| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984, | |
| "eval_steps": 100, | |
| "global_step": 156, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "completion_length": 604.3152841567993, | |
| "epoch": 0.032, | |
| "grad_norm": 0.14503227174282074, | |
| "kl": 0.00012428760528564452, | |
| "learning_rate": 9.375e-07, | |
| "loss": 0.0, | |
| "reward": 0.6548611164093018, | |
| "reward_std": 0.3155763540416956, | |
| "rewards/accuracy_reward": 0.6541666720062494, | |
| "rewards/format_reward": 0.0006944444496184588, | |
| "step": 5 | |
| }, | |
| { | |
| "completion_length": 643.951396560669, | |
| "epoch": 0.064, | |
| "grad_norm": 0.08788644522428513, | |
| "kl": 0.00025680065155029295, | |
| "learning_rate": 1.875e-06, | |
| "loss": 0.0, | |
| "reward": 0.6215277845971287, | |
| "reward_std": 0.30458943024277685, | |
| "rewards/accuracy_reward": 0.6208333398215473, | |
| "rewards/format_reward": 0.0006944444496184588, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 606.6493129730225, | |
| "epoch": 0.096, | |
| "grad_norm": 0.10090751200914383, | |
| "kl": 0.0006760239601135253, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.0, | |
| "reward": 0.6791666748467833, | |
| "reward_std": 0.27497591376304625, | |
| "rewards/accuracy_reward": 0.6791666748467833, | |
| "rewards/format_reward": 0.0, | |
| "step": 15 | |
| }, | |
| { | |
| "completion_length": 628.4958393096924, | |
| "epoch": 0.128, | |
| "grad_norm": 0.10319401323795319, | |
| "kl": 0.0019372463226318359, | |
| "learning_rate": 2.993961440992859e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7437500089406968, | |
| "reward_std": 0.23143401462584734, | |
| "rewards/accuracy_reward": 0.7437500089406968, | |
| "rewards/format_reward": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "completion_length": 637.5111164093017, | |
| "epoch": 0.16, | |
| "grad_norm": 0.0020248512737452984, | |
| "kl": 0.13670992851257324, | |
| "learning_rate": 2.9695130976348534e-06, | |
| "loss": 0.0055, | |
| "reward": 0.7256944484543055, | |
| "reward_std": 0.23655044846236706, | |
| "rewards/accuracy_reward": 0.7256944484543055, | |
| "rewards/format_reward": 0.0, | |
| "step": 25 | |
| }, | |
| { | |
| "completion_length": 628.5451442718506, | |
| "epoch": 0.192, | |
| "grad_norm": 0.09042873978614807, | |
| "kl": 0.0024117946624755858, | |
| "learning_rate": 2.9265847744427307e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7555555619299412, | |
| "reward_std": 0.20260455049574375, | |
| "rewards/accuracy_reward": 0.7555555619299412, | |
| "rewards/format_reward": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "completion_length": 615.7291702270508, | |
| "epoch": 0.224, | |
| "grad_norm": 0.07207166403532028, | |
| "kl": 0.0032260894775390627, | |
| "learning_rate": 2.865716319988224e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7326388933230191, | |
| "reward_std": 0.21108066253364086, | |
| "rewards/accuracy_reward": 0.7326388933230191, | |
| "rewards/format_reward": 0.0, | |
| "step": 35 | |
| }, | |
| { | |
| "completion_length": 626.9736179351806, | |
| "epoch": 0.256, | |
| "grad_norm": 0.054428163915872574, | |
| "kl": 0.0026816368103027345, | |
| "learning_rate": 2.7876731904027993e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7451388964429497, | |
| "reward_std": 0.2157622052356601, | |
| "rewards/accuracy_reward": 0.7451388964429497, | |
| "rewards/format_reward": 0.0, | |
| "step": 40 | |
| }, | |
| { | |
| "completion_length": 594.1208389282226, | |
| "epoch": 0.288, | |
| "grad_norm": 0.08016934990882874, | |
| "kl": 0.002810811996459961, | |
| "learning_rate": 2.6934368233226715e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7694444504100829, | |
| "reward_std": 0.1868146162480116, | |
| "rewards/accuracy_reward": 0.7694444504100829, | |
| "rewards/format_reward": 0.0, | |
| "step": 45 | |
| }, | |
| { | |
| "completion_length": 614.3645858764648, | |
| "epoch": 0.32, | |
| "grad_norm": 0.11283142864704132, | |
| "kl": 0.002644062042236328, | |
| "learning_rate": 2.584192295741087e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7666666714474559, | |
| "reward_std": 0.18280234448611737, | |
| "rewards/accuracy_reward": 0.7666666714474559, | |
| "rewards/format_reward": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "completion_length": 587.173616027832, | |
| "epoch": 0.352, | |
| "grad_norm": 0.052785150706768036, | |
| "kl": 0.003171491622924805, | |
| "learning_rate": 2.461313420977536e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7930555634200573, | |
| "reward_std": 0.17841423936188222, | |
| "rewards/accuracy_reward": 0.7930555634200573, | |
| "rewards/format_reward": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "completion_length": 593.7201442718506, | |
| "epoch": 0.384, | |
| "grad_norm": 0.1886148303747177, | |
| "kl": 0.004308557510375977, | |
| "learning_rate": 2.3263454721781537e-06, | |
| "loss": 0.0002, | |
| "reward": 0.777083340100944, | |
| "reward_std": 0.18415420949459077, | |
| "rewards/accuracy_reward": 0.777083340100944, | |
| "rewards/format_reward": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "completion_length": 600.0354248046875, | |
| "epoch": 0.416, | |
| "grad_norm": 0.019889166578650475, | |
| "kl": 0.0036341667175292967, | |
| "learning_rate": 2.18098574960932e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7277777815237642, | |
| "reward_std": 0.18065068628638983, | |
| "rewards/accuracy_reward": 0.7277777815237642, | |
| "rewards/format_reward": 0.0, | |
| "step": 65 | |
| }, | |
| { | |
| "completion_length": 598.8555599212647, | |
| "epoch": 0.448, | |
| "grad_norm": 0.050428278744220734, | |
| "kl": 0.016699600219726562, | |
| "learning_rate": 2.027062236122014e-06, | |
| "loss": 0.0007, | |
| "reward": 0.733333338610828, | |
| "reward_std": 0.18704659640789031, | |
| "rewards/accuracy_reward": 0.733333338610828, | |
| "rewards/format_reward": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "completion_length": 589.2645866394043, | |
| "epoch": 0.48, | |
| "grad_norm": 0.5229154229164124, | |
| "kl": 0.004603099822998047, | |
| "learning_rate": 1.866510609206841e-06, | |
| "loss": 0.0002, | |
| "reward": 0.7680555649101735, | |
| "reward_std": 0.20577130634337665, | |
| "rewards/accuracy_reward": 0.7680555649101735, | |
| "rewards/format_reward": 0.0, | |
| "step": 75 | |
| }, | |
| { | |
| "completion_length": 613.6562576293945, | |
| "epoch": 0.512, | |
| "grad_norm": 0.03132776543498039, | |
| "kl": 0.0037424564361572266, | |
| "learning_rate": 1.7013498987264833e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7611111178994179, | |
| "reward_std": 0.1536103853955865, | |
| "rewards/accuracy_reward": 0.7611111178994179, | |
| "rewards/format_reward": 0.0, | |
| "step": 80 | |
| }, | |
| { | |
| "completion_length": 580.3000034332275, | |
| "epoch": 0.544, | |
| "grad_norm": 0.041480712592601776, | |
| "kl": 0.0032805442810058595, | |
| "learning_rate": 1.5336570964437077e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7500000051222742, | |
| "reward_std": 0.18460483234375716, | |
| "rewards/accuracy_reward": 0.7500000051222742, | |
| "rewards/format_reward": 0.0, | |
| "step": 85 | |
| }, | |
| { | |
| "completion_length": 577.7583400726319, | |
| "epoch": 0.576, | |
| "grad_norm": 0.0553722009062767, | |
| "kl": 0.003495025634765625, | |
| "learning_rate": 1.3655410366448499e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7687500093132258, | |
| "reward_std": 0.17976610995829106, | |
| "rewards/accuracy_reward": 0.7687500093132258, | |
| "rewards/format_reward": 0.0, | |
| "step": 90 | |
| }, | |
| { | |
| "completion_length": 585.31389503479, | |
| "epoch": 0.608, | |
| "grad_norm": 0.05203121528029442, | |
| "kl": 0.003870105743408203, | |
| "learning_rate": 1.199115876325091e-06, | |
| "loss": 0.0002, | |
| "reward": 0.7888888973742724, | |
| "reward_std": 0.15750880874693393, | |
| "rewards/accuracy_reward": 0.7888888973742724, | |
| "rewards/format_reward": 0.0, | |
| "step": 95 | |
| }, | |
| { | |
| "completion_length": 565.354870223999, | |
| "epoch": 0.64, | |
| "grad_norm": 0.06253661960363388, | |
| "kl": 0.003356456756591797, | |
| "learning_rate": 1.036474508437579e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7930555619299412, | |
| "reward_std": 0.15678047761321068, | |
| "rewards/accuracy_reward": 0.7930555619299412, | |
| "rewards/format_reward": 0.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_completion_length": 580.157574237716, | |
| "eval_kl": 0.007348269420822866, | |
| "eval_loss": 0.00029379926854744554, | |
| "eval_reward": 0.6814303872967167, | |
| "eval_reward_std": 0.20519427888108788, | |
| "eval_rewards/accuracy_reward": 0.6814303872967167, | |
| "eval_rewards/format_reward": 0.0, | |
| "eval_runtime": 15488.5186, | |
| "eval_samples_per_second": 0.323, | |
| "eval_steps_per_second": 0.108, | |
| "step": 100 | |
| }, | |
| { | |
| "completion_length": 574.4409788131713, | |
| "epoch": 0.672, | |
| "grad_norm": 0.034538425505161285, | |
| "kl": 0.003883838653564453, | |
| "learning_rate": 8.796622425502193e-07, | |
| "loss": 0.0002, | |
| "reward": 0.7833333427086473, | |
| "reward_std": 0.1512267516925931, | |
| "rewards/accuracy_reward": 0.7833333427086473, | |
| "rewards/format_reward": 0.0, | |
| "step": 105 | |
| }, | |
| { | |
| "completion_length": 611.8923690795898, | |
| "epoch": 0.704, | |
| "grad_norm": 0.004378489218652248, | |
| "kl": 0.0033061027526855467, | |
| "learning_rate": 7.30651083891141e-07, | |
| "loss": 0.0001, | |
| "reward": 0.7430555634200573, | |
| "reward_std": 0.19789394550025463, | |
| "rewards/accuracy_reward": 0.7430555634200573, | |
| "rewards/format_reward": 0.0, | |
| "step": 110 | |
| }, | |
| { | |
| "completion_length": 608.2687576293945, | |
| "epoch": 0.736, | |
| "grad_norm": 0.0850713923573494, | |
| "kl": 0.0032596588134765625, | |
| "learning_rate": 5.913149342387704e-07, | |
| "loss": 0.0001, | |
| "reward": 0.7694444509223104, | |
| "reward_std": 0.1801286071538925, | |
| "rewards/accuracy_reward": 0.7694444509223104, | |
| "rewards/format_reward": 0.0, | |
| "step": 115 | |
| }, | |
| { | |
| "completion_length": 572.259033203125, | |
| "epoch": 0.768, | |
| "grad_norm": 0.06911196559667587, | |
| "kl": 0.0034391403198242186, | |
| "learning_rate": 4.63406026519703e-07, | |
| "loss": 0.0001, | |
| "reward": 0.7722222298849374, | |
| "reward_std": 0.15622506141662598, | |
| "rewards/accuracy_reward": 0.7722222298849374, | |
| "rewards/format_reward": 0.0, | |
| "step": 120 | |
| }, | |
| { | |
| "completion_length": 604.2020915985107, | |
| "epoch": 0.8, | |
| "grad_norm": 0.09203875064849854, | |
| "kl": 0.003432178497314453, | |
| "learning_rate": 3.4853288946298335e-07, | |
| "loss": 0.0001, | |
| "reward": 0.7388888964429497, | |
| "reward_std": 0.18809502944350243, | |
| "rewards/accuracy_reward": 0.7388888964429497, | |
| "rewards/format_reward": 0.0, | |
| "step": 125 | |
| }, | |
| { | |
| "completion_length": 582.3465305328369, | |
| "epoch": 0.832, | |
| "grad_norm": 0.04850884899497032, | |
| "kl": 0.0036045074462890624, | |
| "learning_rate": 2.48140119418046e-07, | |
| "loss": 0.0001, | |
| "reward": 0.7687500065192581, | |
| "reward_std": 0.15752214156091213, | |
| "rewards/accuracy_reward": 0.7687500065192581, | |
| "rewards/format_reward": 0.0, | |
| "step": 130 | |
| }, | |
| { | |
| "completion_length": 599.1916728973389, | |
| "epoch": 0.864, | |
| "grad_norm": 0.03212558850646019, | |
| "kl": 0.0031602859497070314, | |
| "learning_rate": 1.634902137174483e-07, | |
| "loss": 0.0001, | |
| "reward": 0.7312500088475644, | |
| "reward_std": 0.1761287273839116, | |
| "rewards/accuracy_reward": 0.7312500088475644, | |
| "rewards/format_reward": 0.0, | |
| "step": 135 | |
| }, | |
| { | |
| "completion_length": 577.2263946533203, | |
| "epoch": 0.896, | |
| "grad_norm": 0.03874632343649864, | |
| "kl": 0.003940582275390625, | |
| "learning_rate": 9.564769404039419e-08, | |
| "loss": 0.0002, | |
| "reward": 0.7729166727513075, | |
| "reward_std": 0.189943253621459, | |
| "rewards/accuracy_reward": 0.7729166727513075, | |
| "rewards/format_reward": 0.0, | |
| "step": 140 | |
| }, | |
| { | |
| "completion_length": 575.8805583953857, | |
| "epoch": 0.928, | |
| "grad_norm": 0.07863741368055344, | |
| "kl": 0.004055213928222656, | |
| "learning_rate": 4.546571943496969e-08, | |
| "loss": 0.0002, | |
| "reward": 0.7590277843177319, | |
| "reward_std": 0.19089928902685643, | |
| "rewards/accuracy_reward": 0.7590277843177319, | |
| "rewards/format_reward": 0.0, | |
| "step": 145 | |
| }, | |
| { | |
| "completion_length": 585.861116027832, | |
| "epoch": 0.96, | |
| "grad_norm": 0.02937915176153183, | |
| "kl": 0.004036521911621094, | |
| "learning_rate": 1.357535734809795e-08, | |
| "loss": 0.0002, | |
| "reward": 0.7618055611848831, | |
| "reward_std": 0.169761879183352, | |
| "rewards/accuracy_reward": 0.7618055611848831, | |
| "rewards/format_reward": 0.0, | |
| "step": 150 | |
| }, | |
| { | |
| "completion_length": 580.2944515228271, | |
| "epoch": 0.992, | |
| "grad_norm": 0.04815535247325897, | |
| "kl": 0.003546333312988281, | |
| "learning_rate": 3.77647586240204e-10, | |
| "loss": 0.0001, | |
| "reward": 0.7680555626749992, | |
| "reward_std": 0.18228026237338782, | |
| "rewards/accuracy_reward": 0.7680555626749992, | |
| "rewards/format_reward": 0.0, | |
| "step": 155 | |
| }, | |
| { | |
| "completion_length": 629.1527805328369, | |
| "epoch": 0.9984, | |
| "kl": 0.002917766571044922, | |
| "reward": 0.7187500055879354, | |
| "reward_std": 0.1746763400733471, | |
| "rewards/accuracy_reward": 0.7187500055879354, | |
| "rewards/format_reward": 0.0, | |
| "step": 156, | |
| "total_flos": 0.0, | |
| "train_loss": 0.0003127560741445333, | |
| "train_runtime": 45856.1745, | |
| "train_samples_per_second": 0.164, | |
| "train_steps_per_second": 0.003 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 156, | |
| "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": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |