| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 396, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00506409241968666, | |
| "grad_norm": 5.678173069867947, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3528, | |
| "mean_token_accuracy": 0.9147689286619425, | |
| "num_tokens": 586487.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.05064092419686659, | |
| "grad_norm": 0.3134267771799764, | |
| "learning_rate": 0.0001, | |
| "loss": 0.2555, | |
| "mean_token_accuracy": 0.9199666447109647, | |
| "num_tokens": 5968280.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10128184839373318, | |
| "grad_norm": 0.14841672886560275, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1589, | |
| "mean_token_accuracy": 0.9397787630558014, | |
| "num_tokens": 11930513.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.15192277259059978, | |
| "grad_norm": 0.0924952729966239, | |
| "learning_rate": 0.0001, | |
| "loss": 0.139, | |
| "mean_token_accuracy": 0.946698647364974, | |
| "num_tokens": 17853273.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.20256369678746636, | |
| "grad_norm": 0.08634084330086704, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1277, | |
| "mean_token_accuracy": 0.95089195612818, | |
| "num_tokens": 23792525.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.25320462098433294, | |
| "grad_norm": 0.07548012506474341, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1212, | |
| "mean_token_accuracy": 0.953272457793355, | |
| "num_tokens": 29710036.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.30384554518119955, | |
| "grad_norm": 0.07448026254865987, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1151, | |
| "mean_token_accuracy": 0.9557252813130617, | |
| "num_tokens": 35685902.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.35448646937806616, | |
| "grad_norm": 0.0879124617580733, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1112, | |
| "mean_token_accuracy": 0.957416345924139, | |
| "num_tokens": 41642058.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4051273935749327, | |
| "grad_norm": 0.07273411152988447, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1098, | |
| "mean_token_accuracy": 0.9579035054892302, | |
| "num_tokens": 47564730.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.4557683177717993, | |
| "grad_norm": 0.07959899342781425, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1076, | |
| "mean_token_accuracy": 0.9587001299485565, | |
| "num_tokens": 53504116.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5064092419686659, | |
| "grad_norm": 0.09016384504615442, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1064, | |
| "mean_token_accuracy": 0.9593421731144189, | |
| "num_tokens": 59463168.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5570501661655325, | |
| "grad_norm": 0.08920769596857461, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1061, | |
| "mean_token_accuracy": 0.9594598092138767, | |
| "num_tokens": 65384936.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6076910903623991, | |
| "grad_norm": 0.07392576254888708, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1037, | |
| "mean_token_accuracy": 0.960389680787921, | |
| "num_tokens": 71307379.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6583320145592657, | |
| "grad_norm": 0.07677666218910574, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1034, | |
| "mean_token_accuracy": 0.9603631895035505, | |
| "num_tokens": 77241947.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7089729387561323, | |
| "grad_norm": 0.08726421315898845, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1039, | |
| "mean_token_accuracy": 0.9603854931890965, | |
| "num_tokens": 83153311.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7596138629529989, | |
| "grad_norm": 0.07666737355553582, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1025, | |
| "mean_token_accuracy": 0.9607552452012896, | |
| "num_tokens": 89070746.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8102547871498654, | |
| "grad_norm": 0.07414032432163672, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1015, | |
| "mean_token_accuracy": 0.9611390510573983, | |
| "num_tokens": 95023182.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.860895711346732, | |
| "grad_norm": 0.07359359677738689, | |
| "learning_rate": 0.0001, | |
| "loss": 0.101, | |
| "mean_token_accuracy": 0.9613310528919101, | |
| "num_tokens": 100991573.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9115366355435987, | |
| "grad_norm": 0.07081191483858919, | |
| "learning_rate": 0.0001, | |
| "loss": 0.101, | |
| "mean_token_accuracy": 0.9613639689981938, | |
| "num_tokens": 106911646.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.9621775597404653, | |
| "grad_norm": 0.06862894355772857, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1005, | |
| "mean_token_accuracy": 0.9614966074004769, | |
| "num_tokens": 112858548.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0101281848393733, | |
| "grad_norm": 0.07771738090866837, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0961, | |
| "mean_token_accuracy": 0.9630992827635787, | |
| "num_tokens": 118496916.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0607691090362399, | |
| "grad_norm": 0.07979388553941799, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0857, | |
| "mean_token_accuracy": 0.9667170405387878, | |
| "num_tokens": 124429378.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.1114100332331065, | |
| "grad_norm": 0.08390392921164329, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0862, | |
| "mean_token_accuracy": 0.9665718451142311, | |
| "num_tokens": 130351603.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.162050957429973, | |
| "grad_norm": 0.08073237194411616, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0862, | |
| "mean_token_accuracy": 0.9665358671918511, | |
| "num_tokens": 136297692.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.2126918816268397, | |
| "grad_norm": 0.07936192759076259, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0878, | |
| "mean_token_accuracy": 0.9658019699156284, | |
| "num_tokens": 142214284.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.2633328058237063, | |
| "grad_norm": 0.07759622665900838, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0878, | |
| "mean_token_accuracy": 0.96583257522434, | |
| "num_tokens": 148149286.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.313973730020573, | |
| "grad_norm": 0.07211151901281589, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0872, | |
| "mean_token_accuracy": 0.9660558473318815, | |
| "num_tokens": 154110902.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.3646146542174395, | |
| "grad_norm": 0.09217544647176644, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0884, | |
| "mean_token_accuracy": 0.965622316300869, | |
| "num_tokens": 160053880.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.415255578414306, | |
| "grad_norm": 0.07245289826759833, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0885, | |
| "mean_token_accuracy": 0.9656080398708582, | |
| "num_tokens": 166020913.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.4658965026111725, | |
| "grad_norm": 0.0635278547575967, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0888, | |
| "mean_token_accuracy": 0.9655069762840867, | |
| "num_tokens": 171954396.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.5165374268080392, | |
| "grad_norm": 0.07063266507456424, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0885, | |
| "mean_token_accuracy": 0.9655342735350132, | |
| "num_tokens": 177907629.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.5671783510049058, | |
| "grad_norm": 0.09732972676853666, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0891, | |
| "mean_token_accuracy": 0.9652918018400669, | |
| "num_tokens": 183844004.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.6178192752017724, | |
| "grad_norm": 0.0844837833297714, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0887, | |
| "mean_token_accuracy": 0.9655532678589225, | |
| "num_tokens": 189815446.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.668460199398639, | |
| "grad_norm": 0.08316032543162945, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0904, | |
| "mean_token_accuracy": 0.9649298392236233, | |
| "num_tokens": 195754733.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.7191011235955056, | |
| "grad_norm": 0.08646791215437016, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0902, | |
| "mean_token_accuracy": 0.9649106938391924, | |
| "num_tokens": 201671874.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.7697420477923722, | |
| "grad_norm": 0.0758323031775405, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0901, | |
| "mean_token_accuracy": 0.9649893533438444, | |
| "num_tokens": 207597294.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.8203829719892388, | |
| "grad_norm": 0.07364548136372781, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0894, | |
| "mean_token_accuracy": 0.9652600049972534, | |
| "num_tokens": 213539369.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.8710238961861054, | |
| "grad_norm": 0.06507687587244941, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0901, | |
| "mean_token_accuracy": 0.9650838864967227, | |
| "num_tokens": 219489672.0, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.921664820382972, | |
| "grad_norm": 0.06289308119009687, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0903, | |
| "mean_token_accuracy": 0.9650570031255483, | |
| "num_tokens": 225428629.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.9723057445798386, | |
| "grad_norm": 0.06107047382323621, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0898, | |
| "mean_token_accuracy": 0.9651019360870123, | |
| "num_tokens": 231350509.0, | |
| "step": 390 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 788, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 617049630834688.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |