| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.999534233814625, | |
| "eval_steps": 500, | |
| "global_step": 1073, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004657661853749418, | |
| "grad_norm": 0.7354202100822638, | |
| "learning_rate": 9.259259259259259e-07, | |
| "loss": 0.3196, | |
| "num_tokens": 5242880.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.009315323707498836, | |
| "grad_norm": 0.5952654676340116, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.3174, | |
| "num_tokens": 10485760.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.013972985561248253, | |
| "grad_norm": 0.43877948419088947, | |
| "learning_rate": 2.7777777777777783e-06, | |
| "loss": 0.3134, | |
| "num_tokens": 15691438.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.018630647414997672, | |
| "grad_norm": 0.5019079869033612, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.3059, | |
| "num_tokens": 20915252.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02328830926874709, | |
| "grad_norm": 0.3568474515771396, | |
| "learning_rate": 4.62962962962963e-06, | |
| "loss": 0.3007, | |
| "num_tokens": 26134828.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.027945971122496506, | |
| "grad_norm": 0.32524596110161, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 0.2994, | |
| "num_tokens": 31377708.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.032603632976245925, | |
| "grad_norm": 0.3131487298314187, | |
| "learning_rate": 6.481481481481482e-06, | |
| "loss": 0.2976, | |
| "num_tokens": 36557688.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.037261294829995344, | |
| "grad_norm": 0.295150152816882, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.2949, | |
| "num_tokens": 41800322.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04191895668374476, | |
| "grad_norm": 0.27649297196341205, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.2752, | |
| "num_tokens": 47043202.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.04657661853749418, | |
| "grad_norm": 0.26536425990237617, | |
| "learning_rate": 9.25925925925926e-06, | |
| "loss": 0.2753, | |
| "num_tokens": 52286082.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05123428039124359, | |
| "grad_norm": 0.26054731579822765, | |
| "learning_rate": 1.0185185185185186e-05, | |
| "loss": 0.2753, | |
| "num_tokens": 57528962.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.05589194224499301, | |
| "grad_norm": 0.22392623893482058, | |
| "learning_rate": 1.1111111111111113e-05, | |
| "loss": 0.2678, | |
| "num_tokens": 62771842.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06054960409874243, | |
| "grad_norm": 0.24677568197943178, | |
| "learning_rate": 1.2037037037037039e-05, | |
| "loss": 0.2648, | |
| "num_tokens": 67980354.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.06520726595249185, | |
| "grad_norm": 0.23607027936747194, | |
| "learning_rate": 1.2962962962962964e-05, | |
| "loss": 0.2485, | |
| "num_tokens": 73223234.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06986492780624126, | |
| "grad_norm": 0.25205887201034227, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.2601, | |
| "num_tokens": 78437814.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.07452258965999069, | |
| "grad_norm": 0.24007224732310273, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.2644, | |
| "num_tokens": 83680694.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0791802515137401, | |
| "grad_norm": 0.21255762003515916, | |
| "learning_rate": 1.5740740740740744e-05, | |
| "loss": 0.2575, | |
| "num_tokens": 88923574.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.08383791336748952, | |
| "grad_norm": 0.2634263813918112, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.2571, | |
| "num_tokens": 94166454.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08849557522123894, | |
| "grad_norm": 0.232673109008467, | |
| "learning_rate": 1.7592592592592595e-05, | |
| "loss": 0.2574, | |
| "num_tokens": 99409334.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.09315323707498836, | |
| "grad_norm": 0.22826372859712407, | |
| "learning_rate": 1.851851851851852e-05, | |
| "loss": 0.2583, | |
| "num_tokens": 104652214.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09781089892873777, | |
| "grad_norm": 0.2448192802958887, | |
| "learning_rate": 1.9444444444444445e-05, | |
| "loss": 0.2565, | |
| "num_tokens": 109895094.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.10246856078248719, | |
| "grad_norm": 0.24004339643079584, | |
| "learning_rate": 1.995854922279793e-05, | |
| "loss": 0.2518, | |
| "num_tokens": 115137974.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.10712622263623661, | |
| "grad_norm": 0.27326023849500386, | |
| "learning_rate": 1.985492227979275e-05, | |
| "loss": 0.2596, | |
| "num_tokens": 120380854.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.11178388448998602, | |
| "grad_norm": 0.23485753253200659, | |
| "learning_rate": 1.9751295336787565e-05, | |
| "loss": 0.2602, | |
| "num_tokens": 125623734.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.11644154634373545, | |
| "grad_norm": 0.26201877946619834, | |
| "learning_rate": 1.9647668393782386e-05, | |
| "loss": 0.2515, | |
| "num_tokens": 130866614.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.12109920819748486, | |
| "grad_norm": 0.25613849410111256, | |
| "learning_rate": 1.9544041450777206e-05, | |
| "loss": 0.2539, | |
| "num_tokens": 136068456.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1257568700512343, | |
| "grad_norm": 0.25969468002736057, | |
| "learning_rate": 1.9440414507772023e-05, | |
| "loss": 0.2541, | |
| "num_tokens": 141311336.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.1304145319049837, | |
| "grad_norm": 0.22634422881885577, | |
| "learning_rate": 1.9336787564766843e-05, | |
| "loss": 0.2537, | |
| "num_tokens": 146554216.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1350721937587331, | |
| "grad_norm": 0.2522224868314366, | |
| "learning_rate": 1.923316062176166e-05, | |
| "loss": 0.2576, | |
| "num_tokens": 151797096.0, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.13972985561248252, | |
| "grad_norm": 0.23800652804398692, | |
| "learning_rate": 1.9129533678756477e-05, | |
| "loss": 0.2587, | |
| "num_tokens": 157039976.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.14438751746623196, | |
| "grad_norm": 0.2568784521668884, | |
| "learning_rate": 1.9025906735751297e-05, | |
| "loss": 0.2511, | |
| "num_tokens": 162282856.0, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.14904517931998137, | |
| "grad_norm": 0.2554600443840236, | |
| "learning_rate": 1.8922279792746117e-05, | |
| "loss": 0.2607, | |
| "num_tokens": 167466294.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1537028411737308, | |
| "grad_norm": 0.2711816282826398, | |
| "learning_rate": 1.8818652849740934e-05, | |
| "loss": 0.2527, | |
| "num_tokens": 172703130.0, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.1583605030274802, | |
| "grad_norm": 0.2765740820594681, | |
| "learning_rate": 1.8715025906735754e-05, | |
| "loss": 0.2562, | |
| "num_tokens": 177946010.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1630181648812296, | |
| "grad_norm": 0.24459888224432297, | |
| "learning_rate": 1.861139896373057e-05, | |
| "loss": 0.2605, | |
| "num_tokens": 183145930.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.16767582673497905, | |
| "grad_norm": 0.22491101674717587, | |
| "learning_rate": 1.850777202072539e-05, | |
| "loss": 0.2548, | |
| "num_tokens": 188388810.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.17233348858872846, | |
| "grad_norm": 0.2547998556520989, | |
| "learning_rate": 1.8404145077720208e-05, | |
| "loss": 0.2626, | |
| "num_tokens": 193628148.0, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.17699115044247787, | |
| "grad_norm": 0.22561482520286322, | |
| "learning_rate": 1.8300518134715028e-05, | |
| "loss": 0.2557, | |
| "num_tokens": 198871028.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.18164881229622729, | |
| "grad_norm": 0.29034056879576386, | |
| "learning_rate": 1.8196891191709845e-05, | |
| "loss": 0.2606, | |
| "num_tokens": 204113908.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.18630647414997673, | |
| "grad_norm": 0.25443247266585134, | |
| "learning_rate": 1.8093264248704665e-05, | |
| "loss": 0.2521, | |
| "num_tokens": 209266418.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.19096413600372614, | |
| "grad_norm": 0.24563579984761633, | |
| "learning_rate": 1.7989637305699482e-05, | |
| "loss": 0.257, | |
| "num_tokens": 214355738.0, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.19562179785747555, | |
| "grad_norm": 0.21968633620629222, | |
| "learning_rate": 1.7886010362694302e-05, | |
| "loss": 0.2598, | |
| "num_tokens": 219598618.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.20027945971122496, | |
| "grad_norm": 0.23648866761989712, | |
| "learning_rate": 1.778238341968912e-05, | |
| "loss": 0.2596, | |
| "num_tokens": 224779924.0, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.20493712156497437, | |
| "grad_norm": 0.23777490605559648, | |
| "learning_rate": 1.767875647668394e-05, | |
| "loss": 0.2535, | |
| "num_tokens": 230001220.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2095947834187238, | |
| "grad_norm": 0.24256243884671247, | |
| "learning_rate": 1.757512953367876e-05, | |
| "loss": 0.2652, | |
| "num_tokens": 235244100.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.21425244527247322, | |
| "grad_norm": 0.24176306110223997, | |
| "learning_rate": 1.7471502590673576e-05, | |
| "loss": 0.2601, | |
| "num_tokens": 240459436.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.21891010712622264, | |
| "grad_norm": 0.2486438604804707, | |
| "learning_rate": 1.7367875647668397e-05, | |
| "loss": 0.2581, | |
| "num_tokens": 245702316.0, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.22356776897997205, | |
| "grad_norm": 0.21873926029814306, | |
| "learning_rate": 1.7264248704663214e-05, | |
| "loss": 0.2577, | |
| "num_tokens": 250945196.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.22822543083372146, | |
| "grad_norm": 0.21237096265719993, | |
| "learning_rate": 1.716062176165803e-05, | |
| "loss": 0.2501, | |
| "num_tokens": 256130440.0, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2328830926874709, | |
| "grad_norm": 0.23047577910983572, | |
| "learning_rate": 1.705699481865285e-05, | |
| "loss": 0.252, | |
| "num_tokens": 261373320.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2375407545412203, | |
| "grad_norm": 0.2554824770015439, | |
| "learning_rate": 1.695336787564767e-05, | |
| "loss": 0.2586, | |
| "num_tokens": 266595056.0, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.24219841639496972, | |
| "grad_norm": 0.25106744384452817, | |
| "learning_rate": 1.6849740932642488e-05, | |
| "loss": 0.2608, | |
| "num_tokens": 271837936.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.24685607824871914, | |
| "grad_norm": 0.21461956509187238, | |
| "learning_rate": 1.6746113989637308e-05, | |
| "loss": 0.2552, | |
| "num_tokens": 277080816.0, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2515137401024686, | |
| "grad_norm": 0.2333994074126637, | |
| "learning_rate": 1.6642487046632125e-05, | |
| "loss": 0.2524, | |
| "num_tokens": 282323696.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.25617140195621796, | |
| "grad_norm": 0.2727418904587689, | |
| "learning_rate": 1.6538860103626945e-05, | |
| "loss": 0.262, | |
| "num_tokens": 287548638.0, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.2608290638099674, | |
| "grad_norm": 0.2552685407130407, | |
| "learning_rate": 1.6435233160621765e-05, | |
| "loss": 0.2512, | |
| "num_tokens": 292791518.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.26548672566371684, | |
| "grad_norm": 0.22432212853480846, | |
| "learning_rate": 1.6331606217616582e-05, | |
| "loss": 0.258, | |
| "num_tokens": 298034398.0, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.2701443875174662, | |
| "grad_norm": 0.2412274852289003, | |
| "learning_rate": 1.6227979274611402e-05, | |
| "loss": 0.2513, | |
| "num_tokens": 303277278.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.27480204937121566, | |
| "grad_norm": 0.24631119063401033, | |
| "learning_rate": 1.612435233160622e-05, | |
| "loss": 0.2526, | |
| "num_tokens": 308520158.0, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.27945971122496505, | |
| "grad_norm": 0.2549591348218789, | |
| "learning_rate": 1.6020725388601036e-05, | |
| "loss": 0.2616, | |
| "num_tokens": 313763038.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2841173730787145, | |
| "grad_norm": 0.22557733703619626, | |
| "learning_rate": 1.5917098445595856e-05, | |
| "loss": 0.2679, | |
| "num_tokens": 318941236.0, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.2887750349324639, | |
| "grad_norm": 0.25941400928229763, | |
| "learning_rate": 1.5813471502590673e-05, | |
| "loss": 0.2562, | |
| "num_tokens": 324184116.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2934326967862133, | |
| "grad_norm": 0.21473020825163133, | |
| "learning_rate": 1.5709844559585493e-05, | |
| "loss": 0.2504, | |
| "num_tokens": 329426996.0, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.29809035863996275, | |
| "grad_norm": 0.2582115110100909, | |
| "learning_rate": 1.5606217616580313e-05, | |
| "loss": 0.2566, | |
| "num_tokens": 334669876.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.30274802049371213, | |
| "grad_norm": 0.23341581985658094, | |
| "learning_rate": 1.550259067357513e-05, | |
| "loss": 0.2585, | |
| "num_tokens": 339912756.0, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.3074056823474616, | |
| "grad_norm": 0.25568740020118363, | |
| "learning_rate": 1.539896373056995e-05, | |
| "loss": 0.2569, | |
| "num_tokens": 345155636.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.312063344201211, | |
| "grad_norm": 0.23524769292076683, | |
| "learning_rate": 1.5295336787564767e-05, | |
| "loss": 0.2523, | |
| "num_tokens": 350398516.0, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3167210060549604, | |
| "grad_norm": 0.21901504673475616, | |
| "learning_rate": 1.5191709844559586e-05, | |
| "loss": 0.2596, | |
| "num_tokens": 355641396.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.32137866790870984, | |
| "grad_norm": 0.21218687174498777, | |
| "learning_rate": 1.5088082901554406e-05, | |
| "loss": 0.2508, | |
| "num_tokens": 360884276.0, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.3260363297624592, | |
| "grad_norm": 0.2429427262409178, | |
| "learning_rate": 1.4984455958549225e-05, | |
| "loss": 0.2559, | |
| "num_tokens": 366127156.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.33069399161620866, | |
| "grad_norm": 0.23262133648796862, | |
| "learning_rate": 1.4880829015544043e-05, | |
| "loss": 0.2545, | |
| "num_tokens": 371370036.0, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3353516534699581, | |
| "grad_norm": 0.2661672340086952, | |
| "learning_rate": 1.4777202072538862e-05, | |
| "loss": 0.2573, | |
| "num_tokens": 376600048.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3400093153237075, | |
| "grad_norm": 0.25237609053275833, | |
| "learning_rate": 1.4673575129533678e-05, | |
| "loss": 0.2582, | |
| "num_tokens": 381842928.0, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.3446669771774569, | |
| "grad_norm": 0.27096117370330536, | |
| "learning_rate": 1.4569948186528497e-05, | |
| "loss": 0.2687, | |
| "num_tokens": 387073976.0, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3493246390312063, | |
| "grad_norm": 0.25747433192796804, | |
| "learning_rate": 1.4466321243523317e-05, | |
| "loss": 0.2604, | |
| "num_tokens": 392316856.0, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 0.22767537032556018, | |
| "learning_rate": 1.4362694300518136e-05, | |
| "loss": 0.2507, | |
| "num_tokens": 397559736.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3586399627387052, | |
| "grad_norm": 0.21105495940834915, | |
| "learning_rate": 1.4259067357512954e-05, | |
| "loss": 0.2452, | |
| "num_tokens": 402802616.0, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.36329762459245457, | |
| "grad_norm": 0.2241169206183185, | |
| "learning_rate": 1.4155440414507773e-05, | |
| "loss": 0.2555, | |
| "num_tokens": 408014414.0, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.367955286446204, | |
| "grad_norm": 0.21730493901887643, | |
| "learning_rate": 1.4051813471502591e-05, | |
| "loss": 0.2534, | |
| "num_tokens": 413257294.0, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.37261294829995345, | |
| "grad_norm": 0.22197460277482448, | |
| "learning_rate": 1.394818652849741e-05, | |
| "loss": 0.2517, | |
| "num_tokens": 418500174.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.37727061015370283, | |
| "grad_norm": 0.22146782224116823, | |
| "learning_rate": 1.384455958549223e-05, | |
| "loss": 0.2535, | |
| "num_tokens": 423692782.0, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.3819282720074523, | |
| "grad_norm": 0.2395587667676847, | |
| "learning_rate": 1.3740932642487049e-05, | |
| "loss": 0.2558, | |
| "num_tokens": 428845062.0, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.38658593386120166, | |
| "grad_norm": 0.27096838638774945, | |
| "learning_rate": 1.3637305699481867e-05, | |
| "loss": 0.2558, | |
| "num_tokens": 434087942.0, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.3912435957149511, | |
| "grad_norm": 0.2526407628179118, | |
| "learning_rate": 1.3533678756476684e-05, | |
| "loss": 0.2507, | |
| "num_tokens": 439278078.0, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.39590125756870054, | |
| "grad_norm": 0.24764029701669407, | |
| "learning_rate": 1.3430051813471503e-05, | |
| "loss": 0.2676, | |
| "num_tokens": 444520958.0, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.4005589194224499, | |
| "grad_norm": 0.2364638367260199, | |
| "learning_rate": 1.3326424870466321e-05, | |
| "loss": 0.2525, | |
| "num_tokens": 449763838.0, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.40521658127619936, | |
| "grad_norm": 0.21509051339514587, | |
| "learning_rate": 1.3222797927461141e-05, | |
| "loss": 0.2544, | |
| "num_tokens": 454963270.0, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.40987424312994875, | |
| "grad_norm": 0.2410470274612002, | |
| "learning_rate": 1.311917098445596e-05, | |
| "loss": 0.252, | |
| "num_tokens": 460206150.0, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4145319049836982, | |
| "grad_norm": 0.20592818916644473, | |
| "learning_rate": 1.3015544041450778e-05, | |
| "loss": 0.2521, | |
| "num_tokens": 465378662.0, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.4191895668374476, | |
| "grad_norm": 0.2264587781053658, | |
| "learning_rate": 1.2911917098445597e-05, | |
| "loss": 0.2586, | |
| "num_tokens": 470602566.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.423847228691197, | |
| "grad_norm": 0.25655803861329096, | |
| "learning_rate": 1.2808290155440415e-05, | |
| "loss": 0.266, | |
| "num_tokens": 475800020.0, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.42850489054494645, | |
| "grad_norm": 0.22574245461885561, | |
| "learning_rate": 1.2704663212435234e-05, | |
| "loss": 0.2542, | |
| "num_tokens": 481042900.0, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.43316255239869583, | |
| "grad_norm": 0.22808862494719606, | |
| "learning_rate": 1.2601036269430054e-05, | |
| "loss": 0.261, | |
| "num_tokens": 486220500.0, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.43782021425244527, | |
| "grad_norm": 0.23099682460823345, | |
| "learning_rate": 1.2497409326424873e-05, | |
| "loss": 0.2543, | |
| "num_tokens": 491463380.0, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4424778761061947, | |
| "grad_norm": 0.23489365109623844, | |
| "learning_rate": 1.239378238341969e-05, | |
| "loss": 0.2629, | |
| "num_tokens": 496706260.0, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.4471355379599441, | |
| "grad_norm": 0.20523732690325006, | |
| "learning_rate": 1.2290155440414508e-05, | |
| "loss": 0.2572, | |
| "num_tokens": 501949140.0, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.45179319981369354, | |
| "grad_norm": 0.2099895660197865, | |
| "learning_rate": 1.2186528497409327e-05, | |
| "loss": 0.2513, | |
| "num_tokens": 507192020.0, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.4564508616674429, | |
| "grad_norm": 0.2135358910850049, | |
| "learning_rate": 1.2082901554404145e-05, | |
| "loss": 0.2542, | |
| "num_tokens": 512433054.0, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.46110852352119236, | |
| "grad_norm": 0.20900397680709945, | |
| "learning_rate": 1.1979274611398965e-05, | |
| "loss": 0.2498, | |
| "num_tokens": 517658160.0, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.4657661853749418, | |
| "grad_norm": 0.23907354699307065, | |
| "learning_rate": 1.1875647668393784e-05, | |
| "loss": 0.2547, | |
| "num_tokens": 522901040.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4704238472286912, | |
| "grad_norm": 0.21616411710432162, | |
| "learning_rate": 1.1772020725388602e-05, | |
| "loss": 0.2526, | |
| "num_tokens": 528127586.0, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.4750815090824406, | |
| "grad_norm": 0.2098299301335155, | |
| "learning_rate": 1.1668393782383421e-05, | |
| "loss": 0.2531, | |
| "num_tokens": 533355284.0, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.47973917093619, | |
| "grad_norm": 0.251556887150099, | |
| "learning_rate": 1.1564766839378238e-05, | |
| "loss": 0.2505, | |
| "num_tokens": 538598164.0, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.48439683278993945, | |
| "grad_norm": 0.2339273763266774, | |
| "learning_rate": 1.1461139896373056e-05, | |
| "loss": 0.2568, | |
| "num_tokens": 543841044.0, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.4890544946436889, | |
| "grad_norm": 0.22013897389362086, | |
| "learning_rate": 1.1357512953367878e-05, | |
| "loss": 0.26, | |
| "num_tokens": 549083924.0, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.49371215649743827, | |
| "grad_norm": 0.21805387603575355, | |
| "learning_rate": 1.1253886010362695e-05, | |
| "loss": 0.2582, | |
| "num_tokens": 554323932.0, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4983698183511877, | |
| "grad_norm": 0.2342930652112791, | |
| "learning_rate": 1.1150259067357514e-05, | |
| "loss": 0.2555, | |
| "num_tokens": 559555680.0, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5030274802049371, | |
| "grad_norm": 0.2557255664414066, | |
| "learning_rate": 1.1046632124352332e-05, | |
| "loss": 0.2627, | |
| "num_tokens": 564798560.0, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5076851420586865, | |
| "grad_norm": 0.2400422425402717, | |
| "learning_rate": 1.094300518134715e-05, | |
| "loss": 0.2518, | |
| "num_tokens": 569918486.0, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5123428039124359, | |
| "grad_norm": 0.2802173303892893, | |
| "learning_rate": 1.083937823834197e-05, | |
| "loss": 0.2592, | |
| "num_tokens": 575160720.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5170004657661854, | |
| "grad_norm": 0.2487334570481747, | |
| "learning_rate": 1.073575129533679e-05, | |
| "loss": 0.2607, | |
| "num_tokens": 580403600.0, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.5216581276199348, | |
| "grad_norm": 0.2149721507442794, | |
| "learning_rate": 1.0632124352331608e-05, | |
| "loss": 0.2606, | |
| "num_tokens": 585635440.0, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 0.2233187847444525, | |
| "learning_rate": 1.0528497409326426e-05, | |
| "loss": 0.2594, | |
| "num_tokens": 590878320.0, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.5309734513274337, | |
| "grad_norm": 0.2278330157191538, | |
| "learning_rate": 1.0424870466321243e-05, | |
| "loss": 0.2559, | |
| "num_tokens": 596063606.0, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5356311131811831, | |
| "grad_norm": 0.22928159523756741, | |
| "learning_rate": 1.0321243523316062e-05, | |
| "loss": 0.2518, | |
| "num_tokens": 601306486.0, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.5402887750349324, | |
| "grad_norm": 0.2180906186504879, | |
| "learning_rate": 1.021761658031088e-05, | |
| "loss": 0.2526, | |
| "num_tokens": 606451006.0, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5449464368886818, | |
| "grad_norm": 0.19971415049523444, | |
| "learning_rate": 1.01139896373057e-05, | |
| "loss": 0.2504, | |
| "num_tokens": 611693886.0, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.5496040987424313, | |
| "grad_norm": 0.23365868641736956, | |
| "learning_rate": 1.0010362694300519e-05, | |
| "loss": 0.2583, | |
| "num_tokens": 616936766.0, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5542617605961807, | |
| "grad_norm": 0.24974952391002533, | |
| "learning_rate": 9.906735751295338e-06, | |
| "loss": 0.2548, | |
| "num_tokens": 622179646.0, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.5589194224499301, | |
| "grad_norm": 0.22082467861156416, | |
| "learning_rate": 9.803108808290156e-06, | |
| "loss": 0.2613, | |
| "num_tokens": 627422526.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5635770843036796, | |
| "grad_norm": 0.2007903670054586, | |
| "learning_rate": 9.699481865284975e-06, | |
| "loss": 0.2509, | |
| "num_tokens": 632619928.0, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.568234746157429, | |
| "grad_norm": 0.22445366837916325, | |
| "learning_rate": 9.595854922279793e-06, | |
| "loss": 0.2539, | |
| "num_tokens": 637862808.0, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5728924080111784, | |
| "grad_norm": 0.22967464133749027, | |
| "learning_rate": 9.492227979274612e-06, | |
| "loss": 0.253, | |
| "num_tokens": 643105688.0, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.5775500698649279, | |
| "grad_norm": 0.2516176298289619, | |
| "learning_rate": 9.388601036269432e-06, | |
| "loss": 0.2525, | |
| "num_tokens": 648348568.0, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5822077317186772, | |
| "grad_norm": 0.24426274764470918, | |
| "learning_rate": 9.284974093264249e-06, | |
| "loss": 0.2587, | |
| "num_tokens": 653591448.0, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.5868653935724266, | |
| "grad_norm": 0.22061859638423034, | |
| "learning_rate": 9.181347150259067e-06, | |
| "loss": 0.2565, | |
| "num_tokens": 658834328.0, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.5915230554261761, | |
| "grad_norm": 0.2253838500280039, | |
| "learning_rate": 9.077720207253888e-06, | |
| "loss": 0.2593, | |
| "num_tokens": 664028780.0, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.5961807172799255, | |
| "grad_norm": 0.24259696163395483, | |
| "learning_rate": 8.974093264248706e-06, | |
| "loss": 0.2567, | |
| "num_tokens": 669256250.0, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6008383791336749, | |
| "grad_norm": 0.20792994997638586, | |
| "learning_rate": 8.870466321243523e-06, | |
| "loss": 0.2535, | |
| "num_tokens": 674460550.0, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6054960409874243, | |
| "grad_norm": 0.22514234774379766, | |
| "learning_rate": 8.766839378238343e-06, | |
| "loss": 0.2589, | |
| "num_tokens": 679701528.0, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6101537028411738, | |
| "grad_norm": 0.22252529285070463, | |
| "learning_rate": 8.663212435233162e-06, | |
| "loss": 0.2583, | |
| "num_tokens": 684944408.0, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.6148113646949231, | |
| "grad_norm": 0.22429197369560439, | |
| "learning_rate": 8.55958549222798e-06, | |
| "loss": 0.2594, | |
| "num_tokens": 690187288.0, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6194690265486725, | |
| "grad_norm": 0.215043612295039, | |
| "learning_rate": 8.455958549222799e-06, | |
| "loss": 0.2564, | |
| "num_tokens": 695372640.0, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.624126688402422, | |
| "grad_norm": 0.23264384545455694, | |
| "learning_rate": 8.352331606217617e-06, | |
| "loss": 0.2607, | |
| "num_tokens": 700615520.0, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6287843502561714, | |
| "grad_norm": 0.22834680912110178, | |
| "learning_rate": 8.248704663212436e-06, | |
| "loss": 0.2531, | |
| "num_tokens": 705831004.0, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.6334420121099208, | |
| "grad_norm": 0.22671009140383497, | |
| "learning_rate": 8.145077720207254e-06, | |
| "loss": 0.2572, | |
| "num_tokens": 711015208.0, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6380996739636703, | |
| "grad_norm": 0.2252785168427841, | |
| "learning_rate": 8.041450777202073e-06, | |
| "loss": 0.2612, | |
| "num_tokens": 716205240.0, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.6427573358174197, | |
| "grad_norm": 0.21656916692135098, | |
| "learning_rate": 7.937823834196891e-06, | |
| "loss": 0.2561, | |
| "num_tokens": 721448120.0, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6474149976711691, | |
| "grad_norm": 0.22150705991215874, | |
| "learning_rate": 7.834196891191712e-06, | |
| "loss": 0.253, | |
| "num_tokens": 726691000.0, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.6520726595249184, | |
| "grad_norm": 0.20529604986432817, | |
| "learning_rate": 7.730569948186528e-06, | |
| "loss": 0.2493, | |
| "num_tokens": 731930908.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6567303213786679, | |
| "grad_norm": 0.20637008610419608, | |
| "learning_rate": 7.626943005181348e-06, | |
| "loss": 0.2547, | |
| "num_tokens": 737133156.0, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.6613879832324173, | |
| "grad_norm": 0.21054817618044103, | |
| "learning_rate": 7.523316062176167e-06, | |
| "loss": 0.2593, | |
| "num_tokens": 742336330.0, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6660456450861667, | |
| "grad_norm": 0.21270024962264275, | |
| "learning_rate": 7.419689119170985e-06, | |
| "loss": 0.2582, | |
| "num_tokens": 747579210.0, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.6707033069399162, | |
| "grad_norm": 0.19546074259353546, | |
| "learning_rate": 7.3160621761658035e-06, | |
| "loss": 0.2608, | |
| "num_tokens": 752752000.0, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.6753609687936656, | |
| "grad_norm": 0.21890077846741351, | |
| "learning_rate": 7.212435233160623e-06, | |
| "loss": 0.2554, | |
| "num_tokens": 757994880.0, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.680018630647415, | |
| "grad_norm": 0.22539862112894182, | |
| "learning_rate": 7.108808290155441e-06, | |
| "loss": 0.2552, | |
| "num_tokens": 763237760.0, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6846762925011645, | |
| "grad_norm": 0.20207855236453137, | |
| "learning_rate": 7.005181347150259e-06, | |
| "loss": 0.2488, | |
| "num_tokens": 768480640.0, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.6893339543549138, | |
| "grad_norm": 0.20415283157246739, | |
| "learning_rate": 6.9015544041450784e-06, | |
| "loss": 0.2572, | |
| "num_tokens": 773723520.0, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6939916162086632, | |
| "grad_norm": 0.22081176128637509, | |
| "learning_rate": 6.797927461139897e-06, | |
| "loss": 0.2554, | |
| "num_tokens": 778966400.0, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.6986492780624126, | |
| "grad_norm": 0.20761176823421718, | |
| "learning_rate": 6.6943005181347155e-06, | |
| "loss": 0.2578, | |
| "num_tokens": 784209280.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7033069399161621, | |
| "grad_norm": 0.20641059761415284, | |
| "learning_rate": 6.590673575129535e-06, | |
| "loss": 0.2536, | |
| "num_tokens": 789452160.0, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 0.21141222664298415, | |
| "learning_rate": 6.487046632124353e-06, | |
| "loss": 0.2522, | |
| "num_tokens": 794695040.0, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7126222636236609, | |
| "grad_norm": 0.21470008081540384, | |
| "learning_rate": 6.383419689119171e-06, | |
| "loss": 0.2546, | |
| "num_tokens": 799937676.0, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.7172799254774104, | |
| "grad_norm": 0.2224245999859149, | |
| "learning_rate": 6.2797927461139905e-06, | |
| "loss": 0.2558, | |
| "num_tokens": 805180556.0, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.7219375873311598, | |
| "grad_norm": 0.25633254604263456, | |
| "learning_rate": 6.176165803108809e-06, | |
| "loss": 0.2548, | |
| "num_tokens": 810397834.0, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.7265952491849091, | |
| "grad_norm": 0.22627553864771588, | |
| "learning_rate": 6.0725388601036275e-06, | |
| "loss": 0.2557, | |
| "num_tokens": 815565546.0, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7312529110386586, | |
| "grad_norm": 0.2086026526549718, | |
| "learning_rate": 5.968911917098445e-06, | |
| "loss": 0.2555, | |
| "num_tokens": 820802646.0, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.735910572892408, | |
| "grad_norm": 0.20950901705644956, | |
| "learning_rate": 5.865284974093265e-06, | |
| "loss": 0.2563, | |
| "num_tokens": 826045526.0, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.7405682347461574, | |
| "grad_norm": 0.20425765360730283, | |
| "learning_rate": 5.761658031088083e-06, | |
| "loss": 0.2474, | |
| "num_tokens": 831245076.0, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.7452258965999069, | |
| "grad_norm": 0.22176868184718157, | |
| "learning_rate": 5.658031088082902e-06, | |
| "loss": 0.2553, | |
| "num_tokens": 836487956.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7498835584536563, | |
| "grad_norm": 0.20901479921501584, | |
| "learning_rate": 5.554404145077721e-06, | |
| "loss": 0.2656, | |
| "num_tokens": 841670226.0, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.7545412203074057, | |
| "grad_norm": 0.19627580251663526, | |
| "learning_rate": 5.4507772020725395e-06, | |
| "loss": 0.259, | |
| "num_tokens": 846891280.0, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.759198882161155, | |
| "grad_norm": 0.20804229040608027, | |
| "learning_rate": 5.347150259067357e-06, | |
| "loss": 0.2624, | |
| "num_tokens": 852076096.0, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.7638565440149045, | |
| "grad_norm": 0.201694556902628, | |
| "learning_rate": 5.243523316062177e-06, | |
| "loss": 0.2548, | |
| "num_tokens": 857318976.0, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.7685142058686539, | |
| "grad_norm": 0.20232675891945648, | |
| "learning_rate": 5.139896373056995e-06, | |
| "loss": 0.2565, | |
| "num_tokens": 862561856.0, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.7731718677224033, | |
| "grad_norm": 0.20294268616748656, | |
| "learning_rate": 5.036269430051814e-06, | |
| "loss": 0.2529, | |
| "num_tokens": 867804736.0, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.7778295295761528, | |
| "grad_norm": 0.21461420143759774, | |
| "learning_rate": 4.932642487046633e-06, | |
| "loss": 0.2558, | |
| "num_tokens": 873045328.0, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.7824871914299022, | |
| "grad_norm": 0.20318596772981168, | |
| "learning_rate": 4.829015544041451e-06, | |
| "loss": 0.2501, | |
| "num_tokens": 878249566.0, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.7871448532836516, | |
| "grad_norm": 0.22569842123192363, | |
| "learning_rate": 4.72538860103627e-06, | |
| "loss": 0.2539, | |
| "num_tokens": 883492446.0, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.7918025151374011, | |
| "grad_norm": 0.2135171582667774, | |
| "learning_rate": 4.621761658031089e-06, | |
| "loss": 0.2565, | |
| "num_tokens": 888735326.0, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.7964601769911505, | |
| "grad_norm": 0.20694406745039767, | |
| "learning_rate": 4.518134715025907e-06, | |
| "loss": 0.2547, | |
| "num_tokens": 893978206.0, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.8011178388448998, | |
| "grad_norm": 0.20744719753065266, | |
| "learning_rate": 4.414507772020726e-06, | |
| "loss": 0.2553, | |
| "num_tokens": 899180370.0, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.8057755006986492, | |
| "grad_norm": 0.19522748984864885, | |
| "learning_rate": 4.310880829015544e-06, | |
| "loss": 0.2575, | |
| "num_tokens": 904405480.0, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.8104331625523987, | |
| "grad_norm": 0.23294784646343103, | |
| "learning_rate": 4.207253886010363e-06, | |
| "loss": 0.2509, | |
| "num_tokens": 909648360.0, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.8150908244061481, | |
| "grad_norm": 0.2050630764552645, | |
| "learning_rate": 4.103626943005182e-06, | |
| "loss": 0.2613, | |
| "num_tokens": 914891240.0, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.8197484862598975, | |
| "grad_norm": 0.19176161559505597, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.2512, | |
| "num_tokens": 920134120.0, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.824406148113647, | |
| "grad_norm": 0.19654285773639535, | |
| "learning_rate": 3.896373056994819e-06, | |
| "loss": 0.2533, | |
| "num_tokens": 925348344.0, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.8290638099673964, | |
| "grad_norm": 0.2010009018948378, | |
| "learning_rate": 3.7927461139896377e-06, | |
| "loss": 0.2561, | |
| "num_tokens": 930591224.0, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.8337214718211458, | |
| "grad_norm": 0.2104741532182875, | |
| "learning_rate": 3.6891191709844567e-06, | |
| "loss": 0.2578, | |
| "num_tokens": 935834104.0, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.8383791336748952, | |
| "grad_norm": 0.21493823596738523, | |
| "learning_rate": 3.5854922279792748e-06, | |
| "loss": 0.2515, | |
| "num_tokens": 941076984.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.8430367955286446, | |
| "grad_norm": 0.21807351695749647, | |
| "learning_rate": 3.4818652849740937e-06, | |
| "loss": 0.2539, | |
| "num_tokens": 946319864.0, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.847694457382394, | |
| "grad_norm": 0.1988482880849994, | |
| "learning_rate": 3.3782383419689123e-06, | |
| "loss": 0.2591, | |
| "num_tokens": 951515422.0, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.8523521192361434, | |
| "grad_norm": 0.20318657661837963, | |
| "learning_rate": 3.274611398963731e-06, | |
| "loss": 0.2578, | |
| "num_tokens": 956704248.0, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.8570097810898929, | |
| "grad_norm": 0.20729484609083484, | |
| "learning_rate": 3.1709844559585493e-06, | |
| "loss": 0.2596, | |
| "num_tokens": 961947128.0, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.8616674429436423, | |
| "grad_norm": 0.20160433894055318, | |
| "learning_rate": 3.0673575129533683e-06, | |
| "loss": 0.2565, | |
| "num_tokens": 967190008.0, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.8663251047973917, | |
| "grad_norm": 0.19780649285588722, | |
| "learning_rate": 2.963730569948187e-06, | |
| "loss": 0.2565, | |
| "num_tokens": 972432888.0, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.8709827666511412, | |
| "grad_norm": 0.19348229043943008, | |
| "learning_rate": 2.8601036269430053e-06, | |
| "loss": 0.2527, | |
| "num_tokens": 977675768.0, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.8756404285048905, | |
| "grad_norm": 0.19368600162131364, | |
| "learning_rate": 2.7564766839378243e-06, | |
| "loss": 0.2528, | |
| "num_tokens": 982908826.0, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.8802980903586399, | |
| "grad_norm": 0.19463123376478544, | |
| "learning_rate": 2.6528497409326424e-06, | |
| "loss": 0.2526, | |
| "num_tokens": 988129568.0, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.8849557522123894, | |
| "grad_norm": 0.21523758048888883, | |
| "learning_rate": 2.5492227979274614e-06, | |
| "loss": 0.25, | |
| "num_tokens": 993372448.0, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.8896134140661388, | |
| "grad_norm": 0.1998332679699758, | |
| "learning_rate": 2.44559585492228e-06, | |
| "loss": 0.256, | |
| "num_tokens": 998615328.0, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.8942710759198882, | |
| "grad_norm": 0.2407703669071803, | |
| "learning_rate": 2.3419689119170984e-06, | |
| "loss": 0.2564, | |
| "num_tokens": 1003840426.0, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.8989287377736377, | |
| "grad_norm": 0.20294711945138116, | |
| "learning_rate": 2.2383419689119174e-06, | |
| "loss": 0.2567, | |
| "num_tokens": 1009083306.0, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.9035863996273871, | |
| "grad_norm": 0.2009843934095899, | |
| "learning_rate": 2.134715025906736e-06, | |
| "loss": 0.2599, | |
| "num_tokens": 1014326186.0, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.9082440614811365, | |
| "grad_norm": 0.20111239212224993, | |
| "learning_rate": 2.0310880829015544e-06, | |
| "loss": 0.2496, | |
| "num_tokens": 1019569066.0, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.9129017233348858, | |
| "grad_norm": 0.211111112971013, | |
| "learning_rate": 1.9274611398963734e-06, | |
| "loss": 0.2559, | |
| "num_tokens": 1024811946.0, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.9175593851886353, | |
| "grad_norm": 0.19512540112769755, | |
| "learning_rate": 1.823834196891192e-06, | |
| "loss": 0.2545, | |
| "num_tokens": 1030054826.0, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.9222170470423847, | |
| "grad_norm": 0.19366074847807285, | |
| "learning_rate": 1.7202072538860104e-06, | |
| "loss": 0.2476, | |
| "num_tokens": 1035279144.0, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.9268747088961341, | |
| "grad_norm": 0.19609442419329468, | |
| "learning_rate": 1.6165803108808292e-06, | |
| "loss": 0.2505, | |
| "num_tokens": 1040522024.0, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.9315323707498836, | |
| "grad_norm": 0.19637536941923484, | |
| "learning_rate": 1.5129533678756477e-06, | |
| "loss": 0.2621, | |
| "num_tokens": 1045764904.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.936190032603633, | |
| "grad_norm": 0.19240257426466145, | |
| "learning_rate": 1.4093264248704663e-06, | |
| "loss": 0.2569, | |
| "num_tokens": 1051007784.0, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.9408476944573824, | |
| "grad_norm": 0.21559906809271565, | |
| "learning_rate": 1.3056994818652852e-06, | |
| "loss": 0.256, | |
| "num_tokens": 1056250664.0, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.9455053563111319, | |
| "grad_norm": 0.18577975032596017, | |
| "learning_rate": 1.2020725388601037e-06, | |
| "loss": 0.2578, | |
| "num_tokens": 1061493544.0, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.9501630181648812, | |
| "grad_norm": 0.19056451193579874, | |
| "learning_rate": 1.0984455958549225e-06, | |
| "loss": 0.2569, | |
| "num_tokens": 1066736424.0, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.9548206800186306, | |
| "grad_norm": 0.186349857791698, | |
| "learning_rate": 9.94818652849741e-07, | |
| "loss": 0.2495, | |
| "num_tokens": 1071979304.0, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.95947834187238, | |
| "grad_norm": 0.18365015032875678, | |
| "learning_rate": 8.911917098445596e-07, | |
| "loss": 0.2486, | |
| "num_tokens": 1077222184.0, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.9641360037261295, | |
| "grad_norm": 0.19412525794148613, | |
| "learning_rate": 7.875647668393784e-07, | |
| "loss": 0.2514, | |
| "num_tokens": 1082465064.0, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.9687936655798789, | |
| "grad_norm": 0.19370419262849992, | |
| "learning_rate": 6.839378238341969e-07, | |
| "loss": 0.2526, | |
| "num_tokens": 1087707944.0, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.9734513274336283, | |
| "grad_norm": 0.19588742190290914, | |
| "learning_rate": 5.803108808290156e-07, | |
| "loss": 0.2556, | |
| "num_tokens": 1092923046.0, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.9781089892873778, | |
| "grad_norm": 0.19865808017580125, | |
| "learning_rate": 4.7668393782383424e-07, | |
| "loss": 0.2552, | |
| "num_tokens": 1098165926.0, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.9827666511411272, | |
| "grad_norm": 0.19403189440332239, | |
| "learning_rate": 3.730569948186528e-07, | |
| "loss": 0.2509, | |
| "num_tokens": 1103324696.0, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.9874243129948765, | |
| "grad_norm": 0.20419846254982182, | |
| "learning_rate": 2.694300518134715e-07, | |
| "loss": 0.2582, | |
| "num_tokens": 1108567576.0, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.992081974848626, | |
| "grad_norm": 0.18937813994316932, | |
| "learning_rate": 1.6580310880829015e-07, | |
| "loss": 0.2517, | |
| "num_tokens": 1113810456.0, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.9967396367023754, | |
| "grad_norm": 0.1911183837465565, | |
| "learning_rate": 6.217616580310881e-08, | |
| "loss": 0.2494, | |
| "num_tokens": 1119053336.0, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.999534233814625, | |
| "num_tokens": 1122199064.0, | |
| "step": 1073, | |
| "total_flos": 9.18849215432491e+17, | |
| "train_loss": 0.25810016151465554, | |
| "train_runtime": 63035.3979, | |
| "train_samples_per_second": 0.272, | |
| "train_steps_per_second": 0.017 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1073, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.18849215432491e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |