{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 5.714285714285714, "eval_steps": 500, "global_step": 200, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "completion_length": 259.0625, "epoch": 0.05714285714285714, "grad_norm": 40.40498733520508, "kl": 0.0, "learning_rate": 1.6666666666666665e-07, "loss": -0.0001, "reward": 4.278493709862232, "reward_std": 1.2902945003088462, "rewards/concensus_correctness_reward_func": 1.7047499865293503, "rewards/consensus_reward_func": 1.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0625, "rewards/question_recreation_reward_func": 0.675962442997843, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.1875, "rewards/xmlcount_reward_func": 0.6477812505327165, "step": 2 }, { "completion_length": 131.8125, "epoch": 0.11428571428571428, "grad_norm": 11.698363304138184, "kl": 0.05078949875195282, "learning_rate": 5e-07, "loss": 0.0, "reward": 6.484069406986237, "reward_std": 0.30921973334625363, "rewards/concensus_correctness_reward_func": 1.9382499903440475, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.1875, "rewards/question_recreation_reward_func": 0.866131940856576, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.0859375, "step": 4 }, { "completion_length": 167.59375, "epoch": 0.17142857142857143, "grad_norm": 113298272.0, "kl": 3785875.3213646472, "learning_rate": 8.333333333333333e-07, "loss": 3785.876, "reward": 6.581158250570297, "reward_std": 0.3191240727901459, "rewards/concensus_correctness_reward_func": 2.0427499786019325, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0625, "rewards/question_recreation_reward_func": 0.8977832514792681, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.15625, "step": 6 }, { "completion_length": 139.6875, "epoch": 0.22857142857142856, "grad_norm": 898.8678588867188, "kl": 377.1116261743221, "learning_rate": 9.99934441832816e-07, "loss": 0.3771, "reward": 6.764802664518356, "reward_std": 0.35591571987606585, "rewards/concensus_correctness_reward_func": 2.1679999828338623, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.1875, "rewards/question_recreation_reward_func": 0.8858651723712683, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.1484375, "step": 8 }, { "completion_length": 150.0625, "epoch": 0.2857142857142857, "grad_norm": 2745.2119140625, "kl": 198.94737304830204, "learning_rate": 9.994100796397953e-07, "loss": 0.1989, "reward": 6.304591566324234, "reward_std": 0.26397108286619186, "rewards/concensus_correctness_reward_func": 1.9194999784231186, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8811853062361479, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.12890625, "step": 10 }, { "completion_length": 157.65625, "epoch": 0.34285714285714286, "grad_norm": 47.17015075683594, "kl": 10.171083455644862, "learning_rate": 9.983619052372847e-07, "loss": 0.0102, "reward": 6.150758773088455, "reward_std": 0.26591456681489944, "rewards/concensus_correctness_reward_func": 1.9152499809861183, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8178525845287368, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.34375, "rewards/xmlcount_reward_func": 1.0739062502980232, "step": 12 }, { "completion_length": 127.9375, "epoch": 0.4, "grad_norm": 30.31770896911621, "kl": 1.7561549306603865, "learning_rate": 9.967910180154888e-07, "loss": 0.0018, "reward": 6.409551799297333, "reward_std": 0.0954170327167958, "rewards/concensus_correctness_reward_func": 1.9182499796152115, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8545830636285245, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4375, "rewards/xmlcount_reward_func": 1.19921875, "step": 14 }, { "completion_length": 125.1875, "epoch": 0.45714285714285713, "grad_norm": 21.316240310668945, "kl": 61.69702582350237, "learning_rate": 9.946990656181779e-07, "loss": 0.0617, "reward": 7.006136029958725, "reward_std": 0.3882664368939004, "rewards/concensus_correctness_reward_func": 2.293499991297722, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.3125, "rewards/question_recreation_reward_func": 0.8649797681719065, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.12890625, "step": 16 }, { "completion_length": 117.84375, "epoch": 0.5142857142857142, "grad_norm": 0.30895093083381653, "kl": 3.9632333860547533, "learning_rate": 9.92088242214537e-07, "loss": 0.004, "reward": 7.0876125395298, "reward_std": 0.02687215618789196, "rewards/concensus_correctness_reward_func": 2.1668749675154686, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.9754250645637512, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.2265625, "step": 18 }, { "completion_length": 121.875, "epoch": 0.5714285714285714, "grad_norm": 27.664838790893555, "kl": 20.31627042979403, "learning_rate": 9.889612861977853e-07, "loss": 0.0203, "reward": 6.253143697977066, "reward_std": 0.15424543642438948, "rewards/concensus_correctness_reward_func": 1.9189999774098396, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.9044562540948391, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.328125, "rewards/xmlcount_reward_func": 1.1015625, "step": 20 }, { "completion_length": 116.15625, "epoch": 0.6285714285714286, "grad_norm": 43.6223030090332, "kl": 7.602870227153744, "learning_rate": 9.853214773129795e-07, "loss": 0.0076, "reward": 6.478605628013611, "reward_std": 0.09073211625218391, "rewards/concensus_correctness_reward_func": 1.918249987065792, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.9867619276046753, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.2298437505960464, "step": 22 }, { "completion_length": 144.28125, "epoch": 0.6857142857142857, "grad_norm": 208.64044189453125, "kl": 1171.1961524431879, "learning_rate": 9.81172633217015e-07, "loss": 1.1712, "reward": 6.415135711431503, "reward_std": 0.1869081521872431, "rewards/concensus_correctness_reward_func": 1.9849999640136957, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.3125, "rewards/question_recreation_reward_func": 0.8572919256985188, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.328125, "rewards/xmlcount_reward_func": 1.057218749076128, "step": 24 }, { "completion_length": 125.46875, "epoch": 0.7428571428571429, "grad_norm": 17.30021858215332, "kl": 22246.188865144257, "learning_rate": 9.765191054744304e-07, "loss": 22.2462, "reward": 6.495312124490738, "reward_std": 0.2721763923764229, "rewards/concensus_correctness_reward_func": 1.8801249898970127, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.9621558953076601, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.453125, "rewards/xmlcount_reward_func": 1.1999062523245811, "step": 26 }, { "completion_length": 110.0625, "epoch": 0.8, "grad_norm": 0.00991878192871809, "kl": 0.025511330234024854, "learning_rate": 9.713657749932171e-07, "loss": 0.0, "reward": 6.9107471108436584, "reward_std": 0.0035395766608417034, "rewards/concensus_correctness_reward_func": 2.038249984383583, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.9974971413612366, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.5, "rewards/xmlcount_reward_func": 1.25, "step": 28 }, { "completion_length": 110.5625, "epoch": 0.8571428571428571, "grad_norm": 3189.01806640625, "kl": 254.6465662113756, "learning_rate": 9.657180469054212e-07, "loss": 0.2546, "reward": 7.057701647281647, "reward_std": 0.15232454985380173, "rewards/concensus_correctness_reward_func": 2.16849996894598, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.9595141224563122, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.2109375, "step": 30 }, { "completion_length": 144.03125, "epoch": 0.9142857142857143, "grad_norm": 1612504.375, "kl": 182371.5907895875, "learning_rate": 9.59581844897906e-07, "loss": 182.3716, "reward": 6.731081739068031, "reward_std": 0.1562379876850173, "rewards/concensus_correctness_reward_func": 2.1022499799728394, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.8358630024595186, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.13671875, "step": 32 }, { "completion_length": 120.0, "epoch": 0.9714285714285714, "grad_norm": 0.11929623037576675, "kl": 11.570230805784433, "learning_rate": 9.529636049992233e-07, "loss": 0.0116, "reward": 6.652986168861389, "reward_std": 0.16899008117616177, "rewards/concensus_correctness_reward_func": 2.043749988079071, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0625, "rewards/question_recreation_reward_func": 0.9419549033045769, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.1829062476754189, "step": 34 }, { "completion_length": 155.0625, "epoch": 1.0285714285714285, "grad_norm": 26.967811584472656, "kl": 41.41832308189555, "learning_rate": 9.458702688291071e-07, "loss": 0.0414, "reward": 6.463270157575607, "reward_std": 0.0986387450248003, "rewards/concensus_correctness_reward_func": 1.978499986231327, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.9144576657563448, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.1484375, "step": 36 }, { "completion_length": 172.125, "epoch": 1.0857142857142856, "grad_norm": 9184.3974609375, "kl": 382.0947194062253, "learning_rate": 9.383092763176738e-07, "loss": 0.3821, "reward": 6.269357413053513, "reward_std": 0.37034115626011044, "rewards/concensus_correctness_reward_func": 1.8153749704360962, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0625, "rewards/question_recreation_reward_func": 0.8684198576956987, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.1168124992400408, "step": 38 }, { "completion_length": 162.34375, "epoch": 1.1428571428571428, "grad_norm": 85.47990417480469, "kl": 20.64761355567407, "learning_rate": 9.302885579019626e-07, "loss": 0.0206, "reward": 6.838260769844055, "reward_std": 0.28286098898388445, "rewards/concensus_correctness_reward_func": 2.168374963104725, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.8730107028968632, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.140625, "step": 40 }, { "completion_length": 238.5625, "epoch": 1.2, "grad_norm": 24413.638671875, "kl": 7399.846957861285, "learning_rate": 9.218165262080022e-07, "loss": 7.3998, "reward": 6.1828261613845825, "reward_std": 0.09892372484318912, "rewards/concensus_correctness_reward_func": 1.9538749903440475, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.7445762166171335, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.109375, "step": 42 }, { "completion_length": 121.21875, "epoch": 1.2571428571428571, "grad_norm": 628.685302734375, "kl": 196.11569420348624, "learning_rate": 9.129020672271281e-07, "loss": 0.1961, "reward": 6.581244736909866, "reward_std": 0.20140560899744742, "rewards/concensus_correctness_reward_func": 2.038249984383583, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.875025988323614, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.390625, "rewards/xmlcount_reward_func": 1.15234375, "step": 44 }, { "completion_length": 144.90625, "epoch": 1.3142857142857143, "grad_norm": 38.4938850402832, "kl": 14.245073152454609, "learning_rate": 9.035545309958046e-07, "loss": 0.0142, "reward": 6.480484366416931, "reward_std": 0.2079393861349672, "rewards/concensus_correctness_reward_func": 1.920499972999096, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8879531715065241, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.1407812498509884, "step": 46 }, { "completion_length": 169.46875, "epoch": 1.3714285714285714, "grad_norm": 40.683929443359375, "kl": 40.01735451040594, "learning_rate": 8.937837217887272e-07, "loss": 0.04, "reward": 6.485899716615677, "reward_std": 0.30478080292232335, "rewards/concensus_correctness_reward_func": 2.0423749834299088, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8660247628577054, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.0462500005960464, "step": 48 }, { "completion_length": 323.28125, "epoch": 1.4285714285714286, "grad_norm": 136.50091552734375, "kl": 208.73060331924557, "learning_rate": 8.83599887835493e-07, "loss": 0.2087, "reward": 5.868565648794174, "reward_std": 0.40139913094753865, "rewards/concensus_correctness_reward_func": 1.9188749939203262, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.6675031447084621, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.28125, "rewards/xmlcount_reward_func": 1.0009374991059303, "step": 50 }, { "completion_length": 223.90625, "epoch": 1.4857142857142858, "grad_norm": 3347453.5, "kl": 294342.96547964675, "learning_rate": 8.73013710571623e-07, "loss": 294.343, "reward": 6.5324656665325165, "reward_std": 0.18765212316066027, "rewards/concensus_correctness_reward_func": 2.0436249747872353, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.836496805306524, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.12109375, "step": 52 }, { "completion_length": 108.03125, "epoch": 1.5428571428571427, "grad_norm": 5130.6298828125, "kl": 589.7922593226319, "learning_rate": 8.620362934352108e-07, "loss": 0.5898, "reward": 6.766858085989952, "reward_std": 0.38228555023670197, "rewards/concensus_correctness_reward_func": 2.107999976724386, "rewards/consensus_reward_func": 1.9375, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.8854205666575581, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4375, "rewards/xmlcount_reward_func": 1.1484375, "step": 54 }, { "completion_length": 197.375, "epoch": 1.6, "grad_norm": 57.51034164428711, "kl": 17.006517063246605, "learning_rate": 8.506791502209496e-07, "loss": 0.017, "reward": 7.339930027723312, "reward_std": 0.07264970572759921, "rewards/concensus_correctness_reward_func": 2.541249990463257, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.5, "rewards/question_recreation_reward_func": 0.8182113186921924, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.10546875, "step": 56 }, { "completion_length": 121.5625, "epoch": 1.657142857142857, "grad_norm": 64.63337707519531, "kl": 37.444439904923236, "learning_rate": 8.389541930037516e-07, "loss": 0.0374, "reward": 6.30495747923851, "reward_std": 0.06559963073232211, "rewards/concensus_correctness_reward_func": 1.9192499816417694, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8310199771076441, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.1484375, "step": 58 }, { "completion_length": 266.6875, "epoch": 1.7142857142857144, "grad_norm": 81.60047912597656, "kl": 40.02249676991286, "learning_rate": 8.268737196446263e-07, "loss": 0.04, "reward": 6.112664610147476, "reward_std": 0.12300743258674629, "rewards/concensus_correctness_reward_func": 1.9388749822974205, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.6816021917038597, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.328125, "rewards/xmlcount_reward_func": 1.0390625, "step": 60 }, { "completion_length": 168.25, "epoch": 1.7714285714285714, "grad_norm": 7163.0244140625, "kl": 595.5683929037647, "learning_rate": 8.144504008919222e-07, "loss": 0.5956, "reward": 6.3996052742004395, "reward_std": 0.0010765780170913786, "rewards/concensus_correctness_reward_func": 1.917749971151352, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8724802895449102, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4375, "rewards/xmlcount_reward_func": 1.171875, "step": 62 }, { "completion_length": 110.59375, "epoch": 1.8285714285714287, "grad_norm": 7897.408203125, "kl": 480.5656087023023, "learning_rate": 8.016972670914623e-07, "loss": 0.4806, "reward": 7.056159615516663, "reward_std": 0.06930830026976764, "rewards/concensus_correctness_reward_func": 2.169124998152256, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.953440960496664, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.453125, "rewards/xmlcount_reward_func": 1.23046875, "step": 64 }, { "completion_length": 180.96875, "epoch": 1.8857142857142857, "grad_norm": 168.32589721679688, "kl": 21.501576103312345, "learning_rate": 7.886276945195097e-07, "loss": 0.0215, "reward": 6.382116913795471, "reward_std": 0.37760969676310197, "rewards/concensus_correctness_reward_func": 2.043124981224537, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0625, "rewards/question_recreation_reward_func": 0.7921168450266123, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.109375, "step": 66 }, { "completion_length": 232.65625, "epoch": 1.9428571428571428, "grad_norm": 129.97056579589844, "kl": 17.854988246223172, "learning_rate": 7.752553913529018e-07, "loss": 0.0179, "reward": 6.253960728645325, "reward_std": 0.4535234868526459, "rewards/concensus_correctness_reward_func": 2.0708749908953905, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.1875, "rewards/question_recreation_reward_func": 0.6741795390844345, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.28125, "rewards/xmlcount_reward_func": 1.0401562489569187, "step": 68 }, { "completion_length": 143.1875, "epoch": 2.0, "grad_norm": 72.03203582763672, "kl": 3.0041830078563976, "learning_rate": 7.61594383291065e-07, "loss": 0.003, "reward": 6.1214989721775055, "reward_std": 0.33512147367582656, "rewards/concensus_correctness_reward_func": 2.042374961078167, "rewards/consensus_reward_func": 1.75, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0625, "rewards/question_recreation_reward_func": 0.8236239920370281, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.359375, "rewards/xmlcount_reward_func": 1.0836249999701977, "step": 70 }, { "completion_length": 152.0, "epoch": 2.057142857142857, "grad_norm": 43.813533782958984, "kl": 6.536038895829506, "learning_rate": 7.476589988449938e-07, "loss": 0.0065, "reward": 6.629714250564575, "reward_std": 0.3594768207985908, "rewards/concensus_correctness_reward_func": 2.107249964028597, "rewards/consensus_reward_func": 1.9375, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.8115267266985029, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.390625, "rewards/xmlcount_reward_func": 1.1328125, "step": 72 }, { "completion_length": 178.46875, "epoch": 2.1142857142857143, "grad_norm": 233.30178833007812, "kl": 18.578114512876056, "learning_rate": 7.334638543086203e-07, "loss": 0.0186, "reward": 6.874465793371201, "reward_std": 0.40307298954576254, "rewards/concensus_correctness_reward_func": 2.2929999753832817, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.3125, "rewards/question_recreation_reward_func": 0.8509970791637897, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.390625, "rewards/xmlcount_reward_func": 1.15234375, "step": 74 }, { "completion_length": 132.71875, "epoch": 2.1714285714285713, "grad_norm": 33.67573928833008, "kl": 12.638730006660069, "learning_rate": 7.190238384283412e-07, "loss": 0.0126, "reward": 6.277231350541115, "reward_std": 0.0966361474711448, "rewards/concensus_correctness_reward_func": 1.8519999831914902, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8510125828906894, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.15234375, "step": 76 }, { "completion_length": 119.125, "epoch": 2.2285714285714286, "grad_norm": 71.8940658569336, "kl": 13.95311574424943, "learning_rate": 7.043540967867781e-07, "loss": 0.014, "reward": 6.613084226846695, "reward_std": 0.4508566010626964, "rewards/concensus_correctness_reward_func": 2.167749986052513, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.1875, "rewards/question_recreation_reward_func": 0.7734592199558392, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.109375, "step": 78 }, { "completion_length": 175.90625, "epoch": 2.2857142857142856, "grad_norm": 77835.71875, "kl": 3964.1012081073877, "learning_rate": 6.894700159171534e-07, "loss": 3.9641, "reward": 5.97839692234993, "reward_std": 0.07155295684788143, "rewards/concensus_correctness_reward_func": 1.918999969959259, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.7429906615288928, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.328125, "rewards/xmlcount_reward_func": 1.11328125, "step": 80 }, { "completion_length": 179.65625, "epoch": 2.342857142857143, "grad_norm": 125543.6640625, "kl": 9835.70870141859, "learning_rate": 6.743872071649411e-07, "loss": 9.8357, "reward": 6.020951867103577, "reward_std": 0.06402797414375527, "rewards/concensus_correctness_reward_func": 1.9197499826550484, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.7222956451587379, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.328125, "rewards/xmlcount_reward_func": 1.05078125, "step": 82 }, { "completion_length": 232.0, "epoch": 2.4, "grad_norm": 1412.854736328125, "kl": 446.583844974978, "learning_rate": 6.59121490313722e-07, "loss": 0.4466, "reward": 6.281670957803726, "reward_std": 0.18979688500985503, "rewards/concensus_correctness_reward_func": 2.004124980419874, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.749139740364626, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.34375, "rewards/xmlcount_reward_func": 1.0596562512218952, "step": 84 }, { "completion_length": 118.21875, "epoch": 2.4571428571428573, "grad_norm": 61.02229309082031, "kl": 31.27182410069142, "learning_rate": 6.436888769924141e-07, "loss": 0.0313, "reward": 6.057249546051025, "reward_std": 0.34927656571380794, "rewards/concensus_correctness_reward_func": 1.95499999076128, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.7006870835321024, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.34375, "rewards/xmlcount_reward_func": 1.0578125044703484, "step": 86 }, { "completion_length": 118.75, "epoch": 2.5142857142857142, "grad_norm": 14.912981986999512, "kl": 3.7455607797385255, "learning_rate": 6.281055538812861e-07, "loss": 0.0037, "reward": 6.8723739087581635, "reward_std": 0.13205374777317047, "rewards/concensus_correctness_reward_func": 1.8407499808818102, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.375, "rewards/question_recreation_reward_func": 0.9721551463007927, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.215718749910593, "step": 88 }, { "completion_length": 153.84375, "epoch": 2.571428571428571, "grad_norm": 0.0025314288213849068, "kl": 8.448847692096706, "learning_rate": 6.123878657343647e-07, "loss": 0.0084, "reward": 6.656649470329285, "reward_std": 0.08741115219891071, "rewards/concensus_correctness_reward_func": 2.0427499935030937, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8717119987122715, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.1953125, "step": 90 }, { "completion_length": 251.75, "epoch": 2.6285714285714286, "grad_norm": 118.81327056884766, "kl": 13.05496934192388, "learning_rate": 5.96552298236044e-07, "loss": 0.0131, "reward": 6.694221943616867, "reward_std": 0.4122789396205917, "rewards/concensus_correctness_reward_func": 2.2352499924600124, "rewards/consensus_reward_func": 1.9375, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.375, "rewards/question_recreation_reward_func": 0.7167846110532992, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.34375, "rewards/xmlcount_reward_func": 1.0859375, "step": 92 }, { "completion_length": 163.65625, "epoch": 2.685714285714286, "grad_norm": 87.89823913574219, "kl": 2.4992747763023715, "learning_rate": 5.806154607098799e-07, "loss": 0.0025, "reward": 6.533147543668747, "reward_std": 0.003381799964699894, "rewards/concensus_correctness_reward_func": 1.9176249727606773, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.9358350443653762, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.2109375, "step": 94 }, { "completion_length": 153.03125, "epoch": 2.742857142857143, "grad_norm": 400.9477844238281, "kl": 38.25428117752017, "learning_rate": 5.645940686977032e-07, "loss": 0.0383, "reward": 6.829828351736069, "reward_std": 0.3507880922406912, "rewards/concensus_correctness_reward_func": 2.167749986052513, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.8925471268594265, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.14453125, "step": 96 }, { "completion_length": 144.6875, "epoch": 2.8, "grad_norm": 55.653812408447266, "kl": 107186.28675935966, "learning_rate": 5.485049264273241e-07, "loss": 107.1863, "reward": 6.462806463241577, "reward_std": 0.14959648743388243, "rewards/concensus_correctness_reward_func": 1.9177499786019325, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.9200564883649349, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4375, "rewards/xmlcount_reward_func": 1.1875, "step": 98 }, { "completion_length": 173.96875, "epoch": 2.857142857142857, "grad_norm": 77.20722198486328, "kl": 11.116460184275184, "learning_rate": 5.323649091872178e-07, "loss": 0.0111, "reward": 6.238676607608795, "reward_std": 0.16680637560784817, "rewards/concensus_correctness_reward_func": 1.9146249815821648, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.88452033046633, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.1426562517881393, "step": 100 }, { "completion_length": 127.15625, "epoch": 2.914285714285714, "grad_norm": 18.17213249206543, "kl": 9.415469576871885, "learning_rate": 5.16190945626678e-07, "loss": 0.0094, "reward": 6.7656005918979645, "reward_std": 0.1332016959786415, "rewards/concensus_correctness_reward_func": 2.0424999743700027, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.9379443395882845, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.453125, "rewards/xmlcount_reward_func": 1.20703125, "step": 102 }, { "completion_length": 110.46875, "epoch": 2.9714285714285715, "grad_norm": 22.680030822753906, "kl": 16.711946348986203, "learning_rate": 5e-07, "loss": 0.0167, "reward": 6.9859276711940765, "reward_std": 0.1056629114318639, "rewards/concensus_correctness_reward_func": 2.169499985873699, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.9140838906168938, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.453125, "rewards/xmlcount_reward_func": 1.19921875, "step": 104 }, { "completion_length": 137.53125, "epoch": 3.0285714285714285, "grad_norm": 36.1477165222168, "kl": 13.781904241098744, "learning_rate": 4.838090543733221e-07, "loss": 0.0138, "reward": 6.522283971309662, "reward_std": 0.15269813538179733, "rewards/concensus_correctness_reward_func": 2.0437499657273293, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8847840757807717, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.1875, "step": 106 }, { "completion_length": 124.65625, "epoch": 3.085714285714286, "grad_norm": 0.24235893785953522, "kl": 146.02455591049488, "learning_rate": 4.676350908127821e-07, "loss": 0.146, "reward": 6.913030564785004, "reward_std": 0.17469462309964, "rewards/concensus_correctness_reward_func": 2.167249970138073, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.886405548080802, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4375, "rewards/xmlcount_reward_func": 1.171875, "step": 108 }, { "completion_length": 124.09375, "epoch": 3.142857142857143, "grad_norm": 14.452537536621094, "kl": 9.019021915784492, "learning_rate": 4.5149507357267597e-07, "loss": 0.009, "reward": 6.673075348138809, "reward_std": 0.23445336427539587, "rewards/concensus_correctness_reward_func": 2.043874979019165, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8596691160928458, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.453125, "rewards/xmlcount_reward_func": 1.19140625, "step": 110 }, { "completion_length": 142.78125, "epoch": 3.2, "grad_norm": 23.269039154052734, "kl": 102600.93675793686, "learning_rate": 4.354059313022969e-07, "loss": 102.6009, "reward": 6.444055080413818, "reward_std": 0.13274624338373542, "rewards/concensus_correctness_reward_func": 1.918249972164631, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.9086175980046391, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4375, "rewards/xmlcount_reward_func": 1.1796875, "step": 112 }, { "completion_length": 156.375, "epoch": 3.257142857142857, "grad_norm": 0.0024780540261417627, "kl": 10458499.638899168, "learning_rate": 4.193845392901201e-07, "loss": 10458.5, "reward": 6.469187945127487, "reward_std": 0.13797323405742645, "rewards/concensus_correctness_reward_func": 1.9172499775886536, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.9074067602632567, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.453125, "rewards/xmlcount_reward_func": 1.19140625, "step": 114 }, { "completion_length": 120.1875, "epoch": 3.314285714285714, "grad_norm": 3.623257875442505, "kl": 4.9637036175093385, "learning_rate": 4.0344770176395606e-07, "loss": 0.005, "reward": 6.929087281227112, "reward_std": 0.05713449977338314, "rewards/concensus_correctness_reward_func": 2.166124999523163, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.9309310428798199, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.23828125, "step": 116 }, { "completion_length": 164.125, "epoch": 3.3714285714285714, "grad_norm": 57468.05859375, "kl": 7235.024369265923, "learning_rate": 3.8761213426563543e-07, "loss": 7.235, "reward": 6.663719207048416, "reward_std": 0.16646834812127054, "rewards/concensus_correctness_reward_func": 2.0431249737739563, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8979379505617544, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.17578125, "step": 118 }, { "completion_length": 118.5, "epoch": 3.4285714285714284, "grad_norm": 520.0580444335938, "kl": 67.83392252838627, "learning_rate": 3.718944461187138e-07, "loss": 0.0678, "reward": 6.547982841730118, "reward_std": 0.14045805705245584, "rewards/concensus_correctness_reward_func": 1.9401249885559082, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8695765999145806, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4375, "rewards/xmlcount_reward_func": 1.17578125, "step": 120 }, { "completion_length": 172.46875, "epoch": 3.4857142857142858, "grad_norm": 136.3828887939453, "kl": 66.3491708847132, "learning_rate": 3.563111230075859e-07, "loss": 0.0664, "reward": 6.197802782058716, "reward_std": 0.08567016178858466, "rewards/concensus_correctness_reward_func": 1.8532499894499779, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8328340587904677, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.390625, "rewards/xmlcount_reward_func": 1.12109375, "step": 122 }, { "completion_length": 165.90625, "epoch": 3.5428571428571427, "grad_norm": 0.05435093119740486, "kl": 150.12784510711583, "learning_rate": 3.408785096862782e-07, "loss": 0.1501, "reward": 6.551819205284119, "reward_std": 0.015193770406767726, "rewards/concensus_correctness_reward_func": 2.0423749908804893, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8338817218318582, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.1443125009536743, "step": 124 }, { "completion_length": 119.78125, "epoch": 3.6, "grad_norm": 0.8537346124649048, "kl": 0.3934392740372914, "learning_rate": 3.2561279283505884e-07, "loss": 0.0004, "reward": 7.02690127491951, "reward_std": 0.026761331711895764, "rewards/concensus_correctness_reward_func": 2.0481249764561653, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.9951512552797794, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.484375, "rewards/xmlcount_reward_func": 1.2492499947547913, "step": 126 }, { "completion_length": 151.15625, "epoch": 3.657142857142857, "grad_norm": 0.0007187644951045513, "kl": 35.26266118535864, "learning_rate": 3.105299840828466e-07, "loss": 0.0353, "reward": 6.682261526584625, "reward_std": 0.03159405663609505, "rewards/concensus_correctness_reward_func": 2.044749967753887, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8836052501574159, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4375, "rewards/xmlcount_reward_func": 1.19140625, "step": 128 }, { "completion_length": 175.4375, "epoch": 3.7142857142857144, "grad_norm": 4472.33447265625, "kl": 407.4183887737163, "learning_rate": 2.95645903213222e-07, "loss": 0.4074, "reward": 6.523717999458313, "reward_std": 0.23186340555548668, "rewards/concensus_correctness_reward_func": 1.9444999862462282, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8487492590211332, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4375, "rewards/xmlcount_reward_func": 1.16796875, "step": 130 }, { "completion_length": 119.59375, "epoch": 3.7714285714285714, "grad_norm": 14.829834938049316, "kl": 43.95136128488707, "learning_rate": 2.8097616157165885e-07, "loss": 0.044, "reward": 6.873107701539993, "reward_std": 0.1839166151767131, "rewards/concensus_correctness_reward_func": 2.2936249896883965, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.375, "rewards/question_recreation_reward_func": 0.832482704310678, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.1220000013709068, "step": 132 }, { "completion_length": 132.0625, "epoch": 3.8285714285714287, "grad_norm": 10.728759765625, "kl": 14.55994114780924, "learning_rate": 2.665361456913797e-07, "loss": 0.0146, "reward": 6.96189820766449, "reward_std": 0.15106727764941752, "rewards/concensus_correctness_reward_func": 2.204374983906746, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.375, "rewards/question_recreation_reward_func": 0.8239294798113406, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.390625, "rewards/xmlcount_reward_func": 1.16796875, "step": 134 }, { "completion_length": 167.0625, "epoch": 3.8857142857142857, "grad_norm": 38.03074264526367, "kl": 7.051730461721945, "learning_rate": 2.523410011550064e-07, "loss": 0.0071, "reward": 6.516344875097275, "reward_std": 0.12807103618979454, "rewards/concensus_correctness_reward_func": 1.9381249845027924, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8555636611999944, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.17578125, "step": 136 }, { "completion_length": 118.5625, "epoch": 3.942857142857143, "grad_norm": 28.598356246948242, "kl": 17.839532846181896, "learning_rate": 2.3840561670893495e-07, "loss": 0.0178, "reward": 6.337287962436676, "reward_std": 0.29263460593210766, "rewards/concensus_correctness_reward_func": 1.9114999696612358, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8437567495275289, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.17578125, "step": 138 }, { "completion_length": 134.5625, "epoch": 4.0, "grad_norm": 62.3493537902832, "kl": 14.68970625225824, "learning_rate": 2.247446086470982e-07, "loss": 0.0147, "reward": 6.461593389511108, "reward_std": 0.2324538507964462, "rewards/concensus_correctness_reward_func": 2.0439999774098396, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0625, "rewards/question_recreation_reward_func": 0.8629059121012688, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4375, "rewards/xmlcount_reward_func": 1.1796875, "step": 140 }, { "completion_length": 108.59375, "epoch": 4.057142857142857, "grad_norm": 23.657360076904297, "kl": 0.6967418518526642, "learning_rate": 2.113723054804904e-07, "loss": 0.0007, "reward": 6.604361176490784, "reward_std": 0.08363489806652069, "rewards/concensus_correctness_reward_func": 1.9134999737143517, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.9760174416005611, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.484375, "rewards/xmlcount_reward_func": 1.23046875, "step": 142 }, { "completion_length": 168.4375, "epoch": 4.114285714285714, "grad_norm": 16.927154541015625, "kl": 3.9155555481934243, "learning_rate": 1.9830273290853766e-07, "loss": 0.0039, "reward": 6.76575893163681, "reward_std": 0.10381055303150788, "rewards/concensus_correctness_reward_func": 2.1677499637007713, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.8870714246877469, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.1640625, "step": 144 }, { "completion_length": 207.34375, "epoch": 4.171428571428572, "grad_norm": 230.26246643066406, "kl": 35.77139102048136, "learning_rate": 1.8554959910807772e-07, "loss": 0.0358, "reward": 6.17943412065506, "reward_std": 0.5001923546660691, "rewards/concensus_correctness_reward_func": 1.8548749722540379, "rewards/consensus_reward_func": 1.9375, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8440904202871025, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.13671875, "step": 146 }, { "completion_length": 179.96875, "epoch": 4.228571428571429, "grad_norm": 608.2833862304688, "kl": 41.165939685676896, "learning_rate": 1.7312628035537386e-07, "loss": 0.0412, "reward": 6.023386672139168, "reward_std": 0.5373760673683137, "rewards/concensus_correctness_reward_func": 1.79399998113513, "rewards/consensus_reward_func": 1.9375, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8114491496235132, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.1054375022649765, "step": 148 }, { "completion_length": 119.25, "epoch": 4.285714285714286, "grad_norm": 156.11758422851562, "kl": 14.033162109145678, "learning_rate": 1.6104580699624837e-07, "loss": 0.014, "reward": 6.490307152271271, "reward_std": 0.4207246375735849, "rewards/concensus_correctness_reward_func": 1.9844999872148037, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0625, "rewards/question_recreation_reward_func": 0.9474009126424789, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.1990312486886978, "step": 150 }, { "completion_length": 166.75, "epoch": 4.3428571428571425, "grad_norm": 59.2918815612793, "kl": 15.187666200303994, "learning_rate": 1.493208497790504e-07, "loss": 0.0152, "reward": 6.442292809486389, "reward_std": 0.08591883815824986, "rewards/concensus_correctness_reward_func": 1.9396249949932098, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.834699070896022, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.13671875, "step": 152 }, { "completion_length": 151.84375, "epoch": 4.4, "grad_norm": 236.9644775390625, "kl": 35.47643838436477, "learning_rate": 1.3796370656478934e-07, "loss": 0.0355, "reward": 6.461767286062241, "reward_std": 0.4090730301104486, "rewards/concensus_correctness_reward_func": 2.0437499657273293, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8281734781339765, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.359375, "rewards/xmlcount_reward_func": 1.10546875, "step": 154 }, { "completion_length": 141.0, "epoch": 4.457142857142857, "grad_norm": 6.475616931915283, "kl": 607.9583089863813, "learning_rate": 1.2698628942837697e-07, "loss": 0.608, "reward": 6.360057026147842, "reward_std": 0.10659515461156843, "rewards/concensus_correctness_reward_func": 1.9188749715685844, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.9255570247769356, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4375, "rewards/xmlcount_reward_func": 1.203125, "step": 156 }, { "completion_length": 152.3125, "epoch": 4.514285714285714, "grad_norm": 2432.385986328125, "kl": 223.98980636267672, "learning_rate": 1.1640011216450691e-07, "loss": 0.224, "reward": 6.682069569826126, "reward_std": 0.14451245963573456, "rewards/concensus_correctness_reward_func": 2.0427499935030937, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8815071193967015, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.453125, "rewards/xmlcount_reward_func": 1.1796875, "step": 158 }, { "completion_length": 170.28125, "epoch": 4.571428571428571, "grad_norm": 66.50875091552734, "kl": 179.85483134612986, "learning_rate": 1.0621627821127288e-07, "loss": 0.1799, "reward": 6.952862948179245, "reward_std": 0.34136643912643194, "rewards/concensus_correctness_reward_func": 2.2936249896883965, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.3125, "rewards/question_recreation_reward_func": 0.7928941965801641, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.1475937515497208, "step": 160 }, { "completion_length": 150.375, "epoch": 4.628571428571428, "grad_norm": 69454.9453125, "kl": 3160.4802792127484, "learning_rate": 9.644546900419531e-08, "loss": 3.1605, "reward": 7.1283751130104065, "reward_std": 0.16841159365139902, "rewards/concensus_correctness_reward_func": 2.2938749864697456, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.375, "rewards/question_recreation_reward_func": 0.8977188635617495, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.1399062499403954, "step": 162 }, { "completion_length": 141.03125, "epoch": 4.685714285714286, "grad_norm": 0.005350879859179258, "kl": 40.605473204805264, "learning_rate": 8.70979327728718e-08, "loss": 0.0406, "reward": 6.543893337249756, "reward_std": 0.008683009073138237, "rewards/concensus_correctness_reward_func": 1.9179999753832817, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.9462057938799262, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.2109375, "step": 164 }, { "completion_length": 172.5, "epoch": 4.742857142857143, "grad_norm": 842.9741821289062, "kl": 79.63627462068501, "learning_rate": 7.81834737919978e-08, "loss": 0.0796, "reward": 6.854692578315735, "reward_std": 0.287606913712807, "rewards/concensus_correctness_reward_func": 2.2923749834299088, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.7849738348741084, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.390625, "rewards/xmlcount_reward_func": 1.13671875, "step": 166 }, { "completion_length": 206.53125, "epoch": 4.8, "grad_norm": 32.29534912109375, "kl": 25.739543075352145, "learning_rate": 6.971144209803736e-08, "loss": 0.0257, "reward": 6.660571306943893, "reward_std": 0.22189393499866128, "rewards/concensus_correctness_reward_func": 2.1687499657273293, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.7823212522780523, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.0844999998807907, "step": 168 }, { "completion_length": 158.78125, "epoch": 4.857142857142857, "grad_norm": 48.24919891357422, "kl": 16.90938147829354, "learning_rate": 6.16907236823262e-08, "loss": 0.0169, "reward": 6.540567606687546, "reward_std": 0.11463689978700131, "rewards/concensus_correctness_reward_func": 2.076999992132187, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.8065051874145865, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.390625, "rewards/xmlcount_reward_func": 1.1414375007152557, "step": 170 }, { "completion_length": 140.40625, "epoch": 4.914285714285715, "grad_norm": 1.2339164018630981, "kl": 38.32704227172388, "learning_rate": 5.412973117089287e-08, "loss": 0.0383, "reward": 6.7995350658893585, "reward_std": 0.22941981069743633, "rewards/concensus_correctness_reward_func": 1.9438749924302101, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.3125, "rewards/question_recreation_reward_func": 0.8947226200252771, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.453125, "rewards/xmlcount_reward_func": 1.1953125, "step": 172 }, { "completion_length": 204.125, "epoch": 4.9714285714285715, "grad_norm": 83.86946868896484, "kl": 16.977694245022008, "learning_rate": 4.703639500077655e-08, "loss": 0.017, "reward": 6.165009528398514, "reward_std": 0.3757715789251961, "rewards/concensus_correctness_reward_func": 1.818874979391694, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8160095855710097, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.1551249995827675, "step": 174 }, { "completion_length": 248.21875, "epoch": 5.0285714285714285, "grad_norm": 78.45205688476562, "kl": 123.29459129135614, "learning_rate": 4.041815510209395e-08, "loss": 0.1233, "reward": 6.23151370882988, "reward_std": 0.30146771227009594, "rewards/concensus_correctness_reward_func": 1.9547499865293503, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.7259825635701418, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.34375, "rewards/xmlcount_reward_func": 1.08203125, "step": 176 }, { "completion_length": 115.59375, "epoch": 5.085714285714285, "grad_norm": 30.307924270629883, "kl": 5.359901059065578, "learning_rate": 3.4281953094578875e-08, "loss": 0.0054, "reward": 6.928922921419144, "reward_std": 0.0711733426360297, "rewards/concensus_correctness_reward_func": 2.1689999848604202, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.892735461005941, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.1953125, "step": 178 }, { "completion_length": 155.28125, "epoch": 5.142857142857143, "grad_norm": 733.8907470703125, "kl": 27.44156739033167, "learning_rate": 2.8634225006782864e-08, "loss": 0.0274, "reward": 6.603150278329849, "reward_std": 0.13888305581349414, "rewards/concensus_correctness_reward_func": 2.042874976992607, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8571502412669361, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.171875, "step": 180 }, { "completion_length": 210.46875, "epoch": 5.2, "grad_norm": 193.40077209472656, "kl": 32.62125267219369, "learning_rate": 2.348089452556956e-08, "loss": 0.0326, "reward": 6.1919485330581665, "reward_std": 0.3429363896648283, "rewards/concensus_correctness_reward_func": 2.0476249754428864, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.7224485471379012, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.328125, "rewards/xmlcount_reward_func": 1.09375, "step": 182 }, { "completion_length": 178.65625, "epoch": 5.257142857142857, "grad_norm": 195.80307006835938, "kl": 32.1218818299501, "learning_rate": 1.882736678298491e-08, "loss": 0.0321, "reward": 6.4709332883358, "reward_std": 0.20533737423829734, "rewards/concensus_correctness_reward_func": 2.041749984025955, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8159645728301257, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.390625, "rewards/xmlcount_reward_func": 1.0975937489420176, "step": 184 }, { "completion_length": 202.9375, "epoch": 5.314285714285714, "grad_norm": 6.5529704093933105, "kl": 21.807212316646257, "learning_rate": 1.4678522687020412e-08, "loss": 0.0218, "reward": 6.331148684024811, "reward_std": 0.04444863580283709, "rewards/concensus_correctness_reward_func": 1.9174999818205833, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8277112208306789, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.421875, "rewards/xmlcount_reward_func": 1.1640625, "step": 186 }, { "completion_length": 183.84375, "epoch": 5.371428571428572, "grad_norm": 14.336308479309082, "kl": 11.637659553496746, "learning_rate": 1.1038713802214717e-08, "loss": 0.0116, "reward": 6.26752108335495, "reward_std": 0.21460022201063111, "rewards/concensus_correctness_reward_func": 2.1094999872148037, "rewards/consensus_reward_func": 1.8125, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.7791148920659907, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.34375, "rewards/xmlcount_reward_func": 1.09765625, "step": 188 }, { "completion_length": 143.65625, "epoch": 5.428571428571429, "grad_norm": 35.581268310546875, "kl": 22.53030281640372, "learning_rate": 7.91175778546288e-09, "loss": 0.0225, "reward": 6.6998831778764725, "reward_std": 0.5453806091099977, "rewards/concensus_correctness_reward_func": 2.2333749793469906, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.3125, "rewards/question_recreation_reward_func": 0.7986019123345613, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.375, "rewards/xmlcount_reward_func": 1.1054062508046627, "step": 190 }, { "completion_length": 126.28125, "epoch": 5.485714285714286, "grad_norm": 1261.33740234375, "kl": 216.40358931079027, "learning_rate": 5.3009343818219975e-09, "loss": 0.2164, "reward": 6.315451234579086, "reward_std": 0.19842870347201824, "rewards/concensus_correctness_reward_func": 1.851999968290329, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.0, "rewards/question_recreation_reward_func": 0.8775137448683381, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.40625, "rewards/xmlcount_reward_func": 1.1796875, "step": 192 }, { "completion_length": 113.5625, "epoch": 5.542857142857143, "grad_norm": 38.23686981201172, "kl": 26.032929529420585, "learning_rate": 3.2089819845111944e-09, "loss": 0.026, "reward": 6.602267116308212, "reward_std": 0.1286412663757801, "rewards/concensus_correctness_reward_func": 2.169375002384186, "rewards/consensus_reward_func": 1.875, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.8703921169508249, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.390625, "rewards/xmlcount_reward_func": 1.171875, "step": 194 }, { "completion_length": 151.3125, "epoch": 5.6, "grad_norm": 0.003121931105852127, "kl": 1.4879173506217853, "learning_rate": 1.638094762715314e-09, "loss": 0.0015, "reward": 6.882100820541382, "reward_std": 0.09168865531682968, "rewards/concensus_correctness_reward_func": 1.9496249873191118, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.25, "rewards/question_recreation_reward_func": 0.9676321819424629, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.484375, "rewards/xmlcount_reward_func": 1.23046875, "step": 196 }, { "completion_length": 127.21875, "epoch": 5.6571428571428575, "grad_norm": 0.0015910969814285636, "kl": 1.242647759863985, "learning_rate": 5.899203602046654e-10, "loss": 0.0012, "reward": 6.5998228788375854, "reward_std": 0.05409337143646553, "rewards/concensus_correctness_reward_func": 1.8141249939799309, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.9653854127973318, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.2265625, "step": 198 }, { "completion_length": 109.0625, "epoch": 5.714285714285714, "grad_norm": 0.0088896993547678, "kl": 10.252777249065858, "learning_rate": 6.555816718389895e-11, "loss": 0.0103, "reward": 6.735629111528397, "reward_std": 0.09949736995622516, "rewards/concensus_correctness_reward_func": 2.0439999774098396, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 0.125, "rewards/question_recreation_reward_func": 0.9142853366211057, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.453125, "rewards/xmlcount_reward_func": 1.19921875, "step": 200 }, { "epoch": 5.714285714285714, "step": 200, "total_flos": 0.0, "train_loss": 149.93084794194365, "train_runtime": 1724.5009, "train_samples_per_second": 1.856, "train_steps_per_second": 0.116 } ], "logging_steps": 2, "max_steps": 200, "num_input_tokens_seen": 0, "num_train_epochs": 6, "save_steps": 25, "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": 2, "trial_name": null, "trial_params": null }