| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9963369963369964, | |
| "eval_steps": 500, | |
| "global_step": 34, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3534.5521545410156, | |
| "epoch": 0.029304029304029304, | |
| "grad_norm": 0.08098714798688889, | |
| "kl": 0.0, | |
| "learning_rate": 2.5e-07, | |
| "loss": 0.0471, | |
| "reward": 0.2721354260575026, | |
| "reward_std": 0.2681832159869373, | |
| "rewards/accuracy_reward": 0.16666666697710752, | |
| "rewards/tag_count_reward": 0.10546875349245965, | |
| "step": 1 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3753.3802490234375, | |
| "epoch": 0.05860805860805861, | |
| "grad_norm": 0.10224228352308273, | |
| "kl": 0.0, | |
| "learning_rate": 5e-07, | |
| "loss": 0.065, | |
| "reward": 0.1861979232635349, | |
| "reward_std": 0.22471919422969222, | |
| "rewards/accuracy_reward": 0.10937500325962901, | |
| "rewards/tag_count_reward": 0.07682291802484542, | |
| "step": 2 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3401.057403564453, | |
| "epoch": 0.08791208791208792, | |
| "grad_norm": 0.09934523701667786, | |
| "kl": 3.714859485626221e-05, | |
| "learning_rate": 7.5e-07, | |
| "loss": 0.0568, | |
| "reward": 0.2799479253590107, | |
| "reward_std": 0.30056663788855076, | |
| "rewards/accuracy_reward": 0.16145833674818277, | |
| "rewards/tag_count_reward": 0.1184895858168602, | |
| "step": 3 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3252.7084045410156, | |
| "epoch": 0.11721611721611722, | |
| "grad_norm": 0.08650524169206619, | |
| "kl": 3.603845834732056e-05, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0621, | |
| "reward": 0.36328125500585884, | |
| "reward_std": 0.2626346012111753, | |
| "rewards/accuracy_reward": 0.2447916716337204, | |
| "rewards/tag_count_reward": 0.11848958709742874, | |
| "step": 4 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3478.4375915527344, | |
| "epoch": 0.14652014652014653, | |
| "grad_norm": 0.08601976931095123, | |
| "kl": 3.30507755279541e-05, | |
| "learning_rate": 9.975348529157229e-07, | |
| "loss": 0.0597, | |
| "reward": 0.34114584454800934, | |
| "reward_std": 0.2457886270713061, | |
| "rewards/accuracy_reward": 0.23958333721384406, | |
| "rewards/tag_count_reward": 0.10156250407453626, | |
| "step": 5 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3602.2396850585938, | |
| "epoch": 0.17582417582417584, | |
| "grad_norm": 0.09895998239517212, | |
| "kl": 3.193318843841553e-05, | |
| "learning_rate": 9.901664203302124e-07, | |
| "loss": 0.0823, | |
| "reward": 0.24348959093913436, | |
| "reward_std": 0.2597248964011669, | |
| "rewards/accuracy_reward": 0.15625000884756446, | |
| "rewards/tag_count_reward": 0.08723958628252149, | |
| "step": 6 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3152.7813415527344, | |
| "epoch": 0.20512820512820512, | |
| "grad_norm": 0.10855786502361298, | |
| "kl": 2.0995736122131348e-05, | |
| "learning_rate": 9.779754323328192e-07, | |
| "loss": 0.0637, | |
| "reward": 0.4752604365348816, | |
| "reward_std": 0.3382994793355465, | |
| "rewards/accuracy_reward": 0.34375000931322575, | |
| "rewards/tag_count_reward": 0.1315104216337204, | |
| "step": 7 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3384.385498046875, | |
| "epoch": 0.23443223443223443, | |
| "grad_norm": 0.08194123208522797, | |
| "kl": 1.2666918337345123e-05, | |
| "learning_rate": 9.610954559391704e-07, | |
| "loss": 0.04, | |
| "reward": 0.296875006519258, | |
| "reward_std": 0.24468757305294275, | |
| "rewards/accuracy_reward": 0.19270833814516664, | |
| "rewards/tag_count_reward": 0.10416667116805911, | |
| "step": 8 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3482.776123046875, | |
| "epoch": 0.26373626373626374, | |
| "grad_norm": 0.11475110799074173, | |
| "kl": 1.0907649993896484e-05, | |
| "learning_rate": 9.397114317029974e-07, | |
| "loss": 0.1002, | |
| "reward": 0.2395833432674408, | |
| "reward_std": 0.2327865227125585, | |
| "rewards/accuracy_reward": 0.1510416711680591, | |
| "rewards/tag_count_reward": 0.0885416695382446, | |
| "step": 9 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3433.7032165527344, | |
| "epoch": 0.29304029304029305, | |
| "grad_norm": 0.09596528857946396, | |
| "kl": 2.337619662284851e-06, | |
| "learning_rate": 9.140576474687263e-07, | |
| "loss": 0.0698, | |
| "reward": 0.34895834513008595, | |
| "reward_std": 0.2847986426204443, | |
| "rewards/accuracy_reward": 0.23958333861082792, | |
| "rewards/tag_count_reward": 0.10937500442378223, | |
| "step": 10 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3547.1927795410156, | |
| "epoch": 0.32234432234432236, | |
| "grad_norm": 0.09714619070291519, | |
| "kl": 1.394655555486679e-07, | |
| "learning_rate": 8.844151714648274e-07, | |
| "loss": 0.0661, | |
| "reward": 0.3085937525611371, | |
| "reward_std": 0.2829530159942806, | |
| "rewards/accuracy_reward": 0.2083333358168602, | |
| "rewards/tag_count_reward": 0.10026041860692203, | |
| "step": 11 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3510.7396545410156, | |
| "epoch": 0.3516483516483517, | |
| "grad_norm": 0.09303876012563705, | |
| "kl": 3.088265657424927e-05, | |
| "learning_rate": 8.511087728614862e-07, | |
| "loss": 0.0517, | |
| "reward": 0.27604167291428894, | |
| "reward_std": 0.21292753401212394, | |
| "rewards/accuracy_reward": 0.18750000931322575, | |
| "rewards/tag_count_reward": 0.08854166825767606, | |
| "step": 12 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3502.9532165527344, | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 0.10329490154981613, | |
| "kl": 0.00013720989227294922, | |
| "learning_rate": 8.145033635316128e-07, | |
| "loss": 0.0755, | |
| "reward": 0.2122395852347836, | |
| "reward_std": 0.1669893644284457, | |
| "rewards/accuracy_reward": 0.11979167070239782, | |
| "rewards/tag_count_reward": 0.09244791802484542, | |
| "step": 13 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3517.135498046875, | |
| "epoch": 0.41025641025641024, | |
| "grad_norm": 0.11326687037944794, | |
| "kl": 0.00015026330947875977, | |
| "learning_rate": 7.75e-07, | |
| "loss": 0.0726, | |
| "reward": 0.3020833428017795, | |
| "reward_std": 0.2878086185082793, | |
| "rewards/accuracy_reward": 0.19791667303070426, | |
| "rewards/tag_count_reward": 0.10416666837409139, | |
| "step": 14 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3760.557373046875, | |
| "epoch": 0.43956043956043955, | |
| "grad_norm": 0.06679689884185791, | |
| "kl": 0.00015845894813537598, | |
| "learning_rate": 7.330314893841101e-07, | |
| "loss": 0.0487, | |
| "reward": 0.1692708395421505, | |
| "reward_std": 0.19603674625977874, | |
| "rewards/accuracy_reward": 0.10937500465661287, | |
| "rewards/tag_count_reward": 0.05989583348855376, | |
| "step": 15 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3544.5365600585938, | |
| "epoch": 0.46886446886446886, | |
| "grad_norm": 0.08167487382888794, | |
| "kl": 0.00022110342979431152, | |
| "learning_rate": 6.890576474687263e-07, | |
| "loss": 0.0531, | |
| "reward": 0.2864583423361182, | |
| "reward_std": 0.2276854459196329, | |
| "rewards/accuracy_reward": 0.19270833721384406, | |
| "rewards/tag_count_reward": 0.09375000314321369, | |
| "step": 16 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3696.0521850585938, | |
| "epoch": 0.4981684981684982, | |
| "grad_norm": 0.11336914449930191, | |
| "kl": 0.0003203749656677246, | |
| "learning_rate": 6.435602608679916e-07, | |
| "loss": 0.0661, | |
| "reward": 0.2500000058207661, | |
| "reward_std": 0.22962875291705132, | |
| "rewards/accuracy_reward": 0.1666666679084301, | |
| "rewards/tag_count_reward": 0.08333333558402956, | |
| "step": 17 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3608.9584045410156, | |
| "epoch": 0.5274725274725275, | |
| "grad_norm": 0.08060825616121292, | |
| "kl": 0.0004144906997680664, | |
| "learning_rate": 5.97037808470444e-07, | |
| "loss": 0.0568, | |
| "reward": 0.3007812607102096, | |
| "reward_std": 0.2574492711573839, | |
| "rewards/accuracy_reward": 0.2135416711680591, | |
| "rewards/tag_count_reward": 0.08723958604969084, | |
| "step": 18 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3601.6927795410156, | |
| "epoch": 0.5567765567765568, | |
| "grad_norm": 0.07288842648267746, | |
| "kl": 0.0003364086151123047, | |
| "learning_rate": 5.5e-07, | |
| "loss": 0.0319, | |
| "reward": 0.376302097691223, | |
| "reward_std": 0.2284244946204126, | |
| "rewards/accuracy_reward": 0.281250003259629, | |
| "rewards/tag_count_reward": 0.09505208558402956, | |
| "step": 19 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3625.0208740234375, | |
| "epoch": 0.5860805860805861, | |
| "grad_norm": 0.07371893525123596, | |
| "kl": 0.0006115436553955078, | |
| "learning_rate": 5.02962191529556e-07, | |
| "loss": 0.0335, | |
| "reward": 0.2552083439659327, | |
| "reward_std": 0.25146263325586915, | |
| "rewards/accuracy_reward": 0.17187500791624188, | |
| "rewards/tag_count_reward": 0.08333333511836827, | |
| "step": 20 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3542.1354370117188, | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.10749542713165283, | |
| "kl": 0.00035321712493896484, | |
| "learning_rate": 4.5643973913200837e-07, | |
| "loss": 0.0995, | |
| "reward": 0.373697929084301, | |
| "reward_std": 0.31743163615465164, | |
| "rewards/accuracy_reward": 0.281250006519258, | |
| "rewards/tag_count_reward": 0.09244791977107525, | |
| "step": 21 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3414.6927795410156, | |
| "epoch": 0.6446886446886447, | |
| "grad_norm": 0.08405382186174393, | |
| "kl": 0.00036454200744628906, | |
| "learning_rate": 4.1094235253127374e-07, | |
| "loss": 0.075, | |
| "reward": 0.37760418374091387, | |
| "reward_std": 0.2682782169431448, | |
| "rewards/accuracy_reward": 0.2708333353511989, | |
| "rewards/tag_count_reward": 0.1067708374466747, | |
| "step": 22 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3678.791748046875, | |
| "epoch": 0.673992673992674, | |
| "grad_norm": 0.07961313426494598, | |
| "kl": 0.0004169940948486328, | |
| "learning_rate": 3.6696851061588994e-07, | |
| "loss": 0.0502, | |
| "reward": 0.2786458395421505, | |
| "reward_std": 0.24647203274071217, | |
| "rewards/accuracy_reward": 0.1979166711680591, | |
| "rewards/tag_count_reward": 0.08072916697710752, | |
| "step": 23 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3838.8334045410156, | |
| "epoch": 0.7032967032967034, | |
| "grad_norm": 0.06324659287929535, | |
| "kl": 0.0005202293395996094, | |
| "learning_rate": 3.250000000000001e-07, | |
| "loss": 0.0313, | |
| "reward": 0.13541667081881315, | |
| "reward_std": 0.16657467489130795, | |
| "rewards/accuracy_reward": 0.08854167023673654, | |
| "rewards/tag_count_reward": 0.04687500209547579, | |
| "step": 24 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3507.5677490234375, | |
| "epoch": 0.7326007326007326, | |
| "grad_norm": 0.085300974547863, | |
| "kl": 0.0008640289306640625, | |
| "learning_rate": 2.854966364683872e-07, | |
| "loss": 0.0561, | |
| "reward": 0.29427084082271904, | |
| "reward_std": 0.19792527868412435, | |
| "rewards/accuracy_reward": 0.20833333488553762, | |
| "rewards/tag_count_reward": 0.0859375026775524, | |
| "step": 25 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3683.151123046875, | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 0.06933189183473587, | |
| "kl": 0.0010666847229003906, | |
| "learning_rate": 2.488912271385139e-07, | |
| "loss": 0.0326, | |
| "reward": 0.272135422565043, | |
| "reward_std": 0.2521828315220773, | |
| "rewards/accuracy_reward": 0.1927083353511989, | |
| "rewards/tag_count_reward": 0.0794270858168602, | |
| "step": 26 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3190.5000610351562, | |
| "epoch": 0.7912087912087912, | |
| "grad_norm": 0.12118662893772125, | |
| "kl": 0.0013322830200195312, | |
| "learning_rate": 2.1558482853517253e-07, | |
| "loss": 0.0751, | |
| "reward": 0.40755209885537624, | |
| "reward_std": 0.31804133765399456, | |
| "rewards/accuracy_reward": 0.2604166753590107, | |
| "rewards/tag_count_reward": 0.1471354179084301, | |
| "step": 27 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3750.2240600585938, | |
| "epoch": 0.8205128205128205, | |
| "grad_norm": 0.07822524011135101, | |
| "kl": 0.0015530586242675781, | |
| "learning_rate": 1.8594235253127372e-07, | |
| "loss": 0.0456, | |
| "reward": 0.2773437611758709, | |
| "reward_std": 0.25566962361335754, | |
| "rewards/accuracy_reward": 0.1927083395421505, | |
| "rewards/tag_count_reward": 0.08463541837409139, | |
| "step": 28 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3394.760498046875, | |
| "epoch": 0.8498168498168498, | |
| "grad_norm": 0.12853899598121643, | |
| "kl": 0.0016031265258789062, | |
| "learning_rate": 1.6028856829700258e-07, | |
| "loss": 0.0882, | |
| "reward": 0.43750001303851604, | |
| "reward_std": 0.33361043129116297, | |
| "rewards/accuracy_reward": 0.31770833767950535, | |
| "rewards/tag_count_reward": 0.1197916716337204, | |
| "step": 29 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3737.7344360351562, | |
| "epoch": 0.8791208791208791, | |
| "grad_norm": 0.07847777009010315, | |
| "kl": 0.0011386871337890625, | |
| "learning_rate": 1.3890454406082956e-07, | |
| "loss": 0.0195, | |
| "reward": 0.24479167815297842, | |
| "reward_std": 0.23484968207776546, | |
| "rewards/accuracy_reward": 0.17187500931322575, | |
| "rewards/tag_count_reward": 0.07291666860692203, | |
| "step": 30 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3828.666778564453, | |
| "epoch": 0.9084249084249084, | |
| "grad_norm": 0.06969457864761353, | |
| "kl": 0.0012054443359375, | |
| "learning_rate": 1.220245676671809e-07, | |
| "loss": 0.0396, | |
| "reward": 0.17578125256113708, | |
| "reward_std": 0.19791242526844144, | |
| "rewards/accuracy_reward": 0.12500000558793545, | |
| "rewards/tag_count_reward": 0.050781250931322575, | |
| "step": 31 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3739.2969360351562, | |
| "epoch": 0.9377289377289377, | |
| "grad_norm": 0.08980078995227814, | |
| "kl": 0.001861572265625, | |
| "learning_rate": 1.0983357966978745e-07, | |
| "loss": 0.0435, | |
| "reward": 0.28906251152511686, | |
| "reward_std": 0.2909349126275629, | |
| "rewards/accuracy_reward": 0.20312500838190317, | |
| "rewards/tag_count_reward": 0.08593750221189111, | |
| "step": 32 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3494.8907470703125, | |
| "epoch": 0.967032967032967, | |
| "grad_norm": 0.16050584614276886, | |
| "kl": 0.00262451171875, | |
| "learning_rate": 1.0246514708427701e-07, | |
| "loss": 0.1098, | |
| "reward": 0.281250006868504, | |
| "reward_std": 0.2926723880227655, | |
| "rewards/accuracy_reward": 0.17187500465661287, | |
| "rewards/tag_count_reward": 0.10937500221189111, | |
| "step": 33 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 3986.875, | |
| "epoch": 0.9963369963369964, | |
| "grad_norm": 0.08669467270374298, | |
| "kl": 0.0013146400451660156, | |
| "learning_rate": 1e-07, | |
| "loss": 0.0431, | |
| "reward": 0.2903645932674408, | |
| "reward_std": 0.3158208169043064, | |
| "rewards/accuracy_reward": 0.20312500838190317, | |
| "rewards/tag_count_reward": 0.0872395858168602, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.9963369963369964, | |
| "step": 34, | |
| "total_flos": 0.0, | |
| "train_loss": 0.0, | |
| "train_runtime": 3.2357, | |
| "train_samples_per_second": 252.496, | |
| "train_steps_per_second": 10.508 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 34, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |