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": 8.51627463410803,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.5648,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 8.344813705382741,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.5929,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 8.456049956520184,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.6075,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 8.41251070176363,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 1.6066,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 8.065731413606771,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.5914,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 7.669659139954757,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.5231,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 6.468511610421052,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 1.545,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 5.631293802442801,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 1.5288,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 3.187079905660967,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 1.4472,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 3.0059413152021617,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 1.4336,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.9638366105029523,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 1.4687,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 5.435474511574397,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 1.4067,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 5.807314551667901,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 1.459,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 5.684127862702643,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 1.4085,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 4.9054375899266915,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 1.3748,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 3.7432301990894503,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 1.3285,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 3.84953236202518,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 1.3573,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 3.323634191056838,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 1.3535,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 2.658039401805748,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 1.2984,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 2.304420076398985,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 1.2601,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.912551428655787,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 1.3193,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.8755882061709597,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 1.2808,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 2.0928499278918338,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 1.2872,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.9646361491190016,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 1.2672,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 1.6109808978361764,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 1.2729,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.0584770148537619,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 1.2331,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 1.4782326649249462,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 1.2451,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 1.5224516478486427,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 1.2039,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 1.1363794412509514,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 1.1924,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 1.0727053155032213,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 1.2437,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 1.259466754060013,
+      "learning_rate": 1e-05,
+      "loss": 1.2607,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 1.0782622965135544,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 1.2084,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 1.1654866066400131,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 1.1812,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 1.1039636269891273,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 1.1613,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.8368543907465872,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 1.1936,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.9785529157256355,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 1.21,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.7856336512096475,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 1.174,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 1.131174465439735,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 1.1783,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.8465993986404992,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 1.214,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.9174118418140413,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 1.1848,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.8867213564692444,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 1.1704,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.814196788202363,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 1.17,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.8312766834349802,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 1.1679,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 1.0457972405993645,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 1.1412,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.8083811555188504,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 1.1429,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 1.1535257206564677,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 1.1504,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.8280528379093287,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 1.1611,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.8747592472732715,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 1.1747,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.9242529530028772,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 1.1203,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.904477860930384,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 1.1555,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 1.167235719081674,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 1.1291,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 1.0031694924343981,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 1.122,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 1.1896000103184179,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 1.1277,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 1.1856965553016572,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 1.1834,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 1.0707553350218122,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 1.1633,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.99691215516412,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 1.1289,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.9223843486519595,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 1.1305,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 1.0227487249943885,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 1.1674,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.9301999346405377,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 1.128,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 1.1155496689243005,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 1.1629,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.9302527730393769,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 1.1257,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.939040268084827,
+      "learning_rate": 2e-05,
+      "loss": 1.1103,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 1.403983469467397,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 1.1178,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.8584431134858549,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 1.0964,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 1.2383193548162792,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 1.1091,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.9711025941699734,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 1.1009,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.9430103462262766,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 1.1354,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.9917851190671784,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 1.1186,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 1.2529460980486937,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 1.1419,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.9307553287893967,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 1.1237,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 1.5048679667048097,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 1.1277,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 1.1542716782645621,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 1.1511,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 1.360793813553756,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 1.1107,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 1.2301330932767651,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 1.1381,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 1.2264606230671353,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 1.169,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.137019871717019,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 1.0984,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 1.5606656630160765,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 1.1207,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.2040409674288586,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 1.1155,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 1.2418196820581682,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 1.1226,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 1.2822027170638268,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 1.0958,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 1.059193549663083,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 1.1033,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 1.4358574743885066,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 1.141,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 1.0071141966120516,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 1.1075,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 1.5288856973506793,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 1.1241,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 1.223081237095944,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 1.119,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 1.223487294846226,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 1.1254,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 0.9316254349236368,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 1.1742,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 1.3400486969194705,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 1.0763,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 1.1248982659480344,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 1.1242,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 1.716036047477209,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 1.0848,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.2174277597071748,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 1.1127,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 1.3660641809876792,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 1.0976,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 1.226981185596938,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 1.1244,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 1.2642453959213082,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 1.1061,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.423008347770767,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 1.126,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 1.083672037280691,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 1.1138,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 1.3200729735522878,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 1.119,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 0.872167767610567,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 1.1159,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 1.3673558492081515,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 1.0499,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 1.1377394716655869,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 1.1322,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 1.2908725034149342,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 1.1071,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 0.9957916291207337,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 1.102,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 1.5480435735695595,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 1.1184,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.398754453512048,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 1.0982,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 1.5339036488623723,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 1.1163,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 1.4298208428264325,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 1.1366,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.164808142470854,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 1.115,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 1.3823330932114881,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 1.1122,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 1.2159381824368585,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 1.1114,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 1.6188251924504629,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 1.1082,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 1.7725503743687723,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 1.1572,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 1.2045304507477235,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 1.1211,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.3946526449498604,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 1.1126,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 2.133415348837169,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 1.1228,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 1.173734865853631,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 1.1093,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 1.7006561275609862,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 1.1458,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 1.4925672169858788,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 1.0862,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 1.6536195033984538,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 1.1005,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 1.7445444931389686,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 1.0687,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 1.3310876974488557,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 1.1293,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 2.3084098246851306,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 1.09,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 2.1764821338737894,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 1.1525,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 1.7194434119586472,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 1.0817,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 2.1592755210478525,
+      "learning_rate": 4e-05,
+      "loss": 1.1217,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 1.6059435970687184,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 1.1016,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 1.9342581312092104,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 1.1043,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 2.0054255071979097,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 1.0944,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.3334460585239332,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 1.1175,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 2.0188119293994347,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 1.0728,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 1.7156496498954625,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 1.109,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 1.3489825170702217,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 1.1111,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 2.2730217944661075,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 1.0693,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 1.2369014539553633,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 1.1265,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 2.549512162240363,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 1.0595,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 1.8318592872891202,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 1.1226,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 2.425611824363391,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 1.084,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 1.4548727848242868,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 1.0518,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 2.020571325954832,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 1.0913,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 1.5298329092910241,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 1.0967,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 1.5974608511375459,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 1.0715,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 1.655911165986942,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 1.0685,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 1.3778205669078485,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 1.0908,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 1.617138953647057,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 1.118,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 1.3839656372319313,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 1.0931,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.9019490639346629,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 1.1098,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 1.5106506806445676,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 1.097,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 1.0933788645757494,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 1.12,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 1.412307829076467,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 1.1073,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 1.269436150750781,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 1.1098,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 1.6526412038536686,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 1.0985,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 1.3463502043066504,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 1.086,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 1.6215009009319477,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 1.0717,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 1.2706975782848131,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 1.0811,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 1.6027847461300249,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 1.0879,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 2.030943681646194,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 1.1123,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 1.2623464708274974,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 1.0975,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 3.1256302703270866,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 1.0966,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 2.5781884726336286,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 1.0926,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 2.0738320868031024,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 1.0996,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 2.0126348841071637,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 1.1046,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 1.8659131582251607,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 1.0981,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 1.5439100897718767,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 1.1083,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 1.4006329004565838,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 1.0604,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 1.5003215944554946,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 1.0567,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 1.572854337640819,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 1.1103,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 1.2040961084219581,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 1.1083,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 1.1017476821304852,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 1.1072,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 1.1200592605104909,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 1.0681,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 1.4738526188034218,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 1.0948,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 1.3649588668739512,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 1.1001,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 1.4100861965654627,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 1.0787,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 1.7216150356434639,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 1.1223,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 1.4125308948711557,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 1.0753,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 1.8211063403940997,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 1.0986,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 1.2624417751679098,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 1.0543,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 2.0678361553105415,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 1.1094,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 1.2665924058867444,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 1.0538,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 2.355841686559975,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 1.109,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 1.6544058152117722,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 1.1032,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 1.6696564970727388,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 1.0672,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 1.6587085496566163,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 1.1145,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 1.0701548835024615,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 1.0894,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 1.5349826638694382,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 1.0764,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 1.6560702086958572,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 1.0925,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 1.2620273080317783,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 1.1289,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 1.5316514850190837,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 1.1018,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 1.3038401680416025,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 1.0652,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 1.7884022402879107,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 1.0914,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 1.246457244349869,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 1.0927,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 1.9655429903785309,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 1.1019,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 1.4623712479612183,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 1.0816,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 1.8518567178299277,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 1.1283,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 1.5202334585756092,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 1.0989,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 1.6514564002580554,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 1.1049,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 1.3278497002385818,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 1.0795,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 1.730669438198087,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 1.0984,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 1.1465559089388744,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 1.1,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 1.4310722888532552,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 1.0542,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 1.212178664187285,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 1.0681,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 1.0874222025593303,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 1.0993,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 1.3524925517893807,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 1.0839,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 1.5123389931286497,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 1.1241,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 1.427145506158117,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 1.0787,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.9616051511445216,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 1.0617,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 1.3574087444280942,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 1.0837,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 1.6944783694928933,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 1.0842,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 1.1009663937556142,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 1.0804,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 1.2546046602275538,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 1.0757,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 1.1661077223595369,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 1.09,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 1.3753386368724927,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 1.0852,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.9849388457917506,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 1.0816,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 2.0581412301269224,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 1.0877,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 1.475434356459291,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 1.0914,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 1.7976791267845131,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 1.0945,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 1.3854368450502932,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 1.0635,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 1.6990905311323135,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 1.0384,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 1.245990543457289,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 1.058,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 1.6615445568362333,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 1.0946,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 1.4695523065925906,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 1.1121,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 1.8211972030894845,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 1.0605,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 1.5547443852504237,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 1.0487,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 1.7775765674220505,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 1.0654,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 1.4580441897576966,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 1.0442,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 1.6082729413761796,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 1.0449,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 1.3001728105538146,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 1.0942,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 1.2933924102822063,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 1.0499,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 1.1205711155520848,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 1.0723,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 1.2829715236074326,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 1.0446,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 1.5784163081605247,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 1.0849,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.9501331001270003,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 1.0987,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 1.4758879271765697,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 1.0707,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.9606982606324972,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 1.1127,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 1.7667045025970969,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 1.0756,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 1.1294513583670813,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 1.0858,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 1.5953708287491022,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 1.0496,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 1.333256027542238,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 1.0577,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 1.6307412469832923,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 1.0974,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 1.3628908600748006,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 1.0839,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 1.3488474507316806,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 1.0905,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 1.6039971868184548,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 1.0632,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 1.061486164294291,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 1.086,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 1.195350448740223,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 1.1027,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 1.068243357847597,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 1.0895,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 1.279351699402425,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 1.0794,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.8083127285580332,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 1.0401,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 1.3165276800283663,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 1.12,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.162910640014297,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 1.0938,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 1.6733151804964328,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 1.0157,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 1.1923804236149942,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 1.0167,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 1.5363402500874455,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.9961,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 1.5052881740236221,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.9762,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 1.243123907788557,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 1.0062,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 1.6698055933689748,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.9509,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 1.6022324757929443,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.9971,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 1.6355622063025312,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 1.0228,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 2.007324520669785,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.9846,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 1.5305146267353498,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 1.0063,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 1.3026802138482292,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.965,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 1.6104691255599024,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 1.0047,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 1.0506614765853035,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 1.0016,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 1.1500005658996735,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.9566,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 1.304105928880093,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.9801,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 1.1656236662294228,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.9767,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 1.3001025641969928,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 1.0329,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 1.2958596875653632,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.9381,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 1.0444596678981952,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.9748,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 1.5831816305068607,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.9881,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 1.2882930921938798,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.988,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 1.5078512321582984,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 1.0063,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 1.5201849291416838,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.9922,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.9487663163396786,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 1.0053,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 1.5608443037251987,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.9672,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 1.242311357609997,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.9642,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 1.1698109931416825,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.9396,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 1.4356887888829906,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.966,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 1.1262858239113847,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.9793,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 1.7274080952390585,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.9881,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 1.46124973269521,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.9891,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 1.507364376355238,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.9635,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 1.6959911495910376,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.9922,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 1.3739219348918976,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.9985,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 1.1908298602241303,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.9913,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 1.557293897730811,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 1.0102,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 1.0323786842000098,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 1.0211,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 1.4810552373447716,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 1.0084,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 1.2160994607231819,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 1.0052,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 1.2973951716631515,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.9582,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 1.1796386831167687,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 1.0086,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 1.0330819767733668,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.9725,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 1.7425660106695795,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 1.0358,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 1.2287161078905553,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 1.0033,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 1.9137476106619615,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.9856,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 1.5613025013265993,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 1.01,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 1.5755750054044295,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.9593,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 1.311506788055471,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.9856,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 1.527527076154362,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.9865,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 1.1801217423153536,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 1.0155,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 1.3988869354915303,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.9594,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 1.4080994434083989,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 1.0075,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 1.1689069609834106,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 1.0136,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 1.1689705723470925,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.9754,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 1.4256784291483882,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.9836,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 1.5150345094721325,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.9864,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 1.513930388933527,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 1.0121,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 1.2362497531251955,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.9834,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 1.6337010069359281,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.9517,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 1.3318514166910822,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.9743,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 1.2550917019498324,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.9661,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 1.7177650372247117,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.9915,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 1.0510754426254916,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.9796,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 2.1064564851572047,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.9475,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 1.496326618657796,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 1.0074,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 1.6966800500112955,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.9946,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 1.6873669884990865,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.9658,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 1.3934921748022273,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.9874,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 1.9669839752310967,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.9972,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 1.5771891507163645,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.9771,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 2.2026072666823238,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.9803,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 1.8949436198085612,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 1.0032,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 2.1990373260483578,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.9976,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 1.8178355524530858,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.9775,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 2.034144976680855,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.9866,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 1.844743277663449,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.9784,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 1.6917220182839225,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.9587,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 1.4055782690313363,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.9883,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 1.5397833738280606,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.9638,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 1.4139486800168395,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.9906,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 1.2245208862409076,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.9434,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 1.7227674177094692,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.9837,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 1.1330290588890888,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.9808,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 1.7499112468055902,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.9887,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 1.623720812988408,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.9981,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 1.3990726302811185,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.9827,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 2.237511014702572,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.9881,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 1.6358421978877884,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.9976,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 1.871641093913985,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.9457,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 1.6409534983172056,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 1.0035,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 2.3629269930188013,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 1.0137,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 1.5397008915139794,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.9654,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 2.6305697288234184,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.9773,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 2.307910099803769,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.9914,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 1.8941878367533618,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 1.015,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 2.1414957906001972,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 1.0323,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 1.6418568968977012,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 1.0129,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 1.7983211198847557,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.9709,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 1.8700624396563217,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 1.0027,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 1.5426359025311402,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.9634,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 2.004588802259208,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.9793,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 1.6143157873527316,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.993,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 2.2113407330890227,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.947,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 1.8805230259538477,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.98,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 2.086957664679214,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.948,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 1.870694641712293,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.9511,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 1.8076384083145325,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 1.0037,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 1.8085031174692463,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.976,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 1.9161530536288658,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.9934,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 1.48232684628254,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.9704,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 2.1041434402256787,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 1.006,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 1.8515904411238027,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.9904,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 1.8599037074749938,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.9893,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 1.6999904625087756,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.9994,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 1.9435080468113988,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 1.0388,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 1.5225397218971004,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.9684,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 2.2843302026853087,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.9735,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 2.237715657271571,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.9989,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 1.7925834244386645,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.9922,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 1.9963626550815032,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.9968,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 1.3895023999076919,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.9735,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 1.387548191461799,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 1.0368,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 1.910111016473163,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.9743,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 1.3381704829484844,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.9735,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 2.045969251551876,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.9448,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 1.6969862103436146,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.967,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 2.036904209534844,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.9575,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 2.020111285176102,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.9862,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 1.4551101554696182,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.9768,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 1.5345537655043802,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.9886,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 1.8979219493597304,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.9886,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 1.4888009420859392,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.9517,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 1.8235568400564108,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.9548,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 1.381719805989444,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.969,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 1.990857855715007,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 1.0118,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 1.7833333885932303,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.9846,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 1.9034069637811268,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.9876,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 1.8990895517154243,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.9994,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 1.6474015376489834,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.968,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 1.5103693317970561,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 1.009,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 1.8167042095101864,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.9707,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 1.328050954282212,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.9656,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 2.2561409097694085,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 1.0013,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 2.116252779519774,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 1.0074,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 1.3019817515709502,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.9679,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 1.4963012392863928,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.9454,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 1.4034784183441051,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.9931,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 1.1380650457856911,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.9632,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 1.6350347919039796,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.9929,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 1.1313941210922602,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.9921,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 2.149272475187329,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.9658,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 1.7854597005696269,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.9661,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 1.7327732328244505,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.9738,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 1.6480449091051883,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.9669,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 1.6128493770674501,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.9987,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 1.5013026224089747,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 1.0132,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 1.8333135105398215,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.9707,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 1.323936550850708,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.9951,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 1.9150559348865366,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 1.0012,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 1.7616220619376983,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.922,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 1.417060420764514,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.9893,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 1.462727951581017,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.9776,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 1.3362021504881854,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.927,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 1.1992587337494038,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 1.0151,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 1.1440687615692224,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.9784,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 1.044755189646294,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.9655,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.8208050371534346,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.9609,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 1.1086976959448918,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.9742,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.9873858621868938,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 1.0135,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.9065166802465574,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.9794,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 1.1844267686739485,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.9789,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.9052127353295333,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.9876,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.9956561418446512,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.9633,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 1.2855330274068624,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.9725,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.9244447069662723,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.9918,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 1.1072750603589019,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 1.0158,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 1.0339304140870154,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 1.0045,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 1.0371029957480826,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.9796,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 1.1955089080731036,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.9954,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 1.0599539361867536,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.981,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 1.1049493839577706,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.9883,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 1.2012874327370988,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.98,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 1.013920486723881,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.9814,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.9526757473567122,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.964,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 1.4033130673443364,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.9999,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.9153540641844549,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.988,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 1.161685661596254,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.9733,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 1.0924076407874002,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.9853,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 1.1030789076511878,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.9576,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 1.2930302690298092,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 1.0253,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 1.325427201473766,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.9992,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 1.091951657594921,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.9857,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 1.093745173169124,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.9929,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.8023094454488958,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.9732,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 1.1337351489031309,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.9637,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.8035085068241219,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.9946,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.9764832653438298,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.9673,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.9114431986368879,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.9899,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 1.0920617546412013,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.9957,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 1.1587395648319174,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.9512,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.9882771639149397,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.9906,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 1.2892858332127204,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 1.0168,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.9876540671829679,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.9729,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 1.5386947433130762,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.9885,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 1.0590748026007963,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 1.0167,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 1.1383196360086547,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 1.0049,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 1.1474021464774449,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.9764,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.9943828122315308,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.9529,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 1.725472182170164,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.9947,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 1.2615581349694898,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.981,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 1.1653553563760854,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 1.0038,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 1.619972369280137,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.9958,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 1.3149434604654544,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.963,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 1.3658548292258348,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 1.0082,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 1.1120969350899799,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 1.0325,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 1.1299921269228492,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 1.0034,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 1.1067889308458043,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.998,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.939101506411309,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 1.0008,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 1.5799369048713119,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 1.0033,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 1.4517191699351604,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 1.0017,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 1.6458184575854913,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 1.0037,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 1.4450965172133874,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 1.0022,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 1.785740935884219,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.9864,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 1.451295803361489,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.9839,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 1.4122244665668664,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 1.0001,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 1.132383565628871,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.9422,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 1.1429101117545553,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 1.0092,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 1.1607435529471972,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.9913,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.8811878599440235,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.9983,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 1.299686244034383,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 1.0007,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.7678881846731842,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.9955,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 1.6592686512992612,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.9821,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.9093652114052821,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 1.0093,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 1.9272982813230897,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.993,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 1.3863831881812814,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.9694,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 1.90643394220855,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.9785,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 1.8533785043476827,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.9684,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 1.5736596612039548,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 1.0273,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 1.5090215442638988,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.9967,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 1.3544851706874135,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 1.0062,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 1.3124044311672916,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.9936,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 1.2619312782898617,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.9809,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 1.111899588126423,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.9812,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 1.3503424233260226,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.9966,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 1.2245230933878855,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.99,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 1.0755426352054283,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 1.0087,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.3601980086862502,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.978,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 1.0219572962087002,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.8377,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 1.515224942536625,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.8684,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 1.2024298050639077,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.8383,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 1.3834140061892792,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.8681,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 1.4396416371658038,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.8264,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 1.0402720341003007,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.8231,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 1.454539806058273,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.867,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 1.1815331498311223,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.866,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 1.33065660947295,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.8665,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 1.4564476084212539,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.8498,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.9801839634462834,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.8487,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 1.2447721905842228,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.8451,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 1.1162574046367928,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.8497,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 1.2845022346438029,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.844,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 1.0831030117877303,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.8418,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 1.0845147990663344,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.8183,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 1.098042656024876,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.8581,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.800644757730114,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.8378,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.8235931264743673,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.84,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.7159565569884797,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.8298,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 1.0086107105386488,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.8669,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.8288313591596234,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.8482,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.8844666094829181,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.8795,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.7818871190800502,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.8413,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.9338687119891145,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.8622,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.7715261702055939,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.8425,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.8876944621852836,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.8283,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.8105045077850086,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.8525,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.8899297224946291,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.8318,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.8068069111356132,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.8474,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.9092784011953471,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.8189,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.8869139116010227,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.8691,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.8415889974090254,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.875,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.7779238741946766,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.8589,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.8672352866969267,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.8362,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.7769896573208799,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.8657,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.8692843247091826,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.8605,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.7310635681512708,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.8608,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.9061281136252611,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.873,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.7710871563227552,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.8342,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.7449334831216905,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.8378,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.7001664074618685,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.8454,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.6806523102068011,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.8251,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.7480393284233143,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.8101,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.803616516477838,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.8741,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.8095842276497828,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.8514,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.8235958346563387,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.8262,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.7901047717846242,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.8684,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.7503136866209327,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.8323,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.8320997958782013,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.8583,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.895641108773523,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.8698,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.9679939728278846,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.9267,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.8699219200000989,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.86,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.7496051565397823,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.8464,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.9166681917858566,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.852,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.7627435824545693,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.8683,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.7048097770730517,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.8472,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.8843817649760572,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.8971,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.6253468510991596,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.8712,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.9773610893964202,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.8352,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.9669459200010267,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.894,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.6259070986331479,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.8259,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 1.0797424577441368,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.8722,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.9028650112520679,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.8782,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.7342767173357225,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.852,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.8821295225201886,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.8217,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.6472379234172465,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.8623,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.8996147513750175,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.8953,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.8643776408782667,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.8874,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.8816695873978893,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.8778,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 1.1741549178221904,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.9039,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.7317852398698128,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.8359,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 1.0655181360596746,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.8684,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.7930311130180527,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.8602,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 1.0095381739799973,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.8925,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 1.1645210823281174,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.8721,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.7854382396006054,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.8793,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.9719733241552868,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.874,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.7967948125200054,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.8511,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.8032884306506584,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.8584,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.8223396771103911,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.8851,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.868075488799785,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.8555,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.7314146298995321,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.8743,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.9217282005155795,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.8888,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.8319053200432011,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.8919,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.9323889874953947,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.8805,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.8292617320014698,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.8458,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.9683063725179001,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.8483,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.7495441114176105,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.8973,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.8685142611549241,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.8552,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.711934158311536,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.9007,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.9412578414155035,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.8795,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.8499249026649963,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.8751,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.8454767189796256,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.8775,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.815742774352505,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.8639,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.6906754336659542,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.867,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.7117151367527252,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.8626,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.8162669126476234,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.8529,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.7382713291693197,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.8589,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.8877259604072135,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.8744,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.6940949993325922,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.9078,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.8437562245844137,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.8618,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.730731491595146,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.8647,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.7770686314194437,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.8812,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.7677238243271394,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.8403,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.6233213008984949,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.8826,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.6459329177122247,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.9018,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.623685148316736,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.8587,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.6887875430464138,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.8479,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.8362574833131501,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.8741,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.6597421724847363,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.9001,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.7776113689049389,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.8577,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.750248429531645,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.8519,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.7816997496909648,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.8724,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.6602047526691684,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.8748,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.6627997030894762,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.894,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.7388991323908335,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.8546,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.67773728858182,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.8656,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.7936230587164385,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.8985,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.7825046957323639,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.872,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.7466752958082974,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.8879,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.7571055111821817,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.8494,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.7466034708119432,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.8412,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.7205572333980352,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.8721,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.8017740122502697,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.8768,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.7086627696072771,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.8615,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.9213006806931245,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.8618,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.7064822046855779,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.8645,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 1.021151202849447,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.8829,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.7867121231130412,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.8638,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.7516813576715398,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.9006,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.7513884666850262,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.8504,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.8026201297138028,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.8471,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.7617533804067254,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.8688,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.6827916265410062,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.8599,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.7980143660984791,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.8576,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.8427534298133011,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.9018,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.8338503648746189,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.8721,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.7547382247465682,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.8852,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.8817271189948328,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.8631,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.7746000716774487,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.8972,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.8214975147604839,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.8849,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.8105929205445496,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.8879,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.7752509424796129,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.8741,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.9397230700401901,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.8824,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.6713700419780837,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.8712,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 1.019232851573359,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.8776,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.6648606443774848,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.8682,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.8399759863278496,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.8891,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.696041999210878,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.8827,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.918741590236244,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.8958,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.8782420405496539,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.8404,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.6147823546956954,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.8361,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.7599370085365399,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.8785,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.5872382303778613,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.8667,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.8341481044443629,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.8563,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.5729784280863371,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.846,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.8864289896632419,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.8498,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.623169446130345,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.8961,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.7979709156149613,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.8511,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.7210072119223199,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.8969,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.61393396843562,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.8369,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.7214121737185761,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.8497,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.6586202887220092,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.8695,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.7462757554097607,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.8628,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.6513295463289365,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.8598,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.5842111687855405,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.8528,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.6524003904153478,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.8177,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.7229041004710888,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.8669,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.6165014514794132,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.881,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.775957667672207,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.8926,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.6300197080556684,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.8737,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.5817555029495072,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.8298,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.6557009599591153,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.849,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.6683834188327793,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.8659,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.7101096020847516,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.864,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.724775167878485,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.891,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.7676595898287064,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.8442,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.8263643901949106,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.8942,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.9448408475005116,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.8738,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.7754262426527093,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.8559,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.7290248569961699,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.8616,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.8542313641680126,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.8726,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.6605878999891854,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.8383,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.7736457593524003,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.8601,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.7099566934286069,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.8432,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.725732391491258,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.8836,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.8844359442485307,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.8797,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.5981356539741599,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.9016,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.740608971233623,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.8364,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.7180459247517176,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.8744,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.8122959158675309,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.8961,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.7096565042831814,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.8615,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.8459941831183796,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.8353,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.703039872715493,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.9038,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.7279874125990077,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.8512,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.7096547260179359,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.8769,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.8289701529871943,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.8638,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.7304309352416349,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.8172,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.8675606016928892,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.8678,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.6534633716777438,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.8471,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.6744307261370823,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.8299,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.7441710690812902,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.8658,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.5979881792325631,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.8447,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.7663366446496725,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.8965,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.6701495485438645,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.8407,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.6766456789279983,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.8834,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.64250800307045,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.8891,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.7273535052545503,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.8548,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.8279569595885151,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.8336,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.6002426464599023,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.8911,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.785721077641233,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.8683,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.5843275008000811,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.8702,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.7999161822053524,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.8671,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.6817311731007873,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.8847,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.80348625802985,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.8728,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.8693914184833696,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.8988,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.6741806848713241,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.8616,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.9278187317168175,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.8544,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 1.045100555155749,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.876,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.7130417952189182,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.8407,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 1.101399925192534,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.872,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.6929759275250025,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.8966,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.9519834527767292,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.8726,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.719081838418714,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.8408,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.7720037804125927,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.8774,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.8509052631605446,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.8802,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.7158484962584825,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.8865,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.7462395161540898,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.8853,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.763605866236919,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.8544,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.6303892255783998,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.8961,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.8397047025058372,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.8436,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.7543065257366385,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.8721,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.7860628944033136,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.874,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.9539968048476496,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.8555,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.7628755522126811,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.887,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.9832959276662953,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.8456,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.7636078615585606,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.8831,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.8386024985155873,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.8584,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.7510814892043777,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.8657,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.6521020177709603,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.8709,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.6822811273876713,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.8539,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.574465156004791,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.8749,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.6375370559189582,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.8802,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.6710952613361544,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.8296,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.5961912772893546,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.8681,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6899580764256895,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.8306,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.9268378893958751,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.6861,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.81080434468756,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.7079,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 2.2074164628939283,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.6808,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 4.635117280389213,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.7539,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 1.2657805864133411,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.7028,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.881232217061371,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.6813,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.9673412799879425,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.7196,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.894017612366,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.6979,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.9733152130224468,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.6611,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.7835263858664082,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.6616,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.8338673077831725,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.6822,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.7580834449033592,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.6528,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.8016976527810523,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.6678,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.7747073731053228,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.6838,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.7028895038169658,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.6507,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.7459484238647955,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.6759,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.75692574655556,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.669,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.8296333240313267,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.6863,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.7002485071767994,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.6564,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.7480636615621262,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.6759,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.7695478443321649,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.6919,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.5898553498439647,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.6429,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.8245433750421309,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.7059,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.7096107207486524,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.6816,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.6244491936536304,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.666,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.622725560194113,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.6982,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.7360423523980835,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.6923,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.6225310726192471,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.675,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.6270669377488606,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.687,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.6525600105690812,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.6892,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.6393154849054182,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.7068,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.5595092610944368,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.6651,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.6192768170713002,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.6598,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.6547953255516026,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.6612,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.5690968754952739,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.6785,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.6493838175414285,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.6594,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.6295125463142542,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.637,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.5424303794762491,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.6994,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.6191258967176473,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.6761,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.5483509863484637,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.691,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.6065857358568216,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.6671,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.5569870320067055,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.6688,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.5324675937629244,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.6539,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.627238014292415,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.6924,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.5243421120462258,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.6695,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.6402677151007748,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.6758,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.6015474399904414,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.6797,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.6645210799577735,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.671,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.5713433281311059,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.7019,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.5406521009975211,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.6723,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.6915484682444422,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.6923,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.5458233717267508,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.6996,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.6044775421804485,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.6809,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.5150725287562571,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.6725,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.5046450964127915,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.6897,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.547357324313691,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.6753,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.511595462907422,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.7058,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.5191709659517637,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.6672,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.6588731757874349,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.6735,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.6166523821331987,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.6669,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.6900741922111647,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.7171,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.6060947244730269,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.6545,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.5669361972577129,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.7045,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.6573187776304866,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.6867,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.564249340127258,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.6637,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.5580533571175795,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.6908,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.5861282894648735,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.6768,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.5931315566668564,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.7004,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.5625517532570122,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.6865,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.5832897968578439,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.6939,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.5351901566423408,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.6523,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.5446958573131564,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.677,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.5785576769472207,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.6551,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.6296373497637867,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.6678,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.5527904675065888,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.682,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.6262361231167637,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.6894,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.5611172137837627,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.668,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.5137534883098788,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.6651,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.5142703629343831,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.6998,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.5582168351360727,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.6952,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.5482343186814995,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.68,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.620929667297882,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.6882,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.5334458735794403,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.6766,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.6673228947381801,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.711,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.6144888669270385,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.6956,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.5984733321577961,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.6567,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.6855984170591842,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.7071,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.5089282033148309,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.6623,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.6629237682293958,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.6771,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.5936622322708586,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.6726,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.5848366359653289,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.646,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.6260829450073556,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.6788,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.5390366316622088,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.6684,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.5637851560585899,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.6626,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.6327106781006522,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.6827,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.5270068220042332,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.6838,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.5632689096440515,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.6909,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.5770786508013602,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.6941,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.4942403322914515,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.6408,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.6033723089143271,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.6933,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.5033186028824983,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.6895,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.6384654374873335,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.6624,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.6097210326174723,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.6778,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.518237012542876,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.684,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.5955243308162159,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.696,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.6887399357664438,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.7022,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.5082920201754499,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.6812,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.6066742803441593,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.6733,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.7060562484322589,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.669,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.5326182222442032,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.6987,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.5439465469415274,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.6855,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.6820011619690379,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.6603,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.5989713792128849,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.6548,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.501182814402405,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.6807,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.5640254032616082,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.6685,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.5078945135619715,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.6836,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.5369364741183926,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.6874,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.535023522970385,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.7072,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.5474083543829017,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.7066,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.5406817535498447,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.6964,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.5271933104070997,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.697,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.5382300414032869,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.676,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.5366281063880608,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.6627,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.5655939717041056,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.6886,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.5468295836664104,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.6724,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.5156878136989778,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.6657,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.5307706148401437,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.6709,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.5689313514996069,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.6624,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.5731102846919263,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.6779,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.519741503218015,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.6878,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.5460688324358013,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.6661,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.5560780737737859,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.6745,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.5416526995993799,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.6887,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.4934068987208162,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.6456,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.532317143250822,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.6883,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.4913177647656789,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.6885,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.5072731422942063,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.6933,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.5058414945423347,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.666,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.5610225576367933,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.6763,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.5030001507912527,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.6645,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.5208737430815872,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.7067,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.5341929860807024,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.666,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.5329908796235828,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.6818,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.5045787686608272,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.6764,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.5035670308831107,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.6893,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.520882968249889,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.6949,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.5020969177722187,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.6653,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.504766679111993,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.671,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.49073850669593927,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.6632,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.5416555764391279,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.6893,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.5245798673642333,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.6989,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.5085313034776724,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.6721,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.5101241212930345,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.6961,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.47935888920431996,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.6585,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.527341528956794,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.7146,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.5000954063984554,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.6912,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.4883797425048548,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.6697,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.47502006339736846,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.6659,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.5040227211058186,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.6794,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.5015116373528011,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.6772,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.4903673735480041,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.6691,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.48714381817447905,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.6916,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.5882001149467057,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.6755,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.49789724799248497,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.6662,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.5018824970539147,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.6595,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.5158505396973072,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.664,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.49931138627051624,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.6711,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.4867865344478767,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.6819,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.5161916502367375,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.6732,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.4714882924052319,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.6897,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.48514944045136976,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.6941,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.509446052407022,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.7013,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.4687472227527154,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.6787,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.5252455478948326,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.6929,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.4829532142705618,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.671,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.48661501927058487,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.6847,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.4880211698725606,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.6824,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.4889075139087396,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.6947,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.5203155658815484,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.6662,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.5051745549792195,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.7013,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.49674414804235983,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.6639,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.472495534847283,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.6981,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.5210432772640523,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.7092,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.47446721474653275,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.6885,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.5371059229722521,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.6801,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.5162525592417003,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.6815,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.45914101212838876,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.6772,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.5435382369830968,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.6782,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.5202496735872386,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.6993,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.5006883991299138,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.6698,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.5601320996198362,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.657,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.5398255060450176,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.6505,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.4700319232307948,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.6696,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.5571323553532036,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.6727,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.52951490543799,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.6833,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.5119103236288164,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.6827,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.4760005218330048,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.663,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.5083697286610932,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.6829,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.517674619886402,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.6602,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.49170783534739704,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.6465,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.48774065474696876,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.6585,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.5230017953290332,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.7028,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.4944590955212943,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.6705,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.498242342454282,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.6601,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.4621045324661911,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.6877,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.46255158146904407,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.6424,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.48900756760029984,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.6397,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.5140862011931187,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.6832,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.4622890943708547,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.6812,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.4699959213311207,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.6716,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.5172960895223945,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.6699,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.4869141100188777,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.6879,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.48840669162970674,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.6908,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.47895387035670706,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.6623,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.4830714108774878,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.6551,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.4780330260248538,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.6812,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.49091275485394903,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.6603,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 1.071937423674755,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.6813,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.44903073344724975,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.6625,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.4699455275774622,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.6818,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.5161088414822875,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.6878,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.4585256627690697,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.6999,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.4832742135976854,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.688,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.46803880930561204,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.6875,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.4615960918037735,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.639,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.4794230120231865,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.6826,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.47525680043834034,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.672,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.48468538280279255,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.6998,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.4635108192043249,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.679,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.46492983687220113,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.6649,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.4879502435679885,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.6952,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.48184010785733244,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.6763,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.4885362093082099,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.6585,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.45542173782153744,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.6674,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.5035032552482046,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.6703,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.5268983531669464,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.6933,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.47146709524538943,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.6945,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.5082854202093042,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.7036,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.47352407139321406,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.6877,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.46739720639695814,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.6748,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.48276964565276614,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.6827,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.47450176138656847,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.6636,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.4695475201640276,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.6881,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.45973139439995,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.6627,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.4815751501967175,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.6864,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.49785907689336656,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.6777,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.49201055668012744,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.6215,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.933496089368358,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.5381,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.7881806456455547,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.5329,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.568154600792115,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.5315,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.6029731371383937,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.5041,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 6.0134964374463005,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.6821,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 1.316588462376856,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.5002,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 1.2436776451067733,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.5224,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.8952571168464621,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.5502,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.7059346168565297,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.5545,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.7811118376795825,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.5386,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.8190483440466596,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.5205,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.7619613639054236,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.5411,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.6599641030883541,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.5162,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.5961372830522903,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.5425,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.6370117797539102,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.5261,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.6564938935463092,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.512,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.6594553257198801,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.5352,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.6280292652216142,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.5292,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.6095892072619188,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.5454,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.6055207995637298,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.515,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.6552571156803567,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.5372,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.6550098253023039,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.5308,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.5825902556413192,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.5317,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.5319042727806024,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.5139,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.5359187923612662,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.5105,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.561317329926889,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.5318,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.6270035943764644,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.5295,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.5635566966697159,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.5334,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.5122717008665516,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.528,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.5133123665312612,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.5121,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.5197525514600045,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.5226,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.555520370325026,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.5687,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.5214475353069739,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.5286,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.5223866297253932,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.5063,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.5028518857631812,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.5053,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.49861863537869755,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.5434,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.5380117375971252,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.52,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.5225643483645509,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.5154,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.5487502078745986,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.5122,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.4851325531623411,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.5143,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.49880979050056073,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.5183,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.5114561023512505,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.5482,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.47132517653923584,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.5119,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.4929054695503821,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.5316,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.5070862978090805,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.538,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.49431798631773943,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.5357,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.4840912825096761,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.5134,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.4749287893687528,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.5263,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.4719558759443875,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.5137,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.488150638868349,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.5089,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.484438173164881,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.5259,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.49073491406781594,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.4966,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.4820940538488755,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.5232,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.4803935176838339,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.5462,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.48742389755275567,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.5485,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.48369318473542766,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.5291,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.48234275773881846,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.5291,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.49534835732340043,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.5214,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.49403719365238974,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.53,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.4877200819246285,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.5222,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.4723049843998586,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.5133,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.46852567328868255,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.5429,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.4733289524339881,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.5385,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.4712884957787124,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.5139,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.4778525321018174,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.5164,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.47987671126383047,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.533,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.47079141861944834,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.5286,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.47434310669124985,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.5261,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.48170116651325867,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.5426,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.47724609576778315,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.5316,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.47301563970340377,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.5249,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.4654699003767655,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.5202,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.4688869881880558,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.5442,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.47830064639947056,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.5312,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.4730103777901641,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.5102,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.4696378131554933,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.5393,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.47726025428030916,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.5225,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.4741687467391817,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.5063,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.4872032768007929,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.5359,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.46739750537926283,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.5045,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.4641656472126262,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.5374,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.48042503200005826,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.5305,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.47309266919301074,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.5215,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.46005997385647723,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.504,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.4781814489385533,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.5201,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.46644340245882654,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.5518,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.4799192274851927,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.5555,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.4733694132456095,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.527,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.4697005320254446,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.5363,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.4619529651450419,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.5111,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.4752814087773469,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.5522,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.4704811389430905,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.5044,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.46733583581911475,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.532,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.4649619694867191,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.5154,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.46235537078354444,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.5043,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.4932062897797419,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.5474,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.4860924120514255,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.5414,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.4829292116952551,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.5356,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.47877177184429137,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.5302,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.4601290609581488,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.5097,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.4750161199715454,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.5361,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.47261598075903344,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.5327,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.4783886392391301,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.5323,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.47673190739857413,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.5362,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.48332594855408584,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.5238,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.45810635499084695,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.5107,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.46959058703459433,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.5252,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.4797501410964186,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.5572,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.47254754698211654,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.5208,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.4549139318729733,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.5199,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.4651406376737727,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.5134,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.4597660170409875,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.5461,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.46331435515684183,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.5334,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.4707922997548929,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.527,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.4530744706964159,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.5503,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.473137435028591,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.5396,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.488710753573527,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.536,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.47233456653219213,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.5266,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.4759928648856727,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.5508,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.459663788587944,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.5214,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.4825494737848254,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.5352,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.46245738721200214,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.5199,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.4673050966678573,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.5198,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.4581527725075577,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.5044,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.4646422358834328,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.5173,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.4738835050490392,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.5424,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.4789090730254928,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.5583,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.47358259739931097,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.5276,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.4801301315326722,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.5298,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.48233461953059364,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.5471,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.4779695442061154,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.5371,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.4669662457831725,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.5302,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.4728789903757756,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.5267,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.4897892152327258,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.5479,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.4619597823556659,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.5183,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.46011992715863576,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.5285,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.4562658796719578,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.5453,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.4857756913935323,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.5384,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.4727663717445859,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.5233,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.4632772545020289,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.5441,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.4614897238806748,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.5493,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.4538155788058905,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.4978,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.49214532192231264,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.5238,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.4867759927664385,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.5126,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.4634998653803972,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.5379,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.459998653660628,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.518,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.49205808905160725,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.5573,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.46136061894967867,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.5406,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.48055285964177036,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.5413,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.48605787814334234,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.5347,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.45944047683143907,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.5239,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.4563473494944263,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.5204,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.47363363987522045,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.5242,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.46780606844707684,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.5305,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.4573029863992859,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.5184,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.4701138546164557,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.5167,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.47028578459329395,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.5148,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.4633463500573393,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.5336,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.4673063363224811,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.509,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.464876281695356,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.5406,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.47130099428498073,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.5282,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.4727759749048157,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.5424,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.4613457915360654,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.4963,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.4729311291019276,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.5119,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.4646161600850966,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.5121,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.4553261310130059,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.5319,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.46619057480282006,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.5521,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.45877742369150226,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.5013,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.4627247965418664,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.521,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.45492215344144726,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.5246,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.46918117977555757,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.5241,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.4679125192857745,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.5085,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.4574026710011303,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.497,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.4563488925435044,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.5196,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.4672743082233542,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.5094,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.459179802595566,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.5367,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.46051127100026656,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.5159,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.4657420082472283,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.5297,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.46879510730222,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.5152,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.4604264131599764,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.4967,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.45895167338306847,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.525,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.46613870416675907,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.5463,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.4518879464984705,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.5293,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.4623272105237057,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.5302,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.460940170654815,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.5457,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.4517896526552488,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.5204,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.4623299294885853,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.5145,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.4643493402132099,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.5273,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.4639021736120464,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.552,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.4616561722229895,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.5294,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.47210043906848587,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.5221,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.45220085827792583,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.524,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.47347630048599915,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.5317,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.46028711874586203,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.5109,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.46539227787953125,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.5451,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.47335258699229255,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.5344,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.4494668186730836,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.529,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.46244598684222576,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.518,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.46794683287846406,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.5298,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.45136384446349725,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.5169,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.46761256331828427,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.5142,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.46449344618359784,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.5236,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.45910306624527064,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.5245,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.4529138097559364,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.4879,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.46373273351104105,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.525,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.4614915473960331,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.4994,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.46367833034557926,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.5161,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.47964371586424875,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.5399,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.4666399514608963,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.5036,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.4672648923988541,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.5254,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.45426181781053054,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.5171,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.46520424411450795,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.5137,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.4641959677697355,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.5346,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.45971922048610664,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.5316,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.4531144513604987,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.5291,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.4731645141613083,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.5229,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.46060990674649643,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.5449,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.46519265885720706,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.5454,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.45430345019423074,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.5605,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.461118207476319,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.5248,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.46542410103625664,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.5357,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.45374062071397286,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.4959,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.460328757318496,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.5593,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.45637418653856876,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.531,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.4560232713409067,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.5346,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.4537126690715381,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.5462,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.47568569274042344,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.5199,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.46237152100383994,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.5408,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.46055017208331445,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.5058,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.4589132077788409,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.5168,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.4907700657809176,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.5179,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.4673664364332686,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.5278,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.4640970616986782,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.4946,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.4498200731507754,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.543,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.45992875060033966,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.5202,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.45775454668210985,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.5013,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.4605560197198589,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.5138,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.4661088420232132,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.5296,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.4559018611754207,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.5011,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.46342376692604986,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.5212,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.4682708964647222,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.5355,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.4610662627612966,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.5341,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.44969607685787794,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.5252,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.47047405976907425,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.5444,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.4566039884281351,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.4839,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.4647778817002192,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.5273,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.467036144640842,
+      "learning_rate": 0.0,
+      "loss": 0.524,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 1.5605192251460813e+18,
+      "train_loss": 0.8394831982460099,
+      "train_runtime": 13704.7732,
+      "train_samples_per_second": 11.529,
+      "train_steps_per_second": 0.09
+    }
+  ],
+  "logging_steps": 1.0,
+  "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": 1.5605192251460813e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}