| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.997867803837953, | |
| "eval_steps": 100, | |
| "global_step": 117, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 620.6149749755859, | |
| "epoch": 0.008528784648187633, | |
| "grad_norm": 0.43470078706741333, | |
| "kl": 0.0, | |
| "learning_rate": 2.5e-07, | |
| "loss": 0.0087, | |
| "reward": 0.611607164144516, | |
| "reward_std": 0.3698773570358753, | |
| "rewards/accuracy_reward": 0.611607164144516, | |
| "rewards/format_reward": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 593.875862121582, | |
| "epoch": 0.042643923240938165, | |
| "grad_norm": 0.43776455521583557, | |
| "kl": 0.00015407800674438477, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.0231, | |
| "reward": 0.6205357443541288, | |
| "reward_std": 0.37901367666199803, | |
| "rewards/accuracy_reward": 0.6205357443541288, | |
| "rewards/format_reward": 0.0, | |
| "step": 5 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 616.8395332336426, | |
| "epoch": 0.08528784648187633, | |
| "grad_norm": 0.8126201033592224, | |
| "kl": 0.002016162872314453, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.043, | |
| "reward": 0.6250000275671482, | |
| "reward_std": 0.347636329382658, | |
| "rewards/accuracy_reward": 0.6250000275671482, | |
| "rewards/format_reward": 0.0, | |
| "step": 10 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 610.7120796203614, | |
| "epoch": 0.1279317697228145, | |
| "grad_norm": 0.1547478586435318, | |
| "kl": 0.013087844848632813, | |
| "learning_rate": 2.993961440992859e-06, | |
| "loss": 0.0896, | |
| "reward": 0.7162946730852127, | |
| "reward_std": 0.2822846982628107, | |
| "rewards/accuracy_reward": 0.7162946730852127, | |
| "rewards/format_reward": 0.0, | |
| "step": 15 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 612.700473022461, | |
| "epoch": 0.17057569296375266, | |
| "grad_norm": 0.24769772589206696, | |
| "kl": 0.01591835021972656, | |
| "learning_rate": 2.957235057439301e-06, | |
| "loss": 0.0769, | |
| "reward": 0.7506696745753288, | |
| "reward_std": 0.23058689776808022, | |
| "rewards/accuracy_reward": 0.7506696745753288, | |
| "rewards/format_reward": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 575.4734634399414, | |
| "epoch": 0.21321961620469082, | |
| "grad_norm": 1.411234736442566, | |
| "kl": 0.006292724609375, | |
| "learning_rate": 2.887956450710995e-06, | |
| "loss": 0.0598, | |
| "reward": 0.8000000357627869, | |
| "reward_std": 0.18773337844759225, | |
| "rewards/accuracy_reward": 0.8000000357627869, | |
| "rewards/format_reward": 0.0, | |
| "step": 25 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 579.8424324035644, | |
| "epoch": 0.255863539445629, | |
| "grad_norm": 0.17543800175189972, | |
| "kl": 0.004089736938476562, | |
| "learning_rate": 2.7876731904027993e-06, | |
| "loss": 0.0551, | |
| "reward": 0.7698661036789417, | |
| "reward_std": 0.20191936586052178, | |
| "rewards/accuracy_reward": 0.7698661036789417, | |
| "rewards/format_reward": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 577.5357452392578, | |
| "epoch": 0.29850746268656714, | |
| "grad_norm": 0.14460965991020203, | |
| "kl": 0.004575347900390625, | |
| "learning_rate": 2.6586254388368995e-06, | |
| "loss": 0.0494, | |
| "reward": 0.7982143223285675, | |
| "reward_std": 0.17998763117939234, | |
| "rewards/accuracy_reward": 0.7982143223285675, | |
| "rewards/format_reward": 0.0, | |
| "step": 35 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 573.0772575378418, | |
| "epoch": 0.3411513859275053, | |
| "grad_norm": 0.10168638080358505, | |
| "kl": 0.004691314697265625, | |
| "learning_rate": 2.5036959095382875e-06, | |
| "loss": 0.0403, | |
| "reward": 0.7850446775555611, | |
| "reward_std": 0.1875380737707019, | |
| "rewards/accuracy_reward": 0.7850446775555611, | |
| "rewards/format_reward": 0.0, | |
| "step": 40 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 583.6614151000977, | |
| "epoch": 0.3837953091684435, | |
| "grad_norm": 0.15747234225273132, | |
| "kl": 0.007823562622070313, | |
| "learning_rate": 2.3263454721781537e-06, | |
| "loss": 0.042, | |
| "reward": 0.8042411088943482, | |
| "reward_std": 0.17801097435876728, | |
| "rewards/accuracy_reward": 0.8042411088943482, | |
| "rewards/format_reward": 0.0, | |
| "step": 45 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 582.4127494812012, | |
| "epoch": 0.42643923240938164, | |
| "grad_norm": 0.13581737875938416, | |
| "kl": 0.005602645874023438, | |
| "learning_rate": 2.1305358424643485e-06, | |
| "loss": 0.04, | |
| "reward": 0.7600446760654449, | |
| "reward_std": 0.18583385180681944, | |
| "rewards/accuracy_reward": 0.7600446760654449, | |
| "rewards/format_reward": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 570.5196685791016, | |
| "epoch": 0.4690831556503198, | |
| "grad_norm": 0.14316147565841675, | |
| "kl": 0.006391143798828125, | |
| "learning_rate": 1.9206410839590043e-06, | |
| "loss": 0.0253, | |
| "reward": 0.770982176065445, | |
| "reward_std": 0.1873089073225856, | |
| "rewards/accuracy_reward": 0.770982176065445, | |
| "rewards/format_reward": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 597.8089546203613, | |
| "epoch": 0.511727078891258, | |
| "grad_norm": 0.13278549909591675, | |
| "kl": 0.006211090087890625, | |
| "learning_rate": 1.7013498987264833e-06, | |
| "loss": 0.0323, | |
| "reward": 0.7582589626312256, | |
| "reward_std": 0.19469668436795473, | |
| "rewards/accuracy_reward": 0.7582589626312256, | |
| "rewards/format_reward": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 570.72748336792, | |
| "epoch": 0.5543710021321961, | |
| "grad_norm": 0.11964963376522064, | |
| "kl": 0.0067108154296875, | |
| "learning_rate": 1.4775608894771048e-06, | |
| "loss": 0.0285, | |
| "reward": 0.7604911118745804, | |
| "reward_std": 0.17904917486011981, | |
| "rewards/accuracy_reward": 0.7604911118745804, | |
| "rewards/format_reward": 0.0, | |
| "step": 65 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 583.9893104553223, | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 0.13020634651184082, | |
| "kl": 0.005590057373046875, | |
| "learning_rate": 1.2542731328772936e-06, | |
| "loss": 0.0335, | |
| "reward": 0.7493303909897804, | |
| "reward_std": 0.19070985857397318, | |
| "rewards/accuracy_reward": 0.7493303909897804, | |
| "rewards/format_reward": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 552.5370803833008, | |
| "epoch": 0.6396588486140725, | |
| "grad_norm": 0.193593829870224, | |
| "kl": 0.00690155029296875, | |
| "learning_rate": 1.036474508437579e-06, | |
| "loss": 0.0293, | |
| "reward": 0.7955357491970062, | |
| "reward_std": 0.17065229499712586, | |
| "rewards/accuracy_reward": 0.7955357491970062, | |
| "rewards/format_reward": 0.0, | |
| "step": 75 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 574.2089538574219, | |
| "epoch": 0.6823027718550106, | |
| "grad_norm": 0.26767051219940186, | |
| "kl": 0.006514739990234375, | |
| "learning_rate": 8.290302775265509e-07, | |
| "loss": 0.0359, | |
| "reward": 0.7752232506871224, | |
| "reward_std": 0.17853977270424365, | |
| "rewards/accuracy_reward": 0.7752232506871224, | |
| "rewards/format_reward": 0.0, | |
| "step": 80 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 587.2473434448242, | |
| "epoch": 0.7249466950959488, | |
| "grad_norm": 0.14498883485794067, | |
| "kl": 0.008393096923828124, | |
| "learning_rate": 6.3657440147149e-07, | |
| "loss": 0.0378, | |
| "reward": 0.77857146859169, | |
| "reward_std": 0.1991961758583784, | |
| "rewards/accuracy_reward": 0.77857146859169, | |
| "rewards/format_reward": 0.0, | |
| "step": 85 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 571.0614074707031, | |
| "epoch": 0.767590618336887, | |
| "grad_norm": 0.3927496671676636, | |
| "kl": 0.00813751220703125, | |
| "learning_rate": 4.63406026519703e-07, | |
| "loss": 0.0197, | |
| "reward": 0.753348246216774, | |
| "reward_std": 0.18298904253169895, | |
| "rewards/accuracy_reward": 0.753348246216774, | |
| "rewards/format_reward": 0.0, | |
| "step": 90 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 579.4897560119629, | |
| "epoch": 0.8102345415778252, | |
| "grad_norm": 0.15556038916110992, | |
| "kl": 0.007627105712890625, | |
| "learning_rate": 3.133934480154885e-07, | |
| "loss": 0.0348, | |
| "reward": 0.7566964641213417, | |
| "reward_std": 0.20147906215861439, | |
| "rewards/accuracy_reward": 0.7566964641213417, | |
| "rewards/format_reward": 0.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.8528784648187633, | |
| "grad_norm": 0.397614985704422, | |
| "learning_rate": 1.8988769907430552e-07, | |
| "loss": 0.0301, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8528784648187633, | |
| "eval_clip_ratio": 0.0, | |
| "eval_completion_length": 558.2626213378907, | |
| "eval_kl": 0.02655439453125, | |
| "eval_loss": 0.01617129147052765, | |
| "eval_reward": 0.6867000318527222, | |
| "eval_reward_std": 0.22212331930994988, | |
| "eval_rewards/accuracy_reward": 0.6867000318527222, | |
| "eval_rewards/format_reward": 0.0, | |
| "eval_runtime": 17227.6498, | |
| "eval_samples_per_second": 0.29, | |
| "eval_steps_per_second": 0.003, | |
| "step": 100 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 574.3294898986817, | |
| "epoch": 0.8955223880597015, | |
| "grad_norm": 0.7429274916648865, | |
| "kl": 0.007941818237304688, | |
| "learning_rate": 9.564769404039419e-08, | |
| "loss": 0.0334, | |
| "reward": 0.7816964637488126, | |
| "reward_std": 0.18899818495847284, | |
| "rewards/accuracy_reward": 0.7816964637488126, | |
| "rewards/format_reward": 0.0, | |
| "step": 105 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 574.8442253112793, | |
| "epoch": 0.9381663113006397, | |
| "grad_norm": 0.1576305478811264, | |
| "kl": 0.00836029052734375, | |
| "learning_rate": 3.277859889929147e-08, | |
| "loss": 0.0323, | |
| "reward": 0.8006696775555611, | |
| "reward_std": 0.19058242589235305, | |
| "rewards/accuracy_reward": 0.8006696775555611, | |
| "rewards/format_reward": 0.0, | |
| "step": 110 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 558.2982414245605, | |
| "epoch": 0.9808102345415778, | |
| "grad_norm": 0.16698426008224487, | |
| "kl": 0.0088043212890625, | |
| "learning_rate": 2.684805348397268e-09, | |
| "loss": 0.0311, | |
| "reward": 0.7834821820259095, | |
| "reward_std": 0.17699794424697757, | |
| "rewards/accuracy_reward": 0.7834821820259095, | |
| "rewards/format_reward": 0.0, | |
| "step": 115 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 580.5032634735107, | |
| "epoch": 0.997867803837953, | |
| "kl": 0.008594512939453125, | |
| "reward": 0.7589286100119352, | |
| "reward_std": 0.21703570941463113, | |
| "rewards/accuracy_reward": 0.7589286100119352, | |
| "rewards/format_reward": 0.0, | |
| "step": 117, | |
| "total_flos": 0.0, | |
| "train_loss": 0.03906208295056708, | |
| "train_runtime": 48868.7895, | |
| "train_samples_per_second": 0.153, | |
| "train_steps_per_second": 0.002 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 117, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |