| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 10, | |
| "global_step": 66, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.046511627906976744, | |
| "grad_norm": 3.328749895095825, | |
| "learning_rate": 0.0, | |
| "loss": 4.5983, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.09302325581395349, | |
| "grad_norm": 3.354735851287842, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 4.5836, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.13953488372093023, | |
| "grad_norm": 2.9163668155670166, | |
| "learning_rate": 5.714285714285714e-05, | |
| "loss": 4.397, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.18604651162790697, | |
| "grad_norm": 2.2333481311798096, | |
| "learning_rate": 8.571428571428571e-05, | |
| "loss": 4.1457, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.23255813953488372, | |
| "grad_norm": 1.991350769996643, | |
| "learning_rate": 0.00011428571428571428, | |
| "loss": 3.9002, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.27906976744186046, | |
| "grad_norm": 1.85317063331604, | |
| "learning_rate": 0.00014285714285714287, | |
| "loss": 3.5969, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.32558139534883723, | |
| "grad_norm": 1.9501370191574097, | |
| "learning_rate": 0.00017142857142857143, | |
| "loss": 3.2651, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.37209302325581395, | |
| "grad_norm": 2.216587543487549, | |
| "learning_rate": 0.0002, | |
| "loss": 2.8473, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.4186046511627907, | |
| "grad_norm": 2.0312318801879883, | |
| "learning_rate": 0.0001998582695676762, | |
| "loss": 2.3923, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.46511627906976744, | |
| "grad_norm": 1.8019684553146362, | |
| "learning_rate": 0.00019943348002101371, | |
| "loss": 2.0528, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.46511627906976744, | |
| "eval_loss": 1.894411325454712, | |
| "eval_runtime": 145.4733, | |
| "eval_samples_per_second": 33.601, | |
| "eval_steps_per_second": 0.137, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.5116279069767442, | |
| "grad_norm": 1.951332449913025, | |
| "learning_rate": 0.00019872683547213446, | |
| "loss": 1.7962, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.5581395348837209, | |
| "grad_norm": 1.5264594554901123, | |
| "learning_rate": 0.00019774033898178667, | |
| "loss": 1.5795, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.6046511627906976, | |
| "grad_norm": 1.0167055130004883, | |
| "learning_rate": 0.0001964767868814516, | |
| "loss": 1.4307, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.6511627906976745, | |
| "grad_norm": 1.1928244829177856, | |
| "learning_rate": 0.00019493976084683813, | |
| "loss": 1.3048, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.6976744186046512, | |
| "grad_norm": 1.513029932975769, | |
| "learning_rate": 0.00019313361774523385, | |
| "loss": 1.1708, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.7441860465116279, | |
| "grad_norm": 1.6432111263275146, | |
| "learning_rate": 0.00019106347728549135, | |
| "loss": 1.0173, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.7906976744186046, | |
| "grad_norm": 1.4672132730484009, | |
| "learning_rate": 0.00018873520750565718, | |
| "loss": 0.8399, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.8372093023255814, | |
| "grad_norm": 1.3380573987960815, | |
| "learning_rate": 0.0001861554081393806, | |
| "loss": 0.6508, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.8837209302325582, | |
| "grad_norm": 1.26139235496521, | |
| "learning_rate": 0.0001833313919082515, | |
| "loss": 0.4701, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "grad_norm": 1.0803511142730713, | |
| "learning_rate": 0.00018027116379309638, | |
| "loss": 0.316, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "eval_loss": 0.19302108883857727, | |
| "eval_runtime": 145.4827, | |
| "eval_samples_per_second": 33.598, | |
| "eval_steps_per_second": 0.137, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.9767441860465116, | |
| "grad_norm": 0.8388907313346863, | |
| "learning_rate": 0.00017698339834299061, | |
| "loss": 0.1865, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.6994265913963318, | |
| "learning_rate": 0.00017347741508630672, | |
| "loss": 0.1083, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 1.0465116279069768, | |
| "grad_norm": 0.266382098197937, | |
| "learning_rate": 0.0001697631521134985, | |
| "loss": 0.0619, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 1.0930232558139534, | |
| "grad_norm": 0.17738769948482513, | |
| "learning_rate": 0.00016585113790650388, | |
| "loss": 0.0463, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.1395348837209303, | |
| "grad_norm": 0.12172795832157135, | |
| "learning_rate": 0.0001617524614946192, | |
| "loss": 0.039, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.1860465116279069, | |
| "grad_norm": 0.1264028549194336, | |
| "learning_rate": 0.0001574787410214407, | |
| "loss": 0.0372, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.2325581395348837, | |
| "grad_norm": 0.108629010617733, | |
| "learning_rate": 0.00015304209081197425, | |
| "loss": 0.0332, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.2790697674418605, | |
| "grad_norm": 0.0759243443608284, | |
| "learning_rate": 0.00014845508703326504, | |
| "loss": 0.0323, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.3255813953488373, | |
| "grad_norm": 0.06167895719408989, | |
| "learning_rate": 0.00014373073204588556, | |
| "loss": 0.0292, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.372093023255814, | |
| "grad_norm": 0.05063502490520477, | |
| "learning_rate": 0.00013888241754733208, | |
| "loss": 0.0322, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.372093023255814, | |
| "eval_loss": 0.03178785368800163, | |
| "eval_runtime": 144.6511, | |
| "eval_samples_per_second": 33.792, | |
| "eval_steps_per_second": 0.138, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.4186046511627908, | |
| "grad_norm": 0.05390379950404167, | |
| "learning_rate": 0.00013392388661180303, | |
| "loss": 0.0317, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.4651162790697674, | |
| "grad_norm": 0.05356408655643463, | |
| "learning_rate": 0.0001288691947339621, | |
| "loss": 0.0261, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.5116279069767442, | |
| "grad_norm": 0.039639439433813095, | |
| "learning_rate": 0.0001237326699871115, | |
| "loss": 0.0286, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.558139534883721, | |
| "grad_norm": 0.03714418411254883, | |
| "learning_rate": 0.00011852887240871145, | |
| "loss": 0.0279, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.6046511627906976, | |
| "grad_norm": 0.04030028358101845, | |
| "learning_rate": 0.00011327255272837221, | |
| "loss": 0.0265, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.6511627906976745, | |
| "grad_norm": 0.035650961101055145, | |
| "learning_rate": 0.00010797861055530831, | |
| "loss": 0.0253, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.697674418604651, | |
| "grad_norm": 0.041386678814888, | |
| "learning_rate": 0.00010266205214377748, | |
| "loss": 0.0261, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.744186046511628, | |
| "grad_norm": 0.03318718075752258, | |
| "learning_rate": 9.733794785622253e-05, | |
| "loss": 0.0253, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.7906976744186047, | |
| "grad_norm": 0.03519133850932121, | |
| "learning_rate": 9.202138944469168e-05, | |
| "loss": 0.0259, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.8372093023255816, | |
| "grad_norm": 0.03141423687338829, | |
| "learning_rate": 8.672744727162781e-05, | |
| "loss": 0.0229, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.8372093023255816, | |
| "eval_loss": 0.028911028057336807, | |
| "eval_runtime": 145.5272, | |
| "eval_samples_per_second": 33.588, | |
| "eval_steps_per_second": 0.137, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.8837209302325582, | |
| "grad_norm": 0.04335688054561615, | |
| "learning_rate": 8.147112759128859e-05, | |
| "loss": 0.0246, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.9302325581395348, | |
| "grad_norm": 0.031896017491817474, | |
| "learning_rate": 7.626733001288851e-05, | |
| "loss": 0.0235, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.9767441860465116, | |
| "grad_norm": 0.03081543557345867, | |
| "learning_rate": 7.113080526603792e-05, | |
| "loss": 0.0247, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.04796084016561508, | |
| "learning_rate": 6.607611338819697e-05, | |
| "loss": 0.0266, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 2.046511627906977, | |
| "grad_norm": 0.032496869564056396, | |
| "learning_rate": 6.111758245266794e-05, | |
| "loss": 0.0217, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.0930232558139537, | |
| "grad_norm": 0.03214440867304802, | |
| "learning_rate": 5.626926795411447e-05, | |
| "loss": 0.022, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 2.13953488372093, | |
| "grad_norm": 0.026299171149730682, | |
| "learning_rate": 5.1544912966734994e-05, | |
| "loss": 0.0206, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 2.186046511627907, | |
| "grad_norm": 0.02710675820708275, | |
| "learning_rate": 4.695790918802576e-05, | |
| "loss": 0.0241, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 2.2325581395348837, | |
| "grad_norm": 0.027937039732933044, | |
| "learning_rate": 4.252125897855932e-05, | |
| "loss": 0.0213, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 2.2790697674418605, | |
| "grad_norm": 0.03069477155804634, | |
| "learning_rate": 3.824753850538082e-05, | |
| "loss": 0.0206, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.2790697674418605, | |
| "eval_loss": 0.02796892449259758, | |
| "eval_runtime": 144.1386, | |
| "eval_samples_per_second": 33.912, | |
| "eval_steps_per_second": 0.139, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.3255813953488373, | |
| "grad_norm": 0.027339283376932144, | |
| "learning_rate": 3.414886209349615e-05, | |
| "loss": 0.0202, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.3720930232558137, | |
| "grad_norm": 0.02695722132921219, | |
| "learning_rate": 3.0236847886501542e-05, | |
| "loss": 0.0214, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.4186046511627906, | |
| "grad_norm": 0.022607291117310524, | |
| "learning_rate": 2.6522584913693294e-05, | |
| "loss": 0.0202, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.4651162790697674, | |
| "grad_norm": 0.02922969125211239, | |
| "learning_rate": 2.301660165700936e-05, | |
| "loss": 0.0225, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.511627906976744, | |
| "grad_norm": 0.02776559814810753, | |
| "learning_rate": 1.9728836206903656e-05, | |
| "loss": 0.0188, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.558139534883721, | |
| "grad_norm": 0.027599385008215904, | |
| "learning_rate": 1.6668608091748495e-05, | |
| "loss": 0.0215, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.604651162790698, | |
| "grad_norm": 0.02333148941397667, | |
| "learning_rate": 1.3844591860619383e-05, | |
| "loss": 0.0197, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.6511627906976747, | |
| "grad_norm": 0.02748979814350605, | |
| "learning_rate": 1.1264792494342857e-05, | |
| "loss": 0.023, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.697674418604651, | |
| "grad_norm": 0.023095615208148956, | |
| "learning_rate": 8.936522714508678e-06, | |
| "loss": 0.021, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.744186046511628, | |
| "grad_norm": 0.030044227838516235, | |
| "learning_rate": 6.866382254766157e-06, | |
| "loss": 0.0238, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.744186046511628, | |
| "eval_loss": 0.027968447655439377, | |
| "eval_runtime": 144.4837, | |
| "eval_samples_per_second": 33.831, | |
| "eval_steps_per_second": 0.138, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.7906976744186047, | |
| "grad_norm": 0.026822634041309357, | |
| "learning_rate": 5.060239153161872e-06, | |
| "loss": 0.022, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.8372093023255816, | |
| "grad_norm": 0.02589753456413746, | |
| "learning_rate": 3.5232131185484076e-06, | |
| "loss": 0.0193, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.883720930232558, | |
| "grad_norm": 0.025665169581770897, | |
| "learning_rate": 2.259661018213333e-06, | |
| "loss": 0.0211, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.9302325581395348, | |
| "grad_norm": 0.025219907984137535, | |
| "learning_rate": 1.2731645278655445e-06, | |
| "loss": 0.0205, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.9767441860465116, | |
| "grad_norm": 0.02233767695724964, | |
| "learning_rate": 5.665199789862907e-07, | |
| "loss": 0.0211, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.03987536579370499, | |
| "learning_rate": 1.4173043232380557e-07, | |
| "loss": 0.0221, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 66, | |
| "total_flos": 4.4321989464330076e+18, | |
| "train_loss": 0.7241760994674582, | |
| "train_runtime": 13313.2011, | |
| "train_samples_per_second": 9.912, | |
| "train_steps_per_second": 0.005 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 66, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.4321989464330076e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |