diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8687 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 1235,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004048582995951417,
+      "grad_norm": 6.8419207838082,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.069,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 6.714141674019474,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.0445,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 6.865340298514671,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.071,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 6.742394324423914,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 1.056,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 6.319047583382961,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.0462,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 6.163381684954911,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.0515,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 4.944777476391676,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 1.0046,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 4.467753493818218,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.9912,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 2.9438716703702106,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.9579,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.5995122246516673,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.9525,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.6245086345508852,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.9609,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 4.09311052490772,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.9291,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.296234708359482,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.9223,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 4.057901205982037,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.9074,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 3.7041161967774623,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.9089,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 2.482766574010958,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.8585,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.3561935411820105,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.8405,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.0277391996699388,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.8342,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.5307140671223323,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.8116,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.3566875461513686,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.801,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.2785026702529054,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.8031,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.336305384268318,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.7841,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.1275333112571555,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.7695,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 0.8907873307863878,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.7574,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 0.9883747982465748,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.758,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.0537554210366356,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.7386,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 0.957374571547893,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.725,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 0.7900747618327876,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.7271,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 0.9628513738209611,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.7201,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.9489482595396356,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.7125,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.9193359773400351,
+      "learning_rate": 1e-05,
+      "loss": 0.7205,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.8774978502397276,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.7307,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.9118185471289296,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.6988,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.7782636083654288,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.7061,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.8731130812850824,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.7032,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.6230433891534902,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.6719,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.737470904359871,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.6777,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.6640316897026143,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.6904,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.6786771794588108,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.6917,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.7757758258491871,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.6662,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.6923185216956429,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.6883,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.6611884633328821,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.673,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.631283824486792,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.6696,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.637547514504739,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.6742,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.683714183426591,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.6686,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.6146910012075152,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.6712,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.7701668188100989,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.6519,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.5702428902156962,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.6651,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.7643314030163509,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.663,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.4965329726851037,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.6499,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.6040014947039244,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.6278,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.5648678723847734,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.64,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.5811855655575323,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.637,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.622232826893364,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.6507,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.6605551168438615,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.6535,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.6774224758255734,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.6331,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.5581518370491224,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.6255,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.656083001454514,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.642,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.6655910210109465,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.6433,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.6245475581887411,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.6256,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.716268858280956,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.6457,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.7101826122647915,
+      "learning_rate": 2e-05,
+      "loss": 0.634,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.8275642163738473,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.6371,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.82951244774213,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.6106,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.9639488061064423,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.6306,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.7973913988355769,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.6461,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 1.5457769245837103,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.6336,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.8302201324235244,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.6464,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 1.8495018852965601,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.6238,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.9627055281994674,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.6188,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 2.1413054064134394,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.6227,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 1.7224784283526113,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.6287,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 1.5845768732351915,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.6315,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 1.5184072244538755,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.63,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 1.2594935561994478,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.611,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.0132822848650198,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.6172,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 1.2973649028720322,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.6283,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.0676995345309535,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.6069,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 1.1632282992220246,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.6249,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 0.9528393561195238,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.6089,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 1.2873703811063246,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.6287,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 0.744491029730277,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.6238,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 1.2559865260866965,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.6207,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 0.7686907700575324,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.5944,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 1.104551870717243,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.593,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 0.8197765381811236,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.6076,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 0.9284804710418514,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.6174,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 0.8533372584091243,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.5946,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 0.9745118079879973,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.6056,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 1.147407224595906,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.6022,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.1721899105031324,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.6088,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 1.1407955772640428,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.5813,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 1.131494087136902,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5931,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 0.9610049426237944,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.5894,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.0985926224011293,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.5866,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 1.0366418348906505,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.5807,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 1.3400948534451373,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.6034,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 0.964218592474145,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.5952,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 1.4302519255391442,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.5955,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 0.9619761669995119,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.569,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 1.2774788042568848,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.5921,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 0.8971448079915789,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.5903,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 1.3218920753824595,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.5902,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.1036367027405871,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.5975,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 1.0284439808903332,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.5717,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 1.054039686381172,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.5823,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.0084413150283718,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.5985,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 1.0955454085059568,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.5775,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 0.9656792610405721,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.5734,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 1.078439688343585,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.5945,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 1.2740627171024186,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.5789,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 0.9945693139969629,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.5828,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.1343134532574328,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.5661,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.301506419542637,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.5751,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 1.1327916583749817,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.5767,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 0.9547454468591835,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.5643,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.8035033279992481,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.5688,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 0.8860342921013263,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.5533,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 1.0346580812256874,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.5887,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 1.3036641270291611,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.5701,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 1.000471577816299,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.585,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 0.9007725984769361,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.5692,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 1.2851027077198727,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.5686,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 0.7261806758510909,
+      "learning_rate": 4e-05,
+      "loss": 0.5578,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 0.8555029216503917,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.5615,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 0.9413429694860261,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.5744,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 1.136007202290212,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.5639,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.0195195909390915,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.5536,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 0.848648153776253,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.5672,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.8476331192777602,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.5544,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 0.9234640389640394,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.5625,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 1.0329103496041725,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.5457,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 1.0324369398577824,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.5601,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 1.0725492928127922,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.5635,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 1.2435601186987522,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.5753,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.8421474754753083,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.5669,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 0.8975246133469654,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.5541,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 0.7790485250031598,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.5239,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 1.0146057717539798,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.5303,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 0.8192124326498337,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.5392,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 0.7697841666421429,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.5417,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 0.9669567107845374,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.5356,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.8047658658623704,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.5348,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 0.9334643961413442,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.5313,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.9766149190649573,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.5416,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 0.821278680598867,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.5067,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.7557353662215545,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.5414,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 0.8134991490485881,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.5454,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.6729961925801907,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.528,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.7377678288548787,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.5463,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 0.9367798050660798,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.5297,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.8830385193551918,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.5285,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 0.718408508246984,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.534,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.8975505615679714,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.5403,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.6034272306314863,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.5469,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.6706912352340534,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.5381,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.6402933715782599,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.5277,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.7107627757154072,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.532,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.5929839913104396,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.5195,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.6318565725748744,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.5341,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.7347372875166216,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.5535,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.6778860424645389,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.5325,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.7431354286676835,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.5053,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.5931368045854893,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.5249,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.779087864976502,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.5381,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.6422686625579448,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.5234,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.7561035823386832,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.5139,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.9137502009882038,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.5276,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.8396387610585635,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.5228,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.6245889885650763,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.4981,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.7272286122239915,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.5233,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.7825852349208883,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.514,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.9211248850139325,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.4986,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.661979616118445,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.5018,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.7920306490026219,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.5316,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.7789804162298627,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.5127,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.7692515649705358,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.5206,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.6046226016408913,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.5064,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.8697761981384178,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.4921,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.7383932835816347,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.5015,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.6131833640129614,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.5001,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.6837833743654159,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.5176,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.7149204037417801,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.4944,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.56975740781647,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.5003,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.5908407643243749,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.4951,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.6369762209036236,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.5002,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.8010623361716604,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.5167,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.5124601158867411,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.4933,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.7016336711340103,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.4828,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.6320330280395398,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.4913,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.6035943904794591,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.4817,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.5342314723961504,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.4468,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.6763510992190495,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.4816,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.7503729302571788,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.4658,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.650351404762163,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.4798,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.8704434346882466,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.5018,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.9394045567537891,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.472,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.7512638337677776,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.5055,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.6094964922282291,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.5009,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.9458794515734642,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.4887,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.8632864685161494,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.482,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.9586906761500834,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.4426,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.8315289529737999,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.4959,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.7060474318543982,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.4758,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.7529289414875708,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.4754,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.7080698728273159,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.4798,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.7147386940949388,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.463,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.49638321330135265,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.5049,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.6282437126310544,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.4731,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.6365284469099577,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.4701,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.5806511938689413,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.4897,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.6309054395297136,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.4967,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.5428914516148093,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.4997,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.5233651238011315,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.4602,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.606590187275132,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.4788,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.6883528192233642,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.484,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.5140295591300175,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.4804,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.5148765386008198,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.4626,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.6008377754642922,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.4512,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.6259105840053233,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.5095,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.4508927942894163,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.458,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.5521808408602142,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.4303,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.5020361114956705,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.4368,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.5150225357581372,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.4423,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.47699934043273956,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.4356,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.5772099051736128,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.4574,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.6833268218383349,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.4492,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.5484302530026526,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.454,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.46604012999727124,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.4431,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.5783991381639668,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.4452,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.6288865683178803,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.4207,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.435574385478961,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.4524,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.5021957868516655,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.4639,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.5506018605913123,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.4384,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.6213231443920317,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.4063,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.7499258366512295,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.4495,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.6795622209387177,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.4388,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.45428202428932896,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.4319,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.6301232020551378,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.4617,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.5175025280453827,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.4598,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.5441844071817823,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.441,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.5246852801584035,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.4665,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.48483194012110636,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.4467,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.5004152096028142,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.4299,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.5429819057374166,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.4338,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.579649559597786,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.4252,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5655833867046296,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.442,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.6135074980300468,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.3902,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.5301774054381727,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.3868,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.6398415455456686,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.4027,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.6477199680619679,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.3748,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.5067034923408934,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.3981,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.5954806620434017,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.381,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.5731413485121852,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.3777,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.5128843667296871,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.3649,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.5325092565540996,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.3601,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.5326963753497516,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.3731,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.468385113720668,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.3771,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.478166478654015,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.3821,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.533233069275888,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.3751,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.48395078782709783,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.3929,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.40399665805993196,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.3562,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.5052147855484459,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.3828,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.5238470317121442,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.3677,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.49396910772581537,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.387,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.44270811142021094,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.3395,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.5199978670821648,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.3735,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.5822864705668847,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.3778,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.4814342582026144,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.3634,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.7749052272059241,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.3953,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.5952921262633325,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.3634,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.4904339085171513,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.3475,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.49065073214332466,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.3689,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.5454509237602865,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.3708,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.4653137768998122,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.358,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.4956061194928068,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.3502,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.538180913394385,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.3537,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.5005753461607118,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.3452,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.500121101306834,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.3669,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.5113540091540985,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.3444,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.6608000472145337,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.3571,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.5961076786742423,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.3363,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.4946559886110739,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.3561,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.5456876322180483,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.3699,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.601424849564865,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.3561,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.5591310537869073,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.3349,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.49331910832305226,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.3181,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.5308581458078702,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.346,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.4691103920059366,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.3459,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.4570457823996922,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.3515,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.44635690331900435,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.3611,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.42202933452868646,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.3703,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.42615850922494297,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.3366,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.5332295340259343,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.3387,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.47985320780721574,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.3453,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.4528280775683642,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.3512,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.42408916673779534,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.3555,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.423754844511852,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.3424,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.42261750166677403,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.3064,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.46678733530997235,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.3208,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.6013721273048744,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.3708,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.6403259260099291,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.3185,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.49825286491592274,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.377,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.4175798886029385,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.3073,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.39541326443349195,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.3082,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.5125821953356609,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.3321,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.4688720510797511,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.3248,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.48420595009875605,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.3144,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.5858178843970941,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.3701,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.4195657394053345,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.3273,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.5010264298639663,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.3319,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.500302406226112,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.3271,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.44050247634893736,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.3169,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.5075889304727471,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.3442,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.4516415910007703,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.3438,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.5358742834413363,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.3091,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.44057734332675097,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.3258,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.4618069496185687,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.3355,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.43183527136128175,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.3421,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.45087685799197497,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.2906,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.4614021197404614,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.3188,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.4666425967175284,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.3354,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.5187792416008564,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.341,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.5500101843742847,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.3274,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.5645335292634046,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.3153,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.41402984036297047,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.3401,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.5555958243628459,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.3048,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.5061327677086204,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.3608,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.485058073189369,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.3261,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.5112290110615253,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.3187,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.4055678442207254,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.3164,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.611296218896908,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.3226,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.6050377875379331,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.3124,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.4337450720004949,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.2965,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.5849279655903681,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.3011,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.4493095774978499,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.3211,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.5200847559698538,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.3249,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.5016658024316953,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.2682,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.4459154137702164,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.3187,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.5504371290760967,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.3136,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.5162237292429935,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.3069,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.5920717948764,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.3414,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.5096429857492301,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.2825,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.473976679877754,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.2845,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.5368105020743816,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.2753,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.5652962257245828,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.3228,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.40836756525270446,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.3001,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.5506247156351659,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.3293,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.44244972487292633,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.3219,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.43233114562213926,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.3141,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.4654076108726684,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.2833,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.44752591741033954,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.3384,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.43192093918687435,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.2974,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.4206049067124765,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.2673,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.46266422168150345,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.2944,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.5088247754396241,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.2982,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.48039425965710875,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.3531,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.5826208736790356,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.2886,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.4598394129763192,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.311,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.5461068768012293,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.2734,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.5766892356926656,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.2685,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.4707023523615186,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.282,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.6074624050302415,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.2814,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.4715027535182993,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.3133,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.5825268821061756,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.2739,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.3970368055192549,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.2825,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.759010248517992,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.2767,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.5406113501538095,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.2718,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.6344151524278276,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.3095,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.4604106512601274,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.2735,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.5744900271426252,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.2738,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.4885415602919859,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.2643,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.6181003284296804,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.2868,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.45773000763853594,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.2866,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.6285632406314516,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.2902,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.44041898626603443,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.2924,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.47769126564404185,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.2842,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.4269430370742871,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.3331,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.44379909407896884,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.2701,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.4189658276326786,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.2653,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.46036055207756166,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.2775,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.40829083399258476,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.2615,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.44059801412643174,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.2873,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.4058520847475071,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.2809,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.37685514098956546,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.2754,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.40875125294196435,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.2654,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.4091903359964023,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.3021,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.3877243977716942,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.2595,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.4398178295814742,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.2583,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.42294885751884675,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.2756,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.5060167390750107,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.2828,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.45687446484548205,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.2681,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.4250622993374131,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.2859,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.44579802550555664,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.2589,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.40464815804634685,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.264,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.4022439929089541,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.2689,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.4378192226443805,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.2648,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.40293443168491855,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.2619,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.46653406869294944,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.2685,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.4674596285901746,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.2758,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.39296016445189685,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.239,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.4024747195477908,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.2348,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.4502453215931355,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.255,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.3545075377476257,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.2728,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.4878431467737063,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.2631,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.396774860992172,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.248,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.38059118475915865,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.2299,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.4126433729043668,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.2584,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.38921370079473255,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.258,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.3832313705514346,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.2401,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 1.210020688728795,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.2608,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.35896143984409734,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.249,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.426084449789151,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.2413,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.3966196138072835,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.2513,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.42079272766900805,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.2837,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.43337885714820507,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.2698,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.4828813427261293,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.267,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.4140236503329523,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.2801,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.49129997072088544,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.2621,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.43569127082391396,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.2253,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.3869595290311914,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.1976,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.4169705257165773,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.2292,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.4342938914452435,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.2372,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.3954585834442909,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.2196,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.41977419149951545,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.2292,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.38193557383386045,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.2301,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.44867982145161167,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.2573,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.4244847501369856,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.2403,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.40710188487621773,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.2504,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.4318221085415973,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.2447,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.41727541820814334,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.237,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.40770086138079364,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.2067,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.47401126187554277,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.2491,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.37928928287682145,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.2518,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.5014457264206669,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.2655,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.48152120902004986,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.246,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.4445749864021211,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.2329,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.4731225794083456,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.2253,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.529201997215349,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.2412,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.4185862908863047,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.2198,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.4319281439834333,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.2386,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.5585478008179572,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.2434,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.4624140175041264,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.2179,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.4367578718280598,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.2412,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.5672996432533312,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.208,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.4277027959430167,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.1987,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.4366899443436819,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.2278,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.4888571435080264,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.2333,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.396462065139919,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.2233,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.4338410668124771,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.1996,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.4288430608973884,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.2057,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.43126431078230704,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.2286,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.40021924325266195,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.2221,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.4136291640455241,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.2138,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.40090574659636763,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.2182,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.38665025752287163,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.2069,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.4044721561906851,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.2226,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.4814952553212686,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.2206,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.39638124853981926,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.2067,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.3934790717824548,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.2046,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.40497062683976887,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.2006,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.3844525179705836,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.1999,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.4220468867135266,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.2225,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.39268735233234153,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.2139,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.3976817495887603,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.2245,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.39890323883184425,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.2152,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.45571439045471396,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.2249,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.4881947299540926,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.215,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.39026885324152777,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.1995,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.4149833924470789,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.1946,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.3754205149667385,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.2046,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.43513961049972055,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.2159,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.3789383784122094,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.1959,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.4543770792323818,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.2223,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.39823963647964455,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.1827,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.40684396972896064,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.1974,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.45389413003251317,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.1948,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.37857315914794554,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.1696,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.4179255142890502,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.2045,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.3954162635747155,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.2002,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.4520888492855324,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.2008,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.4126478587118523,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.1998,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.4509641399470838,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.1958,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.4239542400659815,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.2046,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.414588964799034,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.1892,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.4159669698227553,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.1993,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.3567696653971091,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.1918,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.4401260344866331,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.1929,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.40051339479717796,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.172,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.45113694961542533,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.1815,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.39227756890144433,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.184,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.46293196157726096,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.1927,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.4211823736908664,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.1922,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.43671238659870376,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.1927,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.4588603419813441,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.153,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.368071755739681,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.1252,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.5257656833558906,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.1501,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.3946454991956238,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.1523,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.4176194998855526,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.1174,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.3795505774157025,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.1345,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.4396320102726141,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.123,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.3410490805728742,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.1409,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.42316764074592245,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.1355,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.36340961354860685,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.1207,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.4542684930971316,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.1233,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.4041730167448343,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.1253,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.3609779686974304,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.1274,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.4032546689979691,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.1138,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.35300483017307405,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.1397,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.38284756590150754,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.1159,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.34349080852714964,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.1252,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.4051239406158845,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.139,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.36087997651141557,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.1292,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.336951227767331,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.1092,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.3582081052592236,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.133,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.3362857943653978,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.1381,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.329724848185343,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.1034,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.35050772802390207,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.1322,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.3738094217660585,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.1113,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.3732514886033079,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.1356,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.3441629642850979,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.1312,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.3052326815115122,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.1022,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.33082064301335945,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.1197,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.34304857327042315,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.1233,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.3233482308207224,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.1285,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.3250566382292908,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.1167,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.34221186372767476,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.109,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.32699448911638246,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.1095,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.34179354426496394,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.117,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.3500438557324849,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.1384,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.31901920154341373,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.1119,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.3368737679406321,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.1143,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.3376656101382426,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.1159,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.3209711359164756,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.113,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.3268544320235726,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.1121,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.31756809439519973,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.1247,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.32681563841057454,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.1267,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.29982513467437816,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.0999,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.3548082027050364,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.1333,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.3146588943835892,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.1098,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.3263128453715242,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.1108,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.3681099638860991,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.1227,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.3470939759868883,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.1187,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.34142524419973475,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.1164,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.33707661824993573,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.1105,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.3519585481885585,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.1176,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.3327846105359143,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.1245,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.32981800205849726,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.1022,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.3341904935387382,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.1156,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.32658494202730043,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.0991,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.3369092447552172,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.1211,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.34663952346504534,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.1085,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.3099753062118295,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.1105,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.3433362484805364,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.1325,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.3202263110309071,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.1065,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.32691136796733605,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.1213,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.314994979172895,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.0935,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.307693991608665,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.0979,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.339121089196316,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.1046,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.31596476487989417,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.1102,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.3473723062157903,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.0971,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.29560177590683506,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.0888,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.3421593439827455,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.1041,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.32511071693988985,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.1075,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.34571803835120934,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.115,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.29854668907549525,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.0926,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.36622495303007435,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.115,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.33467117960352905,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.1094,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.35662264603360067,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.1054,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.32686479491153164,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.0966,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.34827436488788827,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.1287,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.3283290198252523,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.1081,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.3640967469283413,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.0972,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.3319946903238166,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.1029,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.3708749573248257,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.1115,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.3127558427669913,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.1113,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.33313258080866026,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.1047,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.34499331956190393,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.1211,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.33292410808954676,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.0803,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.3104631688479102,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.0952,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.2998538710061506,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.0909,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.3008106525902097,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.0909,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.34411715872595033,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.0977,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.3307709075029269,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.0981,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.3297464322988875,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.1004,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.3581990185866317,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.0988,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.3428853094070213,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.1098,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.30128273429819175,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.0944,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.33674113147631174,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.0839,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.3240379734039465,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.098,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.3258867968383589,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.1032,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.31598471470409517,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.0893,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.31850980305999227,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.0902,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.30819086929639683,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.0928,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.317630952506878,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.0937,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.3495351127544828,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.0975,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.3367816851771483,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.0992,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.3420555866901988,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.1082,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.32610228672442215,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.0847,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.30732692940910394,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.0741,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.3893908447702343,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.0908,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.32134203520872934,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.0859,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.41816794226988624,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.089,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.3475332226595739,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.0942,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.35615333096295304,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.0957,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.32362666054147265,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.0838,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.31070979924182857,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.0849,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.30047495574344263,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.0876,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.3137092886055063,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.0771,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.3250568744863078,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.1024,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.3037579593502572,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.0842,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.5436202868521243,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.0928,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 1.4076837531471915,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.0793,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.4106580876389566,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.108,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.38295471114388757,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.0745,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.33957267088933357,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.076,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.33510597742360815,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.0779,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.35527313653778203,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.0931,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.3044033705142739,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.0738,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.3168035491552008,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.0818,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.3063095636656837,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.0724,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.31751869905376234,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.0756,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.31962500391129395,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.0978,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.30309292412050043,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.0732,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.3126650960905874,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.0647,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.34565262526621615,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.0906,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.2969038998619364,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.0618,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.3019895947359554,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.0773,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.3202260480014226,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.0866,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.2831542751760487,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.0751,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.3050835144128416,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.0778,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.3297556149475966,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.0725,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.2840132888444444,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.08,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.3237872201116535,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.0753,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.3020086972475046,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.0683,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.28364836140715455,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.067,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.2731323993302286,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.0565,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.30172499909720646,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.0753,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.31473257178457087,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.0719,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.2995444954191081,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.0752,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.2864310654844743,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.0644,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.2756409588840138,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.0648,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.29188606181642385,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.0661,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.27627988518369256,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.0624,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.27799921442097997,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.0675,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.3027915219434075,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.0676,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.2761859918110921,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.0692,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.29603745668203063,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.0731,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.2892919748096696,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.0775,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.2885733959196583,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.0641,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.29867077860206986,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.0731,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.2811620461572549,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.057,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.282050617856572,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.0628,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.31758836923558637,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.0669,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.28890113137560314,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.0685,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.28727305020830396,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.0586,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.28121600279497116,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.0674,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.27423875237681095,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.0558,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.2854704647626315,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.0584,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.2826966279579198,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.065,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.29262999048319743,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.0625,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.2733515471502373,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.0618,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.3525443412905111,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.0715,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.282680294278483,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.0579,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.3162413757898801,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.0638,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.28298984559551427,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.0698,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.2843324974032262,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.0466,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.2904558585541804,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.0657,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.2928102547372279,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.0604,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.28565589537617475,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.0612,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.2874221632809731,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.0665,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.30134019740773915,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.0706,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.2857480537739323,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.0587,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.28981683801672287,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.0574,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.2823357217535641,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.0523,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.3408711114078049,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.0691,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.27517613812831376,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.0607,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.28038496288635884,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.0562,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.2864718708819575,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.0662,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.3073130737176036,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.0644,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.27925281555172476,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.0558,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.28864907763380654,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.0494,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.2837023104927078,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.0719,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.2696454940710617,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.0502,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.26393999279334096,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.056,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.3011508664298952,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.0644,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.2689934723930213,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.0562,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.27291719016518795,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.0565,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.2636166299851386,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.0538,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.2849142200085252,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.0579,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.2474755199975639,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.0536,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.26878339076977287,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.0517,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.24861391541176184,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.0518,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.2679327801756367,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.0481,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.25981631374427955,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.0469,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.24812363066863152,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.0399,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.2778953249676859,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.0614,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.2859366939363036,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.054,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.288981087621676,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.0589,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.3043804662045663,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.0636,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.26591063798811715,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.0508,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.28651438802486917,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.0583,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.25816064578832404,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.0458,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.2876995216170418,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.0538,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.28414816832442785,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.0591,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.2798844325559107,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.0456,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.27416168765306553,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.0505,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.27686268490041804,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.0515,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.3126488111777764,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.0489,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.25834255348401913,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.0503,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.3082873361378363,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.0494,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.2682325002882332,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.0533,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.3241320860474771,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.0541,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.32604761472461957,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.0645,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.2784244188430634,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.0492,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.26803377526605293,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.0434,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.2661307062113881,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.0482,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.26337019936446143,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.0461,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.28849513493645834,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.0359,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.25720283008380873,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.0435,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.25816110494025146,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.0462,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.26676981035696484,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.0406,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.2757901019854911,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.0488,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.25654408195650835,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.0483,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.2786185115455833,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.0394,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.27255827557014733,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.0465,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.2562658739850307,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.0431,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.2693051264383689,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.0413,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.24711092492654047,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.0435,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.2722683621227424,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.0452,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.2976554154908803,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.0373,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.24201245115743394,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.0349,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.25221535443276466,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.0382,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.2575919537975062,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.0413,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.25845344462256975,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.0458,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.2726148006243744,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.0345,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.2551801332238337,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.0371,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.2616502333589699,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.0353,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.23241475181452753,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.0256,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.25634448975147944,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.0384,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2498248133237537,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.041,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.2044943138689192,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.0183,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.2165065323926062,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.022,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.2164104204654174,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.0253,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.2014892767185973,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.017,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.21418078201479115,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.0233,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.21564234260382095,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.0232,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.1984283213318292,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.0179,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.19629600312442871,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.0187,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.20795646901803777,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.0215,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.22092161172959332,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.0244,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.21890614915781278,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.0233,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.21699907167318527,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.0248,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.20063628343658813,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.0209,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.2065713682620512,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.0223,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.2279181168430412,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.0204,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.23347652853308712,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.0201,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.2093400582046256,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.0215,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.22211932033642598,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.0214,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.19339342195346013,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.0184,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.20743370601779407,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.0191,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.24377024464507427,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.0196,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.21451834322382138,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.0204,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.2185457176059151,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.017,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.2137094819888845,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.0214,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.20007384951089346,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.017,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.21264565817401454,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.0191,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.23155838607350968,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.0202,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.1950386880506152,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.0184,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.21134890482508295,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.019,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.2052627022675823,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.0186,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.1983088621535463,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.0225,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.24208296676718788,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.0237,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.21700672014916958,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.0247,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.17290862584807926,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.0171,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.20019056600226545,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.0167,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.19299670903493998,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.0173,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.1999132041583053,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.0193,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.21419696387265938,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.0208,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.22232534588306402,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.0183,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.2041926217865339,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.0201,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.18485198452625112,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.018,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.1792168771627224,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.0136,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.1978233974860807,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.0189,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.2197269641399418,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.0181,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.187207053042676,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.0148,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.19772702521982283,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.0207,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.1915610621050468,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.017,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.20386776658254982,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.0199,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.1910031719819258,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.0177,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.20372083848714487,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.0241,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.21102348175821092,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.0183,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.22445707020176137,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.0223,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.1791095442108656,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.0151,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.19422822769445763,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.0178,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.18880611278067044,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.0151,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.18812832136874588,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.0167,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.18222465702768517,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.0164,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.16955354919383783,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.0138,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.20338833443253607,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.0164,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.20124689665760737,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.02,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.18073516136445508,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.0142,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.1769427029669363,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.0141,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.19590622030750066,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.0179,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.17425729244013974,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.012,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.18659401536923478,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.0159,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.1932450013464998,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.0165,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.18558653431492403,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.0148,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.17911594231819045,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.0141,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.15804508625230174,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.0126,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.18163150782905363,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.0154,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.19317178950431108,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.0173,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.17622452964325797,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.015,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.1706879638939546,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.0137,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.18014820234772685,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.0149,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.1972980219579852,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.0178,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.18380050668669132,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.0171,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.17596309379890973,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.016,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.18972388534667803,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.0169,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.17880047161192977,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.015,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.18273736506772198,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.0152,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.1873568235911105,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.0124,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.17541103230368932,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.0138,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.17579260308381955,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.015,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.17214675112561656,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.0113,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.17177513067199407,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.0115,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.2002574703910759,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.0174,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.1826945504834934,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.016,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.162806902290428,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.0128,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.19805758818672398,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.0154,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.1751417335087378,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.0141,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.16239002018520954,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.0104,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.14382949453054059,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.0074,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.16788066384130065,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.0105,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.1618830343155672,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.0104,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.16617863753695042,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.0128,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.20625878081468504,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.0129,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.18205723938094906,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.0127,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.18464545444127328,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.0144,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.1607296933010911,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.0115,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.18491058182236045,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.014,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.16302925453190859,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.0126,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.16725431208246724,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.0102,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.18889042158302097,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.0147,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.16378212832612576,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.0139,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.17484398662865694,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.0119,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.18531875214005464,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.0117,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.17941449486667982,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.0123,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.14958135179753826,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.0093,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.16365284374032732,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.0137,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.1496716832853256,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.0109,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.16470220119025047,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.0102,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.169389568227208,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.0127,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.14551881481691056,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.0079,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.15012649914364712,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.0085,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.17471754390330602,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.0141,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.1526359563528617,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.0092,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.147494431789975,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.0107,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.16213563914144216,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.0114,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.17409400071991066,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.0148,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.17662973675930643,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.0135,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.17427613216141424,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.0156,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.16012798918465762,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.0117,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.14425407210075497,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.0084,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.1610879722025783,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.01,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.1909775072397994,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.0123,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.1591395594747094,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.0107,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.13190858038219563,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.0072,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.16087666214998136,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.0106,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.15124184191074827,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.0098,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.14886009848772633,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.0078,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.14248471634548857,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.0091,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.15441676255798903,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.0121,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.14467794853257362,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.01,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.15878630459197435,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.0115,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.14063540981527045,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.0101,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.14266275119394686,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.008,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.17897197107199286,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.013,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.17495340476391508,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.0121,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.13848890753132523,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.0092,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.15743780887298905,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.0123,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.14977024707645795,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.0095,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.13877889663985318,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.0096,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.1564618752498682,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.0094,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.13366451076161856,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.006,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.16883380836334444,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.0078,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.17103647643214914,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.0102,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.15123604840315527,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.011,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.1409317561754041,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.0063,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.16175687328057306,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.0114,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.1284807359682832,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.0069,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.1435859379271588,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.0075,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.14940994717518155,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.0096,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.1311627394470833,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.0073,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.13461189747086558,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.0079,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.1302785636657104,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.0065,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.13493084892888793,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.0078,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.13343587346748367,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.006,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.13662406244183758,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.0094,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.14995153597323718,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.0077,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.146001508981887,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.009,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.1339267421380569,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.0083,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.11739534842053492,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.0056,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.11022896622194665,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.0056,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.12644718138438574,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.005,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.12623516625903522,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.008,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.11299384901501569,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.0064,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.13280712612393367,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.0064,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.15286393535523746,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.0075,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.134540252441549,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.0085,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.27916757872284004,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.0058,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.133599111602928,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.0078,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.11786040045605051,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.0061,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.11575475055225143,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.0069,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.13233072330660461,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.0087,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.13158253397525177,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.006,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.13584894663222905,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.0059,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.15170528629099864,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.0075,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.130576781701456,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.0064,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.12296832214239824,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.0067,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.12038835153812869,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.0056,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.133111614509901,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.0069,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.13087234157869437,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.0078,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.1273639738876921,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.0064,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.11165663837138703,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.0042,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.14228101899517306,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.0077,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.1068192771302476,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.0038,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.1356574270461618,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.0057,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.12483116189719787,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.0071,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.13101192185631608,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.0063,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.1260108697374098,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.0075,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.12310644053925805,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.0064,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.13807593772549748,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.0092,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.12656845332152725,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.0048,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.1236079698760979,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.0053,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.12993097168380655,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.0055,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.13530191378244472,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.0071,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.12781853029250928,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.0065,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.11139242040356792,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.006,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.09583419376356632,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.0036,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.11027578026524819,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.0051,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.1159358559972987,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.006,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.11840594661588408,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.0063,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.1464906700961494,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.0068,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.112483915386537,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.0053,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.10082360408823862,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.0043,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.10525155323863188,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.0054,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.10252532039908607,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.0039,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.09288616539881057,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.0038,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.11088062224939021,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.003,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.11243330475640469,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.0061,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.10847872256401742,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.0048,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.10225340374116573,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.0041,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.12589404165352294,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.0052,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.11374339727276123,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.0043,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.1019571732697141,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.0045,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.10970105361869135,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.0049,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.11643275164685175,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.0054,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.09502326149713393,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.0048,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.11448925890549842,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.0045,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.09848562912141172,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.005,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.09760446531924365,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.0034,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.09651621961140472,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.0038,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.10144431051565506,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.0045,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.11624992996051337,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.0058,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.10734224458941953,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.0051,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.08940413593223702,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.0044,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.08905657412735957,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.003,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.09627595571893918,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.004,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.09586102724461155,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.0037,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.10947873473833754,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.0049,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.0963516233300949,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.0041,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.09625112360839344,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.0034,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.08360901345634232,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.0025,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.08878824537131717,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.0039,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.09071794800047392,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.0036,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.10726538679929866,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.0046,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.09952681478691265,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.0043,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.09356799857314299,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.0037,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.09579738155920435,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.0034,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.08734359739586901,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.0033,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.0695095952555885,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.0025,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.09884556332680335,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.004,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.08868328506105598,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.0038,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.102055760401982,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.0036,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.08923375651928588,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.0029,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.1381687781895922,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.0047,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.0817360253219779,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.0033,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.05844469203184074,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.0016,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.05343370047422418,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.0016,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.061503502079780545,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.0019,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.0558566177598099,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.0017,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.06560291350861519,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.0023,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.07692295229671886,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.0024,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.06276499797241099,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.0022,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.06641424780253716,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.0018,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.0556321096588536,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.0014,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.06991132056405301,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.0021,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.053543556706633966,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.0014,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.05546451226072442,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.0013,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.0625045418284395,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.0016,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.06011613380781655,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.0017,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.06406288529478224,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.0021,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.05932617612340494,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.0019,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.05086663498375656,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.0016,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.07859199572014194,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.0016,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.04972384246507023,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.0015,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.06516407550010962,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.0017,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.0623220117756796,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.002,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.062342146458976476,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.0018,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.06302832587044777,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.0016,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.06301388232213519,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.0022,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.058351349644438895,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.0019,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.05759194681727135,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.0014,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.06254565059177704,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.0014,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.06757356559860278,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.0022,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.052732735026790066,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.0012,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.06869659362122482,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.002,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.0662099541411707,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.0015,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.0491917725429393,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.0015,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.05101812801064782,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.0016,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.058565992204640066,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.0016,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.05456170032337532,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.0015,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.056269568001385675,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.0017,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.05673810397989378,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.0016,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.07383227346432791,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.0021,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.052924257679441145,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.0016,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.05245174720793023,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.0017,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.04585376486806312,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.0012,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.0708828906491889,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.002,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.05664061193485882,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.0015,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.05817936525600852,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.0017,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.05554574298635371,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.0014,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.06515670923921028,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.0015,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.0659377743261773,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.0014,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.06257101691137239,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.0017,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.04672633405972623,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.0013,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.045570983078136895,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.0011,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.04292254360671153,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.0011,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.047752358645853886,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.0016,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.04781769309805918,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.0011,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.032242349033290846,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.001,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.04811957504839937,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.0014,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.050758708924673955,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.0011,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.05372969612235076,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.0016,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.057888649422847584,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.0017,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.05589857046750797,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.0014,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.06100013028579527,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.0017,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.0469747672749322,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.0014,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.055756463299578146,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.0012,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.04563496780178556,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.0016,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.04439222393865118,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.001,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.05413744892846423,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.0014,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.04831412915429541,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.0011,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.060794321182460975,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.0019,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.05122997445906406,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.0013,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.03237624463666633,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.0008,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.04618751539051554,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.0012,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.05048137586137536,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.0015,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.049468008203215744,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.0013,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.045532859735278305,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.0012,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.04630657831154439,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.0015,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.05760499363314879,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.0018,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.05338325822033949,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.0016,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.05304965925247524,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.0015,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.053203541569067016,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.0015,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.0562574310358345,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.0017,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.04446832646586928,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.0013,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.04505509551997334,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.0012,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.055289407049551036,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.0013,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.04298165075032528,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.0013,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.03293513420232301,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.001,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.04647047864689705,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.0012,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.04474590534277902,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.0015,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.06209361030845431,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.0017,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.03764737269977762,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.0008,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.04309422424732535,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.0011,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.047107416780463506,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.0012,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.043728432807820156,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.0014,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.031184176728525935,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.0008,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.046206463615926424,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.0011,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.05016735506083668,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.0013,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.05239321176397563,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.0015,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.05106998582791536,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.0015,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.064515591297552,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.0015,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.04528948463000288,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.0014,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.043580813880291204,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.0013,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.03879545283545901,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.001,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.04248367363392264,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.001,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.04938694837880631,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.0014,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.041804385779217884,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.0015,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.05790350015625083,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.0015,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.04185722079028421,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.0011,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.048376137002086386,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.0012,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.03988918632872776,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.0013,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.03462599434732374,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.0011,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.044661871520436804,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.0011,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.04259637241343482,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.0015,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.04743499903273526,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.0013,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.04425861872180874,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.0015,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.03229017474193582,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.0011,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.05104982737564717,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.0016,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.03661598776622372,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.0011,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.0447060538540874,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.0013,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.0405205562959491,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.0011,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.03325752960012097,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.0008,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.04226694791981145,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.0015,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.046904788063477004,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.0013,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.056766189941008086,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.0018,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.056418497289406545,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.0016,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.04595745935678094,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.0014,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.03867771815000152,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.0011,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.03713495859459256,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.0012,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.04335956615979452,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.0011,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.038008161192040056,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.0013,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.03028129800479903,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.0007,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.041464945599717834,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.0011,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.043017327470345666,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.0013,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.03359932109507887,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.0008,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.042846810882208264,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.0011,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.03848089044786708,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.0014,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.02457688422736028,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.0009,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.02960256094108256,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.001,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.031074852460647586,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.0011,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.03638564044526656,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.0011,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.02651338082697707,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.001,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.03910912693785035,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.001,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.03993409673974867,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.0009,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.031265353755079134,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.001,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.040134395548207046,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.0009,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.0352979382829372,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.0011,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.03049784860102486,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.0009,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.029063846486500976,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.0011,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.026438287401541746,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.0009,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.039363674122002645,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.001,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.03591503218764929,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.0013,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.05028406071629587,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.001,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.02204877586611783,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.0007,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.039021447793917266,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.0012,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.03797269149183884,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.0008,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.03908232740423514,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.0011,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.03461322087038239,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.0012,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.028805124679987304,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.0007,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.0376316842666059,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.0013,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.03500038180540507,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.001,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.03759798219773018,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.0014,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.037177517698712984,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.0013,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.04362954270412902,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.001,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.029903520646063966,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.0012,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.03530846668656541,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.0011,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.03911658681665948,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.0009,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.031714772763201095,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.0011,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.039942377673003016,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.0008,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.03804511236135309,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.0012,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.03549585451396993,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.0008,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.027317041958247683,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.0008,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.02988478260552174,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.001,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.034199121026011664,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.0012,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.027112826834280658,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.0011,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.030202891934009657,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.0009,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.03426287935112351,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.0011,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.02896165649734155,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.0009,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.03358426815959833,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.0012,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.03369516832324002,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.0011,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.02414562116250302,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.0008,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.04107836934608845,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.0016,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.030179002867904217,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.001,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.023436289888154126,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.0009,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.03525686105266216,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.0014,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.033039849916936995,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.0009,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.02934403844686922,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.001,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.024626618613387694,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.0008,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.029915549447520305,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.0011,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.02637886531895613,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.0009,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.03177277478383307,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.001,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.033944681750539774,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.0011,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.058650496681608245,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.0013,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.037928558284488795,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.0011,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.025943120026450987,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.001,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.029711952119672095,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.001,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.029249047793481586,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.001,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.032691325331772485,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.0013,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.03559601100863972,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.0014,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.023669451972448914,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.0009,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.034921403337837185,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.0013,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.02919215473521776,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.001,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.028790636333629605,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.0009,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.028490377032194553,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.001,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.025891304493957542,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.0009,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.024975565609812365,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.0009,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.02743406203221405,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.0009,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.02829964239601064,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.001,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.02816814463111912,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.001,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.023617776397787126,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.0008,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.03803473423967851,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.0014,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.02395897384250937,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.0008,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.033697408920820786,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.0011,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.04024617650763686,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.0012,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.0289756462411807,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.0009,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.022475463348476294,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.0008,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.0318706686195332,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.001,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.02854534030921803,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.0009,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.03905625734266262,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.0011,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.029295285749092327,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.0011,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.030093083521443315,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.0011,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.029545210216962204,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.0011,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.03106874660989388,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.0013,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.030255074395219378,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.0009,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.03814469840261027,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.001,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.032280712807751494,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.0011,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.0303080468251903,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.0011,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.029859923654024515,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.001,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.028822334466000844,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.001,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.025775091017652955,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.0009,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.03058056176065662,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.001,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.032231956084000195,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.0012,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.02871793493412294,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.0009,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.023742747356194086,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.0008,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.034971463984121695,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.0012,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.016558322666488445,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.0006,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.03154105854972151,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.0011,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.026550638029888517,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.0009,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.02977246326046892,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.001,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.030691072667718197,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.0012,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.02568121953494789,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.0009,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.034101217882006596,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.0012,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.03269191414079156,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.001,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.02999315242135729,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.0011,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.03135071466558241,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.001,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.028163997845350907,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.0009,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.031062594100668556,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.0012,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.03422846237575765,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.0012,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.036909885150032656,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.001,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.033025570598517975,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.0013,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.026468392263283555,
+      "learning_rate": 0.0,
+      "loss": 0.0009,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 3.854022017808335e+18,
+      "train_loss": 0.19459536587399984,
+      "train_runtime": 27931.0386,
+      "train_samples_per_second": 5.657,
+      "train_steps_per_second": 0.044
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1235,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.854022017808335e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}