diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8687 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 1235,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004048582995951417,
+      "grad_norm": 6.484579185933172,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.1094,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 6.617022941628354,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.1209,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 6.662924201928013,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.1315,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 6.661706257823938,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 1.105,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 6.025688721812807,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.089,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 5.9350161799594945,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.102,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 4.644351394890572,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 1.0793,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 4.512997435651717,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 1.0583,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 2.7993082864808665,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 1.0076,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.5293546485423404,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.9911,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.403341165678143,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.9835,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 3.30440299181442,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.9834,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 3.7010464902841362,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.9684,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 3.2907398235480194,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.9695,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 3.0427907387400497,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.9335,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 2.1700780738389858,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.9227,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.204913857437493,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.8835,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.0441190562697376,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.8897,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.6846992426792988,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.8578,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.4224709699839442,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.8603,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.1681203002888287,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.8387,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.2999851993408402,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.8413,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.3468619174580614,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.8163,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.1053357034527205,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.8219,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 0.9055851535196511,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.8004,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.0170462585352147,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.8112,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 1.0444487107765492,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.7809,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 0.8683644041815828,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.7836,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 0.7857851964751171,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.7989,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.8160726880376635,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.8081,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.8994307790457099,
+      "learning_rate": 1e-05,
+      "loss": 0.7761,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.736886782984379,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.7723,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.703934638601186,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.7589,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.630904136720354,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.7332,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.8036739027867097,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.7433,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.6980293344317294,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.7758,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.6143857235137982,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.7562,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.7034065299862696,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.7351,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.7143269640648959,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.768,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.7028781845638217,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.7363,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.6696561580172864,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.7191,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.6314352827931888,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.7454,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.8008594960500123,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.7288,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.6180713913599598,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.724,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.6869881818187865,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.7434,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.6652255245980918,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.7335,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.6591248933291005,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.7219,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.6865925211122758,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.7222,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.6951463753129005,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.7302,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.7801325507012549,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.7221,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 1.154578277190581,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.7204,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.8083325724981575,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.7161,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.5971852045852191,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.7303,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.7656377572193656,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.7054,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.6312264301685111,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.7283,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.8101147177622084,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.7148,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.6112815556356385,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.7542,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.8648466686152165,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.7409,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.7399529686775608,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.7108,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.7178352729659243,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.7057,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.7796310054132842,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.6929,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.672222369434276,
+      "learning_rate": 2e-05,
+      "loss": 0.6804,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.6886821078906318,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.7,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 1.3618669533476653,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.7684,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.7901178552945111,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.696,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.787314845207934,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.6951,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.7507268725777138,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.725,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.6345103743646076,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.6912,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 0.7362623863400026,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.7132,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.6295914105205163,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.7296,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 0.7330608978298979,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.6915,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 0.6727874815348832,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.6799,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 0.67454478461001,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.6957,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 0.7412010508646389,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.6869,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 0.8278071677362511,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.6946,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.6342307220910951,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.7011,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 0.657770360428558,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.6767,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.6892821947919512,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.6969,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 0.6026633911033683,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.7084,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 0.7309390839975506,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.701,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 0.6679270808850089,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.7129,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 0.7654765394293027,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.7045,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 0.633918499988127,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.7235,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 0.6661305437549405,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.6984,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 0.82635773377921,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.7008,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 0.7069904989929792,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.701,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 0.6615263423033075,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.6972,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 0.6646713838842491,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.6976,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 0.793789133919017,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.7007,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 0.7570016716519147,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.6976,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.7531279744246266,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.6935,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 0.647262645082105,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.7222,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 0.6960519051005202,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.6926,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 0.794446406401357,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.6932,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.0906713214562083,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.6758,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 0.84058388657974,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.6848,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 0.8621993994624154,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.6982,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 1.513631813165153,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.7062,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 0.7290819892746895,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.6998,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 1.691321514723803,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.7009,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 0.9025336387327216,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.7203,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 1.5882019918127517,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.6829,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 1.0247680579445075,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.6728,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.7187568501950077,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.7073,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 1.290878625415404,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.7112,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 1.3269348111159915,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.6874,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.4304888447343338,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.6963,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 1.098387911388081,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.725,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 1.4076166838084547,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.7128,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 0.9519149994182035,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.6939,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 1.5243116185892456,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.683,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 1.117205875673437,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.7162,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.6510920796446602,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.6946,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.3991312193429586,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.6838,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 1.126057686727526,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.7026,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 1.0735231176404372,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.7031,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.9858436920679089,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.6948,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 0.8266277290395218,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.6911,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 0.814264467938799,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.6717,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 1.2690766321358433,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.6949,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 0.7077786664515173,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.6991,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 1.1017877578988333,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.668,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 0.9329027486052942,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.7109,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 0.6781953265388925,
+      "learning_rate": 4e-05,
+      "loss": 0.6975,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 0.8373649141658054,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.6714,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 0.7436557033092351,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.6792,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 0.9191118720715192,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.683,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.0992245479353078,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.687,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 0.6476620667995538,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.7051,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 1.036901021182914,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.7306,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 0.7508671769196622,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.6986,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 0.8281007822464226,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.7046,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.6517487821552678,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.6877,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 0.8419986626422321,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.7058,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 0.8828419313696348,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.6968,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.6951009647124928,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.694,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 0.6695319422272823,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.6716,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 0.7468117523212147,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.6645,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 0.8852678717344173,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.6898,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 0.7242406098675795,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.6584,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 0.8263557459415484,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.7048,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 0.8563202373867324,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.7145,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.773527147014368,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.692,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 0.7895630926736722,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.6792,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.7885932367425502,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.6897,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 0.807128608893747,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.7086,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.6620213509241065,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.687,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 0.5959144737985186,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.675,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.794333359336447,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.6914,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.7869619107999738,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.6777,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 0.7718728487753661,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.6878,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.7760903009045681,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.6894,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 1.00177908258274,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.7059,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.7671554586215122,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.7006,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.7796036276787921,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.6921,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.9697675421624615,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.6704,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.6677957381059069,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.703,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.7744777298830986,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.6872,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.6811123783844539,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.7015,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.7152006954797135,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.6899,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.7050009090751465,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.6985,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.6525354784000442,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.6878,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.69247713652762,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.673,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.604042586828125,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.6738,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.6573705228168397,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.7083,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.6437026080984879,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.6932,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.8127062753424059,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.6547,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.5954252665099279,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.6937,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.6648077973479983,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.7263,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.5700986856297189,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.6986,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 2.311963847239052,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.6963,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.7468508418039908,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.6946,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.6638599328606124,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.7147,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.8487111416586343,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.6844,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.6092433640069226,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.6871,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.7449784967047385,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.6712,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.6313628519724547,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.6658,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.6762154104250356,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.6682,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.6534632945187169,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.688,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.5521618458315415,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.6787,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.776840557090752,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.675,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.5478880808435369,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.656,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.729108043056085,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.6789,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.5866027806662433,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.6843,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.7681788561261357,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.7083,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.7791547090290452,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.6908,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.60547054691224,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.6805,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.5849627425103077,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.684,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.6102591798285169,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.6999,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.5342897144943202,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.7016,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.710896772846417,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.6673,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.6391343145395507,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.6823,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.8519719785826291,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.6557,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.6655423365858544,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.6952,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.8587778497695235,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.7068,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.7368459830058872,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.7142,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.6479575721440707,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.6802,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.629178497999839,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.6732,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.5931292471043306,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.6701,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.6021480183595456,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.6982,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.6022056528354597,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.6776,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.6659697931716496,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.7063,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.511434405104951,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.6917,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.6745310771399383,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.6929,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.6496111091618253,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.6853,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.5483919344328376,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.6756,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.7173615053132074,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.6821,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.5283310429152404,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.6941,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.7011614481619741,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.6751,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.7530847582999003,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.688,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.5235212210684753,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.6933,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.6591033705326661,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.6779,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.5525169356070281,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.6624,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.682270680302631,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.686,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.5788568789785198,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.6699,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.647390072207996,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.6736,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.5501319628106675,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.6763,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.5268629125244858,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.6385,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.6034223504752916,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.7029,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.5658452025240828,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.6554,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.5454312376291289,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.6892,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.505101847952586,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.6641,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.5353142793683752,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.6586,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.4801427668389248,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.6671,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.4756724257975903,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.6735,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.5205083459798991,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.6837,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.48318809494100023,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.6695,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.4772635027059648,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.6878,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.5354330272055516,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.6716,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.5272319468161022,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.6797,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.5471017960926458,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.6729,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.5741296357683981,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.6819,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.5689544114299873,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.6751,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.5983541310926412,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.6707,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.6299402497862715,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.6808,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.6015309275347069,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.6595,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.5682009527422395,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.6701,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.5822615348467381,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.6947,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.6772515706433874,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.6723,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.715047113316924,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.6505,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.5600612536366489,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.6548,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.6861865577279129,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.6817,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.5411167502954892,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.6724,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.711054546045831,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.6843,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.6551012462227631,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.6569,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.6681937843830188,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.6948,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.725495647759067,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.6616,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.8266046278155461,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.5774,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.5676326792056207,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.5414,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.9009656658607629,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.5554,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.7066441224417254,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.5391,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.652352473751489,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.5548,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.6367386652642419,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.5093,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.6518278060366186,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.546,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.5786038884923488,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.5553,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.5480609640839307,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.5533,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.6650915608564529,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.5687,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.6040651568269516,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.5519,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.7184626541811912,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.5762,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.606870418408378,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.5366,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.6574675072374896,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.5444,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.5607313837471323,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.5314,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.8088279028590005,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.5815,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.5756292601491743,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.5421,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.6659241955928588,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.5506,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.6167111193850427,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.5542,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.5374230001885213,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.5499,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.5940053294235037,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.5637,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.6134201065568358,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.5285,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.5395874090508638,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.5504,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.6137850979427271,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.5307,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.5568309781423528,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.5349,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.5648622232727647,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.5408,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.5142139406175281,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.5528,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.5707023239985929,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.5387,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.4841045420809928,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.5472,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.5319763029184214,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.5396,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.5277748814369149,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.5595,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.7145760757382228,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.5667,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.5596550585793025,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.5431,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.6213335431227177,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.5541,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.5635639083685097,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.5517,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.6255812938321925,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.5568,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.5116771827201939,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.5444,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.6263218839991387,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.5504,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.4536488951613987,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.5419,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.5406947388603593,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.5295,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.5717484013110897,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.5525,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.48160387276536504,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.5406,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.5134436134943037,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.5419,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.49029891384881946,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.5296,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.5527906062357564,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.5509,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.4562022962860242,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.5494,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.5184641628801195,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.5448,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.4855879989430507,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.5517,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.5072831404918348,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.5531,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.4611846306027754,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.5508,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.516256301082155,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.529,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.5578937335560691,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.5726,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.5964186993052121,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.5483,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.531936619542025,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.5423,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.5675687191791642,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.5263,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.5950778312763378,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.5479,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.5371770262248664,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.5642,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.5856253533167985,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.5468,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.4769547256126986,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.5526,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.5613300018701136,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.5655,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.509840533177327,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.5535,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.5498889033498204,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.5468,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.48225297279502366,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.5613,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.60549105732761,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.5564,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.5163772200265605,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.5529,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.5586444845932336,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.5532,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.5674254669104768,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.5401,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.5621272808971483,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.553,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.537317757374782,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.5751,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.7000861274654362,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.5669,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.4824849096087633,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.5374,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.5847800000284845,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.5476,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.5836502854714282,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.5679,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.5571533312667496,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.5469,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.5442239400027419,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.5532,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.8189221086782976,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.5368,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.4969141043459774,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.5378,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.5709768881166456,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.5377,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.5294302279630771,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.5399,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.5417625452164385,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.5753,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.5722699135679572,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.5236,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.4893960002176216,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.5281,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.5545368935954407,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.5555,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.523906730593,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.5494,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.4764836390312042,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.5505,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.5667350466020525,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.5636,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.6382955008195805,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.5617,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.5232984321939417,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.5349,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.5272809315937347,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.5614,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.4835519314501256,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.5603,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.5876991709749676,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.5528,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.507626151238949,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.5368,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.5672597441941727,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.5412,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.514233611688224,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.5698,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.5800115972895205,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.5849,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.5028154130352466,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.5633,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.5820552757375486,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.5486,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.5565428604749887,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.5482,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.5841095991442146,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.5601,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.6942958148271219,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.5532,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.6018900929564824,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.5478,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.5312371044894805,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.5569,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.6283736456315372,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.5778,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.5946579684317745,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.5683,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.6022767011383001,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.5582,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.575421337065877,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.5428,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.5404327774325272,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.5413,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.5289565790356882,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.5506,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.5265402031236959,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.5404,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.49412605689684186,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.5321,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.4855099374495374,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.5538,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.5112616592991966,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.5349,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.5011264415230071,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.5369,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.5760382926398183,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.5584,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.5021613396691542,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.5591,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.49747871183215653,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.5798,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.504115720612789,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.5514,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.5355443920992473,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.5603,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.48762650369988436,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.5512,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.532952121094694,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.5402,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.46701666394824565,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.5538,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.5205871067199865,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.5416,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.47652005326837926,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.5544,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.470748860876292,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.5621,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.5642112276527179,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.5517,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.4967242527234315,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.5436,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 1.1894039068841427,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.5744,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.5735052025220339,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.5418,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.5275675291935679,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.55,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.6525385829691661,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.5429,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.4980007177460864,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.569,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.6719186202515635,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.542,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.5166012771361573,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.5511,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.6033172321513353,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.5375,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.5083684940917925,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.5356,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.516561806838242,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.5423,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.5219561578855523,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.5687,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.5128819400795702,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.5684,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.47507144253577044,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.5525,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.5426477442182387,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.5403,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.44406844033117565,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.5452,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.5992589369029162,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.5443,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.44977019801421814,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.5625,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.654556076110489,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.5457,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.4280021495516217,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.5496,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.6780442604973201,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.5523,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.4516001936473416,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.5466,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.5809842703780619,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.5501,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.4841600745738762,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.5652,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.5395300085653645,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.5587,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.5073974373064409,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.5515,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.5548417320959066,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.5501,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.4633546379834556,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.564,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.6022544813349208,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.5794,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.49992900579012095,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.5538,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.5614950574415751,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.559,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.531412891029991,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.5713,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.5128358250192541,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.5677,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.5003798554072322,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.5602,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.5845325670533724,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.5516,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.4956660715778495,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.5347,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.5459674601535282,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.5333,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.5538494554612672,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.5499,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.6361662648328897,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.5712,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.5696119581562389,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.5453,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.5847477756643229,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.5605,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.6090542708593046,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.5518,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.5051898862977982,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.5508,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.5586688914322674,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.5566,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.4790883559148496,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.5241,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.614390015021991,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.5498,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.4985939836516929,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.5649,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.5752038704620823,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.5473,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 12.27769641766638,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.5355,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.9237518686688798,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.5641,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.6806422253957486,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.5592,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.7764895825276286,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.5678,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.8931589228808706,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.5698,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.5996727359150731,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.5496,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.7799577040527093,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.5565,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.6278703885607912,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.5664,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.6001891502792126,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.5431,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.570805949713383,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.5435,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.5400196620679605,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.569,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.5913407414389906,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.5436,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.642216271299973,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.5428,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.524614840788624,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.5482,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.4788532715972934,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.561,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.5390929189258838,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.559,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.4624600094115747,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.5543,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.5593822754459783,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.5579,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.46703819957096593,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.5564,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.5259825599320421,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.5632,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.429578179804886,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.5391,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.5314973014013452,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.5637,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.4398280773186812,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.5455,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.4822172070737945,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.5586,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.4290658181479459,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.5303,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.4992930580122448,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.56,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.43158780107268885,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.5539,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.5046268020284668,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.5468,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.48018135479692264,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.5714,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.5170404782720023,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.537,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.4698812650345225,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.5575,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.48650666722276953,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.5696,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.46047973797879294,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.551,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.4380867219403287,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.5867,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.4702725570848873,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.5622,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.4262075811663244,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.561,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.4292040849193712,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.5326,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.43080142447364517,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.5566,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.44216611490702534,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.5363,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.5216027804511066,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.547,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.4624722554598115,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.5367,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.5494993563684403,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.5423,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.4821509256741741,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.5444,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.5351608782888926,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.5546,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.5142666559625834,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.5308,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.5695775604339967,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.5518,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.48454109587779715,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.5438,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.5917738311340052,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.5527,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.43182064241486634,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.534,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.5727531011904301,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.5419,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.4500779982227152,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.5247,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.4215336279858491,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.5551,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.5685441717051013,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.5464,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.4513146131133903,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.5616,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.4822066939609388,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.5508,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.4598697893521902,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.5423,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.47989322325108913,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.5403,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.47522928479829113,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.5604,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.46824416702952626,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.5756,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.4700338189494635,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.5367,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.5529278146779912,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.5485,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.45085341311043686,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.5498,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.5339892995954562,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.5558,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.47360578345124243,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.5582,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.4720765903965615,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.5615,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.5810223310350013,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.5641,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.5719413965629085,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.5369,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.5102068841738954,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.5414,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.5367808773679363,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.5435,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.5982092542908997,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.5518,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.5225572654529751,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.5497,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.5651730670526596,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.5561,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.48860084958478817,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.5563,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.4680187625871907,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.5639,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.6543359590521888,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.4279,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.5103469478884258,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.4382,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.8411819658402788,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.4228,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.6597831288745851,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.4,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.6267132712424154,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.4359,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.5650948440820962,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.4096,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 1.7968795468760914,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.4799,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.675823574847661,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.3881,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.506383796393,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.4188,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.57776966428706,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.4071,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.4952416438428652,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.4421,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.5416253252506429,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.403,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.4869150373965809,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.4417,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.5582209558309273,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.4294,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.5606068710949789,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.4277,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.5056833593966326,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.4045,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.5276461418901633,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.4156,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.496957006483615,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.415,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.4107456013513719,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.3965,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.4937970862107908,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.4048,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.4370952339670076,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.3877,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.4460869348796453,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.4315,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.41192126916440724,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.4272,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.4571064772832866,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.4136,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.4187993449993884,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.4155,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.4789052101279257,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.4113,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.40434864757414224,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.4058,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.4535952088958238,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.4019,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.41337237177148967,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.3916,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.4698424745631681,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.4396,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.4088028988404592,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.4244,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.42607026510463986,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.4349,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.487019902323066,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.408,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.4112882749015096,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.436,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.467839599076156,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.4265,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.463576133624571,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.4229,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.4643096240052505,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.4325,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.4080245525270853,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.3928,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.43749430726322,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.4244,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.44110824476711435,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.4137,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.3970215300437868,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.4348,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.461410651689528,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.3951,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.39846430468703214,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.3983,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.5009312551429812,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.4357,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.42662069585941986,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.4227,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.4741269116244328,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.422,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.44049600212386075,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.4016,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.4446974325608274,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.4274,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.4175553430685666,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.4013,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.4425101314174278,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.4263,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.4370859660590963,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.4252,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.43077842246520537,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.4024,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.4967943328412564,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.4126,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.4423605099711317,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.42,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.41462561657267394,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.4231,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.45722872123281627,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.4303,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.39009026729082663,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.4384,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.45190535137511323,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.3975,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.43205350869431997,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.4175,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.4323880484339317,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.4095,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.4383131032218113,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.4142,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.4686763617901195,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.4503,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.4406261951129377,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.4141,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.5798771170031229,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.4178,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.4171307253513814,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.4088,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.5707588748826273,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.4224,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.4303339696455761,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.4069,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.5443837725570546,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.4129,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.4084964117457437,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.4428,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.48523184616633963,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.4068,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.41067640596833205,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.4326,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.4239934325815212,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.3991,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.4147515211192254,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.4129,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.4345516536583445,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.4389,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.4152296648917417,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.3988,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.42994279516534706,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.4196,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.3881335952927788,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.4097,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.44343840136840224,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.4018,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.4657718272762125,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.3885,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.4405927724860087,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.398,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.48643162375730553,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.4198,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.39236150950127024,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.4469,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.4812529105861061,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.4215,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.39887302778240147,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.4154,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.4554489662352036,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.4104,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.414958036203445,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.435,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.39995931032883464,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.3959,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.4244446949063031,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.4268,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.4352146449916987,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.4083,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.4100772023751195,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.3885,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.40800689299227316,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.3883,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.4391423395463649,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.3796,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.39129538812709064,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.3907,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.5417559050922992,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.401,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.45332819887248654,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.4104,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.4899011452338315,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.4413,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.5372335118832429,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.415,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.4309687548072833,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.4117,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.4868185705336043,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.4058,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.4027882258058427,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.4115,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.4703035259129809,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.4168,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.415189420766769,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.4109,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.4599148612893004,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.411,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.3986801336105369,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.4197,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.41845653271313016,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.428,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.39862696270002057,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.4232,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.42952422004650487,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.4232,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.43452122964129986,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.4005,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.46490325318679965,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.4072,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.41531561146026685,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.3976,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.41109459113596375,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.3948,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.4178680069897879,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.3869,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.40831199087541076,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.4346,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.43667834548758877,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.4481,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.42758160932289285,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.4182,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.41442786710428337,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.4298,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.44157247809779265,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.4181,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.4882883765072915,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.417,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.41006784295181303,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.3992,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.44712904804042464,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.4223,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.4507737266253266,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.4421,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.4271515204579643,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.3953,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.4509950891471527,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.4082,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.4178920569123137,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.3918,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.44272167343288393,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.4531,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.40244175355118417,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.4107,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.4483581979763629,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.4309,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.42398752379839916,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.3762,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.4462973391128629,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.4241,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.45137401406474464,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.4173,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.3769761880505656,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.43,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.4515342533132135,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.4342,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.41061289100704634,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.4176,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.3966056382585618,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.4293,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.4290096129364181,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.4191,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.4254812033954116,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.4199,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.4133521086205344,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.4003,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.42115596161719493,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.4092,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.44379099049175885,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.4088,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.3729542466309837,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.4376,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.39235767445223313,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.4313,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.42907223166034625,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.4181,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.43808627849782195,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.4028,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.4437739678520603,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.411,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.4196181553399626,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.4288,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.42744551450208984,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.4139,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.3651810106796195,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.4206,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.41340626122325114,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.427,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.38493697561765244,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.4144,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.4071022440710037,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.4078,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.379021556900768,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.4266,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.4085518727338247,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.4195,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.5036096373407274,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.4243,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.35704558496169664,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.4057,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.9884493192034737,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.4,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.3966759148878023,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.4091,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.3909403021432739,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.3851,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.43661994383890357,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.42,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.3979659891246475,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.4124,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.41687610176996154,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.4288,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.3800196835168986,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.3994,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.43240114190070744,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.414,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.40085160832447425,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.418,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.4357254189566814,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.3932,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.4650512852393052,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.4029,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.3890543433681168,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.4197,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.4183566194198221,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.4078,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.3877929647685638,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.426,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.40235313359383174,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.4225,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.3624623047971135,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.4049,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.3846201092679407,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.3945,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.3963652431758373,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.4289,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.42061862998727373,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.3884,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.4569400669115899,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.4158,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.40848386558012956,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.4178,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.4510470062363927,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.412,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.36828541521890584,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.422,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.4158672617159063,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.4191,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.39681962836842144,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.4162,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.4378653534821604,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.3689,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.4380116786030519,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.4076,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.4259563581058558,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.4135,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.4805526516778331,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.4187,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.39438163468442494,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.4161,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.3928632736724274,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.4081,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.38752047141597756,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.4293,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.44103870818625657,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.4139,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.4097021735081435,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.4302,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.37132919063988234,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.4326,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.43976403611720044,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.3997,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.4002471205002559,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.4236,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.3773843517004388,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.4079,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.44127460829942206,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.4201,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.3849728311590267,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.4131,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.40345231069168846,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.4143,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.37153725003869786,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.4036,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.38890420491248245,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.4005,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.37629947330273017,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.4152,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.3805409268795865,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.4049,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.40558596892275633,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.3971,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.4228854783974626,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.4218,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.38901416235108194,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.4337,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.3939710889338497,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.4107,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.39347942713053224,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.3906,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.40171171684170265,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.423,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.41844595903450343,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.3979,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.3963883478177424,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.4124,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.3992500019018458,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.4344,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.4049474984314642,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.4147,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.3505736101844277,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.4338,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.3969008041311653,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.4277,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.39144753777623664,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.4176,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.38650654484034175,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.4151,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.3751077469214528,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.4299,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.38415275996890275,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.4353,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.40291872384034977,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.3972,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.3790019994591881,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.4188,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.39158211608107013,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.4184,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.40559908632496106,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.4049,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.4103989486679989,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.4138,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.3963491788338057,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.4341,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.3977862513016425,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.4188,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.38483545432399435,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.4485,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.41688628972577285,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.4051,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.3945580617435192,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.4133,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.3955496929990669,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.4115,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.39142796801224067,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.4162,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.40580216576132927,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.4171,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.3849486145498053,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.4064,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.39737969891078023,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.4097,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.3925000686791272,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.3974,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.38782564497220634,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.4199,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.39475510533382324,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.3969,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.36989491468140523,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.4294,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.43300467098389556,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.4126,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.3799224737739026,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.4069,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.39970990380105403,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.4101,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.44029624914617965,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.4257,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.3683218219812328,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.4087,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.3705605005915206,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.4199,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.37567730178066056,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.4164,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.3663680896208593,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.4127,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.3619537546723646,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.4016,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.35205458700486275,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.4114,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.37970832115815095,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.4149,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.3823576855004118,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.4255,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.37106000967712366,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.4182,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.5410672660670022,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.3037,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.3910819711912579,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.3098,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.6220318624443132,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.3237,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.5392776483203106,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.3087,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.4419407239320347,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.3245,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.5527849680340257,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.3237,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.4484749593731717,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.3496,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.4763753006526932,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.2978,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.4911529672613926,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.3247,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.4632992375473752,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.3389,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.4437156092800971,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.2941,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.3894756816564006,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.3067,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.4212095504536426,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.3126,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.4304429727768482,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.283,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.35804667491259545,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.3085,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.41140826433300326,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.3057,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.42758691623569456,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.3241,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.36202919668183847,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.3206,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.42507556891560344,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.3148,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.38833785907055957,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.294,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.358185749901586,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.3305,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.42815251223540074,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.3507,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.38654157633533637,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.2988,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.3494794766483416,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.3193,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.3764714058530579,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.3294,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.3932384474545768,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.3308,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.35723498607539644,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.3229,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.3502016550565653,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.2946,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.3650309781839943,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.2924,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.3847927660189986,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.3101,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.3568705715172085,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.3009,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.3850455427955741,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.3205,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.37713059378816416,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.3325,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.3421945094583891,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.3337,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.3664482992290827,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.3101,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.3456330674318976,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.3067,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.3497817081634653,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.3001,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.3633649347718385,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.3147,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.35870077735359673,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.3189,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.3273172428632415,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.3122,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.34957911209187204,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.3213,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.34722797278553913,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.3351,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.33855182459265065,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.3302,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.34274991828080276,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.2955,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.33750356667257186,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.2894,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.33698870362751265,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.3066,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.3319047749504512,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.2978,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.35316471239078806,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.2937,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.31958171952979303,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.3437,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.7384675711934252,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.3422,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.33970576865754354,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.2858,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.3340264501112379,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.3425,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.3453711144257807,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.2822,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.3860621073104159,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.3289,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.33295381737065394,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.3122,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.3392403092610673,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.3073,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.37513892903587714,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.3173,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.3696329393825166,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.299,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.32548325490695545,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.3426,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.3323687035461251,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.2841,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.3609929904398247,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.3304,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.31344980159045477,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.31,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.34184571413836456,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.2895,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.33162059058364657,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.295,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.3360515008827838,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.3008,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.33550614255229316,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.3164,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.3509887332498973,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.3157,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.31241322160602314,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.2957,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.3268688730609065,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.2945,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.3174245100667965,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.3174,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.315069201776678,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.3123,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.32214561021155597,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.3286,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.305718213809787,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.2904,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.3325866284512572,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.3355,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.32142336814261774,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.3325,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.3142480249666165,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.3277,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.34217648966529596,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.3045,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.3083543446768173,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.3348,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.3193420359183561,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.3307,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.321278090920803,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.3219,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.3239110877036613,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.3071,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.31587462589848275,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.3044,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.32541714864108273,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.3333,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.3303523425555252,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.2871,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.3248029915178213,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.3149,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.3151121276554887,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.3213,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.33062732646030835,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.3224,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.3140621205346744,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.3136,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.32721423561566354,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.3065,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.33262259328588284,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.3263,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.3418171809855034,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.3346,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.3316416341681999,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.3146,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.344706802158628,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.314,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.3452879251026902,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.3253,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.32507667872335466,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.2798,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.3454979283347759,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.3013,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.32328529188727706,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.318,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.3322693265055345,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.2991,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.31460708735979503,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.3181,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.3085315363205929,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.3219,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.3258380004939168,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.3188,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.30864795111488036,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.3216,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.3158936331048464,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.3141,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.3288804813052583,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.3108,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.3329267894885101,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.304,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.3499714396377572,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.3237,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.3310607017594345,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.2939,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.32098677614221854,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.3236,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.3419287916823644,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.3308,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.30161571187858954,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.3411,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.3105155140990893,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.3242,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.3334740671862213,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.3334,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.30986371380036554,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.3429,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.32916850920364327,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.3093,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.3148002220619016,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.3101,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.33213623295291594,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.2814,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.3035071105854426,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.3227,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.3081994894789397,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.2967,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.34986731413225935,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.3364,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.3297387495556014,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.3315,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.32985331168268545,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.293,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.3357390216203154,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.29,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.33655608354284466,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.2988,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.32720200163847274,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.3159,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.3220648778331958,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.305,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.316611617305275,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.2895,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.3389101511389327,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.329,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.34487514538188246,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.3293,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.3228571613356772,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.313,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.3156068514087194,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.3168,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.327851857972054,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.3008,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.32602527867637193,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.3138,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.3132875016858712,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.3151,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.3115862525987803,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.3243,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.2975646962583945,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.3092,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.3253668589619297,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.3145,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.3175773879356347,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.3173,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.31390106987595484,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.3045,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.3048323451473413,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.297,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.32301056777957965,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.3163,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.32125109958117665,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.2968,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.308790182905006,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.3508,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.3188042953333013,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.3147,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.3103655238204178,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.2722,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.3113908613823902,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.3237,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.3076506191818626,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.2843,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.3243205425965125,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.331,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 1.3640070200541061,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.309,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.3072820149774452,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.3067,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.30819243979608424,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.3094,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.31939045633714375,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.315,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.3079138544755003,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.317,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.32935232367569117,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.3343,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.3141066583920061,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.301,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.2995505733108559,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.3154,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.31710717583440123,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.2767,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.3144459257653098,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.3083,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.32436250519969895,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.327,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.31819006875167977,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.296,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.332733536032387,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.299,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.34023716529536635,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.2691,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.32177996586311164,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.3268,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.31021298238218564,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.2971,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.3086494934759478,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.3163,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.309976028252943,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.3083,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.3052376656302563,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.329,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.3147876242707729,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.3009,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.3220937769073598,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.2967,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.32714407148434854,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.3272,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.28987147732920016,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.3295,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.31195587746522485,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.3195,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.30816350713241913,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.3119,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.3116116899835858,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.3319,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.2979054635305884,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.3278,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.3216774187236809,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.302,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.30447727656387324,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.3148,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.30715184559815645,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.2857,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.3070228742537361,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.2749,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.30396934650873875,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.3217,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.31479000685036795,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.3183,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.3077009417996169,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.3095,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.30764751040406035,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.2929,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.3252027182147758,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.3273,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.32655931159659213,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.3151,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.2948261818375066,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.3368,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.3033903911684418,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.3129,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.29736832522930556,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.3102,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.29358470972034045,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.3322,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.303027882219726,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.3365,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.30848860156550667,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.306,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.29366052790870845,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.2903,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.3049597566729291,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.3135,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.31983927617980235,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.2986,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.30214091462729203,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.2814,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.3065722740131642,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.2802,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.30319333697984563,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.3129,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.2973891308128764,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.3159,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.31699555597449036,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.2817,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.3136925480942107,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.2957,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.3198810480805276,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.2779,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.3021262209730387,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.3023,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.3011305689532702,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.2957,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.31111873882342855,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.313,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.31581288963877485,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.3283,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.2970306133738865,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.3163,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.30892590325460473,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.2844,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.29870390891932724,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.3196,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.3376270012097709,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.2927,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.30813159524860484,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.2999,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.2868952761286795,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.3559,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.30724719734336525,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.3037,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.31352696569352745,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.2932,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.3083509397572283,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.3042,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.2881834213449572,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.3328,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.29354699893468256,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.2905,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.29537369030152405,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.3123,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.3125404463262918,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.3237,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.3046339540848886,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.3031,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.2928058606501109,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.3193,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.3128479106055163,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.316,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.30224919052185256,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.2969,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.2906598228303833,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.3192,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.30322285195038223,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.3067,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.30982893616226037,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.3176,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.30829851709440853,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.3352,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.32444214732295273,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.2618,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.30917196363951494,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.3051,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.31632363801863306,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.3025,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.297026009719188,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.3411,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.3211588552303353,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.28,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.3011344308979398,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.3258,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.30201311173309525,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.287,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.30076315088288696,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.2991,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.3108122184750787,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.2817,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.38194143114721146,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.3089,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.31472935897950116,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.3183,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.302088040980906,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.2914,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.30565573483146863,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.3194,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.30367517486123985,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.3075,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.31134559203760087,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.3285,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.3267243505300939,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.3006,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.29324175616280745,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.3123,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.2961266461574495,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.3071,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.31036535993096565,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.3003,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.3245253992863963,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.2929,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.31593763785708845,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.2937,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.2934532987919655,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.307,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.45943267957730816,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.2735,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.3175099426393721,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.2658,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.28985762793896147,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.2728,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.2960120434366301,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.2798,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.348503962274494,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.2484,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.40555322571970687,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.2766,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.37147222541877073,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.2682,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.30418399609823543,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.264,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.31657670565693113,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.2802,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.3360094674330084,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.258,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.332161939824591,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.2768,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.3103027910198252,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.2628,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.28315513090429634,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.2707,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.2805568057351615,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.2685,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.291424191650973,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.2599,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.32341929451748325,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.2552,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.2960803932588185,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.234,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.2815776694391159,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.2538,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.2704024825004202,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.2511,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.2875722112512841,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.2399,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.2748927882010255,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.2658,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.2862977209452845,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.2669,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.2806987473596226,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.2571,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.28175465591067095,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.256,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.26938055030931946,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.2506,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.2655697946803227,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.2806,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.2620957137026146,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.2702,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.2502684079710666,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.2238,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.2693574770911405,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.2944,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.2734083016287476,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.269,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.24318546141952802,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.2916,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.24637918451138283,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.2542,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.2596357192139629,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.2803,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.2613524125265808,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.2577,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.2690915077227642,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.2876,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.27221463889285,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.245,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.2565255371262271,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.2804,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.261243408094994,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.2232,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.2796305900851381,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.2719,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.2724573992540577,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.2411,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.26322447585286385,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.2675,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.2646673435331104,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.2684,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.2462656781469421,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.2769,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.25513995566901543,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.2523,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.24666708686797076,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.251,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.2619602143444443,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.2553,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.27296672596888005,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.2479,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.25372139251283155,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.2279,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.23648629150911923,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.2286,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.2460967952913876,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.2258,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.24731648291326402,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.2336,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.24300848203148126,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.2472,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.2525965748470619,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.2572,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.24388342871556676,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.2688,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.2562388438137502,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.2544,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.26704049971921484,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.2493,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.2579315347760817,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.2592,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.2680451865822835,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.2206,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.24632765558448969,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.2405,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.25263588311832563,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.2324,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.2679749974216252,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.2558,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.26708637800541807,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.2736,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.24443418155398114,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.2629,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.24642458219288188,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.2637,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.2606254242993008,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.2621,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.24834272184135564,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.2586,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.26093009740496503,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.2833,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.2504206833948055,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.2295,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.2462808991777672,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.2403,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.2490232377938173,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.2454,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.2544412860477114,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.2495,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.25481075581147616,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.2474,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.2780088460313552,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.2328,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.2545859440394849,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.2521,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.24816467911678347,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.2917,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.26151518644423544,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.2608,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.2631539910933469,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.2718,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.2489229252704048,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.2634,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.26473511989059556,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.2484,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.26255409214843833,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.2655,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.2514436239140505,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.2499,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.2488523028175505,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.2521,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.25944287079566475,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.2399,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.2569385184210285,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.2685,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.24909351990213047,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.2538,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.25616805304769436,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.2736,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.24965349723372238,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.2657,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.25289097262598964,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.2247,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.25252531323283084,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.2486,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.270908428578071,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.2699,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.255425046172118,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.2549,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.2496910116565281,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.2601,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.25060147026443685,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.2586,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.25235568756531646,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.2535,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.25929853481340265,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.2181,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.2563102183529307,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.2598,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.2610648803107209,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.24,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.2582268257161907,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.2366,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.24265329969188054,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.2495,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.24222972788434782,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.2451,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.25361011969736763,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.238,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.2615844293007099,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.2449,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.25296042513305217,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.2874,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.25318993567145104,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.25,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.25625076228336996,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.269,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.2642127517700529,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.2335,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.25594710055287495,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.2523,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.24617093666055145,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.2638,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.24224003839995878,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.2602,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.25013824382054983,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.249,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.23597062628898713,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.2517,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.25984735794068387,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.2344,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.24788661600794148,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.233,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.24820586394842867,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.2606,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.26304664132091155,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.2577,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.2556930129476708,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.2443,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.23592744688614217,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.2544,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.2590344505500613,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.2621,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.2435250631737042,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.2569,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.2392493360642648,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.2456,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.24691677879506577,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.2739,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.26226618844406585,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.2637,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.24549411182114797,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.2726,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.2557484568621365,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.2413,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.23667492854580136,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.2631,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.24769315781444837,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.2304,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.25753820947482575,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.225,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.2541687190595688,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.2616,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.25128486682304796,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.2858,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.2474555292229644,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.2613,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.24969523241532426,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.2876,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.24424427613568966,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.2607,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.2624483579916018,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.2285,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.2497179552140092,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.265,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.25929998109690106,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.2657,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.24830190900013976,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.2382,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.24237334543686734,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.2456,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.24876974985947953,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.2526,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.25585691252808024,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.2621,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.25433448146842924,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.2583,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.24630129964812417,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.248,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.2533117654905944,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.2647,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.25852772670966484,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.248,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.26567919960966363,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.2502,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.2505207943399698,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.2804,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.25401946395253,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.2583,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.2458737736681624,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.2551,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.24410979390194495,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.2278,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.26570259823597137,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.2263,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.24388714405457995,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.2469,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.2522203366647109,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.2622,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.26467071555652893,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.2379,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.25442361381810485,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.2548,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.24651686450999055,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.2495,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.23657403843638364,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.2583,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.2648783966768146,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.2462,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.2500284852012768,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.2801,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.25215714709583864,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.2381,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.24403513959461937,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.2481,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.24049191161309416,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.2463,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.24467391688122667,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.2888,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.26209535477942586,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.295,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.26128024761738516,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.2767,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.30744406069741437,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.2576,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.23743009609608332,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.2683,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.253896455547253,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.2395,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.24627546033418807,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.268,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.2456044083174807,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.2669,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.2418597561510454,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.2579,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.23406390322321638,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.2601,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.24726735798344116,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.2652,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.24293499807394714,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.2636,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.24027607716784685,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.2456,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.23881077553562186,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.2547,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.24188106683307978,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.2263,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.275798927231976,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.2626,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.2510599834252435,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.2267,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.25284079171203144,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.2234,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.24568729283888052,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.2884,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.24370848619268756,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.2457,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.25976790003904193,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.2865,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.2542560100012487,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.2483,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.2521802120390839,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.2524,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.2608825738735974,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.2498,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.2668021958091731,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.2663,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.24592729968004137,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.2409,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.23714800899776536,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.2743,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.2621627469499654,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.2423,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.25627937322351035,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.2614,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.27313690628046405,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.2454,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.23416212452128515,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.2457,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.26018095819333814,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.2312,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.25274808406878574,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.2404,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.25301948127748686,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.2658,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.24320366190306963,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.2683,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.2526028069677529,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.271,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.2571910267036805,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.226,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.23359307816007796,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.2319,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.24893553912308597,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.2473,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.2497145069691861,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.252,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.26012319779360227,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.2769,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.2526068544931641,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.2665,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.24908166133078857,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.225,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.2530178006739668,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.2526,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.24104303215654602,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.2643,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.24649454942063115,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.2405,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.35356029909870246,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.2621,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.250121102775398,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.2363,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.2433292657065562,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.2661,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.2423806505980352,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.2572,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.24002058134216378,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.2334,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.26385306464364056,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.2464,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.25068139264721384,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.234,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.24982427741487812,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.2416,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.24810652518589035,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.2409,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.24784107085348717,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.2362,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.23873023832846055,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.2658,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.2442948127405335,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.2507,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.25008811328658087,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.277,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.25114385850680326,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.256,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.2472165692272086,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.2473,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.24004238747165732,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.1951,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.2449990923041328,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.2663,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.2575247407649802,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.262,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.24219337548187897,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.2569,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.236538193513278,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.2288,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.24192175132125926,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.2468,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.24341360891828157,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.248,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.24203966978888317,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.2785,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.25100978914229727,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.2311,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.25029282286955556,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.2747,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.2506831830123874,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.263,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.2601578424261114,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.2526,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.24254339531045768,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.2483,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.25731515717218895,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.2637,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.2696425146577586,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.2444,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.2539756413758767,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.2628,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.23958153648329425,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.2591,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.24900138251035098,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.2659,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.25029126200764884,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.2275,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.24606064896379057,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.2954,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.23761612223383186,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.2395,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.25817460351296073,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.2448,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.23660763431467924,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.2781,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.24730798215873828,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.256,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.2368105008080205,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.2404,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.24040484472231727,
+      "learning_rate": 0.0,
+      "loss": 0.274,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 2.439643535096414e+18,
+      "train_loss": 0.45167180501256393,
+      "train_runtime": 25923.726,
+      "train_samples_per_second": 6.095,
+      "train_steps_per_second": 0.048
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1235,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.439643535096414e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}