d1_science_gpt / trainer_state.json
neginr's picture
End of training
9d6c538 verified
{
"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": 6.54252503289313,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.0582,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 6.6527955680127375,
"learning_rate": 6.451612903225807e-07,
"loss": 1.0753,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 6.786409984803582,
"learning_rate": 9.67741935483871e-07,
"loss": 1.0603,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 6.560218464841574,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.0678,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 6.338319969380865,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.0528,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 6.250557510037696,
"learning_rate": 1.935483870967742e-06,
"loss": 1.0488,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 5.179864154720911,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.038,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 4.706663774522184,
"learning_rate": 2.580645161290323e-06,
"loss": 0.9885,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.631532355882913,
"learning_rate": 2.903225806451613e-06,
"loss": 0.9617,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.4254009101272227,
"learning_rate": 3.225806451612903e-06,
"loss": 0.9431,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.1880521969068547,
"learning_rate": 3.548387096774194e-06,
"loss": 0.9449,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 4.063164822563402,
"learning_rate": 3.870967741935484e-06,
"loss": 0.9421,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 4.658225214533937,
"learning_rate": 4.193548387096774e-06,
"loss": 0.955,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 4.322598321133886,
"learning_rate": 4.516129032258065e-06,
"loss": 0.9272,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 4.144388209571823,
"learning_rate": 4.838709677419355e-06,
"loss": 0.9332,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.8733765885213316,
"learning_rate": 5.161290322580646e-06,
"loss": 0.8486,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.9488988802923854,
"learning_rate": 5.483870967741935e-06,
"loss": 0.8633,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.623985549570228,
"learning_rate": 5.806451612903226e-06,
"loss": 0.8413,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 2.3661079208530293,
"learning_rate": 6.129032258064517e-06,
"loss": 0.8686,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.8841947449589393,
"learning_rate": 6.451612903225806e-06,
"loss": 0.8382,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.7141189440029052,
"learning_rate": 6.774193548387097e-06,
"loss": 0.831,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.7634599258671566,
"learning_rate": 7.096774193548388e-06,
"loss": 0.8401,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.7506391117802198,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.7619,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.8471574407131688,
"learning_rate": 7.741935483870968e-06,
"loss": 0.799,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.4244594662437151,
"learning_rate": 8.064516129032258e-06,
"loss": 0.7483,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.289887331856195,
"learning_rate": 8.387096774193549e-06,
"loss": 0.7825,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 1.1928050984809988,
"learning_rate": 8.70967741935484e-06,
"loss": 0.7846,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 1.164334915422566,
"learning_rate": 9.03225806451613e-06,
"loss": 0.7759,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 1.0935651542208724,
"learning_rate": 9.35483870967742e-06,
"loss": 0.7617,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.913308025564479,
"learning_rate": 9.67741935483871e-06,
"loss": 0.7535,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.8436513005317946,
"learning_rate": 1e-05,
"loss": 0.7526,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.8903612921186901,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.7548,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.8510450708583713,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.7146,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.7856818783434437,
"learning_rate": 1.096774193548387e-05,
"loss": 0.7478,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.8082559344415756,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.7235,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.7703197356716803,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.7033,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.7937745839252959,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.7166,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.6271256983125751,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.7193,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.8552861497488623,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.7455,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.631965806484086,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.7093,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.629741349746896,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.6907,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.7351936190224178,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.7098,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.7273656726091237,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.7043,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.6333227243741322,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.6993,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.6570254625210553,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.703,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.7352373894966219,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.6893,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.7017962219938786,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.7008,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.7899297829397719,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.7094,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.712273061766724,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.7153,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.7199592840331306,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.6953,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.6864771643303407,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.6847,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.7919878374351934,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.664,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.6944647815402618,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.6727,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.6828500263065087,
"learning_rate": 1.741935483870968e-05,
"loss": 0.6875,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.995710478943789,
"learning_rate": 1.774193548387097e-05,
"loss": 0.6781,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.8216734280993048,
"learning_rate": 1.806451612903226e-05,
"loss": 0.6982,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.7198925432060852,
"learning_rate": 1.838709677419355e-05,
"loss": 0.6877,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.8284632423459102,
"learning_rate": 1.870967741935484e-05,
"loss": 0.6662,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.6622767394201778,
"learning_rate": 1.903225806451613e-05,
"loss": 0.6974,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.7037212774605385,
"learning_rate": 1.935483870967742e-05,
"loss": 0.6941,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.7472323159443776,
"learning_rate": 1.967741935483871e-05,
"loss": 0.6882,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.5927456228714584,
"learning_rate": 2e-05,
"loss": 0.6681,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.7424889060346633,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.6701,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.7072003793451687,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.7227,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.6572138173673795,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.6606,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.7660980490108259,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.6692,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.6691853951007285,
"learning_rate": 2.161290322580645e-05,
"loss": 0.658,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.8612680205922728,
"learning_rate": 2.193548387096774e-05,
"loss": 0.6838,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.8869360330877609,
"learning_rate": 2.225806451612903e-05,
"loss": 0.6914,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.7566225206051927,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.6611,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.7760868006388096,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.6803,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.7433872052436679,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.6701,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.7049673540560645,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.6521,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.9507418588268729,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.675,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.6662069127121751,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.6676,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.953635820829178,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.6903,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 1.0238813855691387,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.6762,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.8001934837232657,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.6564,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.2417497718036872,
"learning_rate": 2.548387096774194e-05,
"loss": 0.6544,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.9262619221629236,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.663,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.1253389211309575,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.668,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 0.7925357240897694,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.6431,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.9015972075391829,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.6417,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.8734209422651077,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.6767,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.9352260017982886,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.6623,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 0.876638587250648,
"learning_rate": 2.774193548387097e-05,
"loss": 0.6278,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 0.9994325332999316,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.6627,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.0505722167006275,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.6774,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.1022666784182678,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.6247,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 0.8288879500712274,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.6369,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.8942386336449706,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.6629,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 0.8187573882331101,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.6659,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 0.8291041817923095,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.6737,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 0.9748346641886536,
"learning_rate": 3.032258064516129e-05,
"loss": 0.6553,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.8519285673204215,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.6653,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.8499221239430167,
"learning_rate": 3.096774193548387e-05,
"loss": 0.686,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.1426509909614588,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.658,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 0.6955283356583173,
"learning_rate": 3.161290322580645e-05,
"loss": 0.6328,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.8449466741271808,
"learning_rate": 3.193548387096774e-05,
"loss": 0.6601,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 0.7546032093541805,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.6779,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 0.7926504663123204,
"learning_rate": 3.258064516129033e-05,
"loss": 0.6564,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.868209076315329,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.6425,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.8465856348376247,
"learning_rate": 3.322580645161291e-05,
"loss": 0.6556,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.860866568264827,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.6763,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 0.7238089632120781,
"learning_rate": 3.387096774193549e-05,
"loss": 0.6393,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.7836597217511073,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.6621,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 0.6844664089079812,
"learning_rate": 3.451612903225807e-05,
"loss": 0.6375,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.9217104145494108,
"learning_rate": 3.483870967741936e-05,
"loss": 0.6719,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 0.9185010907089493,
"learning_rate": 3.516129032258065e-05,
"loss": 0.6621,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 0.8902511912110184,
"learning_rate": 3.548387096774194e-05,
"loss": 0.6129,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 0.7872809217171998,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.6628,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 0.8514960650350535,
"learning_rate": 3.612903225806452e-05,
"loss": 0.6556,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.6116724843497204,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.6563,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.9235347806724912,
"learning_rate": 3.67741935483871e-05,
"loss": 0.6402,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 0.8111398948587443,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.6668,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 0.8121362181406842,
"learning_rate": 3.741935483870968e-05,
"loss": 0.6548,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.7177325960187582,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.6661,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 0.708338181559848,
"learning_rate": 3.806451612903226e-05,
"loss": 0.6351,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.7225992493681584,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.6571,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.7618974528611288,
"learning_rate": 3.870967741935484e-05,
"loss": 0.6253,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 0.7240585040372929,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.649,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 0.6969542933435044,
"learning_rate": 3.935483870967742e-05,
"loss": 0.6425,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 0.7335652466878205,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.6417,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 0.7130452931108866,
"learning_rate": 4e-05,
"loss": 0.6579,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 0.7556937346646709,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.652,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 0.6831184501122739,
"learning_rate": 3.999968016170525e-05,
"loss": 0.648,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 0.8745119319494694,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.625,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.7112357223592055,
"learning_rate": 3.999872065705065e-05,
"loss": 0.6272,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.8304373795998625,
"learning_rate": 3.999800103862945e-05,
"loss": 0.6128,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.7783884849258975,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.6188,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.7095151194354035,
"learning_rate": 3.999608209836944e-05,
"loss": 0.6277,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 0.8365084538448017,
"learning_rate": 3.999488279187443e-05,
"loss": 0.6552,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.8675589333789743,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.6546,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 0.7511022631302737,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.6516,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.6888200773225338,
"learning_rate": 3.999032564583976e-05,
"loss": 0.6294,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 0.7384812291683697,
"learning_rate": 3.998848689546591e-05,
"loss": 0.6329,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.8816249987938365,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.6436,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.0151596360006159,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.6176,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.8470073546705934,
"learning_rate": 3.998201174509306e-05,
"loss": 0.6161,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 1.2704949614216232,
"learning_rate": 3.997953378607961e-05,
"loss": 0.6536,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.8097519288749706,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.6464,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.909550074938196,
"learning_rate": 3.997409862168454e-05,
"loss": 0.6529,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.8185191827158622,
"learning_rate": 3.997114145976237e-05,
"loss": 0.6357,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.8846214461356153,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.6528,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.6587234858067798,
"learning_rate": 3.996474809469545e-05,
"loss": 0.6258,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.7363833641047713,
"learning_rate": 3.996131194267189e-05,
"loss": 0.6579,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.6436745096523887,
"learning_rate": 3.995771618052995e-05,
"loss": 0.6415,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.7108249519922472,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.6332,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.6339201132554361,
"learning_rate": 3.995004594217343e-05,
"loss": 0.6507,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.6904122051962935,
"learning_rate": 3.994597152728987e-05,
"loss": 0.6255,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.6984188839836145,
"learning_rate": 3.994173762494947e-05,
"loss": 0.6281,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.7311464965578766,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.619,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.715879199592676,
"learning_rate": 3.993279149458988e-05,
"loss": 0.6143,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.6777882961425501,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.6178,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.6138983146495601,
"learning_rate": 3.992320783722616e-05,
"loss": 0.6221,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.6313561333585328,
"learning_rate": 3.991817703090965e-05,
"loss": 0.6368,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.7117711208651638,
"learning_rate": 3.991298695938038e-05,
"loss": 0.603,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.5984119547603572,
"learning_rate": 3.9907637664138e-05,
"loss": 0.6542,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.8570654077261398,
"learning_rate": 3.990212918795534e-05,
"loss": 0.6144,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.5801730485851938,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.6209,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.7411782851911445,
"learning_rate": 3.989063487022416e-05,
"loss": 0.575,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.5759382839664651,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.6111,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.6064697288707733,
"learning_rate": 3.987850437381915e-05,
"loss": 0.6439,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.7026804115950247,
"learning_rate": 3.987220067906313e-05,
"loss": 0.6058,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.5364333275096212,
"learning_rate": 3.986573808672002e-05,
"loss": 0.6052,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.6391006970222157,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.6286,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.5403981895192761,
"learning_rate": 3.985233641724152e-05,
"loss": 0.6139,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.7346310655128702,
"learning_rate": 3.984539744726554e-05,
"loss": 0.628,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.6136033875021776,
"learning_rate": 3.983829979402038e-05,
"loss": 0.6207,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.6292208975151847,
"learning_rate": 3.983104351425871e-05,
"loss": 0.6426,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.6246251329675516,
"learning_rate": 3.982362866600154e-05,
"loss": 0.6262,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.6117764247792865,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.6055,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.6643571387891405,
"learning_rate": 3.980832350242388e-05,
"loss": 0.6336,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.6670073876125131,
"learning_rate": 3.980043330948306e-05,
"loss": 0.6393,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.5922662613417837,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.6095,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.6474911004366046,
"learning_rate": 3.978417801674579e-05,
"loss": 0.6269,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.6207824982121042,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.6135,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.5939453258061225,
"learning_rate": 3.976728995023254e-05,
"loss": 0.635,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.7010330035748025,
"learning_rate": 3.975860879481514e-05,
"loss": 0.6184,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.5585513383909848,
"learning_rate": 3.974976965008833e-05,
"loss": 0.6239,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.7226735633691221,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.6084,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.6732850675465403,
"learning_rate": 3.973161767667946e-05,
"loss": 0.6226,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.6398759420403359,
"learning_rate": 3.972230499314009e-05,
"loss": 0.6174,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.6043184097564525,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.6036,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.6723933033665365,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.613,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.6313976904868301,
"learning_rate": 3.969342105253098e-05,
"loss": 0.6111,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.7729315552213893,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.6157,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.6279419302479686,
"learning_rate": 3.967337762346568e-05,
"loss": 0.6216,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.6801394756659801,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.638,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.6959374002861063,
"learning_rate": 3.965270496444528e-05,
"loss": 0.5934,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.6328314777204253,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.6223,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.6630772470635696,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.624,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.6282176319770415,
"learning_rate": 3.962051762160151e-05,
"loss": 0.6051,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.5506229266585054,
"learning_rate": 3.960947462140638e-05,
"loss": 0.5881,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.5683902707879557,
"learning_rate": 3.959827482437472e-05,
"loss": 0.6247,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.6190861603592975,
"learning_rate": 3.958691832005981e-05,
"loss": 0.6162,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.5471655941417504,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.5939,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.6706357881327978,
"learning_rate": 3.956373555405774e-05,
"loss": 0.6049,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.6228778648830049,
"learning_rate": 3.955190947773937e-05,
"loss": 0.6083,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.6950555132227567,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.5859,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.6010133812587987,
"learning_rate": 3.952778841127214e-05,
"loss": 0.6115,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.6278754433095889,
"learning_rate": 3.95154936139947e-05,
"loss": 0.6104,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.6163818425995513,
"learning_rate": 3.950304277135034e-05,
"loss": 0.6233,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.6081704308990832,
"learning_rate": 3.94904359828957e-05,
"loss": 0.6143,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.6365402887318071,
"learning_rate": 3.94776733494343e-05,
"loss": 0.5985,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.7231457131069157,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.6218,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.6032984248886468,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.6126,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.7608431519047664,
"learning_rate": 3.943845140573213e-05,
"loss": 0.6081,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.7164188756434797,
"learning_rate": 3.942506642518952e-05,
"loss": 0.6111,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.7775915026674127,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.6049,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.770933671336052,
"learning_rate": 3.939783060543167e-05,
"loss": 0.6117,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.5356996726724111,
"learning_rate": 3.938397998399333e-05,
"loss": 0.5997,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.5963786658718667,
"learning_rate": 3.936997436876756e-05,
"loss": 0.6244,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.5318862823872476,
"learning_rate": 3.935581387174289e-05,
"loss": 0.5935,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.5437048432221436,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.6084,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.5157541191546005,
"learning_rate": 3.932702868644224e-05,
"loss": 0.5726,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.559132005416482,
"learning_rate": 3.931240422833181e-05,
"loss": 0.578,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.5270105357529334,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.6066,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.5319604375919418,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.6178,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.592361785956154,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.6224,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.572173618570945,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.6295,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.635460555114343,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.6022,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.5441310535280796,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.6041,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.6270270595734365,
"learning_rate": 3.92057159243286e-05,
"loss": 0.607,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.5484982080847627,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.595,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.6898094869667206,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.5931,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.7698323273914051,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.6064,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.6246540403619371,
"learning_rate": 3.914136997982874e-05,
"loss": 0.6058,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.7737064950433888,
"learning_rate": 3.912490053436079e-05,
"loss": 0.6062,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.6357299096337816,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.6299,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.8117674609790301,
"learning_rate": 3.909150300975178e-05,
"loss": 0.5947,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.5548643988269629,
"learning_rate": 3.907457519765642e-05,
"loss": 0.5895,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.6852931787615852,
"learning_rate": 3.905749486576607e-05,
"loss": 0.6071,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.5153024583488073,
"learning_rate": 3.90402621506546e-05,
"loss": 0.6015,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.5710208308664567,
"learning_rate": 3.902287719011435e-05,
"loss": 0.6214,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.464789329698853,
"learning_rate": 3.900534012315499e-05,
"loss": 0.6032,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.5611355542171742,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.5901,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.4686375792814413,
"learning_rate": 3.896981023209777e-05,
"loss": 0.575,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.5433236580124005,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.5902,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.5207076015195774,
"learning_rate": 3.893367361386491e-05,
"loss": 0.5898,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.5282681466405796,
"learning_rate": 3.891537814248417e-05,
"loss": 0.5841,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.5748888622846863,
"learning_rate": 3.889693142424384e-05,
"loss": 0.6025,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.5098383234363987,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.5661,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.5091368846385511,
"learning_rate": 3.885958483839049e-05,
"loss": 0.6075,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.5742717608081999,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.6024,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.5468314855175579,
"learning_rate": 3.882163505079171e-05,
"loss": 0.5844,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.6165135145925439,
"learning_rate": 3.880243433489132e-05,
"loss": 0.5801,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.6933479035449523,
"learning_rate": 3.878308327522702e-05,
"loss": 0.4746,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.5913108815976487,
"learning_rate": 3.876358202652936e-05,
"loss": 0.473,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.6973979504997776,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.4925,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.8074990900280271,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.4868,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.6505083549317171,
"learning_rate": 3.870417871154801e-05,
"loss": 0.4763,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.6445394345656741,
"learning_rate": 3.868407827802192e-05,
"loss": 0.4797,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.6733854534625568,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.4864,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.6644111824574493,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.4725,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.5781819358886734,
"learning_rate": 3.862288124195319e-05,
"loss": 0.4933,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.6867986830315623,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.4918,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.5727223551510962,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.4569,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.6165757348105646,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.4808,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.6452295365773665,
"learning_rate": 3.853920126717703e-05,
"loss": 0.4941,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.5935192066134087,
"learning_rate": 3.851791025492574e-05,
"loss": 0.4908,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.7781053744525859,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.481,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.5909753690192919,
"learning_rate": 3.847488419569866e-05,
"loss": 0.4718,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.607345863295295,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.4824,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.5644627160528835,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.4693,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.6239097840516217,
"learning_rate": 3.840923760917199e-05,
"loss": 0.4815,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.6076539154153098,
"learning_rate": 3.838706077964473e-05,
"loss": 0.4727,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.6055641137000414,
"learning_rate": 3.836473692766944e-05,
"loss": 0.486,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.5720276031093565,
"learning_rate": 3.834226623174705e-05,
"loss": 0.4726,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.5631065679655113,
"learning_rate": 3.831964887155264e-05,
"loss": 0.506,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.6692209479183602,
"learning_rate": 3.829688502793403e-05,
"loss": 0.4586,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.7123563927163225,
"learning_rate": 3.827397488291029e-05,
"loss": 0.4897,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.5537710871940449,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.4644,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.6290187187071296,
"learning_rate": 3.822771642257145e-05,
"loss": 0.4856,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.6049833020612471,
"learning_rate": 3.820436847713776e-05,
"loss": 0.4709,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.5392180679220814,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.4839,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.6872897157900923,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.4752,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.539390136322504,
"learning_rate": 3.813345202354145e-05,
"loss": 0.4716,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.581689940970197,
"learning_rate": 3.810952296329563e-05,
"loss": 0.491,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.6017400157671016,
"learning_rate": 3.808544909978676e-05,
"loss": 0.4625,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.5405988506400956,
"learning_rate": 3.806123062550882e-05,
"loss": 0.4828,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.5592724282447129,
"learning_rate": 3.803686773411208e-05,
"loss": 0.4856,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.5320128848163298,
"learning_rate": 3.801236062040158e-05,
"loss": 0.4818,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.6186954413690161,
"learning_rate": 3.798770948033553e-05,
"loss": 0.4991,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.6827230897797941,
"learning_rate": 3.796291451102381e-05,
"loss": 0.4772,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.5048204243492119,
"learning_rate": 3.793797591072631e-05,
"loss": 0.4586,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.6405349353529254,
"learning_rate": 3.791289387885144e-05,
"loss": 0.4948,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.6732613251951862,
"learning_rate": 3.788766861595443e-05,
"loss": 0.4828,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.6141751758614831,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.4861,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.767767712576583,
"learning_rate": 3.783678920503983e-05,
"loss": 0.4657,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.54219879093076,
"learning_rate": 3.781113546385265e-05,
"loss": 0.4799,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.6866074714627114,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.4955,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.5452723118262983,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.4677,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.5624670373661709,
"learning_rate": 3.77333205623025e-05,
"loss": 0.4802,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.5753081010395208,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.4831,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.6005609479713715,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.4751,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.5829508539511747,
"learning_rate": 3.765422951112704e-05,
"loss": 0.4782,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.645534330272284,
"learning_rate": 3.762758321970216e-05,
"loss": 0.4804,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.6121831021665196,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.4973,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.5317980085217004,
"learning_rate": 3.757386800198538e-05,
"loss": 0.4866,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.6459598850880114,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.4687,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.48294517943038207,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.4853,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.6084584646698084,
"learning_rate": 3.749224181796305e-05,
"loss": 0.5065,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.5622211200581978,
"learning_rate": 3.746475306375545e-05,
"loss": 0.4962,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.559728005192945,
"learning_rate": 3.743712466184774e-05,
"loss": 0.4769,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.5612486022586707,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.4841,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.514972062989895,
"learning_rate": 3.73814497997107e-05,
"loss": 0.4776,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.5368385981879866,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.4702,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.5059980322740075,
"learning_rate": 3.732521901224724e-05,
"loss": 0.5004,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.5840227234510333,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.4802,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.500348613385171,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.5017,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.5707971977102372,
"learning_rate": 3.723983441007882e-05,
"loss": 0.4628,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.5529349444141299,
"learning_rate": 3.721109687296781e-05,
"loss": 0.4753,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.6321808525066425,
"learning_rate": 3.718222171638484e-05,
"loss": 0.4892,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.5733018657459384,
"learning_rate": 3.715320917121488e-05,
"loss": 0.4776,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.5397677470497598,
"learning_rate": 3.712405946944149e-05,
"loss": 0.4854,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.5858876564166573,
"learning_rate": 3.709477284414488e-05,
"loss": 0.4843,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.47550454141786536,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.4642,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.5752377851591357,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.4882,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.4851445896045377,
"learning_rate": 3.70060937743295e-05,
"loss": 0.4763,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.5835262074235729,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.4666,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.5033103121692726,
"learning_rate": 3.694629409915555e-05,
"loss": 0.4965,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.5751099576360243,
"learning_rate": 3.691619088858401e-05,
"loss": 0.5011,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.5390437938203366,
"learning_rate": 3.68859524166009e-05,
"loss": 0.4841,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.46243585429252293,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.4805,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.5383793757009917,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.4647,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.48821808907044983,
"learning_rate": 3.67944278554384e-05,
"loss": 0.505,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.4991786377218267,
"learning_rate": 3.676365076645557e-05,
"loss": 0.4737,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.5028491082067206,
"learning_rate": 3.673273963576791e-05,
"loss": 0.4911,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.5572131223316068,
"learning_rate": 3.670169471054001e-05,
"loss": 0.4665,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.51997182411859,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.497,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.5828787869700721,
"learning_rate": 3.663920447046888e-05,
"loss": 0.4937,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.5289839592116031,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.4872,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.50692359261088,
"learning_rate": 3.657618204491937e-05,
"loss": 0.5032,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.6593522561914353,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.4945,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.44480573773616033,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.4737,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.7699655160509601,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.4891,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.4226249311962829,
"learning_rate": 3.644854871713611e-05,
"loss": 0.4792,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.6675630361080385,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.4661,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.46237700865311926,
"learning_rate": 3.638394189710493e-05,
"loss": 0.4782,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.6388113526752635,
"learning_rate": 3.635144184933412e-05,
"loss": 0.4867,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5348106824709229,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.4897,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.5093694367791612,
"learning_rate": 3.628604977762752e-05,
"loss": 0.504,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.5568496793111625,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.4724,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.5446382688042629,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.5056,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.536395132338955,
"learning_rate": 3.618698565901727e-05,
"loss": 0.4772,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.6963893068720297,
"learning_rate": 3.615370507164654e-05,
"loss": 0.4964,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.46137528765957236,
"learning_rate": 3.61202953196805e-05,
"loss": 0.4717,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.5526230042496636,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.4537,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.45402305696898165,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.4983,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.5757964376361537,
"learning_rate": 3.601929375279662e-05,
"loss": 0.5034,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.4421506044692842,
"learning_rate": 3.598537002418012e-05,
"loss": 0.4699,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.5207461790334952,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.4741,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.5013670742729585,
"learning_rate": 3.591713938344416e-05,
"loss": 0.4871,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.5850134658459442,
"learning_rate": 3.588283301689507e-05,
"loss": 0.4609,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.5200116932042812,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.4894,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.5385477813435035,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.4755,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.5538896646851328,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.4976,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.5317810862122688,
"learning_rate": 3.574434032194994e-05,
"loss": 0.4612,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.6392153353919,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.4884,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.48585190713538,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.4776,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.5303045035592099,
"learning_rate": 3.563914797637216e-05,
"loss": 0.4754,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.5156626960423407,
"learning_rate": 3.560383338592649e-05,
"loss": 0.4852,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.602184571124935,
"learning_rate": 3.556839402764487e-05,
"loss": 0.4779,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.5233011855231123,
"learning_rate": 3.553283018489946e-05,
"loss": 0.4759,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.5426034678603063,
"learning_rate": 3.54971421420578e-05,
"loss": 0.4916,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.4570055226640213,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.456,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.5332014622171376,
"learning_rate": 3.54253945985191e-05,
"loss": 0.5025,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.5137422196466015,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.4586,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.48575860402710896,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.4955,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.5100305622329596,
"learning_rate": 3.531684894865866e-05,
"loss": 0.4907,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.5126295756102172,
"learning_rate": 3.528042173241128e-05,
"loss": 0.497,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.4840035930205478,
"learning_rate": 3.524387233431892e-05,
"loss": 0.49,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.4836449509950142,
"learning_rate": 3.520720104662958e-05,
"loss": 0.4937,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.4976988206160749,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.4643,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.47368089957984183,
"learning_rate": 3.513349397632281e-05,
"loss": 0.4738,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.4595802546615118,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.4701,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.5308032163416854,
"learning_rate": 3.505930287892533e-05,
"loss": 0.4827,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.5045408371877175,
"learning_rate": 3.502202656100099e-05,
"loss": 0.4884,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.4788881228196615,
"learning_rate": 3.498463012735256e-05,
"loss": 0.4787,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.4803941169797675,
"learning_rate": 3.494711387700094e-05,
"loss": 0.4507,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.4629989498524079,
"learning_rate": 3.490947810992505e-05,
"loss": 0.4776,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.5227276216731517,
"learning_rate": 3.48717231270595e-05,
"loss": 0.4957,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.5158633024485508,
"learning_rate": 3.483384923029211e-05,
"loss": 0.4806,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.45834604043508753,
"learning_rate": 3.479585672246156e-05,
"loss": 0.4779,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.5060547750424707,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.4833,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.44226355371070236,
"learning_rate": 3.471951708970527e-05,
"loss": 0.4764,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.5273764091124008,
"learning_rate": 3.468117057518921e-05,
"loss": 0.4668,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.4450786593343081,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.4828,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.5055518717735205,
"learning_rate": 3.460412568296731e-05,
"loss": 0.4694,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.534623917743588,
"learning_rate": 3.456542792131039e-05,
"loss": 0.4825,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.48799902585738414,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.4902,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.47556024212911213,
"learning_rate": 3.448768331403348e-05,
"loss": 0.4798,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.5665218120613021,
"learning_rate": 3.44486370900573e-05,
"loss": 0.4694,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.471224290796191,
"learning_rate": 3.440947533516396e-05,
"loss": 0.4824,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.6033790309950003,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.4945,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.46816234818012825,
"learning_rate": 3.43308064860925e-05,
"loss": 0.4664,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.5110188764485555,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.472,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.5042019936706125,
"learning_rate": 3.425167928295015e-05,
"loss": 0.4951,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.4673875438901899,
"learning_rate": 3.421194458890411e-05,
"loss": 0.4817,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.49155856771230944,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.467,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.4677186922615347,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.4642,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.47327252795979474,
"learning_rate": 3.409205995219561e-05,
"loss": 0.473,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.517735433245937,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.4777,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.4772587513885688,
"learning_rate": 3.40115729298209e-05,
"loss": 0.4975,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.5625493341873776,
"learning_rate": 3.397116120327025e-05,
"loss": 0.4921,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.45685801907580426,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.4851,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.5823967792854199,
"learning_rate": 3.389000293509524e-05,
"loss": 0.476,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.47089001326570473,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.4761,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.4884168316963993,
"learning_rate": 3.380840041143494e-05,
"loss": 0.4759,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.5225627016376682,
"learning_rate": 3.376743336885792e-05,
"loss": 0.4675,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.5224652517555549,
"learning_rate": 3.372635624225055e-05,
"loss": 0.4667,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.48412374915929585,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.4745,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.4749717718136679,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.4879,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.4778607411077151,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.459,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.43704378019395573,
"learning_rate": 3.356095347769799e-05,
"loss": 0.4729,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.4807153289365822,
"learning_rate": 3.351933087522552e-05,
"loss": 0.4646,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.471565964767912,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.4806,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.509934744867631,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.5022,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.48534100760439414,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.4877,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.5411387021590808,
"learning_rate": 3.335176280432308e-05,
"loss": 0.5038,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.4934211918757256,
"learning_rate": 3.330960304618295e-05,
"loss": 0.4753,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.549098814775199,
"learning_rate": 3.326733686481153e-05,
"loss": 0.4777,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.548730173521942,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.4824,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.4591748959052468,
"learning_rate": 3.318248658506012e-05,
"loss": 0.4799,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.462702884524921,
"learning_rate": 3.313990316514069e-05,
"loss": 0.4535,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.4660678464095837,
"learning_rate": 3.309721467890571e-05,
"loss": 0.4724,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.5481403275852751,
"learning_rate": 3.305442146769115e-05,
"loss": 0.4699,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.4882850531662466,
"learning_rate": 3.30115238736704e-05,
"loss": 0.4661,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.49365691101963194,
"learning_rate": 3.296852223985148e-05,
"loss": 0.4506,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.4940645387223427,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.4766,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.4573549359971478,
"learning_rate": 3.288220822900796e-05,
"loss": 0.4569,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.5237459946431507,
"learning_rate": 3.283889654214788e-05,
"loss": 0.471,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.4366436518963983,
"learning_rate": 3.279548219581317e-05,
"loss": 0.4669,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.5115564666989367,
"learning_rate": 3.275196553714379e-05,
"loss": 0.4925,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.4546363260090527,
"learning_rate": 3.270834691409778e-05,
"loss": 0.4795,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.4727866989994553,
"learning_rate": 3.266462667544849e-05,
"loss": 0.4477,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.4230252918783196,
"learning_rate": 3.262080517078178e-05,
"loss": 0.4813,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.4477484680076822,
"learning_rate": 3.257688275049323e-05,
"loss": 0.4772,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.4510476035378918,
"learning_rate": 3.253285976578535e-05,
"loss": 0.4668,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.4628979136719135,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.4606,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.49313284811214336,
"learning_rate": 3.244451351193935e-05,
"loss": 0.4848,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.46901708145100984,
"learning_rate": 3.240019094921551e-05,
"loss": 0.4779,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.53129997632846,
"learning_rate": 3.235576923489527e-05,
"loss": 0.4804,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.4647006719856076,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.466,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.5416475730349204,
"learning_rate": 3.226662977303494e-05,
"loss": 0.4723,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.4835444753133873,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.4837,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.5391237698821085,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.4611,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.49125785123554344,
"learning_rate": 3.21321858487569e-05,
"loss": 0.4853,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.5267429546960046,
"learning_rate": 3.208717671149931e-05,
"loss": 0.4498,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.5101692179354296,
"learning_rate": 3.204207092549882e-05,
"loss": 0.4549,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.5404524899059504,
"learning_rate": 3.19968688514201e-05,
"loss": 0.4712,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.48797927427017496,
"learning_rate": 3.195157085069774e-05,
"loss": 0.4847,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.5181093262207814,
"learning_rate": 3.190617728553332e-05,
"loss": 0.4869,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.4764937893355212,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.4646,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.46641975784058975,
"learning_rate": 3.181510491450254e-05,
"loss": 0.4788,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.47317490350050995,
"learning_rate": 3.176942683684842e-05,
"loss": 0.4882,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.5390162291497009,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.4649,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.49190003796700155,
"learning_rate": 3.167778872346327e-05,
"loss": 0.4703,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.5381568870092988,
"learning_rate": 3.163182942046815e-05,
"loss": 0.4951,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.4530390601359916,
"learning_rate": 3.158577710967494e-05,
"loss": 0.4821,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.4355637928231159,
"learning_rate": 3.153963215931669e-05,
"loss": 0.4693,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.44560397978352206,
"learning_rate": 3.14933949383672e-05,
"loss": 0.4795,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.44463900171277954,
"learning_rate": 3.144706581653805e-05,
"loss": 0.4755,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.46307979110037517,
"learning_rate": 3.140064516427566e-05,
"loss": 0.4694,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.4742046091233585,
"learning_rate": 3.135413335275833e-05,
"loss": 0.4737,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.4915413800809012,
"learning_rate": 3.130753075389327e-05,
"loss": 0.4691,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.4910371827398374,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.4742,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.5070483024617625,
"learning_rate": 3.121405468537547e-05,
"loss": 0.4611,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.4989689523183768,
"learning_rate": 3.116718196315488e-05,
"loss": 0.4564,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.48263756907380007,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.4665,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.5369593659228631,
"learning_rate": 3.107316901675252e-05,
"loss": 0.4946,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.4446998993227199,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.4762,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.4292226474542245,
"learning_rate": 3.097880190800057e-05,
"loss": 0.4734,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.41095056095180116,
"learning_rate": 3.093148648549788e-05,
"loss": 0.4587,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.4702568335417841,
"learning_rate": 3.088408365512055e-05,
"loss": 0.4841,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.47811330124524853,
"learning_rate": 3.083659379590034e-05,
"loss": 0.4707,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.448109519769707,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.4763,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.4611758122526298,
"learning_rate": 3.074135451053475e-05,
"loss": 0.4811,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.4871041865397318,
"learning_rate": 3.069360584592015e-05,
"loss": 0.4878,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.47219326409218954,
"learning_rate": 3.064577167551816e-05,
"loss": 0.4665,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.4748991052903682,
"learning_rate": 3.059785238180953e-05,
"loss": 0.4796,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.43439669977228307,
"learning_rate": 3.054984834795566e-05,
"loss": 0.4977,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.500960283215057,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.4745,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.4301592062660473,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.4866,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.4984974614725582,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.4872,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.43496255114660903,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.4733,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.4088839888840378,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.4796,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.48256834018857175,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.4727,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.4697377467933103,
"learning_rate": 3.021147971398632e-05,
"loss": 0.4535,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.4840334442540311,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.4821,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.4959682049309122,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.4788,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.4391527510123386,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.4601,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.5477297945992382,
"learning_rate": 3.001632138223332e-05,
"loss": 0.4846,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.7297618056160644,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.3421,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.5222359867620878,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.3311,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 1.3128230351225345,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.3505,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.6051168715989582,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.3272,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.5584461053360308,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.3128,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.5074894324445761,
"learning_rate": 2.972118904869376e-05,
"loss": 0.3132,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.5627395281089766,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.3345,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.4789814428584496,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.3229,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.47305195937254435,
"learning_rate": 2.95725682050846e-05,
"loss": 0.3243,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.5477231397280418,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.3289,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.4621895682117664,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.3331,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.5313504898132879,
"learning_rate": 2.942325848714505e-05,
"loss": 0.3417,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.4318114975055604,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.3245,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.43789240646568184,
"learning_rate": 2.932334127129686e-05,
"loss": 0.316,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.4507476753992844,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.329,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.4563057341024407,
"learning_rate": 2.922312585929131e-05,
"loss": 0.3089,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.48425835936956463,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.3272,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.4832987941958981,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.3292,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.4594460386562541,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.3257,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.4311521569233003,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.3238,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.44181240020038054,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.3483,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.4614382253871693,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.3169,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.423796743811679,
"learning_rate": 2.887006998351372e-05,
"loss": 0.3357,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.5109894456201817,
"learning_rate": 2.881934649605535e-05,
"loss": 0.3351,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.42179267333019504,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.3196,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.4407066971502919,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.3224,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.44466184515543455,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.3253,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.4222719928464355,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.3194,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.4657784767421243,
"learning_rate": 2.856467940121025e-05,
"loss": 0.3093,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.4538070716720573,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.3155,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.47085014725970725,
"learning_rate": 2.846233032790149e-05,
"loss": 0.3399,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.40752797951180053,
"learning_rate": 2.84110540896642e-05,
"loss": 0.3315,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.44844732065030213,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.3371,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.43238652126974436,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.3245,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.4485476752360341,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.3468,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.42278425786786505,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.3395,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.43694687340167243,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.3245,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.42908345023126254,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.3519,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.420309810053332,
"learning_rate": 2.805026032173773e-05,
"loss": 0.3122,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.48761828148881353,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.3312,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.453229558915008,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.3306,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.4775114976565495,
"learning_rate": 2.7894660620408e-05,
"loss": 0.3384,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.4434544152045328,
"learning_rate": 2.784266724883173e-05,
"loss": 0.3245,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.4420616530627332,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.3351,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.4828280656054258,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.3335,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.39572208552747956,
"learning_rate": 2.768631254100716e-05,
"loss": 0.3123,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.4553124118400245,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.3305,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.5060022384466235,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.3117,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.4861841196977863,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.3231,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.4712311270652961,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.3303,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.4899901904788088,
"learning_rate": 2.742449775134085e-05,
"loss": 0.3168,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.44601778279904103,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.3212,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.4544232361761732,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.3254,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.4621159731737694,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.3444,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.4343552269418455,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.335,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.5014845704872097,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.3404,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.4260072914368558,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.3468,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.486381202336903,
"learning_rate": 2.705547548999996e-05,
"loss": 0.3226,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.43775388566351736,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.3307,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.47376673110637446,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.3265,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.4473662159391423,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.3584,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.41221792486880043,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.3445,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.4595911363624057,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.3316,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.4060216660026106,
"learning_rate": 2.67369650552155e-05,
"loss": 0.3256,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.41512072871911543,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.3188,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.42287394067669964,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.3316,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.39467348662783264,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.3352,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.4181609108214365,
"learning_rate": 2.652354173593614e-05,
"loss": 0.3252,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.38942749438858004,
"learning_rate": 2.647005443331317e-05,
"loss": 0.3311,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.4031396633306214,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.3295,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.4218040620291566,
"learning_rate": 2.636292505301542e-05,
"loss": 0.3525,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.4137907079094771,
"learning_rate": 2.630928383194431e-05,
"loss": 0.3311,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.4205744406882296,
"learning_rate": 2.625559216200781e-05,
"loss": 0.324,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.4291845609199152,
"learning_rate": 2.620185047252308e-05,
"loss": 0.3333,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.4142972338788534,
"learning_rate": 2.614805919320723e-05,
"loss": 0.3339,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.4219133770401668,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.3202,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.420665505785128,
"learning_rate": 2.60403295859298e-05,
"loss": 0.3395,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.4187412583760787,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.3369,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.40367727388838714,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.3351,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.4282083725654605,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.3412,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.41557905156223823,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.3175,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.43918523921922786,
"learning_rate": 2.577016790136453e-05,
"loss": 0.3511,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.4081811156251241,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.3311,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.4117395889204347,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.3212,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.43308069045289893,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.3243,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.4042822621850307,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.3414,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.4072504038036139,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.337,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.4119097505209542,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.3369,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.4196844648956455,
"learning_rate": 2.539001591449889e-05,
"loss": 0.3471,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.4210314191627971,
"learning_rate": 2.53355326151989e-05,
"loss": 0.3211,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.41184840470269857,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.3227,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.41066063325934393,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.3454,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.4593212848162361,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.346,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.3822856902141647,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.335,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.4306408632487855,
"learning_rate": 2.506248490185627e-05,
"loss": 0.3383,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.4160723820227664,
"learning_rate": 2.500775217241923e-05,
"loss": 0.3303,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.4442695972353914,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.3521,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.4423281466669127,
"learning_rate": 2.4898167025948e-05,
"loss": 0.3506,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.42976473825470096,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.3278,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.4302974217731706,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.3449,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.4235639667408169,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.3214,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.41697808164400735,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.334,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.4434773242626289,
"learning_rate": 2.46235264422994e-05,
"loss": 0.3562,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.4912292137909725,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.3492,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.393852168221721,
"learning_rate": 2.451340834511592e-05,
"loss": 0.3487,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.42593343442471676,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.3344,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.42280438734901,
"learning_rate": 2.440314589184958e-05,
"loss": 0.3467,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.3688333252085214,
"learning_rate": 2.434796163357541e-05,
"loss": 0.3499,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.4236653433081111,
"learning_rate": 2.429274260911588e-05,
"loss": 0.3158,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.41304095905191873,
"learning_rate": 2.423748926000084e-05,
"loss": 0.3413,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.3993104111235559,
"learning_rate": 2.418220202803459e-05,
"loss": 0.3497,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.44235696215646286,
"learning_rate": 2.412688135529237e-05,
"loss": 0.3328,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.4015273974877139,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.3514,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.4456444296298263,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.3124,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.4574467641520772,
"learning_rate": 2.396072311715187e-05,
"loss": 0.329,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.4195068137195505,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.3377,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.5254363541175328,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.3579,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.42176953878192686,
"learning_rate": 2.379427985200223e-05,
"loss": 0.3464,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.420658857033658,
"learning_rate": 2.373873749394972e-05,
"loss": 0.3498,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.4740160548802436,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.3426,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.3978210846213618,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.3306,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.4173330307402403,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.344,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.4103925122924849,
"learning_rate": 2.351627355800233e-05,
"loss": 0.3277,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.4142182315638821,
"learning_rate": 2.346058617160256e-05,
"loss": 0.3374,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.40599059817968225,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.3524,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.4251100042950861,
"learning_rate": 2.334912883203469e-05,
"loss": 0.3253,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.4075967537349475,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.3298,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.4119687936016248,
"learning_rate": 2.323756437450136e-05,
"loss": 0.3474,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.3947979027873917,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.3159,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.4140110834068132,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.326,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.39341376970379105,
"learning_rate": 2.307002464849052e-05,
"loss": 0.326,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.42897399009103737,
"learning_rate": 2.30141283818846e-05,
"loss": 0.325,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.37925917484581123,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.3331,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.41282182407389356,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.335,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.42395812679813216,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.3336,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.43585815765707603,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.3229,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.39600722531665206,
"learning_rate": 2.273429448107257e-05,
"loss": 0.3308,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.4135929528089719,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.3329,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.39992278796177533,
"learning_rate": 2.262220474347102e-05,
"loss": 0.3416,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.4176614809166882,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.3316,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.42807346357800585,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.3334,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.4050892739391034,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.3286,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.44641591006943515,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.3474,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.4250554907738178,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.3256,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.4303704415517529,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.3313,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.4098047659227485,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.3518,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.39883627201992694,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.3424,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.37496904557240046,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.3291,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.43986488392252404,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.3454,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.36433453432115914,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.3186,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.3823132612698826,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.3485,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.4084929089974682,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.332,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.4024486558845084,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.3232,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.39616548836957044,
"learning_rate": 2.17791064266051e-05,
"loss": 0.3348,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.4047500028857613,
"learning_rate": 2.172276926861741e-05,
"loss": 0.3247,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.38023703882838406,
"learning_rate": 2.166641833541255e-05,
"loss": 0.3309,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.3765688564109845,
"learning_rate": 2.161005407757109e-05,
"loss": 0.3259,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.38689116059010376,
"learning_rate": 2.155367694578013e-05,
"loss": 0.3244,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.39883472140817555,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.3251,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.3719654686345985,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.3432,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.42451308356321277,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.3612,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.38733402517676985,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.3225,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.37742236021730796,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.3184,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.40176119925716725,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.328,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.3884092428886,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.3495,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.4344131509555857,
"learning_rate": 2.110225054402733e-05,
"loss": 0.3463,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.39456661594196085,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.3358,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.3973660573292394,
"learning_rate": 2.09892967807724e-05,
"loss": 0.3488,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.39328613149576086,
"learning_rate": 2.093280780774835e-05,
"loss": 0.3248,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.38012613518247995,
"learning_rate": 2.087631137601793e-05,
"loss": 0.333,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.39477488734511,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.3298,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.42237415004955936,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.3117,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.38377005916870816,
"learning_rate": 2.070678184630458e-05,
"loss": 0.3252,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.3959399855325972,
"learning_rate": 2.06502600977305e-05,
"loss": 0.3113,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.4001931748473555,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.3343,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.39942855599920063,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.3379,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.4503056176800424,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.3247,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.39742751161594564,
"learning_rate": 2.042412562887594e-05,
"loss": 0.3308,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.46396043672342524,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.3415,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.4608916716304003,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.3449,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.37611317945651973,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.3478,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.4626687980738363,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.3357,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.39537492810694697,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.3242,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.3931060746221335,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.3425,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.42813674544994385,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.3433,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.4175774082785973,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.3304,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.37889594460819276,
"learning_rate": 1.99151687695709e-05,
"loss": 0.3324,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.3893507288720627,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.3389,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.43427766823055935,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.3304,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.3997193907664502,
"learning_rate": 1.974551241345447e-05,
"loss": 0.3379,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.41681847673478245,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.3298,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.40992828254855784,
"learning_rate": 1.963241759676897e-05,
"loss": 0.3309,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.4132239582130733,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.3262,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.36873719056114374,
"learning_rate": 1.951933453677638e-05,
"loss": 0.328,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.4084352978607618,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.3236,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.3938171310663874,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.3372,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.38271419790142724,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.3076,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.4100780289839541,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.3291,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.3931125104707156,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.3249,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.42456251435815495,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.3239,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.39445164477316796,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.3302,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.4699299825287938,
"learning_rate": 1.906719219225165e-05,
"loss": 0.3419,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.3887238612826242,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.3348,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.38700240055690366,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.3259,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.437225685305747,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.3258,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.420803872295328,
"learning_rate": 1.884128556891707e-05,
"loss": 0.317,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.41257350079883454,
"learning_rate": 1.878483094691119e-05,
"loss": 0.3232,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.4171719059241101,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.3567,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.4175189869289011,
"learning_rate": 1.867195130360947e-05,
"loss": 0.331,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.4116084050613115,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.3364,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.43121901067319746,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.3209,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.40136436456205626,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.3357,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.3778439461948685,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.3332,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.4032057747030609,
"learning_rate": 1.838994592242891e-05,
"loss": 0.3285,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.3901634539830116,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.325,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.38687739146846245,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.3087,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.4311534399934671,
"learning_rate": 1.82208935733949e-05,
"loss": 0.3257,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.39139243106585503,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.3348,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.4130304029112717,
"learning_rate": 1.810826238483892e-05,
"loss": 0.3227,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.40892356824786924,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.3374,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.3891753212280243,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.3249,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.4106321833038679,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.3334,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.4359353710341607,
"learning_rate": 1.788318512320838e-05,
"loss": 0.3422,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.3907244112870797,
"learning_rate": 1.782695699829051e-05,
"loss": 0.3296,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.3933054363576831,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.339,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.4377314289218825,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.3403,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.4035860882581815,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.3285,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.4315983927924383,
"learning_rate": 1.76022227484387e-05,
"loss": 0.3276,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.43902219463222786,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.3414,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.4052133851153242,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.3338,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.444838510655084,
"learning_rate": 1.743387180005809e-05,
"loss": 0.3486,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.39696975201839824,
"learning_rate": 1.737779525652899e-05,
"loss": 0.3233,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.422340379038928,
"learning_rate": 1.732173968007914e-05,
"loss": 0.3028,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.41144371012549824,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.3214,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.40894836780833316,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.3285,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.40768176466547024,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.3171,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.4005481771300986,
"learning_rate": 1.709773600686015e-05,
"loss": 0.3452,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.35920358194200047,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.3234,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.4188812024355619,
"learning_rate": 1.69858716181154e-05,
"loss": 0.3332,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.3876824211142511,
"learning_rate": 1.692997535150948e-05,
"loss": 0.3341,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.3959642226900734,
"learning_rate": 1.687410363273884e-05,
"loss": 0.3362,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.39912154781971587,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.3379,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.3834823736080831,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.32,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.3864444604669746,
"learning_rate": 1.67066402299235e-05,
"loss": 0.323,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.3777035237209204,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.3289,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.3772191461768755,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.3377,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.38335462309921575,
"learning_rate": 1.653941382839745e-05,
"loss": 0.3037,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.6109948646423844,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.2162,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.4433444885079696,
"learning_rate": 1.642806717162757e-05,
"loss": 0.2194,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.41106490888351593,
"learning_rate": 1.637243646233718e-05,
"loss": 0.1991,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.7676429927675327,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.2358,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.5424993871747211,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.2204,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.38136421718019436,
"learning_rate": 1.620572014799777e-05,
"loss": 0.2203,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.5053118851187011,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.2261,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.45877065186593335,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.2194,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.4255316634315317,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.211,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.4038657781227475,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.2342,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.4425990110639577,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.2077,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.44202476852554007,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.2303,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.4104426780140304,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.2123,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.41475801893510145,
"learning_rate": 1.576251073999917e-05,
"loss": 0.2165,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.35716820564965207,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.2285,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.3845508660390427,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.2076,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.40058974221535754,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.1987,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.3882469705114244,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.2178,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.3608077289022688,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.2302,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.3851859936361538,
"learning_rate": 1.543151434154755e-05,
"loss": 0.2036,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.37910144818194347,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.2225,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.37175268430447095,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.2073,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.3620814728391472,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.2235,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.3848310260511229,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.2072,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.34910308955886976,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.2277,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.3631373035122296,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.2267,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.3594480556361167,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.2037,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.3817230529781532,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.22,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.38206357441940303,
"learning_rate": 1.493751509814374e-05,
"loss": 0.2408,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.34572130425069053,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.2249,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.370473673777069,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.231,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.34468683221251534,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.2385,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.35541818937436515,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.2214,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.33202642278595973,
"learning_rate": 1.466446738480111e-05,
"loss": 0.2067,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.343443367856164,
"learning_rate": 1.460998408550112e-05,
"loss": 0.2241,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.36375558508406336,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.1984,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.3592274867735492,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.1971,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.3370611045515599,
"learning_rate": 1.444679451900437e-05,
"loss": 0.2043,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.3462847283168971,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.2213,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.35680568928430834,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.2316,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.36681919357336074,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.2118,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.3337416256710728,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.2094,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.3565996290083235,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.2307,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.3625174661236343,
"learning_rate": 1.412162598317602e-05,
"loss": 0.2178,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.34998529229344294,
"learning_rate": 1.406759321421907e-05,
"loss": 0.2079,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.3163215245324535,
"learning_rate": 1.401360788062871e-05,
"loss": 0.2162,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.3233782964879403,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.2532,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.3225179216943266,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.2184,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.31795716532877527,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.2147,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.3416695993115089,
"learning_rate": 1.379814952747693e-05,
"loss": 0.2232,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.3352819624749138,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.2374,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.33040714883360517,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.2129,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.32935475988420226,
"learning_rate": 1.363707494698459e-05,
"loss": 0.2007,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.32777473277349495,
"learning_rate": 1.358348460369265e-05,
"loss": 0.2263,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.32218241215363724,
"learning_rate": 1.352994556668684e-05,
"loss": 0.22,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.32493331206198517,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.2342,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.3360665455428075,
"learning_rate": 1.342302312350678e-05,
"loss": 0.2078,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.3458085973596219,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.2054,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.33198562420346117,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.2169,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.3350633154638498,
"learning_rate": 1.326303494478451e-05,
"loss": 0.2288,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.34481745988269624,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.2125,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.3414040025842459,
"learning_rate": 1.315664479122898e-05,
"loss": 0.2191,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.34465842669556673,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.2176,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.34709735694242155,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.2203,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.3345513464840583,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.198,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.3405116663475839,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.2118,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.34016623427004355,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.2153,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.3359794831906306,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.2263,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.3255712651787879,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.2248,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.3287279856877731,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.2177,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.3250064597598404,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.2218,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.32275265657670016,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.2201,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.3297659565196311,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.2187,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.32489008661813784,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.2212,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.33323163878628353,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.2292,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.3269712594501887,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.2152,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.3249551432729755,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.2107,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.3408666195907296,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.2138,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.35602195116239344,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.2054,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.33424453438248125,
"learning_rate": 1.220938883250288e-05,
"loss": 0.226,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.33009208850582605,
"learning_rate": 1.215733275116827e-05,
"loss": 0.2146,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.3614007326004468,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.2302,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.3391305583847175,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.2005,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.32665597616178926,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.2265,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.3402436262037722,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.2127,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.3401140006261643,
"learning_rate": 1.189800129803139e-05,
"loss": 0.2213,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.32371050187555955,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.231,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.3290117973847371,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.2261,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.34205894785275465,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.2037,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.3493429391177775,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.2192,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.33559587562363613,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.2238,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.353982247139534,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.2137,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.3685170571891043,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.2249,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.33440387396275156,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.2127,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.3322152489610214,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.2185,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.35021732404409084,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.2075,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.3391741108714113,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.2079,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.33537049418035814,
"learning_rate": 1.128231163049258e-05,
"loss": 0.2107,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.3367316396816032,
"learning_rate": 1.123144751066259e-05,
"loss": 0.2381,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.32568271370009644,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.2278,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.3269552153054297,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.2058,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.3309677188243144,
"learning_rate": 1.107927745387114e-05,
"loss": 0.2222,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.3555674860486538,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.1835,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.3239825135272739,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.2108,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.3469974066711182,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.222,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.3388189752770768,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.2183,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.3356706751480787,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.2278,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.32962072970258527,
"learning_rate": 1.077687414070869e-05,
"loss": 0.2216,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.32132341411128285,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.2052,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.32851682777408114,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.2184,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.32407259675715067,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.2076,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.33668971563178474,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.2007,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.32711056245954234,
"learning_rate": 1.052689572753247e-05,
"loss": 0.2201,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.3338239422402676,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.206,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.33146671004554723,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.2198,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.3297417404350913,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.2047,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.33609153268897096,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.2023,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.3346193171644457,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.1892,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.3487862651682351,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.2212,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.3291018487796825,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.2195,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.32388983888928546,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.2219,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.3378603261902683,
"learning_rate": 1.008173988718005e-05,
"loss": 0.2199,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.3310187350602211,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.2167,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.3299572255864363,
"learning_rate": 9.983678617766677e-06,
"loss": 0.2129,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.32603688586759016,
"learning_rate": 9.934767922553788e-06,
"loss": 0.2344,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.3334427919572295,
"learning_rate": 9.885937708668371e-06,
"loss": 0.2169,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.32269822317761326,
"learning_rate": 9.837188366555524e-06,
"loss": 0.2305,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.33184772120086536,
"learning_rate": 9.788520286013691e-06,
"loss": 0.2068,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.3381960085255103,
"learning_rate": 9.739933856191537e-06,
"loss": 0.2333,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.32496671104457975,
"learning_rate": 9.691429465584866e-06,
"loss": 0.2123,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.3349822581126303,
"learning_rate": 9.643007502033492e-06,
"loss": 0.2182,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.3244137101938614,
"learning_rate": 9.594668352718152e-06,
"loss": 0.2162,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.344679519894668,
"learning_rate": 9.546412404157383e-06,
"loss": 0.2206,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.3236817635374516,
"learning_rate": 9.498240042204471e-06,
"loss": 0.218,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.316819475398691,
"learning_rate": 9.450151652044339e-06,
"loss": 0.1979,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.32193686419564016,
"learning_rate": 9.402147618190472e-06,
"loss": 0.1977,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.3288695330678184,
"learning_rate": 9.354228324481841e-06,
"loss": 0.2163,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.3365867522783033,
"learning_rate": 9.306394154079852e-06,
"loss": 0.2063,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.3240435405476152,
"learning_rate": 9.258645489465254e-06,
"loss": 0.2068,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.337954338228228,
"learning_rate": 9.21098271243509e-06,
"loss": 0.2118,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.3425531004841409,
"learning_rate": 9.163406204099659e-06,
"loss": 0.2095,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.32819201448687857,
"learning_rate": 9.115916344879452e-06,
"loss": 0.2144,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.3345541117450446,
"learning_rate": 9.068513514502126e-06,
"loss": 0.2214,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.33104900196816217,
"learning_rate": 9.021198091999439e-06,
"loss": 0.2177,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.33490389388902636,
"learning_rate": 8.973970455704251e-06,
"loss": 0.2021,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.3241170929796508,
"learning_rate": 8.926830983247495e-06,
"loss": 0.2091,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.31462047901243,
"learning_rate": 8.879780051555116e-06,
"loss": 0.2268,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.32274885265629283,
"learning_rate": 8.83281803684512e-06,
"loss": 0.2142,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.32978820716612234,
"learning_rate": 8.785945314624532e-06,
"loss": 0.2183,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.33260802310398263,
"learning_rate": 8.739162259686385e-06,
"loss": 0.1945,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.31737558122718945,
"learning_rate": 8.692469246106734e-06,
"loss": 0.2158,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.32147573754134245,
"learning_rate": 8.645866647241677e-06,
"loss": 0.1945,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.3475972137468911,
"learning_rate": 8.599354835724348e-06,
"loss": 0.2363,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.3220412605010933,
"learning_rate": 8.552934183461958e-06,
"loss": 0.2293,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.3330760767013158,
"learning_rate": 8.506605061632805e-06,
"loss": 0.2318,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.3350230650361425,
"learning_rate": 8.460367840683308e-06,
"loss": 0.2016,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.342446412611879,
"learning_rate": 8.414222890325065e-06,
"loss": 0.2341,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.3297955427591715,
"learning_rate": 8.368170579531855e-06,
"loss": 0.2007,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.3305520836175077,
"learning_rate": 8.322211276536733e-06,
"loss": 0.2375,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.3252159802544415,
"learning_rate": 8.27634534882907e-06,
"loss": 0.2212,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.330522022375567,
"learning_rate": 8.230573163151585e-06,
"loss": 0.2129,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.31685636303062653,
"learning_rate": 8.184895085497466e-06,
"loss": 0.2149,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.30863269479971006,
"learning_rate": 8.139311481107401e-06,
"loss": 0.2157,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.3262186658299971,
"learning_rate": 8.093822714466686e-06,
"loss": 0.2368,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.3177354945676001,
"learning_rate": 8.048429149302271e-06,
"loss": 0.199,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.3312190446822484,
"learning_rate": 8.003131148579902e-06,
"loss": 0.2122,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.3140074974585432,
"learning_rate": 7.957929074501182e-06,
"loss": 0.2166,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.32002877653949835,
"learning_rate": 7.912823288500695e-06,
"loss": 0.2059,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.335394050168754,
"learning_rate": 7.867814151243102e-06,
"loss": 0.2156,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.33427364658288755,
"learning_rate": 7.822902022620267e-06,
"loss": 0.2226,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.3321006331466735,
"learning_rate": 7.778087261748377e-06,
"loss": 0.2068,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.3408964019088908,
"learning_rate": 7.733370226965056e-06,
"loss": 0.2094,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.3185039279297754,
"learning_rate": 7.688751275826527e-06,
"loss": 0.2123,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.32249845320132287,
"learning_rate": 7.644230765104735e-06,
"loss": 0.2027,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.31282945704667486,
"learning_rate": 7.599809050784502e-06,
"loss": 0.218,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.3245753441072549,
"learning_rate": 7.555486488060661e-06,
"loss": 0.2204,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.32859988683275754,
"learning_rate": 7.511263431335249e-06,
"loss": 0.1991,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.3380163869377996,
"learning_rate": 7.467140234214656e-06,
"loss": 0.2189,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.3313592955950651,
"learning_rate": 7.423117249506777e-06,
"loss": 0.2269,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.3246626467693581,
"learning_rate": 7.379194829218231e-06,
"loss": 0.2128,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.31598895331268667,
"learning_rate": 7.335373324551518e-06,
"loss": 0.2213,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.31391845535505214,
"learning_rate": 7.291653085902224e-06,
"loss": 0.2268,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.3229584142026759,
"learning_rate": 7.248034462856213e-06,
"loss": 0.2203,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.32665298753709865,
"learning_rate": 7.20451780418683e-06,
"loss": 0.1975,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.31144939480969297,
"learning_rate": 7.161103457852128e-06,
"loss": 0.2086,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.3202952333616765,
"learning_rate": 7.117791770992046e-06,
"loss": 0.1942,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.3228000344879362,
"learning_rate": 7.074583089925688e-06,
"loss": 0.2246,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.3175516353508376,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.2193,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.3224203967133674,
"learning_rate": 6.988476126329604e-06,
"loss": 0.2171,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.32743197467263724,
"learning_rate": 6.945578532308857e-06,
"loss": 0.2106,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.3235527221741976,
"learning_rate": 6.902785321094301e-06,
"loss": 0.2168,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.3391505878656938,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.2138,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.32301288868743955,
"learning_rate": 6.817513414939892e-06,
"loss": 0.1938,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.3235151772736185,
"learning_rate": 6.775035401831924e-06,
"loss": 0.2122,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.32598219393421574,
"learning_rate": 6.732663135188473e-06,
"loss": 0.1996,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.33179316813945653,
"learning_rate": 6.690396953817055e-06,
"loss": 0.2139,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.3228078061824639,
"learning_rate": 6.648237195676927e-06,
"loss": 0.2126,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.30826026970200016,
"learning_rate": 6.606184197876395e-06,
"loss": 0.2065,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.317891410872139,
"learning_rate": 6.564238296670109e-06,
"loss": 0.2154,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.3216096880541866,
"learning_rate": 6.52239982745638e-06,
"loss": 0.2286,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.3310052650275855,
"learning_rate": 6.480669124774484e-06,
"loss": 0.2091,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.3185755026984936,
"learning_rate": 6.439046522302013e-06,
"loss": 0.2233,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.32687103920458094,
"learning_rate": 6.397532352852191e-06,
"loss": 0.2202,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.33664177532623935,
"learning_rate": 6.356126948371202e-06,
"loss": 0.196,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.3212947869494803,
"learning_rate": 6.314830639935559e-06,
"loss": 0.2186,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.32262795028183855,
"learning_rate": 6.273643757749448e-06,
"loss": 0.2108,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.32454234674819354,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.191,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.3205394992088339,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.2131,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.3358477792170376,
"learning_rate": 6.150742957589784e-06,
"loss": 0.1988,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.32458635707784395,
"learning_rate": 6.109997064904774e-06,
"loss": 0.2165,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.32206523251485125,
"learning_rate": 6.069362236313092e-06,
"loss": 0.2207,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.33730498563691436,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.2019,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.326768491362698,
"learning_rate": 5.988427070179108e-06,
"loss": 0.2194,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.32077014853696073,
"learning_rate": 5.948127379792234e-06,
"loss": 0.2185,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.32392583928770524,
"learning_rate": 5.90794004780439e-06,
"loss": 0.2005,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.3218708459382538,
"learning_rate": 5.867865395552405e-06,
"loss": 0.2152,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.3186154229914523,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.2078,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.3161173873062872,
"learning_rate": 5.788055411095892e-06,
"loss": 0.2124,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.30945824693338264,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.2113,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.31471875213979406,
"learning_rate": 5.708699979051613e-06,
"loss": 0.2206,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.31217203564491564,
"learning_rate": 5.66919351390751e-06,
"loss": 0.1917,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.30788581825544076,
"learning_rate": 5.629801637510188e-06,
"loss": 0.2104,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.31266192390454994,
"learning_rate": 5.590524664836043e-06,
"loss": 0.2425,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.3343728432123998,
"learning_rate": 5.551362909942708e-06,
"loss": 0.1976,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.3109593598317646,
"learning_rate": 5.51231668596653e-06,
"loss": 0.2204,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.3095610254301087,
"learning_rate": 5.47338630512007e-06,
"loss": 0.2084,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.29834282781674665,
"learning_rate": 5.434572078689617e-06,
"loss": 0.2268,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.3118277167545828,
"learning_rate": 5.395874317032694e-06,
"loss": 0.2059,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.30669778514759616,
"learning_rate": 5.357293329575573e-06,
"loss": 0.2252,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.3220691957892396,
"learning_rate": 5.318829424810797e-06,
"loss": 0.2244,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.30606556205342667,
"learning_rate": 5.280482910294728e-06,
"loss": 0.2088,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.31726729053521163,
"learning_rate": 5.242254092645078e-06,
"loss": 0.2142,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.3016722025055655,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.238,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.3223426964744204,
"learning_rate": 5.166150769707892e-06,
"loss": 0.208,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.30443446343174185,
"learning_rate": 5.128276872940503e-06,
"loss": 0.2143,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.31151355640517986,
"learning_rate": 5.090521890074953e-06,
"loss": 0.2114,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.3173205554829928,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.2097,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.31234284887882285,
"learning_rate": 5.015369872647442e-06,
"loss": 0.2198,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.3241622449470173,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.1975,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.31388921720683843,
"learning_rate": 4.940697121074675e-06,
"loss": 0.2112,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.3168893262231043,
"learning_rate": 4.903541216934864e-06,
"loss": 0.2209,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.3107073987041532,
"learning_rate": 4.8665060236772e-06,
"loss": 0.2106,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.3079669183279688,
"learning_rate": 4.829591837434102e-06,
"loss": 0.2045,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.3221379444703648,
"learning_rate": 4.792798953370424e-06,
"loss": 0.2156,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.3292699431706113,
"learning_rate": 4.756127665681083e-06,
"loss": 0.2052,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.3032608525016488,
"learning_rate": 4.719578267588725e-06,
"loss": 0.2168,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.2967471767205956,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.1948,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.4585442008334314,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.1774,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.36166354920484994,
"learning_rate": 4.610664328486485e-06,
"loss": 0.1395,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.2966970092521419,
"learning_rate": 4.5746054014809e-06,
"loss": 0.1693,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.2551128081167552,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.1613,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.31661406032443706,
"learning_rate": 4.502857857942204e-06,
"loss": 0.1604,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.37263390400762186,
"learning_rate": 4.467169815100549e-06,
"loss": 0.1562,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.3515247730143439,
"learning_rate": 4.431605972355137e-06,
"loss": 0.1427,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.38523900953276785,
"learning_rate": 4.396166614073516e-06,
"loss": 0.172,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.30792916609321075,
"learning_rate": 4.360852023627846e-06,
"loss": 0.1452,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.3091702788311764,
"learning_rate": 4.325662483392656e-06,
"loss": 0.1582,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.2975882135028012,
"learning_rate": 4.29059827474257e-06,
"loss": 0.1801,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.31545710444877356,
"learning_rate": 4.255659678050063e-06,
"loss": 0.146,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.3064066290705381,
"learning_rate": 4.22084697268323e-06,
"loss": 0.1776,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.27096907573385265,
"learning_rate": 4.186160437003535e-06,
"loss": 0.1409,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.28014168099058595,
"learning_rate": 4.151600348363582e-06,
"loss": 0.1565,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.2763384340306842,
"learning_rate": 4.11716698310493e-06,
"loss": 0.1572,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.2779514651640974,
"learning_rate": 4.082860616555848e-06,
"loss": 0.137,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.3022270514453975,
"learning_rate": 4.048681523029125e-06,
"loss": 0.1878,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.27238993385056226,
"learning_rate": 4.014629975819882e-06,
"loss": 0.158,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.27068150469812247,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.1563,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.2821503950064601,
"learning_rate": 3.94691060843287e-06,
"loss": 0.1522,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.27099069108653107,
"learning_rate": 3.913243329737353e-06,
"loss": 0.1563,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.27805625338263057,
"learning_rate": 3.879704680319503e-06,
"loss": 0.163,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.2696860603907764,
"learning_rate": 3.846294928353471e-06,
"loss": 0.1336,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.2657718143946861,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.1395,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.29303180855512956,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.1476,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.26397908766389194,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.1638,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.26267556966356403,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.1533,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.2688164817977203,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.1495,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.27028436353271745,
"learning_rate": 3.648558150665886e-06,
"loss": 0.1775,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.2604244437841048,
"learning_rate": 3.616058102895075e-06,
"loss": 0.1714,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.26658257855476936,
"learning_rate": 3.583689060687079e-06,
"loss": 0.1619,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.26069018587794285,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.1567,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.24845277392226373,
"learning_rate": 3.51934502719796e-06,
"loss": 0.1612,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.25552222888944803,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.173,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.2561041899536199,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.1804,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.25583390614353396,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.1559,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.2520318036897624,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.1428,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.2517707875132337,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.141,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.2581221544432937,
"learning_rate": 3.32948376099375e-06,
"loss": 0.1501,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.2560115492829098,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.163,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.2513059210380708,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.1625,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.24902121647706785,
"learning_rate": 3.236349233544438e-06,
"loss": 0.1294,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.25656267414260386,
"learning_rate": 3.205572144561604e-06,
"loss": 0.1656,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.25989989119326384,
"learning_rate": 3.174929343376374e-06,
"loss": 0.1701,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.26794038966424183,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.1621,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.25408986998618577,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.1353,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.2565254819646904,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.152,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.25084507201699807,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.1674,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.2628053548747913,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.1724,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.2535277569823339,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.1701,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.2504680068521443,
"learning_rate": 2.964210239224696e-06,
"loss": 0.1347,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.25889012092508157,
"learning_rate": 2.934650470499869e-06,
"loss": 0.137,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.25850561399202077,
"learning_rate": 2.905227155855128e-06,
"loss": 0.1547,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.33756284408971365,
"learning_rate": 2.875940530558519e-06,
"loss": 0.1509,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.2569310906354682,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.1621,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.26186993809071396,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.1684,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.25644504917891303,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.1457,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.26206555798751974,
"learning_rate": 2.760165589921184e-06,
"loss": 0.1665,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.26556374859723963,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.1613,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.2536880681396584,
"learning_rate": 2.703104292151164e-06,
"loss": 0.1563,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.2600455289638661,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.1548,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.25606779390815393,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.1508,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.26180801179560353,
"learning_rate": 2.618550200289303e-06,
"loss": 0.1485,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.2615001609893515,
"learning_rate": 2.59064316684412e-06,
"loss": 0.142,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.25811765422641475,
"learning_rate": 2.562875338152264e-06,
"loss": 0.1467,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.2645360046316017,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.1565,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.2605759758346135,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.1636,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.2530058820774413,
"learning_rate": 2.480409295328803e-06,
"loss": 0.1502,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.2560466194812595,
"learning_rate": 2.45320049480108e-06,
"loss": 0.1513,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.25741560220536774,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.157,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.26234577442207574,
"learning_rate": 2.399204021408421e-06,
"loss": 0.1616,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.26078416036645036,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.1571,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.26633637392624115,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.142,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.2664652823314423,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.1388,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.25275741291033216,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.1607,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.2646939970643851,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.1363,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.255432185168855,
"learning_rate": 2.24059906434998e-06,
"loss": 0.1391,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.26208773919977957,
"learning_rate": 2.214660694699091e-06,
"loss": 0.1569,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.25976707597577964,
"learning_rate": 2.188864536147355e-06,
"loss": 0.1589,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.26172162211661754,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.1496,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.2613290119581876,
"learning_rate": 2.137699676264171e-06,
"loss": 0.1702,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.2559184629739678,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.1445,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.263730227132631,
"learning_rate": 2.08710612114857e-06,
"loss": 0.1506,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.254377029751861,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.1409,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.26865401431761254,
"learning_rate": 2.0370854889762e-06,
"loss": 0.149,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.2572469709412178,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.1428,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.2619268430213824,
"learning_rate": 1.987639379598427e-06,
"loss": 0.158,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.25328052757833097,
"learning_rate": 1.963132265887919e-06,
"loss": 0.1675,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.2567316531346918,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.1579,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.2659083895186478,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.1408,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.25471635224970696,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.137,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.25495599351803166,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.1345,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.2620952968342191,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.1552,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.2659116066141839,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.161,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.2521439767360931,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.1711,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.2566600982524618,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.1446,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.2606335169977236,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.139,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.25305358964159563,
"learning_rate": 1.726025117089718e-06,
"loss": 0.1312,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.25214358775511786,
"learning_rate": 1.70311497206598e-06,
"loss": 0.1377,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.2578311182280444,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.158,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.2642088114486923,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.135,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.2612771984658466,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.1486,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.2574433948290398,
"learning_rate": 1.612939220355272e-06,
"loss": 0.1597,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.2549146603907145,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.1521,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.26342652549009726,
"learning_rate": 1.568732761074121e-06,
"loss": 0.1579,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.26869442430895274,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.1436,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.2550339139399324,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.1648,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.2569435608322062,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.1476,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.25576262355400237,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.1393,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.2537736444664676,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.1591,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.2686843060468495,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.1343,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.2617564767843787,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.1535,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.2587668969516281,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.1606,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.2474212070003112,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.146,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.260424976294781,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.1588,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.25409598820958273,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.1395,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.25286688324863077,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.1366,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.25693785263890623,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.1542,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.2619913847636806,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.1954,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.2486029942913162,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.1508,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.2640224447906125,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.1457,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.2503744343508262,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.1517,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.261055362072243,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.1548,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.2603394706260709,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.1324,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.2585197866667702,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.1648,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.25900006347887916,
"learning_rate": 1.140415161609505e-06,
"loss": 0.1628,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.2767971315946937,
"learning_rate": 1.121666393356604e-06,
"loss": 0.1731,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.2576919101195174,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.141,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.26925303958907326,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.1475,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.2544856571780259,
"learning_rate": 1.066326386135097e-06,
"loss": 0.1478,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.2786755973070403,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.1505,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.262398721388194,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.1498,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.2501920047977871,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.1405,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.2599957855828353,
"learning_rate": 9.9465987684501e-07,
"loss": 0.1439,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.24568861298986439,
"learning_rate": 9.771228098856534e-07,
"loss": 0.1595,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.26267705335841124,
"learning_rate": 9.597378493454013e-07,
"loss": 0.1498,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.26239133924786645,
"learning_rate": 9.425051342339353e-07,
"loss": 0.1498,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.25222465765786023,
"learning_rate": 9.254248023435864e-07,
"loss": 0.1542,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.26061362990238973,
"learning_rate": 9.084969902482288e-07,
"loss": 0.1598,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.2550441215918384,
"learning_rate": 8.917218333021993e-07,
"loss": 0.1221,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.26556766384172487,
"learning_rate": 8.750994656392109e-07,
"loss": 0.153,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.25735098419549485,
"learning_rate": 8.58630020171265e-07,
"loss": 0.1406,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.2706302621303061,
"learning_rate": 8.423136285876099e-07,
"loss": 0.1592,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.25200610302958426,
"learning_rate": 8.26150421353682e-07,
"loss": 0.1599,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.26838471129260233,
"learning_rate": 8.101405277100549e-07,
"loss": 0.1473,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.2525891487797046,
"learning_rate": 7.942840756714077e-07,
"loss": 0.1475,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.2577412523971255,
"learning_rate": 7.785811920255093e-07,
"loss": 0.1569,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.24933090739314834,
"learning_rate": 7.630320023322024e-07,
"loss": 0.1471,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.2555613407442317,
"learning_rate": 7.476366309223903e-07,
"loss": 0.1565,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.25561243420947344,
"learning_rate": 7.323952008970537e-07,
"loss": 0.1396,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.25343104257496996,
"learning_rate": 7.173078341262618e-07,
"loss": 0.1413,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.2635571429835534,
"learning_rate": 7.023746512481989e-07,
"loss": 0.157,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.255979013256347,
"learning_rate": 6.875957716681902e-07,
"loss": 0.1428,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.2579413421572432,
"learning_rate": 6.729713135577687e-07,
"loss": 0.1453,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.27561390318112927,
"learning_rate": 6.585013938537143e-07,
"loss": 0.1642,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.2604673454277188,
"learning_rate": 6.441861282571115e-07,
"loss": 0.1679,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.25160962638922935,
"learning_rate": 6.300256312324471e-07,
"loss": 0.1583,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.24916690567546756,
"learning_rate": 6.160200160066776e-07,
"loss": 0.1392,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.2602625399783633,
"learning_rate": 6.021693945683327e-07,
"loss": 0.1404,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.25670842480443584,
"learning_rate": 5.884738776666044e-07,
"loss": 0.1572,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.25613954217550106,
"learning_rate": 5.749335748104878e-07,
"loss": 0.1496,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.2575022257407841,
"learning_rate": 5.615485942678778e-07,
"loss": 0.1582,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.26539299749543893,
"learning_rate": 5.483190430647289e-07,
"loss": 0.1565,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.2507044925919369,
"learning_rate": 5.352450269841747e-07,
"loss": 0.1339,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.24860052539075037,
"learning_rate": 5.22326650565701e-07,
"loss": 0.1661,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.2488942549534064,
"learning_rate": 5.095640171043048e-07,
"loss": 0.1612,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.25371869744544967,
"learning_rate": 4.969572286496594e-07,
"loss": 0.1347,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.25337328920974533,
"learning_rate": 4.845063860053079e-07,
"loss": 0.1659,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.264733632505244,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.1534,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.25545648714500907,
"learning_rate": 4.600729351261857e-07,
"loss": 0.1672,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.25853534965072406,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.1535,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.259063615375163,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.1527,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.2585212679739568,
"learning_rate": 4.245948007320588e-07,
"loss": 0.1596,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.25396964359840857,
"learning_rate": 4.130816799401971e-07,
"loss": 0.1533,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.26323665285550685,
"learning_rate": 4.017251756252827e-07,
"loss": 0.1419,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.256421943877111,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.1549,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.25818576013902134,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.1397,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.25714513137104145,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.1557,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.2661785375907154,
"learning_rate": 3.578671204615347e-07,
"loss": 0.1634,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.2643769776628539,
"learning_rate": 3.472950355547267e-07,
"loss": 0.1462,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.25355533316801826,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.1269,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.23887600923767177,
"learning_rate": 3.266223765343224e-07,
"loss": 0.15,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.2550109484896568,
"learning_rate": 3.16521967718757e-07,
"loss": 0.1493,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.24891009124672034,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.1553,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.2567681107348134,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.1559,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.2462561838958348,
"learning_rate": 2.871653894244597e-07,
"loss": 0.1441,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.25825267095005633,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.1397,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.25345583790385445,
"learning_rate": 2.683823233205485e-07,
"loss": 0.151,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.24958567928872902,
"learning_rate": 2.592274132703287e-07,
"loss": 0.1614,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.2570439175338606,
"learning_rate": 2.502303499116754e-07,
"loss": 0.1458,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.2629592775156749,
"learning_rate": 2.413912051848666e-07,
"loss": 0.1398,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.25093606704925336,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.1589,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.25463383544257967,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.1549,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.2598973075382309,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.1451,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.2448178741381249,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.1283,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.25513763243690973,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.1573,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.25753871844119086,
"learning_rate": 1.916764975761254e-07,
"loss": 0.1517,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.25729170206756813,
"learning_rate": 1.839446914621923e-07,
"loss": 0.149,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.249038434523844,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.1701,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.25045973350588785,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.1496,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.25447863445358015,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.1562,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.25515180437109247,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.1526,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.2690887211976928,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.1571,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.2565851633534115,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.1528,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.25922791941003637,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.1695,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.2522500059960182,
"learning_rate": 1.277993209368744e-07,
"loss": 0.1386,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.2539515632285741,
"learning_rate": 1.214956261808542e-07,
"loss": 0.1327,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.26709194542384784,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.1462,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.2636433306289763,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.1516,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.25922581143443973,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.1728,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.2629837278578024,
"learning_rate": 9.787081204466253e-08,
"loss": 0.1497,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.26090674225525595,
"learning_rate": 9.236233586200627e-08,
"loss": 0.1395,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.2542330465545048,
"learning_rate": 8.70130406196279e-08,
"loss": 0.1433,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.2485901158787996,
"learning_rate": 8.182296909035047e-08,
"loss": 0.1618,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.2661660939316501,
"learning_rate": 7.679216277384439e-08,
"loss": 0.1507,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.2600572707057085,
"learning_rate": 7.19206618963053e-08,
"loss": 0.1505,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.2553966087977119,
"learning_rate": 6.720850541012347e-08,
"loss": 0.1698,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.25975105801657894,
"learning_rate": 6.265573099357714e-08,
"loss": 0.1485,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.2560886829834456,
"learning_rate": 5.826237505053067e-08,
"loss": 0.1433,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.24935909677244658,
"learning_rate": 5.4028472710137e-08,
"loss": 0.1692,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.275794422503228,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.1547,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.25527323900912846,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.1405,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.2531956241512576,
"learning_rate": 4.228381947004812e-08,
"loss": 0.1547,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.25274853021962107,
"learning_rate": 3.868805732811343e-08,
"loss": 0.145,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.2553872364616715,
"learning_rate": 3.525190530455636e-08,
"loss": 0.1388,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.26677095270365336,
"learning_rate": 3.197539087475887e-08,
"loss": 0.1582,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.25646327084341525,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.1749,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.26985242181217817,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.1453,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.2530352071011195,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.1552,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.2549921395622377,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.1401,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.2572071236701915,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.1335,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.2646043227826785,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.1577,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.2532164883187548,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.1386,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.2535270161548347,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.1665,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.26437535355588654,
"learning_rate": 9.67435416023843e-09,
"loss": 0.1579,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.2599099895038438,
"learning_rate": 7.995445897566178e-09,
"loss": 0.1498,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.2546010271337492,
"learning_rate": 6.476393170578021e-09,
"loss": 0.147,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.25520538103420687,
"learning_rate": 5.11720812557348e-09,
"loss": 0.1769,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.24761017369420318,
"learning_rate": 3.917901630563492e-09,
"loss": 0.1541,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.25426016539654483,
"learning_rate": 2.878483275170485e-09,
"loss": 0.1633,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.24880190242743536,
"learning_rate": 1.998961370555108e-09,
"loss": 0.1507,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.2537966549812671,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.1462,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.26055917468352296,
"learning_rate": 7.196337656112384e-10,
"loss": 0.1503,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.24247966643320795,
"learning_rate": 3.198382947511292e-10,
"loss": 0.1508,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.24966942250368204,
"learning_rate": 7.995973352770137e-11,
"loss": 0.1556,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.24423203883459188,
"learning_rate": 0.0,
"loss": 0.1292,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 1.836079948138283e+18,
"train_loss": 0.3704581044342836,
"train_runtime": 21246.0211,
"train_samples_per_second": 7.437,
"train_steps_per_second": 0.058
}
],
"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": 1.836079948138283e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}