e1_math_all_qwq_together / trainer_state.json
neginr's picture
End of training
e2d0586 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": 8.717077057202136,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.3198,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 8.73407974450614,
"learning_rate": 6.451612903225807e-07,
"loss": 1.3248,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 8.563081029118337,
"learning_rate": 9.67741935483871e-07,
"loss": 1.3267,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 8.743982630025517,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.3597,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 8.12190833619046,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.3046,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 7.947333254997,
"learning_rate": 1.935483870967742e-06,
"loss": 1.3257,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 6.741506178392702,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.332,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 5.998594589689949,
"learning_rate": 2.580645161290323e-06,
"loss": 1.2261,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 3.616532571651946,
"learning_rate": 2.903225806451613e-06,
"loss": 1.2239,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.970589148681019,
"learning_rate": 3.225806451612903e-06,
"loss": 1.2392,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.8398355229362977,
"learning_rate": 3.548387096774194e-06,
"loss": 1.2512,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 6.283472587398409,
"learning_rate": 3.870967741935484e-06,
"loss": 1.2103,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 6.931802673942392,
"learning_rate": 4.193548387096774e-06,
"loss": 1.2289,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 6.9647649178879885,
"learning_rate": 4.516129032258065e-06,
"loss": 1.2127,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 7.1688078597767815,
"learning_rate": 4.838709677419355e-06,
"loss": 1.266,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 5.375064437930492,
"learning_rate": 5.161290322580646e-06,
"loss": 1.169,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 4.815380267383903,
"learning_rate": 5.483870967741935e-06,
"loss": 1.1669,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 3.7637388147625988,
"learning_rate": 5.806451612903226e-06,
"loss": 1.1205,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 2.576477695894462,
"learning_rate": 6.129032258064517e-06,
"loss": 1.1277,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.8220593576046689,
"learning_rate": 6.451612903225806e-06,
"loss": 1.0995,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.6287067499342172,
"learning_rate": 6.774193548387097e-06,
"loss": 1.0967,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.6592772271102405,
"learning_rate": 7.096774193548388e-06,
"loss": 1.0788,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.7502774076296277,
"learning_rate": 7.4193548387096784e-06,
"loss": 1.0855,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.449173385598469,
"learning_rate": 7.741935483870968e-06,
"loss": 1.0538,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.0707174955389154,
"learning_rate": 8.064516129032258e-06,
"loss": 1.0519,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.8522195907826695,
"learning_rate": 8.387096774193549e-06,
"loss": 1.0204,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.9366738136263536,
"learning_rate": 8.70967741935484e-06,
"loss": 1.0182,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.9306102379161173,
"learning_rate": 9.03225806451613e-06,
"loss": 1.0013,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.9423102726055625,
"learning_rate": 9.35483870967742e-06,
"loss": 1.0449,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.7966834657821106,
"learning_rate": 9.67741935483871e-06,
"loss": 1.0379,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.8270132158282731,
"learning_rate": 1e-05,
"loss": 1.0151,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.8113538976548847,
"learning_rate": 1.0322580645161291e-05,
"loss": 1.0168,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.8049010848672515,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.989,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.6893911969709962,
"learning_rate": 1.096774193548387e-05,
"loss": 0.9753,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.8233056887799035,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.9889,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.6665630429130324,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.9717,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.6378539765970779,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.9635,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.7739852202182684,
"learning_rate": 1.2258064516129034e-05,
"loss": 1.0132,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.7493912851418406,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.9626,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.6069788144352212,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.9775,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.655470786082102,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.9697,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.5948837423233663,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.9425,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.5697875458427148,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.9702,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.572646680486803,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.9326,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.6101276017478687,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.973,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.7125888112788703,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.9806,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.6745170470877099,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.9572,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.5869207547580559,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.9357,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.5430356196929949,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.9519,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.6650311282004343,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.9365,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.793048785649017,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.9362,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.8627369000705034,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.9215,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.911408425389485,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.934,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.837724388464827,
"learning_rate": 1.741935483870968e-05,
"loss": 0.954,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.8359247323039358,
"learning_rate": 1.774193548387097e-05,
"loss": 0.9319,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 1.0028472372841208,
"learning_rate": 1.806451612903226e-05,
"loss": 0.95,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 1.716215343187442,
"learning_rate": 1.838709677419355e-05,
"loss": 0.9671,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.6490070273889359,
"learning_rate": 1.870967741935484e-05,
"loss": 0.9523,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 2.289121221458783,
"learning_rate": 1.903225806451613e-05,
"loss": 0.9369,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 1.088306767086354,
"learning_rate": 1.935483870967742e-05,
"loss": 0.9687,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 3.080439345715589,
"learning_rate": 1.967741935483871e-05,
"loss": 0.9563,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 2.8674672615228674,
"learning_rate": 2e-05,
"loss": 0.9356,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 1.4695776070788424,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.9425,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 2.231778722031106,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.9426,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 1.5990947266528974,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.9522,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 1.781804124852919,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.9414,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 1.2324123106392861,
"learning_rate": 2.161290322580645e-05,
"loss": 0.9062,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 1.5951391686324958,
"learning_rate": 2.193548387096774e-05,
"loss": 0.9612,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 1.180730760582771,
"learning_rate": 2.225806451612903e-05,
"loss": 0.9329,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 1.3059402812562149,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.9163,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 1.3269451805485164,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.9176,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 1.214885720178411,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.9578,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 1.4850998847415902,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.9199,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 1.4473563844024018,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.9371,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 1.2467111571091667,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.903,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.3871634464861982,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.9234,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 1.172104566792862,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.9179,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.2061588552026858,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.9154,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 0.9785617558494099,
"learning_rate": 2.548387096774194e-05,
"loss": 0.9253,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.048921075061821,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.9365,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.1825482779772607,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.9235,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.2500470022952617,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.9402,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 1.339401802811874,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.9491,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.8878612309673676,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.9333,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.9088872835598792,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.9278,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.2851534457942773,
"learning_rate": 2.774193548387097e-05,
"loss": 0.9257,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.1470759736789802,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.8997,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.3457902878948829,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.9359,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.375309808778082,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.9219,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.484306292461444,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.9521,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.42745875673281,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.9006,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 0.9700511554320095,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.9231,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.8045374401355803,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.8843,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.1417074826866376,
"learning_rate": 3.032258064516129e-05,
"loss": 0.9237,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.4554593553015405,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.9337,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.9811474654150498,
"learning_rate": 3.096774193548387e-05,
"loss": 0.9211,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.8494809287899228,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.9217,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.578405788313454,
"learning_rate": 3.161290322580645e-05,
"loss": 0.9147,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 1.74473404431181,
"learning_rate": 3.193548387096774e-05,
"loss": 0.9131,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.774373704960964,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.9383,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.646006121652112,
"learning_rate": 3.258064516129033e-05,
"loss": 0.8804,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.2176980217396214,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.9001,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 1.4601794247066893,
"learning_rate": 3.322580645161291e-05,
"loss": 0.9123,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.052214163726828,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.8957,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.633678371259503,
"learning_rate": 3.387096774193549e-05,
"loss": 0.8973,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.7639370360247593,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.9015,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.0846720404480863,
"learning_rate": 3.451612903225807e-05,
"loss": 0.9342,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 2.217893588110234,
"learning_rate": 3.483870967741936e-05,
"loss": 0.904,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.5333960454491418,
"learning_rate": 3.516129032258065e-05,
"loss": 0.8859,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.7341128791782932,
"learning_rate": 3.548387096774194e-05,
"loss": 0.9164,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.423408184659622,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.9018,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.5092270262419232,
"learning_rate": 3.612903225806452e-05,
"loss": 0.8843,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 1.181831089931261,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.9271,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.6959665982250978,
"learning_rate": 3.67741935483871e-05,
"loss": 0.9075,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.0384931248980018,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.9235,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 2.0421524224844094,
"learning_rate": 3.741935483870968e-05,
"loss": 0.9024,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 1.3701587011958993,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.9204,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.7441258506445467,
"learning_rate": 3.806451612903226e-05,
"loss": 0.9093,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 1.4918059124841605,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.8781,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 1.6769232283654563,
"learning_rate": 3.870967741935484e-05,
"loss": 0.8972,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.5062502036992729,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.9033,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.5620582290395648,
"learning_rate": 3.935483870967742e-05,
"loss": 0.9172,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.6484015890201227,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.9161,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.1641763144043402,
"learning_rate": 4e-05,
"loss": 0.8789,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.3614869039534756,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.8993,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.0287375822163698,
"learning_rate": 3.999968016170525e-05,
"loss": 0.8841,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.1121118965869534,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.8881,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 2.3478264100678192,
"learning_rate": 3.999872065705065e-05,
"loss": 0.9235,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 1.353620145655632,
"learning_rate": 3.999800103862945e-05,
"loss": 0.9078,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 2.0133741155220557,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.9021,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 1.4204863554244707,
"learning_rate": 3.999608209836944e-05,
"loss": 0.9146,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.5439860929592586,
"learning_rate": 3.999488279187443e-05,
"loss": 0.8994,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.4471852233160087,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.8889,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.1957538365426943,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.9101,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 1.4337823853310103,
"learning_rate": 3.999032564583976e-05,
"loss": 0.8928,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.3349493573824776,
"learning_rate": 3.998848689546591e-05,
"loss": 0.8853,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 1.7269611370101359,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.9146,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.777725282249926,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.9136,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.9422244879036534,
"learning_rate": 3.998201174509306e-05,
"loss": 0.8821,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 1.5877478998225056,
"learning_rate": 3.997953378607961e-05,
"loss": 0.9058,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 2.112213033468576,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.8985,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 1.4248742884217838,
"learning_rate": 3.997409862168454e-05,
"loss": 0.8648,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 2.1403990433962226,
"learning_rate": 3.997114145976237e-05,
"loss": 0.8826,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 1.7010973424648557,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.9011,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 1.8548602280040647,
"learning_rate": 3.996474809469545e-05,
"loss": 0.8888,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 1.4730402524421218,
"learning_rate": 3.996131194267189e-05,
"loss": 0.8829,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 1.8528635917549836,
"learning_rate": 3.995771618052995e-05,
"loss": 0.9128,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 1.2128384979641234,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.8972,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 2.126302192994177,
"learning_rate": 3.995004594217343e-05,
"loss": 0.9028,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 1.5468780192358957,
"learning_rate": 3.994597152728987e-05,
"loss": 0.903,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 1.9823866387322113,
"learning_rate": 3.994173762494947e-05,
"loss": 0.8941,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 1.6584548697165027,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.8944,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 1.4654039028451071,
"learning_rate": 3.993279149458988e-05,
"loss": 0.8948,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 1.4097416088509362,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.8978,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 1.3425362518392807,
"learning_rate": 3.992320783722616e-05,
"loss": 0.9236,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 1.1877457980436656,
"learning_rate": 3.991817703090965e-05,
"loss": 0.907,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 1.1491065439548402,
"learning_rate": 3.991298695938038e-05,
"loss": 0.8786,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 1.2719880082590764,
"learning_rate": 3.9907637664138e-05,
"loss": 0.8801,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 1.3996943111651652,
"learning_rate": 3.990212918795534e-05,
"loss": 0.8972,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 1.1021012703088864,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.876,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 1.0033426531607537,
"learning_rate": 3.989063487022416e-05,
"loss": 0.8827,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 1.500273307129403,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.8842,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.9690805266666519,
"learning_rate": 3.987850437381915e-05,
"loss": 0.8978,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 1.7269305426947235,
"learning_rate": 3.987220067906313e-05,
"loss": 0.914,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.9504552663876357,
"learning_rate": 3.986573808672002e-05,
"loss": 0.895,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 1.6001180456644768,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.8775,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 1.2112908841747483,
"learning_rate": 3.985233641724152e-05,
"loss": 0.8923,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 1.3166451247429332,
"learning_rate": 3.984539744726554e-05,
"loss": 0.8999,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 1.8686817827273638,
"learning_rate": 3.983829979402038e-05,
"loss": 0.8999,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 1.4175756611253294,
"learning_rate": 3.983104351425871e-05,
"loss": 0.8954,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 1.6815385875965778,
"learning_rate": 3.982362866600154e-05,
"loss": 0.9019,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 1.6735046961229267,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.8984,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 1.2843021488014976,
"learning_rate": 3.980832350242388e-05,
"loss": 0.8786,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 1.409422477670674,
"learning_rate": 3.980043330948306e-05,
"loss": 0.8933,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 1.2736756684086796,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.8921,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 1.2081729300483337,
"learning_rate": 3.978417801674579e-05,
"loss": 0.887,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 1.054110171290527,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.866,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 1.5322386473884744,
"learning_rate": 3.976728995023254e-05,
"loss": 0.8809,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 1.0139929683668134,
"learning_rate": 3.975860879481514e-05,
"loss": 0.8792,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 1.035880568984841,
"learning_rate": 3.974976965008833e-05,
"loss": 0.8768,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 1.1933263291479284,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.9079,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.9247379508935909,
"learning_rate": 3.973161767667946e-05,
"loss": 0.8872,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 1.1473813307696483,
"learning_rate": 3.972230499314009e-05,
"loss": 0.8944,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 1.0477709520726064,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.9023,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 1.8293471695976669,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.8885,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 1.123988618363668,
"learning_rate": 3.969342105253098e-05,
"loss": 0.9108,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 1.5080536390466273,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.8892,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 1.757725456440744,
"learning_rate": 3.967337762346568e-05,
"loss": 0.8785,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.9527013315661004,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.8977,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 2.5657281977708166,
"learning_rate": 3.965270496444528e-05,
"loss": 0.9134,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 1.6031731126149673,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.855,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 2.722168638429974,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.8815,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 2.1857286251731387,
"learning_rate": 3.962051762160151e-05,
"loss": 0.9158,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 2.495256122553091,
"learning_rate": 3.960947462140638e-05,
"loss": 0.8776,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 2.1312014182466,
"learning_rate": 3.959827482437472e-05,
"loss": 0.8717,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 2.0699576442569967,
"learning_rate": 3.958691832005981e-05,
"loss": 0.9002,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 1.9253355362639772,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.884,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 1.885291420593827,
"learning_rate": 3.956373555405774e-05,
"loss": 0.9031,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 1.3170908758354056,
"learning_rate": 3.955190947773937e-05,
"loss": 0.8828,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 2.1044006966852917,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.8911,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 1.2984445290804794,
"learning_rate": 3.952778841127214e-05,
"loss": 0.8839,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 2.198937754055747,
"learning_rate": 3.95154936139947e-05,
"loss": 0.8665,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 1.9571132004259255,
"learning_rate": 3.950304277135034e-05,
"loss": 0.9128,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 1.6209314172187055,
"learning_rate": 3.94904359828957e-05,
"loss": 0.8628,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 1.298026047723875,
"learning_rate": 3.94776733494343e-05,
"loss": 0.8741,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 1.3114698896261174,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.8647,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 1.1007760242126852,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.8617,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 1.1640089592218428,
"learning_rate": 3.943845140573213e-05,
"loss": 0.8754,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.8445732971232951,
"learning_rate": 3.942506642518952e-05,
"loss": 0.8672,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 1.076935391930916,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.8781,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 1.0515583327064595,
"learning_rate": 3.939783060543167e-05,
"loss": 0.9038,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 1.004233796391762,
"learning_rate": 3.938397998399333e-05,
"loss": 0.8644,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 1.1250136550571241,
"learning_rate": 3.936997436876756e-05,
"loss": 0.8635,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.9650728339256097,
"learning_rate": 3.935581387174289e-05,
"loss": 0.8783,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 1.3685192787144687,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.8488,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.8679331633553292,
"learning_rate": 3.932702868644224e-05,
"loss": 0.8673,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.7962917556185315,
"learning_rate": 3.931240422833181e-05,
"loss": 0.88,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 1.0901950743442368,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.8854,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 1.0242923916924922,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.879,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 1.8992868891064123,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.8709,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 1.1880593563977462,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.8766,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 2.3146265676885402,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.8932,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 2.074426383822035,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.8813,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 1.528141601937934,
"learning_rate": 3.92057159243286e-05,
"loss": 0.8735,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 1.354659632906594,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.8726,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 1.2441892926629838,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.8829,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 1.4256067496373483,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.8938,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 1.1839492670770757,
"learning_rate": 3.914136997982874e-05,
"loss": 0.8802,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 1.0621588271437892,
"learning_rate": 3.912490053436079e-05,
"loss": 0.8943,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 1.6459950596396813,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.87,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.9312153260186771,
"learning_rate": 3.909150300975178e-05,
"loss": 0.8866,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 1.7807775923954743,
"learning_rate": 3.907457519765642e-05,
"loss": 0.8775,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 1.3906042156128937,
"learning_rate": 3.905749486576607e-05,
"loss": 0.8517,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 1.1713129183903161,
"learning_rate": 3.90402621506546e-05,
"loss": 0.8677,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 1.6612626633500527,
"learning_rate": 3.902287719011435e-05,
"loss": 0.881,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.984734758546614,
"learning_rate": 3.900534012315499e-05,
"loss": 0.8684,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 2.0303347028702743,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.8863,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 1.6905855929898683,
"learning_rate": 3.896981023209777e-05,
"loss": 0.9043,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 1.892120420628357,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.8824,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 1.3106408076101699,
"learning_rate": 3.893367361386491e-05,
"loss": 0.8952,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 2.0538048382545133,
"learning_rate": 3.891537814248417e-05,
"loss": 0.8674,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 1.562144751212541,
"learning_rate": 3.889693142424384e-05,
"loss": 0.9008,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 2.048566352217155,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.8727,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 1.8049556830056386,
"learning_rate": 3.885958483839049e-05,
"loss": 0.8769,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 1.500375531115393,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.8752,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 1.4482930903892683,
"learning_rate": 3.882163505079171e-05,
"loss": 0.8988,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 1.6749707274311791,
"learning_rate": 3.880243433489132e-05,
"loss": 0.8675,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 1.3179178651913364,
"learning_rate": 3.878308327522702e-05,
"loss": 0.8418,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 1.6707271887143647,
"learning_rate": 3.876358202652936e-05,
"loss": 0.8699,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 1.3967218472805116,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.8597,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 1.535517394267547,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.8525,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 1.3207685730045746,
"learning_rate": 3.870417871154801e-05,
"loss": 0.8474,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 1.4877711120146855,
"learning_rate": 3.868407827802192e-05,
"loss": 0.8615,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 1.2355658340924816,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.842,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 1.6152443550012372,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.8671,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 1.2912204620692944,
"learning_rate": 3.862288124195319e-05,
"loss": 0.8587,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 1.4473561894151967,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.8333,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 1.3003329647898312,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.8514,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 1.515084935169108,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.8745,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 1.2944377746635323,
"learning_rate": 3.853920126717703e-05,
"loss": 0.846,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 1.4507679346412465,
"learning_rate": 3.851791025492574e-05,
"loss": 0.8253,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 1.145166362535298,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.8378,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 1.541697085694857,
"learning_rate": 3.847488419569866e-05,
"loss": 0.8731,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 1.4621883074663315,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.851,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 1.5013935465029147,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.8682,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 1.576916791693101,
"learning_rate": 3.840923760917199e-05,
"loss": 0.8507,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 1.0752222490028946,
"learning_rate": 3.838706077964473e-05,
"loss": 0.8322,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 1.3433817773008478,
"learning_rate": 3.836473692766944e-05,
"loss": 0.8691,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 1.057218236136573,
"learning_rate": 3.834226623174705e-05,
"loss": 0.8512,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 1.2248782343670521,
"learning_rate": 3.831964887155264e-05,
"loss": 0.8623,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.8962696314896386,
"learning_rate": 3.829688502793403e-05,
"loss": 0.8498,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 1.0546851382896205,
"learning_rate": 3.827397488291029e-05,
"loss": 0.8623,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.7130179680616302,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.8223,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 1.0465863227300483,
"learning_rate": 3.822771642257145e-05,
"loss": 0.8432,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.7538327675477806,
"learning_rate": 3.820436847713776e-05,
"loss": 0.8399,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 1.2123270740019247,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.8353,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.6787762979915554,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.8219,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.9332317345490105,
"learning_rate": 3.813345202354145e-05,
"loss": 0.8674,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 1.0960418813369737,
"learning_rate": 3.810952296329563e-05,
"loss": 0.8368,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.9048818339238396,
"learning_rate": 3.808544909978676e-05,
"loss": 0.854,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.6355327364267638,
"learning_rate": 3.806123062550882e-05,
"loss": 0.8225,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.6623884651718756,
"learning_rate": 3.803686773411208e-05,
"loss": 0.8471,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.6894776656344076,
"learning_rate": 3.801236062040158e-05,
"loss": 0.8463,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 1.0252005382675664,
"learning_rate": 3.798770948033553e-05,
"loss": 0.8308,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.9120654242327886,
"learning_rate": 3.796291451102381e-05,
"loss": 0.8334,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.741203967053604,
"learning_rate": 3.793797591072631e-05,
"loss": 0.8421,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.5376485040992061,
"learning_rate": 3.791289387885144e-05,
"loss": 0.8551,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.5318613496125442,
"learning_rate": 3.788766861595443e-05,
"loss": 0.8275,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.6776302738426053,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.825,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.6503025734830219,
"learning_rate": 3.783678920503983e-05,
"loss": 0.8225,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.7605085285055009,
"learning_rate": 3.781113546385265e-05,
"loss": 0.8243,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.8910473565115815,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.8406,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.8658377324940847,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.8318,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.8318530324191464,
"learning_rate": 3.77333205623025e-05,
"loss": 0.8303,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.5356006400015801,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.824,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.5824731489485483,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.8372,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.9316349922910493,
"learning_rate": 3.765422951112704e-05,
"loss": 0.8415,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 1.2537224557460798,
"learning_rate": 3.762758321970216e-05,
"loss": 0.8519,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.7537959890179878,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.8666,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.7145359983045113,
"learning_rate": 3.757386800198538e-05,
"loss": 0.8542,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.7230273211481736,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.8511,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.7807958292029413,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.8552,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.83137762587199,
"learning_rate": 3.749224181796305e-05,
"loss": 0.8216,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 1.0783797999233269,
"learning_rate": 3.746475306375545e-05,
"loss": 0.8491,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.7696285658590962,
"learning_rate": 3.743712466184774e-05,
"loss": 0.8306,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.6032082495275205,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.8358,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.5489211267405271,
"learning_rate": 3.73814497997107e-05,
"loss": 0.8436,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.5548397675808098,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.8542,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.7807526039946073,
"learning_rate": 3.732521901224724e-05,
"loss": 0.875,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.8407019345909487,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.8545,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.7547103924814242,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.8369,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.7127685465445234,
"learning_rate": 3.723983441007882e-05,
"loss": 0.8294,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.5543593611819981,
"learning_rate": 3.721109687296781e-05,
"loss": 0.8349,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.5122544896682636,
"learning_rate": 3.718222171638484e-05,
"loss": 0.8256,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.48772907828779055,
"learning_rate": 3.715320917121488e-05,
"loss": 0.8292,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.6467938851599885,
"learning_rate": 3.712405946944149e-05,
"loss": 0.8418,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.5354874901831779,
"learning_rate": 3.709477284414488e-05,
"loss": 0.8339,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.5730854916560109,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.8406,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.5927420805204957,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.865,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.5740203364130949,
"learning_rate": 3.70060937743295e-05,
"loss": 0.8328,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.64732075970761,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.8454,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.63160286744345,
"learning_rate": 3.694629409915555e-05,
"loss": 0.831,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.726614777167876,
"learning_rate": 3.691619088858401e-05,
"loss": 0.8353,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.8477952302003232,
"learning_rate": 3.68859524166009e-05,
"loss": 0.8317,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.8396696765561658,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.8567,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.6248268411287214,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.8373,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.6279262181311798,
"learning_rate": 3.67944278554384e-05,
"loss": 0.8534,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.6335893321081183,
"learning_rate": 3.676365076645557e-05,
"loss": 0.8103,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.6029065546278686,
"learning_rate": 3.673273963576791e-05,
"loss": 0.8267,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.6271276377843555,
"learning_rate": 3.670169471054001e-05,
"loss": 0.8425,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.4830726148037946,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.8266,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.49729539304246045,
"learning_rate": 3.663920447046888e-05,
"loss": 0.8469,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.5804399164500207,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.8401,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.474080578085958,
"learning_rate": 3.657618204491937e-05,
"loss": 0.8257,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.5001479136010148,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.8415,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.6333647202715383,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.8565,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.48555503987486837,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.8408,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.5452329029869084,
"learning_rate": 3.644854871713611e-05,
"loss": 0.8569,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.6005045939074958,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.8511,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.6901418777425293,
"learning_rate": 3.638394189710493e-05,
"loss": 0.8344,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.6918771194416522,
"learning_rate": 3.635144184933412e-05,
"loss": 0.8279,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5391047924409833,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.8354,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.7067585727153842,
"learning_rate": 3.628604977762752e-05,
"loss": 0.8145,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.5530969325552306,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.8235,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.9266458414398645,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.8499,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 1.0860487114507387,
"learning_rate": 3.618698565901727e-05,
"loss": 0.8259,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.7808075409468774,
"learning_rate": 3.615370507164654e-05,
"loss": 0.8415,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.7864989206542845,
"learning_rate": 3.61202953196805e-05,
"loss": 0.8743,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.6730121676642608,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.8465,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.4734753671537761,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.8373,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.4982958045134141,
"learning_rate": 3.601929375279662e-05,
"loss": 0.8464,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.5577358691071804,
"learning_rate": 3.598537002418012e-05,
"loss": 0.8214,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.7984669776348764,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.8363,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.9065245651196449,
"learning_rate": 3.591713938344416e-05,
"loss": 0.8201,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.9200183858696145,
"learning_rate": 3.588283301689507e-05,
"loss": 0.8454,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 1.0101465441081283,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.8611,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.8912457352310165,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.8585,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.7660818343170837,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.8566,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.5800727932401207,
"learning_rate": 3.574434032194994e-05,
"loss": 0.8228,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.809136673257855,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.8276,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.9033549456859421,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.8134,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 1.1108209746167739,
"learning_rate": 3.563914797637216e-05,
"loss": 0.8365,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.7773736973247549,
"learning_rate": 3.560383338592649e-05,
"loss": 0.8165,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.5844827270306644,
"learning_rate": 3.556839402764487e-05,
"loss": 0.8369,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.592140245089408,
"learning_rate": 3.553283018489946e-05,
"loss": 0.8205,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.8659741321647659,
"learning_rate": 3.54971421420578e-05,
"loss": 0.8642,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 1.0903668080309343,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.8062,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 1.0616639499240674,
"learning_rate": 3.54253945985191e-05,
"loss": 0.8598,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.7679938945161079,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.8401,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.48027725553280665,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.8529,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.7284747573835668,
"learning_rate": 3.531684894865866e-05,
"loss": 0.8426,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.8842284370127068,
"learning_rate": 3.528042173241128e-05,
"loss": 0.8705,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 1.0789080689320165,
"learning_rate": 3.524387233431892e-05,
"loss": 0.838,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.9031645997934836,
"learning_rate": 3.520720104662958e-05,
"loss": 0.8283,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.7152953944761786,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.8421,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.6763067573069287,
"learning_rate": 3.513349397632281e-05,
"loss": 0.8518,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.5548502110960519,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.851,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.6715454006487772,
"learning_rate": 3.505930287892533e-05,
"loss": 0.8765,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 1.0424534591759231,
"learning_rate": 3.502202656100099e-05,
"loss": 0.8691,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.799425212568037,
"learning_rate": 3.498463012735256e-05,
"loss": 0.8246,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.5162090824855455,
"learning_rate": 3.494711387700094e-05,
"loss": 0.8497,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.6774655362604096,
"learning_rate": 3.490947810992505e-05,
"loss": 0.834,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.947171384746483,
"learning_rate": 3.48717231270595e-05,
"loss": 0.8515,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.9810717180476648,
"learning_rate": 3.483384923029211e-05,
"loss": 0.834,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.6131710811701973,
"learning_rate": 3.479585672246156e-05,
"loss": 0.8248,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.698452854412634,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.8094,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.7738511681227984,
"learning_rate": 3.471951708970527e-05,
"loss": 0.8353,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.5019052401571946,
"learning_rate": 3.468117057518921e-05,
"loss": 0.8357,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.6002200355732925,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.8366,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.5816456406270768,
"learning_rate": 3.460412568296731e-05,
"loss": 0.813,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.47638480265391264,
"learning_rate": 3.456542792131039e-05,
"loss": 0.8382,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.5717465876769906,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.821,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.5117768704165521,
"learning_rate": 3.448768331403348e-05,
"loss": 0.8446,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.5943712219015852,
"learning_rate": 3.44486370900573e-05,
"loss": 0.8309,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.5362782278444816,
"learning_rate": 3.440947533516396e-05,
"loss": 0.8358,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.6005978573075129,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.8274,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.6656236179592638,
"learning_rate": 3.43308064860925e-05,
"loss": 0.8466,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.6105649943885447,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.8848,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.6858079716990868,
"learning_rate": 3.425167928295015e-05,
"loss": 0.8242,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.6267838360331325,
"learning_rate": 3.421194458890411e-05,
"loss": 0.8311,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.7424238651595483,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.8354,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.6595310244574957,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.8415,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.5316079501820514,
"learning_rate": 3.409205995219561e-05,
"loss": 0.8606,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.5329990804432004,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.8502,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.4462607142532273,
"learning_rate": 3.40115729298209e-05,
"loss": 0.8272,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.6299372194444457,
"learning_rate": 3.397116120327025e-05,
"loss": 0.8595,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.8402747848935657,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.8473,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.8826200223527036,
"learning_rate": 3.389000293509524e-05,
"loss": 0.8756,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.8400757261804322,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.857,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.7982523750103262,
"learning_rate": 3.380840041143494e-05,
"loss": 0.8217,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.7138648954750898,
"learning_rate": 3.376743336885792e-05,
"loss": 0.8067,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.6681012802934984,
"learning_rate": 3.372635624225055e-05,
"loss": 0.8104,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.5720945715618656,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.843,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.5046398613087686,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.8297,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.42142285580524635,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.8457,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.623003409833016,
"learning_rate": 3.356095347769799e-05,
"loss": 0.8262,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.8656329223723417,
"learning_rate": 3.351933087522552e-05,
"loss": 0.8709,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.9661020331529381,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.8284,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.8592745006438893,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.8254,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.5769800908480799,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.8283,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.6795805054283822,
"learning_rate": 3.335176280432308e-05,
"loss": 0.8369,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.8884416769000887,
"learning_rate": 3.330960304618295e-05,
"loss": 0.8411,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.8190439960997065,
"learning_rate": 3.326733686481153e-05,
"loss": 0.8495,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.7110408942794737,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.8379,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.4655931142178711,
"learning_rate": 3.318248658506012e-05,
"loss": 0.8256,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.6821141626203067,
"learning_rate": 3.313990316514069e-05,
"loss": 0.8405,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.6772280049203822,
"learning_rate": 3.309721467890571e-05,
"loss": 0.8345,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.8732133049014631,
"learning_rate": 3.305442146769115e-05,
"loss": 0.8325,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.7669701952469508,
"learning_rate": 3.30115238736704e-05,
"loss": 0.8158,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.49089891251841494,
"learning_rate": 3.296852223985148e-05,
"loss": 0.8003,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.6199852922895752,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.8587,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.6948769780734381,
"learning_rate": 3.288220822900796e-05,
"loss": 0.8322,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.6789554443052713,
"learning_rate": 3.283889654214788e-05,
"loss": 0.8084,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.5679950572282572,
"learning_rate": 3.279548219581317e-05,
"loss": 0.8392,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.5666304816680315,
"learning_rate": 3.275196553714379e-05,
"loss": 0.8187,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.46384569642238194,
"learning_rate": 3.270834691409778e-05,
"loss": 0.8283,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.6188401840229774,
"learning_rate": 3.266462667544849e-05,
"loss": 0.8092,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.4619396132562119,
"learning_rate": 3.262080517078178e-05,
"loss": 0.8293,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.6566508936709775,
"learning_rate": 3.257688275049323e-05,
"loss": 0.822,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.4366606973773963,
"learning_rate": 3.253285976578535e-05,
"loss": 0.8332,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.4659010988067617,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.833,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.542222987677847,
"learning_rate": 3.244451351193935e-05,
"loss": 0.8391,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.9618082525355824,
"learning_rate": 3.240019094921551e-05,
"loss": 0.8344,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.6238340017226341,
"learning_rate": 3.235576923489527e-05,
"loss": 0.8437,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.5360427647051232,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.836,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.5952440715201844,
"learning_rate": 3.226662977303494e-05,
"loss": 0.8308,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.5207423950385281,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.8368,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.6518948701240381,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.8277,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.5922118530015528,
"learning_rate": 3.21321858487569e-05,
"loss": 0.8166,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.6283385850976868,
"learning_rate": 3.208717671149931e-05,
"loss": 0.8407,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.4638992556773813,
"learning_rate": 3.204207092549882e-05,
"loss": 0.8444,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.8017006759056786,
"learning_rate": 3.19968688514201e-05,
"loss": 0.825,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.8263735927587246,
"learning_rate": 3.195157085069774e-05,
"loss": 0.8182,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.5586738836442032,
"learning_rate": 3.190617728553332e-05,
"loss": 0.838,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.5857720940175104,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.8204,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.6950607579267734,
"learning_rate": 3.181510491450254e-05,
"loss": 0.824,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.6319801447064536,
"learning_rate": 3.176942683684842e-05,
"loss": 0.8345,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.6130942038514943,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.8373,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.46978106483631576,
"learning_rate": 3.167778872346327e-05,
"loss": 0.8349,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.5222894931849617,
"learning_rate": 3.163182942046815e-05,
"loss": 0.8361,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.6842989908530157,
"learning_rate": 3.158577710967494e-05,
"loss": 0.8114,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.534281293118863,
"learning_rate": 3.153963215931669e-05,
"loss": 0.8355,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.5240902729594563,
"learning_rate": 3.14933949383672e-05,
"loss": 0.8258,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.5992917091686945,
"learning_rate": 3.144706581653805e-05,
"loss": 0.829,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.6448337187488371,
"learning_rate": 3.140064516427566e-05,
"loss": 0.8403,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.5882886542956104,
"learning_rate": 3.135413335275833e-05,
"loss": 0.8473,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.4354698049105088,
"learning_rate": 3.130753075389327e-05,
"loss": 0.8205,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.5654305470148392,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.8502,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.5055809668192633,
"learning_rate": 3.121405468537547e-05,
"loss": 0.8315,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.5576791448974084,
"learning_rate": 3.116718196315488e-05,
"loss": 0.8544,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.5582824776613439,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.8218,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.3799419447990565,
"learning_rate": 3.107316901675252e-05,
"loss": 0.8202,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.5588553497311747,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.8435,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.588444990256919,
"learning_rate": 3.097880190800057e-05,
"loss": 0.8089,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.5110211432616398,
"learning_rate": 3.093148648549788e-05,
"loss": 0.8259,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.41636871373046025,
"learning_rate": 3.088408365512055e-05,
"loss": 0.8788,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.38783430653330114,
"learning_rate": 3.083659379590034e-05,
"loss": 0.8556,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.4802090592070843,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.8087,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.41101990526244664,
"learning_rate": 3.074135451053475e-05,
"loss": 0.8191,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.48276315222082045,
"learning_rate": 3.069360584592015e-05,
"loss": 0.7936,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.365076726343317,
"learning_rate": 3.064577167551816e-05,
"loss": 0.8504,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.458161583321819,
"learning_rate": 3.059785238180953e-05,
"loss": 0.8237,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.47135968645017773,
"learning_rate": 3.054984834795566e-05,
"loss": 0.8364,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.3904324345028358,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.8196,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.5279395554158428,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.8547,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.5914985307775409,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.8547,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.484945429555782,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.8427,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.4840184713531061,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.8469,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.5286820494672207,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.8452,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.45958031930355375,
"learning_rate": 3.021147971398632e-05,
"loss": 0.8241,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.42879819052486573,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.8175,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.5278152924473862,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.8775,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.6956576765068326,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.8438,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.6851547019435457,
"learning_rate": 3.001632138223332e-05,
"loss": 0.8078,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.5301434237969255,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.8015,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.6843477202540791,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.7785,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.660016910359562,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.7708,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.7760261723393328,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.7845,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.74143647621078,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.7672,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.7041094614630151,
"learning_rate": 2.972118904869376e-05,
"loss": 0.7563,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.7498849746684962,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.7752,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.8575532390520516,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.8033,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.5799296225120721,
"learning_rate": 2.95725682050846e-05,
"loss": 0.7966,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.7577929497165186,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.7785,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.5176675933529341,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.7736,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.804327018835391,
"learning_rate": 2.942325848714505e-05,
"loss": 0.8046,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 1.202181477378266,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.7926,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.6235619350510673,
"learning_rate": 2.932334127129686e-05,
"loss": 0.77,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.9363289616961874,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.7768,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.5667907789295361,
"learning_rate": 2.922312585929131e-05,
"loss": 0.7899,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.6926093359065822,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.7856,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.4473939124306412,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.7818,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.5174792162182881,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.7326,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.5025879605453734,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.7658,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.558796244989256,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.7903,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.39011231797583357,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.7862,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.5423582143356043,
"learning_rate": 2.887006998351372e-05,
"loss": 0.8041,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.5432826399072437,
"learning_rate": 2.881934649605535e-05,
"loss": 0.7944,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.562678727519141,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.7781,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.6637992649832085,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.8051,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.47070202262046407,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.801,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.6376863700266083,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.772,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.6600107857319129,
"learning_rate": 2.856467940121025e-05,
"loss": 0.7937,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.5164483149739486,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.8082,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.4536462131618645,
"learning_rate": 2.846233032790149e-05,
"loss": 0.7763,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.4542316457315252,
"learning_rate": 2.84110540896642e-05,
"loss": 0.78,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.5070885581305077,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.7706,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.5514784181667436,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.7639,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.41859517318500533,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.7855,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.4608316778361714,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.7758,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.38886901121368844,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.7674,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.41236680643356455,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.7896,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.5078856124513287,
"learning_rate": 2.805026032173773e-05,
"loss": 0.7766,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.46013370255838487,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.7813,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.5199558527449399,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.7771,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.3990992103630945,
"learning_rate": 2.7894660620408e-05,
"loss": 0.7786,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.4635091553066939,
"learning_rate": 2.784266724883173e-05,
"loss": 0.7928,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.4445632510612474,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.7748,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.4420686199925253,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.7912,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.4226073388178384,
"learning_rate": 2.768631254100716e-05,
"loss": 0.7873,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.5490103183380579,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.7661,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.46209943326103176,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.7948,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.49154130172693994,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.7845,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.5270847475204204,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.793,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.4725660354592679,
"learning_rate": 2.742449775134085e-05,
"loss": 0.7894,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.5988852511049476,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.8239,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.5114352116524442,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.7894,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.6568159256146845,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.7974,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 1.7312166072722037,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.9052,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.663323323503205,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.7976,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.39646086370179323,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.785,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.6484040849210742,
"learning_rate": 2.705547548999996e-05,
"loss": 0.7853,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.4096337214397728,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.8131,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.7013310771444832,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.7709,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.4713569219625446,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.7697,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.5822729258755129,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.7787,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.4112513525994735,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.7676,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.47647949848525106,
"learning_rate": 2.67369650552155e-05,
"loss": 0.7922,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.41401646004980236,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.7912,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.4564685141265274,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.809,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.43260641756639284,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.8027,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.4384291802473379,
"learning_rate": 2.652354173593614e-05,
"loss": 0.7764,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.3861424652559147,
"learning_rate": 2.647005443331317e-05,
"loss": 0.8219,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.4178093578408944,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.809,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.38412447858989834,
"learning_rate": 2.636292505301542e-05,
"loss": 0.7976,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.41573228257989586,
"learning_rate": 2.630928383194431e-05,
"loss": 0.7556,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.4456580380826303,
"learning_rate": 2.625559216200781e-05,
"loss": 0.7997,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.5725121395834575,
"learning_rate": 2.620185047252308e-05,
"loss": 0.76,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.47929267942344633,
"learning_rate": 2.614805919320723e-05,
"loss": 0.7783,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.5006351576774252,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.8019,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.4068542797224279,
"learning_rate": 2.60403295859298e-05,
"loss": 0.7938,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.49850329846298985,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.792,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.4072659761247321,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.7833,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.40580158320216103,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.8004,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.4336094936231124,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.7901,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.44497782313997064,
"learning_rate": 2.577016790136453e-05,
"loss": 0.7861,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.4468574684391416,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.8024,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.43355996037245814,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.7755,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.45839529797167344,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.7781,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.47084489171735505,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.7823,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.38309585592405365,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.7851,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.46866044981454996,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.777,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.47258413180839465,
"learning_rate": 2.539001591449889e-05,
"loss": 0.7933,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.35040788821937724,
"learning_rate": 2.53355326151989e-05,
"loss": 0.7725,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.423275876257118,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.7965,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.42136718534412443,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.7928,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.4135227289511859,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.7752,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.4200056854676083,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.7911,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.48536601941630714,
"learning_rate": 2.506248490185627e-05,
"loss": 0.7929,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.452796337820375,
"learning_rate": 2.500775217241923e-05,
"loss": 0.7902,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.36424279146154587,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.7812,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.40540913306637166,
"learning_rate": 2.4898167025948e-05,
"loss": 0.7677,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.37569979075130694,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.7613,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.3860917202321348,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.788,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.3881686093403846,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.7858,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.4394188136904339,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.8099,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.41506003487488863,
"learning_rate": 2.46235264422994e-05,
"loss": 0.7837,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.41832238017870044,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.8137,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.4285777930581317,
"learning_rate": 2.451340834511592e-05,
"loss": 0.8003,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.5876411798855117,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.7931,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.3975164814531544,
"learning_rate": 2.440314589184958e-05,
"loss": 0.7709,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.6015291178594853,
"learning_rate": 2.434796163357541e-05,
"loss": 0.7863,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.4215207983692927,
"learning_rate": 2.429274260911588e-05,
"loss": 0.7799,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.49726344776875914,
"learning_rate": 2.423748926000084e-05,
"loss": 0.7608,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.36661832437941183,
"learning_rate": 2.418220202803459e-05,
"loss": 0.7881,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.5760620677429236,
"learning_rate": 2.412688135529237e-05,
"loss": 0.7935,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.3750142421972317,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.7971,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.4811509830568932,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.8083,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.39777680136545396,
"learning_rate": 2.396072311715187e-05,
"loss": 0.7849,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.5038624060499117,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.7893,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.420137079281489,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.7906,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.3691159958168798,
"learning_rate": 2.379427985200223e-05,
"loss": 0.7892,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.429400009643035,
"learning_rate": 2.373873749394972e-05,
"loss": 0.7824,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.364404066969196,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.7828,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.5179915998738095,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.7923,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.5624063647896611,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.7784,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.49130688554267576,
"learning_rate": 2.351627355800233e-05,
"loss": 0.7812,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.36378672924851024,
"learning_rate": 2.346058617160256e-05,
"loss": 0.7929,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.4452443490025012,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.8026,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.38257798192684817,
"learning_rate": 2.334912883203469e-05,
"loss": 0.8103,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.4525759962979123,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.7744,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.3963241391630055,
"learning_rate": 2.323756437450136e-05,
"loss": 0.7879,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.44258128223392096,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.7818,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.562902752537454,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.8037,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.46066081565574035,
"learning_rate": 2.307002464849052e-05,
"loss": 0.7686,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.3910076807967767,
"learning_rate": 2.30141283818846e-05,
"loss": 0.782,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.5042032004752378,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.7732,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.4902985435179134,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.7817,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.4518204939683741,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.8,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.42205447373049587,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.7956,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.40618277380749807,
"learning_rate": 2.273429448107257e-05,
"loss": 0.791,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.32985247313297444,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.7746,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.3658305837308003,
"learning_rate": 2.262220474347102e-05,
"loss": 0.7811,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.4304176747514498,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.7837,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.41371785638946146,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.7947,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.4749837443550939,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.7796,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.43505710273690207,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.7732,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.36607886959065944,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.7984,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.3652778432226744,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.7917,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.36748505743475196,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.805,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.4073456697677083,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.7825,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.3516638022717686,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.7673,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.42968601325557176,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.8302,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.37162007801158853,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.7772,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.43737646774493155,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.8065,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.42375659859892584,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.7714,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.5039308805439984,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.7892,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.4061773040491222,
"learning_rate": 2.17791064266051e-05,
"loss": 0.7727,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.39856321329242517,
"learning_rate": 2.172276926861741e-05,
"loss": 0.787,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.4107610990496439,
"learning_rate": 2.166641833541255e-05,
"loss": 0.7686,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.4648711939293097,
"learning_rate": 2.161005407757109e-05,
"loss": 0.7995,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.34726357975548805,
"learning_rate": 2.155367694578013e-05,
"loss": 0.8047,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.44852701287259933,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.7963,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.39848396261763686,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.7823,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.38872506408914376,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.7895,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.40108664859058524,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.7974,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.40625800384084076,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.789,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.379100548067334,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.7664,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.398932676286566,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.7838,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.40625958480156715,
"learning_rate": 2.110225054402733e-05,
"loss": 0.8013,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.41370567982202955,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.7925,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.38725406775459603,
"learning_rate": 2.09892967807724e-05,
"loss": 0.7759,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.4759104163708301,
"learning_rate": 2.093280780774835e-05,
"loss": 0.7826,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.4979318212247079,
"learning_rate": 2.087631137601793e-05,
"loss": 0.7976,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.5412728447943578,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.8096,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.5242278710902902,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.7852,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.5356878437798293,
"learning_rate": 2.070678184630458e-05,
"loss": 0.7767,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.3977138178057268,
"learning_rate": 2.06502600977305e-05,
"loss": 0.793,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.4454510974589826,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.768,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.4210931394670057,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.7607,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.3712002721409388,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.7814,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.39056886614691366,
"learning_rate": 2.042412562887594e-05,
"loss": 0.7668,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.3272291458853311,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.7619,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.437971560940654,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.7865,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.3490241084143671,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.777,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.41692254485290536,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.7775,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.4038710387174913,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.8096,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.4385604588510723,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.7874,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.45208416508431176,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.7774,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.45121407963400073,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.7828,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.4190483425948251,
"learning_rate": 1.99151687695709e-05,
"loss": 0.7898,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.4477042486511187,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.7961,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.3920132251661349,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.775,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.33027507244004556,
"learning_rate": 1.974551241345447e-05,
"loss": 0.789,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.35119400430629566,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.7875,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.3299853785909362,
"learning_rate": 1.963241759676897e-05,
"loss": 0.7807,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.3626675103761053,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.7573,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.34685665642692465,
"learning_rate": 1.951933453677638e-05,
"loss": 0.7749,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.37620908245925755,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.812,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.3847178402850913,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.7587,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.3035361584053321,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.782,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.4287200304871788,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.7668,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.5348389895922298,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.81,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.36536756798761333,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.8098,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.5229268015329802,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.7662,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.5170110251287309,
"learning_rate": 1.906719219225165e-05,
"loss": 0.7828,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.4135629813786101,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.7857,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.6389681079757368,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.8071,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.3955952532338688,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.7794,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.4744723647485214,
"learning_rate": 1.884128556891707e-05,
"loss": 0.7986,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.4534323411891586,
"learning_rate": 1.878483094691119e-05,
"loss": 0.7819,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.4260724155701338,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.7782,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.4987805436535235,
"learning_rate": 1.867195130360947e-05,
"loss": 0.7994,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.42895636762270195,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.7776,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.4330383204028291,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.7883,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.38727194892759853,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.7822,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.4620690993307156,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.8003,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.45919832292924484,
"learning_rate": 1.838994592242891e-05,
"loss": 0.754,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.4517332344398096,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.7889,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.5629711344458151,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.7811,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.3996100944123458,
"learning_rate": 1.82208935733949e-05,
"loss": 0.7988,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.42849584733750645,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.7887,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.4428542801079424,
"learning_rate": 1.810826238483892e-05,
"loss": 0.7951,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.37215657882585357,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.8047,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.42875447973889214,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.7807,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.35717315774356323,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.795,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.395225069667372,
"learning_rate": 1.788318512320838e-05,
"loss": 0.7841,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.3767610486357836,
"learning_rate": 1.782695699829051e-05,
"loss": 0.7818,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.43908454244849215,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.7802,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.353841302779209,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.792,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.4431064988019766,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.7767,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.3409311823334651,
"learning_rate": 1.76022227484387e-05,
"loss": 0.7921,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.4671165897262597,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.783,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.3858020382648713,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.7602,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.4030590926792235,
"learning_rate": 1.743387180005809e-05,
"loss": 0.781,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.5013982965241162,
"learning_rate": 1.737779525652899e-05,
"loss": 0.7891,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.30716209121575655,
"learning_rate": 1.732173968007914e-05,
"loss": 0.7637,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.4700047126156623,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.7878,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.38770060563658293,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.7602,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.42987112364167845,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.7619,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.39932218928190605,
"learning_rate": 1.709773600686015e-05,
"loss": 0.7736,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.45192125782562853,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.795,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.41614818949325094,
"learning_rate": 1.69858716181154e-05,
"loss": 0.7712,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.44486833199913633,
"learning_rate": 1.692997535150948e-05,
"loss": 0.7927,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.39176889505195117,
"learning_rate": 1.687410363273884e-05,
"loss": 0.7671,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.3850332079508979,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.7875,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.38152500728784117,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.7884,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.31003488658059797,
"learning_rate": 1.67066402299235e-05,
"loss": 0.7783,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.36501126582874666,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.7562,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.34974893099021553,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.7627,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.39666673053038354,
"learning_rate": 1.653941382839745e-05,
"loss": 0.8114,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.422106275469367,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.7426,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.39207407182078874,
"learning_rate": 1.642806717162757e-05,
"loss": 0.7199,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.44764778726100096,
"learning_rate": 1.637243646233718e-05,
"loss": 0.7306,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.47842594522841936,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.7491,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.48513928974358245,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.708,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.5149363519379652,
"learning_rate": 1.620572014799777e-05,
"loss": 0.7243,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.5243757037753638,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.7304,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.5232876764324049,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.7285,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.5583438055950111,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.7499,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.49088030692361034,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.758,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.5357228928274299,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.7301,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.4674397991137305,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.7468,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.56369029745893,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.7215,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.3470004715409878,
"learning_rate": 1.576251073999917e-05,
"loss": 0.7384,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.5258744458693203,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.7347,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.4619687084470259,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.7162,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.40619782295749196,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.7244,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.4894071577919648,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.7437,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.34769548394912553,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.7441,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.49976604203916225,
"learning_rate": 1.543151434154755e-05,
"loss": 0.7291,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.37518857407974526,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.7295,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.4507331277938761,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.7331,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.4364655473961867,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.7465,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.45691756422535607,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.7275,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.41629495513394904,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.7575,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.3597415653228408,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.7377,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.377058350379016,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.7483,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.39472401155180453,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.7286,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.35891387582097584,
"learning_rate": 1.493751509814374e-05,
"loss": 0.711,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.3968048796763944,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.7336,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.34158340841834195,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.7383,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.40934517591011693,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.7322,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.3321043839501841,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.7374,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.35748277955176433,
"learning_rate": 1.466446738480111e-05,
"loss": 0.7426,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.33165970724475974,
"learning_rate": 1.460998408550112e-05,
"loss": 0.7351,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.3619886956732458,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.7373,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.35701179276935835,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.7201,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.39443184332578396,
"learning_rate": 1.444679451900437e-05,
"loss": 0.7561,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.29955748965188767,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.7356,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.4625065105453688,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.7359,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.29642309027579594,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.7485,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.3563430337946367,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.746,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.37234738716418997,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.7502,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.2915888152085597,
"learning_rate": 1.412162598317602e-05,
"loss": 0.7365,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.4196913941071035,
"learning_rate": 1.406759321421907e-05,
"loss": 0.727,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.30320311197964434,
"learning_rate": 1.401360788062871e-05,
"loss": 0.7228,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.3945816531181313,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.7582,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.4376643786066289,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.7424,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.3374435000133517,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.7345,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.4132515173889866,
"learning_rate": 1.379814952747693e-05,
"loss": 0.7425,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.3096642112848635,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.7394,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.3370504218489404,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.7376,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.3557576649433343,
"learning_rate": 1.363707494698459e-05,
"loss": 0.7083,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.2959648796338822,
"learning_rate": 1.358348460369265e-05,
"loss": 0.7531,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.33259598581268573,
"learning_rate": 1.352994556668684e-05,
"loss": 0.7273,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.2819303436901885,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.748,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.36120730266066275,
"learning_rate": 1.342302312350678e-05,
"loss": 0.7339,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.26819491912144516,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.7134,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.3285229869156072,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.7104,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.29909964017659646,
"learning_rate": 1.326303494478451e-05,
"loss": 0.7188,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.35808527053015154,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.7267,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.2747032803527441,
"learning_rate": 1.315664479122898e-05,
"loss": 0.7159,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.31059107646928863,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.7485,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.30863707898541154,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.75,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.2804466076494519,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.7637,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.325449849034207,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.7334,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.2961851939046928,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.7089,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.2910438273479294,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.7393,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.3032691262601605,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.7337,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.3511886101818047,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.7606,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.3420621204784597,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.7381,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.2991265827703402,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.7475,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.35110904144406613,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.7563,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.3155423848553404,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.7407,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.30784506103252074,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.7566,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.3277090275109362,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.7349,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.3377281621133364,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.7329,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.31032337363633034,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.7361,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.31695085191058,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.7207,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.3042062809911249,
"learning_rate": 1.220938883250288e-05,
"loss": 0.7404,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.27091999341173967,
"learning_rate": 1.215733275116827e-05,
"loss": 0.7482,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.32768042785022006,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.7465,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.3012667930316465,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.7311,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.460529170874744,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.758,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.3244014464861419,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.7414,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.3040872807173849,
"learning_rate": 1.189800129803139e-05,
"loss": 0.7452,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.3034588750956295,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.734,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.32428777202906217,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.7545,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.3339120742266572,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.7441,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.2491592586359776,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.7512,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.29613926598434054,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.7544,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.2700024830073618,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.7315,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.2831363534017283,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.756,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.30702242248064765,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.7382,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.27243448165507783,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.73,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.2828764806847012,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.7516,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.3089609044321958,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.7515,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.29233393668111224,
"learning_rate": 1.128231163049258e-05,
"loss": 0.7116,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.2930307582347836,
"learning_rate": 1.123144751066259e-05,
"loss": 0.7189,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.28391789868782347,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.7158,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.32640554219409995,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.7463,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.26914146966020935,
"learning_rate": 1.107927745387114e-05,
"loss": 0.7505,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.33506640413230815,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.749,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.2808516116249565,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.7359,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.3009544462213639,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.7289,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.39339686772204996,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.729,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.307481441377854,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.7361,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.32083384917996055,
"learning_rate": 1.077687414070869e-05,
"loss": 0.7209,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.2743688874465913,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.7442,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.3443633000892172,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.7504,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.2964998094268687,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.7395,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.2849542894798217,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.7181,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.2985123436724555,
"learning_rate": 1.052689572753247e-05,
"loss": 0.7337,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.32217951559818037,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.739,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.2776934630658145,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.7649,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.27517708854694733,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.7208,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.3721821749645142,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.7407,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.2921178306084925,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.7478,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.30594054003651994,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.7433,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.3363007024305207,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.738,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.2785756183437176,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.7162,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.3357573886173712,
"learning_rate": 1.008173988718005e-05,
"loss": 0.7468,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.31214754155669416,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.7626,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.2666803161116507,
"learning_rate": 9.983678617766677e-06,
"loss": 0.745,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.2898021592246684,
"learning_rate": 9.934767922553788e-06,
"loss": 0.7372,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.2909342921092306,
"learning_rate": 9.885937708668371e-06,
"loss": 0.7121,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.2727529533342857,
"learning_rate": 9.837188366555524e-06,
"loss": 0.7314,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.274053071920743,
"learning_rate": 9.788520286013691e-06,
"loss": 0.7452,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.31732490250220247,
"learning_rate": 9.739933856191537e-06,
"loss": 0.7528,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.2663677053302905,
"learning_rate": 9.691429465584866e-06,
"loss": 0.7381,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.2724309057082169,
"learning_rate": 9.643007502033492e-06,
"loss": 0.7439,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.3327546476250683,
"learning_rate": 9.594668352718152e-06,
"loss": 0.727,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.2757036121721533,
"learning_rate": 9.546412404157383e-06,
"loss": 0.7382,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.28767412308690143,
"learning_rate": 9.498240042204471e-06,
"loss": 0.7631,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.3295802411238543,
"learning_rate": 9.450151652044339e-06,
"loss": 0.736,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.30584691514786133,
"learning_rate": 9.402147618190472e-06,
"loss": 0.733,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.27390745129951727,
"learning_rate": 9.354228324481841e-06,
"loss": 0.7584,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.2993799040799575,
"learning_rate": 9.306394154079852e-06,
"loss": 0.7321,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.28807114812437257,
"learning_rate": 9.258645489465254e-06,
"loss": 0.7446,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.28020520071884086,
"learning_rate": 9.21098271243509e-06,
"loss": 0.7363,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.26219236977429816,
"learning_rate": 9.163406204099659e-06,
"loss": 0.7224,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.263775947975912,
"learning_rate": 9.115916344879452e-06,
"loss": 0.7299,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.27289700664046496,
"learning_rate": 9.068513514502126e-06,
"loss": 0.7383,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.2811344012292321,
"learning_rate": 9.021198091999439e-06,
"loss": 0.7538,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.2614514971797044,
"learning_rate": 8.973970455704251e-06,
"loss": 0.7527,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.26705417157834266,
"learning_rate": 8.926830983247495e-06,
"loss": 0.7329,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.2630291650987163,
"learning_rate": 8.879780051555116e-06,
"loss": 0.7249,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.32025365090134567,
"learning_rate": 8.83281803684512e-06,
"loss": 0.7425,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.26567438838003554,
"learning_rate": 8.785945314624532e-06,
"loss": 0.7206,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.2693131491878029,
"learning_rate": 8.739162259686385e-06,
"loss": 0.7287,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.28532405642795294,
"learning_rate": 8.692469246106734e-06,
"loss": 0.7244,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.2531628415909308,
"learning_rate": 8.645866647241677e-06,
"loss": 0.7448,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.30527171685830873,
"learning_rate": 8.599354835724348e-06,
"loss": 0.735,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.28874222337456124,
"learning_rate": 8.552934183461958e-06,
"loss": 0.7579,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.23032172063911385,
"learning_rate": 8.506605061632805e-06,
"loss": 0.7387,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.31544426240464624,
"learning_rate": 8.460367840683308e-06,
"loss": 0.7406,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.278312858399336,
"learning_rate": 8.414222890325065e-06,
"loss": 0.7442,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.3146438914672329,
"learning_rate": 8.368170579531855e-06,
"loss": 0.7483,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.27467827583567966,
"learning_rate": 8.322211276536733e-06,
"loss": 0.7459,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.2835852155522614,
"learning_rate": 8.27634534882907e-06,
"loss": 0.7333,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.2650662034746144,
"learning_rate": 8.230573163151585e-06,
"loss": 0.7429,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.2645722323840944,
"learning_rate": 8.184895085497466e-06,
"loss": 0.7266,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.2517190786369726,
"learning_rate": 8.139311481107401e-06,
"loss": 0.7302,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.2706875558756239,
"learning_rate": 8.093822714466686e-06,
"loss": 0.7635,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.26050137693584813,
"learning_rate": 8.048429149302271e-06,
"loss": 0.73,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.2739879737311377,
"learning_rate": 8.003131148579902e-06,
"loss": 0.7518,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.25147586821422796,
"learning_rate": 7.957929074501182e-06,
"loss": 0.7246,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.25185531201130734,
"learning_rate": 7.912823288500695e-06,
"loss": 0.724,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.2946147261576977,
"learning_rate": 7.867814151243102e-06,
"loss": 0.7368,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.2514068626745249,
"learning_rate": 7.822902022620267e-06,
"loss": 0.7476,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.25574318093507414,
"learning_rate": 7.778087261748377e-06,
"loss": 0.7571,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.2752007323489494,
"learning_rate": 7.733370226965056e-06,
"loss": 0.7523,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.26242301370120724,
"learning_rate": 7.688751275826527e-06,
"loss": 0.7524,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.24038057885305025,
"learning_rate": 7.644230765104735e-06,
"loss": 0.7444,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.2743993139231061,
"learning_rate": 7.599809050784502e-06,
"loss": 0.7589,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.24190178051220804,
"learning_rate": 7.555486488060661e-06,
"loss": 0.7227,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.2783145907384766,
"learning_rate": 7.511263431335249e-06,
"loss": 0.7375,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.24645822880614382,
"learning_rate": 7.467140234214656e-06,
"loss": 0.7383,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.2740083748252997,
"learning_rate": 7.423117249506777e-06,
"loss": 0.7524,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.2779533525876971,
"learning_rate": 7.379194829218231e-06,
"loss": 0.7519,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.23919973446030887,
"learning_rate": 7.335373324551518e-06,
"loss": 0.7254,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.25182860551425834,
"learning_rate": 7.291653085902224e-06,
"loss": 0.7479,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.29266702705634734,
"learning_rate": 7.248034462856213e-06,
"loss": 0.7518,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.26884504821464944,
"learning_rate": 7.20451780418683e-06,
"loss": 0.7382,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.23797925031080577,
"learning_rate": 7.161103457852128e-06,
"loss": 0.7442,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.29851380868599775,
"learning_rate": 7.117791770992046e-06,
"loss": 0.7455,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.26215340582567387,
"learning_rate": 7.074583089925688e-06,
"loss": 0.7528,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.28030694392683325,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.7376,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.27710328804138623,
"learning_rate": 6.988476126329604e-06,
"loss": 0.7626,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.27899480137645993,
"learning_rate": 6.945578532308857e-06,
"loss": 0.7498,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.2921514530184704,
"learning_rate": 6.902785321094301e-06,
"loss": 0.7416,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.25292609016682577,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.7398,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.2840975825012877,
"learning_rate": 6.817513414939892e-06,
"loss": 0.7284,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.25700434166747826,
"learning_rate": 6.775035401831924e-06,
"loss": 0.7409,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.3065117739709514,
"learning_rate": 6.732663135188473e-06,
"loss": 0.7266,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.2543213012957332,
"learning_rate": 6.690396953817055e-06,
"loss": 0.7372,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.25349465278358685,
"learning_rate": 6.648237195676927e-06,
"loss": 0.7336,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.2716540230302821,
"learning_rate": 6.606184197876395e-06,
"loss": 0.7454,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.2935385527241577,
"learning_rate": 6.564238296670109e-06,
"loss": 0.7322,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.2667821862958004,
"learning_rate": 6.52239982745638e-06,
"loss": 0.7296,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.23876431822425082,
"learning_rate": 6.480669124774484e-06,
"loss": 0.7305,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.3032881781362573,
"learning_rate": 6.439046522302013e-06,
"loss": 0.7555,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.27878417112800835,
"learning_rate": 6.397532352852191e-06,
"loss": 0.7214,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.25682290288767373,
"learning_rate": 6.356126948371202e-06,
"loss": 0.7198,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.243021358264394,
"learning_rate": 6.314830639935559e-06,
"loss": 0.7145,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.2863879623131075,
"learning_rate": 6.273643757749448e-06,
"loss": 0.7505,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.2655856300330848,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.7189,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.2616577577809902,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.7367,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.28514363427982237,
"learning_rate": 6.150742957589784e-06,
"loss": 0.748,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.3098385852450441,
"learning_rate": 6.109997064904774e-06,
"loss": 0.7352,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.27683670790799375,
"learning_rate": 6.069362236313092e-06,
"loss": 0.7471,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.28404883796628333,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.7475,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.31243678615804205,
"learning_rate": 5.988427070179108e-06,
"loss": 0.7575,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.29124081048315115,
"learning_rate": 5.948127379792234e-06,
"loss": 0.7418,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.2519642291306823,
"learning_rate": 5.90794004780439e-06,
"loss": 0.7476,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.23277567332425853,
"learning_rate": 5.867865395552405e-06,
"loss": 0.7477,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.28135352314985246,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.7278,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.2752641857065284,
"learning_rate": 5.788055411095892e-06,
"loss": 0.7448,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.26626862607992896,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.7528,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.24752171085004213,
"learning_rate": 5.708699979051613e-06,
"loss": 0.7399,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.23402145233153704,
"learning_rate": 5.66919351390751e-06,
"loss": 0.7085,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.23488763994484402,
"learning_rate": 5.629801637510188e-06,
"loss": 0.7314,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.25911004768066814,
"learning_rate": 5.590524664836043e-06,
"loss": 0.7406,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.26265996814031584,
"learning_rate": 5.551362909942708e-06,
"loss": 0.7191,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.23500682338298282,
"learning_rate": 5.51231668596653e-06,
"loss": 0.7485,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.2619479570739235,
"learning_rate": 5.47338630512007e-06,
"loss": 0.747,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.2541016628523844,
"learning_rate": 5.434572078689617e-06,
"loss": 0.7378,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.27036222616385347,
"learning_rate": 5.395874317032694e-06,
"loss": 0.747,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.21693744952754704,
"learning_rate": 5.357293329575573e-06,
"loss": 0.7384,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.24777349822787756,
"learning_rate": 5.318829424810797e-06,
"loss": 0.7054,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.2408191617585044,
"learning_rate": 5.280482910294728e-06,
"loss": 0.7363,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.23206094541449224,
"learning_rate": 5.242254092645078e-06,
"loss": 0.7393,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.2369443323518668,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.719,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.24888923646754632,
"learning_rate": 5.166150769707892e-06,
"loss": 0.731,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.2501287420977614,
"learning_rate": 5.128276872940503e-06,
"loss": 0.7403,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.2678858887959947,
"learning_rate": 5.090521890074953e-06,
"loss": 0.755,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.24398645697243812,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.7728,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.23649478125182558,
"learning_rate": 5.015369872647442e-06,
"loss": 0.7402,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.24514314080269475,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.7382,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.2548119870459054,
"learning_rate": 4.940697121074675e-06,
"loss": 0.7499,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.251408714649995,
"learning_rate": 4.903541216934864e-06,
"loss": 0.7376,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.25996066812050067,
"learning_rate": 4.8665060236772e-06,
"loss": 0.7259,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.23469239280052936,
"learning_rate": 4.829591837434102e-06,
"loss": 0.7108,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.2447151853399599,
"learning_rate": 4.792798953370424e-06,
"loss": 0.7468,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.26436451775823,
"learning_rate": 4.756127665681083e-06,
"loss": 0.7471,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.2597251613733564,
"learning_rate": 4.719578267588725e-06,
"loss": 0.7499,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.24470388162619644,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.7295,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.31907574918476955,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.7198,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.2804090061941369,
"learning_rate": 4.610664328486485e-06,
"loss": 0.7199,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.28617210479019817,
"learning_rate": 4.5746054014809e-06,
"loss": 0.7152,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.36890405566399204,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.6912,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.29833059264913636,
"learning_rate": 4.502857857942204e-06,
"loss": 0.7078,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.31255797263681095,
"learning_rate": 4.467169815100549e-06,
"loss": 0.7058,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.3455308543835751,
"learning_rate": 4.431605972355137e-06,
"loss": 0.6842,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.3504862120347948,
"learning_rate": 4.396166614073516e-06,
"loss": 0.6799,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.27800298167271065,
"learning_rate": 4.360852023627846e-06,
"loss": 0.7246,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.2846782092663103,
"learning_rate": 4.325662483392656e-06,
"loss": 0.6961,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.31434677891176377,
"learning_rate": 4.29059827474257e-06,
"loss": 0.6899,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.34617731660637824,
"learning_rate": 4.255659678050063e-06,
"loss": 0.7092,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.3205106600391895,
"learning_rate": 4.22084697268323e-06,
"loss": 0.6982,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.2845337522217869,
"learning_rate": 4.186160437003535e-06,
"loss": 0.7151,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.317254108818369,
"learning_rate": 4.151600348363582e-06,
"loss": 0.6989,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.30595739635964603,
"learning_rate": 4.11716698310493e-06,
"loss": 0.6919,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.2519123565175346,
"learning_rate": 4.082860616555848e-06,
"loss": 0.7064,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.2553131557820587,
"learning_rate": 4.048681523029125e-06,
"loss": 0.701,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.27019882570758597,
"learning_rate": 4.014629975819882e-06,
"loss": 0.7162,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.2506152240535997,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.71,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.28423957658907056,
"learning_rate": 3.94691060843287e-06,
"loss": 0.7131,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.26683416614766825,
"learning_rate": 3.913243329737353e-06,
"loss": 0.7133,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.2755063499155414,
"learning_rate": 3.879704680319503e-06,
"loss": 0.6844,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.2404733280283817,
"learning_rate": 3.846294928353471e-06,
"loss": 0.7221,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.231115220940767,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.6983,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.26777987709980045,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.7011,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.2834555150251287,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.7003,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.2560509320315433,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.7246,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.2393879084469264,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.7184,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.2980137029736681,
"learning_rate": 3.648558150665886e-06,
"loss": 0.701,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.2533423088476217,
"learning_rate": 3.616058102895075e-06,
"loss": 0.7096,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.2245706717190705,
"learning_rate": 3.583689060687079e-06,
"loss": 0.7115,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.24122986666253712,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.7098,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.29554421929342556,
"learning_rate": 3.51934502719796e-06,
"loss": 0.6914,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.24033893601374273,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.7013,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.2217569197014847,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.7028,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.25265520819160964,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.7109,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.2403679464672173,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.709,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.23000522501239426,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.7179,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.25168699677400463,
"learning_rate": 3.32948376099375e-06,
"loss": 0.7004,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.22644380641284356,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.6866,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.22856916995327403,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.7007,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.21680972521409697,
"learning_rate": 3.236349233544438e-06,
"loss": 0.7368,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.2434681053305772,
"learning_rate": 3.205572144561604e-06,
"loss": 0.7147,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.237550739029002,
"learning_rate": 3.174929343376374e-06,
"loss": 0.6912,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.22086364277430923,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.714,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.20890494391629033,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.7012,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.216504125718304,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.6963,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.22312102891404614,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.7044,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.22507290212681455,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.6884,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.22026010818398492,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.7115,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.20642147548653217,
"learning_rate": 2.964210239224696e-06,
"loss": 0.6897,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.22413881633519864,
"learning_rate": 2.934650470499869e-06,
"loss": 0.7014,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.2200814842578358,
"learning_rate": 2.905227155855128e-06,
"loss": 0.7003,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.20420329442809732,
"learning_rate": 2.875940530558519e-06,
"loss": 0.6745,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.22397686180988557,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.7196,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.23622753171270694,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.6962,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.22062510083019624,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.6686,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.21470788612271768,
"learning_rate": 2.760165589921184e-06,
"loss": 0.7177,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.6366998006701049,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.6968,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.2159567752686981,
"learning_rate": 2.703104292151164e-06,
"loss": 0.7059,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.2480144062830303,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.7052,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.253959692667571,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.701,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.22100933974587236,
"learning_rate": 2.618550200289303e-06,
"loss": 0.7009,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.20496302079790985,
"learning_rate": 2.59064316684412e-06,
"loss": 0.7144,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.21161727678239048,
"learning_rate": 2.562875338152264e-06,
"loss": 0.6981,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.22670285733122275,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.7053,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.2270514939793645,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.6898,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.22644008795822607,
"learning_rate": 2.480409295328803e-06,
"loss": 0.7159,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.19948873281047147,
"learning_rate": 2.45320049480108e-06,
"loss": 0.7035,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.22171139086168692,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.7079,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.23570661820357125,
"learning_rate": 2.399204021408421e-06,
"loss": 0.6955,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.23009780927472792,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.6987,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.20096287302951218,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.7056,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.2252621946003949,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.7108,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.2157410634844143,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.7023,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.23045952335756578,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.688,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.20290868127661546,
"learning_rate": 2.24059906434998e-06,
"loss": 0.7103,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.22799458948375137,
"learning_rate": 2.214660694699091e-06,
"loss": 0.7203,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.21820725616626427,
"learning_rate": 2.188864536147355e-06,
"loss": 0.7116,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.21861058927833038,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.6949,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.23118689908316759,
"learning_rate": 2.137699676264171e-06,
"loss": 0.7216,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.20008792782920967,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.7127,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.21382477979268866,
"learning_rate": 2.08710612114857e-06,
"loss": 0.7036,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.21656642027545664,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.6916,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.24246522079719532,
"learning_rate": 2.0370854889762e-06,
"loss": 0.6901,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.23183070534372952,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.712,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.21232703836895964,
"learning_rate": 1.987639379598427e-06,
"loss": 0.7059,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.22679671604601961,
"learning_rate": 1.963132265887919e-06,
"loss": 0.6683,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.2142156996237026,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.7135,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.19711237986649188,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.6923,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.21030636862391816,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.7054,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.20757075608346517,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.6828,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.21243886793447833,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.7076,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.2150011223084312,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.7176,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.21341887088219227,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.689,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.208838235310045,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.7064,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.21323845219650914,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.6851,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.2030685940564405,
"learning_rate": 1.726025117089718e-06,
"loss": 0.7142,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.21154892227118907,
"learning_rate": 1.70311497206598e-06,
"loss": 0.6878,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.21571247419625578,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.715,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.23029475648755093,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.7117,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.22638476216535594,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.6735,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.20896060997988328,
"learning_rate": 1.612939220355272e-06,
"loss": 0.716,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.2024659310392718,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.6907,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.20913226699457285,
"learning_rate": 1.568732761074121e-06,
"loss": 0.7006,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.21237698900831567,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.6987,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.21736396189211,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.7167,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.2154688665575124,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.6999,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.23605334607294726,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.7039,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.20157387277054498,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.706,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.2017428457362195,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.696,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.20305173273280147,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.6927,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.2341871613240586,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.7121,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.2141039770718164,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.7025,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.22111253662849353,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.7132,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.19659839531696544,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.7057,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.20936459932534662,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.7009,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.21002509725738727,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.706,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.2192999176381315,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.7012,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.224578115750644,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.7087,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.21808693509603091,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.704,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.21787912814462262,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.7056,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.19951780887924667,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.704,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.21074596962118394,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.7073,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.20369885764926585,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.7021,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.2054071741178589,
"learning_rate": 1.140415161609505e-06,
"loss": 0.6998,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.21589554640535696,
"learning_rate": 1.121666393356604e-06,
"loss": 0.6869,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.21432882854271,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.6923,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.1936943506882312,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.6837,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.21452113762878225,
"learning_rate": 1.066326386135097e-06,
"loss": 0.7102,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.21113162335626703,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.6868,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.19128163166705645,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.6969,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.2072786036845764,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.7229,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.22142313302620467,
"learning_rate": 9.9465987684501e-07,
"loss": 0.6876,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.21312252633562054,
"learning_rate": 9.771228098856534e-07,
"loss": 0.6992,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.1936170678260532,
"learning_rate": 9.597378493454013e-07,
"loss": 0.6952,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.21958555460409648,
"learning_rate": 9.425051342339353e-07,
"loss": 0.7031,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.21961622101985478,
"learning_rate": 9.254248023435864e-07,
"loss": 0.7061,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.20033295053222527,
"learning_rate": 9.084969902482288e-07,
"loss": 0.6863,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.20507191928642524,
"learning_rate": 8.917218333021993e-07,
"loss": 0.6883,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.19553302796224556,
"learning_rate": 8.750994656392109e-07,
"loss": 0.7122,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.20500769416913242,
"learning_rate": 8.58630020171265e-07,
"loss": 0.733,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.20873966847149106,
"learning_rate": 8.423136285876099e-07,
"loss": 0.7096,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.2140224429751762,
"learning_rate": 8.26150421353682e-07,
"loss": 0.6736,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.1983278243351102,
"learning_rate": 8.101405277100549e-07,
"loss": 0.7131,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.19754387938546555,
"learning_rate": 7.942840756714077e-07,
"loss": 0.7156,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.20265524146769315,
"learning_rate": 7.785811920255093e-07,
"loss": 0.6912,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.19192067759556636,
"learning_rate": 7.630320023322024e-07,
"loss": 0.7146,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.20360382560303347,
"learning_rate": 7.476366309223903e-07,
"loss": 0.7111,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.1964921662081666,
"learning_rate": 7.323952008970537e-07,
"loss": 0.7247,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.20980554245469946,
"learning_rate": 7.173078341262618e-07,
"loss": 0.6868,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.20339456726634006,
"learning_rate": 7.023746512481989e-07,
"loss": 0.7113,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.21570715329779128,
"learning_rate": 6.875957716681902e-07,
"loss": 0.6998,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.19568224000026926,
"learning_rate": 6.729713135577687e-07,
"loss": 0.7352,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.2043182872558746,
"learning_rate": 6.585013938537143e-07,
"loss": 0.7026,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.21634920078480513,
"learning_rate": 6.441861282571115e-07,
"loss": 0.714,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.21657692414184077,
"learning_rate": 6.300256312324471e-07,
"loss": 0.697,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.19588749043720366,
"learning_rate": 6.160200160066776e-07,
"loss": 0.7179,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.19869371030755606,
"learning_rate": 6.021693945683327e-07,
"loss": 0.7047,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.19526400896801824,
"learning_rate": 5.884738776666044e-07,
"loss": 0.7231,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.20106907940360594,
"learning_rate": 5.749335748104878e-07,
"loss": 0.71,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.19229735974543874,
"learning_rate": 5.615485942678778e-07,
"loss": 0.7006,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.18599774797397722,
"learning_rate": 5.483190430647289e-07,
"loss": 0.7226,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.20578187051314692,
"learning_rate": 5.352450269841747e-07,
"loss": 0.6801,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.2088087573567637,
"learning_rate": 5.22326650565701e-07,
"loss": 0.6984,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.2018035940557044,
"learning_rate": 5.095640171043048e-07,
"loss": 0.711,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.203323592877248,
"learning_rate": 4.969572286496594e-07,
"loss": 0.7096,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.2022949428817206,
"learning_rate": 4.845063860053079e-07,
"loss": 0.7176,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.20194524628409674,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.7021,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.20345965802113358,
"learning_rate": 4.600729351261857e-07,
"loss": 0.7015,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.19445462470237715,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.7099,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.192557192900845,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.7087,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.19125730984454603,
"learning_rate": 4.245948007320588e-07,
"loss": 0.699,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.2094391798873512,
"learning_rate": 4.130816799401971e-07,
"loss": 0.6986,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.1946471464470821,
"learning_rate": 4.017251756252827e-07,
"loss": 0.715,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.2023006614145204,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.6994,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.19306603904728595,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.7201,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.2094486813197125,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.6947,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.20006077656187377,
"learning_rate": 3.578671204615347e-07,
"loss": 0.7249,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.19713974483681998,
"learning_rate": 3.472950355547267e-07,
"loss": 0.7079,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.1951867765109339,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.6936,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.22584789856898413,
"learning_rate": 3.266223765343224e-07,
"loss": 0.689,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.20071941815426328,
"learning_rate": 3.16521967718757e-07,
"loss": 0.6969,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.19555013992874523,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.6828,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.2062541303124939,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.6954,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.19673918201640056,
"learning_rate": 2.871653894244597e-07,
"loss": 0.6998,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.19940631512158832,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.7106,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.21319054378397706,
"learning_rate": 2.683823233205485e-07,
"loss": 0.7211,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.1973906530369326,
"learning_rate": 2.592274132703287e-07,
"loss": 0.7046,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.19378929529476308,
"learning_rate": 2.502303499116754e-07,
"loss": 0.7012,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.19593377572108678,
"learning_rate": 2.413912051848666e-07,
"loss": 0.7314,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.19234460996130034,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.7286,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.19718310408939946,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.6888,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.1899012390638469,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.7188,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.19362851712502202,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.6984,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.1964513358791438,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.7083,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.19564258141993734,
"learning_rate": 1.916764975761254e-07,
"loss": 0.694,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.21014366535753617,
"learning_rate": 1.839446914621923e-07,
"loss": 0.7117,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.2144437868612429,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.6983,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.1921259125782749,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.7088,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.2127261802538754,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.7019,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.19368066742724574,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.6991,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.1977334513617203,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.6882,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.19173155098594152,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.7032,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.18470729949256529,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.7175,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.1964349755234406,
"learning_rate": 1.277993209368744e-07,
"loss": 0.6985,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.1898504892010915,
"learning_rate": 1.214956261808542e-07,
"loss": 0.7165,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.19972426460230702,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.7005,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.18882231149789988,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.7007,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.199637048837968,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.7277,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.2040655585536339,
"learning_rate": 9.787081204466253e-08,
"loss": 0.6987,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.21825420634357845,
"learning_rate": 9.236233586200627e-08,
"loss": 0.6981,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.2060547146774556,
"learning_rate": 8.70130406196279e-08,
"loss": 0.698,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.20263719719905898,
"learning_rate": 8.182296909035047e-08,
"loss": 0.6808,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.19434861818940008,
"learning_rate": 7.679216277384439e-08,
"loss": 0.7032,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.19494070696366567,
"learning_rate": 7.19206618963053e-08,
"loss": 0.7141,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.1978916576301186,
"learning_rate": 6.720850541012347e-08,
"loss": 0.7163,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.19470157997172866,
"learning_rate": 6.265573099357714e-08,
"loss": 0.705,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.19214138488551308,
"learning_rate": 5.826237505053067e-08,
"loss": 0.7063,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.2003780426279055,
"learning_rate": 5.4028472710137e-08,
"loss": 0.6925,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.19833581135409106,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.7071,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.19544257388705982,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.7309,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.19278104325959658,
"learning_rate": 4.228381947004812e-08,
"loss": 0.7164,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.19296519526671044,
"learning_rate": 3.868805732811343e-08,
"loss": 0.701,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.19050596114182328,
"learning_rate": 3.525190530455636e-08,
"loss": 0.7051,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.2024480918960877,
"learning_rate": 3.197539087475887e-08,
"loss": 0.69,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.19155848542172454,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.7162,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.19176641222547594,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.7047,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.19349426795141292,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.7248,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.19027509347312146,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.7089,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.19050879922532982,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.7093,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.19983460660619382,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.7164,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.19107682156852562,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.6921,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.18902893862309478,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.7007,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.19167855277746612,
"learning_rate": 9.67435416023843e-09,
"loss": 0.6897,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.18938187524734473,
"learning_rate": 7.995445897566178e-09,
"loss": 0.6788,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.20227948009338748,
"learning_rate": 6.476393170578021e-09,
"loss": 0.7236,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.20569968357117382,
"learning_rate": 5.11720812557348e-09,
"loss": 0.7225,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.41527235366061005,
"learning_rate": 3.917901630563492e-09,
"loss": 0.7127,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.19984726914989984,
"learning_rate": 2.878483275170485e-09,
"loss": 0.7136,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.19682153215292258,
"learning_rate": 1.998961370555108e-09,
"loss": 0.7018,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.19725451946154565,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.6963,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.19607826359960237,
"learning_rate": 7.196337656112384e-10,
"loss": 0.6771,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.19095602355984856,
"learning_rate": 3.198382947511292e-10,
"loss": 0.7141,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.19497526538332352,
"learning_rate": 7.995973352770137e-11,
"loss": 0.7166,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.19411296069480166,
"learning_rate": 0.0,
"loss": 0.7189,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 5.937242625395393e+18,
"train_loss": 0.8014366944309188,
"train_runtime": 28785.4861,
"train_samples_per_second": 5.489,
"train_steps_per_second": 0.043
}
],
"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": 5.937242625395393e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}