| { | |
| "best_metric": 1.6479060649871826, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-350", | |
| "epoch": 0.9090909090909091, | |
| "eval_steps": 50, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0022727272727272726, | |
| "grad_norm": 1.5825589895248413, | |
| "learning_rate": 1e-05, | |
| "loss": 2.1455, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0022727272727272726, | |
| "eval_loss": 2.5897326469421387, | |
| "eval_runtime": 56.3066, | |
| "eval_samples_per_second": 13.16, | |
| "eval_steps_per_second": 3.303, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004545454545454545, | |
| "grad_norm": 1.4482614994049072, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1563, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.006818181818181818, | |
| "grad_norm": 1.4861360788345337, | |
| "learning_rate": 3e-05, | |
| "loss": 2.2397, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00909090909090909, | |
| "grad_norm": 1.34442138671875, | |
| "learning_rate": 4e-05, | |
| "loss": 2.2328, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.011363636363636364, | |
| "grad_norm": 1.3896499872207642, | |
| "learning_rate": 5e-05, | |
| "loss": 2.1676, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.013636363636363636, | |
| "grad_norm": 1.3315941095352173, | |
| "learning_rate": 6e-05, | |
| "loss": 2.2849, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.015909090909090907, | |
| "grad_norm": 1.1799923181533813, | |
| "learning_rate": 7e-05, | |
| "loss": 2.1404, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.01818181818181818, | |
| "grad_norm": 1.016822338104248, | |
| "learning_rate": 8e-05, | |
| "loss": 2.0373, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.020454545454545454, | |
| "grad_norm": 1.198225736618042, | |
| "learning_rate": 9e-05, | |
| "loss": 1.8749, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.022727272727272728, | |
| "grad_norm": 1.3495861291885376, | |
| "learning_rate": 0.0001, | |
| "loss": 2.0171, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 1.1111146211624146, | |
| "learning_rate": 9.99983777858264e-05, | |
| "loss": 1.9444, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02727272727272727, | |
| "grad_norm": 0.9873141050338745, | |
| "learning_rate": 9.999351124856874e-05, | |
| "loss": 1.925, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.029545454545454545, | |
| "grad_norm": 1.0353142023086548, | |
| "learning_rate": 9.998540070400966e-05, | |
| "loss": 1.9103, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.031818181818181815, | |
| "grad_norm": 1.1844122409820557, | |
| "learning_rate": 9.997404667843075e-05, | |
| "loss": 1.9329, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03409090909090909, | |
| "grad_norm": 1.1251015663146973, | |
| "learning_rate": 9.995944990857849e-05, | |
| "loss": 1.9096, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03636363636363636, | |
| "grad_norm": 1.1467946767807007, | |
| "learning_rate": 9.994161134161634e-05, | |
| "loss": 1.7946, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.038636363636363635, | |
| "grad_norm": 1.1272140741348267, | |
| "learning_rate": 9.992053213506334e-05, | |
| "loss": 1.8906, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04090909090909091, | |
| "grad_norm": 1.1868035793304443, | |
| "learning_rate": 9.989621365671902e-05, | |
| "loss": 1.9761, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04318181818181818, | |
| "grad_norm": 1.2175211906433105, | |
| "learning_rate": 9.986865748457457e-05, | |
| "loss": 1.8191, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.045454545454545456, | |
| "grad_norm": 1.2086597681045532, | |
| "learning_rate": 9.983786540671051e-05, | |
| "loss": 1.9269, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04772727272727273, | |
| "grad_norm": 1.14625883102417, | |
| "learning_rate": 9.980383942118066e-05, | |
| "loss": 1.8706, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.0759248733520508, | |
| "learning_rate": 9.976658173588244e-05, | |
| "loss": 1.6653, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05227272727272727, | |
| "grad_norm": 1.2646729946136475, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 1.6592, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.05454545454545454, | |
| "grad_norm": 1.09037184715271, | |
| "learning_rate": 9.968238114591566e-05, | |
| "loss": 1.5821, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.056818181818181816, | |
| "grad_norm": 1.5458767414093018, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 1.8794, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05909090909090909, | |
| "grad_norm": 1.3365269899368286, | |
| "learning_rate": 9.95852854910781e-05, | |
| "loss": 1.669, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06136363636363636, | |
| "grad_norm": 1.5855895280838013, | |
| "learning_rate": 9.953190975913647e-05, | |
| "loss": 1.687, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06363636363636363, | |
| "grad_norm": 1.3576229810714722, | |
| "learning_rate": 9.947531997255256e-05, | |
| "loss": 1.8413, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0659090909090909, | |
| "grad_norm": 1.3961035013198853, | |
| "learning_rate": 9.941551980335652e-05, | |
| "loss": 1.433, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.06818181818181818, | |
| "grad_norm": 1.439440369606018, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 1.5848, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07045454545454545, | |
| "grad_norm": 1.7640531063079834, | |
| "learning_rate": 9.928630404658255e-05, | |
| "loss": 1.7603, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.07272727272727272, | |
| "grad_norm": 1.8586909770965576, | |
| "learning_rate": 9.921689684362989e-05, | |
| "loss": 1.822, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 1.8702236413955688, | |
| "learning_rate": 9.914429602677162e-05, | |
| "loss": 1.623, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.07727272727272727, | |
| "grad_norm": 1.9076231718063354, | |
| "learning_rate": 9.906850630697068e-05, | |
| "loss": 1.5713, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.07954545454545454, | |
| "grad_norm": 2.04288649559021, | |
| "learning_rate": 9.898953260211338e-05, | |
| "loss": 1.4979, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08181818181818182, | |
| "grad_norm": 1.918068289756775, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.6285, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.08409090909090909, | |
| "grad_norm": 2.0530202388763428, | |
| "learning_rate": 9.882205394146361e-05, | |
| "loss": 1.5323, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.08636363636363636, | |
| "grad_norm": 2.3702471256256104, | |
| "learning_rate": 9.87335598531214e-05, | |
| "loss": 1.5823, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.08863636363636364, | |
| "grad_norm": 2.307603597640991, | |
| "learning_rate": 9.864190351391822e-05, | |
| "loss": 1.6971, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 2.700122833251953, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 1.6285, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09318181818181819, | |
| "grad_norm": 3.2432916164398193, | |
| "learning_rate": 9.844912807753104e-05, | |
| "loss": 1.5747, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.09545454545454546, | |
| "grad_norm": 2.6756370067596436, | |
| "learning_rate": 9.834802148926882e-05, | |
| "loss": 1.4143, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.09772727272727273, | |
| "grad_norm": 2.664668321609497, | |
| "learning_rate": 9.824377766717759e-05, | |
| "loss": 1.2366, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.7741057872772217, | |
| "learning_rate": 9.813640337548954e-05, | |
| "loss": 1.7072, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.10227272727272728, | |
| "grad_norm": 2.7716453075408936, | |
| "learning_rate": 9.802590558156862e-05, | |
| "loss": 1.3241, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10454545454545454, | |
| "grad_norm": 3.296673536300659, | |
| "learning_rate": 9.791229145545831e-05, | |
| "loss": 1.7961, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.10681818181818181, | |
| "grad_norm": 3.684264898300171, | |
| "learning_rate": 9.779556836941645e-05, | |
| "loss": 1.7892, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.10909090909090909, | |
| "grad_norm": 3.983945846557617, | |
| "learning_rate": 9.767574389743682e-05, | |
| "loss": 1.764, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.11136363636363636, | |
| "grad_norm": 7.055461883544922, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.129, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.11363636363636363, | |
| "grad_norm": 12.737040519714355, | |
| "learning_rate": 9.742682209735727e-05, | |
| "loss": 2.5162, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11363636363636363, | |
| "eval_loss": 2.159236431121826, | |
| "eval_runtime": 57.2811, | |
| "eval_samples_per_second": 12.936, | |
| "eval_steps_per_second": 3.247, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1159090909090909, | |
| "grad_norm": 5.642988681793213, | |
| "learning_rate": 9.729774092143627e-05, | |
| "loss": 2.4571, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.11818181818181818, | |
| "grad_norm": 3.5502121448516846, | |
| "learning_rate": 9.716559066288715e-05, | |
| "loss": 2.2891, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.12045454545454545, | |
| "grad_norm": 1.7564338445663452, | |
| "learning_rate": 9.703037989675087e-05, | |
| "loss": 2.1306, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.12272727272727273, | |
| "grad_norm": 1.1576513051986694, | |
| "learning_rate": 9.689211739666023e-05, | |
| "loss": 2.0528, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 0.8917579054832458, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 2.0089, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12727272727272726, | |
| "grad_norm": 0.774079442024231, | |
| "learning_rate": 9.66064732786784e-05, | |
| "loss": 1.7843, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.12954545454545455, | |
| "grad_norm": 0.8814948797225952, | |
| "learning_rate": 9.645911019582467e-05, | |
| "loss": 1.845, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.1318181818181818, | |
| "grad_norm": 0.7452542781829834, | |
| "learning_rate": 9.630873244788883e-05, | |
| "loss": 1.8975, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1340909090909091, | |
| "grad_norm": 0.7592681050300598, | |
| "learning_rate": 9.615534979266745e-05, | |
| "loss": 1.8472, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.13636363636363635, | |
| "grad_norm": 0.8305085897445679, | |
| "learning_rate": 9.599897218294122e-05, | |
| "loss": 1.8512, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13863636363636364, | |
| "grad_norm": 0.7982023358345032, | |
| "learning_rate": 9.583960976582913e-05, | |
| "loss": 1.8055, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1409090909090909, | |
| "grad_norm": 0.782379150390625, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.7597, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1431818181818182, | |
| "grad_norm": 0.8304200172424316, | |
| "learning_rate": 9.551197206565173e-05, | |
| "loss": 1.7911, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.14545454545454545, | |
| "grad_norm": 0.950410783290863, | |
| "learning_rate": 9.534371804252728e-05, | |
| "loss": 1.8224, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.14772727272727273, | |
| "grad_norm": 0.8509146571159363, | |
| "learning_rate": 9.517252173051911e-05, | |
| "loss": 1.793, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9001911878585815, | |
| "learning_rate": 9.49983942383106e-05, | |
| "loss": 1.7985, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.15227272727272728, | |
| "grad_norm": 0.9097445011138916, | |
| "learning_rate": 9.482134686478519e-05, | |
| "loss": 1.7203, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.15454545454545454, | |
| "grad_norm": 0.9786297678947449, | |
| "learning_rate": 9.464139109829321e-05, | |
| "loss": 1.719, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.15681818181818183, | |
| "grad_norm": 0.9668626189231873, | |
| "learning_rate": 9.445853861590647e-05, | |
| "loss": 1.7404, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.1590909090909091, | |
| "grad_norm": 1.0761997699737549, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 1.6017, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16136363636363638, | |
| "grad_norm": 1.1242194175720215, | |
| "learning_rate": 9.408419115078471e-05, | |
| "loss": 1.824, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.16363636363636364, | |
| "grad_norm": 1.1410738229751587, | |
| "learning_rate": 9.389272045892024e-05, | |
| "loss": 1.6692, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.16590909090909092, | |
| "grad_norm": 1.165338158607483, | |
| "learning_rate": 9.36984016313259e-05, | |
| "loss": 1.7682, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.16818181818181818, | |
| "grad_norm": 1.1751066446304321, | |
| "learning_rate": 9.350124727707197e-05, | |
| "loss": 1.6575, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.17045454545454544, | |
| "grad_norm": 1.1582705974578857, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 1.6029, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17272727272727273, | |
| "grad_norm": 1.2535523176193237, | |
| "learning_rate": 9.309848334400246e-05, | |
| "loss": 1.6107, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 1.5007827281951904, | |
| "learning_rate": 9.289289989996133e-05, | |
| "loss": 1.6491, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.17727272727272728, | |
| "grad_norm": 1.2690426111221313, | |
| "learning_rate": 9.268453319711363e-05, | |
| "loss": 1.5701, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.17954545454545454, | |
| "grad_norm": 1.3464139699935913, | |
| "learning_rate": 9.247339675607605e-05, | |
| "loss": 1.6228, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 1.4133737087249756, | |
| "learning_rate": 9.225950427718975e-05, | |
| "loss": 1.4565, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18409090909090908, | |
| "grad_norm": 1.5410175323486328, | |
| "learning_rate": 9.204286963963111e-05, | |
| "loss": 1.6295, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.18636363636363637, | |
| "grad_norm": 1.8548007011413574, | |
| "learning_rate": 9.182350690051133e-05, | |
| "loss": 1.7913, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.18863636363636363, | |
| "grad_norm": 1.5754772424697876, | |
| "learning_rate": 9.160143029396422e-05, | |
| "loss": 1.6161, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.19090909090909092, | |
| "grad_norm": 1.5569956302642822, | |
| "learning_rate": 9.13766542302225e-05, | |
| "loss": 1.2768, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.19318181818181818, | |
| "grad_norm": 1.8846619129180908, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 1.4766, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19545454545454546, | |
| "grad_norm": 1.877416729927063, | |
| "learning_rate": 9.091906224695935e-05, | |
| "loss": 1.5793, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.19772727272727272, | |
| "grad_norm": 2.0326881408691406, | |
| "learning_rate": 9.068627601992598e-05, | |
| "loss": 1.4267, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 2.376016139984131, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.7734, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.20227272727272727, | |
| "grad_norm": 2.103180408477783, | |
| "learning_rate": 9.021279861989885e-05, | |
| "loss": 1.643, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.20454545454545456, | |
| "grad_norm": 1.992186188697815, | |
| "learning_rate": 8.997213817017507e-05, | |
| "loss": 1.3302, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20681818181818182, | |
| "grad_norm": 2.1027109622955322, | |
| "learning_rate": 8.972888398568772e-05, | |
| "loss": 1.4462, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.20909090909090908, | |
| "grad_norm": 1.9664043188095093, | |
| "learning_rate": 8.948305185085225e-05, | |
| "loss": 1.4575, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.21136363636363636, | |
| "grad_norm": 2.469254970550537, | |
| "learning_rate": 8.92346577173636e-05, | |
| "loss": 1.7019, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.21363636363636362, | |
| "grad_norm": 2.431694507598877, | |
| "learning_rate": 8.898371770316111e-05, | |
| "loss": 1.6776, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2159090909090909, | |
| "grad_norm": 2.6154329776763916, | |
| "learning_rate": 8.873024809138272e-05, | |
| "loss": 1.6518, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21818181818181817, | |
| "grad_norm": 2.664855480194092, | |
| "learning_rate": 8.847426532930831e-05, | |
| "loss": 1.7601, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.22045454545454546, | |
| "grad_norm": 3.687713861465454, | |
| "learning_rate": 8.821578602729242e-05, | |
| "loss": 1.7593, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.22272727272727272, | |
| "grad_norm": 4.494145393371582, | |
| "learning_rate": 8.795482695768658e-05, | |
| "loss": 1.748, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 9.136951446533203, | |
| "learning_rate": 8.769140505375085e-05, | |
| "loss": 1.8634, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 9.54596996307373, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 2.7632, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "eval_loss": 1.9716918468475342, | |
| "eval_runtime": 56.7645, | |
| "eval_samples_per_second": 13.054, | |
| "eval_steps_per_second": 3.277, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22954545454545455, | |
| "grad_norm": 2.4930062294006348, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 2.277, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.2318181818181818, | |
| "grad_norm": 1.7898674011230469, | |
| "learning_rate": 8.688653405904652e-05, | |
| "loss": 2.2272, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2340909090909091, | |
| "grad_norm": 1.2085728645324707, | |
| "learning_rate": 8.661343332988869e-05, | |
| "loss": 2.0419, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.23636363636363636, | |
| "grad_norm": 0.7948428392410278, | |
| "learning_rate": 8.633795680751116e-05, | |
| "loss": 1.8466, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.23863636363636365, | |
| "grad_norm": 0.6310542225837708, | |
| "learning_rate": 8.606012236719073e-05, | |
| "loss": 1.7921, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2409090909090909, | |
| "grad_norm": 0.6660516262054443, | |
| "learning_rate": 8.577994803720606e-05, | |
| "loss": 1.9284, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2431818181818182, | |
| "grad_norm": 0.7869789004325867, | |
| "learning_rate": 8.549745199766792e-05, | |
| "loss": 1.9913, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.24545454545454545, | |
| "grad_norm": 0.7278242111206055, | |
| "learning_rate": 8.521265257933948e-05, | |
| "loss": 1.8148, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.24772727272727274, | |
| "grad_norm": 0.7877641320228577, | |
| "learning_rate": 8.492556826244687e-05, | |
| "loss": 1.8851, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.771149218082428, | |
| "learning_rate": 8.463621767547998e-05, | |
| "loss": 1.784, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25227272727272726, | |
| "grad_norm": 0.7780814170837402, | |
| "learning_rate": 8.434461959398376e-05, | |
| "loss": 1.8547, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2545454545454545, | |
| "grad_norm": 0.8390243053436279, | |
| "learning_rate": 8.405079293933986e-05, | |
| "loss": 1.7756, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.25681818181818183, | |
| "grad_norm": 0.8175789713859558, | |
| "learning_rate": 8.375475677753881e-05, | |
| "loss": 1.6015, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2590909090909091, | |
| "grad_norm": 0.8630351424217224, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 1.809, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.26136363636363635, | |
| "grad_norm": 0.929864227771759, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 1.7913, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2636363636363636, | |
| "grad_norm": 0.9909330010414124, | |
| "learning_rate": 8.285358405218655e-05, | |
| "loss": 1.7055, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.26590909090909093, | |
| "grad_norm": 0.9742305874824524, | |
| "learning_rate": 8.25489033703452e-05, | |
| "loss": 1.7266, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2681818181818182, | |
| "grad_norm": 0.9148551225662231, | |
| "learning_rate": 8.224211063680853e-05, | |
| "loss": 1.7596, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.27045454545454545, | |
| "grad_norm": 1.1245241165161133, | |
| "learning_rate": 8.19332257589174e-05, | |
| "loss": 1.5876, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 0.9692203402519226, | |
| "learning_rate": 8.162226877976887e-05, | |
| "loss": 1.5285, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 1.0414646863937378, | |
| "learning_rate": 8.130925987691569e-05, | |
| "loss": 1.4326, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2772727272727273, | |
| "grad_norm": 1.1119694709777832, | |
| "learning_rate": 8.099421936105702e-05, | |
| "loss": 1.697, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.27954545454545454, | |
| "grad_norm": 1.1254791021347046, | |
| "learning_rate": 8.067716767472045e-05, | |
| "loss": 1.5894, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2818181818181818, | |
| "grad_norm": 1.0397603511810303, | |
| "learning_rate": 8.035812539093557e-05, | |
| "loss": 1.6561, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.2840909090909091, | |
| "grad_norm": 1.1851451396942139, | |
| "learning_rate": 8.003711321189895e-05, | |
| "loss": 1.632, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2863636363636364, | |
| "grad_norm": 1.2362711429595947, | |
| "learning_rate": 7.971415196763088e-05, | |
| "loss": 1.7342, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.28863636363636364, | |
| "grad_norm": 1.1873120069503784, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 1.5524, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 1.407163381576538, | |
| "learning_rate": 7.906246623448183e-05, | |
| "loss": 1.7548, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.29318181818181815, | |
| "grad_norm": 1.3823617696762085, | |
| "learning_rate": 7.873378403255419e-05, | |
| "loss": 1.6648, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.29545454545454547, | |
| "grad_norm": 1.4122381210327148, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 1.6129, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29772727272727273, | |
| "grad_norm": 1.2999695539474487, | |
| "learning_rate": 7.807084759519405e-05, | |
| "loss": 1.5797, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.629199743270874, | |
| "learning_rate": 7.773663637675694e-05, | |
| "loss": 1.8081, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.30227272727272725, | |
| "grad_norm": 1.7925862073898315, | |
| "learning_rate": 7.740062536773352e-05, | |
| "loss": 1.7419, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.30454545454545456, | |
| "grad_norm": 1.5080852508544922, | |
| "learning_rate": 7.706283637139658e-05, | |
| "loss": 1.3884, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3068181818181818, | |
| "grad_norm": 1.4949387311935425, | |
| "learning_rate": 7.672329130639005e-05, | |
| "loss": 1.4689, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3090909090909091, | |
| "grad_norm": 1.7226927280426025, | |
| "learning_rate": 7.638201220530665e-05, | |
| "loss": 1.3717, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.31136363636363634, | |
| "grad_norm": 1.970788836479187, | |
| "learning_rate": 7.603902121325813e-05, | |
| "loss": 1.5973, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.31363636363636366, | |
| "grad_norm": 1.949326515197754, | |
| "learning_rate": 7.569434058643844e-05, | |
| "loss": 1.4445, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3159090909090909, | |
| "grad_norm": 2.491636276245117, | |
| "learning_rate": 7.534799269067953e-05, | |
| "loss": 1.5557, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3181818181818182, | |
| "grad_norm": 2.221522092819214, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.2813, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.32045454545454544, | |
| "grad_norm": 2.3903207778930664, | |
| "learning_rate": 7.465038509514688e-05, | |
| "loss": 1.3909, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.32272727272727275, | |
| "grad_norm": 2.156156301498413, | |
| "learning_rate": 7.42991706621303e-05, | |
| "loss": 1.7252, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 2.55950927734375, | |
| "learning_rate": 7.394637949075154e-05, | |
| "loss": 1.5541, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.32727272727272727, | |
| "grad_norm": 2.2347304821014404, | |
| "learning_rate": 7.35920344731241e-05, | |
| "loss": 1.6432, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.32954545454545453, | |
| "grad_norm": 2.5954465866088867, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 1.496, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.33181818181818185, | |
| "grad_norm": 3.0030901432037354, | |
| "learning_rate": 7.287877497021978e-05, | |
| "loss": 1.8088, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.3340909090909091, | |
| "grad_norm": 5.477200508117676, | |
| "learning_rate": 7.251990676732984e-05, | |
| "loss": 1.8302, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.33636363636363636, | |
| "grad_norm": 4.523463726043701, | |
| "learning_rate": 7.215957727996207e-05, | |
| "loss": 2.1942, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3386363636363636, | |
| "grad_norm": 5.909972190856934, | |
| "learning_rate": 7.179780988938051e-05, | |
| "loss": 2.1971, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3409090909090909, | |
| "grad_norm": 6.263747692108154, | |
| "learning_rate": 7.143462807015271e-05, | |
| "loss": 2.5567, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3409090909090909, | |
| "eval_loss": 1.8857125043869019, | |
| "eval_runtime": 56.8075, | |
| "eval_samples_per_second": 13.044, | |
| "eval_steps_per_second": 3.274, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3431818181818182, | |
| "grad_norm": 1.6396535634994507, | |
| "learning_rate": 7.107005538862646e-05, | |
| "loss": 1.9603, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.34545454545454546, | |
| "grad_norm": 1.813876748085022, | |
| "learning_rate": 7.07041155014006e-05, | |
| "loss": 2.1505, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3477272727272727, | |
| "grad_norm": 1.245876669883728, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 1.9938, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.8962484002113342, | |
| "learning_rate": 6.996822917828477e-05, | |
| "loss": 1.7504, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.3522727272727273, | |
| "grad_norm": 0.8175750970840454, | |
| "learning_rate": 6.959833049300377e-05, | |
| "loss": 1.8551, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.35454545454545455, | |
| "grad_norm": 0.7571668028831482, | |
| "learning_rate": 6.922716010014255e-05, | |
| "loss": 1.8552, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.3568181818181818, | |
| "grad_norm": 0.6910615563392639, | |
| "learning_rate": 6.885474208441603e-05, | |
| "loss": 1.7962, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.35909090909090907, | |
| "grad_norm": 0.6795138120651245, | |
| "learning_rate": 6.848110061149556e-05, | |
| "loss": 1.7103, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.3613636363636364, | |
| "grad_norm": 0.7749749422073364, | |
| "learning_rate": 6.810625992644085e-05, | |
| "loss": 1.8269, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.7578449249267578, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 1.7598, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3659090909090909, | |
| "grad_norm": 0.8355617523193359, | |
| "learning_rate": 6.735307828766515e-05, | |
| "loss": 1.7385, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.36818181818181817, | |
| "grad_norm": 0.8077674508094788, | |
| "learning_rate": 6.697478620682137e-05, | |
| "loss": 1.7478, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.3704545454545455, | |
| "grad_norm": 0.834367573261261, | |
| "learning_rate": 6.659539265642643e-05, | |
| "loss": 1.8464, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.37272727272727274, | |
| "grad_norm": 0.8702117204666138, | |
| "learning_rate": 6.621492225478414e-05, | |
| "loss": 1.8095, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.8388840556144714, | |
| "learning_rate": 6.583339969007363e-05, | |
| "loss": 1.7239, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.37727272727272726, | |
| "grad_norm": 0.8587437272071838, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.7238, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3795454545454545, | |
| "grad_norm": 0.9449232220649719, | |
| "learning_rate": 6.506729716392481e-05, | |
| "loss": 1.4925, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.38181818181818183, | |
| "grad_norm": 1.028063416481018, | |
| "learning_rate": 6.468276691378155e-05, | |
| "loss": 1.4695, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.3840909090909091, | |
| "grad_norm": 1.010267972946167, | |
| "learning_rate": 6.429728391993446e-05, | |
| "loss": 1.6856, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.38636363636363635, | |
| "grad_norm": 0.9631296992301941, | |
| "learning_rate": 6.391087319582264e-05, | |
| "loss": 1.7071, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3886363636363636, | |
| "grad_norm": 1.1068823337554932, | |
| "learning_rate": 6.35235598150842e-05, | |
| "loss": 1.6841, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.39090909090909093, | |
| "grad_norm": 0.9881982803344727, | |
| "learning_rate": 6.313536890992935e-05, | |
| "loss": 1.4997, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.3931818181818182, | |
| "grad_norm": 1.1173704862594604, | |
| "learning_rate": 6.274632566950967e-05, | |
| "loss": 1.7673, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.39545454545454545, | |
| "grad_norm": 1.1943248510360718, | |
| "learning_rate": 6.235645533828349e-05, | |
| "loss": 1.6108, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3977272727272727, | |
| "grad_norm": 1.0729866027832031, | |
| "learning_rate": 6.19657832143779e-05, | |
| "loss": 1.5005, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.389591932296753, | |
| "learning_rate": 6.157433464794716e-05, | |
| "loss": 1.8246, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.4022727272727273, | |
| "grad_norm": 1.1723880767822266, | |
| "learning_rate": 6.118213503952779e-05, | |
| "loss": 1.5835, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.40454545454545454, | |
| "grad_norm": 1.2850761413574219, | |
| "learning_rate": 6.078920983839031e-05, | |
| "loss": 1.5418, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.4068181818181818, | |
| "grad_norm": 1.1876102685928345, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 1.4865, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.4090909090909091, | |
| "grad_norm": 1.4014610052108765, | |
| "learning_rate": 6.0001284688802226e-05, | |
| "loss": 1.7036, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4113636363636364, | |
| "grad_norm": 1.3281017541885376, | |
| "learning_rate": 5.960633586768543e-05, | |
| "loss": 1.4997, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.41363636363636364, | |
| "grad_norm": 1.4098666906356812, | |
| "learning_rate": 5.921076370520058e-05, | |
| "loss": 1.6918, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.4159090909090909, | |
| "grad_norm": 1.551309585571289, | |
| "learning_rate": 5.8814593869458455e-05, | |
| "loss": 1.7399, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.41818181818181815, | |
| "grad_norm": 1.6254075765609741, | |
| "learning_rate": 5.841785206735192e-05, | |
| "loss": 1.5785, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.42045454545454547, | |
| "grad_norm": 1.879658818244934, | |
| "learning_rate": 5.8020564042888015e-05, | |
| "loss": 1.4979, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.42272727272727273, | |
| "grad_norm": 1.7498807907104492, | |
| "learning_rate": 5.762275557551727e-05, | |
| "loss": 1.5391, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 1.4871435165405273, | |
| "learning_rate": 5.7224452478461064e-05, | |
| "loss": 1.396, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.42727272727272725, | |
| "grad_norm": 1.9054415225982666, | |
| "learning_rate": 5.682568059703659e-05, | |
| "loss": 1.699, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.42954545454545456, | |
| "grad_norm": 1.7112065553665161, | |
| "learning_rate": 5.642646580697973e-05, | |
| "loss": 1.3715, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4318181818181818, | |
| "grad_norm": 1.7900882959365845, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 1.2397, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4340909090909091, | |
| "grad_norm": 2.077932357788086, | |
| "learning_rate": 5.562681114593028e-05, | |
| "loss": 1.4125, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.43636363636363634, | |
| "grad_norm": 2.357682466506958, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 1.5047, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.43863636363636366, | |
| "grad_norm": 2.542640209197998, | |
| "learning_rate": 5.482569604572576e-05, | |
| "loss": 1.629, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.4409090909090909, | |
| "grad_norm": 2.49251389503479, | |
| "learning_rate": 5.442465579556793e-05, | |
| "loss": 1.6731, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.4431818181818182, | |
| "grad_norm": 2.617112636566162, | |
| "learning_rate": 5.402332843583631e-05, | |
| "loss": 1.5917, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.44545454545454544, | |
| "grad_norm": 3.3823883533477783, | |
| "learning_rate": 5.3621740008088126e-05, | |
| "loss": 1.744, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.44772727272727275, | |
| "grad_norm": 4.142048358917236, | |
| "learning_rate": 5.321991657082097e-05, | |
| "loss": 1.9433, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 3.9996933937072754, | |
| "learning_rate": 5.281788419778187e-05, | |
| "loss": 1.9832, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.45227272727272727, | |
| "grad_norm": 4.659099578857422, | |
| "learning_rate": 5.2415668976275355e-05, | |
| "loss": 1.7046, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 9.069555282592773, | |
| "learning_rate": 5.201329700547076e-05, | |
| "loss": 2.5753, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "eval_loss": 1.7734038829803467, | |
| "eval_runtime": 57.014, | |
| "eval_samples_per_second": 12.997, | |
| "eval_steps_per_second": 3.262, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.45681818181818185, | |
| "grad_norm": 0.8230122923851013, | |
| "learning_rate": 5.161079439470866e-05, | |
| "loss": 1.8995, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.4590909090909091, | |
| "grad_norm": 1.1307181119918823, | |
| "learning_rate": 5.1208187261806615e-05, | |
| "loss": 2.1963, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.46136363636363636, | |
| "grad_norm": 0.8498021364212036, | |
| "learning_rate": 5.080550173136457e-05, | |
| "loss": 1.992, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.4636363636363636, | |
| "grad_norm": 0.8756141066551208, | |
| "learning_rate": 5.0402763933069496e-05, | |
| "loss": 2.0252, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.4659090909090909, | |
| "grad_norm": 0.7922159433364868, | |
| "learning_rate": 5e-05, | |
| "loss": 1.9845, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4681818181818182, | |
| "grad_norm": 0.7838784456253052, | |
| "learning_rate": 4.9597236066930516e-05, | |
| "loss": 1.8631, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.47045454545454546, | |
| "grad_norm": 0.7158495187759399, | |
| "learning_rate": 4.919449826863544e-05, | |
| "loss": 1.9371, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.4727272727272727, | |
| "grad_norm": 0.725161075592041, | |
| "learning_rate": 4.87918127381934e-05, | |
| "loss": 1.7292, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.475, | |
| "grad_norm": 0.6915109157562256, | |
| "learning_rate": 4.8389205605291365e-05, | |
| "loss": 1.8597, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.4772727272727273, | |
| "grad_norm": 0.7210138440132141, | |
| "learning_rate": 4.798670299452926e-05, | |
| "loss": 1.9027, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.47954545454545455, | |
| "grad_norm": 0.7396478056907654, | |
| "learning_rate": 4.758433102372466e-05, | |
| "loss": 1.8029, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.4818181818181818, | |
| "grad_norm": 0.7768386602401733, | |
| "learning_rate": 4.7182115802218126e-05, | |
| "loss": 1.6535, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.48409090909090907, | |
| "grad_norm": 0.7233967781066895, | |
| "learning_rate": 4.678008342917903e-05, | |
| "loss": 1.627, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.4863636363636364, | |
| "grad_norm": 0.8505319952964783, | |
| "learning_rate": 4.6378259991911886e-05, | |
| "loss": 1.7172, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.48863636363636365, | |
| "grad_norm": 0.8245370388031006, | |
| "learning_rate": 4.597667156416371e-05, | |
| "loss": 1.6162, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4909090909090909, | |
| "grad_norm": 0.8295523524284363, | |
| "learning_rate": 4.5575344204432084e-05, | |
| "loss": 1.5304, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.49318181818181817, | |
| "grad_norm": 0.8710333704948425, | |
| "learning_rate": 4.5174303954274244e-05, | |
| "loss": 1.732, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.4954545454545455, | |
| "grad_norm": 0.8906968235969543, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 1.6825, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.49772727272727274, | |
| "grad_norm": 0.9994188547134399, | |
| "learning_rate": 4.437318885406973e-05, | |
| "loss": 1.7121, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.9736546874046326, | |
| "learning_rate": 4.397316598723385e-05, | |
| "loss": 1.5441, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5022727272727273, | |
| "grad_norm": 1.2273542881011963, | |
| "learning_rate": 4.3573534193020274e-05, | |
| "loss": 1.6355, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5045454545454545, | |
| "grad_norm": 0.982225239276886, | |
| "learning_rate": 4.317431940296343e-05, | |
| "loss": 1.6633, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5068181818181818, | |
| "grad_norm": 1.0423542261123657, | |
| "learning_rate": 4.277554752153895e-05, | |
| "loss": 1.5226, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.509090909090909, | |
| "grad_norm": 1.1639502048492432, | |
| "learning_rate": 4.237724442448273e-05, | |
| "loss": 1.6973, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5113636363636364, | |
| "grad_norm": 1.0587353706359863, | |
| "learning_rate": 4.197943595711198e-05, | |
| "loss": 1.4378, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5136363636363637, | |
| "grad_norm": 1.1679116487503052, | |
| "learning_rate": 4.1582147932648074e-05, | |
| "loss": 1.5652, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.5159090909090909, | |
| "grad_norm": 1.2100685834884644, | |
| "learning_rate": 4.118540613054156e-05, | |
| "loss": 1.5762, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5181818181818182, | |
| "grad_norm": 1.3331880569458008, | |
| "learning_rate": 4.078923629479943e-05, | |
| "loss": 1.6778, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5204545454545455, | |
| "grad_norm": 1.1564544439315796, | |
| "learning_rate": 4.039366413231458e-05, | |
| "loss": 1.4463, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5227272727272727, | |
| "grad_norm": 1.2036336660385132, | |
| "learning_rate": 3.9998715311197785e-05, | |
| "loss": 1.5798, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 1.3497117757797241, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 1.5303, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5272727272727272, | |
| "grad_norm": 1.3475767374038696, | |
| "learning_rate": 3.92107901616097e-05, | |
| "loss": 1.4704, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5295454545454545, | |
| "grad_norm": 1.3444191217422485, | |
| "learning_rate": 3.8817864960472236e-05, | |
| "loss": 1.386, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.5318181818181819, | |
| "grad_norm": 1.5539790391921997, | |
| "learning_rate": 3.842566535205286e-05, | |
| "loss": 1.5666, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5340909090909091, | |
| "grad_norm": 1.810019612312317, | |
| "learning_rate": 3.803421678562213e-05, | |
| "loss": 1.4582, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5363636363636364, | |
| "grad_norm": 1.7602699995040894, | |
| "learning_rate": 3.764354466171652e-05, | |
| "loss": 1.6293, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.5386363636363637, | |
| "grad_norm": 1.8737438917160034, | |
| "learning_rate": 3.725367433049033e-05, | |
| "loss": 1.5442, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.5409090909090909, | |
| "grad_norm": 1.889951467514038, | |
| "learning_rate": 3.6864631090070655e-05, | |
| "loss": 1.3386, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.5431818181818182, | |
| "grad_norm": 2.090937614440918, | |
| "learning_rate": 3.6476440184915815e-05, | |
| "loss": 1.5659, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 1.8041363954544067, | |
| "learning_rate": 3.608912680417737e-05, | |
| "loss": 1.2024, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5477272727272727, | |
| "grad_norm": 2.0270559787750244, | |
| "learning_rate": 3.570271608006555e-05, | |
| "loss": 1.4053, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 2.1177871227264404, | |
| "learning_rate": 3.531723308621847e-05, | |
| "loss": 1.6485, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.5522727272727272, | |
| "grad_norm": 1.9602763652801514, | |
| "learning_rate": 3.493270283607522e-05, | |
| "loss": 1.4901, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.5545454545454546, | |
| "grad_norm": 2.3617732524871826, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 1.9368, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.5568181818181818, | |
| "grad_norm": 3.2929348945617676, | |
| "learning_rate": 3.4166600309926387e-05, | |
| "loss": 1.6233, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5590909090909091, | |
| "grad_norm": 3.4516398906707764, | |
| "learning_rate": 3.3785077745215873e-05, | |
| "loss": 1.5825, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.5613636363636364, | |
| "grad_norm": 3.40051531791687, | |
| "learning_rate": 3.340460734357359e-05, | |
| "loss": 1.8922, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.5636363636363636, | |
| "grad_norm": 4.159510135650635, | |
| "learning_rate": 3.3025213793178646e-05, | |
| "loss": 1.5235, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.5659090909090909, | |
| "grad_norm": 4.598018169403076, | |
| "learning_rate": 3.264692171233485e-05, | |
| "loss": 1.8749, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.5681818181818182, | |
| "grad_norm": 7.766690731048584, | |
| "learning_rate": 3.226975564787322e-05, | |
| "loss": 2.4468, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5681818181818182, | |
| "eval_loss": 1.694659948348999, | |
| "eval_runtime": 57.0659, | |
| "eval_samples_per_second": 12.985, | |
| "eval_steps_per_second": 3.259, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5704545454545454, | |
| "grad_norm": 0.5794286727905273, | |
| "learning_rate": 3.189374007355917e-05, | |
| "loss": 1.9129, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.5727272727272728, | |
| "grad_norm": 0.6516004204750061, | |
| "learning_rate": 3.151889938850445e-05, | |
| "loss": 1.8642, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.575, | |
| "grad_norm": 0.6618034243583679, | |
| "learning_rate": 3.114525791558398e-05, | |
| "loss": 1.8897, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.5772727272727273, | |
| "grad_norm": 0.8190423250198364, | |
| "learning_rate": 3.0772839899857464e-05, | |
| "loss": 1.9876, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.5795454545454546, | |
| "grad_norm": 0.7865899801254272, | |
| "learning_rate": 3.0401669506996256e-05, | |
| "loss": 1.7384, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5818181818181818, | |
| "grad_norm": 0.7890567779541016, | |
| "learning_rate": 3.003177082171523e-05, | |
| "loss": 1.8608, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.5840909090909091, | |
| "grad_norm": 0.6976439356803894, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 1.7961, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.5863636363636363, | |
| "grad_norm": 0.8020011186599731, | |
| "learning_rate": 2.9295884498599414e-05, | |
| "loss": 1.7689, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.5886363636363636, | |
| "grad_norm": 0.7689779996871948, | |
| "learning_rate": 2.8929944611373554e-05, | |
| "loss": 1.8296, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.5909090909090909, | |
| "grad_norm": 0.8270552158355713, | |
| "learning_rate": 2.8565371929847284e-05, | |
| "loss": 1.8922, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5931818181818181, | |
| "grad_norm": 0.8179712891578674, | |
| "learning_rate": 2.8202190110619493e-05, | |
| "loss": 1.5655, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.5954545454545455, | |
| "grad_norm": 0.7286482453346252, | |
| "learning_rate": 2.784042272003794e-05, | |
| "loss": 1.6941, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.5977272727272728, | |
| "grad_norm": 0.7301362156867981, | |
| "learning_rate": 2.7480093232670158e-05, | |
| "loss": 1.6766, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.7769307494163513, | |
| "learning_rate": 2.712122502978024e-05, | |
| "loss": 1.637, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.6022727272727273, | |
| "grad_norm": 0.8360503911972046, | |
| "learning_rate": 2.6763841397811573e-05, | |
| "loss": 1.779, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6045454545454545, | |
| "grad_norm": 0.83577561378479, | |
| "learning_rate": 2.64079655268759e-05, | |
| "loss": 1.6325, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.6068181818181818, | |
| "grad_norm": 0.98736172914505, | |
| "learning_rate": 2.605362050924848e-05, | |
| "loss": 1.6706, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.6090909090909091, | |
| "grad_norm": 0.874660074710846, | |
| "learning_rate": 2.57008293378697e-05, | |
| "loss": 1.6033, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6113636363636363, | |
| "grad_norm": 0.8998573422431946, | |
| "learning_rate": 2.534961490485313e-05, | |
| "loss": 1.6043, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.6136363636363636, | |
| "grad_norm": 0.9336961507797241, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 1.4292, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.615909090909091, | |
| "grad_norm": 0.9164165258407593, | |
| "learning_rate": 2.4652007309320498e-05, | |
| "loss": 1.5082, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.6181818181818182, | |
| "grad_norm": 0.995330810546875, | |
| "learning_rate": 2.430565941356157e-05, | |
| "loss": 1.6186, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.6204545454545455, | |
| "grad_norm": 3.4215948581695557, | |
| "learning_rate": 2.3960978786741877e-05, | |
| "loss": 1.5612, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.6227272727272727, | |
| "grad_norm": 0.937299370765686, | |
| "learning_rate": 2.361798779469336e-05, | |
| "loss": 1.4908, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 1.351189374923706, | |
| "learning_rate": 2.3276708693609943e-05, | |
| "loss": 1.6862, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6272727272727273, | |
| "grad_norm": 1.1487764120101929, | |
| "learning_rate": 2.2937163628603435e-05, | |
| "loss": 1.5817, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.6295454545454545, | |
| "grad_norm": 1.220341444015503, | |
| "learning_rate": 2.259937463226651e-05, | |
| "loss": 1.5732, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.6318181818181818, | |
| "grad_norm": 1.2071298360824585, | |
| "learning_rate": 2.2263363623243054e-05, | |
| "loss": 1.5262, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.634090909090909, | |
| "grad_norm": 1.3270574808120728, | |
| "learning_rate": 2.192915240480596e-05, | |
| "loss": 1.5323, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 1.651341438293457, | |
| "learning_rate": 2.1596762663442218e-05, | |
| "loss": 1.6108, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6386363636363637, | |
| "grad_norm": 1.5319647789001465, | |
| "learning_rate": 2.1266215967445824e-05, | |
| "loss": 1.5921, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.6409090909090909, | |
| "grad_norm": 1.352888822555542, | |
| "learning_rate": 2.0937533765518187e-05, | |
| "loss": 1.7043, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.6431818181818182, | |
| "grad_norm": 1.5980666875839233, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 1.5739, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.6454545454545455, | |
| "grad_norm": 2.0059399604797363, | |
| "learning_rate": 2.0285848032369137e-05, | |
| "loss": 1.4068, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.6477272727272727, | |
| "grad_norm": 1.9330686330795288, | |
| "learning_rate": 1.996288678810105e-05, | |
| "loss": 1.6017, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.5091181993484497, | |
| "learning_rate": 1.9641874609064443e-05, | |
| "loss": 1.6396, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.6522727272727272, | |
| "grad_norm": 1.5990345478057861, | |
| "learning_rate": 1.932283232527956e-05, | |
| "loss": 1.4568, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.6545454545454545, | |
| "grad_norm": 1.9754979610443115, | |
| "learning_rate": 1.9005780638942982e-05, | |
| "loss": 1.4291, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.6568181818181819, | |
| "grad_norm": 1.8010350465774536, | |
| "learning_rate": 1.8690740123084316e-05, | |
| "loss": 1.7167, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.6590909090909091, | |
| "grad_norm": 2.293537139892578, | |
| "learning_rate": 1.837773122023114e-05, | |
| "loss": 1.4793, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6613636363636364, | |
| "grad_norm": 1.9866435527801514, | |
| "learning_rate": 1.8066774241082612e-05, | |
| "loss": 1.7249, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.6636363636363637, | |
| "grad_norm": 1.9991703033447266, | |
| "learning_rate": 1.7757889363191483e-05, | |
| "loss": 1.3803, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.6659090909090909, | |
| "grad_norm": 2.203927755355835, | |
| "learning_rate": 1.745109662965481e-05, | |
| "loss": 1.584, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.6681818181818182, | |
| "grad_norm": 2.0143518447875977, | |
| "learning_rate": 1.714641594781347e-05, | |
| "loss": 1.772, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.6704545454545454, | |
| "grad_norm": 2.390822649002075, | |
| "learning_rate": 1.684386708796025e-05, | |
| "loss": 1.413, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6727272727272727, | |
| "grad_norm": 2.594231128692627, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 1.7485, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 2.5287561416625977, | |
| "learning_rate": 1.62452432224612e-05, | |
| "loss": 1.318, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.6772727272727272, | |
| "grad_norm": 4.2761759757995605, | |
| "learning_rate": 1.5949207060660138e-05, | |
| "loss": 1.9708, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.6795454545454546, | |
| "grad_norm": 9.272795677185059, | |
| "learning_rate": 1.5655380406016235e-05, | |
| "loss": 1.6868, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "grad_norm": 8.554370880126953, | |
| "learning_rate": 1.536378232452003e-05, | |
| "loss": 2.5753, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "eval_loss": 1.6520681381225586, | |
| "eval_runtime": 57.0367, | |
| "eval_samples_per_second": 12.992, | |
| "eval_steps_per_second": 3.261, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6840909090909091, | |
| "grad_norm": 0.3898196816444397, | |
| "learning_rate": 1.5074431737553157e-05, | |
| "loss": 1.9258, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.6863636363636364, | |
| "grad_norm": 0.4636087119579315, | |
| "learning_rate": 1.4787347420660541e-05, | |
| "loss": 1.8041, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.6886363636363636, | |
| "grad_norm": 0.5153092741966248, | |
| "learning_rate": 1.4502548002332088e-05, | |
| "loss": 1.8317, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.6909090909090909, | |
| "grad_norm": 0.5596972107887268, | |
| "learning_rate": 1.422005196279395e-05, | |
| "loss": 1.8383, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.6931818181818182, | |
| "grad_norm": 0.594912588596344, | |
| "learning_rate": 1.3939877632809278e-05, | |
| "loss": 1.7655, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6954545454545454, | |
| "grad_norm": 0.6729958057403564, | |
| "learning_rate": 1.3662043192488849e-05, | |
| "loss": 1.8028, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.6977272727272728, | |
| "grad_norm": 0.7136591672897339, | |
| "learning_rate": 1.338656667011134e-05, | |
| "loss": 1.7534, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.7392699122428894, | |
| "learning_rate": 1.3113465940953495e-05, | |
| "loss": 1.7832, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.7022727272727273, | |
| "grad_norm": 0.7428660988807678, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 1.6537, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.7045454545454546, | |
| "grad_norm": 0.7300150990486145, | |
| "learning_rate": 1.257446259144494e-05, | |
| "loss": 1.8162, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7068181818181818, | |
| "grad_norm": 0.7684038877487183, | |
| "learning_rate": 1.2308594946249163e-05, | |
| "loss": 1.7033, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.7090909090909091, | |
| "grad_norm": 0.8083834052085876, | |
| "learning_rate": 1.204517304231343e-05, | |
| "loss": 1.6916, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.7113636363636363, | |
| "grad_norm": 0.8635165691375732, | |
| "learning_rate": 1.178421397270758e-05, | |
| "loss": 1.6186, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.7136363636363636, | |
| "grad_norm": 0.8064782023429871, | |
| "learning_rate": 1.1525734670691701e-05, | |
| "loss": 1.7021, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.7159090909090909, | |
| "grad_norm": 0.889693021774292, | |
| "learning_rate": 1.1269751908617277e-05, | |
| "loss": 1.7108, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7181818181818181, | |
| "grad_norm": 0.8787115216255188, | |
| "learning_rate": 1.1016282296838887e-05, | |
| "loss": 1.6432, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.7204545454545455, | |
| "grad_norm": 0.8872643113136292, | |
| "learning_rate": 1.0765342282636416e-05, | |
| "loss": 1.8529, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.7227272727272728, | |
| "grad_norm": 0.9299083948135376, | |
| "learning_rate": 1.0516948149147754e-05, | |
| "loss": 1.5756, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.725, | |
| "grad_norm": 0.9978355169296265, | |
| "learning_rate": 1.0271116014312293e-05, | |
| "loss": 1.5607, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.0877996683120728, | |
| "learning_rate": 1.0027861829824952e-05, | |
| "loss": 1.4807, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7295454545454545, | |
| "grad_norm": 1.0490856170654297, | |
| "learning_rate": 9.787201380101157e-06, | |
| "loss": 1.5464, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.7318181818181818, | |
| "grad_norm": 1.1889746189117432, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.559, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.7340909090909091, | |
| "grad_norm": 1.1732399463653564, | |
| "learning_rate": 9.313723980074018e-06, | |
| "loss": 1.796, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.7363636363636363, | |
| "grad_norm": 1.0363281965255737, | |
| "learning_rate": 9.080937753040646e-06, | |
| "loss": 1.6985, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.7386363636363636, | |
| "grad_norm": 0.992416501045227, | |
| "learning_rate": 8.850806705317183e-06, | |
| "loss": 1.2722, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.740909090909091, | |
| "grad_norm": 1.0892759561538696, | |
| "learning_rate": 8.623345769777514e-06, | |
| "loss": 1.5796, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.7431818181818182, | |
| "grad_norm": 1.2749766111373901, | |
| "learning_rate": 8.398569706035792e-06, | |
| "loss": 1.5505, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.7454545454545455, | |
| "grad_norm": 1.1555516719818115, | |
| "learning_rate": 8.176493099488663e-06, | |
| "loss": 1.3703, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.7477272727272727, | |
| "grad_norm": 1.453114628791809, | |
| "learning_rate": 7.957130360368898e-06, | |
| "loss": 1.4758, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.2333486080169678, | |
| "learning_rate": 7.740495722810271e-06, | |
| "loss": 1.6207, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7522727272727273, | |
| "grad_norm": 1.2648553848266602, | |
| "learning_rate": 7.526603243923957e-06, | |
| "loss": 1.586, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.7545454545454545, | |
| "grad_norm": 1.5073436498641968, | |
| "learning_rate": 7.315466802886401e-06, | |
| "loss": 1.3858, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.7568181818181818, | |
| "grad_norm": 1.4304128885269165, | |
| "learning_rate": 7.107100100038671e-06, | |
| "loss": 1.366, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.759090909090909, | |
| "grad_norm": 1.6923023462295532, | |
| "learning_rate": 6.901516655997536e-06, | |
| "loss": 1.6733, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.7613636363636364, | |
| "grad_norm": 1.4528682231903076, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 1.7169, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7636363636363637, | |
| "grad_norm": 1.71442711353302, | |
| "learning_rate": 6.498752722928042e-06, | |
| "loss": 1.6692, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.7659090909090909, | |
| "grad_norm": 1.6300582885742188, | |
| "learning_rate": 6.301598368674105e-06, | |
| "loss": 1.4135, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.7681818181818182, | |
| "grad_norm": 1.709962010383606, | |
| "learning_rate": 6.107279541079769e-06, | |
| "loss": 1.2563, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.7704545454545455, | |
| "grad_norm": 1.762518286705017, | |
| "learning_rate": 5.915808849215304e-06, | |
| "loss": 1.5454, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.7727272727272727, | |
| "grad_norm": 1.9417179822921753, | |
| "learning_rate": 5.727198717339511e-06, | |
| "loss": 1.485, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.775, | |
| "grad_norm": 1.9873510599136353, | |
| "learning_rate": 5.54146138409355e-06, | |
| "loss": 1.5245, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.7772727272727272, | |
| "grad_norm": 1.6412358283996582, | |
| "learning_rate": 5.358608901706802e-06, | |
| "loss": 1.2719, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.7795454545454545, | |
| "grad_norm": 1.994019865989685, | |
| "learning_rate": 5.178653135214812e-06, | |
| "loss": 1.3924, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.7818181818181819, | |
| "grad_norm": 2.177293539047241, | |
| "learning_rate": 5.001605761689398e-06, | |
| "loss": 1.5447, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.7840909090909091, | |
| "grad_norm": 2.519556760787964, | |
| "learning_rate": 4.827478269480895e-06, | |
| "loss": 1.3895, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7863636363636364, | |
| "grad_norm": 3.9570231437683105, | |
| "learning_rate": 4.65628195747273e-06, | |
| "loss": 1.9054, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.7886363636363637, | |
| "grad_norm": 3.9570329189300537, | |
| "learning_rate": 4.488027934348271e-06, | |
| "loss": 1.821, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.7909090909090909, | |
| "grad_norm": 3.8546714782714844, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 1.8442, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.7931818181818182, | |
| "grad_norm": 5.59627103805542, | |
| "learning_rate": 4.16039023417088e-06, | |
| "loss": 2.3136, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.7954545454545454, | |
| "grad_norm": 8.434266090393066, | |
| "learning_rate": 4.001027817058789e-06, | |
| "loss": 2.2889, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7954545454545454, | |
| "eval_loss": 1.6479060649871826, | |
| "eval_runtime": 56.9492, | |
| "eval_samples_per_second": 13.012, | |
| "eval_steps_per_second": 3.266, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7977272727272727, | |
| "grad_norm": 0.3698289394378662, | |
| "learning_rate": 3.844650207332562e-06, | |
| "loss": 1.8769, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.47637778520584106, | |
| "learning_rate": 3.691267552111183e-06, | |
| "loss": 1.9414, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.8022727272727272, | |
| "grad_norm": 0.5020650029182434, | |
| "learning_rate": 3.54088980417534e-06, | |
| "loss": 1.8797, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.8045454545454546, | |
| "grad_norm": 0.5514700412750244, | |
| "learning_rate": 3.393526721321616e-06, | |
| "loss": 1.8749, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.8068181818181818, | |
| "grad_norm": 0.5742571353912354, | |
| "learning_rate": 3.249187865729264e-06, | |
| "loss": 1.7711, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8090909090909091, | |
| "grad_norm": 0.5958347320556641, | |
| "learning_rate": 3.1078826033397843e-06, | |
| "loss": 1.7989, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.8113636363636364, | |
| "grad_norm": 0.5803319811820984, | |
| "learning_rate": 2.9696201032491434e-06, | |
| "loss": 1.5927, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.8136363636363636, | |
| "grad_norm": 0.7168722748756409, | |
| "learning_rate": 2.8344093371128424e-06, | |
| "loss": 1.6328, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.8159090909090909, | |
| "grad_norm": 0.6790624856948853, | |
| "learning_rate": 2.70225907856374e-06, | |
| "loss": 1.6932, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 0.7426336407661438, | |
| "learning_rate": 2.573177902642726e-06, | |
| "loss": 1.859, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8204545454545454, | |
| "grad_norm": 0.7567765712738037, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 1.5385, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.8227272727272728, | |
| "grad_norm": 0.7817090153694153, | |
| "learning_rate": 2.324256102563188e-06, | |
| "loss": 1.6779, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 0.8798091411590576, | |
| "learning_rate": 2.204431630583548e-06, | |
| "loss": 1.5295, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.8272727272727273, | |
| "grad_norm": 0.8708153963088989, | |
| "learning_rate": 2.087708544541689e-06, | |
| "loss": 1.6753, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.8295454545454546, | |
| "grad_norm": 0.8136506080627441, | |
| "learning_rate": 1.974094418431388e-06, | |
| "loss": 1.6369, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8318181818181818, | |
| "grad_norm": 0.906902015209198, | |
| "learning_rate": 1.8635966245104664e-06, | |
| "loss": 1.6952, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.8340909090909091, | |
| "grad_norm": 0.9133580327033997, | |
| "learning_rate": 1.7562223328224325e-06, | |
| "loss": 1.8168, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.8363636363636363, | |
| "grad_norm": 0.9204378724098206, | |
| "learning_rate": 1.6519785107311891e-06, | |
| "loss": 1.7356, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.8386363636363636, | |
| "grad_norm": 0.9639214873313904, | |
| "learning_rate": 1.5508719224689717e-06, | |
| "loss": 1.6889, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.8409090909090909, | |
| "grad_norm": 1.0332034826278687, | |
| "learning_rate": 1.4529091286973995e-06, | |
| "loss": 1.5004, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8431818181818181, | |
| "grad_norm": 0.986321210861206, | |
| "learning_rate": 1.358096486081778e-06, | |
| "loss": 1.6841, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.8454545454545455, | |
| "grad_norm": 1.020476222038269, | |
| "learning_rate": 1.2664401468786114e-06, | |
| "loss": 1.6094, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.8477272727272728, | |
| "grad_norm": 1.0801876783370972, | |
| "learning_rate": 1.1779460585363944e-06, | |
| "loss": 1.5441, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.1377614736557007, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 1.45, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.8522727272727273, | |
| "grad_norm": 1.1650811433792114, | |
| "learning_rate": 1.0104673978866164e-06, | |
| "loss": 1.5581, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8545454545454545, | |
| "grad_norm": 1.099685788154602, | |
| "learning_rate": 9.314936930293283e-07, | |
| "loss": 1.6638, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.8568181818181818, | |
| "grad_norm": 1.6271238327026367, | |
| "learning_rate": 8.557039732283944e-07, | |
| "loss": 1.5905, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.8590909090909091, | |
| "grad_norm": 1.6747556924819946, | |
| "learning_rate": 7.83103156370113e-07, | |
| "loss": 1.6473, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.8613636363636363, | |
| "grad_norm": 1.1051077842712402, | |
| "learning_rate": 7.136959534174592e-07, | |
| "loss": 1.6016, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.8636363636363636, | |
| "grad_norm": 1.4238884449005127, | |
| "learning_rate": 6.474868681043578e-07, | |
| "loss": 1.5379, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.865909090909091, | |
| "grad_norm": 1.5626180171966553, | |
| "learning_rate": 5.844801966434832e-07, | |
| "loss": 1.4908, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.8681818181818182, | |
| "grad_norm": 1.586370825767517, | |
| "learning_rate": 5.246800274474439e-07, | |
| "loss": 1.6281, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.8704545454545455, | |
| "grad_norm": 1.4967386722564697, | |
| "learning_rate": 4.680902408635335e-07, | |
| "loss": 1.3998, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.8727272727272727, | |
| "grad_norm": 1.4333577156066895, | |
| "learning_rate": 4.1471450892189846e-07, | |
| "loss": 1.4366, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 1.4822272062301636, | |
| "learning_rate": 3.6455629509730136e-07, | |
| "loss": 1.4206, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8772727272727273, | |
| "grad_norm": 1.463018536567688, | |
| "learning_rate": 3.1761885408435054e-07, | |
| "loss": 1.286, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.8795454545454545, | |
| "grad_norm": 1.9729580879211426, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 1.7124, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.8818181818181818, | |
| "grad_norm": 1.8522109985351562, | |
| "learning_rate": 2.334182641175686e-07, | |
| "loss": 1.3835, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.884090909090909, | |
| "grad_norm": 1.4895251989364624, | |
| "learning_rate": 1.9616057881935436e-07, | |
| "loss": 1.2827, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.8863636363636364, | |
| "grad_norm": 1.7748496532440186, | |
| "learning_rate": 1.6213459328950352e-07, | |
| "loss": 1.3755, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8886363636363637, | |
| "grad_norm": 1.7273297309875488, | |
| "learning_rate": 1.3134251542544774e-07, | |
| "loss": 1.206, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.8909090909090909, | |
| "grad_norm": 2.185232639312744, | |
| "learning_rate": 1.0378634328099269e-07, | |
| "loss": 1.4596, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.8931818181818182, | |
| "grad_norm": 2.2152626514434814, | |
| "learning_rate": 7.946786493666647e-08, | |
| "loss": 1.5451, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.8954545454545455, | |
| "grad_norm": 2.496037721633911, | |
| "learning_rate": 5.838865838366792e-08, | |
| "loss": 1.6552, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.8977272727272727, | |
| "grad_norm": 2.770740270614624, | |
| "learning_rate": 4.055009142152067e-08, | |
| "loss": 1.5272, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 2.5166406631469727, | |
| "learning_rate": 2.595332156925534e-08, | |
| "loss": 1.8477, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.9022727272727272, | |
| "grad_norm": 3.1613364219665527, | |
| "learning_rate": 1.4599295990352924e-08, | |
| "loss": 1.6615, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.9045454545454545, | |
| "grad_norm": 3.1120591163635254, | |
| "learning_rate": 6.488751431266149e-09, | |
| "loss": 1.4356, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.9068181818181819, | |
| "grad_norm": 4.566768169403076, | |
| "learning_rate": 1.622214173602199e-09, | |
| "loss": 1.6896, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 7.5620036125183105, | |
| "learning_rate": 0.0, | |
| "loss": 2.4552, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 1.6492966413497925, | |
| "eval_runtime": 57.2439, | |
| "eval_samples_per_second": 12.945, | |
| "eval_steps_per_second": 3.249, | |
| "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": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.020304348968387e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |