| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.385211687537269, | |
| "eval_steps": 500, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.011926058437686345, | |
| "grad_norm": 0.5661849975585938, | |
| "learning_rate": 4.999804802810596e-05, | |
| "loss": 1.4421, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02385211687537269, | |
| "grad_norm": 0.5908129811286926, | |
| "learning_rate": 4.999219241723937e-05, | |
| "loss": 1.3306, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03577817531305903, | |
| "grad_norm": 0.6965810060501099, | |
| "learning_rate": 4.998243408179925e-05, | |
| "loss": 1.1873, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04770423375074538, | |
| "grad_norm": 0.5752276182174683, | |
| "learning_rate": 4.9968774545625344e-05, | |
| "loss": 1.0583, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05963029218843172, | |
| "grad_norm": 0.6488135457038879, | |
| "learning_rate": 4.9951215941760075e-05, | |
| "loss": 1.0045, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.07155635062611806, | |
| "grad_norm": 0.7977299094200134, | |
| "learning_rate": 4.992976101211558e-05, | |
| "loss": 0.9836, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08348240906380441, | |
| "grad_norm": 0.8842712044715881, | |
| "learning_rate": 4.99044131070454e-05, | |
| "loss": 0.8925, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09540846750149076, | |
| "grad_norm": 0.798170804977417, | |
| "learning_rate": 4.987517618482142e-05, | |
| "loss": 0.8783, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1073345259391771, | |
| "grad_norm": 0.8196467757225037, | |
| "learning_rate": 4.984205481101565e-05, | |
| "loss": 0.8681, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.11926058437686345, | |
| "grad_norm": 0.9973539113998413, | |
| "learning_rate": 4.980505415778738e-05, | |
| "loss": 0.8251, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13118664281454978, | |
| "grad_norm": 0.8836011290550232, | |
| "learning_rate": 4.97641800030754e-05, | |
| "loss": 0.8079, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.14311270125223613, | |
| "grad_norm": 0.9255298972129822, | |
| "learning_rate": 4.971943872969582e-05, | |
| "loss": 0.7769, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15503875968992248, | |
| "grad_norm": 0.9712087512016296, | |
| "learning_rate": 4.967083732434529e-05, | |
| "loss": 0.777, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.16696481812760883, | |
| "grad_norm": 0.7609415054321289, | |
| "learning_rate": 4.961838337650997e-05, | |
| "loss": 0.7303, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17889087656529518, | |
| "grad_norm": 0.8297658562660217, | |
| "learning_rate": 4.9562085077280443e-05, | |
| "loss": 0.7442, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.19081693500298152, | |
| "grad_norm": 1.01498281955719, | |
| "learning_rate": 4.950195121807251e-05, | |
| "loss": 0.7077, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.20274299344066785, | |
| "grad_norm": 0.9629709720611572, | |
| "learning_rate": 4.943799118925443e-05, | |
| "loss": 0.673, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2146690518783542, | |
| "grad_norm": 0.9121557474136353, | |
| "learning_rate": 4.937021497868047e-05, | |
| "loss": 0.7109, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.22659511031604054, | |
| "grad_norm": 0.841573178768158, | |
| "learning_rate": 4.9298633170131304e-05, | |
| "loss": 0.6798, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2385211687537269, | |
| "grad_norm": 0.9693652391433716, | |
| "learning_rate": 4.922325694166119e-05, | |
| "loss": 0.6471, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2504472271914132, | |
| "grad_norm": 1.1198012828826904, | |
| "learning_rate": 4.9144098063852485e-05, | |
| "loss": 0.6456, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.26237328562909956, | |
| "grad_norm": 1.1725982427597046, | |
| "learning_rate": 4.9061168897977564e-05, | |
| "loss": 0.6462, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2742993440667859, | |
| "grad_norm": 0.9855665564537048, | |
| "learning_rate": 4.8974482394068514e-05, | |
| "loss": 0.6594, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.28622540250447226, | |
| "grad_norm": 0.8321542739868164, | |
| "learning_rate": 4.888405208889486e-05, | |
| "loss": 0.6372, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2981514609421586, | |
| "grad_norm": 0.9451864957809448, | |
| "learning_rate": 4.878989210384972e-05, | |
| "loss": 0.6594, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.31007751937984496, | |
| "grad_norm": 1.1336216926574707, | |
| "learning_rate": 4.869201714274467e-05, | |
| "loss": 0.6544, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3220035778175313, | |
| "grad_norm": 0.8466928601264954, | |
| "learning_rate": 4.8590442489513543e-05, | |
| "loss": 0.6275, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.33392963625521765, | |
| "grad_norm": 0.979365885257721, | |
| "learning_rate": 4.8485184005825815e-05, | |
| "loss": 0.6348, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.345855694692904, | |
| "grad_norm": 0.9159572124481201, | |
| "learning_rate": 4.837625812860961e-05, | |
| "loss": 0.6045, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.35778175313059035, | |
| "grad_norm": 1.0542089939117432, | |
| "learning_rate": 4.8263681867485e-05, | |
| "loss": 0.6487, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3697078115682767, | |
| "grad_norm": 1.0264719724655151, | |
| "learning_rate": 4.814747280210782e-05, | |
| "loss": 0.6026, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.38163387000596305, | |
| "grad_norm": 0.9389889240264893, | |
| "learning_rate": 4.80276490794244e-05, | |
| "loss": 0.6083, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3935599284436494, | |
| "grad_norm": 1.1090476512908936, | |
| "learning_rate": 4.790422941083786e-05, | |
| "loss": 0.5957, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4054859868813357, | |
| "grad_norm": 1.0268605947494507, | |
| "learning_rate": 4.7777233069286154e-05, | |
| "loss": 0.6538, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.41741204531902204, | |
| "grad_norm": 1.240955114364624, | |
| "learning_rate": 4.7646679886232414e-05, | |
| "loss": 0.62, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4293381037567084, | |
| "grad_norm": 1.2029043436050415, | |
| "learning_rate": 4.7512590248568163e-05, | |
| "loss": 0.6461, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.44126416219439474, | |
| "grad_norm": 1.0567660331726074, | |
| "learning_rate": 4.7374985095429725e-05, | |
| "loss": 0.633, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4531902206320811, | |
| "grad_norm": 0.928354024887085, | |
| "learning_rate": 4.723388591492841e-05, | |
| "loss": 0.633, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.46511627906976744, | |
| "grad_norm": 0.8453586101531982, | |
| "learning_rate": 4.708931474079499e-05, | |
| "loss": 0.6082, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4770423375074538, | |
| "grad_norm": 1.33816397190094, | |
| "learning_rate": 4.6941294148938954e-05, | |
| "loss": 0.6176, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.48896839594514013, | |
| "grad_norm": 0.9532414078712463, | |
| "learning_rate": 4.678984725392309e-05, | |
| "loss": 0.6168, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5008944543828264, | |
| "grad_norm": 0.9017360210418701, | |
| "learning_rate": 4.6634997705354024e-05, | |
| "loss": 0.6454, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5128205128205128, | |
| "grad_norm": 1.1095216274261475, | |
| "learning_rate": 4.6476769684189065e-05, | |
| "loss": 0.5738, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5247465712581991, | |
| "grad_norm": 0.8017715811729431, | |
| "learning_rate": 4.631518789896023e-05, | |
| "loss": 0.5978, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5366726296958855, | |
| "grad_norm": 1.0233415365219116, | |
| "learning_rate": 4.6150277581915804e-05, | |
| "loss": 0.6375, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5485986881335718, | |
| "grad_norm": 0.9298461675643921, | |
| "learning_rate": 4.598206448508007e-05, | |
| "loss": 0.606, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5605247465712582, | |
| "grad_norm": 0.9300760626792908, | |
| "learning_rate": 4.581057487623204e-05, | |
| "loss": 0.6261, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5724508050089445, | |
| "grad_norm": 1.4629548788070679, | |
| "learning_rate": 4.5635835534803406e-05, | |
| "loss": 0.5877, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5843768634466309, | |
| "grad_norm": 0.916437029838562, | |
| "learning_rate": 4.545787374769686e-05, | |
| "loss": 0.5821, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5963029218843172, | |
| "grad_norm": 0.9588505625724792, | |
| "learning_rate": 4.527671730502491e-05, | |
| "loss": 0.6126, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6082289803220036, | |
| "grad_norm": 1.068697214126587, | |
| "learning_rate": 4.5092394495770335e-05, | |
| "loss": 0.6034, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6201550387596899, | |
| "grad_norm": 0.816047191619873, | |
| "learning_rate": 4.490493410336857e-05, | |
| "loss": 0.6088, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6320810971973763, | |
| "grad_norm": 0.8465267419815063, | |
| "learning_rate": 4.4714365401213e-05, | |
| "loss": 0.6067, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6440071556350626, | |
| "grad_norm": 0.9819373488426208, | |
| "learning_rate": 4.4520718148083665e-05, | |
| "loss": 0.6334, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.655933214072749, | |
| "grad_norm": 0.7895939350128174, | |
| "learning_rate": 4.43240225835002e-05, | |
| "loss": 0.6093, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6678592725104353, | |
| "grad_norm": 1.0516682863235474, | |
| "learning_rate": 4.41243094229997e-05, | |
| "loss": 0.5939, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6797853309481217, | |
| "grad_norm": 0.8521412014961243, | |
| "learning_rate": 4.392160985334027e-05, | |
| "loss": 0.5483, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.691711389385808, | |
| "grad_norm": 0.847909152507782, | |
| "learning_rate": 4.371595552763093e-05, | |
| "loss": 0.5946, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7036374478234944, | |
| "grad_norm": 1.1502934694290161, | |
| "learning_rate": 4.350737856038878e-05, | |
| "loss": 0.5651, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7155635062611807, | |
| "grad_norm": 1.0434402227401733, | |
| "learning_rate": 4.3295911522524044e-05, | |
| "loss": 0.5965, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.727489564698867, | |
| "grad_norm": 1.1517246961593628, | |
| "learning_rate": 4.308158743625388e-05, | |
| "loss": 0.5751, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.7394156231365534, | |
| "grad_norm": 1.1459076404571533, | |
| "learning_rate": 4.286443976994569e-05, | |
| "loss": 0.6043, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7513416815742398, | |
| "grad_norm": 1.0111889839172363, | |
| "learning_rate": 4.264450243289079e-05, | |
| "loss": 0.5948, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7632677400119261, | |
| "grad_norm": 1.0213602781295776, | |
| "learning_rate": 4.2421809770009225e-05, | |
| "loss": 0.5538, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7751937984496124, | |
| "grad_norm": 1.0203680992126465, | |
| "learning_rate": 4.219639655648651e-05, | |
| "loss": 0.5696, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7871198568872988, | |
| "grad_norm": 0.8697240352630615, | |
| "learning_rate": 4.196829799234321e-05, | |
| "loss": 0.5613, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7990459153249851, | |
| "grad_norm": 0.8935692310333252, | |
| "learning_rate": 4.173754969693826e-05, | |
| "loss": 0.5712, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.8109719737626714, | |
| "grad_norm": 0.7936813831329346, | |
| "learning_rate": 4.1504187703406604e-05, | |
| "loss": 0.5693, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8228980322003577, | |
| "grad_norm": 1.0851740837097168, | |
| "learning_rate": 4.126824845303248e-05, | |
| "loss": 0.5911, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.8348240906380441, | |
| "grad_norm": 0.9009485244750977, | |
| "learning_rate": 4.102976878955869e-05, | |
| "loss": 0.5551, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8467501490757304, | |
| "grad_norm": 1.1172417402267456, | |
| "learning_rate": 4.0788785953433286e-05, | |
| "loss": 0.564, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8586762075134168, | |
| "grad_norm": 1.2822695970535278, | |
| "learning_rate": 4.05453375759941e-05, | |
| "loss": 0.5766, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8706022659511031, | |
| "grad_norm": 1.327506422996521, | |
| "learning_rate": 4.0299461673592376e-05, | |
| "loss": 0.5289, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8825283243887895, | |
| "grad_norm": 1.0519084930419922, | |
| "learning_rate": 4.0051196641656185e-05, | |
| "loss": 0.5839, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8944543828264758, | |
| "grad_norm": 0.8493732810020447, | |
| "learning_rate": 3.980058124869469e-05, | |
| "loss": 0.583, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.9063804412641622, | |
| "grad_norm": 0.9003544449806213, | |
| "learning_rate": 3.9547654630244156e-05, | |
| "loss": 0.5645, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9183064997018485, | |
| "grad_norm": 0.8397880792617798, | |
| "learning_rate": 3.929245628275662e-05, | |
| "loss": 0.6013, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "grad_norm": 1.000611424446106, | |
| "learning_rate": 3.903502605743222e-05, | |
| "loss": 0.5657, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9421586165772212, | |
| "grad_norm": 0.9028176665306091, | |
| "learning_rate": 3.877540415399612e-05, | |
| "loss": 0.6077, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9540846750149076, | |
| "grad_norm": 0.8427708745002747, | |
| "learning_rate": 3.851363111442101e-05, | |
| "loss": 0.5553, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9660107334525939, | |
| "grad_norm": 1.1009222269058228, | |
| "learning_rate": 3.8249747816596136e-05, | |
| "loss": 0.5806, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.9779367918902803, | |
| "grad_norm": 1.3395967483520508, | |
| "learning_rate": 3.7983795467943975e-05, | |
| "loss": 0.5463, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9898628503279666, | |
| "grad_norm": 0.856643557548523, | |
| "learning_rate": 3.77158155989853e-05, | |
| "loss": 0.6032, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.0017889087656529, | |
| "grad_norm": 1.985956072807312, | |
| "learning_rate": 3.74458500568539e-05, | |
| "loss": 0.663, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.0137149672033392, | |
| "grad_norm": 1.099455714225769, | |
| "learning_rate": 3.717394099876182e-05, | |
| "loss": 0.5532, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.0256410256410255, | |
| "grad_norm": 0.8070193529129028, | |
| "learning_rate": 3.690013088541619e-05, | |
| "loss": 0.537, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.037567084078712, | |
| "grad_norm": 1.1234451532363892, | |
| "learning_rate": 3.662446247438867e-05, | |
| "loss": 0.5192, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.0494931425163982, | |
| "grad_norm": 1.0410645008087158, | |
| "learning_rate": 3.6346978813438464e-05, | |
| "loss": 0.5525, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.0614192009540846, | |
| "grad_norm": 1.0676367282867432, | |
| "learning_rate": 3.606772323379017e-05, | |
| "loss": 0.5546, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.073345259391771, | |
| "grad_norm": 1.1925772428512573, | |
| "learning_rate": 3.57867393433672e-05, | |
| "loss": 0.5298, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.0852713178294573, | |
| "grad_norm": 1.0415118932724, | |
| "learning_rate": 3.55040710199821e-05, | |
| "loss": 0.5354, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.0971973762671436, | |
| "grad_norm": 0.9862761497497559, | |
| "learning_rate": 3.521976240448468e-05, | |
| "loss": 0.5173, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.10912343470483, | |
| "grad_norm": 1.0136168003082275, | |
| "learning_rate": 3.493385789386906e-05, | |
| "loss": 0.5424, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.1210494931425163, | |
| "grad_norm": 0.999031126499176, | |
| "learning_rate": 3.464640213434079e-05, | |
| "loss": 0.574, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.1329755515802027, | |
| "grad_norm": 1.0496968030929565, | |
| "learning_rate": 3.435744001434492e-05, | |
| "loss": 0.5152, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.144901610017889, | |
| "grad_norm": 1.0987663269042969, | |
| "learning_rate": 3.40670166575564e-05, | |
| "loss": 0.5504, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.1568276684555754, | |
| "grad_norm": 1.0069026947021484, | |
| "learning_rate": 3.3775177415833605e-05, | |
| "loss": 0.5547, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.1687537268932617, | |
| "grad_norm": 1.194651484489441, | |
| "learning_rate": 3.348196786213633e-05, | |
| "loss": 0.5319, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.180679785330948, | |
| "grad_norm": 0.9666855335235596, | |
| "learning_rate": 3.3187433783409216e-05, | |
| "loss": 0.5541, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.1926058437686344, | |
| "grad_norm": 0.8826591372489929, | |
| "learning_rate": 3.289162117343173e-05, | |
| "loss": 0.5329, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2045319022063208, | |
| "grad_norm": 1.1290160417556763, | |
| "learning_rate": 3.259457622563593e-05, | |
| "loss": 0.5964, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.2164579606440071, | |
| "grad_norm": 1.078009009361267, | |
| "learning_rate": 3.229634532589296e-05, | |
| "loss": 0.5315, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.2283840190816935, | |
| "grad_norm": 1.1144967079162598, | |
| "learning_rate": 3.199697504526955e-05, | |
| "loss": 0.5698, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.2403100775193798, | |
| "grad_norm": 1.3256494998931885, | |
| "learning_rate": 3.169651213275562e-05, | |
| "loss": 0.5234, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.2522361359570662, | |
| "grad_norm": 1.0332247018814087, | |
| "learning_rate": 3.139500350796397e-05, | |
| "loss": 0.5386, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.2641621943947525, | |
| "grad_norm": 1.2091331481933594, | |
| "learning_rate": 3.1092496253803546e-05, | |
| "loss": 0.5136, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.2760882528324389, | |
| "grad_norm": 1.0742837190628052, | |
| "learning_rate": 3.078903760912695e-05, | |
| "loss": 0.5341, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.2880143112701252, | |
| "grad_norm": 1.5551700592041016, | |
| "learning_rate": 3.048467496135384e-05, | |
| "loss": 0.5613, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.2999403697078116, | |
| "grad_norm": 1.0054267644882202, | |
| "learning_rate": 3.017945583907092e-05, | |
| "loss": 0.5076, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.311866428145498, | |
| "grad_norm": 1.081240177154541, | |
| "learning_rate": 2.9873427904610057e-05, | |
| "loss": 0.5115, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.3237924865831843, | |
| "grad_norm": 1.120842456817627, | |
| "learning_rate": 2.956663894660539e-05, | |
| "loss": 0.5467, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.3357185450208706, | |
| "grad_norm": 1.022741675376892, | |
| "learning_rate": 2.9259136872530812e-05, | |
| "loss": 0.523, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.347644603458557, | |
| "grad_norm": 1.0805561542510986, | |
| "learning_rate": 2.8950969701218783e-05, | |
| "loss": 0.5541, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.3595706618962433, | |
| "grad_norm": 1.1941672563552856, | |
| "learning_rate": 2.864218555536188e-05, | |
| "loss": 0.5303, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.3714967203339297, | |
| "grad_norm": 1.0475640296936035, | |
| "learning_rate": 2.833283265399801e-05, | |
| "loss": 0.544, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.383422778771616, | |
| "grad_norm": 1.258143424987793, | |
| "learning_rate": 2.8022959304980695e-05, | |
| "loss": 0.5596, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.3953488372093024, | |
| "grad_norm": 1.044134259223938, | |
| "learning_rate": 2.7712613897435357e-05, | |
| "loss": 0.5003, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.4072748956469887, | |
| "grad_norm": 1.100220799446106, | |
| "learning_rate": 2.7401844894203056e-05, | |
| "loss": 0.5629, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.419200954084675, | |
| "grad_norm": 0.9259990453720093, | |
| "learning_rate": 2.7090700824272557e-05, | |
| "loss": 0.4516, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.4311270125223614, | |
| "grad_norm": 0.8607438206672668, | |
| "learning_rate": 2.6779230275202243e-05, | |
| "loss": 0.538, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.4430530709600478, | |
| "grad_norm": 1.2486215829849243, | |
| "learning_rate": 2.6467481885532704e-05, | |
| "loss": 0.5314, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.454979129397734, | |
| "grad_norm": 1.1418797969818115, | |
| "learning_rate": 2.6155504337191516e-05, | |
| "loss": 0.4841, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.4669051878354205, | |
| "grad_norm": 1.3612860441207886, | |
| "learning_rate": 2.5843346347891163e-05, | |
| "loss": 0.5133, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.4788312462731068, | |
| "grad_norm": 1.1785390377044678, | |
| "learning_rate": 2.5531056663521362e-05, | |
| "loss": 0.5309, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.4907573047107932, | |
| "grad_norm": 0.8908374905586243, | |
| "learning_rate": 2.521868405053706e-05, | |
| "loss": 0.5022, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.5026833631484795, | |
| "grad_norm": 1.031840443611145, | |
| "learning_rate": 2.4906277288343123e-05, | |
| "loss": 0.5254, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.5146094215861656, | |
| "grad_norm": 1.5957837104797363, | |
| "learning_rate": 2.459388516167711e-05, | |
| "loss": 0.5507, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.5265354800238522, | |
| "grad_norm": 1.2121431827545166, | |
| "learning_rate": 2.428155645299111e-05, | |
| "loss": 0.4943, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.9426692724227905, | |
| "learning_rate": 2.3969339934834012e-05, | |
| "loss": 0.5111, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.550387596899225, | |
| "grad_norm": 1.2159509658813477, | |
| "learning_rate": 2.3657284362235274e-05, | |
| "loss": 0.5302, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.562313655336911, | |
| "grad_norm": 1.1508508920669556, | |
| "learning_rate": 2.3345438465091455e-05, | |
| "loss": 0.5299, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.5742397137745976, | |
| "grad_norm": 1.2826497554779053, | |
| "learning_rate": 2.303385094055669e-05, | |
| "loss": 0.5453, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.5861657722122837, | |
| "grad_norm": 1.520505666732788, | |
| "learning_rate": 2.2722570445438214e-05, | |
| "loss": 0.5804, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.5980918306499703, | |
| "grad_norm": 1.1682236194610596, | |
| "learning_rate": 2.2411645588598232e-05, | |
| "loss": 0.5072, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.6100178890876564, | |
| "grad_norm": 1.1223719120025635, | |
| "learning_rate": 2.2101124923363267e-05, | |
| "loss": 0.5287, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.621943947525343, | |
| "grad_norm": 1.0129245519638062, | |
| "learning_rate": 2.1791056939942228e-05, | |
| "loss": 0.5343, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.6338700059630291, | |
| "grad_norm": 1.00519597530365, | |
| "learning_rate": 2.1481490057854217e-05, | |
| "loss": 0.5329, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.6457960644007157, | |
| "grad_norm": 1.1941732168197632, | |
| "learning_rate": 2.1172472618367483e-05, | |
| "loss": 0.5591, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.6577221228384018, | |
| "grad_norm": 0.9398036599159241, | |
| "learning_rate": 2.0864052876950552e-05, | |
| "loss": 0.5254, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.6696481812760884, | |
| "grad_norm": 0.8543565273284912, | |
| "learning_rate": 2.0556278995736782e-05, | |
| "loss": 0.5212, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.6815742397137745, | |
| "grad_norm": 1.0366357564926147, | |
| "learning_rate": 2.024919903600344e-05, | |
| "loss": 0.5276, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.693500298151461, | |
| "grad_norm": 1.2778667211532593, | |
| "learning_rate": 1.9942860950666574e-05, | |
| "loss": 0.5203, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.7054263565891472, | |
| "grad_norm": 1.2391997575759888, | |
| "learning_rate": 1.9637312576792776e-05, | |
| "loss": 0.5291, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.7173524150268338, | |
| "grad_norm": 1.0726789236068726, | |
| "learning_rate": 1.9332601628129128e-05, | |
| "loss": 0.5299, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.72927847346452, | |
| "grad_norm": 1.1761037111282349, | |
| "learning_rate": 1.9028775687652217e-05, | |
| "loss": 0.5476, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.7412045319022065, | |
| "grad_norm": 1.257358431816101, | |
| "learning_rate": 1.8725882200137762e-05, | |
| "loss": 0.5333, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.7531305903398926, | |
| "grad_norm": 1.121039628982544, | |
| "learning_rate": 1.8423968464751722e-05, | |
| "loss": 0.4962, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.7650566487775792, | |
| "grad_norm": 1.1800827980041504, | |
| "learning_rate": 1.812308162766418e-05, | |
| "loss": 0.4986, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.7769827072152653, | |
| "grad_norm": 1.262669324874878, | |
| "learning_rate": 1.7823268674687077e-05, | |
| "loss": 0.5461, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.7889087656529516, | |
| "grad_norm": 1.1524207592010498, | |
| "learning_rate": 1.7524576423937025e-05, | |
| "loss": 0.5392, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.800834824090638, | |
| "grad_norm": 0.9975520968437195, | |
| "learning_rate": 1.7227051518524286e-05, | |
| "loss": 0.5162, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.8127608825283243, | |
| "grad_norm": 1.1616145372390747, | |
| "learning_rate": 1.6930740419269132e-05, | |
| "loss": 0.5431, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.8246869409660107, | |
| "grad_norm": 0.9507209658622742, | |
| "learning_rate": 1.6635689397446562e-05, | |
| "loss": 0.5037, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.836612999403697, | |
| "grad_norm": 1.0409976243972778, | |
| "learning_rate": 1.6341944527560736e-05, | |
| "loss": 0.5164, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.8485390578413834, | |
| "grad_norm": 0.8729709386825562, | |
| "learning_rate": 1.6049551680150047e-05, | |
| "loss": 0.495, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.8604651162790697, | |
| "grad_norm": 1.223853588104248, | |
| "learning_rate": 1.5758556514624118e-05, | |
| "loss": 0.5834, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.872391174716756, | |
| "grad_norm": 1.184877634048462, | |
| "learning_rate": 1.5469004472133696e-05, | |
| "loss": 0.5459, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.8843172331544424, | |
| "grad_norm": 1.163509488105774, | |
| "learning_rate": 1.5180940768474689e-05, | |
| "loss": 0.516, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.8962432915921288, | |
| "grad_norm": 1.2189282178878784, | |
| "learning_rate": 1.489441038702735e-05, | |
| "loss": 0.5519, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.9081693500298151, | |
| "grad_norm": 1.1891826391220093, | |
| "learning_rate": 1.4609458071731796e-05, | |
| "loss": 0.5173, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.9200954084675015, | |
| "grad_norm": 1.3610830307006836, | |
| "learning_rate": 1.4326128320100867e-05, | |
| "loss": 0.517, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.9320214669051878, | |
| "grad_norm": 1.148544192314148, | |
| "learning_rate": 1.4044465376271532e-05, | |
| "loss": 0.4906, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.9439475253428742, | |
| "grad_norm": 1.404956340789795, | |
| "learning_rate": 1.3764513224095762e-05, | |
| "loss": 0.5598, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.9558735837805605, | |
| "grad_norm": 1.206635594367981, | |
| "learning_rate": 1.3486315580272202e-05, | |
| "loss": 0.5276, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.9677996422182469, | |
| "grad_norm": 0.8971341848373413, | |
| "learning_rate": 1.320991588751938e-05, | |
| "loss": 0.4966, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.9797257006559332, | |
| "grad_norm": 1.1437649726867676, | |
| "learning_rate": 1.2935357307791826e-05, | |
| "loss": 0.5133, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.9916517590936196, | |
| "grad_norm": 1.2897300720214844, | |
| "learning_rate": 1.2662682715540031e-05, | |
| "loss": 0.4814, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 2.0035778175313057, | |
| "grad_norm": 1.0928142070770264, | |
| "learning_rate": 1.2391934691015213e-05, | |
| "loss": 0.6167, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.0155038759689923, | |
| "grad_norm": 1.1107832193374634, | |
| "learning_rate": 1.2123155513620108e-05, | |
| "loss": 0.4915, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 2.0274299344066784, | |
| "grad_norm": 1.2382787466049194, | |
| "learning_rate": 1.1856387155306715e-05, | |
| "loss": 0.5116, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.039355992844365, | |
| "grad_norm": 1.46440589427948, | |
| "learning_rate": 1.1591671274022035e-05, | |
| "loss": 0.4754, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 2.051282051282051, | |
| "grad_norm": 1.4522420167922974, | |
| "learning_rate": 1.1329049207202904e-05, | |
| "loss": 0.4426, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.0632081097197377, | |
| "grad_norm": 1.3376868963241577, | |
| "learning_rate": 1.1068561965320764e-05, | |
| "loss": 0.5026, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 2.075134168157424, | |
| "grad_norm": 1.1326589584350586, | |
| "learning_rate": 1.0810250225477611e-05, | |
| "loss": 0.4872, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.0870602265951104, | |
| "grad_norm": 1.5386810302734375, | |
| "learning_rate": 1.055415432505393e-05, | |
| "loss": 0.5022, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.0989862850327965, | |
| "grad_norm": 1.2370398044586182, | |
| "learning_rate": 1.0300314255409704e-05, | |
| "loss": 0.5388, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.110912343470483, | |
| "grad_norm": 1.2834645509719849, | |
| "learning_rate": 1.004876965563945e-05, | |
| "loss": 0.481, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 2.122838401908169, | |
| "grad_norm": 1.3606491088867188, | |
| "learning_rate": 9.79955980638229e-06, | |
| "loss": 0.4829, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.1347644603458558, | |
| "grad_norm": 1.3206753730773926, | |
| "learning_rate": 9.552723623687934e-06, | |
| "loss": 0.4902, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 2.146690518783542, | |
| "grad_norm": 1.1643896102905273, | |
| "learning_rate": 9.308299652939666e-06, | |
| "loss": 0.5297, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.1586165772212285, | |
| "grad_norm": 1.0991599559783936, | |
| "learning_rate": 9.066326062835179e-06, | |
| "loss": 0.4845, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 2.1705426356589146, | |
| "grad_norm": 1.1832717657089233, | |
| "learning_rate": 8.826840639426218e-06, | |
| "loss": 0.5086, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.182468694096601, | |
| "grad_norm": 1.385024070739746, | |
| "learning_rate": 8.589880780218049e-06, | |
| "loss": 0.5136, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 2.1943947525342873, | |
| "grad_norm": 1.1657204627990723, | |
| "learning_rate": 8.355483488329471e-06, | |
| "loss": 0.5137, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.206320810971974, | |
| "grad_norm": 1.0423134565353394, | |
| "learning_rate": 8.123685366714556e-06, | |
| "loss": 0.5032, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.21824686940966, | |
| "grad_norm": 1.2326374053955078, | |
| "learning_rate": 7.89452261244677e-06, | |
| "loss": 0.4658, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.2301729278473466, | |
| "grad_norm": 1.3008525371551514, | |
| "learning_rate": 7.66803101106657e-06, | |
| "loss": 0.5261, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.2420989862850327, | |
| "grad_norm": 1.296058177947998, | |
| "learning_rate": 7.44424593099316e-06, | |
| "loss": 0.5216, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.2540250447227193, | |
| "grad_norm": 1.3323789834976196, | |
| "learning_rate": 7.223202318001465e-06, | |
| "loss": 0.4949, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.2659511031604054, | |
| "grad_norm": 1.2150744199752808, | |
| "learning_rate": 7.0049346897650745e-06, | |
| "loss": 0.464, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.277877161598092, | |
| "grad_norm": 1.1837234497070312, | |
| "learning_rate": 6.789477130466057e-06, | |
| "loss": 0.5057, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.289803220035778, | |
| "grad_norm": 1.0720417499542236, | |
| "learning_rate": 6.576863285472415e-06, | |
| "loss": 0.4875, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.3017292784734646, | |
| "grad_norm": 1.4399152994155884, | |
| "learning_rate": 6.367126356084127e-06, | |
| "loss": 0.4753, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.3136553369111508, | |
| "grad_norm": 1.3286051750183105, | |
| "learning_rate": 6.160299094348488e-06, | |
| "loss": 0.4725, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.3255813953488373, | |
| "grad_norm": 1.3082605600357056, | |
| "learning_rate": 5.956413797945657e-06, | |
| "loss": 0.5095, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.3375074537865235, | |
| "grad_norm": 1.4694650173187256, | |
| "learning_rate": 5.755502305145089e-06, | |
| "loss": 0.5099, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.34943351222421, | |
| "grad_norm": 1.1760421991348267, | |
| "learning_rate": 5.557595989833747e-06, | |
| "loss": 0.4744, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.361359570661896, | |
| "grad_norm": 1.5401874780654907, | |
| "learning_rate": 5.36272575661684e-06, | |
| "loss": 0.5126, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.3732856290995827, | |
| "grad_norm": 1.114990234375, | |
| "learning_rate": 5.170922035991838e-06, | |
| "loss": 0.4636, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.385211687537269, | |
| "grad_norm": 1.2573448419570923, | |
| "learning_rate": 4.9822147795964805e-06, | |
| "loss": 0.4817, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1257, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.48527209218048e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |