 2.5191826707533173e-07,
+      "loss": 0.3653,
+      "step": 3702
+    },
+    {
+      "epoch": 2.7268041237113403,
+      "grad_norm": 0.3472428619861603,
+      "learning_rate": 2.505771009255714e-07,
+      "loss": 0.4057,
+      "step": 3703
+    },
+    {
+      "epoch": 2.727540500736377,
+      "grad_norm": 0.36920398473739624,
+      "learning_rate": 2.492394226084666e-07,
+      "loss": 0.396,
+      "step": 3704
+    },
+    {
+      "epoch": 2.728276877761414,
+      "grad_norm": 0.3479515314102173,
+      "learning_rate": 2.479052331063658e-07,
+      "loss": 0.3834,
+      "step": 3705
+    },
+    {
+      "epoch": 2.7290132547864507,
+      "grad_norm": 0.3214065730571747,
+      "learning_rate": 2.465745333990588e-07,
+      "loss": 0.3581,
+      "step": 3706
+    },
+    {
+      "epoch": 2.7297496318114876,
+      "grad_norm": 0.29844558238983154,
+      "learning_rate": 2.4524732446377154e-07,
+      "loss": 0.3715,
+      "step": 3707
+    },
+    {
+      "epoch": 2.7304860088365244,
+      "grad_norm": 0.2953944504261017,
+      "learning_rate": 2.439236072751644e-07,
+      "loss": 0.3972,
+      "step": 3708
+    },
+    {
+      "epoch": 2.731222385861561,
+      "grad_norm": 0.322376012802124,
+      "learning_rate": 2.426033828053381e-07,
+      "loss": 0.3785,
+      "step": 3709
+    },
+    {
+      "epoch": 2.731958762886598,
+      "grad_norm": 0.32461515069007874,
+      "learning_rate": 2.4128665202382327e-07,
+      "loss": 0.3594,
+      "step": 3710
+    },
+    {
+      "epoch": 2.732695139911635,
+      "grad_norm": 0.3040623664855957,
+      "learning_rate": 2.3997341589758694e-07,
+      "loss": 0.3436,
+      "step": 3711
+    },
+    {
+      "epoch": 2.7334315169366716,
+      "grad_norm": 0.32611605525016785,
+      "learning_rate": 2.3866367539103206e-07,
+      "loss": 0.3653,
+      "step": 3712
+    },
+    {
+      "epoch": 2.7341678939617085,
+      "grad_norm": 0.32935771346092224,
+      "learning_rate": 2.37357431465991e-07,
+      "loss": 0.3611,
+      "step": 3713
+    },
+    {
+      "epoch": 2.7349042709867453,
+      "grad_norm": 0.3235970139503479,
+      "learning_rate": 2.3605468508172968e-07,
+      "loss": 0.3786,
+      "step": 3714
+    },
+    {
+      "epoch": 2.735640648011782,
+      "grad_norm": 0.33503854274749756,
+      "learning_rate": 2.3475543719494676e-07,
+      "loss": 0.3587,
+      "step": 3715
+    },
+    {
+      "epoch": 2.736377025036819,
+      "grad_norm": 0.324275404214859,
+      "learning_rate": 2.3345968875977008e-07,
+      "loss": 0.3482,
+      "step": 3716
+    },
+    {
+      "epoch": 2.7371134020618557,
+      "grad_norm": 0.33510270714759827,
+      "learning_rate": 2.3216744072775797e-07,
+      "loss": 0.3949,
+      "step": 3717
+    },
+    {
+      "epoch": 2.7378497790868925,
+      "grad_norm": 0.35356804728507996,
+      "learning_rate": 2.3087869404789854e-07,
+      "loss": 0.3781,
+      "step": 3718
+    },
+    {
+      "epoch": 2.7385861561119293,
+      "grad_norm": 0.3195352554321289,
+      "learning_rate": 2.2959344966660802e-07,
+      "loss": 0.3679,
+      "step": 3719
+    },
+    {
+      "epoch": 2.739322533136966,
+      "grad_norm": 0.3420256972312927,
+      "learning_rate": 2.2831170852773198e-07,
+      "loss": 0.3873,
+      "step": 3720
+    },
+    {
+      "epoch": 2.740058910162003,
+      "grad_norm": 0.3301747441291809,
+      "learning_rate": 2.2703347157254142e-07,
+      "loss": 0.3652,
+      "step": 3721
+    },
+    {
+      "epoch": 2.74079528718704,
+      "grad_norm": 0.3048621714115143,
+      "learning_rate": 2.2575873973973485e-07,
+      "loss": 0.3808,
+      "step": 3722
+    },
+    {
+      "epoch": 2.7415316642120766,
+      "grad_norm": 0.31739065051078796,
+      "learning_rate": 2.2448751396543788e-07,
+      "loss": 0.3668,
+      "step": 3723
+    },
+    {
+      "epoch": 2.7422680412371134,
+      "grad_norm": 0.3357614278793335,
+      "learning_rate": 2.2321979518319992e-07,
+      "loss": 0.4031,
+      "step": 3724
+    },
+    {
+      "epoch": 2.7430044182621502,
+      "grad_norm": 0.3367345929145813,
+      "learning_rate": 2.21955584323994e-07,
+      "loss": 0.3808,
+      "step": 3725
+    },
+    {
+      "epoch": 2.743740795287187,
+      "grad_norm": 0.3369562029838562,
+      "learning_rate": 2.2069488231622083e-07,
+      "loss": 0.3616,
+      "step": 3726
+    },
+    {
+      "epoch": 2.744477172312224,
+      "grad_norm": 0.30571091175079346,
+      "learning_rate": 2.1943769008569927e-07,
+      "loss": 0.3605,
+      "step": 3727
+    },
+    {
+      "epoch": 2.7452135493372607,
+      "grad_norm": 0.32707011699676514,
+      "learning_rate": 2.1818400855567523e-07,
+      "loss": 0.3632,
+      "step": 3728
+    },
+    {
+      "epoch": 2.7459499263622975,
+      "grad_norm": 0.3019167482852936,
+      "learning_rate": 2.1693383864681394e-07,
+      "loss": 0.3751,
+      "step": 3729
+    },
+    {
+      "epoch": 2.7466863033873343,
+      "grad_norm": 0.32666751742362976,
+      "learning_rate": 2.1568718127720155e-07,
+      "loss": 0.3796,
+      "step": 3730
+    },
+    {
+      "epoch": 2.747422680412371,
+      "grad_norm": 0.3546668291091919,
+      "learning_rate": 2.1444403736234686e-07,
+      "loss": 0.3746,
+      "step": 3731
+    },
+    {
+      "epoch": 2.748159057437408,
+      "grad_norm": 0.31046637892723083,
+      "learning_rate": 2.132044078151768e-07,
+      "loss": 0.3346,
+      "step": 3732
+    },
+    {
+      "epoch": 2.7488954344624448,
+      "grad_norm": 0.3377501666545868,
+      "learning_rate": 2.119682935460371e-07,
+      "loss": 0.3892,
+      "step": 3733
+    },
+    {
+      "epoch": 2.7496318114874816,
+      "grad_norm": 0.3156897723674774,
+      "learning_rate": 2.1073569546269434e-07,
+      "loss": 0.3731,
+      "step": 3734
+    },
+    {
+      "epoch": 2.7503681885125184,
+      "grad_norm": 0.33260810375213623,
+      "learning_rate": 2.095066144703295e-07,
+      "loss": 0.3553,
+      "step": 3735
+    },
+    {
+      "epoch": 2.7511045655375552,
+      "grad_norm": 0.351136177778244,
+      "learning_rate": 2.0828105147154275e-07,
+      "loss": 0.3744,
+      "step": 3736
+    },
+    {
+      "epoch": 2.751840942562592,
+      "grad_norm": 0.3310754597187042,
+      "learning_rate": 2.07059007366352e-07,
+      "loss": 0.3813,
+      "step": 3737
+    },
+    {
+      "epoch": 2.752577319587629,
+      "grad_norm": 0.3328613340854645,
+      "learning_rate": 2.0584048305218874e-07,
+      "loss": 0.3609,
+      "step": 3738
+    },
+    {
+      "epoch": 2.7533136966126657,
+      "grad_norm": 0.31846439838409424,
+      "learning_rate": 2.0462547942389942e-07,
+      "loss": 0.3687,
+      "step": 3739
+    },
+    {
+      "epoch": 2.7540500736377025,
+      "grad_norm": 0.29258909821510315,
+      "learning_rate": 2.03413997373747e-07,
+      "loss": 0.3837,
+      "step": 3740
+    },
+    {
+      "epoch": 2.7547864506627393,
+      "grad_norm": 0.31229260563850403,
+      "learning_rate": 2.0220603779140759e-07,
+      "loss": 0.3641,
+      "step": 3741
+    },
+    {
+      "epoch": 2.755522827687776,
+      "grad_norm": 0.31942838430404663,
+      "learning_rate": 2.0100160156396986e-07,
+      "loss": 0.3746,
+      "step": 3742
+    },
+    {
+      "epoch": 2.756259204712813,
+      "grad_norm": 0.3302987515926361,
+      "learning_rate": 1.998006895759347e-07,
+      "loss": 0.4101,
+      "step": 3743
+    },
+    {
+      "epoch": 2.7569955817378498,
+      "grad_norm": 0.32852962613105774,
+      "learning_rate": 1.98603302709216e-07,
+      "loss": 0.3463,
+      "step": 3744
+    },
+    {
+      "epoch": 2.7577319587628866,
+      "grad_norm": 0.34027835726737976,
+      "learning_rate": 1.9740944184313882e-07,
+      "loss": 0.3922,
+      "step": 3745
+    },
+    {
+      "epoch": 2.7584683357879234,
+      "grad_norm": 0.33216574788093567,
+      "learning_rate": 1.9621910785443843e-07,
+      "loss": 0.3951,
+      "step": 3746
+    },
+    {
+      "epoch": 2.75920471281296,
+      "grad_norm": 0.3349026143550873,
+      "learning_rate": 1.950323016172595e-07,
+      "loss": 0.3729,
+      "step": 3747
+    },
+    {
+      "epoch": 2.759941089837997,
+      "grad_norm": 0.31592321395874023,
+      "learning_rate": 1.9384902400315764e-07,
+      "loss": 0.3497,
+      "step": 3748
+    },
+    {
+      "epoch": 2.760677466863034,
+      "grad_norm": 0.33262524008750916,
+      "learning_rate": 1.926692758810955e-07,
+      "loss": 0.3656,
+      "step": 3749
+    },
+    {
+      "epoch": 2.7614138438880707,
+      "grad_norm": 0.318095326423645,
+      "learning_rate": 1.9149305811744456e-07,
+      "loss": 0.3944,
+      "step": 3750
+    },
+    {
+      "epoch": 2.7621502209131075,
+      "grad_norm": 0.32870233058929443,
+      "learning_rate": 1.9032037157598494e-07,
+      "loss": 0.3801,
+      "step": 3751
+    },
+    {
+      "epoch": 2.7628865979381443,
+      "grad_norm": 0.30096235871315,
+      "learning_rate": 1.891512171178994e-07,
+      "loss": 0.3364,
+      "step": 3752
+    },
+    {
+      "epoch": 2.763622974963181,
+      "grad_norm": 0.33072760701179504,
+      "learning_rate": 1.8798559560178174e-07,
+      "loss": 0.3867,
+      "step": 3753
+    },
+    {
+      "epoch": 2.764359351988218,
+      "grad_norm": 0.34106144309043884,
+      "learning_rate": 1.8682350788362892e-07,
+      "loss": 0.3824,
+      "step": 3754
+    },
+    {
+      "epoch": 2.7650957290132547,
+      "grad_norm": 0.3276694416999817,
+      "learning_rate": 1.856649548168421e-07,
+      "loss": 0.3463,
+      "step": 3755
+    },
+    {
+      "epoch": 2.7658321060382915,
+      "grad_norm": 0.3228393793106079,
+      "learning_rate": 1.8450993725222856e-07,
+      "loss": 0.3746,
+      "step": 3756
+    },
+    {
+      "epoch": 2.7665684830633284,
+      "grad_norm": 0.3612479567527771,
+      "learning_rate": 1.8335845603799806e-07,
+      "loss": 0.3635,
+      "step": 3757
+    },
+    {
+      "epoch": 2.767304860088365,
+      "grad_norm": 0.31587934494018555,
+      "learning_rate": 1.8221051201976315e-07,
+      "loss": 0.3564,
+      "step": 3758
+    },
+    {
+      "epoch": 2.768041237113402,
+      "grad_norm": 0.3175513446331024,
+      "learning_rate": 1.810661060405411e-07,
+      "loss": 0.3706,
+      "step": 3759
+    },
+    {
+      "epoch": 2.768777614138439,
+      "grad_norm": 0.3018471896648407,
+      "learning_rate": 1.7992523894074688e-07,
+      "loss": 0.3492,
+      "step": 3760
+    },
+    {
+      "epoch": 2.7695139911634756,
+      "grad_norm": 0.3039242625236511,
+      "learning_rate": 1.7878791155819918e-07,
+      "loss": 0.3739,
+      "step": 3761
+    },
+    {
+      "epoch": 2.7702503681885124,
+      "grad_norm": 0.306921124458313,
+      "learning_rate": 1.776541247281177e-07,
+      "loss": 0.3801,
+      "step": 3762
+    },
+    {
+      "epoch": 2.7709867452135493,
+      "grad_norm": 0.3310092091560364,
+      "learning_rate": 1.7652387928311977e-07,
+      "loss": 0.381,
+      "step": 3763
+    },
+    {
+      "epoch": 2.771723122238586,
+      "grad_norm": 0.3500930368900299,
+      "learning_rate": 1.7539717605322527e-07,
+      "loss": 0.3488,
+      "step": 3764
+    },
+    {
+      "epoch": 2.772459499263623,
+      "grad_norm": 0.3258472681045532,
+      "learning_rate": 1.7427401586585068e-07,
+      "loss": 0.3674,
+      "step": 3765
+    },
+    {
+      "epoch": 2.7731958762886597,
+      "grad_norm": 0.30289342999458313,
+      "learning_rate": 1.731543995458096e-07,
+      "loss": 0.3477,
+      "step": 3766
+    },
+    {
+      "epoch": 2.7739322533136965,
+      "grad_norm": 0.29461464285850525,
+      "learning_rate": 1.7203832791531594e-07,
+      "loss": 0.3893,
+      "step": 3767
+    },
+    {
+      "epoch": 2.7746686303387333,
+      "grad_norm": 0.3397156298160553,
+      "learning_rate": 1.7092580179397856e-07,
+      "loss": 0.3684,
+      "step": 3768
+    },
+    {
+      "epoch": 2.77540500736377,
+      "grad_norm": 0.3199734687805176,
+      "learning_rate": 1.6981682199880167e-07,
+      "loss": 0.3924,
+      "step": 3769
+    },
+    {
+      "epoch": 2.776141384388807,
+      "grad_norm": 0.3441019654273987,
+      "learning_rate": 1.6871138934418884e-07,
+      "loss": 0.3684,
+      "step": 3770
+    },
+    {
+      "epoch": 2.776877761413844,
+      "grad_norm": 0.30693432688713074,
+      "learning_rate": 1.676095046419346e-07,
+      "loss": 0.3689,
+      "step": 3771
+    },
+    {
+      "epoch": 2.7776141384388806,
+      "grad_norm": 0.3343988358974457,
+      "learning_rate": 1.6651116870122997e-07,
+      "loss": 0.3952,
+      "step": 3772
+    },
+    {
+      "epoch": 2.7783505154639174,
+      "grad_norm": 0.3290632665157318,
+      "learning_rate": 1.654163823286603e-07,
+      "loss": 0.3819,
+      "step": 3773
+    },
+    {
+      "epoch": 2.7790868924889542,
+      "grad_norm": 0.3326128423213959,
+      "learning_rate": 1.6432514632820363e-07,
+      "loss": 0.3808,
+      "step": 3774
+    },
+    {
+      "epoch": 2.779823269513991,
+      "grad_norm": 0.3110780417919159,
+      "learning_rate": 1.6323746150123e-07,
+      "loss": 0.3911,
+      "step": 3775
+    },
+    {
+      "epoch": 2.780559646539028,
+      "grad_norm": 0.3054400682449341,
+      "learning_rate": 1.6215332864650434e-07,
+      "loss": 0.3856,
+      "step": 3776
+    },
+    {
+      "epoch": 2.7812960235640647,
+      "grad_norm": 0.3271074891090393,
+      "learning_rate": 1.6107274856017763e-07,
+      "loss": 0.3531,
+      "step": 3777
+    },
+    {
+      "epoch": 2.7820324005891015,
+      "grad_norm": 0.3214995563030243,
+      "learning_rate": 1.5999572203579783e-07,
+      "loss": 0.3729,
+      "step": 3778
+    },
+    {
+      "epoch": 2.7827687776141383,
+      "grad_norm": 0.3266492486000061,
+      "learning_rate": 1.5892224986430006e-07,
+      "loss": 0.384,
+      "step": 3779
+    },
+    {
+      "epoch": 2.783505154639175,
+      "grad_norm": 0.2639772295951843,
+      "learning_rate": 1.578523328340087e-07,
+      "loss": 0.3665,
+      "step": 3780
+    },
+    {
+      "epoch": 2.784241531664212,
+      "grad_norm": 0.36343199014663696,
+      "learning_rate": 1.5678597173064026e-07,
+      "loss": 0.3718,
+      "step": 3781
+    },
+    {
+      "epoch": 2.7849779086892488,
+      "grad_norm": 0.36233747005462646,
+      "learning_rate": 1.5572316733729775e-07,
+      "loss": 0.3556,
+      "step": 3782
+    },
+    {
+      "epoch": 2.7857142857142856,
+      "grad_norm": 0.33394989371299744,
+      "learning_rate": 1.5466392043447132e-07,
+      "loss": 0.3592,
+      "step": 3783
+    },
+    {
+      "epoch": 2.7864506627393224,
+      "grad_norm": 0.3197415769100189,
+      "learning_rate": 1.5360823180004146e-07,
+      "loss": 0.3903,
+      "step": 3784
+    },
+    {
+      "epoch": 2.787187039764359,
+      "grad_norm": 0.3251337707042694,
+      "learning_rate": 1.5255610220927252e-07,
+      "loss": 0.3953,
+      "step": 3785
+    },
+    {
+      "epoch": 2.787923416789396,
+      "grad_norm": 0.33090558648109436,
+      "learning_rate": 1.515075324348181e-07,
+      "loss": 0.3659,
+      "step": 3786
+    },
+    {
+      "epoch": 2.788659793814433,
+      "grad_norm": 0.31915581226348877,
+      "learning_rate": 1.504625232467155e-07,
+      "loss": 0.3764,
+      "step": 3787
+    },
+    {
+      "epoch": 2.7893961708394697,
+      "grad_norm": 0.34577444195747375,
+      "learning_rate": 1.4942107541238705e-07,
+      "loss": 0.3593,
+      "step": 3788
+    },
+    {
+      "epoch": 2.7901325478645065,
+      "grad_norm": 0.3278156816959381,
+      "learning_rate": 1.48383189696642e-07,
+      "loss": 0.3688,
+      "step": 3789
+    },
+    {
+      "epoch": 2.7908689248895433,
+      "grad_norm": 0.31039220094680786,
+      "learning_rate": 1.4734886686167182e-07,
+      "loss": 0.3732,
+      "step": 3790
+    },
+    {
+      "epoch": 2.79160530191458,
+      "grad_norm": 0.31084826588630676,
+      "learning_rate": 1.4631810766705112e-07,
+      "loss": 0.35,
+      "step": 3791
+    },
+    {
+      "epoch": 2.792341678939617,
+      "grad_norm": 0.29743996262550354,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.3907,
+      "step": 3792
+    },
+    {
+      "epoch": 2.7930780559646537,
+      "grad_norm": 0.30235204100608826,
+      "learning_rate": 1.4426728322407822e-07,
+      "loss": 0.4054,
+      "step": 3793
+    },
+    {
+      "epoch": 2.7938144329896906,
+      "grad_norm": 0.31857481598854065,
+      "learning_rate": 1.4324721948178743e-07,
+      "loss": 0.377,
+      "step": 3794
+    },
+    {
+      "epoch": 2.7945508100147274,
+      "grad_norm": 0.3209660053253174,
+      "learning_rate": 1.4223072239197333e-07,
+      "loss": 0.3558,
+      "step": 3795
+    },
+    {
+      "epoch": 2.795287187039764,
+      "grad_norm": 0.3144548833370209,
+      "learning_rate": 1.412177927011199e-07,
+      "loss": 0.3617,
+      "step": 3796
+    },
+    {
+      "epoch": 2.796023564064801,
+      "grad_norm": 0.32282963395118713,
+      "learning_rate": 1.4020843115309213e-07,
+      "loss": 0.4057,
+      "step": 3797
+    },
+    {
+      "epoch": 2.796759941089838,
+      "grad_norm": 0.3219519555568695,
+      "learning_rate": 1.3920263848913484e-07,
+      "loss": 0.3917,
+      "step": 3798
+    },
+    {
+      "epoch": 2.7974963181148746,
+      "grad_norm": 0.32549378275871277,
+      "learning_rate": 1.3820041544787167e-07,
+      "loss": 0.3979,
+      "step": 3799
+    },
+    {
+      "epoch": 2.7982326951399115,
+      "grad_norm": 0.36029815673828125,
+      "learning_rate": 1.372017627653044e-07,
+      "loss": 0.36,
+      "step": 3800
+    },
+    {
+      "epoch": 2.7989690721649483,
+      "grad_norm": 0.29594072699546814,
+      "learning_rate": 1.3620668117481471e-07,
+      "loss": 0.3597,
+      "step": 3801
+    },
+    {
+      "epoch": 2.799705449189985,
+      "grad_norm": 0.30883079767227173,
+      "learning_rate": 1.3521517140715867e-07,
+      "loss": 0.3621,
+      "step": 3802
+    },
+    {
+      "epoch": 2.800441826215022,
+      "grad_norm": 0.29919102787971497,
+      "learning_rate": 1.3422723419047267e-07,
+      "loss": 0.3504,
+      "step": 3803
+    },
+    {
+      "epoch": 2.8011782032400587,
+      "grad_norm": 0.3039191663265228,
+      "learning_rate": 1.332428702502675e-07,
+      "loss": 0.37,
+      "step": 3804
+    },
+    {
+      "epoch": 2.8019145802650955,
+      "grad_norm": 0.31316202878952026,
+      "learning_rate": 1.3226208030942934e-07,
+      "loss": 0.3754,
+      "step": 3805
+    },
+    {
+      "epoch": 2.8026509572901324,
+      "grad_norm": 0.2829436957836151,
+      "learning_rate": 1.3128486508822202e-07,
+      "loss": 0.3824,
+      "step": 3806
+    },
+    {
+      "epoch": 2.803387334315169,
+      "grad_norm": 0.3167963922023773,
+      "learning_rate": 1.3031122530428264e-07,
+      "loss": 0.3699,
+      "step": 3807
+    },
+    {
+      "epoch": 2.804123711340206,
+      "grad_norm": 0.3057669997215271,
+      "learning_rate": 1.2934116167262145e-07,
+      "loss": 0.3758,
+      "step": 3808
+    },
+    {
+      "epoch": 2.804860088365243,
+      "grad_norm": 0.3250826895236969,
+      "learning_rate": 1.2837467490562583e-07,
+      "loss": 0.3437,
+      "step": 3809
+    },
+    {
+      "epoch": 2.8055964653902796,
+      "grad_norm": 0.3518430292606354,
+      "learning_rate": 1.274117657130536e-07,
+      "loss": 0.3829,
+      "step": 3810
+    },
+    {
+      "epoch": 2.8063328424153164,
+      "grad_norm": 0.32420122623443604,
+      "learning_rate": 1.2645243480203574e-07,
+      "loss": 0.3743,
+      "step": 3811
+    },
+    {
+      "epoch": 2.8070692194403533,
+      "grad_norm": 0.3204830586910248,
+      "learning_rate": 1.254966828770765e-07,
+      "loss": 0.3848,
+      "step": 3812
+    },
+    {
+      "epoch": 2.80780559646539,
+      "grad_norm": 0.3180425763130188,
+      "learning_rate": 1.2454451064005058e-07,
+      "loss": 0.39,
+      "step": 3813
+    },
+    {
+      "epoch": 2.808541973490427,
+      "grad_norm": 0.32775330543518066,
+      "learning_rate": 1.2359591879020528e-07,
+      "loss": 0.376,
+      "step": 3814
+    },
+    {
+      "epoch": 2.8092783505154637,
+      "grad_norm": 0.34739986062049866,
+      "learning_rate": 1.2265090802415724e-07,
+      "loss": 0.3723,
+      "step": 3815
+    },
+    {
+      "epoch": 2.8100147275405005,
+      "grad_norm": 0.3330659866333008,
+      "learning_rate": 1.217094790358936e-07,
+      "loss": 0.3697,
+      "step": 3816
+    },
+    {
+      "epoch": 2.8107511045655373,
+      "grad_norm": 0.30431681871414185,
+      "learning_rate": 1.2077163251677182e-07,
+      "loss": 0.3748,
+      "step": 3817
+    },
+    {
+      "epoch": 2.811487481590574,
+      "grad_norm": 0.33776891231536865,
+      "learning_rate": 1.1983736915551824e-07,
+      "loss": 0.3889,
+      "step": 3818
+    },
+    {
+      "epoch": 2.812223858615611,
+      "grad_norm": 0.3244916498661041,
+      "learning_rate": 1.1890668963822793e-07,
+      "loss": 0.379,
+      "step": 3819
+    },
+    {
+      "epoch": 2.812960235640648,
+      "grad_norm": 0.3201790750026703,
+      "learning_rate": 1.179795946483625e-07,
+      "loss": 0.3578,
+      "step": 3820
+    },
+    {
+      "epoch": 2.8136966126656846,
+      "grad_norm": 0.32924962043762207,
+      "learning_rate": 1.170560848667529e-07,
+      "loss": 0.3561,
+      "step": 3821
+    },
+    {
+      "epoch": 2.8144329896907214,
+      "grad_norm": 0.318311870098114,
+      "learning_rate": 1.1613616097159774e-07,
+      "loss": 0.3588,
+      "step": 3822
+    },
+    {
+      "epoch": 2.8151693667157582,
+      "grad_norm": 0.30944523215293884,
+      "learning_rate": 1.1521982363846051e-07,
+      "loss": 0.3534,
+      "step": 3823
+    },
+    {
+      "epoch": 2.815905743740795,
+      "grad_norm": 0.3113093078136444,
+      "learning_rate": 1.1430707354027182e-07,
+      "loss": 0.3749,
+      "step": 3824
+    },
+    {
+      "epoch": 2.816642120765832,
+      "grad_norm": 0.298728346824646,
+      "learning_rate": 1.1339791134732769e-07,
+      "loss": 0.3704,
+      "step": 3825
+    },
+    {
+      "epoch": 2.8173784977908687,
+      "grad_norm": 0.30721381306648254,
+      "learning_rate": 1.1249233772729018e-07,
+      "loss": 0.403,
+      "step": 3826
+    },
+    {
+      "epoch": 2.8181148748159055,
+      "grad_norm": 0.3090912997722626,
+      "learning_rate": 1.1159035334518343e-07,
+      "loss": 0.3721,
+      "step": 3827
+    },
+    {
+      "epoch": 2.8188512518409423,
+      "grad_norm": 0.30627208948135376,
+      "learning_rate": 1.1069195886339923e-07,
+      "loss": 0.3677,
+      "step": 3828
+    },
+    {
+      "epoch": 2.819587628865979,
+      "grad_norm": 0.32268771529197693,
+      "learning_rate": 1.0979715494169096e-07,
+      "loss": 0.3591,
+      "step": 3829
+    },
+    {
+      "epoch": 2.820324005891016,
+      "grad_norm": 0.33059263229370117,
+      "learning_rate": 1.089059422371741e-07,
+      "loss": 0.3579,
+      "step": 3830
+    },
+    {
+      "epoch": 2.8210603829160528,
+      "grad_norm": 0.3322978913784027,
+      "learning_rate": 1.0801832140433066e-07,
+      "loss": 0.3856,
+      "step": 3831
+    },
+    {
+      "epoch": 2.8217967599410896,
+      "grad_norm": 0.3347814977169037,
+      "learning_rate": 1.071342930950009e-07,
+      "loss": 0.3722,
+      "step": 3832
+    },
+    {
+      "epoch": 2.8225331369661264,
+      "grad_norm": 0.34562426805496216,
+      "learning_rate": 1.0625385795838883e-07,
+      "loss": 0.3687,
+      "step": 3833
+    },
+    {
+      "epoch": 2.823269513991163,
+      "grad_norm": 0.3513891100883484,
+      "learning_rate": 1.0537701664106003e-07,
+      "loss": 0.3737,
+      "step": 3834
+    },
+    {
+      "epoch": 2.8240058910162,
+      "grad_norm": 0.31909000873565674,
+      "learning_rate": 1.0450376978693999e-07,
+      "loss": 0.3695,
+      "step": 3835
+    },
+    {
+      "epoch": 2.824742268041237,
+      "grad_norm": 0.3165963292121887,
+      "learning_rate": 1.0363411803731404e-07,
+      "loss": 0.3876,
+      "step": 3836
+    },
+    {
+      "epoch": 2.8254786450662737,
+      "grad_norm": 0.32449856400489807,
+      "learning_rate": 1.0276806203082967e-07,
+      "loss": 0.3678,
+      "step": 3837
+    },
+    {
+      "epoch": 2.8262150220913105,
+      "grad_norm": 0.3418518006801605,
+      "learning_rate": 1.0190560240349035e-07,
+      "loss": 0.3594,
+      "step": 3838
+    },
+    {
+      "epoch": 2.8269513991163473,
+      "grad_norm": 0.32669028639793396,
+      "learning_rate": 1.0104673978866164e-07,
+      "loss": 0.3804,
+      "step": 3839
+    },
+    {
+      "epoch": 2.827687776141384,
+      "grad_norm": 0.2968961298465729,
+      "learning_rate": 1.0019147481706626e-07,
+      "loss": 0.3726,
+      "step": 3840
+    },
+    {
+      "epoch": 2.8284241531664214,
+      "grad_norm": 0.3117745816707611,
+      "learning_rate": 9.933980811678401e-08,
+      "loss": 0.3972,
+      "step": 3841
+    },
+    {
+      "epoch": 2.829160530191458,
+      "grad_norm": 0.3052343428134918,
+      "learning_rate": 9.84917403132546e-08,
+      "loss": 0.3743,
+      "step": 3842
+    },
+    {
+      "epoch": 2.829896907216495,
+      "grad_norm": 0.3451398015022278,
+      "learning_rate": 9.764727202927259e-08,
+      "loss": 0.3921,
+      "step": 3843
+    },
+    {
+      "epoch": 2.830633284241532,
+      "grad_norm": 0.3272383213043213,
+      "learning_rate": 9.680640388498974e-08,
+      "loss": 0.3752,
+      "step": 3844
+    },
+    {
+      "epoch": 2.8313696612665686,
+      "grad_norm": 0.3403890132904053,
+      "learning_rate": 9.596913649791484e-08,
+      "loss": 0.3564,
+      "step": 3845
+    },
+    {
+      "epoch": 2.8321060382916055,
+      "grad_norm": 0.3162234425544739,
+      "learning_rate": 9.51354704829105e-08,
+      "loss": 0.3515,
+      "step": 3846
+    },
+    {
+      "epoch": 2.8328424153166423,
+      "grad_norm": 0.31846678256988525,
+      "learning_rate": 9.430540645219755e-08,
+      "loss": 0.374,
+      "step": 3847
+    },
+    {
+      "epoch": 2.833578792341679,
+      "grad_norm": 0.32757052779197693,
+      "learning_rate": 9.347894501534949e-08,
+      "loss": 0.3537,
+      "step": 3848
+    },
+    {
+      "epoch": 2.834315169366716,
+      "grad_norm": 0.32147282361984253,
+      "learning_rate": 9.26560867792936e-08,
+      "loss": 0.3824,
+      "step": 3849
+    },
+    {
+      "epoch": 2.8350515463917527,
+      "grad_norm": 0.34874793887138367,
+      "learning_rate": 9.18368323483132e-08,
+      "loss": 0.3895,
+      "step": 3850
+    },
+    {
+      "epoch": 2.8357879234167895,
+      "grad_norm": 0.29230934381484985,
+      "learning_rate": 9.102118232404311e-08,
+      "loss": 0.391,
+      "step": 3851
+    },
+    {
+      "epoch": 2.8365243004418264,
+      "grad_norm": 0.33275964856147766,
+      "learning_rate": 9.020913730547309e-08,
+      "loss": 0.3728,
+      "step": 3852
+    },
+    {
+      "epoch": 2.837260677466863,
+      "grad_norm": 0.36540740728378296,
+      "learning_rate": 8.940069788894389e-08,
+      "loss": 0.3738,
+      "step": 3853
+    },
+    {
+      "epoch": 2.8379970544919,
+      "grad_norm": 0.29792320728302,
+      "learning_rate": 8.859586466814895e-08,
+      "loss": 0.3478,
+      "step": 3854
+    },
+    {
+      "epoch": 2.838733431516937,
+      "grad_norm": 0.35100606083869934,
+      "learning_rate": 8.77946382341327e-08,
+      "loss": 0.3724,
+      "step": 3855
+    },
+    {
+      "epoch": 2.8394698085419736,
+      "grad_norm": 0.31920912861824036,
+      "learning_rate": 8.699701917529335e-08,
+      "loss": 0.3789,
+      "step": 3856
+    },
+    {
+      "epoch": 2.8402061855670104,
+      "grad_norm": 0.3164528012275696,
+      "learning_rate": 8.62030080773768e-08,
+      "loss": 0.3715,
+      "step": 3857
+    },
+    {
+      "epoch": 2.8409425625920472,
+      "grad_norm": 0.3255048990249634,
+      "learning_rate": 8.541260552348107e-08,
+      "loss": 0.3844,
+      "step": 3858
+    },
+    {
+      "epoch": 2.841678939617084,
+      "grad_norm": 0.3301268517971039,
+      "learning_rate": 8.462581209405519e-08,
+      "loss": 0.4026,
+      "step": 3859
+    },
+    {
+      "epoch": 2.842415316642121,
+      "grad_norm": 0.3406490087509155,
+      "learning_rate": 8.384262836689472e-08,
+      "loss": 0.3798,
+      "step": 3860
+    },
+    {
+      "epoch": 2.8431516936671577,
+      "grad_norm": 0.30102699995040894,
+      "learning_rate": 8.306305491714683e-08,
+      "loss": 0.3826,
+      "step": 3861
+    },
+    {
+      "epoch": 2.8438880706921945,
+      "grad_norm": 0.30409371852874756,
+      "learning_rate": 8.228709231730747e-08,
+      "loss": 0.3754,
+      "step": 3862
+    },
+    {
+      "epoch": 2.8446244477172313,
+      "grad_norm": 0.30450791120529175,
+      "learning_rate": 8.151474113721803e-08,
+      "loss": 0.381,
+      "step": 3863
+    },
+    {
+      "epoch": 2.845360824742268,
+      "grad_norm": 0.3098084628582001,
+      "learning_rate": 8.074600194407257e-08,
+      "loss": 0.3558,
+      "step": 3864
+    },
+    {
+      "epoch": 2.846097201767305,
+      "grad_norm": 0.315789133310318,
+      "learning_rate": 7.998087530240784e-08,
+      "loss": 0.3601,
+      "step": 3865
+    },
+    {
+      "epoch": 2.846833578792342,
+      "grad_norm": 0.3192322552204132,
+      "learning_rate": 7.921936177411049e-08,
+      "loss": 0.4126,
+      "step": 3866
+    },
+    {
+      "epoch": 2.8475699558173786,
+      "grad_norm": 0.3339972198009491,
+      "learning_rate": 7.846146191841319e-08,
+      "loss": 0.4005,
+      "step": 3867
+    },
+    {
+      "epoch": 2.8483063328424154,
+      "grad_norm": 0.31077709794044495,
+      "learning_rate": 7.770717629189462e-08,
+      "loss": 0.3869,
+      "step": 3868
+    },
+    {
+      "epoch": 2.8490427098674522,
+      "grad_norm": 0.327124685049057,
+      "learning_rate": 7.695650544847888e-08,
+      "loss": 0.3628,
+      "step": 3869
+    },
+    {
+      "epoch": 2.849779086892489,
+      "grad_norm": 0.32845091819763184,
+      "learning_rate": 7.620944993943669e-08,
+      "loss": 0.37,
+      "step": 3870
+    },
+    {
+      "epoch": 2.850515463917526,
+      "grad_norm": 0.32456862926483154,
+      "learning_rate": 7.546601031338252e-08,
+      "loss": 0.3459,
+      "step": 3871
+    },
+    {
+      "epoch": 2.8512518409425627,
+      "grad_norm": 0.31279054284095764,
+      "learning_rate": 7.472618711627577e-08,
+      "loss": 0.3533,
+      "step": 3872
+    },
+    {
+      "epoch": 2.8519882179675995,
+      "grad_norm": 0.31263667345046997,
+      "learning_rate": 7.398998089142128e-08,
+      "loss": 0.3837,
+      "step": 3873
+    },
+    {
+      "epoch": 2.8527245949926363,
+      "grad_norm": 0.31096217036247253,
+      "learning_rate": 7.325739217946547e-08,
+      "loss": 0.4039,
+      "step": 3874
+    },
+    {
+      "epoch": 2.853460972017673,
+      "grad_norm": 0.31137025356292725,
+      "learning_rate": 7.252842151839967e-08,
+      "loss": 0.3707,
+      "step": 3875
+    },
+    {
+      "epoch": 2.85419734904271,
+      "grad_norm": 0.3208453953266144,
+      "learning_rate": 7.180306944355896e-08,
+      "loss": 0.3662,
+      "step": 3876
+    },
+    {
+      "epoch": 2.8549337260677468,
+      "grad_norm": 0.3515404164791107,
+      "learning_rate": 7.108133648761839e-08,
+      "loss": 0.3336,
+      "step": 3877
+    },
+    {
+      "epoch": 2.8556701030927836,
+      "grad_norm": 0.32757940888404846,
+      "learning_rate": 7.036322318059785e-08,
+      "loss": 0.3916,
+      "step": 3878
+    },
+    {
+      "epoch": 2.8564064801178204,
+      "grad_norm": 0.3318811357021332,
+      "learning_rate": 6.964873004985717e-08,
+      "loss": 0.339,
+      "step": 3879
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.28970828652381897,
+      "learning_rate": 6.893785762009942e-08,
+      "loss": 0.376,
+      "step": 3880
+    },
+    {
+      "epoch": 2.857879234167894,
+      "grad_norm": 0.30351996421813965,
+      "learning_rate": 6.823060641336809e-08,
+      "loss": 0.3666,
+      "step": 3881
+    },
+    {
+      "epoch": 2.858615611192931,
+      "grad_norm": 0.2939937710762024,
+      "learning_rate": 6.752697694904553e-08,
+      "loss": 0.3697,
+      "step": 3882
+    },
+    {
+      "epoch": 2.8593519882179677,
+      "grad_norm": 0.31340742111206055,
+      "learning_rate": 6.682696974385727e-08,
+      "loss": 0.3581,
+      "step": 3883
+    },
+    {
+      "epoch": 2.8600883652430045,
+      "grad_norm": 0.3075888156890869,
+      "learning_rate": 6.613058531186767e-08,
+      "loss": 0.4048,
+      "step": 3884
+    },
+    {
+      "epoch": 2.8608247422680413,
+      "grad_norm": 0.3178377151489258,
+      "learning_rate": 6.54378241644793e-08,
+      "loss": 0.3927,
+      "step": 3885
+    },
+    {
+      "epoch": 2.861561119293078,
+      "grad_norm": 0.31362468004226685,
+      "learning_rate": 6.474868681043578e-08,
+      "loss": 0.3953,
+      "step": 3886
+    },
+    {
+      "epoch": 2.862297496318115,
+      "grad_norm": 0.3346673250198364,
+      "learning_rate": 6.406317375581839e-08,
+      "loss": 0.3578,
+      "step": 3887
+    },
+    {
+      "epoch": 2.8630338733431517,
+      "grad_norm": 0.31339147686958313,
+      "learning_rate": 6.338128550404721e-08,
+      "loss": 0.377,
+      "step": 3888
+    },
+    {
+      "epoch": 2.8637702503681886,
+      "grad_norm": 0.31569039821624756,
+      "learning_rate": 6.270302255588112e-08,
+      "loss": 0.372,
+      "step": 3889
+    },
+    {
+      "epoch": 2.8645066273932254,
+      "grad_norm": 0.33136993646621704,
+      "learning_rate": 6.202838540941503e-08,
+      "loss": 0.4067,
+      "step": 3890
+    },
+    {
+      "epoch": 2.865243004418262,
+      "grad_norm": 0.3213663101196289,
+      "learning_rate": 6.135737456008207e-08,
+      "loss": 0.3936,
+      "step": 3891
+    },
+    {
+      "epoch": 2.865979381443299,
+      "grad_norm": 0.29569748044013977,
+      "learning_rate": 6.06899905006525e-08,
+      "loss": 0.3815,
+      "step": 3892
+    },
+    {
+      "epoch": 2.866715758468336,
+      "grad_norm": 0.31531018018722534,
+      "learning_rate": 6.002623372123373e-08,
+      "loss": 0.377,
+      "step": 3893
+    },
+    {
+      "epoch": 2.8674521354933726,
+      "grad_norm": 0.31117159128189087,
+      "learning_rate": 5.9366104709267515e-08,
+      "loss": 0.3674,
+      "step": 3894
+    },
+    {
+      "epoch": 2.8681885125184094,
+      "grad_norm": 0.3713977634906769,
+      "learning_rate": 5.8709603949533844e-08,
+      "loss": 0.3724,
+      "step": 3895
+    },
+    {
+      "epoch": 2.8689248895434463,
+      "grad_norm": 0.30198606848716736,
+      "learning_rate": 5.805673192414596e-08,
+      "loss": 0.4026,
+      "step": 3896
+    },
+    {
+      "epoch": 2.869661266568483,
+      "grad_norm": 0.30127283930778503,
+      "learning_rate": 5.740748911255367e-08,
+      "loss": 0.3914,
+      "step": 3897
+    },
+    {
+      "epoch": 2.87039764359352,
+      "grad_norm": 0.3306448757648468,
+      "learning_rate": 5.6761875991541704e-08,
+      "loss": 0.3497,
+      "step": 3898
+    },
+    {
+      "epoch": 2.8711340206185567,
+      "grad_norm": 0.3268984258174896,
+      "learning_rate": 5.611989303522858e-08,
+      "loss": 0.3916,
+      "step": 3899
+    },
+    {
+      "epoch": 2.8718703976435935,
+      "grad_norm": 0.3211970627307892,
+      "learning_rate": 5.5481540715066616e-08,
+      "loss": 0.365,
+      "step": 3900
+    },
+    {
+      "epoch": 2.8726067746686303,
+      "grad_norm": 0.336363822221756,
+      "learning_rate": 5.4846819499843605e-08,
+      "loss": 0.3842,
+      "step": 3901
+    },
+    {
+      "epoch": 2.873343151693667,
+      "grad_norm": 0.3143905997276306,
+      "learning_rate": 5.4215729855678914e-08,
+      "loss": 0.3844,
+      "step": 3902
+    },
+    {
+      "epoch": 2.874079528718704,
+      "grad_norm": 0.35480746626853943,
+      "learning_rate": 5.35882722460257e-08,
+      "loss": 0.3772,
+      "step": 3903
+    },
+    {
+      "epoch": 2.874815905743741,
+      "grad_norm": 0.3486427366733551,
+      "learning_rate": 5.296444713166981e-08,
+      "loss": 0.3757,
+      "step": 3904
+    },
+    {
+      "epoch": 2.8755522827687776,
+      "grad_norm": 0.3225955367088318,
+      "learning_rate": 5.234425497072981e-08,
+      "loss": 0.3417,
+      "step": 3905
+    },
+    {
+      "epoch": 2.8762886597938144,
+      "grad_norm": 0.31462594866752625,
+      "learning_rate": 5.172769621865637e-08,
+      "loss": 0.4036,
+      "step": 3906
+    },
+    {
+      "epoch": 2.8770250368188512,
+      "grad_norm": 0.3221474885940552,
+      "learning_rate": 5.1114771328230615e-08,
+      "loss": 0.3568,
+      "step": 3907
+    },
+    {
+      "epoch": 2.877761413843888,
+      "grad_norm": 0.319514662027359,
+      "learning_rate": 5.050548074956696e-08,
+      "loss": 0.359,
+      "step": 3908
+    },
+    {
+      "epoch": 2.878497790868925,
+      "grad_norm": 0.32837367057800293,
+      "learning_rate": 4.9899824930109694e-08,
+      "loss": 0.3702,
+      "step": 3909
+    },
+    {
+      "epoch": 2.8792341678939617,
+      "grad_norm": 0.30610325932502747,
+      "learning_rate": 4.9297804314633604e-08,
+      "loss": 0.3641,
+      "step": 3910
+    },
+    {
+      "epoch": 2.8799705449189985,
+      "grad_norm": 0.32444047927856445,
+      "learning_rate": 4.869941934524613e-08,
+      "loss": 0.3677,
+      "step": 3911
+    },
+    {
+      "epoch": 2.8807069219440353,
+      "grad_norm": 0.3461378216743469,
+      "learning_rate": 4.810467046138134e-08,
+      "loss": 0.3876,
+      "step": 3912
+    },
+    {
+      "epoch": 2.881443298969072,
+      "grad_norm": 0.32242852449417114,
+      "learning_rate": 4.75135580998054e-08,
+      "loss": 0.349,
+      "step": 3913
+    },
+    {
+      "epoch": 2.882179675994109,
+      "grad_norm": 0.31807756423950195,
+      "learning_rate": 4.69260826946133e-08,
+      "loss": 0.3642,
+      "step": 3914
+    },
+    {
+      "epoch": 2.8829160530191458,
+      "grad_norm": 0.31694477796554565,
+      "learning_rate": 4.634224467722992e-08,
+      "loss": 0.3758,
+      "step": 3915
+    },
+    {
+      "epoch": 2.8836524300441826,
+      "grad_norm": 0.343872994184494,
+      "learning_rate": 4.576204447640675e-08,
+      "loss": 0.3576,
+      "step": 3916
+    },
+    {
+      "epoch": 2.8843888070692194,
+      "grad_norm": 0.3494185209274292,
+      "learning_rate": 4.518548251822685e-08,
+      "loss": 0.3636,
+      "step": 3917
+    },
+    {
+      "epoch": 2.8851251840942562,
+      "grad_norm": 0.32793089747428894,
+      "learning_rate": 4.461255922609986e-08,
+      "loss": 0.3672,
+      "step": 3918
+    },
+    {
+      "epoch": 2.885861561119293,
+      "grad_norm": 0.36390119791030884,
+      "learning_rate": 4.4043275020762e-08,
+      "loss": 0.3504,
+      "step": 3919
+    },
+    {
+      "epoch": 2.88659793814433,
+      "grad_norm": 0.3462429344654083,
+      "learning_rate": 4.3477630320279405e-08,
+      "loss": 0.3632,
+      "step": 3920
+    },
+    {
+      "epoch": 2.8873343151693667,
+      "grad_norm": 0.3237389326095581,
+      "learning_rate": 4.291562554004369e-08,
+      "loss": 0.3802,
+      "step": 3921
+    },
+    {
+      "epoch": 2.8880706921944035,
+      "grad_norm": 0.3111920654773712,
+      "learning_rate": 4.235726109277527e-08,
+      "loss": 0.3409,
+      "step": 3922
+    },
+    {
+      "epoch": 2.8888070692194403,
+      "grad_norm": 0.2972390949726105,
+      "learning_rate": 4.180253738851947e-08,
+      "loss": 0.3955,
+      "step": 3923
+    },
+    {
+      "epoch": 2.889543446244477,
+      "grad_norm": 0.32099413871765137,
+      "learning_rate": 4.125145483464821e-08,
+      "loss": 0.3673,
+      "step": 3924
+    },
+    {
+      "epoch": 2.890279823269514,
+      "grad_norm": 0.31523367762565613,
+      "learning_rate": 4.070401383586109e-08,
+      "loss": 0.354,
+      "step": 3925
+    },
+    {
+      "epoch": 2.8910162002945508,
+      "grad_norm": 0.3093765079975128,
+      "learning_rate": 4.0160214794180976e-08,
+      "loss": 0.38,
+      "step": 3926
+    },
+    {
+      "epoch": 2.8917525773195876,
+      "grad_norm": 0.3105464577674866,
+      "learning_rate": 3.962005810895786e-08,
+      "loss": 0.3578,
+      "step": 3927
+    },
+    {
+      "epoch": 2.8924889543446244,
+      "grad_norm": 0.3610590994358063,
+      "learning_rate": 3.908354417686722e-08,
+      "loss": 0.3708,
+      "step": 3928
+    },
+    {
+      "epoch": 2.893225331369661,
+      "grad_norm": 0.328524112701416,
+      "learning_rate": 3.855067339190721e-08,
+      "loss": 0.3849,
+      "step": 3929
+    },
+    {
+      "epoch": 2.893961708394698,
+      "grad_norm": 0.3390013873577118,
+      "learning_rate": 3.802144614540315e-08,
+      "loss": 0.3637,
+      "step": 3930
+    },
+    {
+      "epoch": 2.894698085419735,
+      "grad_norm": 0.3347131907939911,
+      "learning_rate": 3.749586282600359e-08,
+      "loss": 0.3575,
+      "step": 3931
+    },
+    {
+      "epoch": 2.8954344624447717,
+      "grad_norm": 0.3055760860443115,
+      "learning_rate": 3.6973923819680344e-08,
+      "loss": 0.3632,
+      "step": 3932
+    },
+    {
+      "epoch": 2.8961708394698085,
+      "grad_norm": 0.31726714968681335,
+      "learning_rate": 3.645562950973014e-08,
+      "loss": 0.3791,
+      "step": 3933
+    },
+    {
+      "epoch": 2.8969072164948453,
+      "grad_norm": 0.34960147738456726,
+      "learning_rate": 3.5940980276772394e-08,
+      "loss": 0.3802,
+      "step": 3934
+    },
+    {
+      "epoch": 2.897643593519882,
+      "grad_norm": 0.3189311921596527,
+      "learning_rate": 3.5429976498749794e-08,
+      "loss": 0.3893,
+      "step": 3935
+    },
+    {
+      "epoch": 2.898379970544919,
+      "grad_norm": 0.3356260359287262,
+      "learning_rate": 3.492261855092938e-08,
+      "loss": 0.3665,
+      "step": 3936
+    },
+    {
+      "epoch": 2.8991163475699557,
+      "grad_norm": 0.31776583194732666,
+      "learning_rate": 3.441890680589754e-08,
+      "loss": 0.3619,
+      "step": 3937
+    },
+    {
+      "epoch": 2.8998527245949925,
+      "grad_norm": 0.3313503563404083,
+      "learning_rate": 3.391884163356618e-08,
+      "loss": 0.3877,
+      "step": 3938
+    },
+    {
+      "epoch": 2.9005891016200294,
+      "grad_norm": 0.38583609461784363,
+      "learning_rate": 3.3422423401167634e-08,
+      "loss": 0.3696,
+      "step": 3939
+    },
+    {
+      "epoch": 2.901325478645066,
+      "grad_norm": 0.33413827419281006,
+      "learning_rate": 3.292965247325641e-08,
+      "loss": 0.373,
+      "step": 3940
+    },
+    {
+      "epoch": 2.902061855670103,
+      "grad_norm": 0.331826388835907,
+      "learning_rate": 3.2440529211709146e-08,
+      "loss": 0.3597,
+      "step": 3941
+    },
+    {
+      "epoch": 2.90279823269514,
+      "grad_norm": 0.33604195713996887,
+      "learning_rate": 3.19550539757224e-08,
+      "loss": 0.3729,
+      "step": 3942
+    },
+    {
+      "epoch": 2.9035346097201766,
+      "grad_norm": 0.3193381130695343,
+      "learning_rate": 3.147322712181489e-08,
+      "loss": 0.351,
+      "step": 3943
+    },
+    {
+      "epoch": 2.9042709867452134,
+      "grad_norm": 0.30179837346076965,
+      "learning_rate": 3.0995049003826325e-08,
+      "loss": 0.3728,
+      "step": 3944
+    },
+    {
+      "epoch": 2.9050073637702503,
+      "grad_norm": 0.32953178882598877,
+      "learning_rate": 3.052051997291527e-08,
+      "loss": 0.3682,
+      "step": 3945
+    },
+    {
+      "epoch": 2.905743740795287,
+      "grad_norm": 0.3493231236934662,
+      "learning_rate": 3.0049640377561865e-08,
+      "loss": 0.36,
+      "step": 3946
+    },
+    {
+      "epoch": 2.906480117820324,
+      "grad_norm": 0.332707941532135,
+      "learning_rate": 2.9582410563565587e-08,
+      "loss": 0.3774,
+      "step": 3947
+    },
+    {
+      "epoch": 2.9072164948453607,
+      "grad_norm": 0.326961874961853,
+      "learning_rate": 2.9118830874046988e-08,
+      "loss": 0.3633,
+      "step": 3948
+    },
+    {
+      "epoch": 2.9079528718703975,
+      "grad_norm": 0.34812161326408386,
+      "learning_rate": 2.8658901649443183e-08,
+      "loss": 0.3605,
+      "step": 3949
+    },
+    {
+      "epoch": 2.9086892488954343,
+      "grad_norm": 0.32744961977005005,
+      "learning_rate": 2.8202623227513993e-08,
+      "loss": 0.3688,
+      "step": 3950
+    },
+    {
+      "epoch": 2.909425625920471,
+      "grad_norm": 0.3743055462837219,
+      "learning_rate": 2.7749995943335272e-08,
+      "loss": 0.3543,
+      "step": 3951
+    },
+    {
+      "epoch": 2.910162002945508,
+      "grad_norm": 0.3317681550979614,
+      "learning_rate": 2.730102012930336e-08,
+      "loss": 0.3662,
+      "step": 3952
+    },
+    {
+      "epoch": 2.910898379970545,
+      "grad_norm": 0.36964574456214905,
+      "learning_rate": 2.6855696115133388e-08,
+      "loss": 0.3509,
+      "step": 3953
+    },
+    {
+      "epoch": 2.9116347569955816,
+      "grad_norm": 0.3326326608657837,
+      "learning_rate": 2.6414024227855994e-08,
+      "loss": 0.3931,
+      "step": 3954
+    },
+    {
+      "epoch": 2.9123711340206184,
+      "grad_norm": 0.3078947961330414,
+      "learning_rate": 2.597600479182283e-08,
+      "loss": 0.3566,
+      "step": 3955
+    },
+    {
+      "epoch": 2.9131075110456552,
+      "grad_norm": 0.3276365399360657,
+      "learning_rate": 2.5541638128702694e-08,
+      "loss": 0.3809,
+      "step": 3956
+    },
+    {
+      "epoch": 2.913843888070692,
+      "grad_norm": 0.31601178646087646,
+      "learning_rate": 2.511092455747932e-08,
+      "loss": 0.4119,
+      "step": 3957
+    },
+    {
+      "epoch": 2.914580265095729,
+      "grad_norm": 0.34153667092323303,
+      "learning_rate": 2.4683864394458023e-08,
+      "loss": 0.3436,
+      "step": 3958
+    },
+    {
+      "epoch": 2.9153166421207657,
+      "grad_norm": 0.3269186019897461,
+      "learning_rate": 2.4260457953257377e-08,
+      "loss": 0.3722,
+      "step": 3959
+    },
+    {
+      "epoch": 2.9160530191458025,
+      "grad_norm": 0.3348276913166046,
+      "learning_rate": 2.3840705544815324e-08,
+      "loss": 0.3758,
+      "step": 3960
+    },
+    {
+      "epoch": 2.9167893961708393,
+      "grad_norm": 0.3003559708595276,
+      "learning_rate": 2.3424607477384176e-08,
+      "loss": 0.3738,
+      "step": 3961
+    },
+    {
+      "epoch": 2.917525773195876,
+      "grad_norm": 0.316893607378006,
+      "learning_rate": 2.3012164056534503e-08,
+      "loss": 0.3652,
+      "step": 3962
+    },
+    {
+      "epoch": 2.918262150220913,
+      "grad_norm": 0.3250206410884857,
+      "learning_rate": 2.260337558515291e-08,
+      "loss": 0.3752,
+      "step": 3963
+    },
+    {
+      "epoch": 2.9189985272459498,
+      "grad_norm": 0.34431540966033936,
+      "learning_rate": 2.2198242363439814e-08,
+      "loss": 0.3506,
+      "step": 3964
+    },
+    {
+      "epoch": 2.9197349042709866,
+      "grad_norm": 0.3138362765312195,
+      "learning_rate": 2.179676468891334e-08,
+      "loss": 0.3998,
+      "step": 3965
+    },
+    {
+      "epoch": 2.9204712812960234,
+      "grad_norm": 0.33245036005973816,
+      "learning_rate": 2.1398942856407646e-08,
+      "loss": 0.3787,
+      "step": 3966
+    },
+    {
+      "epoch": 2.92120765832106,
+      "grad_norm": 0.34872448444366455,
+      "learning_rate": 2.100477715806959e-08,
+      "loss": 0.3701,
+      "step": 3967
+    },
+    {
+      "epoch": 2.9219440353460975,
+      "grad_norm": 0.3229193091392517,
+      "learning_rate": 2.061426788336318e-08,
+      "loss": 0.3846,
+      "step": 3968
+    },
+    {
+      "epoch": 2.9226804123711343,
+      "grad_norm": 0.3165433406829834,
+      "learning_rate": 2.0227415319067355e-08,
+      "loss": 0.3617,
+      "step": 3969
+    },
+    {
+      "epoch": 2.923416789396171,
+      "grad_norm": 0.34778398275375366,
+      "learning_rate": 1.984421974927375e-08,
+      "loss": 0.3858,
+      "step": 3970
+    },
+    {
+      "epoch": 2.924153166421208,
+      "grad_norm": 0.31038519740104675,
+      "learning_rate": 1.946468145538949e-08,
+      "loss": 0.3647,
+      "step": 3971
+    },
+    {
+      "epoch": 2.9248895434462447,
+      "grad_norm": 0.2943412959575653,
+      "learning_rate": 1.908880071613717e-08,
+      "loss": 0.3628,
+      "step": 3972
+    },
+    {
+      "epoch": 2.9256259204712816,
+      "grad_norm": 0.29908487200737,
+      "learning_rate": 1.871657780755154e-08,
+      "loss": 0.3674,
+      "step": 3973
+    },
+    {
+      "epoch": 2.9263622974963184,
+      "grad_norm": 0.3493654131889343,
+      "learning_rate": 1.8348013002982278e-08,
+      "loss": 0.3899,
+      "step": 3974
+    },
+    {
+      "epoch": 2.927098674521355,
+      "grad_norm": 0.33574777841567993,
+      "learning_rate": 1.798310657309177e-08,
+      "loss": 0.3621,
+      "step": 3975
+    },
+    {
+      "epoch": 2.927835051546392,
+      "grad_norm": 0.3295884430408478,
+      "learning_rate": 1.7621858785856206e-08,
+      "loss": 0.3744,
+      "step": 3976
+    },
+    {
+      "epoch": 2.928571428571429,
+      "grad_norm": 0.31094932556152344,
+      "learning_rate": 1.72642699065656e-08,
+      "loss": 0.3566,
+      "step": 3977
+    },
+    {
+      "epoch": 2.9293078055964656,
+      "grad_norm": 0.34367263317108154,
+      "learning_rate": 1.6910340197822116e-08,
+      "loss": 0.3691,
+      "step": 3978
+    },
+    {
+      "epoch": 2.9300441826215025,
+      "grad_norm": 0.32079145312309265,
+      "learning_rate": 1.6560069919541177e-08,
+      "loss": 0.3693,
+      "step": 3979
+    },
+    {
+      "epoch": 2.9307805596465393,
+      "grad_norm": 0.3374221920967102,
+      "learning_rate": 1.6213459328950355e-08,
+      "loss": 0.3858,
+      "step": 3980
+    },
+    {
+      "epoch": 2.931516936671576,
+      "grad_norm": 0.34414032101631165,
+      "learning_rate": 1.5870508680589923e-08,
+      "loss": 0.349,
+      "step": 3981
+    },
+    {
+      "epoch": 2.932253313696613,
+      "grad_norm": 0.3187008202075958,
+      "learning_rate": 1.5531218226312872e-08,
+      "loss": 0.3836,
+      "step": 3982
+    },
+    {
+      "epoch": 2.9329896907216497,
+      "grad_norm": 0.3022221624851227,
+      "learning_rate": 1.5195588215283773e-08,
+      "loss": 0.3954,
+      "step": 3983
+    },
+    {
+      "epoch": 2.9337260677466865,
+      "grad_norm": 0.29533877968788147,
+      "learning_rate": 1.4863618893979359e-08,
+      "loss": 0.3874,
+      "step": 3984
+    },
+    {
+      "epoch": 2.9344624447717234,
+      "grad_norm": 0.3319675624370575,
+      "learning_rate": 1.4535310506187394e-08,
+      "loss": 0.3379,
+      "step": 3985
+    },
+    {
+      "epoch": 2.93519882179676,
+      "grad_norm": 0.3372650444507599,
+      "learning_rate": 1.4210663293008353e-08,
+      "loss": 0.3825,
+      "step": 3986
+    },
+    {
+      "epoch": 2.935935198821797,
+      "grad_norm": 0.3170906901359558,
+      "learning_rate": 1.3889677492852083e-08,
+      "loss": 0.3742,
+      "step": 3987
+    },
+    {
+      "epoch": 2.936671575846834,
+      "grad_norm": 0.2827453315258026,
+      "learning_rate": 1.3572353341442246e-08,
+      "loss": 0.3538,
+      "step": 3988
+    },
+    {
+      "epoch": 2.9374079528718706,
+      "grad_norm": 0.31764787435531616,
+      "learning_rate": 1.3258691071811325e-08,
+      "loss": 0.357,
+      "step": 3989
+    },
+    {
+      "epoch": 2.9381443298969074,
+      "grad_norm": 0.31900376081466675,
+      "learning_rate": 1.2948690914303397e-08,
+      "loss": 0.3661,
+      "step": 3990
+    },
+    {
+      "epoch": 2.9388807069219443,
+      "grad_norm": 0.32145196199417114,
+      "learning_rate": 1.2642353096573578e-08,
+      "loss": 0.3865,
+      "step": 3991
+    },
+    {
+      "epoch": 2.939617083946981,
+      "grad_norm": 0.3384080231189728,
+      "learning_rate": 1.2339677843586917e-08,
+      "loss": 0.3477,
+      "step": 3992
+    },
+    {
+      "epoch": 2.940353460972018,
+      "grad_norm": 0.30081725120544434,
+      "learning_rate": 1.2040665377618944e-08,
+      "loss": 0.38,
+      "step": 3993
+    },
+    {
+      "epoch": 2.9410898379970547,
+      "grad_norm": 0.3142537772655487,
+      "learning_rate": 1.1745315918255118e-08,
+      "loss": 0.3647,
+      "step": 3994
+    },
+    {
+      "epoch": 2.9418262150220915,
+      "grad_norm": 0.3099748194217682,
+      "learning_rate": 1.1453629682391943e-08,
+      "loss": 0.3372,
+      "step": 3995
+    },
+    {
+      "epoch": 2.9425625920471283,
+      "grad_norm": 0.32146158814430237,
+      "learning_rate": 1.1165606884234182e-08,
+      "loss": 0.3711,
+      "step": 3996
+    },
+    {
+      "epoch": 2.943298969072165,
+      "grad_norm": 0.29835745692253113,
+      "learning_rate": 1.088124773529764e-08,
+      "loss": 0.3736,
+      "step": 3997
+    },
+    {
+      "epoch": 2.944035346097202,
+      "grad_norm": 0.32554247975349426,
+      "learning_rate": 1.0600552444406387e-08,
+      "loss": 0.3506,
+      "step": 3998
+    },
+    {
+      "epoch": 2.944771723122239,
+      "grad_norm": 0.3305593430995941,
+      "learning_rate": 1.032352121769553e-08,
+      "loss": 0.3846,
+      "step": 3999
+    },
+    {
+      "epoch": 2.9455081001472756,
+      "grad_norm": 0.3391202688217163,
+      "learning_rate": 1.0050154258607336e-08,
+      "loss": 0.3827,
+      "step": 4000
+    },
+    {
+      "epoch": 2.9462444771723124,
+      "grad_norm": 0.3027651309967041,
+      "learning_rate": 9.780451767895104e-09,
+      "loss": 0.358,
+      "step": 4001
+    },
+    {
+      "epoch": 2.9469808541973492,
+      "grad_norm": 0.3510773181915283,
+      "learning_rate": 9.514413943619849e-09,
+      "loss": 0.3837,
+      "step": 4002
+    },
+    {
+      "epoch": 2.947717231222386,
+      "grad_norm": 0.30630064010620117,
+      "learning_rate": 9.252040981151956e-09,
+      "loss": 0.3691,
+      "step": 4003
+    },
+    {
+      "epoch": 2.948453608247423,
+      "grad_norm": 0.31158211827278137,
+      "learning_rate": 8.993333073169519e-09,
+      "loss": 0.3981,
+      "step": 4004
+    },
+    {
+      "epoch": 2.9491899852724597,
+      "grad_norm": 0.31192636489868164,
+      "learning_rate": 8.738290409660566e-09,
+      "loss": 0.3855,
+      "step": 4005
+    },
+    {
+      "epoch": 2.9499263622974965,
+      "grad_norm": 0.33482906222343445,
+      "learning_rate": 8.486913177920275e-09,
+      "loss": 0.3726,
+      "step": 4006
+    },
+    {
+      "epoch": 2.9506627393225333,
+      "grad_norm": 0.32285335659980774,
+      "learning_rate": 8.239201562553201e-09,
+      "loss": 0.3276,
+      "step": 4007
+    },
+    {
+      "epoch": 2.95139911634757,
+      "grad_norm": 0.31974491477012634,
+      "learning_rate": 7.99515574546994e-09,
+      "loss": 0.4194,
+      "step": 4008
+    },
+    {
+      "epoch": 2.952135493372607,
+      "grad_norm": 0.33110713958740234,
+      "learning_rate": 7.754775905891576e-09,
+      "loss": 0.3572,
+      "step": 4009
+    },
+    {
+      "epoch": 2.9528718703976438,
+      "grad_norm": 0.3111792802810669,
+      "learning_rate": 7.518062220345235e-09,
+      "loss": 0.3705,
+      "step": 4010
+    },
+    {
+      "epoch": 2.9536082474226806,
+      "grad_norm": 0.29877275228500366,
+      "learning_rate": 7.285014862666862e-09,
+      "loss": 0.3592,
+      "step": 4011
+    },
+    {
+      "epoch": 2.9543446244477174,
+      "grad_norm": 0.3104023039340973,
+      "learning_rate": 7.055634003998446e-09,
+      "loss": 0.3524,
+      "step": 4012
+    },
+    {
+      "epoch": 2.955081001472754,
+      "grad_norm": 0.3165355920791626,
+      "learning_rate": 6.829919812790797e-09,
+      "loss": 0.3818,
+      "step": 4013
+    },
+    {
+      "epoch": 2.955817378497791,
+      "grad_norm": 0.32550013065338135,
+      "learning_rate": 6.607872454801878e-09,
+      "loss": 0.3589,
+      "step": 4014
+    },
+    {
+      "epoch": 2.956553755522828,
+      "grad_norm": 0.3074639141559601,
+      "learning_rate": 6.38949209309625e-09,
+      "loss": 0.3638,
+      "step": 4015
+    },
+    {
+      "epoch": 2.9572901325478647,
+      "grad_norm": 0.316501259803772,
+      "learning_rate": 6.174778888046184e-09,
+      "loss": 0.3775,
+      "step": 4016
+    },
+    {
+      "epoch": 2.9580265095729015,
+      "grad_norm": 0.33413442969322205,
+      "learning_rate": 5.963732997329996e-09,
+      "loss": 0.3577,
+      "step": 4017
+    },
+    {
+      "epoch": 2.9587628865979383,
+      "grad_norm": 0.3284623622894287,
+      "learning_rate": 5.756354575934265e-09,
+      "loss": 0.3684,
+      "step": 4018
+    },
+    {
+      "epoch": 2.959499263622975,
+      "grad_norm": 0.32929477095603943,
+      "learning_rate": 5.552643776150501e-09,
+      "loss": 0.3902,
+      "step": 4019
+    },
+    {
+      "epoch": 2.960235640648012,
+      "grad_norm": 0.3296683728694916,
+      "learning_rate": 5.352600747577929e-09,
+      "loss": 0.3649,
+      "step": 4020
+    },
+    {
+      "epoch": 2.9609720176730487,
+      "grad_norm": 0.32667848467826843,
+      "learning_rate": 5.1562256371229245e-09,
+      "loss": 0.3842,
+      "step": 4021
+    },
+    {
+      "epoch": 2.9617083946980856,
+      "grad_norm": 0.33496248722076416,
+      "learning_rate": 4.9635185889967966e-09,
+      "loss": 0.3862,
+      "step": 4022
+    },
+    {
+      "epoch": 2.9624447717231224,
+      "grad_norm": 0.2943912744522095,
+      "learning_rate": 4.774479744717453e-09,
+      "loss": 0.3649,
+      "step": 4023
+    },
+    {
+      "epoch": 2.963181148748159,
+      "grad_norm": 0.3493749499320984,
+      "learning_rate": 4.589109243109957e-09,
+      "loss": 0.3731,
+      "step": 4024
+    },
+    {
+      "epoch": 2.963917525773196,
+      "grad_norm": 0.35220539569854736,
+      "learning_rate": 4.4074072203048605e-09,
+      "loss": 0.387,
+      "step": 4025
+    },
+    {
+      "epoch": 2.964653902798233,
+      "grad_norm": 0.3216170370578766,
+      "learning_rate": 4.2293738097376465e-09,
+      "loss": 0.403,
+      "step": 4026
+    },
+    {
+      "epoch": 2.9653902798232696,
+      "grad_norm": 0.3366626799106598,
+      "learning_rate": 4.055009142152066e-09,
+      "loss": 0.3539,
+      "step": 4027
+    },
+    {
+      "epoch": 2.9661266568483065,
+      "grad_norm": 0.3413197994232178,
+      "learning_rate": 3.884313345595137e-09,
+      "loss": 0.3747,
+      "step": 4028
+    },
+    {
+      "epoch": 2.9668630338733433,
+      "grad_norm": 0.3060031831264496,
+      "learning_rate": 3.7172865454210282e-09,
+      "loss": 0.3525,
+      "step": 4029
+    },
+    {
+      "epoch": 2.96759941089838,
+      "grad_norm": 0.3200681507587433,
+      "learning_rate": 3.553928864289402e-09,
+      "loss": 0.3714,
+      "step": 4030
+    },
+    {
+      "epoch": 2.968335787923417,
+      "grad_norm": 0.3404831886291504,
+      "learning_rate": 3.394240422164852e-09,
+      "loss": 0.3466,
+      "step": 4031
+    },
+    {
+      "epoch": 2.9690721649484537,
+      "grad_norm": 0.3348984122276306,
+      "learning_rate": 3.238221336318015e-09,
+      "loss": 0.3704,
+      "step": 4032
+    },
+    {
+      "epoch": 2.9698085419734905,
+      "grad_norm": 0.3363956809043884,
+      "learning_rate": 3.0858717213250176e-09,
+      "loss": 0.3904,
+      "step": 4033
+    },
+    {
+      "epoch": 2.9705449189985274,
+      "grad_norm": 0.3241819441318512,
+      "learning_rate": 2.9371916890658105e-09,
+      "loss": 0.3679,
+      "step": 4034
+    },
+    {
+      "epoch": 2.971281296023564,
+      "grad_norm": 0.34104812145233154,
+      "learning_rate": 2.792181348726941e-09,
+      "loss": 0.3592,
+      "step": 4035
+    },
+    {
+      "epoch": 2.972017673048601,
+      "grad_norm": 0.3131769597530365,
+      "learning_rate": 2.6508408067998926e-09,
+      "loss": 0.3541,
+      "step": 4036
+    },
+    {
+      "epoch": 2.972754050073638,
+      "grad_norm": 0.3056182265281677,
+      "learning_rate": 2.5131701670805252e-09,
+      "loss": 0.3641,
+      "step": 4037
+    },
+    {
+      "epoch": 2.9734904270986746,
+      "grad_norm": 0.3001095652580261,
+      "learning_rate": 2.379169530670744e-09,
+      "loss": 0.3596,
+      "step": 4038
+    },
+    {
+      "epoch": 2.9742268041237114,
+      "grad_norm": 0.33647263050079346,
+      "learning_rate": 2.2488389959751666e-09,
+      "loss": 0.3457,
+      "step": 4039
+    },
+    {
+      "epoch": 2.9749631811487482,
+      "grad_norm": 0.31432968378067017,
+      "learning_rate": 2.12217865870612e-09,
+      "loss": 0.3862,
+      "step": 4040
+    },
+    {
+      "epoch": 2.975699558173785,
+      "grad_norm": 0.3569965362548828,
+      "learning_rate": 1.999188611878089e-09,
+      "loss": 0.3556,
+      "step": 4041
+    },
+    {
+      "epoch": 2.976435935198822,
+      "grad_norm": 0.31356170773506165,
+      "learning_rate": 1.8798689458116025e-09,
+      "loss": 0.3892,
+      "step": 4042
+    },
+    {
+      "epoch": 2.9771723122238587,
+      "grad_norm": 0.31526947021484375,
+      "learning_rate": 1.7642197481315682e-09,
+      "loss": 0.3534,
+      "step": 4043
+    },
+    {
+      "epoch": 2.9779086892488955,
+      "grad_norm": 0.29720044136047363,
+      "learning_rate": 1.6522411037667162e-09,
+      "loss": 0.3851,
+      "step": 4044
+    },
+    {
+      "epoch": 2.9786450662739323,
+      "grad_norm": 0.2992621064186096,
+      "learning_rate": 1.5439330949518216e-09,
+      "loss": 0.3677,
+      "step": 4045
+    },
+    {
+      "epoch": 2.979381443298969,
+      "grad_norm": 0.3420858681201935,
+      "learning_rate": 1.4392958012238167e-09,
+      "loss": 0.3943,
+      "step": 4046
+    },
+    {
+      "epoch": 2.980117820324006,
+      "grad_norm": 0.3347738981246948,
+      "learning_rate": 1.338329299425678e-09,
+      "loss": 0.3472,
+      "step": 4047
+    },
+    {
+      "epoch": 2.9808541973490428,
+      "grad_norm": 0.3168458640575409,
+      "learning_rate": 1.2410336637047604e-09,
+      "loss": 0.3886,
+      "step": 4048
+    },
+    {
+      "epoch": 2.9815905743740796,
+      "grad_norm": 0.3121738135814667,
+      "learning_rate": 1.147408965511132e-09,
+      "loss": 0.3921,
+      "step": 4049
+    },
+    {
+      "epoch": 2.9823269513991164,
+      "grad_norm": 0.30733251571655273,
+      "learning_rate": 1.0574552735997945e-09,
+      "loss": 0.3648,
+      "step": 4050
+    },
+    {
+      "epoch": 2.9830633284241532,
+      "grad_norm": 0.32336708903312683,
+      "learning_rate": 9.711726540312383e-10,
+      "loss": 0.3551,
+      "step": 4051
+    },
+    {
+      "epoch": 2.98379970544919,
+      "grad_norm": 0.3032035529613495,
+      "learning_rate": 8.885611701675567e-10,
+      "loss": 0.35,
+      "step": 4052
+    },
+    {
+      "epoch": 2.984536082474227,
+      "grad_norm": 0.33436474204063416,
+      "learning_rate": 8.09620882676887e-10,
+      "loss": 0.3755,
+      "step": 4053
+    },
+    {
+      "epoch": 2.9852724594992637,
+      "grad_norm": 0.3163788318634033,
+      "learning_rate": 7.343518495300794e-10,
+      "loss": 0.372,
+      "step": 4054
+    },
+    {
+      "epoch": 2.9860088365243005,
+      "grad_norm": 0.31104227900505066,
+      "learning_rate": 6.62754126002918e-10,
+      "loss": 0.3509,
+      "step": 4055
+    },
+    {
+      "epoch": 2.9867452135493373,
+      "grad_norm": 0.3354119062423706,
+      "learning_rate": 5.948277646744549e-10,
+      "loss": 0.3447,
+      "step": 4056
+    },
+    {
+      "epoch": 2.987481590574374,
+      "grad_norm": 0.32740065455436707,
+      "learning_rate": 5.305728154275658e-10,
+      "loss": 0.3559,
+      "step": 4057
+    },
+    {
+      "epoch": 2.988217967599411,
+      "grad_norm": 0.32162371277809143,
+      "learning_rate": 4.699893254495047e-10,
+      "loss": 0.3998,
+      "step": 4058
+    },
+    {
+      "epoch": 2.9889543446244478,
+      "grad_norm": 0.3500117063522339,
+      "learning_rate": 4.1307733923079407e-10,
+      "loss": 0.3904,
+      "step": 4059
+    },
+    {
+      "epoch": 2.9896907216494846,
+      "grad_norm": 0.34684303402900696,
+      "learning_rate": 3.5983689856522453e-10,
+      "loss": 0.3776,
+      "step": 4060
+    },
+    {
+      "epoch": 2.9904270986745214,
+      "grad_norm": 0.31505316495895386,
+      "learning_rate": 3.1026804255207544e-10,
+      "loss": 0.3636,
+      "step": 4061
+    },
+    {
+      "epoch": 2.991163475699558,
+      "grad_norm": 0.3074565529823303,
+      "learning_rate": 2.643708075922291e-10,
+      "loss": 0.3477,
+      "step": 4062
+    },
+    {
+      "epoch": 2.991899852724595,
+      "grad_norm": 0.3139492869377136,
+      "learning_rate": 2.2214522739205657e-10,
+      "loss": 0.3694,
+      "step": 4063
+    },
+    {
+      "epoch": 2.992636229749632,
+      "grad_norm": 0.33408764004707336,
+      "learning_rate": 1.835913329600869e-10,
+      "loss": 0.4098,
+      "step": 4064
+    },
+    {
+      "epoch": 2.9933726067746687,
+      "grad_norm": 0.33919352293014526,
+      "learning_rate": 1.487091526097828e-10,
+      "loss": 0.3638,
+      "step": 4065
+    },
+    {
+      "epoch": 2.9941089837997055,
+      "grad_norm": 0.30452626943588257,
+      "learning_rate": 1.174987119573201e-10,
+      "loss": 0.356,
+      "step": 4066
+    },
+    {
+      "epoch": 2.9948453608247423,
+      "grad_norm": 0.32732394337654114,
+      "learning_rate": 8.996003392214292e-11,
+      "loss": 0.374,
+      "step": 4067
+    },
+    {
+      "epoch": 2.995581737849779,
+      "grad_norm": 0.2973012626171112,
+      "learning_rate": 6.609313872862899e-11,
+      "loss": 0.3513,
+      "step": 4068
+    },
+    {
+      "epoch": 2.996318114874816,
+      "grad_norm": 0.30825671553611755,
+      "learning_rate": 4.5898043903314096e-11,
+      "loss": 0.3687,
+      "step": 4069
+    },
+    {
+      "epoch": 2.9970544918998527,
+      "grad_norm": 0.3054065406322479,
+      "learning_rate": 2.93747642771125e-11,
+      "loss": 0.3703,
+      "step": 4070
+    },
+    {
+      "epoch": 2.9977908689248896,
+      "grad_norm": 0.3413209617137909,
+      "learning_rate": 1.6523311984206757e-11,
+      "loss": 0.3665,
+      "step": 4071
+    },
+    {
+      "epoch": 2.9985272459499264,
+      "grad_norm": 0.3274330496788025,
+      "learning_rate": 7.343696462047689e-12,
+      "loss": 0.3562,
+      "step": 4072
+    },
+    {
+      "epoch": 2.999263622974963,
+      "grad_norm": 0.31068921089172363,
+      "learning_rate": 1.8359244524646103e-12,
+      "loss": 0.3785,
+      "step": 4073
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.31857830286026,
+      "learning_rate": 0.0,
+      "loss": 0.344,
+      "step": 4074
+    },
+    {
+      "epoch": 3.0,
+      "step": 4074,
+      "total_flos": 4823744474644480.0,
+      "train_loss": 0.42331435592460115,
+      "train_runtime": 136035.2407,
+      "train_samples_per_second": 2.875,
+      "train_steps_per_second": 0.03
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4074,
+  "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": 4823744474644480.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}