| { |
| "best_global_step": 800, |
| "best_metric": 0.24719083309173584, |
| "best_model_checkpoint": "./checkpoints/qwen253-lora-leduc_random/checkpoint-800", |
| "epoch": 1.0, |
| "eval_steps": 200, |
| "global_step": 1401, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.007140307033202428, |
| "grad_norm": 12.807258605957031, |
| "learning_rate": 8.999999999999999e-06, |
| "loss": 0.7525, |
| "mean_token_accuracy": 0.8291666507720947, |
| "num_tokens": 77176.0, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.014280614066404856, |
| "grad_norm": 1.4289218187332153, |
| "learning_rate": 1.8999999999999998e-05, |
| "loss": 0.3601, |
| "mean_token_accuracy": 0.8393361955881119, |
| "num_tokens": 151284.0, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.021420921099607283, |
| "grad_norm": 0.5016309022903442, |
| "learning_rate": 2.9e-05, |
| "loss": 0.2532, |
| "mean_token_accuracy": 0.8274722695350647, |
| "num_tokens": 226991.0, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.028561228132809712, |
| "grad_norm": 0.36768677830696106, |
| "learning_rate": 3.4999259502972436e-05, |
| "loss": 0.2514, |
| "mean_token_accuracy": 0.8576099902391434, |
| "num_tokens": 300775.0, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.03570153516601214, |
| "grad_norm": 0.3122566342353821, |
| "learning_rate": 3.4990929631086414e-05, |
| "loss": 0.2575, |
| "mean_token_accuracy": 0.8320898920297622, |
| "num_tokens": 376466.0, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.042841842199214566, |
| "grad_norm": 0.2976522445678711, |
| "learning_rate": 3.4973348686443995e-05, |
| "loss": 0.2506, |
| "mean_token_accuracy": 0.8405988067388535, |
| "num_tokens": 452975.0, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.04998214923241699, |
| "grad_norm": 0.3062177896499634, |
| "learning_rate": 3.4946525967727475e-05, |
| "loss": 0.2555, |
| "mean_token_accuracy": 0.842251580953598, |
| "num_tokens": 528108.0, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.057122456265619424, |
| "grad_norm": 0.22260648012161255, |
| "learning_rate": 3.491047566165693e-05, |
| "loss": 0.2475, |
| "mean_token_accuracy": 0.8489305555820466, |
| "num_tokens": 604469.0, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.06426276329882184, |
| "grad_norm": 0.3016267418861389, |
| "learning_rate": 3.486521683548673e-05, |
| "loss": 0.261, |
| "mean_token_accuracy": 0.8330651551485062, |
| "num_tokens": 680099.0, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.07140307033202428, |
| "grad_norm": 0.13674069941043854, |
| "learning_rate": 3.481077342692073e-05, |
| "loss": 0.2484, |
| "mean_token_accuracy": 0.8475112229585647, |
| "num_tokens": 756304.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.07854337736522671, |
| "grad_norm": 0.2582288384437561, |
| "learning_rate": 3.474717423145147e-05, |
| "loss": 0.2448, |
| "mean_token_accuracy": 0.8493792057037354, |
| "num_tokens": 835073.0, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.08568368439842913, |
| "grad_norm": 0.2136859893798828, |
| "learning_rate": 3.467445288713004e-05, |
| "loss": 0.2484, |
| "mean_token_accuracy": 0.8472158521413803, |
| "num_tokens": 910851.0, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.09282399143163156, |
| "grad_norm": 0.3934606909751892, |
| "learning_rate": 3.459264785677468e-05, |
| "loss": 0.2478, |
| "mean_token_accuracy": 0.8415243119001389, |
| "num_tokens": 986863.0, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.09996429846483398, |
| "grad_norm": 0.12868095934391022, |
| "learning_rate": 3.450180240762754e-05, |
| "loss": 0.2503, |
| "mean_token_accuracy": 0.844350454211235, |
| "num_tokens": 1063159.0, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.10710460549803641, |
| "grad_norm": 0.1570904701948166, |
| "learning_rate": 3.440196458847032e-05, |
| "loss": 0.2531, |
| "mean_token_accuracy": 0.8415277153253555, |
| "num_tokens": 1138174.0, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.11424491253123885, |
| "grad_norm": 0.1334053874015808, |
| "learning_rate": 3.429318720421095e-05, |
| "loss": 0.2506, |
| "mean_token_accuracy": 0.8386974811553956, |
| "num_tokens": 1213391.0, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.12138521956444127, |
| "grad_norm": 0.10832671821117401, |
| "learning_rate": 3.417552778795473e-05, |
| "loss": 0.2514, |
| "mean_token_accuracy": 0.8428560703992843, |
| "num_tokens": 1288233.0, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.12852552659764369, |
| "grad_norm": 0.12491905689239502, |
| "learning_rate": 3.404904857057463e-05, |
| "loss": 0.2481, |
| "mean_token_accuracy": 0.8515459626913071, |
| "num_tokens": 1364933.0, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.13566583363084614, |
| "grad_norm": 0.24591131508350372, |
| "learning_rate": 3.3913816447797016e-05, |
| "loss": 0.2554, |
| "mean_token_accuracy": 0.8417961895465851, |
| "num_tokens": 1440110.0, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.14280614066404856, |
| "grad_norm": 0.18424130976200104, |
| "learning_rate": 3.376990294481997e-05, |
| "loss": 0.2489, |
| "mean_token_accuracy": 0.8421024352312088, |
| "num_tokens": 1513857.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.14280614066404856, |
| "eval_loss": 0.25120916962623596, |
| "eval_num_tokens": 1513857.0, |
| "eval_runtime": 23.3172, |
| "eval_samples_per_second": 9.735, |
| "eval_steps_per_second": 9.735, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.149946447697251, |
| "grad_norm": 0.20695431530475616, |
| "learning_rate": 3.3617384178483175e-05, |
| "loss": 0.2488, |
| "mean_token_accuracy": 0.8430075138807297, |
| "num_tokens": 1590575.0, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.15708675473045342, |
| "grad_norm": 0.17747662961483002, |
| "learning_rate": 3.3456340817009197e-05, |
| "loss": 0.2509, |
| "mean_token_accuracy": 0.8476816356182099, |
| "num_tokens": 1664982.0, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.16422706176365584, |
| "grad_norm": 0.212553471326828, |
| "learning_rate": 3.328685803733747e-05, |
| "loss": 0.2491, |
| "mean_token_accuracy": 0.8427877694368362, |
| "num_tokens": 1740131.0, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.17136736879685827, |
| "grad_norm": 0.2259141355752945, |
| "learning_rate": 3.310902548007365e-05, |
| "loss": 0.2466, |
| "mean_token_accuracy": 0.8501875162124634, |
| "num_tokens": 1816043.0, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.1785076758300607, |
| "grad_norm": 0.17231041193008423, |
| "learning_rate": 3.292293720207809e-05, |
| "loss": 0.244, |
| "mean_token_accuracy": 0.8425901532173157, |
| "num_tokens": 1891899.0, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.18564798286326312, |
| "grad_norm": 0.13625052571296692, |
| "learning_rate": 3.272869162671851e-05, |
| "loss": 0.243, |
| "mean_token_accuracy": 0.8644366472959518, |
| "num_tokens": 1966931.0, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.19278828989646554, |
| "grad_norm": 0.23541058599948883, |
| "learning_rate": 3.252639149181322e-05, |
| "loss": 0.2425, |
| "mean_token_accuracy": 0.8397614926099777, |
| "num_tokens": 2043639.0, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.19992859692966797, |
| "grad_norm": 0.09341134876012802, |
| "learning_rate": 3.2316143795292324e-05, |
| "loss": 0.25, |
| "mean_token_accuracy": 0.8420886158943176, |
| "num_tokens": 2116914.0, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.2070689039628704, |
| "grad_norm": 0.19534029066562653, |
| "learning_rate": 3.209805973860588e-05, |
| "loss": 0.2533, |
| "mean_token_accuracy": 0.8413200944662094, |
| "num_tokens": 2191711.0, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.21420921099607282, |
| "grad_norm": 0.18656742572784424, |
| "learning_rate": 3.187225466790859e-05, |
| "loss": 0.2452, |
| "mean_token_accuracy": 0.8507013201713562, |
| "num_tokens": 2269204.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.22134951802927527, |
| "grad_norm": 0.29077425599098206, |
| "learning_rate": 3.1638848013052504e-05, |
| "loss": 0.2505, |
| "mean_token_accuracy": 0.8318937063217163, |
| "num_tokens": 2345922.0, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.2284898250624777, |
| "grad_norm": 0.15692247450351715, |
| "learning_rate": 3.1397963224419684e-05, |
| "loss": 0.2502, |
| "mean_token_accuracy": 0.8415091365575791, |
| "num_tokens": 2421757.0, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.23563013209568012, |
| "grad_norm": 0.24284499883651733, |
| "learning_rate": 3.114972770762853e-05, |
| "loss": 0.2456, |
| "mean_token_accuracy": 0.8416744440793991, |
| "num_tokens": 2497767.0, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.24277043912888255, |
| "grad_norm": 0.307827353477478, |
| "learning_rate": 3.089427275614796e-05, |
| "loss": 0.241, |
| "mean_token_accuracy": 0.8526032984256744, |
| "num_tokens": 2574923.0, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.24991074616208497, |
| "grad_norm": 0.181515172123909, |
| "learning_rate": 3.063173348185554e-05, |
| "loss": 0.254, |
| "mean_token_accuracy": 0.8498575508594512, |
| "num_tokens": 2650402.0, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.25705105319528737, |
| "grad_norm": 0.16170750558376312, |
| "learning_rate": 3.036224874357575e-05, |
| "loss": 0.2567, |
| "mean_token_accuracy": 0.8391913861036301, |
| "num_tokens": 2725845.0, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.2641913602284898, |
| "grad_norm": 0.13518409430980682, |
| "learning_rate": 3.0085961073636656e-05, |
| "loss": 0.248, |
| "mean_token_accuracy": 0.8520790785551071, |
| "num_tokens": 2800019.0, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.2713316672616923, |
| "grad_norm": 0.1753694862127304, |
| "learning_rate": 2.9803016602483608e-05, |
| "loss": 0.2466, |
| "mean_token_accuracy": 0.8495361566543579, |
| "num_tokens": 2875654.0, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.2784719742948947, |
| "grad_norm": 0.12919923663139343, |
| "learning_rate": 2.9513564981389787e-05, |
| "loss": 0.2449, |
| "mean_token_accuracy": 0.840579941868782, |
| "num_tokens": 2950462.0, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.28561228132809713, |
| "grad_norm": 0.14216065406799316, |
| "learning_rate": 2.921775930330468e-05, |
| "loss": 0.2464, |
| "mean_token_accuracy": 0.8327349931001663, |
| "num_tokens": 3025459.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.28561228132809713, |
| "eval_loss": 0.25034797191619873, |
| "eval_num_tokens": 3025459.0, |
| "eval_runtime": 23.1643, |
| "eval_samples_per_second": 9.8, |
| "eval_steps_per_second": 9.8, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.2927525883612995, |
| "grad_norm": 0.1596086025238037, |
| "learning_rate": 2.8915756021882105e-05, |
| "loss": 0.2379, |
| "mean_token_accuracy": 0.8618583887815475, |
| "num_tokens": 3101469.0, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.299892895394502, |
| "grad_norm": 0.15996187925338745, |
| "learning_rate": 2.8607714868730857e-05, |
| "loss": 0.2452, |
| "mean_token_accuracy": 0.8441025376319885, |
| "num_tokens": 3177343.0, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.3070332024277044, |
| "grad_norm": 0.17053034901618958, |
| "learning_rate": 2.8293798768931527e-05, |
| "loss": 0.2493, |
| "mean_token_accuracy": 0.8420908480882645, |
| "num_tokens": 3252653.0, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.31417350946090683, |
| "grad_norm": 0.30058109760284424, |
| "learning_rate": 2.797417375486431e-05, |
| "loss": 0.2376, |
| "mean_token_accuracy": 0.8616828382015228, |
| "num_tokens": 3327836.0, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.32131381649410923, |
| "grad_norm": 0.14107173681259155, |
| "learning_rate": 2.7649008878393374e-05, |
| "loss": 0.2502, |
| "mean_token_accuracy": 0.8440912365913391, |
| "num_tokens": 3402120.0, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.3284541235273117, |
| "grad_norm": 0.17678247392177582, |
| "learning_rate": 2.7318476121454117e-05, |
| "loss": 0.2483, |
| "mean_token_accuracy": 0.8335580289363861, |
| "num_tokens": 3477148.0, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.3355944305605141, |
| "grad_norm": 0.10473231971263885, |
| "learning_rate": 2.6982750305090776e-05, |
| "loss": 0.2447, |
| "mean_token_accuracy": 0.837072229385376, |
| "num_tokens": 3552114.0, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.34273473759371653, |
| "grad_norm": 0.16203787922859192, |
| "learning_rate": 2.6642008996992387e-05, |
| "loss": 0.2392, |
| "mean_token_accuracy": 0.8442226320505142, |
| "num_tokens": 3626876.0, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.349875044626919, |
| "grad_norm": 0.2136528640985489, |
| "learning_rate": 2.629643241757596e-05, |
| "loss": 0.2461, |
| "mean_token_accuracy": 0.8237904250621796, |
| "num_tokens": 3701915.0, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.3570153516601214, |
| "grad_norm": 0.11235684901475906, |
| "learning_rate": 2.5946203344666726e-05, |
| "loss": 0.2498, |
| "mean_token_accuracy": 0.8354511678218841, |
| "num_tokens": 3776580.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.36415565869332384, |
| "grad_norm": 0.12197854369878769, |
| "learning_rate": 2.5591507016825652e-05, |
| "loss": 0.2433, |
| "mean_token_accuracy": 0.8561921536922454, |
| "num_tokens": 3852836.0, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.37129596572652623, |
| "grad_norm": 0.15267807245254517, |
| "learning_rate": 2.523253103537551e-05, |
| "loss": 0.2516, |
| "mean_token_accuracy": 0.839010488986969, |
| "num_tokens": 3928224.0, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.3784362727597287, |
| "grad_norm": 0.14171989262104034, |
| "learning_rate": 2.4869465265177244e-05, |
| "loss": 0.2455, |
| "mean_token_accuracy": 0.8406431436538696, |
| "num_tokens": 4005075.0, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.3855765797929311, |
| "grad_norm": 0.12645554542541504, |
| "learning_rate": 2.4502501734209142e-05, |
| "loss": 0.2442, |
| "mean_token_accuracy": 0.8290857791900634, |
| "num_tokens": 4079639.0, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.39271688682613354, |
| "grad_norm": 0.17190751433372498, |
| "learning_rate": 2.4131834532001937e-05, |
| "loss": 0.2499, |
| "mean_token_accuracy": 0.8448965847492218, |
| "num_tokens": 4153117.0, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.39985719385933594, |
| "grad_norm": 0.17903761565685272, |
| "learning_rate": 2.3757659706983472e-05, |
| "loss": 0.2482, |
| "mean_token_accuracy": 0.8377207100391388, |
| "num_tokens": 4229717.0, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.4069975008925384, |
| "grad_norm": 0.1580619513988495, |
| "learning_rate": 2.3380175162787408e-05, |
| "loss": 0.2437, |
| "mean_token_accuracy": 0.8535458654165268, |
| "num_tokens": 4305929.0, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.4141378079257408, |
| "grad_norm": 0.25296780467033386, |
| "learning_rate": 2.2999580553580573e-05, |
| "loss": 0.2443, |
| "mean_token_accuracy": 0.8413390308618546, |
| "num_tokens": 4380511.0, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.42127811495894324, |
| "grad_norm": 0.10766352713108063, |
| "learning_rate": 2.261607717846456e-05, |
| "loss": 0.2507, |
| "mean_token_accuracy": 0.8452677458524704, |
| "num_tokens": 4456247.0, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.42841842199214564, |
| "grad_norm": 0.17321763932704926, |
| "learning_rate": 2.2229867875007188e-05, |
| "loss": 0.2393, |
| "mean_token_accuracy": 0.8409263521432877, |
| "num_tokens": 4533226.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.42841842199214564, |
| "eval_loss": 0.24729590117931366, |
| "eval_num_tokens": 4533226.0, |
| "eval_runtime": 22.6919, |
| "eval_samples_per_second": 10.004, |
| "eval_steps_per_second": 10.004, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.4355587290253481, |
| "grad_norm": 0.17038655281066895, |
| "learning_rate": 2.1841156911960374e-05, |
| "loss": 0.2452, |
| "mean_token_accuracy": 0.8418275892734528, |
| "num_tokens": 4608335.0, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.44269903605855054, |
| "grad_norm": 0.18034692108631134, |
| "learning_rate": 2.1450149881220956e-05, |
| "loss": 0.2437, |
| "mean_token_accuracy": 0.839489859342575, |
| "num_tokens": 4684280.0, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.44983934309175294, |
| "grad_norm": 0.11238481104373932, |
| "learning_rate": 2.105705358909175e-05, |
| "loss": 0.2598, |
| "mean_token_accuracy": 0.8357042014598847, |
| "num_tokens": 4761794.0, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.4569796501249554, |
| "grad_norm": 0.09184237569570541, |
| "learning_rate": 2.0662075946900234e-05, |
| "loss": 0.2457, |
| "mean_token_accuracy": 0.8580617070198059, |
| "num_tokens": 4837724.0, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.4641199571581578, |
| "grad_norm": 0.15013352036476135, |
| "learning_rate": 2.0265425861032887e-05, |
| "loss": 0.249, |
| "mean_token_accuracy": 0.8412411540746689, |
| "num_tokens": 4913233.0, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.47126026419136025, |
| "grad_norm": 0.14245887100696564, |
| "learning_rate": 1.9867313122443096e-05, |
| "loss": 0.2507, |
| "mean_token_accuracy": 0.8382949590682983, |
| "num_tokens": 4987339.0, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.47840057122456264, |
| "grad_norm": 0.24093951284885406, |
| "learning_rate": 1.9467948295691343e-05, |
| "loss": 0.245, |
| "mean_token_accuracy": 0.8535878717899322, |
| "num_tokens": 5062116.0, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.4855408782577651, |
| "grad_norm": 0.4231225550174713, |
| "learning_rate": 1.906754260757611e-05, |
| "loss": 0.2427, |
| "mean_token_accuracy": 0.8463588684797287, |
| "num_tokens": 5137441.0, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.4926811852909675, |
| "grad_norm": 0.14312925934791565, |
| "learning_rate": 1.8666307835414613e-05, |
| "loss": 0.246, |
| "mean_token_accuracy": 0.8349683076143265, |
| "num_tokens": 5211726.0, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.49982149232416995, |
| "grad_norm": 0.08626345545053482, |
| "learning_rate": 1.8264456195032308e-05, |
| "loss": 0.2473, |
| "mean_token_accuracy": 0.8358709186315536, |
| "num_tokens": 5287537.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.5069617993573724, |
| "grad_norm": 0.1616782546043396, |
| "learning_rate": 1.7862200228520494e-05, |
| "loss": 0.2411, |
| "mean_token_accuracy": 0.8455326408147812, |
| "num_tokens": 5362681.0, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.5141021063905747, |
| "grad_norm": 0.10026481747627258, |
| "learning_rate": 1.7459752691821326e-05, |
| "loss": 0.249, |
| "mean_token_accuracy": 0.8556286782026291, |
| "num_tokens": 5436979.0, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.5212424134237772, |
| "grad_norm": 0.19526486098766327, |
| "learning_rate": 1.705732644219974e-05, |
| "loss": 0.2437, |
| "mean_token_accuracy": 0.8449528038501739, |
| "num_tokens": 5512627.0, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.5283827204569796, |
| "grad_norm": 0.10900211334228516, |
| "learning_rate": 1.6655134325661816e-05, |
| "loss": 0.2374, |
| "mean_token_accuracy": 0.8406415730714798, |
| "num_tokens": 5588305.0, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.5355230274901821, |
| "grad_norm": 0.13656799495220184, |
| "learning_rate": 1.625338906437902e-05, |
| "loss": 0.2376, |
| "mean_token_accuracy": 0.8520337104797363, |
| "num_tokens": 5663626.0, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.5426633345233846, |
| "grad_norm": 0.14665727317333221, |
| "learning_rate": 1.5852303144178014e-05, |
| "loss": 0.2463, |
| "mean_token_accuracy": 0.8331857204437256, |
| "num_tokens": 5739869.0, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.5498036415565869, |
| "grad_norm": 0.16017290949821472, |
| "learning_rate": 1.5452088702155413e-05, |
| "loss": 0.2447, |
| "mean_token_accuracy": 0.8558647125959397, |
| "num_tokens": 5815701.0, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.5569439485897894, |
| "grad_norm": 0.16595233976840973, |
| "learning_rate": 1.5052957414477072e-05, |
| "loss": 0.2416, |
| "mean_token_accuracy": 0.8570067435503006, |
| "num_tokens": 5892093.0, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.5640842556229918, |
| "grad_norm": 0.16811279952526093, |
| "learning_rate": 1.4655120384421066e-05, |
| "loss": 0.2459, |
| "mean_token_accuracy": 0.8582445293664932, |
| "num_tokens": 5968703.0, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.5712245626561943, |
| "grad_norm": 0.1217021569609642, |
| "learning_rate": 1.4258788030723725e-05, |
| "loss": 0.2396, |
| "mean_token_accuracy": 0.8545576363801957, |
| "num_tokens": 6043916.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.5712245626561943, |
| "eval_loss": 0.24719083309173584, |
| "eval_num_tokens": 6043916.0, |
| "eval_runtime": 23.1714, |
| "eval_samples_per_second": 9.797, |
| "eval_steps_per_second": 9.797, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.5783648696893966, |
| "grad_norm": 0.2356659471988678, |
| "learning_rate": 1.3864169976287719e-05, |
| "loss": 0.2499, |
| "mean_token_accuracy": 0.8494862943887711, |
| "num_tokens": 6119439.0, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.585505176722599, |
| "grad_norm": 0.16304591298103333, |
| "learning_rate": 1.347147493731108e-05, |
| "loss": 0.2393, |
| "mean_token_accuracy": 0.8492362052202225, |
| "num_tokens": 6194667.0, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.5926454837558015, |
| "grad_norm": 0.1561768501996994, |
| "learning_rate": 1.3080910612895759e-05, |
| "loss": 0.239, |
| "mean_token_accuracy": 0.8470070034265518, |
| "num_tokens": 6270318.0, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.599785790789004, |
| "grad_norm": 0.14306160807609558, |
| "learning_rate": 1.2692683575194155e-05, |
| "loss": 0.2446, |
| "mean_token_accuracy": 0.8527232110500336, |
| "num_tokens": 6344479.0, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.6069260978222063, |
| "grad_norm": 0.09408228099346161, |
| "learning_rate": 1.2306999160151655e-05, |
| "loss": 0.2461, |
| "mean_token_accuracy": 0.8437263667583466, |
| "num_tokens": 6419982.0, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.6140664048554088, |
| "grad_norm": 0.1325320303440094, |
| "learning_rate": 1.1924061358903091e-05, |
| "loss": 0.2392, |
| "mean_token_accuracy": 0.8650959640741348, |
| "num_tokens": 6497373.0, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.6212067118886112, |
| "grad_norm": 0.1599915474653244, |
| "learning_rate": 1.1544072709880399e-05, |
| "loss": 0.2468, |
| "mean_token_accuracy": 0.8459037929773331, |
| "num_tokens": 6574671.0, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.6283470189218137, |
| "grad_norm": 0.1079430803656578, |
| "learning_rate": 1.1167234191688655e-05, |
| "loss": 0.2371, |
| "mean_token_accuracy": 0.8470400452613831, |
| "num_tokens": 6650119.0, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.6354873259550161, |
| "grad_norm": 0.10566870123147964, |
| "learning_rate": 1.0793745116807102e-05, |
| "loss": 0.2403, |
| "mean_token_accuracy": 0.8443713188171387, |
| "num_tokens": 6727945.0, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.6426276329882185, |
| "grad_norm": 0.16900038719177246, |
| "learning_rate": 1.0423803026171438e-05, |
| "loss": 0.2294, |
| "mean_token_accuracy": 0.8559036642313004, |
| "num_tokens": 6805156.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.6497679400214209, |
| "grad_norm": 0.16677226126194, |
| "learning_rate": 1.0057603584693038e-05, |
| "loss": 0.2422, |
| "mean_token_accuracy": 0.8480152815580368, |
| "num_tokens": 6880060.0, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.6569082470546234, |
| "grad_norm": 0.10336356610059738, |
| "learning_rate": 9.695340477770451e-06, |
| "loss": 0.2502, |
| "mean_token_accuracy": 0.8547079175710678, |
| "num_tokens": 6956427.0, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.6640485540878258, |
| "grad_norm": 0.13533741235733032, |
| "learning_rate": 9.337205308847836e-06, |
| "loss": 0.2535, |
| "mean_token_accuracy": 0.8299418836832047, |
| "num_tokens": 7031132.0, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.6711888611210282, |
| "grad_norm": 0.11439425498247147, |
| "learning_rate": 8.983387498074633e-06, |
| "loss": 0.2475, |
| "mean_token_accuracy": 0.8546998083591462, |
| "num_tokens": 7107493.0, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.6783291681542306, |
| "grad_norm": 0.11505307257175446, |
| "learning_rate": 8.634074182119893e-06, |
| "loss": 0.2469, |
| "mean_token_accuracy": 0.8413250893354416, |
| "num_tokens": 7182676.0, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.6854694751874331, |
| "grad_norm": 0.12734931707382202, |
| "learning_rate": 8.289450115194459e-06, |
| "loss": 0.2437, |
| "mean_token_accuracy": 0.8464201092720032, |
| "num_tokens": 7257594.0, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.6926097822206355, |
| "grad_norm": 0.10180549323558807, |
| "learning_rate": 7.949697571333143e-06, |
| "loss": 0.2445, |
| "mean_token_accuracy": 0.8510806947946549, |
| "num_tokens": 7333106.0, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.699750089253838, |
| "grad_norm": 0.13846957683563232, |
| "learning_rate": 7.614996247988788e-06, |
| "loss": 0.2424, |
| "mean_token_accuracy": 0.8456225216388702, |
| "num_tokens": 7409348.0, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.7068903962870403, |
| "grad_norm": 0.09437349438667297, |
| "learning_rate": 7.2855231709890236e-06, |
| "loss": 0.253, |
| "mean_token_accuracy": 0.8329681754112244, |
| "num_tokens": 7485199.0, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.7140307033202428, |
| "grad_norm": 0.07914232462644577, |
| "learning_rate": 6.9614526009061e-06, |
| "loss": 0.2454, |
| "mean_token_accuracy": 0.854900935292244, |
| "num_tokens": 7561901.0, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.7140307033202428, |
| "eval_loss": 0.24819402396678925, |
| "eval_num_tokens": 7561901.0, |
| "eval_runtime": 23.4229, |
| "eval_samples_per_second": 9.691, |
| "eval_steps_per_second": 9.691, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.7211710103534452, |
| "grad_norm": 0.1411924809217453, |
| "learning_rate": 6.642955940889251e-06, |
| "loss": 0.2456, |
| "mean_token_accuracy": 0.8437583088874817, |
| "num_tokens": 7637941.0, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.7283113173866477, |
| "grad_norm": 0.19878213107585907, |
| "learning_rate": 6.330201646008442e-06, |
| "loss": 0.241, |
| "mean_token_accuracy": 0.8387387841939926, |
| "num_tokens": 7713694.0, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.73545162441985, |
| "grad_norm": 0.11221861094236374, |
| "learning_rate": 6.023355134157266e-06, |
| "loss": 0.2434, |
| "mean_token_accuracy": 0.8500531584024429, |
| "num_tokens": 7787558.0, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.7425919314530525, |
| "grad_norm": 0.1809459924697876, |
| "learning_rate": 5.722578698562343e-06, |
| "loss": 0.2365, |
| "mean_token_accuracy": 0.8351176798343658, |
| "num_tokens": 7863004.0, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.7497322384862549, |
| "grad_norm": 0.11574816703796387, |
| "learning_rate": 5.428031421945264e-06, |
| "loss": 0.2468, |
| "mean_token_accuracy": 0.8377744346857071, |
| "num_tokens": 7936809.0, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.7568725455194574, |
| "grad_norm": 0.2032371312379837, |
| "learning_rate": 5.139869092382695e-06, |
| "loss": 0.242, |
| "mean_token_accuracy": 0.8558525443077087, |
| "num_tokens": 8012517.0, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.7640128525526597, |
| "grad_norm": 0.2128075659275055, |
| "learning_rate": 4.858244120908964e-06, |
| "loss": 0.2471, |
| "mean_token_accuracy": 0.8502045452594758, |
| "num_tokens": 8088039.0, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.7711531595858622, |
| "grad_norm": 0.3007623851299286, |
| "learning_rate": 4.583305460904831e-06, |
| "loss": 0.247, |
| "mean_token_accuracy": 0.834997683763504, |
| "num_tokens": 8163897.0, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.7782934666190646, |
| "grad_norm": 0.1440867930650711, |
| "learning_rate": 4.315198529315006e-06, |
| "loss": 0.2413, |
| "mean_token_accuracy": 0.8360197603702545, |
| "num_tokens": 8240262.0, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.7854337736522671, |
| "grad_norm": 0.16527746617794037, |
| "learning_rate": 4.054065129736162e-06, |
| "loss": 0.2399, |
| "mean_token_accuracy": 0.8457112580537796, |
| "num_tokens": 8315981.0, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.7925740806854695, |
| "grad_norm": 0.10186518728733063, |
| "learning_rate": 3.800043377415992e-06, |
| "loss": 0.2425, |
| "mean_token_accuracy": 0.8427910953760147, |
| "num_tokens": 8391626.0, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.7997143877186719, |
| "grad_norm": 0.14122404158115387, |
| "learning_rate": 3.55326762620316e-06, |
| "loss": 0.2571, |
| "mean_token_accuracy": 0.8383598476648331, |
| "num_tokens": 8468014.0, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.8068546947518743, |
| "grad_norm": 0.0955263152718544, |
| "learning_rate": 3.31386839748658e-06, |
| "loss": 0.2498, |
| "mean_token_accuracy": 0.8421018213033676, |
| "num_tokens": 8543756.0, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.8139950017850768, |
| "grad_norm": 0.11213722825050354, |
| "learning_rate": 3.081972311161786e-06, |
| "loss": 0.2449, |
| "mean_token_accuracy": 0.8530462324619293, |
| "num_tokens": 8618675.0, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.8211353088182792, |
| "grad_norm": 0.10946866124868393, |
| "learning_rate": 2.8577020186607968e-06, |
| "loss": 0.2334, |
| "mean_token_accuracy": 0.8431146681308747, |
| "num_tokens": 8696120.0, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.8282756158514816, |
| "grad_norm": 0.10328326374292374, |
| "learning_rate": 2.641176138080925e-06, |
| "loss": 0.2361, |
| "mean_token_accuracy": 0.8506006717681884, |
| "num_tokens": 8772797.0, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.835415922884684, |
| "grad_norm": 0.2858479917049408, |
| "learning_rate": 2.432509191446844e-06, |
| "loss": 0.2501, |
| "mean_token_accuracy": 0.8453244566917419, |
| "num_tokens": 8848681.0, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.8425562299178865, |
| "grad_norm": 0.12414267659187317, |
| "learning_rate": 2.2318115441391242e-06, |
| "loss": 0.2496, |
| "mean_token_accuracy": 0.843856155872345, |
| "num_tokens": 8924564.0, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.8496965369510889, |
| "grad_norm": 0.11058641970157623, |
| "learning_rate": 2.039189346521201e-06, |
| "loss": 0.2468, |
| "mean_token_accuracy": 0.8599821984767914, |
| "num_tokens": 9000065.0, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.8568368439842913, |
| "grad_norm": 0.10353564471006393, |
| "learning_rate": 1.8547444777957463e-06, |
| "loss": 0.247, |
| "mean_token_accuracy": 0.8582412838935852, |
| "num_tokens": 9076265.0, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.8568368439842913, |
| "eval_loss": 0.24870693683624268, |
| "eval_num_tokens": 9076265.0, |
| "eval_runtime": 22.8496, |
| "eval_samples_per_second": 9.935, |
| "eval_steps_per_second": 9.935, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.8639771510174937, |
| "grad_norm": 0.17142567038536072, |
| "learning_rate": 1.6785744921200291e-06, |
| "loss": 0.2468, |
| "mean_token_accuracy": 0.8505592197179794, |
| "num_tokens": 9150578.0, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.8711174580506962, |
| "grad_norm": 0.1010131761431694, |
| "learning_rate": 1.5107725670088751e-06, |
| "loss": 0.248, |
| "mean_token_accuracy": 0.8410425752401351, |
| "num_tokens": 9225319.0, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.8782577650838986, |
| "grad_norm": 0.1269058734178543, |
| "learning_rate": 1.3514274540524292e-06, |
| "loss": 0.242, |
| "mean_token_accuracy": 0.8544023931026459, |
| "num_tokens": 9301989.0, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.8853980721171011, |
| "grad_norm": 0.1056295782327652, |
| "learning_rate": 1.2006234319748362e-06, |
| "loss": 0.2393, |
| "mean_token_accuracy": 0.8628782987594604, |
| "num_tokens": 9377759.0, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.8925383791503034, |
| "grad_norm": 0.14275795221328735, |
| "learning_rate": 1.058440262058628e-06, |
| "loss": 0.2443, |
| "mean_token_accuracy": 0.8523239493370056, |
| "num_tokens": 9455774.0, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.8996786861835059, |
| "grad_norm": 0.10365872085094452, |
| "learning_rate": 9.249531459584625e-07, |
| "loss": 0.248, |
| "mean_token_accuracy": 0.851286780834198, |
| "num_tokens": 9531579.0, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.9068189932167083, |
| "grad_norm": 0.10703132301568985, |
| "learning_rate": 8.002326859264233e-07, |
| "loss": 0.2444, |
| "mean_token_accuracy": 0.8504409581422806, |
| "num_tokens": 9607022.0, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.9139593002499108, |
| "grad_norm": 0.12737736105918884, |
| "learning_rate": 6.843448474700275e-07, |
| "loss": 0.2451, |
| "mean_token_accuracy": 0.8450479298830033, |
| "num_tokens": 9682771.0, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.9210996072831131, |
| "grad_norm": 0.14681857824325562, |
| "learning_rate": 5.773509244625948e-07, |
| "loss": 0.2412, |
| "mean_token_accuracy": 0.8388412922620774, |
| "num_tokens": 9760900.0, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.9282399143163156, |
| "grad_norm": 0.11775557696819305, |
| "learning_rate": 4.793075067245142e-07, |
| "loss": 0.2524, |
| "mean_token_accuracy": 0.8410770803689956, |
| "num_tokens": 9835951.0, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.935380221349518, |
| "grad_norm": 0.174688920378685, |
| "learning_rate": 3.9026645009249404e-07, |
| "loss": 0.2397, |
| "mean_token_accuracy": 0.8388120770454407, |
| "num_tokens": 9910832.0, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.9425205283827205, |
| "grad_norm": 0.16480177640914917, |
| "learning_rate": 3.1027484899264876e-07, |
| "loss": 0.2525, |
| "mean_token_accuracy": 0.8406508564949036, |
| "num_tokens": 9985285.0, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.9496608354159228, |
| "grad_norm": 0.11839897930622101, |
| "learning_rate": 2.393750115319379e-07, |
| "loss": 0.2438, |
| "mean_token_accuracy": 0.8284848064184189, |
| "num_tokens": 10061386.0, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.9568011424491253, |
| "grad_norm": 0.1104646697640419, |
| "learning_rate": 1.7760443712112132e-07, |
| "loss": 0.2363, |
| "mean_token_accuracy": 0.8469839125871659, |
| "num_tokens": 10138726.0, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.9639414494823277, |
| "grad_norm": 0.17600639164447784, |
| "learning_rate": 1.249957966410553e-07, |
| "loss": 0.2455, |
| "mean_token_accuracy": 0.8401055574417114, |
| "num_tokens": 10213173.0, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.9710817565155302, |
| "grad_norm": 0.25562846660614014, |
| "learning_rate": 8.157691516284759e-08, |
| "loss": 0.2542, |
| "mean_token_accuracy": 0.8507506132125855, |
| "num_tokens": 10288483.0, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.9782220635487326, |
| "grad_norm": 0.12573958933353424, |
| "learning_rate": 4.737075723097839e-08, |
| "loss": 0.2492, |
| "mean_token_accuracy": 0.8376883298158646, |
| "num_tokens": 10364207.0, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.985362370581935, |
| "grad_norm": 0.0995529517531395, |
| "learning_rate": 2.2395414717205073e-08, |
| "loss": 0.2396, |
| "mean_token_accuracy": 0.8528329610824585, |
| "num_tokens": 10439424.0, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.9925026776151374, |
| "grad_norm": 0.14106954634189606, |
| "learning_rate": 6.664097251635192e-09, |
| "loss": 0.2439, |
| "mean_token_accuracy": 0.8580034106969834, |
| "num_tokens": 10514957.0, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.9996429846483399, |
| "grad_norm": 0.15564119815826416, |
| "learning_rate": 1.8512523607155937e-10, |
| "loss": 0.2479, |
| "mean_token_accuracy": 0.8343614429235459, |
| "num_tokens": 10590673.0, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.9996429846483399, |
| "eval_loss": 0.24886475503444672, |
| "eval_num_tokens": 10590673.0, |
| "eval_runtime": 22.7556, |
| "eval_samples_per_second": 9.976, |
| "eval_steps_per_second": 9.976, |
| "step": 1400 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 1401, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 200, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.8018463453733683e+17, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
