| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 240, | |
| "global_step": 958, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0010438413361169101, | |
| "grad_norm": 1.4100897312164307, | |
| "learning_rate": 2e-05, | |
| "loss": 1.0427, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0010438413361169101, | |
| "eval_loss": 1.9999854564666748, | |
| "eval_runtime": 81.2923, | |
| "eval_samples_per_second": 19.854, | |
| "eval_steps_per_second": 9.927, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0020876826722338203, | |
| "grad_norm": 2.131267547607422, | |
| "learning_rate": 4e-05, | |
| "loss": 1.2661, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.003131524008350731, | |
| "grad_norm": 2.1016671657562256, | |
| "learning_rate": 6e-05, | |
| "loss": 1.0965, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0041753653444676405, | |
| "grad_norm": 2.6869776248931885, | |
| "learning_rate": 8e-05, | |
| "loss": 1.1197, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.005219206680584551, | |
| "grad_norm": 2.6327548027038574, | |
| "learning_rate": 0.0001, | |
| "loss": 0.9191, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.006263048016701462, | |
| "grad_norm": 1.9485589265823364, | |
| "learning_rate": 0.00012, | |
| "loss": 0.7709, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.007306889352818371, | |
| "grad_norm": 1.7759828567504883, | |
| "learning_rate": 0.00014, | |
| "loss": 0.6302, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.008350730688935281, | |
| "grad_norm": 1.4292073249816895, | |
| "learning_rate": 0.00016, | |
| "loss": 0.604, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.009394572025052192, | |
| "grad_norm": 1.8719030618667603, | |
| "learning_rate": 0.00018, | |
| "loss": 0.5034, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.010438413361169102, | |
| "grad_norm": 2.183128833770752, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5065, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.011482254697286013, | |
| "grad_norm": 2.187234401702881, | |
| "learning_rate": 0.00019999945089843994, | |
| "loss": 0.4966, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.012526096033402923, | |
| "grad_norm": 2.7468483448028564, | |
| "learning_rate": 0.00019999780359979, | |
| "loss": 0.3204, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.013569937369519834, | |
| "grad_norm": 2.2195403575897217, | |
| "learning_rate": 0.00019999505812214085, | |
| "loss": 0.2784, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.014613778705636743, | |
| "grad_norm": 2.0797836780548096, | |
| "learning_rate": 0.00019999121449564347, | |
| "loss": 0.3631, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.015657620041753653, | |
| "grad_norm": 2.080904245376587, | |
| "learning_rate": 0.00019998627276250858, | |
| "loss": 0.2425, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.016701461377870562, | |
| "grad_norm": 1.358616590499878, | |
| "learning_rate": 0.00019998023297700658, | |
| "loss": 0.2799, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.017745302713987474, | |
| "grad_norm": 1.4436254501342773, | |
| "learning_rate": 0.00019997309520546647, | |
| "loss": 0.3314, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.018789144050104383, | |
| "grad_norm": 1.1498231887817383, | |
| "learning_rate": 0.00019996485952627552, | |
| "loss": 0.2713, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.019832985386221295, | |
| "grad_norm": 1.3239034414291382, | |
| "learning_rate": 0.00019995552602987827, | |
| "loss": 0.2347, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.020876826722338204, | |
| "grad_norm": 1.1880042552947998, | |
| "learning_rate": 0.00019994509481877537, | |
| "loss": 0.3311, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.021920668058455117, | |
| "grad_norm": 1.1707823276519775, | |
| "learning_rate": 0.00019993356600752276, | |
| "loss": 0.4601, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.022964509394572025, | |
| "grad_norm": 1.6230024099349976, | |
| "learning_rate": 0.00019992093972273018, | |
| "loss": 0.4187, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.024008350730688934, | |
| "grad_norm": 1.6155649423599243, | |
| "learning_rate": 0.00019990721610305996, | |
| "loss": 0.4661, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.025052192066805846, | |
| "grad_norm": 1.4404484033584595, | |
| "learning_rate": 0.0001998923952992252, | |
| "loss": 0.2662, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.026096033402922755, | |
| "grad_norm": 1.0615559816360474, | |
| "learning_rate": 0.00019987647747398852, | |
| "loss": 0.3144, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.027139874739039668, | |
| "grad_norm": 2.5796847343444824, | |
| "learning_rate": 0.00019985946280215994, | |
| "loss": 0.9688, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.028183716075156576, | |
| "grad_norm": 3.5270519256591797, | |
| "learning_rate": 0.00019984135147059514, | |
| "loss": 1.1566, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.029227557411273485, | |
| "grad_norm": 3.404811143875122, | |
| "learning_rate": 0.00019982214367819328, | |
| "loss": 0.9788, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.030271398747390398, | |
| "grad_norm": 2.5648179054260254, | |
| "learning_rate": 0.00019980183963589504, | |
| "loss": 0.9095, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.031315240083507306, | |
| "grad_norm": 2.203162670135498, | |
| "learning_rate": 0.0001997804395666799, | |
| "loss": 0.6835, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.032359081419624215, | |
| "grad_norm": 2.2330455780029297, | |
| "learning_rate": 0.00019975794370556417, | |
| "loss": 0.7176, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.033402922755741124, | |
| "grad_norm": 1.7634390592575073, | |
| "learning_rate": 0.00019973435229959813, | |
| "loss": 0.6717, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.03444676409185804, | |
| "grad_norm": 1.7378387451171875, | |
| "learning_rate": 0.00019970966560786324, | |
| "loss": 0.6573, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.03549060542797495, | |
| "grad_norm": 1.515835165977478, | |
| "learning_rate": 0.0001996838839014696, | |
| "loss": 0.5787, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.03653444676409186, | |
| "grad_norm": 1.6372636556625366, | |
| "learning_rate": 0.0001996570074635527, | |
| "loss": 0.5711, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.037578288100208766, | |
| "grad_norm": 1.449986457824707, | |
| "learning_rate": 0.00019962903658927037, | |
| "loss": 0.4957, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.038622129436325675, | |
| "grad_norm": 1.544659972190857, | |
| "learning_rate": 0.00019959997158579967, | |
| "loss": 0.5339, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.03966597077244259, | |
| "grad_norm": 1.497573733329773, | |
| "learning_rate": 0.0001995698127723334, | |
| "loss": 0.3736, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0407098121085595, | |
| "grad_norm": 1.7716635465621948, | |
| "learning_rate": 0.00019953856048007652, | |
| "loss": 0.5047, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.04175365344467641, | |
| "grad_norm": 1.439935326576233, | |
| "learning_rate": 0.00019950621505224273, | |
| "loss": 0.3242, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04279749478079332, | |
| "grad_norm": 1.5024681091308594, | |
| "learning_rate": 0.00019947277684405056, | |
| "loss": 0.3943, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.04384133611691023, | |
| "grad_norm": 1.411978006362915, | |
| "learning_rate": 0.00019943824622271935, | |
| "loss": 0.2535, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.04488517745302714, | |
| "grad_norm": 1.6485614776611328, | |
| "learning_rate": 0.00019940262356746554, | |
| "loss": 0.3503, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.04592901878914405, | |
| "grad_norm": 2.0574934482574463, | |
| "learning_rate": 0.0001993659092694982, | |
| "loss": 0.3788, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.04697286012526096, | |
| "grad_norm": 1.5055749416351318, | |
| "learning_rate": 0.00019932810373201495, | |
| "loss": 0.285, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.04801670146137787, | |
| "grad_norm": 1.3977243900299072, | |
| "learning_rate": 0.00019928920737019733, | |
| "loss": 0.2997, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.049060542797494784, | |
| "grad_norm": 1.5702697038650513, | |
| "learning_rate": 0.00019924922061120644, | |
| "loss": 0.2235, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.05010438413361169, | |
| "grad_norm": 1.1728659868240356, | |
| "learning_rate": 0.0001992081438941781, | |
| "loss": 0.277, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.0511482254697286, | |
| "grad_norm": 1.6709052324295044, | |
| "learning_rate": 0.00019916597767021807, | |
| "loss": 0.2712, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.05219206680584551, | |
| "grad_norm": 2.5871024131774902, | |
| "learning_rate": 0.00019912272240239716, | |
| "loss": 0.3964, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05323590814196242, | |
| "grad_norm": 2.782623291015625, | |
| "learning_rate": 0.00019907837856574607, | |
| "loss": 0.7383, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.054279749478079335, | |
| "grad_norm": 2.9450387954711914, | |
| "learning_rate": 0.0001990329466472502, | |
| "loss": 0.7801, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.055323590814196244, | |
| "grad_norm": 2.3782992362976074, | |
| "learning_rate": 0.00019898642714584428, | |
| "loss": 0.6978, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.05636743215031315, | |
| "grad_norm": 2.7798867225646973, | |
| "learning_rate": 0.000198938820572407, | |
| "loss": 0.6635, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.05741127348643006, | |
| "grad_norm": 1.2441174983978271, | |
| "learning_rate": 0.00019889012744975508, | |
| "loss": 0.3866, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.05845511482254697, | |
| "grad_norm": 1.9317642450332642, | |
| "learning_rate": 0.00019884034831263808, | |
| "loss": 0.5378, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.059498956158663886, | |
| "grad_norm": 1.48700749874115, | |
| "learning_rate": 0.00019878948370773193, | |
| "loss": 0.416, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.060542797494780795, | |
| "grad_norm": 1.6028392314910889, | |
| "learning_rate": 0.00019873753419363336, | |
| "loss": 0.4354, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.061586638830897704, | |
| "grad_norm": 1.2003154754638672, | |
| "learning_rate": 0.00019868450034085352, | |
| "loss": 0.289, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.06263048016701461, | |
| "grad_norm": 0.8975642919540405, | |
| "learning_rate": 0.00019863038273181186, | |
| "loss": 0.3655, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06367432150313153, | |
| "grad_norm": 1.0003724098205566, | |
| "learning_rate": 0.00019857518196082964, | |
| "loss": 0.3517, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.06471816283924843, | |
| "grad_norm": 0.7653176784515381, | |
| "learning_rate": 0.00019851889863412345, | |
| "loss": 0.2623, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.06576200417536535, | |
| "grad_norm": 0.8752651810646057, | |
| "learning_rate": 0.00019846153336979856, | |
| "loss": 0.295, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.06680584551148225, | |
| "grad_norm": 0.8633437752723694, | |
| "learning_rate": 0.00019840308679784207, | |
| "loss": 0.198, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.06784968684759916, | |
| "grad_norm": 0.9436039924621582, | |
| "learning_rate": 0.00019834355956011606, | |
| "loss": 0.2345, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.06889352818371608, | |
| "grad_norm": 0.6003718376159668, | |
| "learning_rate": 0.00019828295231035051, | |
| "loss": 0.1637, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.06993736951983298, | |
| "grad_norm": 1.460316777229309, | |
| "learning_rate": 0.00019822126571413616, | |
| "loss": 0.2315, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.0709812108559499, | |
| "grad_norm": 0.840602457523346, | |
| "learning_rate": 0.00019815850044891707, | |
| "loss": 0.2587, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.0720250521920668, | |
| "grad_norm": 0.9615221619606018, | |
| "learning_rate": 0.0001980946572039834, | |
| "loss": 0.3806, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.07306889352818371, | |
| "grad_norm": 1.013310432434082, | |
| "learning_rate": 0.00019802973668046363, | |
| "loss": 0.2961, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07411273486430063, | |
| "grad_norm": 0.9203840494155884, | |
| "learning_rate": 0.00019796373959131698, | |
| "loss": 0.2715, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.07515657620041753, | |
| "grad_norm": 0.8724762201309204, | |
| "learning_rate": 0.00019789666666132554, | |
| "loss": 0.2774, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.07620041753653445, | |
| "grad_norm": 0.854420006275177, | |
| "learning_rate": 0.00019782851862708634, | |
| "loss": 0.2466, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.07724425887265135, | |
| "grad_norm": 0.880692720413208, | |
| "learning_rate": 0.00019775929623700318, | |
| "loss": 0.3145, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.07828810020876827, | |
| "grad_norm": 1.4334697723388672, | |
| "learning_rate": 0.00019768900025127851, | |
| "loss": 0.6141, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.07933194154488518, | |
| "grad_norm": 1.6129218339920044, | |
| "learning_rate": 0.0001976176314419051, | |
| "loss": 0.6577, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.08037578288100208, | |
| "grad_norm": 1.4365684986114502, | |
| "learning_rate": 0.00019754519059265736, | |
| "loss": 0.6501, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.081419624217119, | |
| "grad_norm": 1.1411486864089966, | |
| "learning_rate": 0.00019747167849908304, | |
| "loss": 0.5629, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.0824634655532359, | |
| "grad_norm": 0.9845925569534302, | |
| "learning_rate": 0.00019739709596849417, | |
| "loss": 0.4142, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.08350730688935282, | |
| "grad_norm": 1.0630574226379395, | |
| "learning_rate": 0.00019732144381995846, | |
| "loss": 0.4291, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08455114822546973, | |
| "grad_norm": 1.059112548828125, | |
| "learning_rate": 0.0001972447228842902, | |
| "loss": 0.4116, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.08559498956158663, | |
| "grad_norm": 0.99774569272995, | |
| "learning_rate": 0.000197166934004041, | |
| "loss": 0.423, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.08663883089770355, | |
| "grad_norm": 1.2038483619689941, | |
| "learning_rate": 0.00019708807803349088, | |
| "loss": 0.4696, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.08768267223382047, | |
| "grad_norm": 1.5743309259414673, | |
| "learning_rate": 0.00019700815583863852, | |
| "loss": 0.4688, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.08872651356993737, | |
| "grad_norm": 1.0845067501068115, | |
| "learning_rate": 0.00019692716829719194, | |
| "loss": 0.3629, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.08977035490605428, | |
| "grad_norm": 1.1626365184783936, | |
| "learning_rate": 0.00019684511629855888, | |
| "loss": 0.4327, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.09081419624217119, | |
| "grad_norm": 1.1431448459625244, | |
| "learning_rate": 0.00019676200074383692, | |
| "loss": 0.2941, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.0918580375782881, | |
| "grad_norm": 1.0742226839065552, | |
| "learning_rate": 0.00019667782254580374, | |
| "loss": 0.365, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.09290187891440502, | |
| "grad_norm": 1.0580729246139526, | |
| "learning_rate": 0.00019659258262890683, | |
| "loss": 0.351, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.09394572025052192, | |
| "grad_norm": 1.1420572996139526, | |
| "learning_rate": 0.0001965062819292537, | |
| "loss": 0.3829, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09498956158663883, | |
| "grad_norm": 0.9046992063522339, | |
| "learning_rate": 0.0001964189213946013, | |
| "loss": 0.3192, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.09603340292275574, | |
| "grad_norm": 0.94282066822052, | |
| "learning_rate": 0.00019633050198434576, | |
| "loss": 0.251, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.09707724425887265, | |
| "grad_norm": 0.9391703009605408, | |
| "learning_rate": 0.0001962410246695118, | |
| "loss": 0.2545, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.09812108559498957, | |
| "grad_norm": 1.068588137626648, | |
| "learning_rate": 0.00019615049043274205, | |
| "loss": 0.2889, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.09916492693110647, | |
| "grad_norm": 0.8007175922393799, | |
| "learning_rate": 0.00019605890026828634, | |
| "loss": 0.1906, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10020876826722339, | |
| "grad_norm": 1.195952296257019, | |
| "learning_rate": 0.00019596625518199077, | |
| "loss": 0.3045, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.10125260960334029, | |
| "grad_norm": 1.1844594478607178, | |
| "learning_rate": 0.00019587255619128648, | |
| "loss": 0.2509, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.1022964509394572, | |
| "grad_norm": 1.0775648355484009, | |
| "learning_rate": 0.00019577780432517879, | |
| "loss": 0.1594, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.10334029227557412, | |
| "grad_norm": 1.0487576723098755, | |
| "learning_rate": 0.00019568200062423555, | |
| "loss": 0.1619, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.10438413361169102, | |
| "grad_norm": 3.752366304397583, | |
| "learning_rate": 0.00019558514614057609, | |
| "loss": 0.4774, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10542797494780794, | |
| "grad_norm": 1.1260536909103394, | |
| "learning_rate": 0.00019548724193785933, | |
| "loss": 0.4538, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.10647181628392484, | |
| "grad_norm": 1.2162113189697266, | |
| "learning_rate": 0.0001953882890912723, | |
| "loss": 0.4851, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.10751565762004175, | |
| "grad_norm": 1.2823134660720825, | |
| "learning_rate": 0.00019528828868751818, | |
| "loss": 0.5302, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.10855949895615867, | |
| "grad_norm": 1.2430353164672852, | |
| "learning_rate": 0.0001951872418248046, | |
| "loss": 0.4463, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.10960334029227557, | |
| "grad_norm": 0.8766555786132812, | |
| "learning_rate": 0.00019508514961283138, | |
| "loss": 0.3673, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.11064718162839249, | |
| "grad_norm": 1.0745108127593994, | |
| "learning_rate": 0.00019498201317277828, | |
| "loss": 0.4313, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.11169102296450939, | |
| "grad_norm": 0.904514729976654, | |
| "learning_rate": 0.00019487783363729294, | |
| "loss": 0.3183, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.1127348643006263, | |
| "grad_norm": 1.1046079397201538, | |
| "learning_rate": 0.00019477261215047835, | |
| "loss": 0.389, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.11377870563674322, | |
| "grad_norm": 0.9722571969032288, | |
| "learning_rate": 0.00019466634986788005, | |
| "loss": 0.3647, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.11482254697286012, | |
| "grad_norm": 0.8949522376060486, | |
| "learning_rate": 0.0001945590479564738, | |
| "loss": 0.3807, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.11586638830897704, | |
| "grad_norm": 0.8489611744880676, | |
| "learning_rate": 0.00019445070759465253, | |
| "loss": 0.3148, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.11691022964509394, | |
| "grad_norm": 0.8171373009681702, | |
| "learning_rate": 0.00019434132997221345, | |
| "loss": 0.2941, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.11795407098121086, | |
| "grad_norm": 0.8785704970359802, | |
| "learning_rate": 0.00019423091629034507, | |
| "loss": 0.3087, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.11899791231732777, | |
| "grad_norm": 0.8450028896331787, | |
| "learning_rate": 0.00019411946776161387, | |
| "loss": 0.2428, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.12004175365344467, | |
| "grad_norm": 0.7037304639816284, | |
| "learning_rate": 0.00019400698560995103, | |
| "loss": 0.1534, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.12108559498956159, | |
| "grad_norm": 0.8595078587532043, | |
| "learning_rate": 0.00019389347107063912, | |
| "loss": 0.1996, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.12212943632567849, | |
| "grad_norm": 0.7243440747261047, | |
| "learning_rate": 0.00019377892539029827, | |
| "loss": 0.1407, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.12317327766179541, | |
| "grad_norm": 0.7662886381149292, | |
| "learning_rate": 0.0001936633498268728, | |
| "loss": 0.187, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.12421711899791232, | |
| "grad_norm": 0.9693275094032288, | |
| "learning_rate": 0.0001935467456496171, | |
| "loss": 0.2622, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.12526096033402923, | |
| "grad_norm": 0.827397882938385, | |
| "learning_rate": 0.0001934291141390819, | |
| "loss": 0.2819, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.12630480167014613, | |
| "grad_norm": 0.7069506645202637, | |
| "learning_rate": 0.0001933104565871001, | |
| "loss": 0.2907, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.12734864300626306, | |
| "grad_norm": 0.7466704845428467, | |
| "learning_rate": 0.00019319077429677268, | |
| "loss": 0.209, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.12839248434237996, | |
| "grad_norm": 0.8960225582122803, | |
| "learning_rate": 0.00019307006858245424, | |
| "loss": 0.219, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.12943632567849686, | |
| "grad_norm": 0.7968412041664124, | |
| "learning_rate": 0.0001929483407697387, | |
| "loss": 0.292, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.1304801670146138, | |
| "grad_norm": 0.7965363264083862, | |
| "learning_rate": 0.00019282559219544477, | |
| "loss": 0.3051, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1315240083507307, | |
| "grad_norm": 1.490718126296997, | |
| "learning_rate": 0.00019270182420760102, | |
| "loss": 0.485, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.1325678496868476, | |
| "grad_norm": 1.431773066520691, | |
| "learning_rate": 0.00019257703816543144, | |
| "loss": 0.4818, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.1336116910229645, | |
| "grad_norm": 1.4318815469741821, | |
| "learning_rate": 0.00019245123543934017, | |
| "loss": 0.5472, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.13465553235908143, | |
| "grad_norm": 1.2134075164794922, | |
| "learning_rate": 0.00019232441741089676, | |
| "loss": 0.4773, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.13569937369519833, | |
| "grad_norm": 1.110347032546997, | |
| "learning_rate": 0.00019219658547282067, | |
| "loss": 0.4516, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.13674321503131523, | |
| "grad_norm": 1.122799277305603, | |
| "learning_rate": 0.00019206774102896627, | |
| "loss": 0.5222, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.13778705636743216, | |
| "grad_norm": 0.9539543390274048, | |
| "learning_rate": 0.00019193788549430724, | |
| "loss": 0.3994, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.13883089770354906, | |
| "grad_norm": 0.9659099578857422, | |
| "learning_rate": 0.00019180702029492118, | |
| "loss": 0.3946, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.13987473903966596, | |
| "grad_norm": 1.0640462636947632, | |
| "learning_rate": 0.00019167514686797369, | |
| "loss": 0.4409, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.1409185803757829, | |
| "grad_norm": 1.1047799587249756, | |
| "learning_rate": 0.00019154226666170295, | |
| "loss": 0.4123, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.1419624217118998, | |
| "grad_norm": 0.9255096912384033, | |
| "learning_rate": 0.00019140838113540346, | |
| "loss": 0.2524, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.1430062630480167, | |
| "grad_norm": 0.9960671663284302, | |
| "learning_rate": 0.00019127349175941032, | |
| "loss": 0.3268, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.1440501043841336, | |
| "grad_norm": 1.0808758735656738, | |
| "learning_rate": 0.0001911376000150828, | |
| "loss": 0.4802, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.14509394572025053, | |
| "grad_norm": 0.9004728198051453, | |
| "learning_rate": 0.00019100070739478832, | |
| "loss": 0.3092, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.14613778705636743, | |
| "grad_norm": 0.9561731219291687, | |
| "learning_rate": 0.00019086281540188588, | |
| "loss": 0.2572, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.14718162839248433, | |
| "grad_norm": 1.1557111740112305, | |
| "learning_rate": 0.00019072392555070965, | |
| "loss": 0.3391, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.14822546972860126, | |
| "grad_norm": 1.0732665061950684, | |
| "learning_rate": 0.00019058403936655233, | |
| "loss": 0.2717, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.14926931106471816, | |
| "grad_norm": 1.2010210752487183, | |
| "learning_rate": 0.00019044315838564834, | |
| "loss": 0.3122, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.15031315240083507, | |
| "grad_norm": 1.0502979755401611, | |
| "learning_rate": 0.000190301284155157, | |
| "loss": 0.2789, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.151356993736952, | |
| "grad_norm": 0.835246205329895, | |
| "learning_rate": 0.0001901584182331456, | |
| "loss": 0.1788, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1524008350730689, | |
| "grad_norm": 0.9552949666976929, | |
| "learning_rate": 0.00019001456218857208, | |
| "loss": 0.2785, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.1534446764091858, | |
| "grad_norm": 0.6930809617042542, | |
| "learning_rate": 0.00018986971760126805, | |
| "loss": 0.1823, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.1544885177453027, | |
| "grad_norm": 0.6930037140846252, | |
| "learning_rate": 0.00018972388606192125, | |
| "loss": 0.1895, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.15553235908141963, | |
| "grad_norm": 0.8468357920646667, | |
| "learning_rate": 0.0001895770691720582, | |
| "loss": 0.1717, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.15657620041753653, | |
| "grad_norm": 0.9056650996208191, | |
| "learning_rate": 0.0001894292685440266, | |
| "loss": 0.1854, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.15762004175365343, | |
| "grad_norm": 1.131759524345398, | |
| "learning_rate": 0.00018928048580097757, | |
| "loss": 0.441, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.15866388308977036, | |
| "grad_norm": 1.426763892173767, | |
| "learning_rate": 0.00018913072257684778, | |
| "loss": 0.5142, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.15970772442588727, | |
| "grad_norm": 1.4198706150054932, | |
| "learning_rate": 0.00018897998051634166, | |
| "loss": 0.5212, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.16075156576200417, | |
| "grad_norm": 1.228197693824768, | |
| "learning_rate": 0.0001888282612749132, | |
| "loss": 0.4639, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.1617954070981211, | |
| "grad_norm": 0.7881720662117004, | |
| "learning_rate": 0.0001886755665187479, | |
| "loss": 0.3905, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.162839248434238, | |
| "grad_norm": 0.7964795231819153, | |
| "learning_rate": 0.00018852189792474425, | |
| "loss": 0.3802, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.1638830897703549, | |
| "grad_norm": 1.1133593320846558, | |
| "learning_rate": 0.00018836725718049562, | |
| "loss": 0.3962, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.1649269311064718, | |
| "grad_norm": 1.182551383972168, | |
| "learning_rate": 0.00018821164598427145, | |
| "loss": 0.3305, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.16597077244258873, | |
| "grad_norm": 0.8984887003898621, | |
| "learning_rate": 0.0001880550660449988, | |
| "loss": 0.2847, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.16701461377870563, | |
| "grad_norm": 0.8994187116622925, | |
| "learning_rate": 0.00018789751908224338, | |
| "loss": 0.346, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.16805845511482254, | |
| "grad_norm": 0.7379996180534363, | |
| "learning_rate": 0.0001877390068261909, | |
| "loss": 0.3183, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.16910229645093947, | |
| "grad_norm": 0.719083845615387, | |
| "learning_rate": 0.00018757953101762787, | |
| "loss": 0.2136, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.17014613778705637, | |
| "grad_norm": 1.2782928943634033, | |
| "learning_rate": 0.00018741909340792262, | |
| "loss": 0.1879, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.17118997912317327, | |
| "grad_norm": 0.6968748569488525, | |
| "learning_rate": 0.000187257695759006, | |
| "loss": 0.1594, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.1722338204592902, | |
| "grad_norm": 0.8432148098945618, | |
| "learning_rate": 0.00018709533984335192, | |
| "loss": 0.1767, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.1732776617954071, | |
| "grad_norm": 0.6877685785293579, | |
| "learning_rate": 0.00018693202744395827, | |
| "loss": 0.1787, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.174321503131524, | |
| "grad_norm": 0.9011303782463074, | |
| "learning_rate": 0.0001867677603543268, | |
| "loss": 0.1692, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.17536534446764093, | |
| "grad_norm": 0.8391153216362, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 0.2701, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.17640918580375783, | |
| "grad_norm": 0.7215454578399658, | |
| "learning_rate": 0.0001864363693307604, | |
| "loss": 0.2727, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.17745302713987474, | |
| "grad_norm": 0.9198095202445984, | |
| "learning_rate": 0.000186269249036172, | |
| "loss": 0.2709, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.17849686847599164, | |
| "grad_norm": 0.6714828014373779, | |
| "learning_rate": 0.00018610118132999896, | |
| "loss": 0.2343, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.17954070981210857, | |
| "grad_norm": 0.7863060235977173, | |
| "learning_rate": 0.00018593216805796612, | |
| "loss": 0.3066, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.18058455114822547, | |
| "grad_norm": 0.8642680644989014, | |
| "learning_rate": 0.00018576221107618243, | |
| "loss": 0.3599, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.18162839248434237, | |
| "grad_norm": 1.3609846830368042, | |
| "learning_rate": 0.00018559131225112085, | |
| "loss": 0.435, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.1826722338204593, | |
| "grad_norm": 1.3618521690368652, | |
| "learning_rate": 0.00018541947345959754, | |
| "loss": 0.5144, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1837160751565762, | |
| "grad_norm": 1.10430908203125, | |
| "learning_rate": 0.00018524669658875152, | |
| "loss": 0.465, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.1847599164926931, | |
| "grad_norm": 0.8763125538825989, | |
| "learning_rate": 0.00018507298353602375, | |
| "loss": 0.4001, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.18580375782881003, | |
| "grad_norm": 0.979162335395813, | |
| "learning_rate": 0.00018489833620913642, | |
| "loss": 0.4281, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.18684759916492694, | |
| "grad_norm": 0.9191274642944336, | |
| "learning_rate": 0.00018472275652607186, | |
| "loss": 0.4362, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.18789144050104384, | |
| "grad_norm": 1.0854690074920654, | |
| "learning_rate": 0.0001845462464150517, | |
| "loss": 0.4294, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.18893528183716074, | |
| "grad_norm": 1.0451358556747437, | |
| "learning_rate": 0.00018436880781451544, | |
| "loss": 0.4065, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.18997912317327767, | |
| "grad_norm": 1.0197910070419312, | |
| "learning_rate": 0.00018419044267309939, | |
| "loss": 0.3585, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.19102296450939457, | |
| "grad_norm": 0.9235432744026184, | |
| "learning_rate": 0.000184011152949615, | |
| "loss": 0.4618, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.19206680584551147, | |
| "grad_norm": 0.9856287837028503, | |
| "learning_rate": 0.00018383094061302766, | |
| "loss": 0.3951, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.1931106471816284, | |
| "grad_norm": 0.851546585559845, | |
| "learning_rate": 0.0001836498076424349, | |
| "loss": 0.3907, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.1941544885177453, | |
| "grad_norm": 0.936446487903595, | |
| "learning_rate": 0.00018346775602704464, | |
| "loss": 0.3234, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.1951983298538622, | |
| "grad_norm": 0.8742521405220032, | |
| "learning_rate": 0.00018328478776615334, | |
| "loss": 0.3065, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.19624217118997914, | |
| "grad_norm": 0.6365717053413391, | |
| "learning_rate": 0.00018310090486912424, | |
| "loss": 0.1979, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.19728601252609604, | |
| "grad_norm": 0.7701941132545471, | |
| "learning_rate": 0.000182916109355365, | |
| "loss": 0.2322, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.19832985386221294, | |
| "grad_norm": 0.8491252064704895, | |
| "learning_rate": 0.00018273040325430574, | |
| "loss": 0.2477, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.19937369519832984, | |
| "grad_norm": 0.8160005807876587, | |
| "learning_rate": 0.00018254378860537666, | |
| "loss": 0.2554, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.20041753653444677, | |
| "grad_norm": 0.7356383800506592, | |
| "learning_rate": 0.0001823562674579856, | |
| "loss": 0.2209, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.20146137787056367, | |
| "grad_norm": 0.734779953956604, | |
| "learning_rate": 0.0001821678418714957, | |
| "loss": 0.2162, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.20250521920668058, | |
| "grad_norm": 0.8594037294387817, | |
| "learning_rate": 0.00018197851391520264, | |
| "loss": 0.2831, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.2035490605427975, | |
| "grad_norm": 0.6627790331840515, | |
| "learning_rate": 0.00018178828566831192, | |
| "loss": 0.1818, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2045929018789144, | |
| "grad_norm": 0.9464810490608215, | |
| "learning_rate": 0.00018159715921991612, | |
| "loss": 0.2376, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.2056367432150313, | |
| "grad_norm": 0.8601820468902588, | |
| "learning_rate": 0.00018140513666897178, | |
| "loss": 0.1824, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.20668058455114824, | |
| "grad_norm": 0.7504302263259888, | |
| "learning_rate": 0.00018121222012427665, | |
| "loss": 0.2081, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.20772442588726514, | |
| "grad_norm": 0.6832506060600281, | |
| "learning_rate": 0.00018101841170444616, | |
| "loss": 0.1742, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.20876826722338204, | |
| "grad_norm": 1.3405510187149048, | |
| "learning_rate": 0.00018082371353789046, | |
| "loss": 0.2519, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.20981210855949894, | |
| "grad_norm": 0.8375347852706909, | |
| "learning_rate": 0.00018062812776279086, | |
| "loss": 0.3868, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.21085594989561587, | |
| "grad_norm": 0.9397875070571899, | |
| "learning_rate": 0.00018043165652707649, | |
| "loss": 0.4824, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.21189979123173278, | |
| "grad_norm": 0.9745060205459595, | |
| "learning_rate": 0.00018023430198840058, | |
| "loss": 0.4726, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.21294363256784968, | |
| "grad_norm": 0.9175118207931519, | |
| "learning_rate": 0.00018003606631411678, | |
| "loss": 0.3722, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.2139874739039666, | |
| "grad_norm": 0.831420361995697, | |
| "learning_rate": 0.0001798369516812555, | |
| "loss": 0.387, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.2150313152400835, | |
| "grad_norm": 0.7558404207229614, | |
| "learning_rate": 0.00017963696027649986, | |
| "loss": 0.3522, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.2160751565762004, | |
| "grad_norm": 0.8819952607154846, | |
| "learning_rate": 0.0001794360942961617, | |
| "loss": 0.3665, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.21711899791231734, | |
| "grad_norm": 0.9681791663169861, | |
| "learning_rate": 0.00017923435594615744, | |
| "loss": 0.3571, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.21816283924843424, | |
| "grad_norm": 1.2287527322769165, | |
| "learning_rate": 0.000179031747441984, | |
| "loss": 0.269, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.21920668058455114, | |
| "grad_norm": 0.9387004375457764, | |
| "learning_rate": 0.0001788282710086942, | |
| "loss": 0.2254, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.22025052192066805, | |
| "grad_norm": 0.6556700468063354, | |
| "learning_rate": 0.0001786239288808727, | |
| "loss": 0.2438, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.22129436325678498, | |
| "grad_norm": 0.8871288895606995, | |
| "learning_rate": 0.00017841872330261101, | |
| "loss": 0.2408, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.22233820459290188, | |
| "grad_norm": 0.6930763125419617, | |
| "learning_rate": 0.00017821265652748328, | |
| "loss": 0.1755, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.22338204592901878, | |
| "grad_norm": 0.6196439862251282, | |
| "learning_rate": 0.00017800573081852122, | |
| "loss": 0.1714, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.2244258872651357, | |
| "grad_norm": 0.4868805408477783, | |
| "learning_rate": 0.00017779794844818944, | |
| "loss": 0.1688, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.2254697286012526, | |
| "grad_norm": 0.5908803343772888, | |
| "learning_rate": 0.0001775893116983604, | |
| "loss": 0.1653, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.2265135699373695, | |
| "grad_norm": 0.7646514177322388, | |
| "learning_rate": 0.00017737982286028937, | |
| "loss": 0.3009, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.22755741127348644, | |
| "grad_norm": 0.7528402209281921, | |
| "learning_rate": 0.00017716948423458938, | |
| "loss": 0.3261, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.22860125260960334, | |
| "grad_norm": 0.6268967390060425, | |
| "learning_rate": 0.0001769582981312057, | |
| "loss": 0.2322, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.22964509394572025, | |
| "grad_norm": 0.7152937650680542, | |
| "learning_rate": 0.0001767462668693908, | |
| "loss": 0.2756, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.23068893528183715, | |
| "grad_norm": 0.7480033040046692, | |
| "learning_rate": 0.00017653339277767853, | |
| "loss": 0.2613, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.23173277661795408, | |
| "grad_norm": 0.6832448840141296, | |
| "learning_rate": 0.00017631967819385885, | |
| "loss": 0.1719, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.23277661795407098, | |
| "grad_norm": 0.8552375435829163, | |
| "learning_rate": 0.00017610512546495195, | |
| "loss": 0.3867, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.23382045929018788, | |
| "grad_norm": 1.193741798400879, | |
| "learning_rate": 0.00017588973694718262, | |
| "loss": 0.4765, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.2348643006263048, | |
| "grad_norm": 1.2147541046142578, | |
| "learning_rate": 0.00017567351500595425, | |
| "loss": 0.4603, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2359081419624217, | |
| "grad_norm": 1.028960108757019, | |
| "learning_rate": 0.00017545646201582303, | |
| "loss": 0.4966, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.23695198329853862, | |
| "grad_norm": 1.021507978439331, | |
| "learning_rate": 0.00017523858036047157, | |
| "loss": 0.3946, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.23799582463465555, | |
| "grad_norm": 0.9800359606742859, | |
| "learning_rate": 0.00017501987243268306, | |
| "loss": 0.4061, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.23903966597077245, | |
| "grad_norm": 1.0288289785385132, | |
| "learning_rate": 0.00017480034063431477, | |
| "loss": 0.4, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.24008350730688935, | |
| "grad_norm": 0.9607250690460205, | |
| "learning_rate": 0.00017457998737627182, | |
| "loss": 0.3849, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.24112734864300625, | |
| "grad_norm": 0.9097132086753845, | |
| "learning_rate": 0.00017435881507848046, | |
| "loss": 0.4232, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.24217118997912318, | |
| "grad_norm": 0.9774672985076904, | |
| "learning_rate": 0.00017413682616986185, | |
| "loss": 0.4199, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.24321503131524008, | |
| "grad_norm": 0.9056701064109802, | |
| "learning_rate": 0.00017391402308830503, | |
| "loss": 0.3844, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.24425887265135698, | |
| "grad_norm": 0.9123631119728088, | |
| "learning_rate": 0.00017369040828064047, | |
| "loss": 0.3614, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.2453027139874739, | |
| "grad_norm": 0.8399245142936707, | |
| "learning_rate": 0.00017346598420261294, | |
| "loss": 0.3572, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.24634655532359082, | |
| "grad_norm": 1.1125223636627197, | |
| "learning_rate": 0.00017324075331885466, | |
| "loss": 0.3155, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.24739039665970772, | |
| "grad_norm": 0.7240562438964844, | |
| "learning_rate": 0.00017301471810285822, | |
| "loss": 0.281, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.24843423799582465, | |
| "grad_norm": 0.6814681887626648, | |
| "learning_rate": 0.00017278788103694943, | |
| "loss": 0.2605, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.24947807933194155, | |
| "grad_norm": 0.7790616154670715, | |
| "learning_rate": 0.00017256024461226001, | |
| "loss": 0.3214, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.25052192066805845, | |
| "grad_norm": 0.7958056926727295, | |
| "learning_rate": 0.00017233181132870028, | |
| "loss": 0.2668, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.25052192066805845, | |
| "eval_loss": 0.28679800033569336, | |
| "eval_runtime": 81.356, | |
| "eval_samples_per_second": 19.839, | |
| "eval_steps_per_second": 9.919, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.25156576200417535, | |
| "grad_norm": 0.7587823271751404, | |
| "learning_rate": 0.0001721025836949317, | |
| "loss": 0.2911, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.25260960334029225, | |
| "grad_norm": 0.8206672668457031, | |
| "learning_rate": 0.00017187256422833929, | |
| "loss": 0.2602, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.2536534446764092, | |
| "grad_norm": 0.7913001179695129, | |
| "learning_rate": 0.000171641755455004, | |
| "loss": 0.242, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.2546972860125261, | |
| "grad_norm": 0.6242060661315918, | |
| "learning_rate": 0.000171410159909675, | |
| "loss": 0.2098, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.255741127348643, | |
| "grad_norm": 0.9254711270332336, | |
| "learning_rate": 0.00017117778013574174, | |
| "loss": 0.2028, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2567849686847599, | |
| "grad_norm": 0.804876446723938, | |
| "learning_rate": 0.00017094461868520622, | |
| "loss": 0.215, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.2578288100208768, | |
| "grad_norm": 0.786629319190979, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.2075, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.2588726513569937, | |
| "grad_norm": 0.9948318004608154, | |
| "learning_rate": 0.0001704759610052299, | |
| "loss": 0.243, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.2599164926931106, | |
| "grad_norm": 1.256179928779602, | |
| "learning_rate": 0.00017024046992260237, | |
| "loss": 0.2296, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.2609603340292276, | |
| "grad_norm": 0.9201668500900269, | |
| "learning_rate": 0.00017000420745694254, | |
| "loss": 0.1739, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2620041753653445, | |
| "grad_norm": 0.8208956122398376, | |
| "learning_rate": 0.0001697671762028922, | |
| "loss": 0.4128, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.2630480167014614, | |
| "grad_norm": 0.8502789735794067, | |
| "learning_rate": 0.00016952937876353597, | |
| "loss": 0.4016, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.2640918580375783, | |
| "grad_norm": 1.1290831565856934, | |
| "learning_rate": 0.00016929081775037276, | |
| "loss": 0.442, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.2651356993736952, | |
| "grad_norm": 0.9651896953582764, | |
| "learning_rate": 0.00016905149578328702, | |
| "loss": 0.4346, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.2661795407098121, | |
| "grad_norm": 0.9524455070495605, | |
| "learning_rate": 0.00016881141549052013, | |
| "loss": 0.3441, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.267223382045929, | |
| "grad_norm": 0.8862738013267517, | |
| "learning_rate": 0.00016857057950864132, | |
| "loss": 0.3336, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.26826722338204595, | |
| "grad_norm": 0.8053174614906311, | |
| "learning_rate": 0.00016832899048251884, | |
| "loss": 0.3265, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.26931106471816285, | |
| "grad_norm": 0.7121575474739075, | |
| "learning_rate": 0.00016808665106529094, | |
| "loss": 0.3, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.27035490605427975, | |
| "grad_norm": 0.8516458868980408, | |
| "learning_rate": 0.00016784356391833665, | |
| "loss": 0.2941, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.27139874739039666, | |
| "grad_norm": 0.9382047653198242, | |
| "learning_rate": 0.0001675997317112466, | |
| "loss": 0.3625, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.27244258872651356, | |
| "grad_norm": 1.0630011558532715, | |
| "learning_rate": 0.00016735515712179368, | |
| "loss": 0.2875, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.27348643006263046, | |
| "grad_norm": 0.8051387071609497, | |
| "learning_rate": 0.0001671098428359037, | |
| "loss": 0.2498, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.2745302713987474, | |
| "grad_norm": 0.898259162902832, | |
| "learning_rate": 0.00016686379154762574, | |
| "loss": 0.2293, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.2755741127348643, | |
| "grad_norm": 0.6109540462493896, | |
| "learning_rate": 0.00016661700595910285, | |
| "loss": 0.2029, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.2766179540709812, | |
| "grad_norm": 0.5780929327011108, | |
| "learning_rate": 0.0001663694887805419, | |
| "loss": 0.1858, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2776617954070981, | |
| "grad_norm": 0.5590531826019287, | |
| "learning_rate": 0.0001661212427301844, | |
| "loss": 0.1911, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.278705636743215, | |
| "grad_norm": 0.382966548204422, | |
| "learning_rate": 0.00016587227053427612, | |
| "loss": 0.114, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.2797494780793319, | |
| "grad_norm": 0.7528713941574097, | |
| "learning_rate": 0.00016562257492703757, | |
| "loss": 0.2135, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.2807933194154488, | |
| "grad_norm": 0.5933663845062256, | |
| "learning_rate": 0.00016537215865063365, | |
| "loss": 0.1993, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.2818371607515658, | |
| "grad_norm": 0.7138686776161194, | |
| "learning_rate": 0.00016512102445514375, | |
| "loss": 0.2874, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2828810020876827, | |
| "grad_norm": 0.6350995898246765, | |
| "learning_rate": 0.0001648691750985314, | |
| "loss": 0.2225, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.2839248434237996, | |
| "grad_norm": 0.7503290772438049, | |
| "learning_rate": 0.00016461661334661414, | |
| "loss": 0.2687, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.2849686847599165, | |
| "grad_norm": 0.633063018321991, | |
| "learning_rate": 0.00016436334197303295, | |
| "loss": 0.1867, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.2860125260960334, | |
| "grad_norm": 0.750601053237915, | |
| "learning_rate": 0.000164109363759222, | |
| "loss": 0.1644, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.2870563674321503, | |
| "grad_norm": 0.9629106521606445, | |
| "learning_rate": 0.000163854681494378, | |
| "loss": 0.463, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.2881002087682672, | |
| "grad_norm": 1.2181180715560913, | |
| "learning_rate": 0.00016359929797542944, | |
| "loss": 0.497, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.28914405010438415, | |
| "grad_norm": 1.0051687955856323, | |
| "learning_rate": 0.00016334321600700613, | |
| "loss": 0.3734, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.29018789144050106, | |
| "grad_norm": 0.8457571268081665, | |
| "learning_rate": 0.00016308643840140828, | |
| "loss": 0.3634, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.29123173277661796, | |
| "grad_norm": 0.8827991485595703, | |
| "learning_rate": 0.00016282896797857548, | |
| "loss": 0.4218, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.29227557411273486, | |
| "grad_norm": 0.767482578754425, | |
| "learning_rate": 0.000162570807566056, | |
| "loss": 0.3099, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.29331941544885176, | |
| "grad_norm": 0.9973590970039368, | |
| "learning_rate": 0.00016231195999897558, | |
| "loss": 0.3751, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.29436325678496866, | |
| "grad_norm": 1.094586730003357, | |
| "learning_rate": 0.00016205242812000617, | |
| "loss": 0.3329, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.2954070981210856, | |
| "grad_norm": 1.1051408052444458, | |
| "learning_rate": 0.0001617922147793351, | |
| "loss": 0.3888, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.2964509394572025, | |
| "grad_norm": 1.1265891790390015, | |
| "learning_rate": 0.00016153132283463326, | |
| "loss": 0.3881, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.2974947807933194, | |
| "grad_norm": 0.9808762669563293, | |
| "learning_rate": 0.00016126975515102422, | |
| "loss": 0.3608, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.2985386221294363, | |
| "grad_norm": 0.9096065759658813, | |
| "learning_rate": 0.00016100751460105243, | |
| "loss": 0.3036, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.29958246346555323, | |
| "grad_norm": 0.7918301224708557, | |
| "learning_rate": 0.0001607446040646518, | |
| "loss": 0.265, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.30062630480167013, | |
| "grad_norm": 0.8781275153160095, | |
| "learning_rate": 0.00016048102642911397, | |
| "loss": 0.3192, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.30167014613778703, | |
| "grad_norm": 0.7994233965873718, | |
| "learning_rate": 0.00016021678458905684, | |
| "loss": 0.2619, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.302713987473904, | |
| "grad_norm": 0.7719682455062866, | |
| "learning_rate": 0.0001599518814463925, | |
| "loss": 0.2373, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3037578288100209, | |
| "grad_norm": 0.7163065075874329, | |
| "learning_rate": 0.00015968631991029555, | |
| "loss": 0.2461, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.3048016701461378, | |
| "grad_norm": 0.6648116707801819, | |
| "learning_rate": 0.00015942010289717105, | |
| "loss": 0.2033, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.3058455114822547, | |
| "grad_norm": 0.7089317440986633, | |
| "learning_rate": 0.00015915323333062255, | |
| "loss": 0.2308, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.3068893528183716, | |
| "grad_norm": 0.6273900270462036, | |
| "learning_rate": 0.00015888571414141996, | |
| "loss": 0.1762, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.3079331941544885, | |
| "grad_norm": 0.570709228515625, | |
| "learning_rate": 0.00015861754826746734, | |
| "loss": 0.1612, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3089770354906054, | |
| "grad_norm": 0.7131444811820984, | |
| "learning_rate": 0.00015834873865377077, | |
| "loss": 0.1748, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.31002087682672236, | |
| "grad_norm": 0.857348620891571, | |
| "learning_rate": 0.00015807928825240566, | |
| "loss": 0.2076, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.31106471816283926, | |
| "grad_norm": 0.6636834740638733, | |
| "learning_rate": 0.00015780920002248484, | |
| "loss": 0.1577, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.31210855949895616, | |
| "grad_norm": 0.8303975462913513, | |
| "learning_rate": 0.00015753847693012566, | |
| "loss": 0.1722, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.31315240083507306, | |
| "grad_norm": 1.009426474571228, | |
| "learning_rate": 0.00015726712194841756, | |
| "loss": 0.181, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.31419624217118997, | |
| "grad_norm": 0.873030960559845, | |
| "learning_rate": 0.0001569951380573894, | |
| "loss": 0.4061, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.31524008350730687, | |
| "grad_norm": 0.8714029788970947, | |
| "learning_rate": 0.0001567225282439768, | |
| "loss": 0.4541, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.3162839248434238, | |
| "grad_norm": 1.2070977687835693, | |
| "learning_rate": 0.0001564492955019892, | |
| "loss": 0.444, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.3173277661795407, | |
| "grad_norm": 1.1216182708740234, | |
| "learning_rate": 0.0001561754428320771, | |
| "loss": 0.378, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.31837160751565763, | |
| "grad_norm": 1.0232008695602417, | |
| "learning_rate": 0.00015590097324169909, | |
| "loss": 0.3044, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.31941544885177453, | |
| "grad_norm": 0.8639291524887085, | |
| "learning_rate": 0.00015562588974508872, | |
| "loss": 0.3222, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.32045929018789143, | |
| "grad_norm": 0.7577306032180786, | |
| "learning_rate": 0.00015535019536322157, | |
| "loss": 0.3346, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.32150313152400833, | |
| "grad_norm": 0.7879360914230347, | |
| "learning_rate": 0.00015507389312378197, | |
| "loss": 0.2889, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.32254697286012524, | |
| "grad_norm": 0.8477923274040222, | |
| "learning_rate": 0.0001547969860611297, | |
| "loss": 0.2948, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.3235908141962422, | |
| "grad_norm": 0.8615967631340027, | |
| "learning_rate": 0.00015451947721626676, | |
| "loss": 0.3135, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3246346555323591, | |
| "grad_norm": 1.0251150131225586, | |
| "learning_rate": 0.00015424136963680397, | |
| "loss": 0.2807, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.325678496868476, | |
| "grad_norm": 0.8408343195915222, | |
| "learning_rate": 0.00015396266637692743, | |
| "loss": 0.2879, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.3267223382045929, | |
| "grad_norm": 0.7773711681365967, | |
| "learning_rate": 0.00015368337049736502, | |
| "loss": 0.2457, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.3277661795407098, | |
| "grad_norm": 0.797435462474823, | |
| "learning_rate": 0.00015340348506535283, | |
| "loss": 0.259, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.3288100208768267, | |
| "grad_norm": 0.8681854605674744, | |
| "learning_rate": 0.00015312301315460137, | |
| "loss": 0.2163, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.3298538622129436, | |
| "grad_norm": 0.8305045962333679, | |
| "learning_rate": 0.00015284195784526195, | |
| "loss": 0.1331, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.33089770354906056, | |
| "grad_norm": 0.4855158030986786, | |
| "learning_rate": 0.00015256032222389277, | |
| "loss": 0.153, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.33194154488517746, | |
| "grad_norm": 0.4702201783657074, | |
| "learning_rate": 0.00015227810938342492, | |
| "loss": 0.1527, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.33298538622129437, | |
| "grad_norm": 0.6600713133811951, | |
| "learning_rate": 0.0001519953224231287, | |
| "loss": 0.2761, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.33402922755741127, | |
| "grad_norm": 0.5896007418632507, | |
| "learning_rate": 0.00015171196444857933, | |
| "loss": 0.1978, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.33507306889352817, | |
| "grad_norm": 0.7142235040664673, | |
| "learning_rate": 0.0001514280385716229, | |
| "loss": 0.2759, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.33611691022964507, | |
| "grad_norm": 0.6415271759033203, | |
| "learning_rate": 0.00015114354791034225, | |
| "loss": 0.265, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.33716075156576203, | |
| "grad_norm": 0.6545302867889404, | |
| "learning_rate": 0.00015085849558902264, | |
| "loss": 0.2221, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.33820459290187893, | |
| "grad_norm": 0.7755191922187805, | |
| "learning_rate": 0.00015057288473811772, | |
| "loss": 0.2827, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.33924843423799583, | |
| "grad_norm": 0.960443913936615, | |
| "learning_rate": 0.00015028671849421464, | |
| "loss": 0.4552, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.34029227557411273, | |
| "grad_norm": 1.016601324081421, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.4142, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.34133611691022964, | |
| "grad_norm": 1.2163599729537964, | |
| "learning_rate": 0.00014971273240422535, | |
| "loss": 0.5009, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.34237995824634654, | |
| "grad_norm": 1.181193470954895, | |
| "learning_rate": 0.0001494249188616723, | |
| "loss": 0.5031, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.34342379958246344, | |
| "grad_norm": 0.8927332758903503, | |
| "learning_rate": 0.0001491365625331182, | |
| "loss": 0.3835, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.3444676409185804, | |
| "grad_norm": 0.9175090789794922, | |
| "learning_rate": 0.00014884766658530125, | |
| "loss": 0.4822, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3455114822546973, | |
| "grad_norm": 0.8441628813743591, | |
| "learning_rate": 0.00014855823419088576, | |
| "loss": 0.3455, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.3465553235908142, | |
| "grad_norm": 0.7784302234649658, | |
| "learning_rate": 0.00014826826852842726, | |
| "loss": 0.3359, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.3475991649269311, | |
| "grad_norm": 1.0156643390655518, | |
| "learning_rate": 0.00014797777278233778, | |
| "loss": 0.4488, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.348643006263048, | |
| "grad_norm": 0.9029203057289124, | |
| "learning_rate": 0.00014768675014285062, | |
| "loss": 0.3175, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.3496868475991649, | |
| "grad_norm": 0.8379835486412048, | |
| "learning_rate": 0.0001473952038059855, | |
| "loss": 0.352, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.35073068893528186, | |
| "grad_norm": 0.807041585445404, | |
| "learning_rate": 0.00014710313697351341, | |
| "loss": 0.3151, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.35177453027139877, | |
| "grad_norm": 0.7760060429573059, | |
| "learning_rate": 0.0001468105528529214, | |
| "loss": 0.3397, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.35281837160751567, | |
| "grad_norm": 0.7737391591072083, | |
| "learning_rate": 0.00014651745465737737, | |
| "loss": 0.3054, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.35386221294363257, | |
| "grad_norm": 0.7160333395004272, | |
| "learning_rate": 0.00014622384560569493, | |
| "loss": 0.2271, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.35490605427974947, | |
| "grad_norm": 0.7895155549049377, | |
| "learning_rate": 0.00014592972892229778, | |
| "loss": 0.2717, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3559498956158664, | |
| "grad_norm": 0.6400478482246399, | |
| "learning_rate": 0.00014563510783718457, | |
| "loss": 0.2212, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.3569937369519833, | |
| "grad_norm": 0.7159572243690491, | |
| "learning_rate": 0.0001453399855858932, | |
| "loss": 0.2192, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.35803757828810023, | |
| "grad_norm": 0.8888924717903137, | |
| "learning_rate": 0.00014504436540946548, | |
| "loss": 0.269, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.35908141962421714, | |
| "grad_norm": 0.8276658058166504, | |
| "learning_rate": 0.00014474825055441136, | |
| "loss": 0.2466, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.36012526096033404, | |
| "grad_norm": 0.690984308719635, | |
| "learning_rate": 0.00014445164427267344, | |
| "loss": 0.1649, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.36116910229645094, | |
| "grad_norm": 0.7650169730186462, | |
| "learning_rate": 0.0001441545498215912, | |
| "loss": 0.2323, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.36221294363256784, | |
| "grad_norm": 0.8575247526168823, | |
| "learning_rate": 0.00014385697046386512, | |
| "loss": 0.1856, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.36325678496868474, | |
| "grad_norm": 0.9500517249107361, | |
| "learning_rate": 0.00014355890946752102, | |
| "loss": 0.2059, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.36430062630480164, | |
| "grad_norm": 0.7111496925354004, | |
| "learning_rate": 0.00014326037010587404, | |
| "loss": 0.144, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.3653444676409186, | |
| "grad_norm": 0.9617753624916077, | |
| "learning_rate": 0.0001429613556574928, | |
| "loss": 0.2208, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3663883089770355, | |
| "grad_norm": 0.7692716717720032, | |
| "learning_rate": 0.00014266186940616328, | |
| "loss": 0.4445, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.3674321503131524, | |
| "grad_norm": 0.742679238319397, | |
| "learning_rate": 0.00014236191464085286, | |
| "loss": 0.3707, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.3684759916492693, | |
| "grad_norm": 0.928932785987854, | |
| "learning_rate": 0.00014206149465567403, | |
| "loss": 0.4158, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.3695198329853862, | |
| "grad_norm": 1.0475050210952759, | |
| "learning_rate": 0.00014176061274984858, | |
| "loss": 0.4644, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.3705636743215031, | |
| "grad_norm": 0.9185068607330322, | |
| "learning_rate": 0.0001414592722276709, | |
| "loss": 0.4066, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.37160751565762007, | |
| "grad_norm": 0.8016102910041809, | |
| "learning_rate": 0.00014115747639847204, | |
| "loss": 0.3827, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.37265135699373697, | |
| "grad_norm": 0.689323365688324, | |
| "learning_rate": 0.0001408552285765832, | |
| "loss": 0.3004, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.3736951983298539, | |
| "grad_norm": 0.721184253692627, | |
| "learning_rate": 0.00014055253208129938, | |
| "loss": 0.311, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.3747390396659708, | |
| "grad_norm": 0.8813337683677673, | |
| "learning_rate": 0.00014024939023684298, | |
| "loss": 0.3493, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.3757828810020877, | |
| "grad_norm": 0.8543792366981506, | |
| "learning_rate": 0.00013994580637232716, | |
| "loss": 0.2636, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3768267223382046, | |
| "grad_norm": 0.8040288686752319, | |
| "learning_rate": 0.00013964178382171942, | |
| "loss": 0.2661, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.3778705636743215, | |
| "grad_norm": 0.7446701526641846, | |
| "learning_rate": 0.00013933732592380483, | |
| "loss": 0.2464, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.37891440501043844, | |
| "grad_norm": 0.7692604064941406, | |
| "learning_rate": 0.0001390324360221496, | |
| "loss": 0.2179, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.37995824634655534, | |
| "grad_norm": 0.6927527785301208, | |
| "learning_rate": 0.00013872711746506413, | |
| "loss": 0.1913, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.38100208768267224, | |
| "grad_norm": 0.750212550163269, | |
| "learning_rate": 0.00013842137360556628, | |
| "loss": 0.1723, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.38204592901878914, | |
| "grad_norm": 0.7114387154579163, | |
| "learning_rate": 0.0001381152078013447, | |
| "loss": 0.1899, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.38308977035490605, | |
| "grad_norm": 0.671464741230011, | |
| "learning_rate": 0.00013780862341472182, | |
| "loss": 0.2059, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.38413361169102295, | |
| "grad_norm": 0.6177671551704407, | |
| "learning_rate": 0.00013750162381261693, | |
| "loss": 0.1888, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.38517745302713985, | |
| "grad_norm": 0.7235206365585327, | |
| "learning_rate": 0.0001371942123665092, | |
| "loss": 0.2612, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.3862212943632568, | |
| "grad_norm": 0.5639315843582153, | |
| "learning_rate": 0.00013688639245240078, | |
| "loss": 0.2214, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3872651356993737, | |
| "grad_norm": 0.5648091435432434, | |
| "learning_rate": 0.00013657816745077955, | |
| "loss": 0.2087, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.3883089770354906, | |
| "grad_norm": 0.6048844456672668, | |
| "learning_rate": 0.0001362695407465821, | |
| "loss": 0.2312, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.3893528183716075, | |
| "grad_norm": 0.48515263199806213, | |
| "learning_rate": 0.0001359605157291565, | |
| "loss": 0.1624, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.3903966597077244, | |
| "grad_norm": 0.6982168555259705, | |
| "learning_rate": 0.0001356510957922251, | |
| "loss": 0.2632, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.3914405010438413, | |
| "grad_norm": 0.9829317331314087, | |
| "learning_rate": 0.0001353412843338474, | |
| "loss": 0.3995, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3924843423799583, | |
| "grad_norm": 1.1811579465866089, | |
| "learning_rate": 0.00013503108475638244, | |
| "loss": 0.4381, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.3935281837160752, | |
| "grad_norm": 1.0763869285583496, | |
| "learning_rate": 0.00013472050046645166, | |
| "loss": 0.4366, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.3945720250521921, | |
| "grad_norm": 0.8847516179084778, | |
| "learning_rate": 0.00013440953487490144, | |
| "loss": 0.3637, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.395615866388309, | |
| "grad_norm": 0.8159587979316711, | |
| "learning_rate": 0.0001340981913967656, | |
| "loss": 0.3771, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.3966597077244259, | |
| "grad_norm": 0.8111384510993958, | |
| "learning_rate": 0.00013378647345122795, | |
| "loss": 0.3473, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3977035490605428, | |
| "grad_norm": 0.8355801105499268, | |
| "learning_rate": 0.00013347438446158466, | |
| "loss": 0.381, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.3987473903966597, | |
| "grad_norm": 0.7971011400222778, | |
| "learning_rate": 0.0001331619278552068, | |
| "loss": 0.3019, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.39979123173277664, | |
| "grad_norm": 0.8738229274749756, | |
| "learning_rate": 0.00013284910706350247, | |
| "loss": 0.3766, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.40083507306889354, | |
| "grad_norm": 0.8685609698295593, | |
| "learning_rate": 0.0001325359255218795, | |
| "loss": 0.3371, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.40187891440501045, | |
| "grad_norm": 0.8119929432868958, | |
| "learning_rate": 0.00013222238666970728, | |
| "loss": 0.3047, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.40292275574112735, | |
| "grad_norm": 0.831425130367279, | |
| "learning_rate": 0.00013190849395027928, | |
| "loss": 0.3241, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.40396659707724425, | |
| "grad_norm": 0.8819262981414795, | |
| "learning_rate": 0.0001315942508107751, | |
| "loss": 0.2986, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.40501043841336115, | |
| "grad_norm": 0.8911377787590027, | |
| "learning_rate": 0.00013127966070222274, | |
| "loss": 0.3614, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.40605427974947805, | |
| "grad_norm": 0.7478219866752625, | |
| "learning_rate": 0.00013096472707946056, | |
| "loss": 0.2861, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.407098121085595, | |
| "grad_norm": 0.8377799391746521, | |
| "learning_rate": 0.00013064945340109948, | |
| "loss": 0.2854, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.4081419624217119, | |
| "grad_norm": 0.7409958839416504, | |
| "learning_rate": 0.00013033384312948488, | |
| "loss": 0.2589, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.4091858037578288, | |
| "grad_norm": 0.6093199849128723, | |
| "learning_rate": 0.00013001789973065853, | |
| "loss": 0.2124, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.4102296450939457, | |
| "grad_norm": 0.5660393834114075, | |
| "learning_rate": 0.00012970162667432075, | |
| "loss": 0.1718, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.4112734864300626, | |
| "grad_norm": 0.7063373923301697, | |
| "learning_rate": 0.00012938502743379212, | |
| "loss": 0.2245, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.4123173277661795, | |
| "grad_norm": 0.8215602040290833, | |
| "learning_rate": 0.00012906810548597532, | |
| "loss": 0.2442, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.4133611691022965, | |
| "grad_norm": 0.6678823828697205, | |
| "learning_rate": 0.00012875086431131716, | |
| "loss": 0.1891, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.4144050104384134, | |
| "grad_norm": 0.6394354701042175, | |
| "learning_rate": 0.00012843330739377, | |
| "loss": 0.2203, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.4154488517745303, | |
| "grad_norm": 0.7983390688896179, | |
| "learning_rate": 0.00012811543822075397, | |
| "loss": 0.2113, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.4164926931106472, | |
| "grad_norm": 0.611224889755249, | |
| "learning_rate": 0.0001277972602831181, | |
| "loss": 0.1616, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.4175365344467641, | |
| "grad_norm": 0.9979553818702698, | |
| "learning_rate": 0.00012747877707510252, | |
| "loss": 0.1953, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.418580375782881, | |
| "grad_norm": 0.5957475304603577, | |
| "learning_rate": 0.00012715999209429973, | |
| "loss": 0.3356, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.4196242171189979, | |
| "grad_norm": 0.6433320641517639, | |
| "learning_rate": 0.00012684090884161636, | |
| "loss": 0.3177, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.42066805845511485, | |
| "grad_norm": 0.8637453317642212, | |
| "learning_rate": 0.00012652153082123456, | |
| "loss": 0.4008, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.42171189979123175, | |
| "grad_norm": 0.6944557428359985, | |
| "learning_rate": 0.00012620186154057382, | |
| "loss": 0.3198, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.42275574112734865, | |
| "grad_norm": 0.8894799947738647, | |
| "learning_rate": 0.00012588190451025207, | |
| "loss": 0.3917, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.42379958246346555, | |
| "grad_norm": 0.7825080156326294, | |
| "learning_rate": 0.0001255616632440475, | |
| "loss": 0.3679, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.42484342379958245, | |
| "grad_norm": 0.6816397309303284, | |
| "learning_rate": 0.00012524114125885957, | |
| "loss": 0.2855, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.42588726513569936, | |
| "grad_norm": 0.7824950218200684, | |
| "learning_rate": 0.0001249203420746708, | |
| "loss": 0.305, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.42693110647181626, | |
| "grad_norm": 0.7584050297737122, | |
| "learning_rate": 0.0001245992692145078, | |
| "loss": 0.3123, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.4279749478079332, | |
| "grad_norm": 0.6384229063987732, | |
| "learning_rate": 0.00012427792620440278, | |
| "loss": 0.233, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4290187891440501, | |
| "grad_norm": 0.7239426374435425, | |
| "learning_rate": 0.00012395631657335468, | |
| "loss": 0.216, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.430062630480167, | |
| "grad_norm": 0.6760838627815247, | |
| "learning_rate": 0.0001236344438532905, | |
| "loss": 0.2092, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.4311064718162839, | |
| "grad_norm": 0.5736236572265625, | |
| "learning_rate": 0.00012331231157902648, | |
| "loss": 0.1725, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.4321503131524008, | |
| "grad_norm": 0.5566690564155579, | |
| "learning_rate": 0.00012298992328822937, | |
| "loss": 0.1645, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.4331941544885177, | |
| "grad_norm": 0.5182932019233704, | |
| "learning_rate": 0.00012266728252137733, | |
| "loss": 0.1484, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.4342379958246347, | |
| "grad_norm": 0.6892213821411133, | |
| "learning_rate": 0.00012234439282172142, | |
| "loss": 0.1422, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.4352818371607516, | |
| "grad_norm": 0.46982264518737793, | |
| "learning_rate": 0.0001220212577352464, | |
| "loss": 0.1292, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.4363256784968685, | |
| "grad_norm": 0.5547206401824951, | |
| "learning_rate": 0.0001216978808106318, | |
| "loss": 0.1813, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.4373695198329854, | |
| "grad_norm": 0.6126968264579773, | |
| "learning_rate": 0.00012137426559921316, | |
| "loss": 0.1804, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.4384133611691023, | |
| "grad_norm": 0.6041930317878723, | |
| "learning_rate": 0.0001210504156549428, | |
| "loss": 0.2519, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4394572025052192, | |
| "grad_norm": 0.6091843843460083, | |
| "learning_rate": 0.00012072633453435091, | |
| "loss": 0.2423, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.4405010438413361, | |
| "grad_norm": 0.5664049983024597, | |
| "learning_rate": 0.00012040202579650648, | |
| "loss": 0.2357, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.44154488517745305, | |
| "grad_norm": 0.4431195855140686, | |
| "learning_rate": 0.00012007749300297817, | |
| "loss": 0.1704, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.44258872651356995, | |
| "grad_norm": 0.5842788815498352, | |
| "learning_rate": 0.00011975273971779528, | |
| "loss": 0.1979, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.44363256784968685, | |
| "grad_norm": 0.8385937809944153, | |
| "learning_rate": 0.00011942776950740848, | |
| "loss": 0.4167, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.44467640918580376, | |
| "grad_norm": 0.8559632301330566, | |
| "learning_rate": 0.00011910258594065078, | |
| "loss": 0.4068, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.44572025052192066, | |
| "grad_norm": 0.8821555972099304, | |
| "learning_rate": 0.00011877719258869826, | |
| "loss": 0.3839, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.44676409185803756, | |
| "grad_norm": 0.8119932413101196, | |
| "learning_rate": 0.00011845159302503086, | |
| "loss": 0.3336, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.44780793319415446, | |
| "grad_norm": 0.8578853011131287, | |
| "learning_rate": 0.00011812579082539317, | |
| "loss": 0.4184, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.4488517745302714, | |
| "grad_norm": 0.7293173670768738, | |
| "learning_rate": 0.00011779978956775506, | |
| "loss": 0.3733, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4498956158663883, | |
| "grad_norm": 0.8195939064025879, | |
| "learning_rate": 0.00011747359283227251, | |
| "loss": 0.3646, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.4509394572025052, | |
| "grad_norm": 0.7359249591827393, | |
| "learning_rate": 0.00011714720420124831, | |
| "loss": 0.3201, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.4519832985386221, | |
| "grad_norm": 0.8084685206413269, | |
| "learning_rate": 0.00011682062725909258, | |
| "loss": 0.3537, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.453027139874739, | |
| "grad_norm": 0.7745420336723328, | |
| "learning_rate": 0.00011649386559228341, | |
| "loss": 0.2933, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.45407098121085593, | |
| "grad_norm": 0.81271892786026, | |
| "learning_rate": 0.00011616692278932772, | |
| "loss": 0.2821, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.4551148225469729, | |
| "grad_norm": 0.7679653167724609, | |
| "learning_rate": 0.0001158398024407215, | |
| "loss": 0.2973, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.4561586638830898, | |
| "grad_norm": 0.826492190361023, | |
| "learning_rate": 0.00011551250813891066, | |
| "loss": 0.311, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.4572025052192067, | |
| "grad_norm": 0.766703724861145, | |
| "learning_rate": 0.00011518504347825145, | |
| "loss": 0.2606, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.4582463465553236, | |
| "grad_norm": 0.7974645495414734, | |
| "learning_rate": 0.00011485741205497094, | |
| "loss": 0.2843, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.4592901878914405, | |
| "grad_norm": 0.6922892928123474, | |
| "learning_rate": 0.0001145296174671277, | |
| "loss": 0.2178, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4603340292275574, | |
| "grad_norm": 0.6793475151062012, | |
| "learning_rate": 0.00011420166331457207, | |
| "loss": 0.2221, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.4613778705636743, | |
| "grad_norm": 0.9168251752853394, | |
| "learning_rate": 0.00011387355319890685, | |
| "loss": 0.2686, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.46242171189979125, | |
| "grad_norm": 0.7042071223258972, | |
| "learning_rate": 0.00011354529072344748, | |
| "loss": 0.2703, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.46346555323590816, | |
| "grad_norm": 0.6087216734886169, | |
| "learning_rate": 0.00011321687949318276, | |
| "loss": 0.2095, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.46450939457202506, | |
| "grad_norm": 0.5309197306632996, | |
| "learning_rate": 0.00011288832311473508, | |
| "loss": 0.188, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.46555323590814196, | |
| "grad_norm": 0.5840495824813843, | |
| "learning_rate": 0.00011255962519632081, | |
| "loss": 0.1569, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.46659707724425886, | |
| "grad_norm": 0.541627049446106, | |
| "learning_rate": 0.00011223078934771079, | |
| "loss": 0.1778, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.46764091858037576, | |
| "grad_norm": 0.4946768581867218, | |
| "learning_rate": 0.00011190181918019049, | |
| "loss": 0.1729, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.46868475991649267, | |
| "grad_norm": 0.4716700613498688, | |
| "learning_rate": 0.00011157271830652062, | |
| "loss": 0.1518, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.4697286012526096, | |
| "grad_norm": 0.8532614707946777, | |
| "learning_rate": 0.00011124349034089723, | |
| "loss": 0.1282, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4707724425887265, | |
| "grad_norm": 0.7531268000602722, | |
| "learning_rate": 0.00011091413889891211, | |
| "loss": 0.3468, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.4718162839248434, | |
| "grad_norm": 0.7047679424285889, | |
| "learning_rate": 0.00011058466759751302, | |
| "loss": 0.3667, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.47286012526096033, | |
| "grad_norm": 0.7256395220756531, | |
| "learning_rate": 0.00011025508005496417, | |
| "loss": 0.3224, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.47390396659707723, | |
| "grad_norm": 0.7575612664222717, | |
| "learning_rate": 0.00010992537989080618, | |
| "loss": 0.3498, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.47494780793319413, | |
| "grad_norm": 0.7202345132827759, | |
| "learning_rate": 0.00010959557072581652, | |
| "loss": 0.3282, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.4759916492693111, | |
| "grad_norm": 0.6914469599723816, | |
| "learning_rate": 0.00010926565618196978, | |
| "loss": 0.2925, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.477035490605428, | |
| "grad_norm": 0.7601653337478638, | |
| "learning_rate": 0.00010893563988239772, | |
| "loss": 0.3728, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.4780793319415449, | |
| "grad_norm": 0.756959080696106, | |
| "learning_rate": 0.0001086055254513497, | |
| "loss": 0.3186, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.4791231732776618, | |
| "grad_norm": 0.6831420063972473, | |
| "learning_rate": 0.00010827531651415266, | |
| "loss": 0.2786, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.4801670146137787, | |
| "grad_norm": 0.6854783296585083, | |
| "learning_rate": 0.00010794501669717145, | |
| "loss": 0.2334, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.4812108559498956, | |
| "grad_norm": 0.667158305644989, | |
| "learning_rate": 0.00010761462962776897, | |
| "loss": 0.1972, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.4822546972860125, | |
| "grad_norm": 0.6852616667747498, | |
| "learning_rate": 0.00010728415893426635, | |
| "loss": 0.2318, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.48329853862212946, | |
| "grad_norm": 0.6210921406745911, | |
| "learning_rate": 0.00010695360824590303, | |
| "loss": 0.2239, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.48434237995824636, | |
| "grad_norm": 0.5915560722351074, | |
| "learning_rate": 0.00010662298119279701, | |
| "loss": 0.1759, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.48538622129436326, | |
| "grad_norm": 0.5121853351593018, | |
| "learning_rate": 0.00010629228140590486, | |
| "loss": 0.1742, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.48643006263048016, | |
| "grad_norm": 0.6679707169532776, | |
| "learning_rate": 0.00010596151251698199, | |
| "loss": 0.1989, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.48747390396659707, | |
| "grad_norm": 0.5263128280639648, | |
| "learning_rate": 0.00010563067815854266, | |
| "loss": 0.1344, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.48851774530271397, | |
| "grad_norm": 0.6040886640548706, | |
| "learning_rate": 0.00010529978196382011, | |
| "loss": 0.1759, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.48956158663883087, | |
| "grad_norm": 0.4196336269378662, | |
| "learning_rate": 0.00010496882756672666, | |
| "loss": 0.1153, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.4906054279749478, | |
| "grad_norm": 0.6484777927398682, | |
| "learning_rate": 0.00010463781860181385, | |
| "loss": 0.2986, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.49164926931106473, | |
| "grad_norm": 0.651329755783081, | |
| "learning_rate": 0.00010430675870423246, | |
| "loss": 0.2466, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.49269311064718163, | |
| "grad_norm": 0.5095123648643494, | |
| "learning_rate": 0.0001039756515096926, | |
| "loss": 0.199, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.49373695198329853, | |
| "grad_norm": 0.5218927264213562, | |
| "learning_rate": 0.00010364450065442377, | |
| "loss": 0.1569, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.49478079331941544, | |
| "grad_norm": 0.5320420861244202, | |
| "learning_rate": 0.00010331330977513509, | |
| "loss": 0.1954, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.49582463465553234, | |
| "grad_norm": 0.7471246719360352, | |
| "learning_rate": 0.00010298208250897503, | |
| "loss": 0.3619, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.4968684759916493, | |
| "grad_norm": 0.7712672352790833, | |
| "learning_rate": 0.00010265082249349187, | |
| "loss": 0.2926, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.4979123173277662, | |
| "grad_norm": 0.8972386717796326, | |
| "learning_rate": 0.00010231953336659334, | |
| "loss": 0.3387, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.4989561586638831, | |
| "grad_norm": 0.827797532081604, | |
| "learning_rate": 0.00010198821876650701, | |
| "loss": 0.3206, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.8829711675643921, | |
| "learning_rate": 0.00010165688233174017, | |
| "loss": 0.3861, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.5010438413361169, | |
| "grad_norm": 0.8328503370285034, | |
| "learning_rate": 0.00010132552770103987, | |
| "loss": 0.4158, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5010438413361169, | |
| "eval_loss": 0.2552998960018158, | |
| "eval_runtime": 81.2486, | |
| "eval_samples_per_second": 19.865, | |
| "eval_steps_per_second": 9.932, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5020876826722338, | |
| "grad_norm": 0.7070721387863159, | |
| "learning_rate": 0.00010099415851335299, | |
| "loss": 0.3174, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.5031315240083507, | |
| "grad_norm": 0.7391024827957153, | |
| "learning_rate": 0.00010066277840778626, | |
| "loss": 0.3442, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.5041753653444676, | |
| "grad_norm": 0.7629324793815613, | |
| "learning_rate": 0.00010033139102356642, | |
| "loss": 0.3439, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.5052192066805845, | |
| "grad_norm": 0.7324389219284058, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3063, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.5062630480167014, | |
| "grad_norm": 0.6402798295021057, | |
| "learning_rate": 9.966860897643359e-05, | |
| "loss": 0.2383, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5073068893528184, | |
| "grad_norm": 0.7618774771690369, | |
| "learning_rate": 9.933722159221376e-05, | |
| "loss": 0.3004, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.5083507306889353, | |
| "grad_norm": 0.8296042680740356, | |
| "learning_rate": 9.900584148664704e-05, | |
| "loss": 0.3208, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.5093945720250522, | |
| "grad_norm": 0.7663673162460327, | |
| "learning_rate": 9.867447229896018e-05, | |
| "loss": 0.3204, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.5104384133611691, | |
| "grad_norm": 0.7188003063201904, | |
| "learning_rate": 9.834311766825985e-05, | |
| "loss": 0.2645, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.511482254697286, | |
| "grad_norm": 0.6017361879348755, | |
| "learning_rate": 9.801178123349298e-05, | |
| "loss": 0.2076, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5125260960334029, | |
| "grad_norm": 0.6702793836593628, | |
| "learning_rate": 9.768046663340669e-05, | |
| "loss": 0.2207, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.5135699373695198, | |
| "grad_norm": 0.6283150911331177, | |
| "learning_rate": 9.734917750650816e-05, | |
| "loss": 0.2246, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.5146137787056367, | |
| "grad_norm": 0.6348150968551636, | |
| "learning_rate": 9.701791749102495e-05, | |
| "loss": 0.1896, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.5156576200417536, | |
| "grad_norm": 0.6684585213661194, | |
| "learning_rate": 9.668669022486494e-05, | |
| "loss": 0.2409, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.5167014613778705, | |
| "grad_norm": 0.678677499294281, | |
| "learning_rate": 9.635549934557625e-05, | |
| "loss": 0.216, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5177453027139874, | |
| "grad_norm": 0.6523580551147461, | |
| "learning_rate": 9.602434849030745e-05, | |
| "loss": 0.1894, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.5187891440501043, | |
| "grad_norm": 0.5122499465942383, | |
| "learning_rate": 9.569324129576757e-05, | |
| "loss": 0.1579, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.5198329853862212, | |
| "grad_norm": 0.5820009112358093, | |
| "learning_rate": 9.536218139818614e-05, | |
| "loss": 0.1766, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.5208768267223383, | |
| "grad_norm": 0.5032172203063965, | |
| "learning_rate": 9.503117243327337e-05, | |
| "loss": 0.1519, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.5219206680584552, | |
| "grad_norm": 0.7425169944763184, | |
| "learning_rate": 9.47002180361799e-05, | |
| "loss": 0.1333, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5229645093945721, | |
| "grad_norm": 0.7141383290290833, | |
| "learning_rate": 9.436932184145737e-05, | |
| "loss": 0.4269, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.524008350730689, | |
| "grad_norm": 0.662886917591095, | |
| "learning_rate": 9.403848748301802e-05, | |
| "loss": 0.2939, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.5250521920668059, | |
| "grad_norm": 0.6695585250854492, | |
| "learning_rate": 9.370771859409513e-05, | |
| "loss": 0.3167, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.5260960334029228, | |
| "grad_norm": 0.7355087399482727, | |
| "learning_rate": 9.337701880720303e-05, | |
| "loss": 0.3414, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.5271398747390397, | |
| "grad_norm": 0.6738423705101013, | |
| "learning_rate": 9.304639175409698e-05, | |
| "loss": 0.3176, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5281837160751566, | |
| "grad_norm": 0.7200036644935608, | |
| "learning_rate": 9.271584106573364e-05, | |
| "loss": 0.3276, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.5292275574112735, | |
| "grad_norm": 0.6562217473983765, | |
| "learning_rate": 9.238537037223104e-05, | |
| "loss": 0.3353, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.5302713987473904, | |
| "grad_norm": 0.6065823435783386, | |
| "learning_rate": 9.205498330282856e-05, | |
| "loss": 0.3062, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.5313152400835073, | |
| "grad_norm": 0.6631349921226501, | |
| "learning_rate": 9.172468348584739e-05, | |
| "loss": 0.2497, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.5323590814196242, | |
| "grad_norm": 0.613549530506134, | |
| "learning_rate": 9.139447454865033e-05, | |
| "loss": 0.2625, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5334029227557411, | |
| "grad_norm": 0.6613410711288452, | |
| "learning_rate": 9.106436011760229e-05, | |
| "loss": 0.244, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.534446764091858, | |
| "grad_norm": 0.5914390087127686, | |
| "learning_rate": 9.073434381803024e-05, | |
| "loss": 0.2109, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.535490605427975, | |
| "grad_norm": 0.5324224829673767, | |
| "learning_rate": 9.04044292741835e-05, | |
| "loss": 0.1926, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.5365344467640919, | |
| "grad_norm": 0.652651846408844, | |
| "learning_rate": 9.007462010919386e-05, | |
| "loss": 0.2254, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.5375782881002088, | |
| "grad_norm": 0.6112195253372192, | |
| "learning_rate": 8.974491994503584e-05, | |
| "loss": 0.205, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5386221294363257, | |
| "grad_norm": 0.5404685735702515, | |
| "learning_rate": 8.941533240248699e-05, | |
| "loss": 0.1928, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.5396659707724426, | |
| "grad_norm": 0.47451335191726685, | |
| "learning_rate": 8.908586110108794e-05, | |
| "loss": 0.1307, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.5407098121085595, | |
| "grad_norm": 0.8480343818664551, | |
| "learning_rate": 8.875650965910279e-05, | |
| "loss": 0.1533, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.5417536534446764, | |
| "grad_norm": 0.6759589314460754, | |
| "learning_rate": 8.842728169347939e-05, | |
| "loss": 0.1468, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.5427974947807933, | |
| "grad_norm": 0.5591132640838623, | |
| "learning_rate": 8.809818081980953e-05, | |
| "loss": 0.2246, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5438413361169102, | |
| "grad_norm": 0.6184394955635071, | |
| "learning_rate": 8.776921065228924e-05, | |
| "loss": 0.2189, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.5448851774530271, | |
| "grad_norm": 0.5175319910049438, | |
| "learning_rate": 8.744037480367921e-05, | |
| "loss": 0.1886, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.545929018789144, | |
| "grad_norm": 0.645250678062439, | |
| "learning_rate": 8.711167688526493e-05, | |
| "loss": 0.2297, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.5469728601252609, | |
| "grad_norm": 0.6044825315475464, | |
| "learning_rate": 8.678312050681724e-05, | |
| "loss": 0.2029, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.5480167014613778, | |
| "grad_norm": 0.5178519487380981, | |
| "learning_rate": 8.645470927655255e-05, | |
| "loss": 0.1447, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5490605427974948, | |
| "grad_norm": 0.8626076579093933, | |
| "learning_rate": 8.612644680109319e-05, | |
| "loss": 0.4495, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.5501043841336117, | |
| "grad_norm": 0.8363009691238403, | |
| "learning_rate": 8.579833668542796e-05, | |
| "loss": 0.3709, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.5511482254697286, | |
| "grad_norm": 0.872733473777771, | |
| "learning_rate": 8.547038253287233e-05, | |
| "loss": 0.3226, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.5521920668058455, | |
| "grad_norm": 0.865210235118866, | |
| "learning_rate": 8.514258794502905e-05, | |
| "loss": 0.3579, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.5532359081419624, | |
| "grad_norm": 0.7914073467254639, | |
| "learning_rate": 8.481495652174859e-05, | |
| "loss": 0.3216, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5542797494780793, | |
| "grad_norm": 0.8779425024986267, | |
| "learning_rate": 8.448749186108935e-05, | |
| "loss": 0.3532, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.5553235908141962, | |
| "grad_norm": 0.8584082722663879, | |
| "learning_rate": 8.416019755927851e-05, | |
| "loss": 0.4009, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.5563674321503131, | |
| "grad_norm": 0.8470184206962585, | |
| "learning_rate": 8.383307721067231e-05, | |
| "loss": 0.4198, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.55741127348643, | |
| "grad_norm": 0.7338582277297974, | |
| "learning_rate": 8.35061344077166e-05, | |
| "loss": 0.2725, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.558455114822547, | |
| "grad_norm": 0.7652982473373413, | |
| "learning_rate": 8.317937274090747e-05, | |
| "loss": 0.3007, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5594989561586639, | |
| "grad_norm": 0.7415357232093811, | |
| "learning_rate": 8.28527957987517e-05, | |
| "loss": 0.3274, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.5605427974947808, | |
| "grad_norm": 0.6662179231643677, | |
| "learning_rate": 8.252640716772749e-05, | |
| "loss": 0.2606, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.5615866388308977, | |
| "grad_norm": 0.6139498353004456, | |
| "learning_rate": 8.2200210432245e-05, | |
| "loss": 0.2382, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.5626304801670147, | |
| "grad_norm": 0.7014831900596619, | |
| "learning_rate": 8.187420917460686e-05, | |
| "loss": 0.2542, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.5636743215031316, | |
| "grad_norm": 0.7138461470603943, | |
| "learning_rate": 8.154840697496917e-05, | |
| "loss": 0.2392, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5647181628392485, | |
| "grad_norm": 0.7507902979850769, | |
| "learning_rate": 8.122280741130176e-05, | |
| "loss": 0.2554, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.5657620041753654, | |
| "grad_norm": 0.8535422086715698, | |
| "learning_rate": 8.089741405934922e-05, | |
| "loss": 0.2433, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.5668058455114823, | |
| "grad_norm": 0.660111129283905, | |
| "learning_rate": 8.057223049259155e-05, | |
| "loss": 0.1921, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.5678496868475992, | |
| "grad_norm": 0.49220120906829834, | |
| "learning_rate": 8.024726028220474e-05, | |
| "loss": 0.1793, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.5688935281837161, | |
| "grad_norm": 0.5934033393859863, | |
| "learning_rate": 7.992250699702182e-05, | |
| "loss": 0.1686, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.569937369519833, | |
| "grad_norm": 0.6598916053771973, | |
| "learning_rate": 7.959797420349355e-05, | |
| "loss": 0.1981, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.5709812108559499, | |
| "grad_norm": 0.6087566018104553, | |
| "learning_rate": 7.927366546564911e-05, | |
| "loss": 0.1845, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.5720250521920668, | |
| "grad_norm": 0.4998890459537506, | |
| "learning_rate": 7.894958434505725e-05, | |
| "loss": 0.1524, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.5730688935281837, | |
| "grad_norm": 0.5460024476051331, | |
| "learning_rate": 7.862573440078686e-05, | |
| "loss": 0.1808, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.5741127348643006, | |
| "grad_norm": 0.7462297677993774, | |
| "learning_rate": 7.83021191893682e-05, | |
| "loss": 0.1723, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5751565762004175, | |
| "grad_norm": 0.5173273086547852, | |
| "learning_rate": 7.797874226475361e-05, | |
| "loss": 0.3054, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.5762004175365344, | |
| "grad_norm": 0.6547046303749084, | |
| "learning_rate": 7.765560717827858e-05, | |
| "loss": 0.3101, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.5772442588726514, | |
| "grad_norm": 0.676986575126648, | |
| "learning_rate": 7.733271747862265e-05, | |
| "loss": 0.3376, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.5782881002087683, | |
| "grad_norm": 0.7368578910827637, | |
| "learning_rate": 7.701007671177067e-05, | |
| "loss": 0.3517, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.5793319415448852, | |
| "grad_norm": 0.7136873006820679, | |
| "learning_rate": 7.668768842097353e-05, | |
| "loss": 0.3077, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.5803757828810021, | |
| "grad_norm": 0.7121712565422058, | |
| "learning_rate": 7.636555614670953e-05, | |
| "loss": 0.3271, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.581419624217119, | |
| "grad_norm": 0.7123695611953735, | |
| "learning_rate": 7.604368342664533e-05, | |
| "loss": 0.3356, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.5824634655532359, | |
| "grad_norm": 0.7206712961196899, | |
| "learning_rate": 7.572207379559721e-05, | |
| "loss": 0.2915, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.5835073068893528, | |
| "grad_norm": 0.6520224809646606, | |
| "learning_rate": 7.540073078549221e-05, | |
| "loss": 0.2657, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.5845511482254697, | |
| "grad_norm": 0.5960420370101929, | |
| "learning_rate": 7.507965792532921e-05, | |
| "loss": 0.1952, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5855949895615866, | |
| "grad_norm": 0.5834378004074097, | |
| "learning_rate": 7.475885874114047e-05, | |
| "loss": 0.1878, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.5866388308977035, | |
| "grad_norm": 0.6201406121253967, | |
| "learning_rate": 7.443833675595255e-05, | |
| "loss": 0.1927, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.5876826722338204, | |
| "grad_norm": 0.5904473662376404, | |
| "learning_rate": 7.411809548974792e-05, | |
| "loss": 0.1804, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.5887265135699373, | |
| "grad_norm": 0.5292779803276062, | |
| "learning_rate": 7.379813845942623e-05, | |
| "loss": 0.1379, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.5897703549060542, | |
| "grad_norm": 0.618929922580719, | |
| "learning_rate": 7.347846917876544e-05, | |
| "loss": 0.1926, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.5908141962421712, | |
| "grad_norm": 0.5607888698577881, | |
| "learning_rate": 7.315909115838367e-05, | |
| "loss": 0.1845, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.5918580375782881, | |
| "grad_norm": 0.56803297996521, | |
| "learning_rate": 7.284000790570029e-05, | |
| "loss": 0.1762, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.592901878914405, | |
| "grad_norm": 0.5143932700157166, | |
| "learning_rate": 7.252122292489747e-05, | |
| "loss": 0.1514, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.593945720250522, | |
| "grad_norm": 0.6080281734466553, | |
| "learning_rate": 7.220273971688192e-05, | |
| "loss": 0.2516, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.5949895615866388, | |
| "grad_norm": 0.6721866130828857, | |
| "learning_rate": 7.188456177924605e-05, | |
| "loss": 0.2707, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5960334029227558, | |
| "grad_norm": 0.5116624236106873, | |
| "learning_rate": 7.156669260622996e-05, | |
| "loss": 0.2083, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.5970772442588727, | |
| "grad_norm": 0.5874140858650208, | |
| "learning_rate": 7.124913568868287e-05, | |
| "loss": 0.1598, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.5981210855949896, | |
| "grad_norm": 0.506820559501648, | |
| "learning_rate": 7.093189451402469e-05, | |
| "loss": 0.1572, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.5991649269311065, | |
| "grad_norm": 0.7033873796463013, | |
| "learning_rate": 7.061497256620793e-05, | |
| "loss": 0.2867, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.6002087682672234, | |
| "grad_norm": 0.7932460904121399, | |
| "learning_rate": 7.029837332567927e-05, | |
| "loss": 0.2998, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6012526096033403, | |
| "grad_norm": 0.8427619934082031, | |
| "learning_rate": 6.998210026934148e-05, | |
| "loss": 0.426, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.6022964509394572, | |
| "grad_norm": 0.7818666100502014, | |
| "learning_rate": 6.966615687051516e-05, | |
| "loss": 0.3559, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.6033402922755741, | |
| "grad_norm": 0.821897566318512, | |
| "learning_rate": 6.935054659890052e-05, | |
| "loss": 0.3928, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.6043841336116911, | |
| "grad_norm": 0.7375624179840088, | |
| "learning_rate": 6.903527292053942e-05, | |
| "loss": 0.3203, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.605427974947808, | |
| "grad_norm": 0.7323412299156189, | |
| "learning_rate": 6.87203392977773e-05, | |
| "loss": 0.3014, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6064718162839249, | |
| "grad_norm": 0.8238475322723389, | |
| "learning_rate": 6.840574918922493e-05, | |
| "loss": 0.3447, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.6075156576200418, | |
| "grad_norm": 0.7970190644264221, | |
| "learning_rate": 6.809150604972079e-05, | |
| "loss": 0.3556, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.6085594989561587, | |
| "grad_norm": 0.718948483467102, | |
| "learning_rate": 6.777761333029275e-05, | |
| "loss": 0.318, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.6096033402922756, | |
| "grad_norm": 0.8113385438919067, | |
| "learning_rate": 6.746407447812049e-05, | |
| "loss": 0.2928, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.6106471816283925, | |
| "grad_norm": 0.732028603553772, | |
| "learning_rate": 6.715089293649752e-05, | |
| "loss": 0.2688, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6116910229645094, | |
| "grad_norm": 0.693304181098938, | |
| "learning_rate": 6.683807214479323e-05, | |
| "loss": 0.2421, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.6127348643006263, | |
| "grad_norm": 0.8302125334739685, | |
| "learning_rate": 6.652561553841537e-05, | |
| "loss": 0.2625, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.6137787056367432, | |
| "grad_norm": 0.6446481943130493, | |
| "learning_rate": 6.621352654877207e-05, | |
| "loss": 0.2261, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.6148225469728601, | |
| "grad_norm": 0.7838292717933655, | |
| "learning_rate": 6.59018086032344e-05, | |
| "loss": 0.2619, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.615866388308977, | |
| "grad_norm": 0.6317050457000732, | |
| "learning_rate": 6.55904651250986e-05, | |
| "loss": 0.1996, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6169102296450939, | |
| "grad_norm": 0.6110920310020447, | |
| "learning_rate": 6.527949953354835e-05, | |
| "loss": 0.2078, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.6179540709812108, | |
| "grad_norm": 0.5425273180007935, | |
| "learning_rate": 6.496891524361757e-05, | |
| "loss": 0.1852, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.6189979123173278, | |
| "grad_norm": 0.7897228002548218, | |
| "learning_rate": 6.465871566615263e-05, | |
| "loss": 0.2217, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.6200417536534447, | |
| "grad_norm": 0.6448274254798889, | |
| "learning_rate": 6.434890420777491e-05, | |
| "loss": 0.1918, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.6210855949895616, | |
| "grad_norm": 0.690799355506897, | |
| "learning_rate": 6.403948427084356e-05, | |
| "loss": 0.2129, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.6221294363256785, | |
| "grad_norm": 0.6128472685813904, | |
| "learning_rate": 6.373045925341794e-05, | |
| "loss": 0.196, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.6231732776617954, | |
| "grad_norm": 0.4894169270992279, | |
| "learning_rate": 6.342183254922046e-05, | |
| "loss": 0.1341, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.6242171189979123, | |
| "grad_norm": 0.5475450754165649, | |
| "learning_rate": 6.311360754759923e-05, | |
| "loss": 0.1655, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.6252609603340292, | |
| "grad_norm": 0.5742066502571106, | |
| "learning_rate": 6.280578763349078e-05, | |
| "loss": 0.2158, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.6263048016701461, | |
| "grad_norm": 0.6302378177642822, | |
| "learning_rate": 6.249837618738311e-05, | |
| "loss": 0.1211, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.627348643006263, | |
| "grad_norm": 0.6498920321464539, | |
| "learning_rate": 6.219137658527818e-05, | |
| "loss": 0.3415, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.6283924843423799, | |
| "grad_norm": 0.5803071856498718, | |
| "learning_rate": 6.188479219865529e-05, | |
| "loss": 0.3446, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.6294363256784968, | |
| "grad_norm": 0.7242146134376526, | |
| "learning_rate": 6.157862639443374e-05, | |
| "loss": 0.3406, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.6304801670146137, | |
| "grad_norm": 0.741543710231781, | |
| "learning_rate": 6.127288253493591e-05, | |
| "loss": 0.2851, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.6315240083507306, | |
| "grad_norm": 0.6710807681083679, | |
| "learning_rate": 6.09675639778504e-05, | |
| "loss": 0.2975, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6325678496868476, | |
| "grad_norm": 0.6401992440223694, | |
| "learning_rate": 6.0662674076195194e-05, | |
| "loss": 0.2934, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.6336116910229646, | |
| "grad_norm": 0.7866775393486023, | |
| "learning_rate": 6.03582161782806e-05, | |
| "loss": 0.3303, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.6346555323590815, | |
| "grad_norm": 0.6878888607025146, | |
| "learning_rate": 6.005419362767286e-05, | |
| "loss": 0.2885, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.6356993736951984, | |
| "grad_norm": 0.667226254940033, | |
| "learning_rate": 5.975060976315703e-05, | |
| "loss": 0.2663, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.6367432150313153, | |
| "grad_norm": 0.5909189581871033, | |
| "learning_rate": 5.9447467918700614e-05, | |
| "loss": 0.2023, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6377870563674322, | |
| "grad_norm": 0.6986932158470154, | |
| "learning_rate": 5.9144771423416826e-05, | |
| "loss": 0.2354, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.6388308977035491, | |
| "grad_norm": 0.5562401413917542, | |
| "learning_rate": 5.8842523601528e-05, | |
| "loss": 0.1928, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.639874739039666, | |
| "grad_norm": 0.5309166312217712, | |
| "learning_rate": 5.854072777232914e-05, | |
| "loss": 0.1611, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.6409185803757829, | |
| "grad_norm": 0.4029114842414856, | |
| "learning_rate": 5.823938725015148e-05, | |
| "loss": 0.14, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.6419624217118998, | |
| "grad_norm": 0.4246949851512909, | |
| "learning_rate": 5.793850534432599e-05, | |
| "loss": 0.1194, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6430062630480167, | |
| "grad_norm": 0.44221794605255127, | |
| "learning_rate": 5.763808535914723e-05, | |
| "loss": 0.121, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.6440501043841336, | |
| "grad_norm": 0.40379002690315247, | |
| "learning_rate": 5.7338130593836755e-05, | |
| "loss": 0.1215, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.6450939457202505, | |
| "grad_norm": 0.6201443076133728, | |
| "learning_rate": 5.7038644342507205e-05, | |
| "loss": 0.1686, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.6461377870563675, | |
| "grad_norm": 0.7134044766426086, | |
| "learning_rate": 5.673962989412599e-05, | |
| "loss": 0.3048, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.6471816283924844, | |
| "grad_norm": 0.5085525512695312, | |
| "learning_rate": 5.644109053247901e-05, | |
| "loss": 0.2137, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6482254697286013, | |
| "grad_norm": 0.5813112854957581, | |
| "learning_rate": 5.614302953613489e-05, | |
| "loss": 0.2164, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.6492693110647182, | |
| "grad_norm": 0.5314549803733826, | |
| "learning_rate": 5.584545017840885e-05, | |
| "loss": 0.1781, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.6503131524008351, | |
| "grad_norm": 0.4466283619403839, | |
| "learning_rate": 5.5548355727326574e-05, | |
| "loss": 0.1564, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.651356993736952, | |
| "grad_norm": 0.7003150582313538, | |
| "learning_rate": 5.525174944558866e-05, | |
| "loss": 0.3604, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.6524008350730689, | |
| "grad_norm": 0.7904312610626221, | |
| "learning_rate": 5.4955634590534545e-05, | |
| "loss": 0.3483, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6534446764091858, | |
| "grad_norm": 0.7673099637031555, | |
| "learning_rate": 5.466001441410682e-05, | |
| "loss": 0.3912, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.6544885177453027, | |
| "grad_norm": 0.823867678642273, | |
| "learning_rate": 5.4364892162815436e-05, | |
| "loss": 0.3618, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.6555323590814196, | |
| "grad_norm": 0.6855948567390442, | |
| "learning_rate": 5.407027107770219e-05, | |
| "loss": 0.2816, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.6565762004175365, | |
| "grad_norm": 0.7723731994628906, | |
| "learning_rate": 5.377615439430508e-05, | |
| "loss": 0.3292, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.6576200417536534, | |
| "grad_norm": 0.7081869840621948, | |
| "learning_rate": 5.348254534262262e-05, | |
| "loss": 0.3232, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6586638830897703, | |
| "grad_norm": 0.7101826071739197, | |
| "learning_rate": 5.318944714707861e-05, | |
| "loss": 0.3557, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.6597077244258872, | |
| "grad_norm": 0.7043560147285461, | |
| "learning_rate": 5.289686302648661e-05, | |
| "loss": 0.3251, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.6607515657620042, | |
| "grad_norm": 0.8865169286727905, | |
| "learning_rate": 5.2604796194014507e-05, | |
| "loss": 0.3514, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.6617954070981211, | |
| "grad_norm": 0.8106626868247986, | |
| "learning_rate": 5.2313249857149414e-05, | |
| "loss": 0.3226, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.662839248434238, | |
| "grad_norm": 0.7511535286903381, | |
| "learning_rate": 5.202222721766226e-05, | |
| "loss": 0.3186, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6638830897703549, | |
| "grad_norm": 0.7110910415649414, | |
| "learning_rate": 5.1731731471572755e-05, | |
| "loss": 0.29, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.6649269311064718, | |
| "grad_norm": 0.7598642110824585, | |
| "learning_rate": 5.144176580911431e-05, | |
| "loss": 0.2552, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.6659707724425887, | |
| "grad_norm": 0.5399108529090881, | |
| "learning_rate": 5.115233341469877e-05, | |
| "loss": 0.2105, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.6670146137787056, | |
| "grad_norm": 0.6335327625274658, | |
| "learning_rate": 5.0863437466881836e-05, | |
| "loss": 0.2272, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.6680584551148225, | |
| "grad_norm": 0.6470877528190613, | |
| "learning_rate": 5.0575081138327715e-05, | |
| "loss": 0.2329, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6691022964509394, | |
| "grad_norm": 0.5501940250396729, | |
| "learning_rate": 5.028726759577467e-05, | |
| "loss": 0.2057, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.6701461377870563, | |
| "grad_norm": 0.48475509881973267, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.1643, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.6711899791231732, | |
| "grad_norm": 0.8148300051689148, | |
| "learning_rate": 4.97132815057854e-05, | |
| "loss": 0.2454, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.6722338204592901, | |
| "grad_norm": 0.47534969449043274, | |
| "learning_rate": 4.942711526188229e-05, | |
| "loss": 0.1713, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.673277661795407, | |
| "grad_norm": 0.5151733160018921, | |
| "learning_rate": 4.914150441097736e-05, | |
| "loss": 0.1701, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6743215031315241, | |
| "grad_norm": 0.5037069916725159, | |
| "learning_rate": 4.885645208965779e-05, | |
| "loss": 0.1814, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.675365344467641, | |
| "grad_norm": 0.4882695972919464, | |
| "learning_rate": 4.857196142837716e-05, | |
| "loss": 0.1685, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.6764091858037579, | |
| "grad_norm": 0.614020586013794, | |
| "learning_rate": 4.8288035551420697e-05, | |
| "loss": 0.1878, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.6774530271398748, | |
| "grad_norm": 0.44983476400375366, | |
| "learning_rate": 4.80046775768713e-05, | |
| "loss": 0.1455, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.6784968684759917, | |
| "grad_norm": 0.7257928252220154, | |
| "learning_rate": 4.7721890616575103e-05, | |
| "loss": 0.1417, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6795407098121086, | |
| "grad_norm": 0.5455628633499146, | |
| "learning_rate": 4.743967777610727e-05, | |
| "loss": 0.298, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.6805845511482255, | |
| "grad_norm": 0.6038815379142761, | |
| "learning_rate": 4.715804215473809e-05, | |
| "loss": 0.322, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.6816283924843424, | |
| "grad_norm": 0.6264936327934265, | |
| "learning_rate": 4.687698684539866e-05, | |
| "loss": 0.3185, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.6826722338204593, | |
| "grad_norm": 0.6083415150642395, | |
| "learning_rate": 4.659651493464721e-05, | |
| "loss": 0.2614, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.6837160751565762, | |
| "grad_norm": 0.616062343120575, | |
| "learning_rate": 4.6316629502635025e-05, | |
| "loss": 0.2663, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.6847599164926931, | |
| "grad_norm": 0.6474526524543762, | |
| "learning_rate": 4.603733362307261e-05, | |
| "loss": 0.3216, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.68580375782881, | |
| "grad_norm": 0.6374465227127075, | |
| "learning_rate": 4.575863036319604e-05, | |
| "loss": 0.2995, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.6868475991649269, | |
| "grad_norm": 0.627227246761322, | |
| "learning_rate": 4.548052278373327e-05, | |
| "loss": 0.2399, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.6878914405010439, | |
| "grad_norm": 0.6535525321960449, | |
| "learning_rate": 4.520301393887032e-05, | |
| "loss": 0.2309, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.6889352818371608, | |
| "grad_norm": 0.5703950524330139, | |
| "learning_rate": 4.492610687621804e-05, | |
| "loss": 0.2128, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6899791231732777, | |
| "grad_norm": 0.5039522647857666, | |
| "learning_rate": 4.4649804636778456e-05, | |
| "loss": 0.1812, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.6910229645093946, | |
| "grad_norm": 0.6118776798248291, | |
| "learning_rate": 4.4374110254911306e-05, | |
| "loss": 0.225, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.6920668058455115, | |
| "grad_norm": 0.4194796085357666, | |
| "learning_rate": 4.4099026758300944e-05, | |
| "loss": 0.1415, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.6931106471816284, | |
| "grad_norm": 0.5777245163917542, | |
| "learning_rate": 4.382455716792291e-05, | |
| "loss": 0.2032, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.6941544885177453, | |
| "grad_norm": 0.41527435183525085, | |
| "learning_rate": 4.355070449801083e-05, | |
| "loss": 0.1205, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.6951983298538622, | |
| "grad_norm": 0.502178430557251, | |
| "learning_rate": 4.32774717560232e-05, | |
| "loss": 0.1514, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.6962421711899791, | |
| "grad_norm": 0.5909983515739441, | |
| "learning_rate": 4.300486194261057e-05, | |
| "loss": 0.176, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.697286012526096, | |
| "grad_norm": 0.48378539085388184, | |
| "learning_rate": 4.273287805158245e-05, | |
| "loss": 0.1627, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.6983298538622129, | |
| "grad_norm": 0.6273384094238281, | |
| "learning_rate": 4.2461523069874346e-05, | |
| "loss": 0.2523, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.6993736951983298, | |
| "grad_norm": 0.5280055403709412, | |
| "learning_rate": 4.219079997751515e-05, | |
| "loss": 0.2035, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7004175365344467, | |
| "grad_norm": 0.539364755153656, | |
| "learning_rate": 4.192071174759435e-05, | |
| "loss": 0.1896, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.7014613778705637, | |
| "grad_norm": 0.5403777956962585, | |
| "learning_rate": 4.165126134622926e-05, | |
| "loss": 0.1624, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.7025052192066806, | |
| "grad_norm": 0.448177695274353, | |
| "learning_rate": 4.1382451732532665e-05, | |
| "loss": 0.1308, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.7035490605427975, | |
| "grad_norm": 0.6774344444274902, | |
| "learning_rate": 4.1114285858580045e-05, | |
| "loss": 0.3161, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.7045929018789144, | |
| "grad_norm": 0.7080472707748413, | |
| "learning_rate": 4.0846766669377446e-05, | |
| "loss": 0.3357, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7056367432150313, | |
| "grad_norm": 0.6589325666427612, | |
| "learning_rate": 4.0579897102828966e-05, | |
| "loss": 0.2815, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.7066805845511482, | |
| "grad_norm": 0.7149707078933716, | |
| "learning_rate": 4.0313680089704454e-05, | |
| "loss": 0.3612, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.7077244258872651, | |
| "grad_norm": 0.6252415776252747, | |
| "learning_rate": 4.004811855360748e-05, | |
| "loss": 0.2726, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.708768267223382, | |
| "grad_norm": 0.7726844549179077, | |
| "learning_rate": 3.9783215410943174e-05, | |
| "loss": 0.3229, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.7098121085594989, | |
| "grad_norm": 0.7369757890701294, | |
| "learning_rate": 3.951897357088602e-05, | |
| "loss": 0.3436, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7108559498956158, | |
| "grad_norm": 0.788517951965332, | |
| "learning_rate": 3.925539593534824e-05, | |
| "loss": 0.312, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.7118997912317327, | |
| "grad_norm": 0.7983633875846863, | |
| "learning_rate": 3.899248539894757e-05, | |
| "loss": 0.361, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.7129436325678496, | |
| "grad_norm": 0.7084015607833862, | |
| "learning_rate": 3.873024484897576e-05, | |
| "loss": 0.2836, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.7139874739039666, | |
| "grad_norm": 0.7584156394004822, | |
| "learning_rate": 3.8468677165366754e-05, | |
| "loss": 0.2955, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.7150313152400835, | |
| "grad_norm": 0.7465482354164124, | |
| "learning_rate": 3.820778522066494e-05, | |
| "loss": 0.2564, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.7160751565762005, | |
| "grad_norm": 0.6189156770706177, | |
| "learning_rate": 3.794757187999386e-05, | |
| "loss": 0.221, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.7171189979123174, | |
| "grad_norm": 0.6960480809211731, | |
| "learning_rate": 3.7688040001024475e-05, | |
| "loss": 0.2522, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.7181628392484343, | |
| "grad_norm": 0.572296679019928, | |
| "learning_rate": 3.7429192433944014e-05, | |
| "loss": 0.1997, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.7192066805845512, | |
| "grad_norm": 0.6478104591369629, | |
| "learning_rate": 3.717103202142457e-05, | |
| "loss": 0.2126, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.7202505219206681, | |
| "grad_norm": 0.7389695644378662, | |
| "learning_rate": 3.691356159859177e-05, | |
| "loss": 0.2333, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.721294363256785, | |
| "grad_norm": 0.5489468574523926, | |
| "learning_rate": 3.665678399299388e-05, | |
| "loss": 0.1701, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.7223382045929019, | |
| "grad_norm": 0.7258986830711365, | |
| "learning_rate": 3.64007020245706e-05, | |
| "loss": 0.2457, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.7233820459290188, | |
| "grad_norm": 0.6785321235656738, | |
| "learning_rate": 3.614531850562203e-05, | |
| "loss": 0.1936, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.7244258872651357, | |
| "grad_norm": 0.612426221370697, | |
| "learning_rate": 3.589063624077802e-05, | |
| "loss": 0.2403, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.7254697286012526, | |
| "grad_norm": 0.5947864651679993, | |
| "learning_rate": 3.563665802696707e-05, | |
| "loss": 0.1743, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7265135699373695, | |
| "grad_norm": 0.6579543352127075, | |
| "learning_rate": 3.538338665338589e-05, | |
| "loss": 0.1928, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.7275574112734864, | |
| "grad_norm": 0.5460782647132874, | |
| "learning_rate": 3.513082490146864e-05, | |
| "loss": 0.1655, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.7286012526096033, | |
| "grad_norm": 0.7640422582626343, | |
| "learning_rate": 3.487897554485628e-05, | |
| "loss": 0.1659, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.7296450939457203, | |
| "grad_norm": 0.7361250519752502, | |
| "learning_rate": 3.462784134936636e-05, | |
| "loss": 0.1992, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.7306889352818372, | |
| "grad_norm": 1.2653623819351196, | |
| "learning_rate": 3.4377425072962465e-05, | |
| "loss": 0.1184, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7317327766179541, | |
| "grad_norm": 0.6173591613769531, | |
| "learning_rate": 3.412772946572389e-05, | |
| "loss": 0.4121, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.732776617954071, | |
| "grad_norm": 0.5525224208831787, | |
| "learning_rate": 3.387875726981563e-05, | |
| "loss": 0.2601, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.7338204592901879, | |
| "grad_norm": 0.6985558867454529, | |
| "learning_rate": 3.363051121945809e-05, | |
| "loss": 0.3448, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.7348643006263048, | |
| "grad_norm": 0.580680251121521, | |
| "learning_rate": 3.3382994040897196e-05, | |
| "loss": 0.2642, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.7359081419624217, | |
| "grad_norm": 0.5876568555831909, | |
| "learning_rate": 3.3136208452374254e-05, | |
| "loss": 0.271, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7369519832985386, | |
| "grad_norm": 0.6181269884109497, | |
| "learning_rate": 3.289015716409631e-05, | |
| "loss": 0.2522, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.7379958246346555, | |
| "grad_norm": 0.6312392354011536, | |
| "learning_rate": 3.264484287820634e-05, | |
| "loss": 0.2735, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.7390396659707724, | |
| "grad_norm": 0.6163091063499451, | |
| "learning_rate": 3.2400268288753425e-05, | |
| "loss": 0.2415, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.7400835073068893, | |
| "grad_norm": 0.6303150057792664, | |
| "learning_rate": 3.2156436081663356e-05, | |
| "loss": 0.2495, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.7411273486430062, | |
| "grad_norm": 0.6544148325920105, | |
| "learning_rate": 3.191334893470907e-05, | |
| "loss": 0.2445, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7421711899791231, | |
| "grad_norm": 0.468227744102478, | |
| "learning_rate": 3.167100951748115e-05, | |
| "loss": 0.1481, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.7432150313152401, | |
| "grad_norm": 0.5340932607650757, | |
| "learning_rate": 3.14294204913587e-05, | |
| "loss": 0.1431, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.744258872651357, | |
| "grad_norm": 0.5851957201957703, | |
| "learning_rate": 3.1188584509479866e-05, | |
| "loss": 0.1737, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.7453027139874739, | |
| "grad_norm": 0.5840248465538025, | |
| "learning_rate": 3.094850421671295e-05, | |
| "loss": 0.1752, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.7463465553235908, | |
| "grad_norm": 0.5659369826316833, | |
| "learning_rate": 3.0709182249627255e-05, | |
| "loss": 0.1967, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7473903966597077, | |
| "grad_norm": 0.457015722990036, | |
| "learning_rate": 3.0470621236464036e-05, | |
| "loss": 0.1544, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.7484342379958246, | |
| "grad_norm": 0.5182324647903442, | |
| "learning_rate": 3.023282379710779e-05, | |
| "loss": 0.1414, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.7494780793319415, | |
| "grad_norm": 0.5334721207618713, | |
| "learning_rate": 2.9995792543057478e-05, | |
| "loss": 0.1615, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.7505219206680585, | |
| "grad_norm": 0.6061464548110962, | |
| "learning_rate": 2.9759530077397636e-05, | |
| "loss": 0.2452, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.7515657620041754, | |
| "grad_norm": 0.5774762630462646, | |
| "learning_rate": 2.9524038994770107e-05, | |
| "loss": 0.2234, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7515657620041754, | |
| "eval_loss": 0.21803000569343567, | |
| "eval_runtime": 81.2533, | |
| "eval_samples_per_second": 19.864, | |
| "eval_steps_per_second": 9.932, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7526096033402923, | |
| "grad_norm": 0.43510791659355164, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 0.1719, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.7536534446764092, | |
| "grad_norm": 0.4775535762310028, | |
| "learning_rate": 2.905538131479376e-05, | |
| "loss": 0.1874, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.7546972860125261, | |
| "grad_norm": 0.5398250818252563, | |
| "learning_rate": 2.8822219864258272e-05, | |
| "loss": 0.1667, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.755741127348643, | |
| "grad_norm": 0.4013484716415405, | |
| "learning_rate": 2.8589840090325027e-05, | |
| "loss": 0.1252, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.7567849686847599, | |
| "grad_norm": 0.8314927816390991, | |
| "learning_rate": 2.8358244544996038e-05, | |
| "loss": 0.3924, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7578288100208769, | |
| "grad_norm": 0.8482892513275146, | |
| "learning_rate": 2.8127435771660747e-05, | |
| "loss": 0.3675, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.7588726513569938, | |
| "grad_norm": 0.6500058770179749, | |
| "learning_rate": 2.7897416305068323e-05, | |
| "loss": 0.3016, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.7599164926931107, | |
| "grad_norm": 0.6976576447486877, | |
| "learning_rate": 2.7668188671299755e-05, | |
| "loss": 0.3133, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.7609603340292276, | |
| "grad_norm": 0.7735400795936584, | |
| "learning_rate": 2.743975538774002e-05, | |
| "loss": 0.3557, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.7620041753653445, | |
| "grad_norm": 0.7905307412147522, | |
| "learning_rate": 2.7212118963050592e-05, | |
| "loss": 0.3316, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7630480167014614, | |
| "grad_norm": 0.7098974585533142, | |
| "learning_rate": 2.6985281897141812e-05, | |
| "loss": 0.2877, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.7640918580375783, | |
| "grad_norm": 0.6374404430389404, | |
| "learning_rate": 2.675924668114537e-05, | |
| "loss": 0.2589, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.7651356993736952, | |
| "grad_norm": 0.7561594247817993, | |
| "learning_rate": 2.65340157973871e-05, | |
| "loss": 0.2953, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.7661795407098121, | |
| "grad_norm": 0.7232580780982971, | |
| "learning_rate": 2.630959171935956e-05, | |
| "loss": 0.3002, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.767223382045929, | |
| "grad_norm": 0.7140358686447144, | |
| "learning_rate": 2.6085976911694987e-05, | |
| "loss": 0.2545, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.7682672233820459, | |
| "grad_norm": 0.749450147151947, | |
| "learning_rate": 2.586317383013821e-05, | |
| "loss": 0.25, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.7693110647181628, | |
| "grad_norm": 0.847427248954773, | |
| "learning_rate": 2.564118492151957e-05, | |
| "loss": 0.2887, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.7703549060542797, | |
| "grad_norm": 0.5772815346717834, | |
| "learning_rate": 2.5420012623728208e-05, | |
| "loss": 0.2101, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.7713987473903967, | |
| "grad_norm": 0.6374946236610413, | |
| "learning_rate": 2.5199659365685235e-05, | |
| "loss": 0.2193, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.7724425887265136, | |
| "grad_norm": 0.642776370048523, | |
| "learning_rate": 2.4980127567316948e-05, | |
| "loss": 0.2168, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7734864300626305, | |
| "grad_norm": 0.6233210563659668, | |
| "learning_rate": 2.4761419639528437e-05, | |
| "loss": 0.1986, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.7745302713987474, | |
| "grad_norm": 0.47856077551841736, | |
| "learning_rate": 2.4543537984176978e-05, | |
| "loss": 0.1681, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.7755741127348643, | |
| "grad_norm": 0.5664119720458984, | |
| "learning_rate": 2.4326484994045752e-05, | |
| "loss": 0.1847, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.7766179540709812, | |
| "grad_norm": 0.5382654666900635, | |
| "learning_rate": 2.4110263052817394e-05, | |
| "loss": 0.1765, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.7776617954070981, | |
| "grad_norm": 0.46047693490982056, | |
| "learning_rate": 2.3894874535048063e-05, | |
| "loss": 0.1487, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.778705636743215, | |
| "grad_norm": 0.5149843096733093, | |
| "learning_rate": 2.368032180614118e-05, | |
| "loss": 0.1817, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.7797494780793319, | |
| "grad_norm": 0.4655948579311371, | |
| "learning_rate": 2.346660722232148e-05, | |
| "loss": 0.1548, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.7807933194154488, | |
| "grad_norm": 0.5313979387283325, | |
| "learning_rate": 2.325373313060919e-05, | |
| "loss": 0.1504, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.7818371607515657, | |
| "grad_norm": 0.5336787700653076, | |
| "learning_rate": 2.3041701868794287e-05, | |
| "loss": 0.1306, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.7828810020876826, | |
| "grad_norm": 0.7393002510070801, | |
| "learning_rate": 2.2830515765410622e-05, | |
| "loss": 0.1137, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7839248434237995, | |
| "grad_norm": 0.5598475933074951, | |
| "learning_rate": 2.262017713971063e-05, | |
| "loss": 0.3354, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.7849686847599165, | |
| "grad_norm": 0.5945207476615906, | |
| "learning_rate": 2.2410688301639616e-05, | |
| "loss": 0.2804, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.7860125260960334, | |
| "grad_norm": 0.4905988574028015, | |
| "learning_rate": 2.2202051551810565e-05, | |
| "loss": 0.2229, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.7870563674321504, | |
| "grad_norm": 0.607475996017456, | |
| "learning_rate": 2.19942691814788e-05, | |
| "loss": 0.289, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.7881002087682673, | |
| "grad_norm": 0.6597141027450562, | |
| "learning_rate": 2.178734347251673e-05, | |
| "loss": 0.2965, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.7891440501043842, | |
| "grad_norm": 0.6289554238319397, | |
| "learning_rate": 2.1581276697388975e-05, | |
| "loss": 0.2582, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.7901878914405011, | |
| "grad_norm": 0.6147776246070862, | |
| "learning_rate": 2.1376071119127338e-05, | |
| "loss": 0.2848, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.791231732776618, | |
| "grad_norm": 0.5135255455970764, | |
| "learning_rate": 2.1171728991305795e-05, | |
| "loss": 0.2306, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.7922755741127349, | |
| "grad_norm": 0.5475291013717651, | |
| "learning_rate": 2.0968252558016055e-05, | |
| "loss": 0.2394, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.7933194154488518, | |
| "grad_norm": 0.6322019696235657, | |
| "learning_rate": 2.076564405384258e-05, | |
| "loss": 0.2066, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7943632567849687, | |
| "grad_norm": 0.5838301181793213, | |
| "learning_rate": 2.0563905703838316e-05, | |
| "loss": 0.2321, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.7954070981210856, | |
| "grad_norm": 0.47901853919029236, | |
| "learning_rate": 2.0363039723500156e-05, | |
| "loss": 0.1445, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.7964509394572025, | |
| "grad_norm": 0.6136653423309326, | |
| "learning_rate": 2.0163048318744493e-05, | |
| "loss": 0.2168, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.7974947807933194, | |
| "grad_norm": 0.49966001510620117, | |
| "learning_rate": 1.9963933685883253e-05, | |
| "loss": 0.1414, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.7985386221294363, | |
| "grad_norm": 0.5253435373306274, | |
| "learning_rate": 1.9765698011599466e-05, | |
| "loss": 0.1513, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.7995824634655533, | |
| "grad_norm": 0.3400777578353882, | |
| "learning_rate": 1.9568343472923524e-05, | |
| "loss": 0.1112, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.8006263048016702, | |
| "grad_norm": 0.5851226449012756, | |
| "learning_rate": 1.9371872237209165e-05, | |
| "loss": 0.1619, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.8016701461377871, | |
| "grad_norm": 0.30553382635116577, | |
| "learning_rate": 1.917628646210957e-05, | |
| "loss": 0.0872, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.802713987473904, | |
| "grad_norm": 0.5676819682121277, | |
| "learning_rate": 1.8981588295553853e-05, | |
| "loss": 0.2251, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.8037578288100209, | |
| "grad_norm": 0.5648460984230042, | |
| "learning_rate": 1.878777987572339e-05, | |
| "loss": 0.2181, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.8048016701461378, | |
| "grad_norm": 0.607913076877594, | |
| "learning_rate": 1.8594863331028224e-05, | |
| "loss": 0.2216, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.8058455114822547, | |
| "grad_norm": 0.49821627140045166, | |
| "learning_rate": 1.840284078008393e-05, | |
| "loss": 0.1845, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.8068893528183716, | |
| "grad_norm": 0.42891865968704224, | |
| "learning_rate": 1.821171433168809e-05, | |
| "loss": 0.1302, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.8079331941544885, | |
| "grad_norm": 0.8176518678665161, | |
| "learning_rate": 1.8021486084797368e-05, | |
| "loss": 0.3043, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.8089770354906054, | |
| "grad_norm": 0.6164413690567017, | |
| "learning_rate": 1.7832158128504328e-05, | |
| "loss": 0.2472, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8100208768267223, | |
| "grad_norm": 0.7422143220901489, | |
| "learning_rate": 1.7643732542014434e-05, | |
| "loss": 0.3395, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.8110647181628392, | |
| "grad_norm": 0.7129305601119995, | |
| "learning_rate": 1.7456211394623378e-05, | |
| "loss": 0.2934, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.8121085594989561, | |
| "grad_norm": 0.7638242840766907, | |
| "learning_rate": 1.7269596745694295e-05, | |
| "loss": 0.3399, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.8131524008350731, | |
| "grad_norm": 0.624290943145752, | |
| "learning_rate": 1.7083890644635014e-05, | |
| "loss": 0.2525, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.81419624217119, | |
| "grad_norm": 0.6761390566825867, | |
| "learning_rate": 1.6899095130875774e-05, | |
| "loss": 0.2753, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.8152400835073069, | |
| "grad_norm": 0.8017570972442627, | |
| "learning_rate": 1.6715212233846655e-05, | |
| "loss": 0.3149, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.8162839248434238, | |
| "grad_norm": 0.6984527111053467, | |
| "learning_rate": 1.6532243972955398e-05, | |
| "loss": 0.3103, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.8173277661795407, | |
| "grad_norm": 0.6139991283416748, | |
| "learning_rate": 1.635019235756511e-05, | |
| "loss": 0.2274, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.8183716075156576, | |
| "grad_norm": 0.7596770524978638, | |
| "learning_rate": 1.616905938697234e-05, | |
| "loss": 0.2612, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.8194154488517745, | |
| "grad_norm": 0.685114324092865, | |
| "learning_rate": 1.5988847050385037e-05, | |
| "loss": 0.2542, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.8204592901878914, | |
| "grad_norm": 0.6803342700004578, | |
| "learning_rate": 1.5809557326900647e-05, | |
| "loss": 0.2267, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.8215031315240083, | |
| "grad_norm": 0.6194560527801514, | |
| "learning_rate": 1.5631192185484554e-05, | |
| "loss": 0.2029, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.8225469728601252, | |
| "grad_norm": 0.8063942193984985, | |
| "learning_rate": 1.5453753584948328e-05, | |
| "loss": 0.2735, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.8235908141962421, | |
| "grad_norm": 0.6718615293502808, | |
| "learning_rate": 1.527724347392815e-05, | |
| "loss": 0.2278, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.824634655532359, | |
| "grad_norm": 0.6082893013954163, | |
| "learning_rate": 1.5101663790863596e-05, | |
| "loss": 0.1969, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.8256784968684759, | |
| "grad_norm": 0.7025957703590393, | |
| "learning_rate": 1.4927016463976263e-05, | |
| "loss": 0.2071, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.826722338204593, | |
| "grad_norm": 0.5405511260032654, | |
| "learning_rate": 1.4753303411248475e-05, | |
| "loss": 0.191, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.8277661795407099, | |
| "grad_norm": 0.6413715481758118, | |
| "learning_rate": 1.4580526540402461e-05, | |
| "loss": 0.185, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.8288100208768268, | |
| "grad_norm": 0.6367815136909485, | |
| "learning_rate": 1.4408687748879156e-05, | |
| "loss": 0.221, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.8298538622129437, | |
| "grad_norm": 0.4684351086616516, | |
| "learning_rate": 1.4237788923817553e-05, | |
| "loss": 0.1415, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.8308977035490606, | |
| "grad_norm": 0.5200543403625488, | |
| "learning_rate": 1.4067831942033904e-05, | |
| "loss": 0.1648, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.8319415448851775, | |
| "grad_norm": 0.44216129183769226, | |
| "learning_rate": 1.3898818670001034e-05, | |
| "loss": 0.1608, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.8329853862212944, | |
| "grad_norm": 0.5650377869606018, | |
| "learning_rate": 1.3730750963828032e-05, | |
| "loss": 0.1585, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.8340292275574113, | |
| "grad_norm": 0.5171220898628235, | |
| "learning_rate": 1.3563630669239624e-05, | |
| "loss": 0.1575, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.8350730688935282, | |
| "grad_norm": 0.5441738367080688, | |
| "learning_rate": 1.339745962155613e-05, | |
| "loss": 0.0949, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8361169102296451, | |
| "grad_norm": 0.6055110096931458, | |
| "learning_rate": 1.3232239645673217e-05, | |
| "loss": 0.4181, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.837160751565762, | |
| "grad_norm": 0.6455709338188171, | |
| "learning_rate": 1.3067972556041752e-05, | |
| "loss": 0.3383, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.8382045929018789, | |
| "grad_norm": 0.5849418640136719, | |
| "learning_rate": 1.2904660156648074e-05, | |
| "loss": 0.2803, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.8392484342379958, | |
| "grad_norm": 0.538429856300354, | |
| "learning_rate": 1.2742304240994053e-05, | |
| "loss": 0.2402, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.8402922755741128, | |
| "grad_norm": 0.6266717314720154, | |
| "learning_rate": 1.2580906592077402e-05, | |
| "loss": 0.3009, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.8413361169102297, | |
| "grad_norm": 0.5892521142959595, | |
| "learning_rate": 1.2420468982372158e-05, | |
| "loss": 0.249, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.8423799582463466, | |
| "grad_norm": 0.5853463411331177, | |
| "learning_rate": 1.226099317380912e-05, | |
| "loss": 0.2465, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.8434237995824635, | |
| "grad_norm": 0.640910267829895, | |
| "learning_rate": 1.210248091775663e-05, | |
| "loss": 0.1872, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.8444676409185804, | |
| "grad_norm": 0.6560840606689453, | |
| "learning_rate": 1.1944933955001225e-05, | |
| "loss": 0.2542, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.8455114822546973, | |
| "grad_norm": 0.5866312980651855, | |
| "learning_rate": 1.1788354015728543e-05, | |
| "loss": 0.1911, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8465553235908142, | |
| "grad_norm": 0.6191656589508057, | |
| "learning_rate": 1.1632742819504405e-05, | |
| "loss": 0.2215, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.8475991649269311, | |
| "grad_norm": 0.5251643657684326, | |
| "learning_rate": 1.147810207525577e-05, | |
| "loss": 0.1583, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.848643006263048, | |
| "grad_norm": 0.4927600622177124, | |
| "learning_rate": 1.132443348125214e-05, | |
| "loss": 0.1512, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.8496868475991649, | |
| "grad_norm": 0.4139147698879242, | |
| "learning_rate": 1.1171738725086833e-05, | |
| "loss": 0.1172, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.8507306889352818, | |
| "grad_norm": 0.5602164268493652, | |
| "learning_rate": 1.1020019483658384e-05, | |
| "loss": 0.1821, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.8517745302713987, | |
| "grad_norm": 0.4220430850982666, | |
| "learning_rate": 1.0869277423152246e-05, | |
| "loss": 0.1263, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.8528183716075156, | |
| "grad_norm": 0.38943833112716675, | |
| "learning_rate": 1.0719514199022473e-05, | |
| "loss": 0.1141, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.8538622129436325, | |
| "grad_norm": 0.5523675084114075, | |
| "learning_rate": 1.0570731455973414e-05, | |
| "loss": 0.1578, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.8549060542797495, | |
| "grad_norm": 0.6071298718452454, | |
| "learning_rate": 1.04229308279418e-05, | |
| "loss": 0.2323, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.8559498956158664, | |
| "grad_norm": 0.5393807291984558, | |
| "learning_rate": 1.0276113938078769e-05, | |
| "loss": 0.1967, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8569937369519833, | |
| "grad_norm": 0.6537972688674927, | |
| "learning_rate": 1.0130282398731982e-05, | |
| "loss": 0.1953, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.8580375782881002, | |
| "grad_norm": 0.4417877197265625, | |
| "learning_rate": 9.985437811427933e-06, | |
| "loss": 0.1453, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.8590814196242171, | |
| "grad_norm": 0.6434723734855652, | |
| "learning_rate": 9.841581766854401e-06, | |
| "loss": 0.1683, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.860125260960334, | |
| "grad_norm": 0.7254697680473328, | |
| "learning_rate": 9.698715844842988e-06, | |
| "loss": 0.3499, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.8611691022964509, | |
| "grad_norm": 0.6909394264221191, | |
| "learning_rate": 9.556841614351664e-06, | |
| "loss": 0.2757, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8622129436325678, | |
| "grad_norm": 0.7137805223464966, | |
| "learning_rate": 9.415960633447674e-06, | |
| "loss": 0.3122, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.8632567849686847, | |
| "grad_norm": 0.7450171709060669, | |
| "learning_rate": 9.276074449290361e-06, | |
| "loss": 0.2928, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.8643006263048016, | |
| "grad_norm": 0.6502891778945923, | |
| "learning_rate": 9.137184598114134e-06, | |
| "loss": 0.2517, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.8653444676409185, | |
| "grad_norm": 0.6984295845031738, | |
| "learning_rate": 8.999292605211695e-06, | |
| "loss": 0.3062, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.8663883089770354, | |
| "grad_norm": 0.7169867753982544, | |
| "learning_rate": 8.862399984917213e-06, | |
| "loss": 0.2748, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8674321503131524, | |
| "grad_norm": 0.7259141802787781, | |
| "learning_rate": 8.726508240589692e-06, | |
| "loss": 0.3033, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.8684759916492694, | |
| "grad_norm": 0.7949566841125488, | |
| "learning_rate": 8.59161886459654e-06, | |
| "loss": 0.3149, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.8695198329853863, | |
| "grad_norm": 0.7562083005905151, | |
| "learning_rate": 8.457733338297069e-06, | |
| "loss": 0.3192, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.8705636743215032, | |
| "grad_norm": 0.6112555265426636, | |
| "learning_rate": 8.3248531320263e-06, | |
| "loss": 0.2175, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.8716075156576201, | |
| "grad_norm": 0.7050125598907471, | |
| "learning_rate": 8.192979705078852e-06, | |
| "loss": 0.243, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.872651356993737, | |
| "grad_norm": 0.6470485925674438, | |
| "learning_rate": 8.062114505692742e-06, | |
| "loss": 0.2384, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.8736951983298539, | |
| "grad_norm": 0.7082952260971069, | |
| "learning_rate": 7.932258971033746e-06, | |
| "loss": 0.2795, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.8747390396659708, | |
| "grad_norm": 0.843268096446991, | |
| "learning_rate": 7.803414527179343e-06, | |
| "loss": 0.2991, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.8757828810020877, | |
| "grad_norm": 0.6356431245803833, | |
| "learning_rate": 7.675582589103247e-06, | |
| "loss": 0.1963, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.8768267223382046, | |
| "grad_norm": 0.569520890712738, | |
| "learning_rate": 7.548764560659816e-06, | |
| "loss": 0.1703, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.8778705636743215, | |
| "grad_norm": 0.6984921097755432, | |
| "learning_rate": 7.422961834568565e-06, | |
| "loss": 0.2231, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.8789144050104384, | |
| "grad_norm": 0.6111634969711304, | |
| "learning_rate": 7.2981757923989755e-06, | |
| "loss": 0.1825, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.8799582463465553, | |
| "grad_norm": 0.666388213634491, | |
| "learning_rate": 7.174407804555261e-06, | |
| "loss": 0.1775, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.8810020876826722, | |
| "grad_norm": 0.7088585495948792, | |
| "learning_rate": 7.051659230261298e-06, | |
| "loss": 0.1873, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.8820459290187892, | |
| "grad_norm": 0.6127867102622986, | |
| "learning_rate": 6.929931417545788e-06, | |
| "loss": 0.1732, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.8830897703549061, | |
| "grad_norm": 0.637973964214325, | |
| "learning_rate": 6.809225703227351e-06, | |
| "loss": 0.1856, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.884133611691023, | |
| "grad_norm": 0.5888153910636902, | |
| "learning_rate": 6.689543412899913e-06, | |
| "loss": 0.1872, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.8851774530271399, | |
| "grad_norm": 0.6536146402359009, | |
| "learning_rate": 6.57088586091813e-06, | |
| "loss": 0.2259, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.8862212943632568, | |
| "grad_norm": 0.4231550693511963, | |
| "learning_rate": 6.45325435038292e-06, | |
| "loss": 0.1388, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.8872651356993737, | |
| "grad_norm": 0.8168404698371887, | |
| "learning_rate": 6.336650173127223e-06, | |
| "loss": 0.1743, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.8883089770354906, | |
| "grad_norm": 0.5078144073486328, | |
| "learning_rate": 6.221074609701738e-06, | |
| "loss": 0.2793, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.8893528183716075, | |
| "grad_norm": 0.7190085053443909, | |
| "learning_rate": 6.106528929360911e-06, | |
| "loss": 0.4109, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.8903966597077244, | |
| "grad_norm": 0.5939377546310425, | |
| "learning_rate": 5.99301439004899e-06, | |
| "loss": 0.2797, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.8914405010438413, | |
| "grad_norm": 0.7328153848648071, | |
| "learning_rate": 5.880532238386161e-06, | |
| "loss": 0.3652, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.8924843423799582, | |
| "grad_norm": 0.6011344194412231, | |
| "learning_rate": 5.769083709654932e-06, | |
| "loss": 0.2699, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.8935281837160751, | |
| "grad_norm": 0.7004411220550537, | |
| "learning_rate": 5.658670027786561e-06, | |
| "loss": 0.3191, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.894572025052192, | |
| "grad_norm": 0.5985621809959412, | |
| "learning_rate": 5.549292405347495e-06, | |
| "loss": 0.2593, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.8956158663883089, | |
| "grad_norm": 0.6498935222625732, | |
| "learning_rate": 5.440952043526215e-06, | |
| "loss": 0.3027, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.8966597077244259, | |
| "grad_norm": 0.6431671380996704, | |
| "learning_rate": 5.3336501321199714e-06, | |
| "loss": 0.237, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.8977035490605428, | |
| "grad_norm": 0.6603933572769165, | |
| "learning_rate": 5.22738784952167e-06, | |
| "loss": 0.2533, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.8987473903966597, | |
| "grad_norm": 0.5379349589347839, | |
| "learning_rate": 5.1221663627070485e-06, | |
| "loss": 0.1883, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.8997912317327766, | |
| "grad_norm": 0.4264977276325226, | |
| "learning_rate": 5.017986827221733e-06, | |
| "loss": 0.1174, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.9008350730688935, | |
| "grad_norm": 0.5029094815254211, | |
| "learning_rate": 4.914850387168657e-06, | |
| "loss": 0.1564, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.9018789144050104, | |
| "grad_norm": 0.5162425637245178, | |
| "learning_rate": 4.812758175195397e-06, | |
| "loss": 0.1547, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.9029227557411273, | |
| "grad_norm": 0.6308012008666992, | |
| "learning_rate": 4.711711312481815e-06, | |
| "loss": 0.1844, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.9039665970772442, | |
| "grad_norm": 0.46069368720054626, | |
| "learning_rate": 4.61171090872774e-06, | |
| "loss": 0.1363, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.9050104384133612, | |
| "grad_norm": 0.40711909532546997, | |
| "learning_rate": 4.512758062140687e-06, | |
| "loss": 0.1252, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.906054279749478, | |
| "grad_norm": 0.5450437664985657, | |
| "learning_rate": 4.4148538594239174e-06, | |
| "loss": 0.1885, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.907098121085595, | |
| "grad_norm": 0.5693588852882385, | |
| "learning_rate": 4.317999375764459e-06, | |
| "loss": 0.2161, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.9081419624217119, | |
| "grad_norm": 0.45915868878364563, | |
| "learning_rate": 4.2221956748212384e-06, | |
| "loss": 0.1642, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9091858037578288, | |
| "grad_norm": 0.3877635896205902, | |
| "learning_rate": 4.127443808713527e-06, | |
| "loss": 0.1424, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.9102296450939458, | |
| "grad_norm": 0.5209342241287231, | |
| "learning_rate": 4.033744818009244e-06, | |
| "loss": 0.1703, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.9112734864300627, | |
| "grad_norm": 0.4651091396808624, | |
| "learning_rate": 3.941099731713637e-06, | |
| "loss": 0.1584, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.9123173277661796, | |
| "grad_norm": 0.8967810869216919, | |
| "learning_rate": 3.849509567257959e-06, | |
| "loss": 0.3558, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.9133611691022965, | |
| "grad_norm": 0.8381048440933228, | |
| "learning_rate": 3.7589753304882124e-06, | |
| "loss": 0.3886, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9144050104384134, | |
| "grad_norm": 0.6149895787239075, | |
| "learning_rate": 3.669498015654249e-06, | |
| "loss": 0.31, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.9154488517745303, | |
| "grad_norm": 0.7612007856369019, | |
| "learning_rate": 3.5810786053987023e-06, | |
| "loss": 0.359, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.9164926931106472, | |
| "grad_norm": 0.6727755069732666, | |
| "learning_rate": 3.493718070746299e-06, | |
| "loss": 0.2748, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.9175365344467641, | |
| "grad_norm": 0.734786868095398, | |
| "learning_rate": 3.40741737109318e-06, | |
| "loss": 0.3066, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.918580375782881, | |
| "grad_norm": 0.6576768159866333, | |
| "learning_rate": 3.3221774541962845e-06, | |
| "loss": 0.2677, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9196242171189979, | |
| "grad_norm": 0.6196028590202332, | |
| "learning_rate": 3.2379992561630712e-06, | |
| "loss": 0.2484, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.9206680584551148, | |
| "grad_norm": 0.7527311444282532, | |
| "learning_rate": 3.1548837014411357e-06, | |
| "loss": 0.2787, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.9217118997912317, | |
| "grad_norm": 0.7063425779342651, | |
| "learning_rate": 3.0728317028080657e-06, | |
| "loss": 0.303, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.9227557411273486, | |
| "grad_norm": 0.6437200307846069, | |
| "learning_rate": 2.9918441613615123e-06, | |
| "loss": 0.2501, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.9237995824634656, | |
| "grad_norm": 0.8236239552497864, | |
| "learning_rate": 2.9119219665091344e-06, | |
| "loss": 0.3234, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.9248434237995825, | |
| "grad_norm": 0.686543881893158, | |
| "learning_rate": 2.8330659959589946e-06, | |
| "loss": 0.2407, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.9258872651356994, | |
| "grad_norm": 0.7093439698219299, | |
| "learning_rate": 2.755277115709842e-06, | |
| "loss": 0.246, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.9269311064718163, | |
| "grad_norm": 0.7115840315818787, | |
| "learning_rate": 2.678556180041547e-06, | |
| "loss": 0.2551, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.9279749478079332, | |
| "grad_norm": 0.6106806993484497, | |
| "learning_rate": 2.6029040315058485e-06, | |
| "loss": 0.2137, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.9290187891440501, | |
| "grad_norm": 0.5461225509643555, | |
| "learning_rate": 2.5283215009169857e-06, | |
| "loss": 0.1847, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.930062630480167, | |
| "grad_norm": 0.6097748875617981, | |
| "learning_rate": 2.4548094073426398e-06, | |
| "loss": 0.209, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.9311064718162839, | |
| "grad_norm": 0.6483787298202515, | |
| "learning_rate": 2.3823685580949273e-06, | |
| "loss": 0.2244, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.9321503131524008, | |
| "grad_norm": 0.5993359088897705, | |
| "learning_rate": 2.3109997487214983e-06, | |
| "loss": 0.189, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.9331941544885177, | |
| "grad_norm": 0.8005963563919067, | |
| "learning_rate": 2.240703762996843e-06, | |
| "loss": 0.2385, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.9342379958246346, | |
| "grad_norm": 0.4756294786930084, | |
| "learning_rate": 2.1714813729136975e-06, | |
| "loss": 0.1531, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.9352818371607515, | |
| "grad_norm": 0.577684760093689, | |
| "learning_rate": 2.1033333386744846e-06, | |
| "loss": 0.1722, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.9363256784968684, | |
| "grad_norm": 0.5431109666824341, | |
| "learning_rate": 2.036260408683033e-06, | |
| "loss": 0.1796, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.9373695198329853, | |
| "grad_norm": 0.5910576581954956, | |
| "learning_rate": 1.9702633195363917e-06, | |
| "loss": 0.1689, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.9384133611691023, | |
| "grad_norm": 0.6282055377960205, | |
| "learning_rate": 1.9053427960166182e-06, | |
| "loss": 0.1681, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.9394572025052192, | |
| "grad_norm": 0.46189793944358826, | |
| "learning_rate": 1.8414995510829368e-06, | |
| "loss": 0.0832, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9405010438413361, | |
| "grad_norm": 0.7074732780456543, | |
| "learning_rate": 1.778734285863859e-06, | |
| "loss": 0.3342, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.941544885177453, | |
| "grad_norm": 0.6299365162849426, | |
| "learning_rate": 1.717047689649487e-06, | |
| "loss": 0.3618, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.94258872651357, | |
| "grad_norm": 0.6098471879959106, | |
| "learning_rate": 1.6564404398839439e-06, | |
| "loss": 0.3235, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.9436325678496869, | |
| "grad_norm": 0.5733586549758911, | |
| "learning_rate": 1.5969132021579347e-06, | |
| "loss": 0.2646, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.9446764091858038, | |
| "grad_norm": 0.6624009609222412, | |
| "learning_rate": 1.5384666302014406e-06, | |
| "loss": 0.3138, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.9457202505219207, | |
| "grad_norm": 0.7231829762458801, | |
| "learning_rate": 1.481101365876547e-06, | |
| "loss": 0.3016, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.9467640918580376, | |
| "grad_norm": 0.634404718875885, | |
| "learning_rate": 1.4248180391703614e-06, | |
| "loss": 0.2247, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.9478079331941545, | |
| "grad_norm": 0.664415180683136, | |
| "learning_rate": 1.3696172681881503e-06, | |
| "loss": 0.3176, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.9488517745302714, | |
| "grad_norm": 0.5892297625541687, | |
| "learning_rate": 1.3154996591464908e-06, | |
| "loss": 0.2327, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.9498956158663883, | |
| "grad_norm": 0.5772663354873657, | |
| "learning_rate": 1.2624658063666639e-06, | |
| "loss": 0.2104, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.9509394572025052, | |
| "grad_norm": 0.5238860845565796, | |
| "learning_rate": 1.2105162922680824e-06, | |
| "loss": 0.1807, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.9519832985386222, | |
| "grad_norm": 0.5960121750831604, | |
| "learning_rate": 1.15965168736194e-06, | |
| "loss": 0.2065, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.9530271398747391, | |
| "grad_norm": 0.47030001878738403, | |
| "learning_rate": 1.109872550244917e-06, | |
| "loss": 0.1519, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.954070981210856, | |
| "grad_norm": 0.5340988636016846, | |
| "learning_rate": 1.0611794275930399e-06, | |
| "loss": 0.1603, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.9551148225469729, | |
| "grad_norm": 0.3915995955467224, | |
| "learning_rate": 1.01357285415572e-06, | |
| "loss": 0.1179, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.9561586638830898, | |
| "grad_norm": 0.3927573561668396, | |
| "learning_rate": 9.670533527498137e-07, | |
| "loss": 0.1151, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.9572025052192067, | |
| "grad_norm": 0.3486252725124359, | |
| "learning_rate": 9.216214342539386e-07, | |
| "loss": 0.0942, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.9582463465553236, | |
| "grad_norm": 0.3841801583766937, | |
| "learning_rate": 8.772775976028546e-07, | |
| "loss": 0.1185, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.9592901878914405, | |
| "grad_norm": 0.49043968319892883, | |
| "learning_rate": 8.340223297819471e-07, | |
| "loss": 0.1571, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.9603340292275574, | |
| "grad_norm": 0.5569294691085815, | |
| "learning_rate": 7.918561058219198e-07, | |
| "loss": 0.1835, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9613778705636743, | |
| "grad_norm": 0.39632856845855713, | |
| "learning_rate": 7.507793887935654e-07, | |
| "loss": 0.1649, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.9624217118997912, | |
| "grad_norm": 0.46687746047973633, | |
| "learning_rate": 7.10792629802659e-07, | |
| "loss": 0.1609, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.9634655532359081, | |
| "grad_norm": 0.3641255795955658, | |
| "learning_rate": 6.718962679850505e-07, | |
| "loss": 0.1387, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.964509394572025, | |
| "grad_norm": 0.5133196115493774, | |
| "learning_rate": 6.340907305017907e-07, | |
| "loss": 0.1752, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.965553235908142, | |
| "grad_norm": 0.7964520454406738, | |
| "learning_rate": 5.973764325344688e-07, | |
| "loss": 0.3131, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.9665970772442589, | |
| "grad_norm": 0.7928893566131592, | |
| "learning_rate": 5.617537772806602e-07, | |
| "loss": 0.3533, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.9676409185803758, | |
| "grad_norm": 0.7127247452735901, | |
| "learning_rate": 5.272231559494634e-07, | |
| "loss": 0.3272, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.9686847599164927, | |
| "grad_norm": 0.6839384436607361, | |
| "learning_rate": 4.937849477572587e-07, | |
| "loss": 0.3335, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.9697286012526096, | |
| "grad_norm": 0.6431897878646851, | |
| "learning_rate": 4.614395199234678e-07, | |
| "loss": 0.279, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.9707724425887265, | |
| "grad_norm": 0.6917023658752441, | |
| "learning_rate": 4.3018722766661193e-07, | |
| "loss": 0.3009, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9718162839248434, | |
| "grad_norm": 0.6915965676307678, | |
| "learning_rate": 4.0002841420032634e-07, | |
| "loss": 0.2842, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.9728601252609603, | |
| "grad_norm": 0.6639567017555237, | |
| "learning_rate": 3.7096341072964113e-07, | |
| "loss": 0.2613, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.9739039665970772, | |
| "grad_norm": 0.5585394501686096, | |
| "learning_rate": 3.4299253644732855e-07, | |
| "loss": 0.2097, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.9749478079331941, | |
| "grad_norm": 0.7092443704605103, | |
| "learning_rate": 3.161160985304168e-07, | |
| "loss": 0.2731, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.975991649269311, | |
| "grad_norm": 0.8628230094909668, | |
| "learning_rate": 2.903343921367707e-07, | |
| "loss": 0.3337, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.9770354906054279, | |
| "grad_norm": 0.7503064870834351, | |
| "learning_rate": 2.6564770040190535e-07, | |
| "loss": 0.2528, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.9780793319415448, | |
| "grad_norm": 0.6571477651596069, | |
| "learning_rate": 2.420562944358329e-07, | |
| "loss": 0.2097, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.9791231732776617, | |
| "grad_norm": 0.7250248193740845, | |
| "learning_rate": 2.1956043332010955e-07, | |
| "loss": 0.2418, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.9801670146137788, | |
| "grad_norm": 0.5199112892150879, | |
| "learning_rate": 1.9816036410499338e-07, | |
| "loss": 0.1988, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.9812108559498957, | |
| "grad_norm": 0.6716915369033813, | |
| "learning_rate": 1.7785632180670198e-07, | |
| "loss": 0.2049, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.9822546972860126, | |
| "grad_norm": 0.6581385731697083, | |
| "learning_rate": 1.5864852940485898e-07, | |
| "loss": 0.2065, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.9832985386221295, | |
| "grad_norm": 0.5942964553833008, | |
| "learning_rate": 1.405371978400516e-07, | |
| "loss": 0.188, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.9843423799582464, | |
| "grad_norm": 0.6144523620605469, | |
| "learning_rate": 1.2352252601147697e-07, | |
| "loss": 0.1848, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.9853862212943633, | |
| "grad_norm": 0.6265504956245422, | |
| "learning_rate": 1.0760470077479934e-07, | |
| "loss": 0.2032, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.9864300626304802, | |
| "grad_norm": 0.5609498023986816, | |
| "learning_rate": 9.278389694006296e-08, | |
| "loss": 0.1657, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.9874739039665971, | |
| "grad_norm": 0.5039533972740173, | |
| "learning_rate": 7.906027726981568e-08, | |
| "loss": 0.158, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.988517745302714, | |
| "grad_norm": 0.5719618201255798, | |
| "learning_rate": 6.643399247725502e-08, | |
| "loss": 0.151, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.9895615866388309, | |
| "grad_norm": 0.5313311815261841, | |
| "learning_rate": 5.490518122465149e-08, | |
| "loss": 0.1563, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.9906054279749478, | |
| "grad_norm": 0.48120567202568054, | |
| "learning_rate": 4.447397012177224e-08, | |
| "loss": 0.1376, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.9916492693110647, | |
| "grad_norm": 0.44713348150253296, | |
| "learning_rate": 3.5140473724482034e-08, | |
| "loss": 0.1198, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.9926931106471816, | |
| "grad_norm": 0.6056554913520813, | |
| "learning_rate": 2.6904794533544332e-08, | |
| "loss": 0.2989, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.9937369519832986, | |
| "grad_norm": 0.6082270741462708, | |
| "learning_rate": 1.976702299344435e-08, | |
| "loss": 0.2356, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.9947807933194155, | |
| "grad_norm": 0.46945539116859436, | |
| "learning_rate": 1.3727237491412137e-08, | |
| "loss": 0.1697, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.9958246346555324, | |
| "grad_norm": 0.5799762010574341, | |
| "learning_rate": 8.785504356556563e-09, | |
| "loss": 0.2196, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.9968684759916493, | |
| "grad_norm": 0.6873183846473694, | |
| "learning_rate": 4.941877859143684e-09, | |
| "loss": 0.2819, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.9979123173277662, | |
| "grad_norm": 0.6708718538284302, | |
| "learning_rate": 2.1964002100083312e-09, | |
| "loss": 0.2342, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.9989561586638831, | |
| "grad_norm": 0.6889147758483887, | |
| "learning_rate": 5.491015600656013e-10, | |
| "loss": 0.2048, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.7045766115188599, | |
| "learning_rate": 0.0, | |
| "loss": 0.1947, | |
| "step": 958 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 958, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 240, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.355154213385011e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |