| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.937369519832985, | |
| "eval_steps": 500, | |
| "global_step": 145, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.033402922755741124, | |
| "grad_norm": 7.154098089649519, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 1.2049, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.06680584551148225, | |
| "grad_norm": 7.224671814367719, | |
| "learning_rate": 1.0666666666666667e-05, | |
| "loss": 1.2046, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.10020876826722339, | |
| "grad_norm": 5.112010906482035, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.1276, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1336116910229645, | |
| "grad_norm": 5.240191362417293, | |
| "learning_rate": 2.1333333333333335e-05, | |
| "loss": 1.0958, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.16701461377870563, | |
| "grad_norm": 4.398708169023894, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 1.0242, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.20041753653444677, | |
| "grad_norm": 4.9473080352678895, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.0277, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.23382045929018788, | |
| "grad_norm": 3.8837845230573835, | |
| "learning_rate": 3.733333333333334e-05, | |
| "loss": 0.9755, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.267223382045929, | |
| "grad_norm": 2.957475416120971, | |
| "learning_rate": 4.266666666666667e-05, | |
| "loss": 0.9334, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.30062630480167013, | |
| "grad_norm": 2.187999537773017, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.9211, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.33402922755741127, | |
| "grad_norm": 2.3155958654718614, | |
| "learning_rate": 5.333333333333333e-05, | |
| "loss": 0.8983, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3674321503131524, | |
| "grad_norm": 2.8444701447464436, | |
| "learning_rate": 5.8666666666666665e-05, | |
| "loss": 0.8975, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.40083507306889354, | |
| "grad_norm": 3.0982586639870213, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.8856, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.4342379958246347, | |
| "grad_norm": 1.8874954111166966, | |
| "learning_rate": 6.933333333333334e-05, | |
| "loss": 0.872, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.46764091858037576, | |
| "grad_norm": 3.038189077699479, | |
| "learning_rate": 7.466666666666667e-05, | |
| "loss": 0.8892, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5010438413361169, | |
| "grad_norm": 1.7381979961116139, | |
| "learning_rate": 8e-05, | |
| "loss": 0.8535, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.534446764091858, | |
| "grad_norm": 2188.323143153608, | |
| "learning_rate": 7.998832056320773e-05, | |
| "loss": 1.0923, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5678496868475992, | |
| "grad_norm": 5.568753243553315, | |
| "learning_rate": 7.995328907329308e-05, | |
| "loss": 0.9433, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.6012526096033403, | |
| "grad_norm": 3.003707178203899, | |
| "learning_rate": 7.989492598765966e-05, | |
| "loss": 0.8783, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.6346555323590815, | |
| "grad_norm": 3.3787729580945367, | |
| "learning_rate": 7.981326538870596e-05, | |
| "loss": 0.8657, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6680584551148225, | |
| "grad_norm": 3.0096085489081923, | |
| "learning_rate": 7.970835496392216e-05, | |
| "loss": 0.8705, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7014613778705637, | |
| "grad_norm": 2.2251853165136186, | |
| "learning_rate": 7.958025597804205e-05, | |
| "loss": 0.8591, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.7348643006263048, | |
| "grad_norm": 1.393655838420696, | |
| "learning_rate": 7.942904323726604e-05, | |
| "loss": 0.8202, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.7682672233820459, | |
| "grad_norm": 1.709595774468825, | |
| "learning_rate": 7.925480504557654e-05, | |
| "loss": 0.8239, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.8016701461377871, | |
| "grad_norm": 1.04923819217978, | |
| "learning_rate": 7.90576431531709e-05, | |
| "loss": 0.8236, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.8350730688935282, | |
| "grad_norm": 1.4676306786694173, | |
| "learning_rate": 7.883767269704209e-05, | |
| "loss": 0.8083, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.8684759916492694, | |
| "grad_norm": 417.78005391043996, | |
| "learning_rate": 7.859502213374207e-05, | |
| "loss": 1.1719, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.9018789144050104, | |
| "grad_norm": 12.203128331470229, | |
| "learning_rate": 7.832983316436666e-05, | |
| "loss": 0.8597, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.9352818371607515, | |
| "grad_norm": 4.957392164656034, | |
| "learning_rate": 7.804226065180615e-05, | |
| "loss": 0.9382, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.9686847599164927, | |
| "grad_norm": 10.61998316626802, | |
| "learning_rate": 7.773247253030973e-05, | |
| "loss": 0.9939, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.0083507306889352, | |
| "grad_norm": 54.87377316789775, | |
| "learning_rate": 7.740064970741661e-05, | |
| "loss": 0.8724, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.0417536534446765, | |
| "grad_norm": 88.85368950052036, | |
| "learning_rate": 7.704698595831107e-05, | |
| "loss": 0.9805, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.0751565762004176, | |
| "grad_norm": 12.07917493536492, | |
| "learning_rate": 7.667168781266331e-05, | |
| "loss": 0.9689, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.1085594989561587, | |
| "grad_norm": 30.730151095536375, | |
| "learning_rate": 7.627497443402182e-05, | |
| "loss": 1.0908, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.1419624217118998, | |
| "grad_norm": 4.303465706734318, | |
| "learning_rate": 7.585707749182816e-05, | |
| "loss": 0.8883, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.1753653444676408, | |
| "grad_norm": 1.3597826636785535, | |
| "learning_rate": 7.541824102612839e-05, | |
| "loss": 0.8376, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.2087682672233822, | |
| "grad_norm": 2.028071696746282, | |
| "learning_rate": 7.495872130506072e-05, | |
| "loss": 0.8018, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.2421711899791232, | |
| "grad_norm": 1.5668710744698326, | |
| "learning_rate": 7.447878667520198e-05, | |
| "loss": 0.7901, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.2755741127348643, | |
| "grad_norm": 0.9166659557892114, | |
| "learning_rate": 7.397871740486085e-05, | |
| "loss": 0.7699, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.3089770354906054, | |
| "grad_norm": 8.207728871032339, | |
| "learning_rate": 7.345880552040907e-05, | |
| "loss": 0.7735, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.3423799582463465, | |
| "grad_norm": 2.508524585534657, | |
| "learning_rate": 7.291935463574626e-05, | |
| "loss": 0.8447, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.3757828810020878, | |
| "grad_norm": 1.364586902470756, | |
| "learning_rate": 7.236067977499791e-05, | |
| "loss": 0.7856, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.4091858037578289, | |
| "grad_norm": 1.8720653087352772, | |
| "learning_rate": 7.178310718855018e-05, | |
| "loss": 0.7829, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.44258872651357, | |
| "grad_norm": 1.8178558573775088, | |
| "learning_rate": 7.11869741625289e-05, | |
| "loss": 0.7737, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.475991649269311, | |
| "grad_norm": 1.6638629849138615, | |
| "learning_rate": 7.057262882183393e-05, | |
| "loss": 0.7737, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.5093945720250521, | |
| "grad_norm": 1.1958951695778888, | |
| "learning_rate": 6.994042992684406e-05, | |
| "loss": 0.7499, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.5427974947807934, | |
| "grad_norm": 1.1237749762548175, | |
| "learning_rate": 6.929074666391095e-05, | |
| "loss": 0.7457, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.5762004175365343, | |
| "grad_norm": 0.9523363043794499, | |
| "learning_rate": 6.862395842976484e-05, | |
| "loss": 0.7449, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.6096033402922756, | |
| "grad_norm": 0.7794493625394828, | |
| "learning_rate": 6.79404546099575e-05, | |
| "loss": 0.7471, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.6430062630480167, | |
| "grad_norm": 2.1007938107258224, | |
| "learning_rate": 6.724063435147189e-05, | |
| "loss": 0.738, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.6764091858037578, | |
| "grad_norm": 0.8397590052899939, | |
| "learning_rate": 6.652490632963182e-05, | |
| "loss": 0.7366, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.709812108559499, | |
| "grad_norm": 1.5244169148841136, | |
| "learning_rate": 6.579368850944683e-05, | |
| "loss": 0.7518, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.7432150313152401, | |
| "grad_norm": 0.97500410064134, | |
| "learning_rate": 6.504740790153255e-05, | |
| "loss": 0.7365, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.7766179540709812, | |
| "grad_norm": 1.833109071141926, | |
| "learning_rate": 6.428650031274845e-05, | |
| "loss": 0.7327, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.8100208768267223, | |
| "grad_norm": 1.4707510085946327, | |
| "learning_rate": 6.351141009169893e-05, | |
| "loss": 0.7227, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.8434237995824634, | |
| "grad_norm": 1.2363917202252765, | |
| "learning_rate": 6.272258986924624e-05, | |
| "loss": 0.7405, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.8768267223382047, | |
| "grad_norm": 1.0298920741813498, | |
| "learning_rate": 6.192050029418682e-05, | |
| "loss": 0.7241, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.9102296450939458, | |
| "grad_norm": 0.9097363351471279, | |
| "learning_rate": 6.110560976424531e-05, | |
| "loss": 0.7167, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.9436325678496869, | |
| "grad_norm": 0.8471695952793523, | |
| "learning_rate": 6.027839415254362e-05, | |
| "loss": 0.7181, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.977035490605428, | |
| "grad_norm": 0.6602662698524506, | |
| "learning_rate": 5.943933652970424e-05, | |
| "loss": 0.7088, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.0167014613778704, | |
| "grad_norm": 0.624041177687339, | |
| "learning_rate": 5.858892688175075e-05, | |
| "loss": 0.6922, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.0501043841336117, | |
| "grad_norm": 0.731560229530671, | |
| "learning_rate": 5.772766182396966e-05, | |
| "loss": 0.6655, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.083507306889353, | |
| "grad_norm": 0.5160825456760252, | |
| "learning_rate": 5.685604431090117e-05, | |
| "loss": 0.6624, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.116910229645094, | |
| "grad_norm": 0.6466642583190281, | |
| "learning_rate": 5.597458334262782e-05, | |
| "loss": 0.6474, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.150313152400835, | |
| "grad_norm": 0.6905839273768964, | |
| "learning_rate": 5.508379366753282e-05, | |
| "loss": 0.6512, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.183716075156576, | |
| "grad_norm": 0.3760316450742919, | |
| "learning_rate": 5.4184195481701425e-05, | |
| "loss": 0.6523, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.2171189979123174, | |
| "grad_norm": 0.606234562718693, | |
| "learning_rate": 5.3276314125141144e-05, | |
| "loss": 0.6487, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.2505219206680582, | |
| "grad_norm": 0.44809718292050676, | |
| "learning_rate": 5.23606797749979e-05, | |
| "loss": 0.649, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.2839248434237995, | |
| "grad_norm": 0.40244410097202155, | |
| "learning_rate": 5.1437827135947566e-05, | |
| "loss": 0.6468, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.317327766179541, | |
| "grad_norm": 0.359719180741915, | |
| "learning_rate": 5.050829512794348e-05, | |
| "loss": 0.6409, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.3507306889352817, | |
| "grad_norm": 0.40415638024369727, | |
| "learning_rate": 4.9572626571502316e-05, | |
| "loss": 0.639, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.384133611691023, | |
| "grad_norm": 0.3340843503248373, | |
| "learning_rate": 4.8631367870712254e-05, | |
| "loss": 0.6326, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.4175365344467643, | |
| "grad_norm": 0.3262882595570267, | |
| "learning_rate": 4.768506869414834e-05, | |
| "loss": 0.6298, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.450939457202505, | |
| "grad_norm": 0.3253891492249243, | |
| "learning_rate": 4.6734281653881536e-05, | |
| "loss": 0.6326, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.4843423799582465, | |
| "grad_norm": 0.35311540573233735, | |
| "learning_rate": 4.577956198276886e-05, | |
| "loss": 0.6291, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.5177453027139873, | |
| "grad_norm": 0.3440383701499157, | |
| "learning_rate": 4.4821467210212924e-05, | |
| "loss": 0.6332, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.5511482254697286, | |
| "grad_norm": 0.30978369513311105, | |
| "learning_rate": 4.386055683658061e-05, | |
| "loss": 0.6408, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.5845511482254695, | |
| "grad_norm": 0.3823149004105222, | |
| "learning_rate": 4.2897392006470503e-05, | |
| "loss": 0.6246, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.617954070981211, | |
| "grad_norm": 0.2810880790539587, | |
| "learning_rate": 4.1932535181020286e-05, | |
| "loss": 0.6293, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.651356993736952, | |
| "grad_norm": 0.2835535239751324, | |
| "learning_rate": 4.096654980944529e-05, | |
| "loss": 0.6252, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.684759916492693, | |
| "grad_norm": 0.336833154001104, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6305, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.7181628392484343, | |
| "grad_norm": 0.23274589850456745, | |
| "learning_rate": 3.903345019055472e-05, | |
| "loss": 0.6298, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.7515657620041756, | |
| "grad_norm": 0.2420684628004819, | |
| "learning_rate": 3.806746481897973e-05, | |
| "loss": 0.6241, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.7849686847599164, | |
| "grad_norm": 0.23622928619950834, | |
| "learning_rate": 3.710260799352951e-05, | |
| "loss": 0.6167, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.8183716075156577, | |
| "grad_norm": 0.21286687906297902, | |
| "learning_rate": 3.6139443163419394e-05, | |
| "loss": 0.6268, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.8517745302713986, | |
| "grad_norm": 0.20113400910479923, | |
| "learning_rate": 3.517853278978708e-05, | |
| "loss": 0.622, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.88517745302714, | |
| "grad_norm": 0.19296938971649688, | |
| "learning_rate": 3.422043801723116e-05, | |
| "loss": 0.6167, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.9185803757828808, | |
| "grad_norm": 0.17640926051127553, | |
| "learning_rate": 3.3265718346118464e-05, | |
| "loss": 0.6251, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.951983298538622, | |
| "grad_norm": 0.17760201524918323, | |
| "learning_rate": 3.231493130585167e-05, | |
| "loss": 0.6195, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.9853862212943634, | |
| "grad_norm": 0.18267169419590248, | |
| "learning_rate": 3.136863212928776e-05, | |
| "loss": 0.6214, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.1002087682672235, | |
| "grad_norm": 0.2479134339023779, | |
| "learning_rate": 3.0427373428497704e-05, | |
| "loss": 0.5792, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.1336116910229643, | |
| "grad_norm": 0.17829804990091588, | |
| "learning_rate": 2.9491704872056525e-05, | |
| "loss": 0.571, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.1670146137787056, | |
| "grad_norm": 0.2102957726786887, | |
| "learning_rate": 2.8562172864052437e-05, | |
| "loss": 0.5665, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.200417536534447, | |
| "grad_norm": 0.18138996143773695, | |
| "learning_rate": 2.7639320225002108e-05, | |
| "loss": 0.5734, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.233820459290188, | |
| "grad_norm": 0.18231114685106467, | |
| "learning_rate": 2.6723685874858873e-05, | |
| "loss": 0.5665, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.267223382045929, | |
| "grad_norm": 0.1891068826294468, | |
| "learning_rate": 2.5815804518298575e-05, | |
| "loss": 0.5649, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.30062630480167, | |
| "grad_norm": 0.1449193634467542, | |
| "learning_rate": 2.4916206332467184e-05, | |
| "loss": 0.5626, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.3340292275574113, | |
| "grad_norm": 0.17521384625415576, | |
| "learning_rate": 2.4025416657372186e-05, | |
| "loss": 0.5672, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.3674321503131526, | |
| "grad_norm": 0.17060274829594732, | |
| "learning_rate": 2.3143955689098844e-05, | |
| "loss": 0.5701, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.4008350730688934, | |
| "grad_norm": 0.16427792254004098, | |
| "learning_rate": 2.2272338176030354e-05, | |
| "loss": 0.5648, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.4342379958246347, | |
| "grad_norm": 0.16851785214921267, | |
| "learning_rate": 2.141107311824926e-05, | |
| "loss": 0.5637, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.4676409185803756, | |
| "grad_norm": 0.1647295715319099, | |
| "learning_rate": 2.056066347029576e-05, | |
| "loss": 0.5698, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 3.501043841336117, | |
| "grad_norm": 0.14383360405355872, | |
| "learning_rate": 1.9721605847456397e-05, | |
| "loss": 0.5678, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.534446764091858, | |
| "grad_norm": 0.16369393007489977, | |
| "learning_rate": 1.8894390235754686e-05, | |
| "loss": 0.5687, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 3.567849686847599, | |
| "grad_norm": 0.1484533364671656, | |
| "learning_rate": 1.807949970581321e-05, | |
| "loss": 0.5612, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.6012526096033404, | |
| "grad_norm": 0.13327395767499348, | |
| "learning_rate": 1.7277410130753775e-05, | |
| "loss": 0.5621, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.6346555323590817, | |
| "grad_norm": 0.14483989970799924, | |
| "learning_rate": 1.648858990830108e-05, | |
| "loss": 0.5602, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.6680584551148225, | |
| "grad_norm": 0.11501467177953302, | |
| "learning_rate": 1.5713499687251554e-05, | |
| "loss": 0.5625, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.701461377870564, | |
| "grad_norm": 0.12471522633663724, | |
| "learning_rate": 1.4952592098467453e-05, | |
| "loss": 0.5566, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.7348643006263047, | |
| "grad_norm": 0.12841415317626956, | |
| "learning_rate": 1.4206311490553187e-05, | |
| "loss": 0.5563, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.768267223382046, | |
| "grad_norm": 0.13024977809323665, | |
| "learning_rate": 1.3475093670368202e-05, | |
| "loss": 0.5642, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.801670146137787, | |
| "grad_norm": 0.12141142140280577, | |
| "learning_rate": 1.275936564852811e-05, | |
| "loss": 0.5619, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.835073068893528, | |
| "grad_norm": 0.1189681977822036, | |
| "learning_rate": 1.2059545390042526e-05, | |
| "loss": 0.5627, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.8684759916492695, | |
| "grad_norm": 0.11637565722872692, | |
| "learning_rate": 1.1376041570235162e-05, | |
| "loss": 0.5597, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.9018789144050103, | |
| "grad_norm": 0.11126444342562675, | |
| "learning_rate": 1.070925333608907e-05, | |
| "loss": 0.5646, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.9352818371607516, | |
| "grad_norm": 0.11144727795080511, | |
| "learning_rate": 1.0059570073155953e-05, | |
| "loss": 0.5663, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.968684759916493, | |
| "grad_norm": 0.11568625785765184, | |
| "learning_rate": 9.427371178166065e-06, | |
| "loss": 0.5628, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.002087682672234, | |
| "grad_norm": 0.1172051146855964, | |
| "learning_rate": 8.81302583747111e-06, | |
| "loss": 0.5657, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.035490605427975, | |
| "grad_norm": 0.1386962246589997, | |
| "learning_rate": 8.216892811449834e-06, | |
| "loss": 0.5431, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.068893528183716, | |
| "grad_norm": 0.12227065820674828, | |
| "learning_rate": 7.639320225002106e-06, | |
| "loss": 0.5386, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.102296450939457, | |
| "grad_norm": 0.11676189189940173, | |
| "learning_rate": 7.080645364253747e-06, | |
| "loss": 0.5341, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.135699373695198, | |
| "grad_norm": 0.1086671391408473, | |
| "learning_rate": 6.541194479590931e-06, | |
| "loss": 0.5472, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.16910229645094, | |
| "grad_norm": 0.11780635228612878, | |
| "learning_rate": 6.021282595139167e-06, | |
| "loss": 0.5376, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.202505219206681, | |
| "grad_norm": 0.11122819389457546, | |
| "learning_rate": 5.521213324798029e-06, | |
| "loss": 0.5405, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.235908141962422, | |
| "grad_norm": 0.11302950909483094, | |
| "learning_rate": 5.0412786949392845e-06, | |
| "loss": 0.5389, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.2693110647181625, | |
| "grad_norm": 0.10897006989347469, | |
| "learning_rate": 4.581758973871609e-06, | |
| "loss": 0.5443, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.302713987473904, | |
| "grad_norm": 0.10317913683812792, | |
| "learning_rate": 4.142922508171849e-06, | |
| "loss": 0.5363, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 4.336116910229645, | |
| "grad_norm": 0.10003486708202455, | |
| "learning_rate": 3.7250255659781844e-06, | |
| "loss": 0.5364, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 4.369519832985386, | |
| "grad_norm": 0.1071731871255614, | |
| "learning_rate": 3.3283121873367043e-06, | |
| "loss": 0.5432, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 4.402922755741128, | |
| "grad_norm": 0.10539727253992291, | |
| "learning_rate": 2.9530140416889465e-06, | |
| "loss": 0.5373, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 4.4363256784968685, | |
| "grad_norm": 0.09606764766200912, | |
| "learning_rate": 2.5993502925834115e-06, | |
| "loss": 0.5333, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 4.469728601252609, | |
| "grad_norm": 0.09219554498498256, | |
| "learning_rate": 2.2675274696902737e-06, | |
| "loss": 0.5315, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 4.503131524008351, | |
| "grad_norm": 0.08960509523269163, | |
| "learning_rate": 1.957739348193859e-06, | |
| "loss": 0.5334, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.536534446764092, | |
| "grad_norm": 0.09160224849384657, | |
| "learning_rate": 1.670166835633351e-06, | |
| "loss": 0.5384, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 4.569937369519833, | |
| "grad_norm": 0.08852713488345453, | |
| "learning_rate": 1.4049778662579462e-06, | |
| "loss": 0.53, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 4.603340292275574, | |
| "grad_norm": 0.09132315616256415, | |
| "learning_rate": 1.1623273029579195e-06, | |
| "loss": 0.538, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 4.6367432150313155, | |
| "grad_norm": 0.09194633813127549, | |
| "learning_rate": 9.423568468291156e-07, | |
| "loss": 0.541, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.670146137787056, | |
| "grad_norm": 0.09237238332398756, | |
| "learning_rate": 7.451949544234627e-07, | |
| "loss": 0.5379, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 4.703549060542797, | |
| "grad_norm": 0.08974432368849375, | |
| "learning_rate": 5.709567627339674e-07, | |
| "loss": 0.5443, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 4.736951983298539, | |
| "grad_norm": 0.09013165116820136, | |
| "learning_rate": 4.1974402195795514e-07, | |
| "loss": 0.535, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 4.77035490605428, | |
| "grad_norm": 0.09022158123863006, | |
| "learning_rate": 2.916450360778411e-07, | |
| "loss": 0.5333, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.803757828810021, | |
| "grad_norm": 0.08981410735542258, | |
| "learning_rate": 1.867346112940549e-07, | |
| "loss": 0.5462, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 4.8371607515657615, | |
| "grad_norm": 0.09141559902413697, | |
| "learning_rate": 1.0507401234035819e-07, | |
| "loss": 0.5377, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 4.870563674321503, | |
| "grad_norm": 0.08951832053531815, | |
| "learning_rate": 4.6710926706934336e-08, | |
| "loss": 0.5305, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 4.903966597077244, | |
| "grad_norm": 0.08926873218925392, | |
| "learning_rate": 1.1679436792282339e-08, | |
| "loss": 0.54, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 4.937369519832985, | |
| "grad_norm": 0.08723888890212035, | |
| "learning_rate": 0.0, | |
| "loss": 0.54, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 4.937369519832985, | |
| "step": 145, | |
| "total_flos": 3.738141667979428e+18, | |
| "train_loss": 0.2130410626016814, | |
| "train_runtime": 6079.0514, | |
| "train_samples_per_second": 12.591, | |
| "train_steps_per_second": 0.024 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 145, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.738141667979428e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |