| { | |
| "best_metric": 0.2426377534866333, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-400", | |
| "epoch": 0.3284746458632724, | |
| "eval_steps": 50, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.000821186614658181, | |
| "grad_norm": 2.0305542945861816, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7261, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.000821186614658181, | |
| "eval_loss": 1.465772271156311, | |
| "eval_runtime": 288.2262, | |
| "eval_samples_per_second": 7.116, | |
| "eval_steps_per_second": 1.78, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.001642373229316362, | |
| "grad_norm": 2.114654541015625, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7726, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.002463559843974543, | |
| "grad_norm": 2.2381458282470703, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8371, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.003284746458632724, | |
| "grad_norm": 1.9415267705917358, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7885, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0041059330732909054, | |
| "grad_norm": 1.484900712966919, | |
| "learning_rate": 5e-05, | |
| "loss": 0.7522, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.004927119687949086, | |
| "grad_norm": 1.259929895401001, | |
| "learning_rate": 6e-05, | |
| "loss": 0.7248, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.005748306302607267, | |
| "grad_norm": 0.9762012958526611, | |
| "learning_rate": 7e-05, | |
| "loss": 0.5922, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.006569492917265448, | |
| "grad_norm": 1.0204635858535767, | |
| "learning_rate": 8e-05, | |
| "loss": 0.5452, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00739067953192363, | |
| "grad_norm": 0.8520223498344421, | |
| "learning_rate": 9e-05, | |
| "loss": 0.4675, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.008211866146581811, | |
| "grad_norm": 1.470017433166504, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4164, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.009033052761239991, | |
| "grad_norm": 0.8858131766319275, | |
| "learning_rate": 9.99983777858264e-05, | |
| "loss": 0.3847, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.009854239375898173, | |
| "grad_norm": 1.0039604902267456, | |
| "learning_rate": 9.999351124856874e-05, | |
| "loss": 0.3836, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.010675425990556354, | |
| "grad_norm": 0.6317289471626282, | |
| "learning_rate": 9.998540070400966e-05, | |
| "loss": 0.3325, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.011496612605214535, | |
| "grad_norm": 0.680807888507843, | |
| "learning_rate": 9.997404667843075e-05, | |
| "loss": 0.3493, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.012317799219872716, | |
| "grad_norm": 0.6124119758605957, | |
| "learning_rate": 9.995944990857849e-05, | |
| "loss": 0.3204, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.013138985834530896, | |
| "grad_norm": 0.6300580501556396, | |
| "learning_rate": 9.994161134161634e-05, | |
| "loss": 0.3341, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.013960172449189078, | |
| "grad_norm": 0.608344554901123, | |
| "learning_rate": 9.992053213506334e-05, | |
| "loss": 0.3386, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.01478135906384726, | |
| "grad_norm": 0.5785073041915894, | |
| "learning_rate": 9.989621365671902e-05, | |
| "loss": 0.3035, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.01560254567850544, | |
| "grad_norm": 0.6188136339187622, | |
| "learning_rate": 9.986865748457457e-05, | |
| "loss": 0.3162, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.016423732293163622, | |
| "grad_norm": 0.5587109923362732, | |
| "learning_rate": 9.983786540671051e-05, | |
| "loss": 0.273, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.017244918907821802, | |
| "grad_norm": 0.5472216606140137, | |
| "learning_rate": 9.980383942118066e-05, | |
| "loss": 0.3215, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.018066105522479982, | |
| "grad_norm": 0.4810258150100708, | |
| "learning_rate": 9.976658173588244e-05, | |
| "loss": 0.307, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.018887292137138165, | |
| "grad_norm": 0.5424895882606506, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 0.3065, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.019708478751796345, | |
| "grad_norm": 0.5074129700660706, | |
| "learning_rate": 9.968238114591566e-05, | |
| "loss": 0.2774, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.020529665366454525, | |
| "grad_norm": 0.5599948167800903, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 0.2691, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02135085198111271, | |
| "grad_norm": 0.6742061972618103, | |
| "learning_rate": 9.95852854910781e-05, | |
| "loss": 0.2898, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.02217203859577089, | |
| "grad_norm": 0.5526089072227478, | |
| "learning_rate": 9.953190975913647e-05, | |
| "loss": 0.3071, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.02299322521042907, | |
| "grad_norm": 0.5390534996986389, | |
| "learning_rate": 9.947531997255256e-05, | |
| "loss": 0.2906, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.023814411825087253, | |
| "grad_norm": 0.504539430141449, | |
| "learning_rate": 9.941551980335652e-05, | |
| "loss": 0.2688, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.024635598439745433, | |
| "grad_norm": 0.4898300766944885, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 0.2845, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.025456785054403613, | |
| "grad_norm": 0.49974775314331055, | |
| "learning_rate": 9.928630404658255e-05, | |
| "loss": 0.2702, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.026277971669061793, | |
| "grad_norm": 0.5819835662841797, | |
| "learning_rate": 9.921689684362989e-05, | |
| "loss": 0.2699, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.027099158283719976, | |
| "grad_norm": 0.6171815395355225, | |
| "learning_rate": 9.914429602677162e-05, | |
| "loss": 0.3477, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.027920344898378156, | |
| "grad_norm": 0.6020647883415222, | |
| "learning_rate": 9.906850630697068e-05, | |
| "loss": 0.2901, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.028741531513036336, | |
| "grad_norm": 0.7458943724632263, | |
| "learning_rate": 9.898953260211338e-05, | |
| "loss": 0.2498, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02956271812769452, | |
| "grad_norm": 0.6588975787162781, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.2599, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0303839047423527, | |
| "grad_norm": 0.6839740872383118, | |
| "learning_rate": 9.882205394146361e-05, | |
| "loss": 0.337, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.03120509135701088, | |
| "grad_norm": 0.6753020882606506, | |
| "learning_rate": 9.87335598531214e-05, | |
| "loss": 0.3069, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.032026277971669063, | |
| "grad_norm": 0.8709951043128967, | |
| "learning_rate": 9.864190351391822e-05, | |
| "loss": 0.3396, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.032847464586327244, | |
| "grad_norm": 0.6819374561309814, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 0.2963, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.033668651200985424, | |
| "grad_norm": 0.856820821762085, | |
| "learning_rate": 9.844912807753104e-05, | |
| "loss": 0.289, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.034489837815643604, | |
| "grad_norm": 0.7196516990661621, | |
| "learning_rate": 9.834802148926882e-05, | |
| "loss": 0.2858, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.035311024430301784, | |
| "grad_norm": 0.7757695913314819, | |
| "learning_rate": 9.824377766717759e-05, | |
| "loss": 0.2882, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.036132211044959964, | |
| "grad_norm": 0.7751405835151672, | |
| "learning_rate": 9.813640337548954e-05, | |
| "loss": 0.3174, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.03695339765961815, | |
| "grad_norm": 0.8659068942070007, | |
| "learning_rate": 9.802590558156862e-05, | |
| "loss": 0.2818, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03777458427427633, | |
| "grad_norm": 1.336848258972168, | |
| "learning_rate": 9.791229145545831e-05, | |
| "loss": 0.3209, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.03859577088893451, | |
| "grad_norm": 0.9144354462623596, | |
| "learning_rate": 9.779556836941645e-05, | |
| "loss": 0.28, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.03941695750359269, | |
| "grad_norm": 0.8315229415893555, | |
| "learning_rate": 9.767574389743682e-05, | |
| "loss": 0.276, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.04023814411825087, | |
| "grad_norm": 0.967755913734436, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.3298, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.04105933073290905, | |
| "grad_norm": 1.0207139253616333, | |
| "learning_rate": 9.742682209735727e-05, | |
| "loss": 0.3058, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04105933073290905, | |
| "eval_loss": 0.44632911682128906, | |
| "eval_runtime": 291.2748, | |
| "eval_samples_per_second": 7.041, | |
| "eval_steps_per_second": 1.761, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04188051734756724, | |
| "grad_norm": 1.5864020586013794, | |
| "learning_rate": 9.729774092143627e-05, | |
| "loss": 0.4838, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.04270170396222542, | |
| "grad_norm": 1.2497475147247314, | |
| "learning_rate": 9.716559066288715e-05, | |
| "loss": 0.4411, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0435228905768836, | |
| "grad_norm": 0.9487363696098328, | |
| "learning_rate": 9.703037989675087e-05, | |
| "loss": 0.3657, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.04434407719154178, | |
| "grad_norm": 0.33357444405555725, | |
| "learning_rate": 9.689211739666023e-05, | |
| "loss": 0.2727, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.04516526380619996, | |
| "grad_norm": 0.4681488573551178, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 0.315, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.04598645042085814, | |
| "grad_norm": 0.4412896931171417, | |
| "learning_rate": 9.66064732786784e-05, | |
| "loss": 0.3051, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.04680763703551632, | |
| "grad_norm": 0.42806002497673035, | |
| "learning_rate": 9.645911019582467e-05, | |
| "loss": 0.296, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.047628823650174505, | |
| "grad_norm": 0.3795586824417114, | |
| "learning_rate": 9.630873244788883e-05, | |
| "loss": 0.3184, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.048450010264832685, | |
| "grad_norm": 0.34351328015327454, | |
| "learning_rate": 9.615534979266745e-05, | |
| "loss": 0.2832, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.049271196879490865, | |
| "grad_norm": 0.35215750336647034, | |
| "learning_rate": 9.599897218294122e-05, | |
| "loss": 0.3083, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.050092383494149045, | |
| "grad_norm": 0.3156924843788147, | |
| "learning_rate": 9.583960976582913e-05, | |
| "loss": 0.2577, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.050913570108807225, | |
| "grad_norm": 0.4058922231197357, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.3156, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.051734756723465405, | |
| "grad_norm": 0.3673064410686493, | |
| "learning_rate": 9.551197206565173e-05, | |
| "loss": 0.2932, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.052555943338123585, | |
| "grad_norm": 0.3226417601108551, | |
| "learning_rate": 9.534371804252728e-05, | |
| "loss": 0.2404, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.05337712995278177, | |
| "grad_norm": 0.3365015685558319, | |
| "learning_rate": 9.517252173051911e-05, | |
| "loss": 0.263, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.05419831656743995, | |
| "grad_norm": 0.38001781702041626, | |
| "learning_rate": 9.49983942383106e-05, | |
| "loss": 0.2973, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.05501950318209813, | |
| "grad_norm": 0.5203686356544495, | |
| "learning_rate": 9.482134686478519e-05, | |
| "loss": 0.2693, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.05584068979675631, | |
| "grad_norm": 0.32921281456947327, | |
| "learning_rate": 9.464139109829321e-05, | |
| "loss": 0.2742, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.05666187641141449, | |
| "grad_norm": 0.43800121545791626, | |
| "learning_rate": 9.445853861590647e-05, | |
| "loss": 0.2915, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.05748306302607267, | |
| "grad_norm": 0.32359007000923157, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 0.218, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05830424964073085, | |
| "grad_norm": 0.39324089884757996, | |
| "learning_rate": 9.408419115078471e-05, | |
| "loss": 0.2635, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.05912543625538904, | |
| "grad_norm": 0.45099887251853943, | |
| "learning_rate": 9.389272045892024e-05, | |
| "loss": 0.2402, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.05994662287004722, | |
| "grad_norm": 0.4052051603794098, | |
| "learning_rate": 9.36984016313259e-05, | |
| "loss": 0.2775, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.0607678094847054, | |
| "grad_norm": 0.41129499673843384, | |
| "learning_rate": 9.350124727707197e-05, | |
| "loss": 0.2564, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.06158899609936358, | |
| "grad_norm": 0.3723108470439911, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.2614, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.06241018271402176, | |
| "grad_norm": 0.39635589718818665, | |
| "learning_rate": 9.309848334400246e-05, | |
| "loss": 0.2825, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.06323136932867994, | |
| "grad_norm": 0.4818005859851837, | |
| "learning_rate": 9.289289989996133e-05, | |
| "loss": 0.2539, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.06405255594333813, | |
| "grad_norm": 0.4484040439128876, | |
| "learning_rate": 9.268453319711363e-05, | |
| "loss": 0.2857, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.0648737425579963, | |
| "grad_norm": 0.46060118079185486, | |
| "learning_rate": 9.247339675607605e-05, | |
| "loss": 0.2745, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.06569492917265449, | |
| "grad_norm": 0.3952708840370178, | |
| "learning_rate": 9.225950427718975e-05, | |
| "loss": 0.2954, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06651611578731266, | |
| "grad_norm": 0.4594082534313202, | |
| "learning_rate": 9.204286963963111e-05, | |
| "loss": 0.2647, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.06733730240197085, | |
| "grad_norm": 0.498677134513855, | |
| "learning_rate": 9.182350690051133e-05, | |
| "loss": 0.2741, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.06815848901662903, | |
| "grad_norm": 0.48790332674980164, | |
| "learning_rate": 9.160143029396422e-05, | |
| "loss": 0.217, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.06897967563128721, | |
| "grad_norm": 0.4500594437122345, | |
| "learning_rate": 9.13766542302225e-05, | |
| "loss": 0.2478, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.0698008622459454, | |
| "grad_norm": 0.477200984954834, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 0.276, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.07062204886060357, | |
| "grad_norm": 0.5421808362007141, | |
| "learning_rate": 9.091906224695935e-05, | |
| "loss": 0.3018, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.07144323547526175, | |
| "grad_norm": 0.473712295293808, | |
| "learning_rate": 9.068627601992598e-05, | |
| "loss": 0.2687, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.07226442208991993, | |
| "grad_norm": 0.5594393610954285, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.2515, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.07308560870457811, | |
| "grad_norm": 0.6386433839797974, | |
| "learning_rate": 9.021279861989885e-05, | |
| "loss": 0.2529, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.0739067953192363, | |
| "grad_norm": 0.5319457650184631, | |
| "learning_rate": 8.997213817017507e-05, | |
| "loss": 0.2646, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07472798193389447, | |
| "grad_norm": 0.595516562461853, | |
| "learning_rate": 8.972888398568772e-05, | |
| "loss": 0.2756, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.07554916854855266, | |
| "grad_norm": 0.6601677536964417, | |
| "learning_rate": 8.948305185085225e-05, | |
| "loss": 0.2805, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.07637035516321083, | |
| "grad_norm": 0.6162546873092651, | |
| "learning_rate": 8.92346577173636e-05, | |
| "loss": 0.2435, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.07719154177786902, | |
| "grad_norm": 0.6918389797210693, | |
| "learning_rate": 8.898371770316111e-05, | |
| "loss": 0.3159, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.07801272839252721, | |
| "grad_norm": 0.7181240916252136, | |
| "learning_rate": 8.873024809138272e-05, | |
| "loss": 0.2413, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.07883391500718538, | |
| "grad_norm": 1.1472549438476562, | |
| "learning_rate": 8.847426532930831e-05, | |
| "loss": 0.2949, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.07965510162184357, | |
| "grad_norm": 0.9359822273254395, | |
| "learning_rate": 8.821578602729242e-05, | |
| "loss": 0.316, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.08047628823650174, | |
| "grad_norm": 0.7444966435432434, | |
| "learning_rate": 8.795482695768658e-05, | |
| "loss": 0.3119, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.08129747485115993, | |
| "grad_norm": 1.2565944194793701, | |
| "learning_rate": 8.769140505375085e-05, | |
| "loss": 0.2877, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.0821186614658181, | |
| "grad_norm": 0.9129545092582703, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 0.3209, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0821186614658181, | |
| "eval_loss": 0.4635893702507019, | |
| "eval_runtime": 290.5387, | |
| "eval_samples_per_second": 7.059, | |
| "eval_steps_per_second": 1.766, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08293984808047629, | |
| "grad_norm": 1.5546329021453857, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 0.5163, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.08376103469513448, | |
| "grad_norm": 1.350873589515686, | |
| "learning_rate": 8.688653405904652e-05, | |
| "loss": 0.4791, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.08458222130979265, | |
| "grad_norm": 0.984677255153656, | |
| "learning_rate": 8.661343332988869e-05, | |
| "loss": 0.3377, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.08540340792445084, | |
| "grad_norm": 0.3391437530517578, | |
| "learning_rate": 8.633795680751116e-05, | |
| "loss": 0.2633, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.08622459453910901, | |
| "grad_norm": 0.2502461373806, | |
| "learning_rate": 8.606012236719073e-05, | |
| "loss": 0.2567, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0870457811537672, | |
| "grad_norm": 0.28781577944755554, | |
| "learning_rate": 8.577994803720606e-05, | |
| "loss": 0.2769, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.08786696776842537, | |
| "grad_norm": 0.30542901158332825, | |
| "learning_rate": 8.549745199766792e-05, | |
| "loss": 0.2965, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.08868815438308356, | |
| "grad_norm": 0.3351757526397705, | |
| "learning_rate": 8.521265257933948e-05, | |
| "loss": 0.2774, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.08950934099774174, | |
| "grad_norm": 0.2978907525539398, | |
| "learning_rate": 8.492556826244687e-05, | |
| "loss": 0.2744, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.09033052761239992, | |
| "grad_norm": 0.2558261752128601, | |
| "learning_rate": 8.463621767547998e-05, | |
| "loss": 0.2312, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.0911517142270581, | |
| "grad_norm": 0.30879148840904236, | |
| "learning_rate": 8.434461959398376e-05, | |
| "loss": 0.2845, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.09197290084171628, | |
| "grad_norm": 0.309334397315979, | |
| "learning_rate": 8.405079293933986e-05, | |
| "loss": 0.266, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.09279408745637446, | |
| "grad_norm": 0.3099028170108795, | |
| "learning_rate": 8.375475677753881e-05, | |
| "loss": 0.2692, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.09361527407103264, | |
| "grad_norm": 0.30877748131752014, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.284, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.09443646068569082, | |
| "grad_norm": 0.2932473123073578, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 0.268, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.09525764730034901, | |
| "grad_norm": 0.337277889251709, | |
| "learning_rate": 8.285358405218655e-05, | |
| "loss": 0.2448, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.09607883391500718, | |
| "grad_norm": 0.3811333477497101, | |
| "learning_rate": 8.25489033703452e-05, | |
| "loss": 0.293, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.09690002052966537, | |
| "grad_norm": 0.34305429458618164, | |
| "learning_rate": 8.224211063680853e-05, | |
| "loss": 0.2596, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.09772120714432354, | |
| "grad_norm": 0.3743656277656555, | |
| "learning_rate": 8.19332257589174e-05, | |
| "loss": 0.277, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.09854239375898173, | |
| "grad_norm": 0.2987310290336609, | |
| "learning_rate": 8.162226877976887e-05, | |
| "loss": 0.2484, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.0993635803736399, | |
| "grad_norm": 0.2911909222602844, | |
| "learning_rate": 8.130925987691569e-05, | |
| "loss": 0.2297, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.10018476698829809, | |
| "grad_norm": 0.29610538482666016, | |
| "learning_rate": 8.099421936105702e-05, | |
| "loss": 0.2224, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.10100595360295628, | |
| "grad_norm": 0.31938186287879944, | |
| "learning_rate": 8.067716767472045e-05, | |
| "loss": 0.2158, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.10182714021761445, | |
| "grad_norm": 0.3307250440120697, | |
| "learning_rate": 8.035812539093557e-05, | |
| "loss": 0.2443, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.10264832683227264, | |
| "grad_norm": 0.33900371193885803, | |
| "learning_rate": 8.003711321189895e-05, | |
| "loss": 0.2455, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.10346951344693081, | |
| "grad_norm": 0.33954647183418274, | |
| "learning_rate": 7.971415196763088e-05, | |
| "loss": 0.2413, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.104290700061589, | |
| "grad_norm": 0.3721504211425781, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.2724, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.10511188667624717, | |
| "grad_norm": 0.3788248300552368, | |
| "learning_rate": 7.906246623448183e-05, | |
| "loss": 0.2727, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.10593307329090536, | |
| "grad_norm": 0.3605785667896271, | |
| "learning_rate": 7.873378403255419e-05, | |
| "loss": 0.2781, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.10675425990556354, | |
| "grad_norm": 0.33415114879608154, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 0.2226, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.10757544652022172, | |
| "grad_norm": 0.43883877992630005, | |
| "learning_rate": 7.807084759519405e-05, | |
| "loss": 0.2474, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.1083966331348799, | |
| "grad_norm": 0.3597790598869324, | |
| "learning_rate": 7.773663637675694e-05, | |
| "loss": 0.2375, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.10921781974953808, | |
| "grad_norm": 0.41835817694664, | |
| "learning_rate": 7.740062536773352e-05, | |
| "loss": 0.2869, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.11003900636419627, | |
| "grad_norm": 0.40062186121940613, | |
| "learning_rate": 7.706283637139658e-05, | |
| "loss": 0.2451, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.11086019297885444, | |
| "grad_norm": 0.46534448862075806, | |
| "learning_rate": 7.672329130639005e-05, | |
| "loss": 0.2834, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.11168137959351263, | |
| "grad_norm": 0.4524616599082947, | |
| "learning_rate": 7.638201220530665e-05, | |
| "loss": 0.2579, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.11250256620817081, | |
| "grad_norm": 0.42796769738197327, | |
| "learning_rate": 7.603902121325813e-05, | |
| "loss": 0.2507, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.11332375282282899, | |
| "grad_norm": 0.6147335767745972, | |
| "learning_rate": 7.569434058643844e-05, | |
| "loss": 0.2547, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.11414493943748717, | |
| "grad_norm": 0.5195381045341492, | |
| "learning_rate": 7.534799269067953e-05, | |
| "loss": 0.2429, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.11496612605214535, | |
| "grad_norm": 0.549555242061615, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.2826, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11578731266680353, | |
| "grad_norm": 0.5704975128173828, | |
| "learning_rate": 7.465038509514688e-05, | |
| "loss": 0.2789, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.1166084992814617, | |
| "grad_norm": 0.6339702606201172, | |
| "learning_rate": 7.42991706621303e-05, | |
| "loss": 0.2289, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.11742968589611989, | |
| "grad_norm": 0.728171706199646, | |
| "learning_rate": 7.394637949075154e-05, | |
| "loss": 0.3104, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.11825087251077808, | |
| "grad_norm": 0.5731471180915833, | |
| "learning_rate": 7.35920344731241e-05, | |
| "loss": 0.2612, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.11907205912543625, | |
| "grad_norm": 0.7613599896430969, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 0.3024, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.11989324574009444, | |
| "grad_norm": 0.5790632963180542, | |
| "learning_rate": 7.287877497021978e-05, | |
| "loss": 0.2493, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.12071443235475261, | |
| "grad_norm": 0.746548593044281, | |
| "learning_rate": 7.251990676732984e-05, | |
| "loss": 0.2999, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.1215356189694108, | |
| "grad_norm": 0.7720208168029785, | |
| "learning_rate": 7.215957727996207e-05, | |
| "loss": 0.2836, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.12235680558406897, | |
| "grad_norm": 0.6100975275039673, | |
| "learning_rate": 7.179780988938051e-05, | |
| "loss": 0.2187, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.12317799219872716, | |
| "grad_norm": 0.7999358177185059, | |
| "learning_rate": 7.143462807015271e-05, | |
| "loss": 0.2709, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12317799219872716, | |
| "eval_loss": 0.38091349601745605, | |
| "eval_runtime": 291.0342, | |
| "eval_samples_per_second": 7.047, | |
| "eval_steps_per_second": 1.763, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12399917881338535, | |
| "grad_norm": 1.2095229625701904, | |
| "learning_rate": 7.107005538862646e-05, | |
| "loss": 0.4237, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.12482036542804352, | |
| "grad_norm": 1.172702670097351, | |
| "learning_rate": 7.07041155014006e-05, | |
| "loss": 0.351, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.1256415520427017, | |
| "grad_norm": 1.0826219320297241, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 0.3055, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.12646273865735988, | |
| "grad_norm": 0.3956534266471863, | |
| "learning_rate": 6.996822917828477e-05, | |
| "loss": 0.2814, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.12728392527201807, | |
| "grad_norm": 0.26742252707481384, | |
| "learning_rate": 6.959833049300377e-05, | |
| "loss": 0.2178, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.12810511188667625, | |
| "grad_norm": 0.2690823972225189, | |
| "learning_rate": 6.922716010014255e-05, | |
| "loss": 0.2396, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.12892629850133444, | |
| "grad_norm": 0.3007110357284546, | |
| "learning_rate": 6.885474208441603e-05, | |
| "loss": 0.2573, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.1297474851159926, | |
| "grad_norm": 0.27441394329071045, | |
| "learning_rate": 6.848110061149556e-05, | |
| "loss": 0.2418, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.1305686717306508, | |
| "grad_norm": 0.3037412166595459, | |
| "learning_rate": 6.810625992644085e-05, | |
| "loss": 0.2762, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.13138985834530897, | |
| "grad_norm": 0.31967610120773315, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 0.2674, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.13221104495996716, | |
| "grad_norm": 0.2789151668548584, | |
| "learning_rate": 6.735307828766515e-05, | |
| "loss": 0.254, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.13303223157462532, | |
| "grad_norm": 0.3466937839984894, | |
| "learning_rate": 6.697478620682137e-05, | |
| "loss": 0.2806, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.1338534181892835, | |
| "grad_norm": 0.31567007303237915, | |
| "learning_rate": 6.659539265642643e-05, | |
| "loss": 0.2611, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.1346746048039417, | |
| "grad_norm": 0.3090943992137909, | |
| "learning_rate": 6.621492225478414e-05, | |
| "loss": 0.2634, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.13549579141859988, | |
| "grad_norm": 0.30518144369125366, | |
| "learning_rate": 6.583339969007363e-05, | |
| "loss": 0.2314, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.13631697803325807, | |
| "grad_norm": 0.32459190487861633, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.2531, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.13713816464791623, | |
| "grad_norm": 0.3627863824367523, | |
| "learning_rate": 6.506729716392481e-05, | |
| "loss": 0.2555, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.13795935126257441, | |
| "grad_norm": 0.44058528542518616, | |
| "learning_rate": 6.468276691378155e-05, | |
| "loss": 0.2623, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.1387805378772326, | |
| "grad_norm": 0.43444401025772095, | |
| "learning_rate": 6.429728391993446e-05, | |
| "loss": 0.2796, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.1396017244918908, | |
| "grad_norm": 0.3341348469257355, | |
| "learning_rate": 6.391087319582264e-05, | |
| "loss": 0.2392, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.14042291110654898, | |
| "grad_norm": 0.339946448802948, | |
| "learning_rate": 6.35235598150842e-05, | |
| "loss": 0.2923, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.14124409772120713, | |
| "grad_norm": 0.308361291885376, | |
| "learning_rate": 6.313536890992935e-05, | |
| "loss": 0.2172, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.14206528433586532, | |
| "grad_norm": 0.36301669478416443, | |
| "learning_rate": 6.274632566950967e-05, | |
| "loss": 0.2436, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.1428864709505235, | |
| "grad_norm": 0.37350887060165405, | |
| "learning_rate": 6.235645533828349e-05, | |
| "loss": 0.2499, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.1437076575651817, | |
| "grad_norm": 0.33035343885421753, | |
| "learning_rate": 6.19657832143779e-05, | |
| "loss": 0.2233, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.14452884417983985, | |
| "grad_norm": 0.3460734486579895, | |
| "learning_rate": 6.157433464794716e-05, | |
| "loss": 0.238, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.14535003079449804, | |
| "grad_norm": 0.3977898359298706, | |
| "learning_rate": 6.118213503952779e-05, | |
| "loss": 0.2669, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.14617121740915623, | |
| "grad_norm": 0.442679226398468, | |
| "learning_rate": 6.078920983839031e-05, | |
| "loss": 0.2615, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.14699240402381442, | |
| "grad_norm": 0.3600601553916931, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 0.2502, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.1478135906384726, | |
| "grad_norm": 0.3798171877861023, | |
| "learning_rate": 6.0001284688802226e-05, | |
| "loss": 0.2574, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.14863477725313076, | |
| "grad_norm": 0.3883853256702423, | |
| "learning_rate": 5.960633586768543e-05, | |
| "loss": 0.2288, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.14945596386778895, | |
| "grad_norm": 0.48229384422302246, | |
| "learning_rate": 5.921076370520058e-05, | |
| "loss": 0.313, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.15027715048244714, | |
| "grad_norm": 0.43548011779785156, | |
| "learning_rate": 5.8814593869458455e-05, | |
| "loss": 0.2908, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.15109833709710532, | |
| "grad_norm": 0.4758455455303192, | |
| "learning_rate": 5.841785206735192e-05, | |
| "loss": 0.2527, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.1519195237117635, | |
| "grad_norm": 0.38382488489151, | |
| "learning_rate": 5.8020564042888015e-05, | |
| "loss": 0.2506, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.15274071032642167, | |
| "grad_norm": 0.4354500472545624, | |
| "learning_rate": 5.762275557551727e-05, | |
| "loss": 0.2753, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.15356189694107986, | |
| "grad_norm": 0.4720785319805145, | |
| "learning_rate": 5.7224452478461064e-05, | |
| "loss": 0.2799, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.15438308355573804, | |
| "grad_norm": 0.44493159651756287, | |
| "learning_rate": 5.682568059703659e-05, | |
| "loss": 0.2565, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.15520427017039623, | |
| "grad_norm": 0.430894136428833, | |
| "learning_rate": 5.642646580697973e-05, | |
| "loss": 0.2542, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.15602545678505442, | |
| "grad_norm": 0.4223925769329071, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 0.2331, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.15684664339971258, | |
| "grad_norm": 0.44736889004707336, | |
| "learning_rate": 5.562681114593028e-05, | |
| "loss": 0.2476, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.15766783001437076, | |
| "grad_norm": 0.5344662666320801, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.3117, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.15848901662902895, | |
| "grad_norm": 0.4708305895328522, | |
| "learning_rate": 5.482569604572576e-05, | |
| "loss": 0.213, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.15931020324368714, | |
| "grad_norm": 0.5267696976661682, | |
| "learning_rate": 5.442465579556793e-05, | |
| "loss": 0.2536, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.1601313898583453, | |
| "grad_norm": 0.5681753754615784, | |
| "learning_rate": 5.402332843583631e-05, | |
| "loss": 0.3035, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.16095257647300348, | |
| "grad_norm": 0.6239408850669861, | |
| "learning_rate": 5.3621740008088126e-05, | |
| "loss": 0.3042, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.16177376308766167, | |
| "grad_norm": 0.5667141079902649, | |
| "learning_rate": 5.321991657082097e-05, | |
| "loss": 0.2375, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.16259494970231986, | |
| "grad_norm": 0.6225094795227051, | |
| "learning_rate": 5.281788419778187e-05, | |
| "loss": 0.2812, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.16341613631697804, | |
| "grad_norm": 0.8053914904594421, | |
| "learning_rate": 5.2415668976275355e-05, | |
| "loss": 0.2652, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.1642373229316362, | |
| "grad_norm": 0.7740500569343567, | |
| "learning_rate": 5.201329700547076e-05, | |
| "loss": 0.3169, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1642373229316362, | |
| "eval_loss": 0.3369649350643158, | |
| "eval_runtime": 290.8356, | |
| "eval_samples_per_second": 7.052, | |
| "eval_steps_per_second": 1.764, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1650585095462944, | |
| "grad_norm": 0.6902033686637878, | |
| "learning_rate": 5.161079439470866e-05, | |
| "loss": 0.3221, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.16587969616095258, | |
| "grad_norm": 0.8175435066223145, | |
| "learning_rate": 5.1208187261806615e-05, | |
| "loss": 0.3839, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.16670088277561076, | |
| "grad_norm": 0.7495336532592773, | |
| "learning_rate": 5.080550173136457e-05, | |
| "loss": 0.3008, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.16752206939026895, | |
| "grad_norm": 0.533630907535553, | |
| "learning_rate": 5.0402763933069496e-05, | |
| "loss": 0.2779, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.1683432560049271, | |
| "grad_norm": 0.2528562843799591, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2441, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.1691644426195853, | |
| "grad_norm": 0.25667861104011536, | |
| "learning_rate": 4.9597236066930516e-05, | |
| "loss": 0.2325, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.16998562923424348, | |
| "grad_norm": 0.23454199731349945, | |
| "learning_rate": 4.919449826863544e-05, | |
| "loss": 0.2039, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.17080681584890167, | |
| "grad_norm": 0.2983744740486145, | |
| "learning_rate": 4.87918127381934e-05, | |
| "loss": 0.2786, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.17162800246355983, | |
| "grad_norm": 0.31790727376937866, | |
| "learning_rate": 4.8389205605291365e-05, | |
| "loss": 0.292, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.17244918907821802, | |
| "grad_norm": 0.28540918231010437, | |
| "learning_rate": 4.798670299452926e-05, | |
| "loss": 0.2687, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.1732703756928762, | |
| "grad_norm": 0.306003212928772, | |
| "learning_rate": 4.758433102372466e-05, | |
| "loss": 0.244, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.1740915623075344, | |
| "grad_norm": 0.3234851658344269, | |
| "learning_rate": 4.7182115802218126e-05, | |
| "loss": 0.2581, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.17491274892219258, | |
| "grad_norm": 0.30604276061058044, | |
| "learning_rate": 4.678008342917903e-05, | |
| "loss": 0.2566, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.17573393553685074, | |
| "grad_norm": 0.2953539192676544, | |
| "learning_rate": 4.6378259991911886e-05, | |
| "loss": 0.213, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.17655512215150893, | |
| "grad_norm": 0.3371827304363251, | |
| "learning_rate": 4.597667156416371e-05, | |
| "loss": 0.2913, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.1773763087661671, | |
| "grad_norm": 0.3420858085155487, | |
| "learning_rate": 4.5575344204432084e-05, | |
| "loss": 0.2476, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.1781974953808253, | |
| "grad_norm": 0.377582848072052, | |
| "learning_rate": 4.5174303954274244e-05, | |
| "loss": 0.296, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.1790186819954835, | |
| "grad_norm": 0.3134099841117859, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.2529, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.17983986861014165, | |
| "grad_norm": 0.35592809319496155, | |
| "learning_rate": 4.437318885406973e-05, | |
| "loss": 0.2735, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.18066105522479983, | |
| "grad_norm": 0.3550760746002197, | |
| "learning_rate": 4.397316598723385e-05, | |
| "loss": 0.2486, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.18148224183945802, | |
| "grad_norm": 0.388317346572876, | |
| "learning_rate": 4.3573534193020274e-05, | |
| "loss": 0.2972, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.1823034284541162, | |
| "grad_norm": 0.35952627658843994, | |
| "learning_rate": 4.317431940296343e-05, | |
| "loss": 0.2656, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.18312461506877437, | |
| "grad_norm": 0.33761507272720337, | |
| "learning_rate": 4.277554752153895e-05, | |
| "loss": 0.2287, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.18394580168343255, | |
| "grad_norm": 0.3288724422454834, | |
| "learning_rate": 4.237724442448273e-05, | |
| "loss": 0.2278, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.18476698829809074, | |
| "grad_norm": 0.33114880323410034, | |
| "learning_rate": 4.197943595711198e-05, | |
| "loss": 0.2446, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.18558817491274893, | |
| "grad_norm": 0.3646623492240906, | |
| "learning_rate": 4.1582147932648074e-05, | |
| "loss": 0.2331, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.1864093615274071, | |
| "grad_norm": 0.3394116759300232, | |
| "learning_rate": 4.118540613054156e-05, | |
| "loss": 0.2588, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.18723054814206527, | |
| "grad_norm": 0.3472703993320465, | |
| "learning_rate": 4.078923629479943e-05, | |
| "loss": 0.2748, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.18805173475672346, | |
| "grad_norm": 0.38195887207984924, | |
| "learning_rate": 4.039366413231458e-05, | |
| "loss": 0.2917, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.18887292137138165, | |
| "grad_norm": 0.38919854164123535, | |
| "learning_rate": 3.9998715311197785e-05, | |
| "loss": 0.2582, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.18969410798603983, | |
| "grad_norm": 0.43928471207618713, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 0.255, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.19051529460069802, | |
| "grad_norm": 0.40218767523765564, | |
| "learning_rate": 3.92107901616097e-05, | |
| "loss": 0.26, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.19133648121535618, | |
| "grad_norm": 0.4302116930484772, | |
| "learning_rate": 3.8817864960472236e-05, | |
| "loss": 0.2519, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.19215766783001437, | |
| "grad_norm": 0.5646070837974548, | |
| "learning_rate": 3.842566535205286e-05, | |
| "loss": 0.2924, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.19297885444467255, | |
| "grad_norm": 0.40696802735328674, | |
| "learning_rate": 3.803421678562213e-05, | |
| "loss": 0.2742, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.19380004105933074, | |
| "grad_norm": 0.41830089688301086, | |
| "learning_rate": 3.764354466171652e-05, | |
| "loss": 0.2315, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.1946212276739889, | |
| "grad_norm": 0.41818952560424805, | |
| "learning_rate": 3.725367433049033e-05, | |
| "loss": 0.2572, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.1954424142886471, | |
| "grad_norm": 0.4201239049434662, | |
| "learning_rate": 3.6864631090070655e-05, | |
| "loss": 0.2373, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.19626360090330527, | |
| "grad_norm": 0.4650125801563263, | |
| "learning_rate": 3.6476440184915815e-05, | |
| "loss": 0.254, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.19708478751796346, | |
| "grad_norm": 0.514137327671051, | |
| "learning_rate": 3.608912680417737e-05, | |
| "loss": 0.2697, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.19790597413262165, | |
| "grad_norm": 0.45526084303855896, | |
| "learning_rate": 3.570271608006555e-05, | |
| "loss": 0.2491, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.1987271607472798, | |
| "grad_norm": 0.5392455458641052, | |
| "learning_rate": 3.531723308621847e-05, | |
| "loss": 0.2705, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.199548347361938, | |
| "grad_norm": 0.5906926393508911, | |
| "learning_rate": 3.493270283607522e-05, | |
| "loss": 0.2359, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.20036953397659618, | |
| "grad_norm": 0.586770236492157, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.2819, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.20119072059125437, | |
| "grad_norm": 0.6232577562332153, | |
| "learning_rate": 3.4166600309926387e-05, | |
| "loss": 0.3246, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.20201190720591256, | |
| "grad_norm": 0.5822532176971436, | |
| "learning_rate": 3.3785077745215873e-05, | |
| "loss": 0.2804, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.20283309382057071, | |
| "grad_norm": 0.7183935046195984, | |
| "learning_rate": 3.340460734357359e-05, | |
| "loss": 0.2539, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.2036542804352289, | |
| "grad_norm": 0.9526737928390503, | |
| "learning_rate": 3.3025213793178646e-05, | |
| "loss": 0.3401, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.2044754670498871, | |
| "grad_norm": 0.578902542591095, | |
| "learning_rate": 3.264692171233485e-05, | |
| "loss": 0.2102, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.20529665366454528, | |
| "grad_norm": 0.7595329284667969, | |
| "learning_rate": 3.226975564787322e-05, | |
| "loss": 0.2384, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.20529665366454528, | |
| "eval_loss": 0.2933620810508728, | |
| "eval_runtime": 291.0703, | |
| "eval_samples_per_second": 7.046, | |
| "eval_steps_per_second": 1.762, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.20611784027920346, | |
| "grad_norm": 0.702744722366333, | |
| "learning_rate": 3.189374007355917e-05, | |
| "loss": 0.2989, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.20693902689386162, | |
| "grad_norm": 0.7356978058815002, | |
| "learning_rate": 3.151889938850445e-05, | |
| "loss": 0.2862, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.2077602135085198, | |
| "grad_norm": 0.8224905133247375, | |
| "learning_rate": 3.114525791558398e-05, | |
| "loss": 0.3193, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.208581400123178, | |
| "grad_norm": 0.9090282917022705, | |
| "learning_rate": 3.0772839899857464e-05, | |
| "loss": 0.2938, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.20940258673783618, | |
| "grad_norm": 0.8460712432861328, | |
| "learning_rate": 3.0401669506996256e-05, | |
| "loss": 0.2923, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.21022377335249434, | |
| "grad_norm": 0.8029319643974304, | |
| "learning_rate": 3.003177082171523e-05, | |
| "loss": 0.2789, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.21104495996715253, | |
| "grad_norm": 0.7324627041816711, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 0.2719, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.21186614658181072, | |
| "grad_norm": 0.5806877017021179, | |
| "learning_rate": 2.9295884498599414e-05, | |
| "loss": 0.2901, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.2126873331964689, | |
| "grad_norm": 0.3797169029712677, | |
| "learning_rate": 2.8929944611373554e-05, | |
| "loss": 0.2232, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.2135085198111271, | |
| "grad_norm": 0.28451111912727356, | |
| "learning_rate": 2.8565371929847284e-05, | |
| "loss": 0.2239, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.21432970642578525, | |
| "grad_norm": 0.2512100636959076, | |
| "learning_rate": 2.8202190110619493e-05, | |
| "loss": 0.2492, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.21515089304044344, | |
| "grad_norm": 0.24946503341197968, | |
| "learning_rate": 2.784042272003794e-05, | |
| "loss": 0.2228, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.21597207965510162, | |
| "grad_norm": 0.26216790080070496, | |
| "learning_rate": 2.7480093232670158e-05, | |
| "loss": 0.2501, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.2167932662697598, | |
| "grad_norm": 0.2938695549964905, | |
| "learning_rate": 2.712122502978024e-05, | |
| "loss": 0.2504, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.217614452884418, | |
| "grad_norm": 0.2901502847671509, | |
| "learning_rate": 2.6763841397811573e-05, | |
| "loss": 0.226, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.21843563949907616, | |
| "grad_norm": 0.3004022538661957, | |
| "learning_rate": 2.64079655268759e-05, | |
| "loss": 0.2446, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.21925682611373434, | |
| "grad_norm": 0.31173473596572876, | |
| "learning_rate": 2.605362050924848e-05, | |
| "loss": 0.2347, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.22007801272839253, | |
| "grad_norm": 0.31835272908210754, | |
| "learning_rate": 2.57008293378697e-05, | |
| "loss": 0.2462, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.22089919934305072, | |
| "grad_norm": 0.3530433475971222, | |
| "learning_rate": 2.534961490485313e-05, | |
| "loss": 0.2561, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.22172038595770888, | |
| "grad_norm": 0.31918856501579285, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.2428, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.22254157257236706, | |
| "grad_norm": 0.3554997444152832, | |
| "learning_rate": 2.4652007309320498e-05, | |
| "loss": 0.2632, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.22336275918702525, | |
| "grad_norm": 0.3261239230632782, | |
| "learning_rate": 2.430565941356157e-05, | |
| "loss": 0.2075, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.22418394580168344, | |
| "grad_norm": 0.36336517333984375, | |
| "learning_rate": 2.3960978786741877e-05, | |
| "loss": 0.2612, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.22500513241634162, | |
| "grad_norm": 0.38886624574661255, | |
| "learning_rate": 2.361798779469336e-05, | |
| "loss": 0.2701, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.22582631903099978, | |
| "grad_norm": 0.3287777900695801, | |
| "learning_rate": 2.3276708693609943e-05, | |
| "loss": 0.237, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.22664750564565797, | |
| "grad_norm": 0.3637790381908417, | |
| "learning_rate": 2.2937163628603435e-05, | |
| "loss": 0.2608, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.22746869226031616, | |
| "grad_norm": 0.37705734372138977, | |
| "learning_rate": 2.259937463226651e-05, | |
| "loss": 0.2654, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.22828987887497434, | |
| "grad_norm": 0.3406853675842285, | |
| "learning_rate": 2.2263363623243054e-05, | |
| "loss": 0.2395, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.22911106548963253, | |
| "grad_norm": 0.3554440438747406, | |
| "learning_rate": 2.192915240480596e-05, | |
| "loss": 0.2174, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.2299322521042907, | |
| "grad_norm": 0.40587833523750305, | |
| "learning_rate": 2.1596762663442218e-05, | |
| "loss": 0.2378, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.23075343871894888, | |
| "grad_norm": 0.3761412501335144, | |
| "learning_rate": 2.1266215967445824e-05, | |
| "loss": 0.2294, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.23157462533360706, | |
| "grad_norm": 0.4104063808917999, | |
| "learning_rate": 2.0937533765518187e-05, | |
| "loss": 0.2378, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.23239581194826525, | |
| "grad_norm": 0.41846057772636414, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.2895, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.2332169985629234, | |
| "grad_norm": 0.3801814019680023, | |
| "learning_rate": 2.0285848032369137e-05, | |
| "loss": 0.2124, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.2340381851775816, | |
| "grad_norm": 0.37622514367103577, | |
| "learning_rate": 1.996288678810105e-05, | |
| "loss": 0.2238, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.23485937179223978, | |
| "grad_norm": 0.41953378915786743, | |
| "learning_rate": 1.9641874609064443e-05, | |
| "loss": 0.2421, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.23568055840689797, | |
| "grad_norm": 0.4523312747478485, | |
| "learning_rate": 1.932283232527956e-05, | |
| "loss": 0.2796, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.23650174502155616, | |
| "grad_norm": 0.40526655316352844, | |
| "learning_rate": 1.9005780638942982e-05, | |
| "loss": 0.2674, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.23732293163621432, | |
| "grad_norm": 0.4408579170703888, | |
| "learning_rate": 1.8690740123084316e-05, | |
| "loss": 0.2404, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.2381441182508725, | |
| "grad_norm": 0.44801774621009827, | |
| "learning_rate": 1.837773122023114e-05, | |
| "loss": 0.2451, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2389653048655307, | |
| "grad_norm": 0.5268024802207947, | |
| "learning_rate": 1.8066774241082612e-05, | |
| "loss": 0.2969, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.23978649148018888, | |
| "grad_norm": 0.4801257252693176, | |
| "learning_rate": 1.7757889363191483e-05, | |
| "loss": 0.2497, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.24060767809484707, | |
| "grad_norm": 0.4917178750038147, | |
| "learning_rate": 1.745109662965481e-05, | |
| "loss": 0.2554, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.24142886470950523, | |
| "grad_norm": 0.5635613799095154, | |
| "learning_rate": 1.714641594781347e-05, | |
| "loss": 0.2485, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.2422500513241634, | |
| "grad_norm": 0.5717937350273132, | |
| "learning_rate": 1.684386708796025e-05, | |
| "loss": 0.2788, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.2430712379388216, | |
| "grad_norm": 0.6237139105796814, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 0.2782, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.2438924245534798, | |
| "grad_norm": 0.5460456609725952, | |
| "learning_rate": 1.62452432224612e-05, | |
| "loss": 0.2279, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.24471361116813795, | |
| "grad_norm": 0.6361614465713501, | |
| "learning_rate": 1.5949207060660138e-05, | |
| "loss": 0.2781, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.24553479778279613, | |
| "grad_norm": 0.7341341376304626, | |
| "learning_rate": 1.5655380406016235e-05, | |
| "loss": 0.2449, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.24635598439745432, | |
| "grad_norm": 0.8992396593093872, | |
| "learning_rate": 1.536378232452003e-05, | |
| "loss": 0.3146, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.24635598439745432, | |
| "eval_loss": 0.26980382204055786, | |
| "eval_runtime": 290.3497, | |
| "eval_samples_per_second": 7.064, | |
| "eval_steps_per_second": 1.767, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2471771710121125, | |
| "grad_norm": 0.5283815264701843, | |
| "learning_rate": 1.5074431737553157e-05, | |
| "loss": 0.2707, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.2479983576267707, | |
| "grad_norm": 0.643700897693634, | |
| "learning_rate": 1.4787347420660541e-05, | |
| "loss": 0.2902, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.24881954424142885, | |
| "grad_norm": 0.5973144769668579, | |
| "learning_rate": 1.4502548002332088e-05, | |
| "loss": 0.2595, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.24964073085608704, | |
| "grad_norm": 0.6217362880706787, | |
| "learning_rate": 1.422005196279395e-05, | |
| "loss": 0.2573, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.25046191747074525, | |
| "grad_norm": 0.5987569689750671, | |
| "learning_rate": 1.3939877632809278e-05, | |
| "loss": 0.263, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.2512831040854034, | |
| "grad_norm": 0.6554332971572876, | |
| "learning_rate": 1.3662043192488849e-05, | |
| "loss": 0.2771, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.2521042907000616, | |
| "grad_norm": 0.5994774699211121, | |
| "learning_rate": 1.338656667011134e-05, | |
| "loss": 0.2652, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.25292547731471976, | |
| "grad_norm": 0.5521206259727478, | |
| "learning_rate": 1.3113465940953495e-05, | |
| "loss": 0.2352, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.25374666392937795, | |
| "grad_norm": 0.48421710729599, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 0.2611, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.25456785054403613, | |
| "grad_norm": 0.5118895769119263, | |
| "learning_rate": 1.257446259144494e-05, | |
| "loss": 0.2242, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2553890371586943, | |
| "grad_norm": 0.4489380717277527, | |
| "learning_rate": 1.2308594946249163e-05, | |
| "loss": 0.2649, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.2562102237733525, | |
| "grad_norm": 0.40071746706962585, | |
| "learning_rate": 1.204517304231343e-05, | |
| "loss": 0.2177, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.2570314103880107, | |
| "grad_norm": 0.3643863797187805, | |
| "learning_rate": 1.178421397270758e-05, | |
| "loss": 0.272, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.2578525970026689, | |
| "grad_norm": 0.31282395124435425, | |
| "learning_rate": 1.1525734670691701e-05, | |
| "loss": 0.2439, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.258673783617327, | |
| "grad_norm": 0.2850761413574219, | |
| "learning_rate": 1.1269751908617277e-05, | |
| "loss": 0.2273, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.2594949702319852, | |
| "grad_norm": 0.26169025897979736, | |
| "learning_rate": 1.1016282296838887e-05, | |
| "loss": 0.2033, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.2603161568466434, | |
| "grad_norm": 0.25896409153938293, | |
| "learning_rate": 1.0765342282636416e-05, | |
| "loss": 0.2239, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.2611373434613016, | |
| "grad_norm": 0.2623019516468048, | |
| "learning_rate": 1.0516948149147754e-05, | |
| "loss": 0.2327, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.26195853007595976, | |
| "grad_norm": 0.2750350832939148, | |
| "learning_rate": 1.0271116014312293e-05, | |
| "loss": 0.2214, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.26277971669061795, | |
| "grad_norm": 0.2905277609825134, | |
| "learning_rate": 1.0027861829824952e-05, | |
| "loss": 0.2083, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.26360090330527614, | |
| "grad_norm": 0.2894083857536316, | |
| "learning_rate": 9.787201380101157e-06, | |
| "loss": 0.2385, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.2644220899199343, | |
| "grad_norm": 0.3011109530925751, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.228, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.2652432765345925, | |
| "grad_norm": 0.3072446584701538, | |
| "learning_rate": 9.313723980074018e-06, | |
| "loss": 0.2302, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.26606446314925064, | |
| "grad_norm": 0.3350919783115387, | |
| "learning_rate": 9.080937753040646e-06, | |
| "loss": 0.2577, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.2668856497639088, | |
| "grad_norm": 0.35746726393699646, | |
| "learning_rate": 8.850806705317183e-06, | |
| "loss": 0.2655, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.267706836378567, | |
| "grad_norm": 0.3702896535396576, | |
| "learning_rate": 8.623345769777514e-06, | |
| "loss": 0.2523, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.2685280229932252, | |
| "grad_norm": 0.3639756739139557, | |
| "learning_rate": 8.398569706035792e-06, | |
| "loss": 0.2651, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.2693492096078834, | |
| "grad_norm": 0.39260295033454895, | |
| "learning_rate": 8.176493099488663e-06, | |
| "loss": 0.2864, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.2701703962225416, | |
| "grad_norm": 0.43910595774650574, | |
| "learning_rate": 7.957130360368898e-06, | |
| "loss": 0.2596, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.27099158283719976, | |
| "grad_norm": 0.3868727385997772, | |
| "learning_rate": 7.740495722810271e-06, | |
| "loss": 0.2309, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.27181276945185795, | |
| "grad_norm": 0.4218393564224243, | |
| "learning_rate": 7.526603243923957e-06, | |
| "loss": 0.2609, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.27263395606651614, | |
| "grad_norm": 0.4130231738090515, | |
| "learning_rate": 7.315466802886401e-06, | |
| "loss": 0.2362, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.2734551426811743, | |
| "grad_norm": 0.48627179861068726, | |
| "learning_rate": 7.107100100038671e-06, | |
| "loss": 0.2794, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.27427632929583246, | |
| "grad_norm": 0.4429982900619507, | |
| "learning_rate": 6.901516655997536e-06, | |
| "loss": 0.2077, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.27509751591049064, | |
| "grad_norm": 0.4805629551410675, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 0.1976, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.27591870252514883, | |
| "grad_norm": 0.5085516571998596, | |
| "learning_rate": 6.498752722928042e-06, | |
| "loss": 0.2826, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.276739889139807, | |
| "grad_norm": 0.5215482115745544, | |
| "learning_rate": 6.301598368674105e-06, | |
| "loss": 0.248, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.2775610757544652, | |
| "grad_norm": 0.622835636138916, | |
| "learning_rate": 6.107279541079769e-06, | |
| "loss": 0.2981, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.2783822623691234, | |
| "grad_norm": 0.5559107661247253, | |
| "learning_rate": 5.915808849215304e-06, | |
| "loss": 0.2702, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.2792034489837816, | |
| "grad_norm": 0.6636733412742615, | |
| "learning_rate": 5.727198717339511e-06, | |
| "loss": 0.2506, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.28002463559843976, | |
| "grad_norm": 0.6330378651618958, | |
| "learning_rate": 5.54146138409355e-06, | |
| "loss": 0.2406, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.28084582221309795, | |
| "grad_norm": 0.6448156833648682, | |
| "learning_rate": 5.358608901706802e-06, | |
| "loss": 0.249, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.2816670088277561, | |
| "grad_norm": 0.6033381819725037, | |
| "learning_rate": 5.178653135214812e-06, | |
| "loss": 0.2416, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.28248819544241427, | |
| "grad_norm": 0.7269361615180969, | |
| "learning_rate": 5.001605761689398e-06, | |
| "loss": 0.304, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.28330938205707246, | |
| "grad_norm": 0.6423282027244568, | |
| "learning_rate": 4.827478269480895e-06, | |
| "loss": 0.237, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.28413056867173064, | |
| "grad_norm": 0.7654911875724792, | |
| "learning_rate": 4.65628195747273e-06, | |
| "loss": 0.2883, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.28495175528638883, | |
| "grad_norm": 0.820427656173706, | |
| "learning_rate": 4.488027934348271e-06, | |
| "loss": 0.2854, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.285772941901047, | |
| "grad_norm": 0.8790760636329651, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 0.2876, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.2865941285157052, | |
| "grad_norm": 0.9669967889785767, | |
| "learning_rate": 4.16039023417088e-06, | |
| "loss": 0.3004, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.2874153151303634, | |
| "grad_norm": 1.0310205221176147, | |
| "learning_rate": 4.001027817058789e-06, | |
| "loss": 0.3123, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2874153151303634, | |
| "eval_loss": 0.24323655664920807, | |
| "eval_runtime": 290.8902, | |
| "eval_samples_per_second": 7.051, | |
| "eval_steps_per_second": 1.764, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2882365017450216, | |
| "grad_norm": 0.24831348657608032, | |
| "learning_rate": 3.844650207332562e-06, | |
| "loss": 0.2173, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.2890576883596797, | |
| "grad_norm": 0.29482340812683105, | |
| "learning_rate": 3.691267552111183e-06, | |
| "loss": 0.243, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.2898788749743379, | |
| "grad_norm": 0.35965675115585327, | |
| "learning_rate": 3.54088980417534e-06, | |
| "loss": 0.2659, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.2907000615889961, | |
| "grad_norm": 0.3628787100315094, | |
| "learning_rate": 3.393526721321616e-06, | |
| "loss": 0.2395, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.29152124820365427, | |
| "grad_norm": 0.36598777770996094, | |
| "learning_rate": 3.249187865729264e-06, | |
| "loss": 0.2255, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.29234243481831246, | |
| "grad_norm": 0.4101616144180298, | |
| "learning_rate": 3.1078826033397843e-06, | |
| "loss": 0.2461, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.29316362143297064, | |
| "grad_norm": 0.3846442699432373, | |
| "learning_rate": 2.9696201032491434e-06, | |
| "loss": 0.2377, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.29398480804762883, | |
| "grad_norm": 0.3726769983768463, | |
| "learning_rate": 2.8344093371128424e-06, | |
| "loss": 0.241, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.294805994662287, | |
| "grad_norm": 0.4052610695362091, | |
| "learning_rate": 2.70225907856374e-06, | |
| "loss": 0.2495, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.2956271812769452, | |
| "grad_norm": 0.4084397554397583, | |
| "learning_rate": 2.573177902642726e-06, | |
| "loss": 0.2458, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2964483678916034, | |
| "grad_norm": 0.42254287004470825, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 0.2568, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.2972695545062615, | |
| "grad_norm": 0.3660500943660736, | |
| "learning_rate": 2.324256102563188e-06, | |
| "loss": 0.2142, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.2980907411209197, | |
| "grad_norm": 0.3869646191596985, | |
| "learning_rate": 2.204431630583548e-06, | |
| "loss": 0.2636, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.2989119277355779, | |
| "grad_norm": 0.34974104166030884, | |
| "learning_rate": 2.087708544541689e-06, | |
| "loss": 0.1984, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.2997331143502361, | |
| "grad_norm": 0.39927589893341064, | |
| "learning_rate": 1.974094418431388e-06, | |
| "loss": 0.275, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.30055430096489427, | |
| "grad_norm": 0.37916234135627747, | |
| "learning_rate": 1.8635966245104664e-06, | |
| "loss": 0.2321, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.30137548757955246, | |
| "grad_norm": 0.3782179355621338, | |
| "learning_rate": 1.7562223328224325e-06, | |
| "loss": 0.2227, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.30219667419421065, | |
| "grad_norm": 0.37233802676200867, | |
| "learning_rate": 1.6519785107311891e-06, | |
| "loss": 0.2032, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.30301786080886883, | |
| "grad_norm": 0.3644430935382843, | |
| "learning_rate": 1.5508719224689717e-06, | |
| "loss": 0.2159, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.303839047423527, | |
| "grad_norm": 0.3714001476764679, | |
| "learning_rate": 1.4529091286973995e-06, | |
| "loss": 0.2397, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.30466023403818515, | |
| "grad_norm": 0.36238205432891846, | |
| "learning_rate": 1.358096486081778e-06, | |
| "loss": 0.2204, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.30548142065284334, | |
| "grad_norm": 0.36938953399658203, | |
| "learning_rate": 1.2664401468786114e-06, | |
| "loss": 0.2478, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.3063026072675015, | |
| "grad_norm": 0.39639028906822205, | |
| "learning_rate": 1.1779460585363944e-06, | |
| "loss": 0.263, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.3071237938821597, | |
| "grad_norm": 0.3775864541530609, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 0.2503, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.3079449804968179, | |
| "grad_norm": 0.36484062671661377, | |
| "learning_rate": 1.0104673978866164e-06, | |
| "loss": 0.2593, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3087661671114761, | |
| "grad_norm": 0.3943125307559967, | |
| "learning_rate": 9.314936930293283e-07, | |
| "loss": 0.2753, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.3095873537261343, | |
| "grad_norm": 0.41774439811706543, | |
| "learning_rate": 8.557039732283944e-07, | |
| "loss": 0.2385, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.31040854034079246, | |
| "grad_norm": 0.3497304916381836, | |
| "learning_rate": 7.83103156370113e-07, | |
| "loss": 0.2494, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.31122972695545065, | |
| "grad_norm": 0.35694608092308044, | |
| "learning_rate": 7.136959534174592e-07, | |
| "loss": 0.2479, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.31205091357010883, | |
| "grad_norm": 0.3703640103340149, | |
| "learning_rate": 6.474868681043578e-07, | |
| "loss": 0.2361, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.31287210018476697, | |
| "grad_norm": 0.3687499463558197, | |
| "learning_rate": 5.844801966434832e-07, | |
| "loss": 0.2544, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.31369328679942515, | |
| "grad_norm": 0.3702276945114136, | |
| "learning_rate": 5.246800274474439e-07, | |
| "loss": 0.2288, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.31451447341408334, | |
| "grad_norm": 0.35617271065711975, | |
| "learning_rate": 4.680902408635335e-07, | |
| "loss": 0.2311, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.3153356600287415, | |
| "grad_norm": 0.3268205225467682, | |
| "learning_rate": 4.1471450892189846e-07, | |
| "loss": 0.202, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.3161568466433997, | |
| "grad_norm": 0.3945155441761017, | |
| "learning_rate": 3.6455629509730136e-07, | |
| "loss": 0.2682, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.3169780332580579, | |
| "grad_norm": 0.40699857473373413, | |
| "learning_rate": 3.1761885408435054e-07, | |
| "loss": 0.2526, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.3177992198727161, | |
| "grad_norm": 0.4202795624732971, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 0.2616, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.3186204064873743, | |
| "grad_norm": 0.4231754243373871, | |
| "learning_rate": 2.334182641175686e-07, | |
| "loss": 0.2372, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.31944159310203246, | |
| "grad_norm": 0.44029873609542847, | |
| "learning_rate": 1.9616057881935436e-07, | |
| "loss": 0.2181, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.3202627797166906, | |
| "grad_norm": 0.4859876334667206, | |
| "learning_rate": 1.6213459328950352e-07, | |
| "loss": 0.2382, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.3210839663313488, | |
| "grad_norm": 0.49845609068870544, | |
| "learning_rate": 1.3134251542544774e-07, | |
| "loss": 0.2548, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.32190515294600697, | |
| "grad_norm": 0.5106012225151062, | |
| "learning_rate": 1.0378634328099269e-07, | |
| "loss": 0.2547, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.32272633956066515, | |
| "grad_norm": 0.48110705614089966, | |
| "learning_rate": 7.946786493666647e-08, | |
| "loss": 0.2322, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.32354752617532334, | |
| "grad_norm": 0.5826103687286377, | |
| "learning_rate": 5.838865838366792e-08, | |
| "loss": 0.3049, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.32436871278998153, | |
| "grad_norm": 0.6699414253234863, | |
| "learning_rate": 4.055009142152067e-08, | |
| "loss": 0.264, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.3251898994046397, | |
| "grad_norm": 0.6277973651885986, | |
| "learning_rate": 2.595332156925534e-08, | |
| "loss": 0.2696, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.3260110860192979, | |
| "grad_norm": 0.6204254627227783, | |
| "learning_rate": 1.4599295990352924e-08, | |
| "loss": 0.3054, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.3268322726339561, | |
| "grad_norm": 0.6760455965995789, | |
| "learning_rate": 6.488751431266149e-09, | |
| "loss": 0.2705, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.3276534592486142, | |
| "grad_norm": 0.8175690770149231, | |
| "learning_rate": 1.622214173602199e-09, | |
| "loss": 0.2644, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.3284746458632724, | |
| "grad_norm": 1.2129104137420654, | |
| "learning_rate": 0.0, | |
| "loss": 0.329, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3284746458632724, | |
| "eval_loss": 0.2426377534866333, | |
| "eval_runtime": 290.7324, | |
| "eval_samples_per_second": 7.055, | |
| "eval_steps_per_second": 1.765, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.947750106397409e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |