| { | |
| "best_metric": 2.0210015773773193, | |
| "best_model_checkpoint": "./whisper-LoRA-small-ha/checkpoint-2000", | |
| "epoch": 14.331210191082803, | |
| "eval_steps": 250, | |
| "global_step": 2250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1592356687898089, | |
| "grad_norm": 10.639625549316406, | |
| "learning_rate": 1.3800000000000002e-05, | |
| "loss": 5.8923, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3184713375796178, | |
| "grad_norm": 3.0570621490478516, | |
| "learning_rate": 2.88e-05, | |
| "loss": 5.1217, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.47770700636942676, | |
| "grad_norm": 1.420814037322998, | |
| "learning_rate": 3e-05, | |
| "loss": 4.1046, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "grad_norm": 1.6677359342575073, | |
| "learning_rate": 3e-05, | |
| "loss": 3.5633, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7961783439490446, | |
| "grad_norm": 1.1571909189224243, | |
| "learning_rate": 3e-05, | |
| "loss": 3.2634, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "grad_norm": 1.4416701793670654, | |
| "learning_rate": 3e-05, | |
| "loss": 3.2152, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.1146496815286624, | |
| "grad_norm": 1.3764861822128296, | |
| "learning_rate": 3e-05, | |
| "loss": 3.1248, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2738853503184713, | |
| "grad_norm": 1.537541389465332, | |
| "learning_rate": 3e-05, | |
| "loss": 3.0425, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.4331210191082802, | |
| "grad_norm": 1.239444613456726, | |
| "learning_rate": 3e-05, | |
| "loss": 3.0277, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.5923566878980893, | |
| "grad_norm": 1.4676295518875122, | |
| "learning_rate": 3e-05, | |
| "loss": 3.0616, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.5923566878980893, | |
| "eval_loss": 2.9989981651306152, | |
| "eval_runtime": 130.5989, | |
| "eval_samples_per_second": 5.054, | |
| "eval_steps_per_second": 0.636, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.7515923566878981, | |
| "grad_norm": 1.5393909215927124, | |
| "learning_rate": 3e-05, | |
| "loss": 2.9165, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.910828025477707, | |
| "grad_norm": 1.4913767576217651, | |
| "learning_rate": 3e-05, | |
| "loss": 2.8888, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.070063694267516, | |
| "grad_norm": 1.8603297472000122, | |
| "learning_rate": 3e-05, | |
| "loss": 2.8715, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.229299363057325, | |
| "grad_norm": 1.5909651517868042, | |
| "learning_rate": 3e-05, | |
| "loss": 2.7833, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.388535031847134, | |
| "grad_norm": 33.26897048950195, | |
| "learning_rate": 3e-05, | |
| "loss": 2.6853, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.5477707006369426, | |
| "grad_norm": 1.984903335571289, | |
| "learning_rate": 3e-05, | |
| "loss": 2.7053, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.7070063694267517, | |
| "grad_norm": 1.6659773588180542, | |
| "learning_rate": 3e-05, | |
| "loss": 2.6967, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.8662420382165603, | |
| "grad_norm": 2.684084415435791, | |
| "learning_rate": 3e-05, | |
| "loss": 2.63, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.0254777070063694, | |
| "grad_norm": 1.7889983654022217, | |
| "learning_rate": 3e-05, | |
| "loss": 2.1224, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 3.1847133757961785, | |
| "grad_norm": 2.101982831954956, | |
| "learning_rate": 3e-05, | |
| "loss": 2.1053, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.1847133757961785, | |
| "eval_loss": 2.131725311279297, | |
| "eval_runtime": 131.015, | |
| "eval_samples_per_second": 5.038, | |
| "eval_steps_per_second": 0.634, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.343949044585987, | |
| "grad_norm": 1.9597238302230835, | |
| "learning_rate": 3e-05, | |
| "loss": 2.1066, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 3.5031847133757963, | |
| "grad_norm": 1.7918522357940674, | |
| "learning_rate": 3e-05, | |
| "loss": 2.043, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.662420382165605, | |
| "grad_norm": 1.6857051849365234, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9991, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 3.821656050955414, | |
| "grad_norm": 1.8661901950836182, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0209, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.980891719745223, | |
| "grad_norm": 2.1587560176849365, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0404, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 4.140127388535032, | |
| "grad_norm": 1.7673060894012451, | |
| "learning_rate": 3e-05, | |
| "loss": 2.052, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.2993630573248405, | |
| "grad_norm": 1.8025519847869873, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9958, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 4.45859872611465, | |
| "grad_norm": 2.0931341648101807, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9864, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.617834394904459, | |
| "grad_norm": 1.7860947847366333, | |
| "learning_rate": 3e-05, | |
| "loss": 1.998, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 4.777070063694268, | |
| "grad_norm": 2.118542432785034, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0123, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.777070063694268, | |
| "eval_loss": 2.090299606323242, | |
| "eval_runtime": 131.9058, | |
| "eval_samples_per_second": 5.004, | |
| "eval_steps_per_second": 0.629, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.936305732484076, | |
| "grad_norm": 2.0395212173461914, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0432, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 5.095541401273885, | |
| "grad_norm": 2.1469058990478516, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9556, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.254777070063694, | |
| "grad_norm": 2.0194575786590576, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9874, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 5.414012738853503, | |
| "grad_norm": 2.0585408210754395, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9637, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.573248407643312, | |
| "grad_norm": 1.9950058460235596, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9731, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 5.732484076433121, | |
| "grad_norm": 2.102336883544922, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9443, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.89171974522293, | |
| "grad_norm": 1.905841588973999, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9753, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 6.050955414012739, | |
| "grad_norm": 2.1166951656341553, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9176, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 6.210191082802548, | |
| "grad_norm": 2.2197279930114746, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8839, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 6.369426751592357, | |
| "grad_norm": 2.578101873397827, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9218, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.369426751592357, | |
| "eval_loss": 2.0661463737487793, | |
| "eval_runtime": 131.3996, | |
| "eval_samples_per_second": 5.023, | |
| "eval_steps_per_second": 0.632, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.528662420382165, | |
| "grad_norm": 2.2647619247436523, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9609, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 6.687898089171974, | |
| "grad_norm": 2.4119601249694824, | |
| "learning_rate": 3e-05, | |
| "loss": 1.961, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 6.8471337579617835, | |
| "grad_norm": 2.380333662033081, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9162, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 7.006369426751593, | |
| "grad_norm": 2.429640293121338, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9468, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 7.165605095541402, | |
| "grad_norm": 2.362521171569824, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9068, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 7.32484076433121, | |
| "grad_norm": 2.479595899581909, | |
| "learning_rate": 3e-05, | |
| "loss": 1.908, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 7.484076433121019, | |
| "grad_norm": 2.53534197807312, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8769, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 7.643312101910828, | |
| "grad_norm": 2.207059144973755, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8427, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 7.802547770700637, | |
| "grad_norm": 2.4016318321228027, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9285, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 7.961783439490446, | |
| "grad_norm": 2.836793899536133, | |
| "learning_rate": 3e-05, | |
| "loss": 1.894, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 7.961783439490446, | |
| "eval_loss": 2.0512537956237793, | |
| "eval_runtime": 131.0055, | |
| "eval_samples_per_second": 5.038, | |
| "eval_steps_per_second": 0.634, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 8.121019108280255, | |
| "grad_norm": 2.402125358581543, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8639, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 8.280254777070065, | |
| "grad_norm": 2.61692214012146, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8427, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 8.439490445859873, | |
| "grad_norm": 2.6917548179626465, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8541, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 8.598726114649681, | |
| "grad_norm": 3.4319992065429688, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8926, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 8.757961783439491, | |
| "grad_norm": 2.572566032409668, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7997, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 8.9171974522293, | |
| "grad_norm": 2.649146795272827, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8534, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 9.07643312101911, | |
| "grad_norm": 2.9375128746032715, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8279, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 9.235668789808917, | |
| "grad_norm": 2.6020314693450928, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9072, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 9.394904458598726, | |
| "grad_norm": 2.7787463665008545, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8113, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 9.554140127388536, | |
| "grad_norm": 2.8579261302948, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7961, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.554140127388536, | |
| "eval_loss": 2.034783363342285, | |
| "eval_runtime": 131.7061, | |
| "eval_samples_per_second": 5.011, | |
| "eval_steps_per_second": 0.63, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.713375796178344, | |
| "grad_norm": 2.863452911376953, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7903, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 9.872611464968152, | |
| "grad_norm": 3.1351847648620605, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7899, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 10.031847133757962, | |
| "grad_norm": 2.9801595211029053, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8365, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 10.19108280254777, | |
| "grad_norm": 2.9951841831207275, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7694, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 10.35031847133758, | |
| "grad_norm": 3.2015936374664307, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7946, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 10.509554140127388, | |
| "grad_norm": 2.942176103591919, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7126, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 10.668789808917197, | |
| "grad_norm": 3.4911484718322754, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7643, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 10.828025477707007, | |
| "grad_norm": 3.2129969596862793, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8453, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 10.987261146496815, | |
| "grad_norm": 2.9529011249542236, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7673, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 11.146496815286625, | |
| "grad_norm": 2.8384742736816406, | |
| "learning_rate": 3e-05, | |
| "loss": 1.764, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 11.146496815286625, | |
| "eval_loss": 2.0300416946411133, | |
| "eval_runtime": 132.059, | |
| "eval_samples_per_second": 4.998, | |
| "eval_steps_per_second": 0.629, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 11.305732484076433, | |
| "grad_norm": 3.514625072479248, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7332, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 11.464968152866241, | |
| "grad_norm": 3.1534969806671143, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8226, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 11.624203821656051, | |
| "grad_norm": 3.1041505336761475, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7624, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 11.78343949044586, | |
| "grad_norm": 3.0769126415252686, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7711, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 11.94267515923567, | |
| "grad_norm": 3.060081958770752, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6919, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 12.101910828025478, | |
| "grad_norm": 3.662893056869507, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7149, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 12.261146496815286, | |
| "grad_norm": 3.292971134185791, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6796, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 12.420382165605096, | |
| "grad_norm": 3.3103177547454834, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7545, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 12.579617834394904, | |
| "grad_norm": 3.5080769062042236, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7648, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 12.738853503184714, | |
| "grad_norm": 3.3745033740997314, | |
| "learning_rate": 3e-05, | |
| "loss": 1.748, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 12.738853503184714, | |
| "eval_loss": 2.0210015773773193, | |
| "eval_runtime": 131.5955, | |
| "eval_samples_per_second": 5.015, | |
| "eval_steps_per_second": 0.631, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 12.898089171974522, | |
| "grad_norm": 3.3616280555725098, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6712, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 13.05732484076433, | |
| "grad_norm": 3.6867527961730957, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6881, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 13.21656050955414, | |
| "grad_norm": 3.1726772785186768, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6976, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 13.375796178343949, | |
| "grad_norm": 3.53255558013916, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7178, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 13.535031847133759, | |
| "grad_norm": 3.7484970092773438, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6847, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 13.694267515923567, | |
| "grad_norm": 3.8939905166625977, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6349, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 13.853503184713375, | |
| "grad_norm": 3.3592910766601562, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7276, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 14.012738853503185, | |
| "grad_norm": 3.772204875946045, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6284, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 14.171974522292993, | |
| "grad_norm": 4.016607761383057, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6573, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 14.331210191082803, | |
| "grad_norm": 3.5693182945251465, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6554, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 14.331210191082803, | |
| "eval_loss": 2.0353081226348877, | |
| "eval_runtime": 133.6329, | |
| "eval_samples_per_second": 4.939, | |
| "eval_steps_per_second": 0.621, | |
| "step": 2250 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 3140, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 250, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.055608710561792e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |