{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.999266862170088, "eval_steps": 500, "global_step": 1363, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "clip_ratio": 0.0, "completion_length": 509.375, "epoch": 0.0007331378299120235, "grad_norm": 2.6448973612837636, "kl": 0.0, "learning_rate": 7.2992700729927e-09, "loss": -0.0482, "reward": 0.68954798579216, "reward_std": 0.25520985573530197, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.263888917863369, "rewards/repetition_penalty_reward": -0.07434091717004776, "rewards/tag_count_reward": 0.5, "step": 1 }, { "clip_ratio": 0.0, "completion_length": 498.87501525878906, "epoch": 0.001466275659824047, "grad_norm": 2.76096812642669, "kl": 0.0, "learning_rate": 1.45985401459854e-08, "loss": -0.0176, "reward": 0.6545102000236511, "reward_std": 0.22549808025360107, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2083333507180214, "rewards/repetition_penalty_reward": -0.06423980556428432, "rewards/tag_count_reward": 0.5104166865348816, "step": 2 }, { "clip_ratio": 0.0, "completion_length": 488.4791717529297, "epoch": 0.0021994134897360706, "grad_norm": 2.7544190296661046, "kl": 0.0001468658447265625, "learning_rate": 2.1897810218978102e-08, "loss": -0.1206, "reward": 0.6828474402427673, "reward_std": 0.21180886030197144, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2430555745959282, "rewards/repetition_penalty_reward": -0.07062481716275215, "rewards/tag_count_reward": 0.5104166865348816, "step": 3 }, { "clip_ratio": 0.0, "completion_length": 490.04168701171875, "epoch": 0.002932551319648094, "grad_norm": 2.679496733955865, "kl": 0.00016498565673828125, "learning_rate": 2.91970802919708e-08, "loss": -0.0511, "reward": 0.7352340519428253, "reward_std": 0.3041670620441437, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.284722238779068, "rewards/repetition_penalty_reward": -0.07552988082170486, "rewards/tag_count_reward": 0.5260416865348816, "step": 4 }, { "clip_ratio": 0.0, "completion_length": 431.62501525878906, "epoch": 0.0036656891495601175, "grad_norm": 2.7710748320387943, "kl": 0.00017213821411132812, "learning_rate": 3.64963503649635e-08, "loss": -0.0125, "reward": 0.6680092215538025, "reward_std": 0.24565115571022034, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2152777835726738, "rewards/repetition_penalty_reward": -0.06289358995854855, "rewards/tag_count_reward": 0.515625, "step": 5 }, { "clip_ratio": 0.0, "completion_length": 561.625, "epoch": 0.004398826979472141, "grad_norm": 2.6419901655907245, "kl": 0.00017642974853515625, "learning_rate": 4.3795620437956203e-08, "loss": -0.0065, "reward": 0.7031907737255096, "reward_std": 0.24719806760549545, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2777777761220932, "rewards/repetition_penalty_reward": -0.07458702102303505, "rewards/tag_count_reward": 0.5, "step": 6 }, { "clip_ratio": 0.0, "completion_length": 471.1458435058594, "epoch": 0.005131964809384164, "grad_norm": 2.575204146630783, "kl": 0.00015974044799804688, "learning_rate": 5.10948905109489e-08, "loss": 0.0184, "reward": 0.5842337310314178, "reward_std": 0.22514298558235168, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.166666679084301, "rewards/repetition_penalty_reward": -0.08243293687701225, "rewards/tag_count_reward": 0.5, "step": 7 }, { "clip_ratio": 0.0, "completion_length": 457.72918701171875, "epoch": 0.005865102639296188, "grad_norm": 2.7932110624723743, "kl": 0.00015878677368164062, "learning_rate": 5.83941605839416e-08, "loss": 0.0322, "reward": 0.5849673748016357, "reward_std": 0.23537325859069824, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1666666716337204, "rewards/repetition_penalty_reward": -0.08169934153556824, "rewards/tag_count_reward": 0.5, "step": 8 }, { "clip_ratio": 0.0, "completion_length": 468.81251525878906, "epoch": 0.006598240469208211, "grad_norm": 3.1047208356259914, "kl": 0.00018215179443359375, "learning_rate": 6.569343065693431e-08, "loss": -0.0461, "reward": 0.6473296880722046, "reward_std": 0.25502097606658936, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.208333358168602, "rewards/repetition_penalty_reward": -0.07142036035656929, "rewards/tag_count_reward": 0.5104166865348816, "step": 9 }, { "clip_ratio": 0.0, "completion_length": 458.18751525878906, "epoch": 0.007331378299120235, "grad_norm": 2.627504343626485, "kl": 0.00014257431030273438, "learning_rate": 7.2992700729927e-08, "loss": -0.0509, "reward": 0.5702269971370697, "reward_std": 0.17610464990139008, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1458333358168602, "rewards/repetition_penalty_reward": -0.08602304756641388, "rewards/tag_count_reward": 0.5104166865348816, "step": 10 }, { "clip_ratio": 0.0, "completion_length": 529.75, "epoch": 0.008064516129032258, "grad_norm": 2.551343107524344, "kl": 0.00016307830810546875, "learning_rate": 8.029197080291971e-08, "loss": -0.0545, "reward": 0.6536203324794769, "reward_std": 0.2523871883749962, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.243055559694767, "rewards/repetition_penalty_reward": -0.08943524211645126, "rewards/tag_count_reward": 0.5, "step": 11 }, { "clip_ratio": 0.0, "completion_length": 445.2291717529297, "epoch": 0.008797653958944282, "grad_norm": 3.0366791357232406, "kl": 0.00017404556274414062, "learning_rate": 8.759124087591241e-08, "loss": 0.0279, "reward": 0.583998829126358, "reward_std": 0.21335972100496292, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1458333432674408, "rewards/repetition_penalty_reward": -0.06183452531695366, "rewards/tag_count_reward": 0.5, "step": 12 }, { "clip_ratio": 0.0, "completion_length": 474.5625, "epoch": 0.009530791788856305, "grad_norm": 3.1395073617735614, "kl": 0.00017595291137695312, "learning_rate": 9.48905109489051e-08, "loss": -0.1077, "reward": 0.5609289705753326, "reward_std": 0.21441183984279633, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1250000037252903, "rewards/repetition_penalty_reward": -0.06407104432582855, "rewards/tag_count_reward": 0.5, "step": 13 }, { "clip_ratio": 0.0, "completion_length": 485.56251525878906, "epoch": 0.010263929618768328, "grad_norm": 2.758079388838085, "kl": 0.00017261505126953125, "learning_rate": 1.021897810218978e-07, "loss": -0.0345, "reward": 0.5964665710926056, "reward_std": 0.17032359540462494, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1597222313284874, "rewards/repetition_penalty_reward": -0.07367238402366638, "rewards/tag_count_reward": 0.5104166865348816, "step": 14 }, { "clip_ratio": 0.0, "completion_length": 470.3125, "epoch": 0.010997067448680353, "grad_norm": 2.605821618375983, "kl": 0.00021219253540039062, "learning_rate": 1.0948905109489052e-07, "loss": -0.0017, "reward": 0.715076208114624, "reward_std": 0.2355574294924736, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2777777910232544, "rewards/repetition_penalty_reward": -0.06270160153508186, "rewards/tag_count_reward": 0.5, "step": 15 }, { "clip_ratio": 0.0, "completion_length": 483.5833435058594, "epoch": 0.011730205278592375, "grad_norm": 2.722382736115358, "kl": 0.00017833709716796875, "learning_rate": 1.167883211678832e-07, "loss": -0.0496, "reward": 0.621931403875351, "reward_std": 0.23212532699108124, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1875000149011612, "rewards/repetition_penalty_reward": -0.07598524540662766, "rewards/tag_count_reward": 0.5104166865348816, "step": 16 }, { "clip_ratio": 0.0, "completion_length": 509.0625305175781, "epoch": 0.012463343108504398, "grad_norm": 2.7955490988819687, "kl": 0.00018405914306640625, "learning_rate": 1.240875912408759e-07, "loss": -0.0281, "reward": 0.6549810469150543, "reward_std": 0.27128641307353973, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.222222238779068, "rewards/repetition_penalty_reward": -0.06724120303988457, "rewards/tag_count_reward": 0.5, "step": 17 }, { "clip_ratio": 0.0, "completion_length": 518.6041870117188, "epoch": 0.013196480938416423, "grad_norm": 2.723325734305699, "kl": 0.00018548965454101562, "learning_rate": 1.3138686131386862e-07, "loss": -0.0081, "reward": 0.584599107503891, "reward_std": 0.20393361896276474, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1527777872979641, "rewards/repetition_penalty_reward": -0.06817871518433094, "rewards/tag_count_reward": 0.5, "step": 18 }, { "clip_ratio": 0.0, "completion_length": 409.0, "epoch": 0.013929618768328446, "grad_norm": 4.779504705709629, "kl": 0.00019550323486328125, "learning_rate": 1.386861313868613e-07, "loss": 0.0876, "reward": 0.5774352252483368, "reward_std": 0.16354358941316605, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1527777835726738, "rewards/repetition_penalty_reward": -0.0753425844013691, "rewards/tag_count_reward": 0.5, "step": 19 }, { "clip_ratio": 0.0, "completion_length": 485.2083435058594, "epoch": 0.01466275659824047, "grad_norm": 3.0626779341878616, "kl": 0.00018215179443359375, "learning_rate": 1.45985401459854e-07, "loss": -0.0717, "reward": 0.5419801771640778, "reward_std": 0.16230151802301407, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1180555634200573, "rewards/repetition_penalty_reward": -0.07607538625597954, "rewards/tag_count_reward": 0.5, "step": 20 }, { "clip_ratio": 0.0, "completion_length": 527.2083435058594, "epoch": 0.015395894428152493, "grad_norm": 2.413897203039519, "kl": 0.0001659393310546875, "learning_rate": 1.532846715328467e-07, "loss": 0.0023, "reward": 0.5849822759628296, "reward_std": 0.27357739210128784, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1805555671453476, "rewards/repetition_penalty_reward": -0.09557333588600159, "rewards/tag_count_reward": 0.5, "step": 21 }, { "clip_ratio": 0.0, "completion_length": 486.56251525878906, "epoch": 0.016129032258064516, "grad_norm": 2.408879678071363, "kl": 0.00018405914306640625, "learning_rate": 1.6058394160583942e-07, "loss": 0.0098, "reward": 0.6279599368572235, "reward_std": 0.20339544862508774, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1944444552063942, "rewards/repetition_penalty_reward": -0.07690121605992317, "rewards/tag_count_reward": 0.5104166865348816, "step": 22 }, { "clip_ratio": 0.0, "completion_length": 555.5625305175781, "epoch": 0.01686217008797654, "grad_norm": 2.6019168865864306, "kl": 0.00014972686767578125, "learning_rate": 1.678832116788321e-07, "loss": -0.0692, "reward": 0.5722634494304657, "reward_std": 0.21821296960115433, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.166666679084301, "rewards/repetition_penalty_reward": -0.10481992736458778, "rewards/tag_count_reward": 0.5104166865348816, "step": 23 }, { "clip_ratio": 0.0, "completion_length": 467.3541717529297, "epoch": 0.017595307917888565, "grad_norm": 3.0166610353722354, "kl": 0.00017309188842773438, "learning_rate": 1.7518248175182481e-07, "loss": -0.0261, "reward": 0.6123618483543396, "reward_std": 0.19541333615779877, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.173611119389534, "rewards/repetition_penalty_reward": -0.06124930642545223, "rewards/tag_count_reward": 0.5, "step": 24 }, { "clip_ratio": 0.0, "completion_length": 428.9166717529297, "epoch": 0.018328445747800588, "grad_norm": 3.3196057185136856, "kl": 0.0001983642578125, "learning_rate": 1.824817518248175e-07, "loss": -0.0708, "reward": 0.6493120789527893, "reward_std": 0.24062193930149078, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2083333432674408, "rewards/repetition_penalty_reward": -0.06943796388804913, "rewards/tag_count_reward": 0.5104166865348816, "step": 25 }, { "clip_ratio": 0.0, "completion_length": 470.0625, "epoch": 0.01906158357771261, "grad_norm": 2.8082986672044177, "kl": 0.00018548965454101562, "learning_rate": 1.897810218978102e-07, "loss": -0.0244, "reward": 0.78911292552948, "reward_std": 0.3358127474784851, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.2986111044883728, "rewards/repetition_penalty_reward": -0.05637324042618275, "rewards/tag_count_reward": 0.5260416865348816, "step": 26 }, { "clip_ratio": 0.0, "completion_length": 495.79168701171875, "epoch": 0.019794721407624633, "grad_norm": 2.6824722929765668, "kl": 0.0001621246337890625, "learning_rate": 1.9708029197080292e-07, "loss": -0.0727, "reward": 0.5723992586135864, "reward_std": 0.2219652161002159, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.173611119389534, "rewards/repetition_penalty_reward": -0.10121183842420578, "rewards/tag_count_reward": 0.5, "step": 27 }, { "clip_ratio": 0.0, "completion_length": 533.0, "epoch": 0.020527859237536656, "grad_norm": 2.870814349622882, "kl": 0.0001850128173828125, "learning_rate": 2.043795620437956e-07, "loss": -0.0336, "reward": 0.6264206767082214, "reward_std": 0.21734951436519623, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2152777910232544, "rewards/repetition_penalty_reward": -0.08885711058974266, "rewards/tag_count_reward": 0.5, "step": 28 }, { "clip_ratio": 0.0, "completion_length": 496.7708435058594, "epoch": 0.02126099706744868, "grad_norm": 2.5069460426178423, "kl": 0.000171661376953125, "learning_rate": 2.116788321167883e-07, "loss": -0.0277, "reward": 0.6581361889839172, "reward_std": 0.23858975619077682, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2222222313284874, "rewards/repetition_penalty_reward": -0.07450271770358086, "rewards/tag_count_reward": 0.5104166865348816, "step": 29 }, { "clip_ratio": 0.0, "completion_length": 475.8750305175781, "epoch": 0.021994134897360705, "grad_norm": 2.648249474912757, "kl": 0.0001735687255859375, "learning_rate": 2.1897810218978103e-07, "loss": -0.0397, "reward": 0.6791504621505737, "reward_std": 0.2867739200592041, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2222222313284874, "rewards/repetition_penalty_reward": -0.07432174310088158, "rewards/tag_count_reward": 0.5312500298023224, "step": 30 }, { "clip_ratio": 0.0, "completion_length": 443.3958435058594, "epoch": 0.022727272727272728, "grad_norm": 2.890033559674892, "kl": 0.00019073486328125, "learning_rate": 2.2627737226277372e-07, "loss": -0.0641, "reward": 0.6335480809211731, "reward_std": 0.2339138686656952, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2013889104127884, "rewards/repetition_penalty_reward": -0.06784084439277649, "rewards/tag_count_reward": 0.5, "step": 31 }, { "clip_ratio": 0.0, "completion_length": 526.6250152587891, "epoch": 0.02346041055718475, "grad_norm": 2.5853064468226976, "kl": 0.00019741058349609375, "learning_rate": 2.335766423357664e-07, "loss": 0.0082, "reward": 0.6064836084842682, "reward_std": 0.21031419932842255, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1875000111758709, "rewards/repetition_penalty_reward": -0.08101639151573181, "rewards/tag_count_reward": 0.5, "step": 32 }, { "clip_ratio": 0.0, "completion_length": 443.12501525878906, "epoch": 0.024193548387096774, "grad_norm": 3.0609420622300325, "kl": 0.00016927719116210938, "learning_rate": 2.408759124087591e-07, "loss": -0.1695, "reward": 0.6632957458496094, "reward_std": 0.24054750055074692, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1944444589316845, "rewards/repetition_penalty_reward": -0.062398696318268776, "rewards/tag_count_reward": 0.5312500298023224, "step": 33 }, { "clip_ratio": 0.0, "completion_length": 528.1458435058594, "epoch": 0.024926686217008796, "grad_norm": 2.564707811238524, "kl": 0.00021886825561523438, "learning_rate": 2.481751824817518e-07, "loss": -0.0154, "reward": 0.7042053937911987, "reward_std": 0.2845487892627716, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2638888955116272, "rewards/repetition_penalty_reward": -0.07010022550821304, "rewards/tag_count_reward": 0.5104166865348816, "step": 34 }, { "clip_ratio": 0.0, "completion_length": 499.60418701171875, "epoch": 0.025659824046920823, "grad_norm": 2.646106845035658, "kl": 0.00020170211791992188, "learning_rate": 2.5547445255474454e-07, "loss": -0.0509, "reward": 0.7097957134246826, "reward_std": 0.21529126912355423, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2638889104127884, "rewards/repetition_penalty_reward": -0.07492654211819172, "rewards/tag_count_reward": 0.5208333730697632, "step": 35 }, { "clip_ratio": 0.0, "completion_length": 521.0625305175781, "epoch": 0.026392961876832845, "grad_norm": 2.856616633787257, "kl": 0.00019311904907226562, "learning_rate": 2.6277372262773725e-07, "loss": 0.0176, "reward": 0.6456855833530426, "reward_std": 0.27881423383951187, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.215277798473835, "rewards/repetition_penalty_reward": -0.08000890910625458, "rewards/tag_count_reward": 0.5104166865348816, "step": 36 }, { "clip_ratio": 0.0, "completion_length": 503.14585876464844, "epoch": 0.027126099706744868, "grad_norm": 2.665887051686472, "kl": 0.00020599365234375, "learning_rate": 2.700729927007299e-07, "loss": -0.0497, "reward": 0.6908453702926636, "reward_std": 0.23081645369529724, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2569444552063942, "rewards/repetition_penalty_reward": -0.07651573792099953, "rewards/tag_count_reward": 0.5104166865348816, "step": 37 }, { "clip_ratio": 0.0, "completion_length": 414.3958435058594, "epoch": 0.02785923753665689, "grad_norm": 2.9595822498097104, "kl": 0.00022125244140625, "learning_rate": 2.773722627737226e-07, "loss": -0.0005, "reward": 0.5840825140476227, "reward_std": 0.2131967693567276, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1527777872979641, "rewards/repetition_penalty_reward": -0.0686953105032444, "rewards/tag_count_reward": 0.5, "step": 38 }, { "clip_ratio": 0.0, "completion_length": 523.2500152587891, "epoch": 0.028592375366568914, "grad_norm": 2.71164611926968, "kl": 0.0002307891845703125, "learning_rate": 2.846715328467153e-07, "loss": -0.0273, "reward": 0.624500185251236, "reward_std": 0.23453929275274277, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1944444626569748, "rewards/repetition_penalty_reward": -0.08036095649003983, "rewards/tag_count_reward": 0.5104166865348816, "step": 39 }, { "clip_ratio": 0.0, "completion_length": 480.1458435058594, "epoch": 0.02932551319648094, "grad_norm": 2.8793970958085917, "kl": 0.0002193450927734375, "learning_rate": 2.91970802919708e-07, "loss": -0.0961, "reward": 0.7346545159816742, "reward_std": 0.2827245742082596, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.270833358168602, "rewards/repetition_penalty_reward": -0.057012153789401054, "rewards/tag_count_reward": 0.5208333730697632, "step": 40 }, { "clip_ratio": 0.0, "completion_length": 495.5208435058594, "epoch": 0.030058651026392963, "grad_norm": 2.7202320117264907, "kl": 0.00029754638671875, "learning_rate": 2.9927007299270075e-07, "loss": 0.0366, "reward": 0.618072509765625, "reward_std": 0.23927630484104156, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.166666679084301, "rewards/repetition_penalty_reward": -0.06942753493785858, "rewards/tag_count_reward": 0.5208333432674408, "step": 41 }, { "clip_ratio": 0.0, "completion_length": 563.6250305175781, "epoch": 0.030791788856304986, "grad_norm": 2.558370045223577, "kl": 0.000270843505859375, "learning_rate": 3.065693430656934e-07, "loss": -0.0406, "reward": 0.698936402797699, "reward_std": 0.2520362436771393, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.270833358168602, "rewards/repetition_penalty_reward": -0.08231363818049431, "rewards/tag_count_reward": 0.5104166865348816, "step": 42 }, { "clip_ratio": 0.0, "completion_length": 475.97918701171875, "epoch": 0.03152492668621701, "grad_norm": 2.727560907345565, "kl": 0.0002727508544921875, "learning_rate": 3.138686131386861e-07, "loss": -0.0998, "reward": 0.688592404127121, "reward_std": 0.2755677103996277, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2430555745959282, "rewards/repetition_penalty_reward": -0.07529650256037712, "rewards/tag_count_reward": 0.5208333432674408, "step": 43 }, { "clip_ratio": 0.0, "completion_length": 519.9791870117188, "epoch": 0.03225806451612903, "grad_norm": 2.6698796736155446, "kl": 0.0003490447998046875, "learning_rate": 3.2116788321167883e-07, "loss": -0.0122, "reward": 0.7924045026302338, "reward_std": 0.28541746735572815, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.333333358168602, "rewards/repetition_penalty_reward": -0.06176219508051872, "rewards/tag_count_reward": 0.5208333432674408, "step": 44 }, { "clip_ratio": 0.0, "completion_length": 504.58335876464844, "epoch": 0.032991202346041054, "grad_norm": 2.452278139401127, "kl": 0.00032806396484375, "learning_rate": 3.284671532846715e-07, "loss": -0.0191, "reward": 0.6062445044517517, "reward_std": 0.21765749156475067, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.1944444626569748, "rewards/repetition_penalty_reward": -0.08819994330406189, "rewards/tag_count_reward": 0.5, "step": 45 }, { "clip_ratio": 0.0, "completion_length": 495.10418701171875, "epoch": 0.03372434017595308, "grad_norm": 2.6384968933757307, "kl": 0.00043201446533203125, "learning_rate": 3.357664233576642e-07, "loss": -0.009, "reward": 0.7084953188896179, "reward_std": 0.29200321435928345, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2569444626569748, "rewards/repetition_penalty_reward": -0.07969915866851807, "rewards/tag_count_reward": 0.5312500298023224, "step": 46 }, { "clip_ratio": 0.0, "completion_length": 502.0208435058594, "epoch": 0.0344574780058651, "grad_norm": 2.662788904527241, "kl": 0.0004863739013671875, "learning_rate": 3.4306569343065697e-07, "loss": 0.0136, "reward": 0.6565942764282227, "reward_std": 0.2551596984267235, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2361111268401146, "rewards/repetition_penalty_reward": -0.10555854439735413, "rewards/tag_count_reward": 0.5260416865348816, "step": 47 }, { "clip_ratio": 0.0, "completion_length": 546.5208740234375, "epoch": 0.03519061583577713, "grad_norm": 2.7029202066676996, "kl": 0.0005130767822265625, "learning_rate": 3.5036496350364963e-07, "loss": -0.0636, "reward": 0.7792027294635773, "reward_std": 0.25155991315841675, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.347222238779068, "rewards/repetition_penalty_reward": -0.06801950931549072, "rewards/tag_count_reward": 0.5, "step": 48 }, { "clip_ratio": 0.0, "completion_length": 443.06251525878906, "epoch": 0.03592375366568915, "grad_norm": 2.7589483849725522, "kl": 0.0008373260498046875, "learning_rate": 3.5766423357664234e-07, "loss": -0.0326, "reward": 0.6618194282054901, "reward_std": 0.27763861417770386, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.25, "rewards/repetition_penalty_reward": -0.09859726577997208, "rewards/tag_count_reward": 0.5104166865348816, "step": 49 }, { "clip_ratio": 0.0, "completion_length": 462.0625305175781, "epoch": 0.036656891495601175, "grad_norm": 2.8648632394348303, "kl": 0.0007076263427734375, "learning_rate": 3.64963503649635e-07, "loss": -0.0441, "reward": 0.6885200440883636, "reward_std": 0.2172483429312706, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2777777835726738, "rewards/repetition_penalty_reward": -0.08925776183605194, "rewards/tag_count_reward": 0.5, "step": 50 }, { "clip_ratio": 0.0, "completion_length": 501.91668701171875, "epoch": 0.0373900293255132, "grad_norm": 2.7944345979797833, "kl": 0.0007419586181640625, "learning_rate": 3.722627737226277e-07, "loss": -0.0331, "reward": 0.7741267681121826, "reward_std": 0.26387324929237366, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.3402777761220932, "rewards/repetition_penalty_reward": -0.07656766846776009, "rewards/tag_count_reward": 0.5104166865348816, "step": 51 }, { "clip_ratio": 0.0, "completion_length": 472.18751525878906, "epoch": 0.03812316715542522, "grad_norm": 2.6423146438087497, "kl": 0.00083160400390625, "learning_rate": 3.795620437956204e-07, "loss": 0.0049, "reward": 0.7249100506305695, "reward_std": 0.2071620374917984, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.2986111342906952, "rewards/repetition_penalty_reward": -0.07370108738541603, "rewards/tag_count_reward": 0.5, "step": 52 }, { "clip_ratio": 0.0, "completion_length": 505.04168701171875, "epoch": 0.038856304985337244, "grad_norm": 2.6137555315943026, "kl": 0.001148223876953125, "learning_rate": 3.8686131386861313e-07, "loss": -0.0186, "reward": 0.7216832935810089, "reward_std": 0.2934323847293854, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.3055555671453476, "rewards/repetition_penalty_reward": -0.09428896009922028, "rewards/tag_count_reward": 0.5104166865348816, "step": 53 }, { "clip_ratio": 0.0, "completion_length": 489.9375, "epoch": 0.039589442815249266, "grad_norm": 2.9424116740549664, "kl": 0.00109100341796875, "learning_rate": 3.9416058394160584e-07, "loss": -0.022, "reward": 0.9018063545227051, "reward_std": 0.3424327075481415, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.4375000298023224, "rewards/repetition_penalty_reward": -0.06694366224110126, "rewards/tag_count_reward": 0.5104166865348816, "step": 54 }, { "clip_ratio": 0.0, "completion_length": 528.9583435058594, "epoch": 0.04032258064516129, "grad_norm": 2.4966957649374684, "kl": 0.0010433197021484375, "learning_rate": 4.0145985401459856e-07, "loss": -0.0116, "reward": 0.8005258738994598, "reward_std": 0.24090076982975006, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.4166667014360428, "rewards/repetition_penalty_reward": -0.11614080145955086, "rewards/tag_count_reward": 0.5, "step": 55 }, { "clip_ratio": 0.0, "completion_length": 517.4583740234375, "epoch": 0.04105571847507331, "grad_norm": 2.6094881117590223, "kl": 0.001262664794921875, "learning_rate": 4.087591240875912e-07, "loss": -0.0711, "reward": 0.6848195195198059, "reward_std": 0.23144569993019104, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.236111119389534, "rewards/repetition_penalty_reward": -0.07212491519749165, "rewards/tag_count_reward": 0.5208333432674408, "step": 56 }, { "clip_ratio": 0.0, "completion_length": 534.0000305175781, "epoch": 0.041788856304985335, "grad_norm": 2.5627446751027807, "kl": 0.001556396484375, "learning_rate": 4.160583941605839e-07, "loss": -0.0015, "reward": 0.8210827708244324, "reward_std": 0.26339250057935715, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.4027778059244156, "rewards/repetition_penalty_reward": -0.09211170673370361, "rewards/tag_count_reward": 0.5104166865348816, "step": 57 }, { "clip_ratio": 0.0, "completion_length": 521.6041870117188, "epoch": 0.04252199413489736, "grad_norm": 2.787620990751815, "kl": 0.001438140869140625, "learning_rate": 4.233576642335766e-07, "loss": 0.0411, "reward": 0.8132742345333099, "reward_std": 0.2763310372829437, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.3611111342906952, "rewards/repetition_penalty_reward": -0.07908686250448227, "rewards/tag_count_reward": 0.5312500298023224, "step": 58 }, { "clip_ratio": 0.0, "completion_length": 536.4375, "epoch": 0.04325513196480939, "grad_norm": 2.5844944772028726, "kl": 0.00209808349609375, "learning_rate": 4.306569343065693e-07, "loss": -0.1141, "reward": 0.7877600789070129, "reward_std": 0.23524152487516403, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.3750000298023224, "rewards/repetition_penalty_reward": -0.10807328298687935, "rewards/tag_count_reward": 0.5208333730697632, "step": 59 }, { "clip_ratio": 0.0, "completion_length": 526.8333435058594, "epoch": 0.04398826979472141, "grad_norm": 2.4453183077687943, "kl": 0.002227783203125, "learning_rate": 4.3795620437956206e-07, "loss": -0.0189, "reward": 0.8026378750801086, "reward_std": 0.2494141310453415, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.381944477558136, "rewards/repetition_penalty_reward": -0.08972325921058655, "rewards/tag_count_reward": 0.5104166865348816, "step": 60 }, { "clip_ratio": 0.0, "completion_length": 538.2916870117188, "epoch": 0.04472140762463343, "grad_norm": 2.496023501476716, "kl": 0.00217437744140625, "learning_rate": 4.452554744525547e-07, "loss": -0.0263, "reward": 0.7900694012641907, "reward_std": 0.19147495925426483, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.3888889104127884, "rewards/repetition_penalty_reward": -0.10402785614132881, "rewards/tag_count_reward": 0.5052083432674408, "step": 61 }, { "clip_ratio": 0.0, "completion_length": 502.1041717529297, "epoch": 0.045454545454545456, "grad_norm": 2.4902095639273654, "kl": 0.00255584716796875, "learning_rate": 4.5255474452554743e-07, "loss": 0.0029, "reward": 0.8510425984859467, "reward_std": 0.2535877972841263, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.4236111342906952, "rewards/repetition_penalty_reward": -0.07256851345300674, "rewards/tag_count_reward": 0.5, "step": 62 }, { "clip_ratio": 0.0, "completion_length": 469.2083435058594, "epoch": 0.04618768328445748, "grad_norm": 2.5574432170277417, "kl": 0.0037078857421875, "learning_rate": 4.5985401459854014e-07, "loss": -0.04, "reward": 0.7994670569896698, "reward_std": 0.21547260135412216, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.4027777910232544, "rewards/repetition_penalty_reward": -0.10331075266003609, "rewards/tag_count_reward": 0.5, "step": 63 }, { "clip_ratio": 0.0, "completion_length": 480.7083435058594, "epoch": 0.0469208211143695, "grad_norm": 2.725179825178671, "kl": 0.00246429443359375, "learning_rate": 4.671532846715328e-07, "loss": -0.0296, "reward": 0.9697330594062805, "reward_std": 0.295375719666481, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.5208333432674408, "rewards/repetition_penalty_reward": -0.07193369045853615, "rewards/tag_count_reward": 0.5208333432674408, "step": 64 }, { "clip_ratio": 0.0, "completion_length": 473.29168701171875, "epoch": 0.047653958944281524, "grad_norm": 2.7484533646327804, "kl": 0.005218505859375, "learning_rate": 4.744525547445255e-07, "loss": -0.0117, "reward": 0.8709261417388916, "reward_std": 0.3177572637796402, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.4305555522441864, "rewards/repetition_penalty_reward": -0.08046277612447739, "rewards/tag_count_reward": 0.5208333730697632, "step": 65 }, { "clip_ratio": 0.0, "completion_length": 528.1250305175781, "epoch": 0.04838709677419355, "grad_norm": 2.754230706652707, "kl": 0.0030059814453125, "learning_rate": 4.817518248175182e-07, "loss": 0.0394, "reward": 0.8668566048145294, "reward_std": 0.21845827251672745, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.423611119389534, "rewards/repetition_penalty_reward": -0.07758788764476776, "rewards/tag_count_reward": 0.5208333730697632, "step": 66 }, { "clip_ratio": 0.0, "completion_length": 464.5833435058594, "epoch": 0.04912023460410557, "grad_norm": 2.785853687350145, "kl": 0.00362396240234375, "learning_rate": 4.89051094890511e-07, "loss": -0.0255, "reward": 0.9353344142436981, "reward_std": 0.24234941601753235, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.4861111342906952, "rewards/repetition_penalty_reward": -0.07161007076501846, "rewards/tag_count_reward": 0.5208333432674408, "step": 67 }, { "clip_ratio": 0.0, "completion_length": 542.4583435058594, "epoch": 0.04985337243401759, "grad_norm": 2.430982163529705, "kl": 0.00330352783203125, "learning_rate": 4.963503649635036e-07, "loss": -0.0115, "reward": 0.9278749823570251, "reward_std": 0.19417473673820496, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.520833358168602, "rewards/repetition_penalty_reward": -0.09295839816331863, "rewards/tag_count_reward": 0.5, "step": 68 }, { "clip_ratio": 0.0, "completion_length": 521.8541870117188, "epoch": 0.050586510263929615, "grad_norm": 2.5489498982709047, "kl": 0.00344085693359375, "learning_rate": 5.036496350364964e-07, "loss": -0.0328, "reward": 0.9113207161426544, "reward_std": 0.2536340802907944, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.4861111342906952, "rewards/repetition_penalty_reward": -0.08520709350705147, "rewards/tag_count_reward": 0.5104166865348816, "step": 69 }, { "clip_ratio": 0.0, "completion_length": 559.4791870117188, "epoch": 0.051319648093841645, "grad_norm": 2.4000158911131897, "kl": 0.00331878662109375, "learning_rate": 5.109489051094891e-07, "loss": -0.0455, "reward": 0.9332021176815033, "reward_std": 0.20038650184869766, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.5347222685813904, "rewards/repetition_penalty_reward": -0.10152019187808037, "rewards/tag_count_reward": 0.5, "step": 70 }, { "clip_ratio": 0.0, "completion_length": 528.9583587646484, "epoch": 0.05205278592375367, "grad_norm": 2.6501302736560426, "kl": 0.00391387939453125, "learning_rate": 5.182481751824817e-07, "loss": -0.0121, "reward": 1.0358231365680695, "reward_std": 0.25363823771476746, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.6041666865348816, "rewards/repetition_penalty_reward": -0.0891769677400589, "rewards/tag_count_reward": 0.5208333432674408, "step": 71 }, { "clip_ratio": 0.0, "completion_length": 588.2916870117188, "epoch": 0.05278592375366569, "grad_norm": 2.4485936157125074, "kl": 0.00411224365234375, "learning_rate": 5.255474452554745e-07, "loss": -0.0353, "reward": 0.9984837472438812, "reward_std": 0.23911786824464798, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.5486111640930176, "rewards/repetition_penalty_reward": -0.07096070051193237, "rewards/tag_count_reward": 0.5208333730697632, "step": 72 }, { "clip_ratio": 0.0, "completion_length": 556.4375, "epoch": 0.053519061583577714, "grad_norm": 2.490917274894813, "kl": 0.0072174072265625, "learning_rate": 5.328467153284672e-07, "loss": -0.0127, "reward": 0.9354804754257202, "reward_std": 0.2260909304022789, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.534722238779068, "rewards/repetition_penalty_reward": -0.09924176707863808, "rewards/tag_count_reward": 0.5, "step": 73 }, { "clip_ratio": 0.0, "completion_length": 521.7500305175781, "epoch": 0.054252199413489736, "grad_norm": 2.6783458493753733, "kl": 0.005462646484375, "learning_rate": 5.401459854014598e-07, "loss": -0.0287, "reward": 0.9104893207550049, "reward_std": 0.21142099052667618, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.486111119389534, "rewards/repetition_penalty_reward": -0.1068718284368515, "rewards/tag_count_reward": 0.5312500298023224, "step": 74 }, { "clip_ratio": 0.0, "completion_length": 512.2500305175781, "epoch": 0.05498533724340176, "grad_norm": 2.4199982126361843, "kl": 0.0055389404296875, "learning_rate": 5.474452554744526e-07, "loss": -0.0385, "reward": 0.9597910344600677, "reward_std": 0.2526519149541855, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.5486111640930176, "rewards/repetition_penalty_reward": -0.09923676028847694, "rewards/tag_count_reward": 0.5104166865348816, "step": 75 }, { "clip_ratio": 0.0, "completion_length": 510.5000305175781, "epoch": 0.05571847507331378, "grad_norm": 2.567874458479384, "kl": 0.0051422119140625, "learning_rate": 5.547445255474452e-07, "loss": -0.0047, "reward": 1.0856189727783203, "reward_std": 0.24321593344211578, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.6250000298023224, "rewards/repetition_penalty_reward": -0.07063117437064648, "rewards/tag_count_reward": 0.5104166865348816, "step": 76 }, { "clip_ratio": 0.0, "completion_length": 578.2083435058594, "epoch": 0.056451612903225805, "grad_norm": 2.550528253133134, "kl": 0.0053863525390625, "learning_rate": 5.620437956204379e-07, "loss": 0.0184, "reward": 1.1505069136619568, "reward_std": 0.20630817860364914, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.7152778208255768, "rewards/repetition_penalty_reward": -0.07518759742379189, "rewards/tag_count_reward": 0.5104166865348816, "step": 77 }, { "clip_ratio": 0.0, "completion_length": 490.1458435058594, "epoch": 0.05718475073313783, "grad_norm": 2.5568022714676157, "kl": 0.005523681640625, "learning_rate": 5.693430656934306e-07, "loss": -0.0381, "reward": 1.0803128480911255, "reward_std": 0.27904535830020905, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.638888955116272, "rewards/repetition_penalty_reward": -0.0794093906879425, "rewards/tag_count_reward": 0.5208333730697632, "step": 78 }, { "clip_ratio": 0.0, "completion_length": 570.5000305175781, "epoch": 0.05791788856304985, "grad_norm": 2.7737108529156207, "kl": 0.0061187744140625, "learning_rate": 5.766423357664233e-07, "loss": -0.0359, "reward": 1.125738263130188, "reward_std": 0.2861868143081665, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.6527777910232544, "rewards/repetition_penalty_reward": -0.0791228711605072, "rewards/tag_count_reward": 0.5520833730697632, "step": 79 }, { "clip_ratio": 0.0, "completion_length": 658.3125305175781, "epoch": 0.05865102639296188, "grad_norm": 2.3815187987627113, "kl": 0.0054473876953125, "learning_rate": 5.83941605839416e-07, "loss": 0.0008, "reward": 1.0200362801551819, "reward_std": 0.21494220197200775, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.6458333730697632, "rewards/repetition_penalty_reward": -0.13621379435062408, "rewards/tag_count_reward": 0.5104166865348816, "step": 80 }, { "clip_ratio": 0.0, "completion_length": 573.5416717529297, "epoch": 0.0593841642228739, "grad_norm": 2.6130817675171194, "kl": 0.0069732666015625, "learning_rate": 5.912408759124087e-07, "loss": 0.0508, "reward": 1.1462955176830292, "reward_std": 0.28369559347629547, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.6527778208255768, "rewards/repetition_penalty_reward": -0.08981562405824661, "rewards/tag_count_reward": 0.5416666865348816, "step": 81 }, { "clip_ratio": 0.0, "completion_length": 582.2291870117188, "epoch": 0.060117302052785926, "grad_norm": 2.25648691082032, "kl": 0.006988525390625, "learning_rate": 5.985401459854015e-07, "loss": -0.0314, "reward": 1.0307790040969849, "reward_std": 0.2735845670104027, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.5972222685813904, "rewards/repetition_penalty_reward": -0.09769327193498611, "rewards/tag_count_reward": 0.53125, "step": 82 }, { "clip_ratio": 0.0, "completion_length": 542.8958435058594, "epoch": 0.06085043988269795, "grad_norm": 2.5748740896100406, "kl": 0.00799560546875, "learning_rate": 6.058394160583942e-07, "loss": -0.0757, "reward": 1.1210277676582336, "reward_std": 0.22354952991008759, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.694444477558136, "rewards/repetition_penalty_reward": -0.08383342623710632, "rewards/tag_count_reward": 0.5104166865348816, "step": 83 }, { "clip_ratio": 0.0, "completion_length": 540.6666870117188, "epoch": 0.06158357771260997, "grad_norm": 2.671125025547467, "kl": 0.009521484375, "learning_rate": 6.131386861313868e-07, "loss": -0.0565, "reward": 1.1098045110702515, "reward_std": 0.22741875797510147, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.6805556118488312, "rewards/repetition_penalty_reward": -0.07075115293264389, "rewards/tag_count_reward": 0.5, "step": 84 }, { "clip_ratio": 0.0, "completion_length": 489.3333435058594, "epoch": 0.062316715542521994, "grad_norm": 2.7047116143217607, "kl": 0.008575439453125, "learning_rate": 6.204379562043796e-07, "loss": -0.034, "reward": 1.031806230545044, "reward_std": 0.23993152379989624, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.597222238779068, "rewards/repetition_penalty_reward": -0.08624938875436783, "rewards/tag_count_reward": 0.5208333432674408, "step": 85 }, { "clip_ratio": 0.0, "completion_length": 542.9791870117188, "epoch": 0.06304985337243402, "grad_norm": 2.3811852124777007, "kl": 0.011016845703125, "learning_rate": 6.277372262773722e-07, "loss": -0.059, "reward": 1.1634512543678284, "reward_std": 0.1828886717557907, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.7430555820465088, "rewards/repetition_penalty_reward": -0.09002107009291649, "rewards/tag_count_reward": 0.5104166865348816, "step": 86 }, { "clip_ratio": 0.0, "completion_length": 560.7916870117188, "epoch": 0.06378299120234604, "grad_norm": 2.6708744424748825, "kl": 0.009613037109375, "learning_rate": 6.350364963503649e-07, "loss": 0.0257, "reward": 1.2677294611930847, "reward_std": 0.17205430567264557, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8263888955116272, "rewards/repetition_penalty_reward": -0.06907607242465019, "rewards/tag_count_reward": 0.5104166865348816, "step": 87 }, { "clip_ratio": 0.0, "completion_length": 500.62501525878906, "epoch": 0.06451612903225806, "grad_norm": 2.6527083044097615, "kl": 0.013580322265625, "learning_rate": 6.423357664233577e-07, "loss": -0.0816, "reward": 1.157953143119812, "reward_std": 0.24211852997541428, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.722222238779068, "rewards/repetition_penalty_reward": -0.08510247990489006, "rewards/tag_count_reward": 0.5208333432674408, "step": 88 }, { "clip_ratio": 0.0, "completion_length": 585.2083740234375, "epoch": 0.06524926686217009, "grad_norm": 2.254851707913722, "kl": 0.01177978515625, "learning_rate": 6.496350364963503e-07, "loss": -0.0616, "reward": 1.2707419395446777, "reward_std": 0.23218122124671936, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8263888955116272, "rewards/repetition_penalty_reward": -0.08689698204398155, "rewards/tag_count_reward": 0.5312500298023224, "step": 89 }, { "clip_ratio": 0.0, "completion_length": 553.0625, "epoch": 0.06598240469208211, "grad_norm": 2.466044255250126, "kl": 0.01318359375, "learning_rate": 6.56934306569343e-07, "loss": 0.0192, "reward": 1.2977005243301392, "reward_std": 0.2469094917178154, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.8472222685813904, "rewards/repetition_penalty_reward": -0.08077173680067062, "rewards/tag_count_reward": 0.5104166865348816, "step": 90 }, { "clip_ratio": 0.0, "completion_length": 546.8750305175781, "epoch": 0.06671554252199413, "grad_norm": 2.4627490545315553, "kl": 0.012664794921875, "learning_rate": 6.642335766423358e-07, "loss": -0.0433, "reward": 1.2336487770080566, "reward_std": 0.1906793713569641, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8055556118488312, "rewards/repetition_penalty_reward": -0.09274015948176384, "rewards/tag_count_reward": 0.5208333730697632, "step": 91 }, { "clip_ratio": 0.0, "completion_length": 571.0416870117188, "epoch": 0.06744868035190615, "grad_norm": 2.4773692933804274, "kl": 0.01470947265625, "learning_rate": 6.715328467153284e-07, "loss": -0.0215, "reward": 1.2830377221107483, "reward_std": 0.24985270202159882, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8402778506278992, "rewards/repetition_penalty_reward": -0.09890671819448471, "rewards/tag_count_reward": 0.5416666865348816, "step": 92 }, { "clip_ratio": 0.0, "completion_length": 631.0208435058594, "epoch": 0.06818181818181818, "grad_norm": 2.252251791829593, "kl": 0.015655517578125, "learning_rate": 6.788321167883211e-07, "loss": 0.0148, "reward": 1.3440173268318176, "reward_std": 0.15536434948444366, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9305555820465088, "rewards/repetition_penalty_reward": -0.10737163573503494, "rewards/tag_count_reward": 0.5208333432674408, "step": 93 }, { "clip_ratio": 0.0, "completion_length": 568.8958435058594, "epoch": 0.0689149560117302, "grad_norm": 2.371964666409008, "kl": 0.018798828125, "learning_rate": 6.861313868613139e-07, "loss": -0.065, "reward": 1.3092040419578552, "reward_std": 0.21491578966379166, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8888889253139496, "rewards/repetition_penalty_reward": -0.10051822662353516, "rewards/tag_count_reward": 0.5208333432674408, "step": 94 }, { "clip_ratio": 0.0, "completion_length": 638.7083740234375, "epoch": 0.06964809384164222, "grad_norm": 2.102439408725684, "kl": 0.01702880859375, "learning_rate": 6.934306569343066e-07, "loss": 0.0856, "reward": 1.3951058387756348, "reward_std": 0.23365377634763718, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.944444477558136, "rewards/repetition_penalty_reward": -0.12225543707609177, "rewards/tag_count_reward": 0.5312500298023224, "step": 95 }, { "clip_ratio": 0.0, "completion_length": 564.2708740234375, "epoch": 0.07038123167155426, "grad_norm": 2.327185590119073, "kl": 0.0244140625, "learning_rate": 7.007299270072993e-07, "loss": -0.0365, "reward": 1.4423339366912842, "reward_std": 0.1464700922369957, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.07849938794970512, "rewards/tag_count_reward": 0.5625000298023224, "step": 96 }, { "clip_ratio": 0.0, "completion_length": 676.8750305175781, "epoch": 0.07111436950146628, "grad_norm": 2.3497921909453523, "kl": 0.021240234375, "learning_rate": 7.080291970802919e-07, "loss": 0.2078, "reward": 1.3351240158081055, "reward_std": 0.14220446348190308, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.14057040959596634, "rewards/tag_count_reward": 0.5104166865348816, "step": 97 }, { "clip_ratio": 0.0, "completion_length": 649.0416870117188, "epoch": 0.0718475073313783, "grad_norm": 2.2789434818818886, "kl": 0.0260009765625, "learning_rate": 7.153284671532847e-07, "loss": 0.2727, "reward": 1.3547228574752808, "reward_std": 0.15613484382629395, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9374999701976776, "rewards/repetition_penalty_reward": -0.13486044853925705, "rewards/tag_count_reward": 0.5520833730697632, "step": 98 }, { "clip_ratio": 0.0, "completion_length": 621.8333740234375, "epoch": 0.07258064516129033, "grad_norm": 2.239779876944864, "kl": 0.0238037109375, "learning_rate": 7.226277372262773e-07, "loss": -0.04, "reward": 1.4298600554466248, "reward_std": 0.12703320011496544, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.08055667206645012, "rewards/tag_count_reward": 0.5520833432674408, "step": 99 }, { "clip_ratio": 0.0, "completion_length": 571.75, "epoch": 0.07331378299120235, "grad_norm": 2.8804572295856166, "kl": 0.02581787109375, "learning_rate": 7.2992700729927e-07, "loss": 0.0164, "reward": 1.377421259880066, "reward_std": 0.2720048576593399, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8819444179534912, "rewards/repetition_penalty_reward": -0.10348153859376907, "rewards/tag_count_reward": 0.5989583432674408, "step": 100 }, { "clip_ratio": 0.0, "completion_length": 829.5625305175781, "epoch": 0.07404692082111437, "grad_norm": 2.345441719178201, "kl": 0.02825927734375, "learning_rate": 7.372262773722628e-07, "loss": 0.2908, "reward": 1.3878263235092163, "reward_std": 0.27931635081768036, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.17467374354600906, "rewards/tag_count_reward": 0.5625000298023224, "step": 101 }, { "clip_ratio": 0.0, "completion_length": 810.2708435058594, "epoch": 0.0747800586510264, "grad_norm": 2.029032638001561, "kl": 0.0281982421875, "learning_rate": 7.445255474452554e-07, "loss": 0.2321, "reward": 1.4438632726669312, "reward_std": 0.2681792825460434, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.944444477558136, "rewards/repetition_penalty_reward": -0.14120623841881752, "rewards/tag_count_reward": 0.640625, "step": 102 }, { "clip_ratio": 0.0, "completion_length": 716.1875305175781, "epoch": 0.07551319648093842, "grad_norm": 2.150055244776667, "kl": 0.03033447265625, "learning_rate": 7.518248175182481e-07, "loss": 0.1382, "reward": 1.4972922205924988, "reward_std": 0.3046337366104126, "rewards/accuracy_reward": 0.0625, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.12423554062843323, "rewards/tag_count_reward": 0.5729166865348816, "step": 103 }, { "clip_ratio": 0.0, "completion_length": 712.7916870117188, "epoch": 0.07624633431085044, "grad_norm": 2.314344363733894, "kl": 0.0306396484375, "learning_rate": 7.591240875912408e-07, "loss": 0.3068, "reward": 1.419297456741333, "reward_std": 0.2410280853509903, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.1414663940668106, "rewards/tag_count_reward": 0.5677083432674408, "step": 104 }, { "clip_ratio": 0.0, "completion_length": 708.4791870117188, "epoch": 0.07697947214076246, "grad_norm": 2.227593566299099, "kl": 0.02899169921875, "learning_rate": 7.664233576642335e-07, "loss": 0.2322, "reward": 1.5039226412773132, "reward_std": 0.3817121684551239, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.1523272842168808, "rewards/tag_count_reward": 0.6354166865348816, "step": 105 }, { "clip_ratio": 0.0, "completion_length": 741.1875305175781, "epoch": 0.07771260997067449, "grad_norm": 2.2189255233187275, "kl": 0.03448486328125, "learning_rate": 7.737226277372263e-07, "loss": 0.0745, "reward": 1.5402050018310547, "reward_std": 0.3311127871274948, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.13687820360064507, "rewards/tag_count_reward": 0.6354166865348816, "step": 106 }, { "clip_ratio": 0.0, "completion_length": 725.7708435058594, "epoch": 0.07844574780058651, "grad_norm": 2.3727721162936293, "kl": 0.03070068359375, "learning_rate": 7.81021897810219e-07, "loss": 0.0747, "reward": 1.601328730583191, "reward_std": 0.3444855064153671, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.8888889253139496, "rewards/repetition_penalty_reward": -0.11568531394004822, "rewards/tag_count_reward": 0.8072916865348816, "step": 107 }, { "clip_ratio": 0.0, "completion_length": 731.2500305175781, "epoch": 0.07917888563049853, "grad_norm": 2.208602964719147, "kl": 0.0318603515625, "learning_rate": 7.883211678832117e-07, "loss": 0.1092, "reward": 1.816445231437683, "reward_std": 0.4272291660308838, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 0.9513888955116272, "rewards/repetition_penalty_reward": -0.11931872367858887, "rewards/tag_count_reward": 0.8177083432674408, "step": 108 }, { "clip_ratio": 0.0, "completion_length": 648.1041870117188, "epoch": 0.07991202346041056, "grad_norm": 2.180090722902742, "kl": 0.031005859375, "learning_rate": 7.956204379562043e-07, "loss": 0.0962, "reward": 1.7373812198638916, "reward_std": 0.31745412945747375, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9027777910232544, "rewards/repetition_penalty_reward": -0.11331330239772797, "rewards/tag_count_reward": 0.9062500298023224, "step": 109 }, { "clip_ratio": 0.0, "completion_length": 628.875, "epoch": 0.08064516129032258, "grad_norm": 2.24336862203659, "kl": 0.03515625, "learning_rate": 8.029197080291971e-07, "loss": 0.0259, "reward": 1.707243025302887, "reward_std": 0.2569514065980911, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8958333730697632, "rewards/repetition_penalty_reward": -0.11567378789186478, "rewards/tag_count_reward": 0.9270833432674408, "step": 110 }, { "clip_ratio": 0.0, "completion_length": 668.2916870117188, "epoch": 0.0813782991202346, "grad_norm": 2.1467393949027946, "kl": 0.0316162109375, "learning_rate": 8.102189781021898e-07, "loss": 0.1118, "reward": 1.7002267837524414, "reward_std": 0.3253027945756912, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9027777910232544, "rewards/repetition_penalty_reward": -0.15046779811382294, "rewards/tag_count_reward": 0.90625, "step": 111 }, { "clip_ratio": 0.0, "completion_length": 571.2291870117188, "epoch": 0.08211143695014662, "grad_norm": 2.3046724261447245, "kl": 0.03759765625, "learning_rate": 8.175182481751824e-07, "loss": -0.098, "reward": 1.6816839575767517, "reward_std": 0.24633130431175232, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8333334028720856, "rewards/repetition_penalty_reward": -0.10998266562819481, "rewards/tag_count_reward": 0.9583333432674408, "step": 112 }, { "clip_ratio": 0.0, "completion_length": 544.375, "epoch": 0.08284457478005865, "grad_norm": 2.5306732105186605, "kl": 0.034423828125, "learning_rate": 8.248175182481751e-07, "loss": -0.0392, "reward": 2.261577606201172, "reward_std": 0.3669068068265915, "rewards/accuracy_reward": 0.5416666716337204, "rewards/reasoning_steps_reward": 0.8194445371627808, "rewards/repetition_penalty_reward": -0.08390852063894272, "rewards/tag_count_reward": 0.9843750298023224, "step": 113 }, { "clip_ratio": 0.0, "completion_length": 649.6041870117188, "epoch": 0.08357771260997067, "grad_norm": 2.244871151106607, "kl": 0.0389404296875, "learning_rate": 8.321167883211679e-07, "loss": -0.0034, "reward": 1.8438727259635925, "reward_std": 0.42509177327156067, "rewards/accuracy_reward": 0.14583333395421505, "rewards/reasoning_steps_reward": 0.8333333134651184, "rewards/repetition_penalty_reward": -0.11446066945791245, "rewards/tag_count_reward": 0.9791666865348816, "step": 114 }, { "clip_ratio": 0.0, "completion_length": 634.2916870117188, "epoch": 0.08431085043988269, "grad_norm": 2.491100748223219, "kl": 0.0379638671875, "learning_rate": 8.394160583941605e-07, "loss": 0.0772, "reward": 1.9845237731933594, "reward_std": 0.4105287790298462, "rewards/accuracy_reward": 0.2083333395421505, "rewards/reasoning_steps_reward": 0.8958333134651184, "rewards/repetition_penalty_reward": -0.08318460360169411, "rewards/tag_count_reward": 0.9635416865348816, "step": 115 }, { "clip_ratio": 0.0, "completion_length": 685.6875, "epoch": 0.08504398826979472, "grad_norm": 2.344273362836253, "kl": 0.0421142578125, "learning_rate": 8.467153284671532e-07, "loss": 0.1762, "reward": 1.7933751940727234, "reward_std": 0.20816757529973984, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9375000298023224, "rewards/repetition_penalty_reward": -0.10766644030809402, "rewards/tag_count_reward": 0.9635416865348816, "step": 116 }, { "clip_ratio": 0.0, "completion_length": 576.9583740234375, "epoch": 0.08577712609970674, "grad_norm": 2.2662508812576356, "kl": 0.042724609375, "learning_rate": 8.540145985401459e-07, "loss": -0.0667, "reward": 1.9130470752716064, "reward_std": 0.31153056025505066, "rewards/accuracy_reward": 0.10416666977107525, "rewards/reasoning_steps_reward": 0.9166666865348816, "rewards/repetition_penalty_reward": -0.10257799923419952, "rewards/tag_count_reward": 0.9947916865348816, "step": 117 }, { "clip_ratio": 0.0, "completion_length": 654.1875305175781, "epoch": 0.08651026392961877, "grad_norm": 2.3816109059086035, "kl": 0.0408935546875, "learning_rate": 8.613138686131386e-07, "loss": 0.0516, "reward": 1.8271796703338623, "reward_std": 0.26171302795410156, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9166666865348816, "rewards/repetition_penalty_reward": -0.08948708325624466, "rewards/tag_count_reward": 0.9791666865348816, "step": 118 }, { "clip_ratio": 0.0, "completion_length": 602.9583435058594, "epoch": 0.0872434017595308, "grad_norm": 2.2625493199191613, "kl": 0.050537109375, "learning_rate": 8.686131386861314e-07, "loss": -0.02, "reward": 1.8098711371421814, "reward_std": 0.2411314696073532, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9375000298023224, "rewards/repetition_penalty_reward": -0.10158723592758179, "rewards/tag_count_reward": 0.9531250298023224, "step": 119 }, { "clip_ratio": 0.0, "completion_length": 607.9583740234375, "epoch": 0.08797653958944282, "grad_norm": 2.491338730976297, "kl": 0.0498046875, "learning_rate": 8.759124087591241e-07, "loss": 0.0994, "reward": 1.9433124661445618, "reward_std": 0.28877225518226624, "rewards/accuracy_reward": 0.1041666716337204, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.1052987314760685, "rewards/tag_count_reward": 0.9791666865348816, "step": 120 }, { "clip_ratio": 0.0, "completion_length": 602.4166870117188, "epoch": 0.08870967741935484, "grad_norm": 2.100723178631127, "kl": 0.049560546875, "learning_rate": 8.832116788321168e-07, "loss": -0.0021, "reward": 1.8328325152397156, "reward_std": 0.14558201283216476, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.944444477558136, "rewards/repetition_penalty_reward": -0.09077860787510872, "rewards/tag_count_reward": 0.9791666865348816, "step": 121 }, { "clip_ratio": 0.0, "completion_length": 617.0416870117188, "epoch": 0.08944281524926687, "grad_norm": 2.0146574407917996, "kl": 0.0506591796875, "learning_rate": 8.905109489051094e-07, "loss": -0.0551, "reward": 1.9449394941329956, "reward_std": 0.26772307604551315, "rewards/accuracy_reward": 0.125, "rewards/reasoning_steps_reward": 0.9236111044883728, "rewards/repetition_penalty_reward": -0.09325498715043068, "rewards/tag_count_reward": 0.9895833432674408, "step": 122 }, { "clip_ratio": 0.0, "completion_length": 647.875, "epoch": 0.09017595307917889, "grad_norm": 2.1647631689137072, "kl": 0.055908203125, "learning_rate": 8.978102189781022e-07, "loss": -0.0092, "reward": 2.107422709465027, "reward_std": 0.4037548154592514, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.09570235759019852, "rewards/tag_count_reward": 0.9947916865348816, "step": 123 }, { "clip_ratio": 0.0, "completion_length": 606.6458435058594, "epoch": 0.09090909090909091, "grad_norm": 2.3175998277197434, "kl": 0.0531005859375, "learning_rate": 9.051094890510949e-07, "loss": 0.0093, "reward": 2.0177698135375977, "reward_std": 0.21625616401433945, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.11243857070803642, "rewards/tag_count_reward": 0.9843750298023224, "step": 124 }, { "clip_ratio": 0.0, "completion_length": 547.5208435058594, "epoch": 0.09164222873900293, "grad_norm": 2.3635455743051135, "kl": 0.0628662109375, "learning_rate": 9.124087591240875e-07, "loss": 0.0655, "reward": 1.8874244093894958, "reward_std": 0.06788370944559574, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.09521446004509926, "rewards/tag_count_reward": 0.9895833432674408, "step": 125 }, { "clip_ratio": 0.0, "completion_length": 578.6458435058594, "epoch": 0.09237536656891496, "grad_norm": 4.714739215359856, "kl": 0.09521484375, "learning_rate": 9.197080291970803e-07, "loss": 0.0332, "reward": 1.857076346874237, "reward_std": 0.10263842344284058, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.1029931865632534, "rewards/tag_count_reward": 0.9739583432674408, "step": 126 }, { "clip_ratio": 0.0, "completion_length": 616.3958435058594, "epoch": 0.09310850439882698, "grad_norm": 2.242822379316528, "kl": 0.0604248046875, "learning_rate": 9.270072992700729e-07, "loss": 0.0292, "reward": 2.2475364208221436, "reward_std": 0.2705245167016983, "rewards/accuracy_reward": 0.3333333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.07538021355867386, "rewards/tag_count_reward": 0.9895833432674408, "step": 127 }, { "clip_ratio": 0.0, "completion_length": 566.9791870117188, "epoch": 0.093841642228739, "grad_norm": 2.2965764822384434, "kl": 0.069580078125, "learning_rate": 9.343065693430656e-07, "loss": 0.0055, "reward": 2.1329785585403442, "reward_std": 0.1943557783961296, "rewards/accuracy_reward": 0.20833333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.07535484433174133, "rewards/tag_count_reward": 1.0, "step": 128 }, { "clip_ratio": 0.0, "completion_length": 639.4375, "epoch": 0.09457478005865103, "grad_norm": 2.7047332776537085, "kl": 0.066162109375, "learning_rate": 9.416058394160583e-07, "loss": 0.1901, "reward": 1.9127892851829529, "reward_std": 0.08384755812585354, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.07158574648201466, "rewards/tag_count_reward": 0.9843750298023224, "step": 129 }, { "clip_ratio": 0.0, "completion_length": 609.5000305175781, "epoch": 0.09530791788856305, "grad_norm": 2.4797993637993776, "kl": 0.069091796875, "learning_rate": 9.48905109489051e-07, "loss": -0.0213, "reward": 1.9225295186042786, "reward_std": 0.13330717384815216, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.07747054100036621, "rewards/tag_count_reward": 1.0, "step": 130 }, { "clip_ratio": 0.0, "completion_length": 595.3750305175781, "epoch": 0.09604105571847507, "grad_norm": 2.3781148023687093, "kl": 0.07080078125, "learning_rate": 9.562043795620438e-07, "loss": 0.1265, "reward": 1.9733397364616394, "reward_std": 0.22801223397254944, "rewards/accuracy_reward": 0.10416666977107525, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.08916043862700462, "rewards/tag_count_reward": 0.9791666865348816, "step": 131 }, { "clip_ratio": 0.0, "completion_length": 558.0625, "epoch": 0.0967741935483871, "grad_norm": 2.2188909082146813, "kl": 0.074462890625, "learning_rate": 9.635036496350364e-07, "loss": 0.042, "reward": 1.9614204168319702, "reward_std": 0.18102481961250305, "rewards/accuracy_reward": 0.0833333358168602, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.10976015031337738, "rewards/tag_count_reward": 0.9947916865348816, "step": 132 }, { "clip_ratio": 0.0, "completion_length": 525.6666870117188, "epoch": 0.09750733137829912, "grad_norm": 2.3429168691260935, "kl": 0.077392578125, "learning_rate": 9.708029197080291e-07, "loss": 0.043, "reward": 1.9318488240242004, "reward_std": 0.02291060145944357, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.06815127283334732, "rewards/tag_count_reward": 1.0, "step": 133 }, { "clip_ratio": 0.0, "completion_length": 562.2708435058594, "epoch": 0.09824046920821114, "grad_norm": 2.523107372125059, "kl": 0.077392578125, "learning_rate": 9.78102189781022e-07, "loss": 0.0068, "reward": 1.9358287453651428, "reward_std": 0.18595868349075317, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.08326847851276398, "rewards/tag_count_reward": 0.984375, "step": 134 }, { "clip_ratio": 0.0, "completion_length": 547.5416870117188, "epoch": 0.09897360703812316, "grad_norm": 2.3009439072227766, "kl": 0.08154296875, "learning_rate": 9.854014598540146e-07, "loss": 0.0738, "reward": 1.959082305431366, "reward_std": 0.16214194893836975, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.07737604528665543, "rewards/tag_count_reward": 0.9947916865348816, "step": 135 }, { "clip_ratio": 0.0, "completion_length": 560.7916870117188, "epoch": 0.09970674486803519, "grad_norm": 2.3099658884043768, "kl": 0.07763671875, "learning_rate": 9.927007299270073e-07, "loss": 0.02, "reward": 1.9437535405158997, "reward_std": 0.18613886833190918, "rewards/accuracy_reward": 0.06250000186264515, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.08055209368467331, "rewards/tag_count_reward": 0.9895833432674408, "step": 136 }, { "clip_ratio": 0.0, "completion_length": 571.7500305175781, "epoch": 0.10043988269794721, "grad_norm": 2.361601327653796, "kl": 0.08251953125, "learning_rate": 1e-06, "loss": 0.0423, "reward": 1.9818828105926514, "reward_std": 0.11443111579865217, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.059783823788166046, "rewards/tag_count_reward": 1.0, "step": 137 }, { "clip_ratio": 0.0, "completion_length": 608.9375, "epoch": 0.10117302052785923, "grad_norm": 2.454515004494713, "kl": 0.07861328125, "learning_rate": 9.999985249970096e-07, "loss": -0.0004, "reward": 1.9380502700805664, "reward_std": 0.10458014532923698, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.061949726194143295, "rewards/tag_count_reward": 1.0, "step": 138 }, { "clip_ratio": 0.0, "completion_length": 560.3958435058594, "epoch": 0.10190615835777127, "grad_norm": 2.269901600994151, "kl": 0.082763671875, "learning_rate": 9.999940999977079e-07, "loss": 0.0515, "reward": 1.9492172002792358, "reward_std": 0.1285417703911662, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.07508841529488564, "rewards/tag_count_reward": 0.9895833432674408, "step": 139 }, { "clip_ratio": 0.0, "completion_length": 557.4583435058594, "epoch": 0.10263929618768329, "grad_norm": 2.567033517686166, "kl": 0.08251953125, "learning_rate": 9.999867250311034e-07, "loss": 0.0618, "reward": 2.2221978902816772, "reward_std": 0.2952120155096054, "rewards/accuracy_reward": 0.3541666865348816, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.07641324028372765, "rewards/tag_count_reward": 0.9583333432674408, "step": 140 }, { "clip_ratio": 0.0, "completion_length": 520.4791870117188, "epoch": 0.10337243401759531, "grad_norm": 9.621459270704188, "kl": 0.117919921875, "learning_rate": 9.99976400145543e-07, "loss": 0.0766, "reward": 1.9341915845870972, "reward_std": 0.0798512976616621, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.04671124555170536, "rewards/tag_count_reward": 0.9947916865348816, "step": 141 }, { "clip_ratio": 0.0, "completion_length": 576.3125, "epoch": 0.10410557184750734, "grad_norm": 2.309493750807184, "kl": 0.083251953125, "learning_rate": 9.999631254087124e-07, "loss": 0.0645, "reward": 2.1059885025024414, "reward_std": 0.15556566044688225, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.06762255355715752, "rewards/tag_count_reward": 1.0, "step": 142 }, { "clip_ratio": 0.0, "completion_length": 488.9166717529297, "epoch": 0.10483870967741936, "grad_norm": 2.492592510007256, "kl": 0.09814453125, "learning_rate": 9.99946900907635e-07, "loss": 0.054, "reward": 1.9212931394577026, "reward_std": 0.08056900650262833, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.05787358991801739, "rewards/tag_count_reward": 0.9791666865348816, "step": 143 }, { "clip_ratio": 0.0, "completion_length": 517.125, "epoch": 0.10557184750733138, "grad_norm": 2.641625952434204, "kl": 0.092041015625, "learning_rate": 9.999277267486716e-07, "loss": -0.0104, "reward": 1.895661175251007, "reward_std": 0.09707108698785305, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.050519492477178574, "rewards/tag_count_reward": 0.9739583730697632, "step": 144 }, { "clip_ratio": 0.0, "completion_length": 546.3958435058594, "epoch": 0.1063049853372434, "grad_norm": 2.4813074698854085, "kl": 0.09912109375, "learning_rate": 9.999056030575195e-07, "loss": -0.0252, "reward": 2.3760710954666138, "reward_std": 0.5173959732055664, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.06663743779063225, "rewards/tag_count_reward": 0.984375, "step": 145 }, { "clip_ratio": 0.0, "completion_length": 566.9375305175781, "epoch": 0.10703812316715543, "grad_norm": 2.546987058241153, "kl": 0.097412109375, "learning_rate": 9.998805299792124e-07, "loss": 0.0225, "reward": 1.9302016496658325, "reward_std": 0.15545649453997612, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.08195120096206665, "rewards/tag_count_reward": 0.984375, "step": 146 }, { "clip_ratio": 0.0, "completion_length": 523.3333435058594, "epoch": 0.10777126099706745, "grad_norm": 2.279115521679075, "kl": 0.09375, "learning_rate": 9.998525076781186e-07, "loss": 0.0485, "reward": 2.1790050268173218, "reward_std": 0.343727208673954, "rewards/accuracy_reward": 0.2708333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0709951352328062, "rewards/tag_count_reward": 0.9791666865348816, "step": 147 }, { "clip_ratio": 0.0, "completion_length": 505.3333435058594, "epoch": 0.10850439882697947, "grad_norm": 2.450437856988281, "kl": 0.109130859375, "learning_rate": 9.9982153633794e-07, "loss": 0.0185, "reward": 2.1389353275299072, "reward_std": 0.12949354946613312, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.05724521912634373, "rewards/tag_count_reward": 0.9531250298023224, "step": 148 }, { "clip_ratio": 0.0, "completion_length": 552.6666870117188, "epoch": 0.1092375366568915, "grad_norm": 2.392731425965375, "kl": 0.105224609375, "learning_rate": 9.997876161617116e-07, "loss": -0.0438, "reward": 1.9390915632247925, "reward_std": 0.14268473163247108, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.052227944135665894, "rewards/tag_count_reward": 0.984375, "step": 149 }, { "clip_ratio": 0.0, "completion_length": 547.4375305175781, "epoch": 0.10997067448680352, "grad_norm": 2.6243429866910577, "kl": 0.1064453125, "learning_rate": 9.997507473717993e-07, "loss": -0.0684, "reward": 1.918170690536499, "reward_std": 0.1779511570930481, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.05752386339008808, "rewards/tag_count_reward": 0.96875, "step": 150 }, { "clip_ratio": 0.0, "completion_length": 526.9791870117188, "epoch": 0.11070381231671554, "grad_norm": 2.429431812645537, "kl": 0.110107421875, "learning_rate": 9.997109302098988e-07, "loss": 0.0351, "reward": 2.0930378437042236, "reward_std": 0.24497440457344055, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.04064269922673702, "rewards/tag_count_reward": 0.9531250298023224, "step": 151 }, { "clip_ratio": 0.0, "completion_length": 551.1041870117188, "epoch": 0.11143695014662756, "grad_norm": 2.735990047546789, "kl": 0.111572265625, "learning_rate": 9.996681649370347e-07, "loss": 0.0556, "reward": 2.1636458039283752, "reward_std": 0.18747984617948532, "rewards/accuracy_reward": 0.25000000558793545, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0551042165607214, "rewards/tag_count_reward": 0.9687500298023224, "step": 152 }, { "clip_ratio": 0.0, "completion_length": 578.4583435058594, "epoch": 0.11217008797653959, "grad_norm": 2.325307078829229, "kl": 0.09814453125, "learning_rate": 9.996224518335572e-07, "loss": 0.0049, "reward": 1.9978720545768738, "reward_std": 0.13420861586928368, "rewards/accuracy_reward": 0.0625, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.05941970832645893, "rewards/tag_count_reward": 0.9947916865348816, "step": 153 }, { "clip_ratio": 0.0, "completion_length": 491.75, "epoch": 0.11290322580645161, "grad_norm": 2.4201064670759, "kl": 0.103515625, "learning_rate": 9.995737911991421e-07, "loss": -0.0234, "reward": 1.9454825520515442, "reward_std": 0.1356247467920184, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.05798974260687828, "rewards/tag_count_reward": 0.96875, "step": 154 }, { "clip_ratio": 0.0, "completion_length": 608.2916870117188, "epoch": 0.11363636363636363, "grad_norm": 2.250316333226382, "kl": 0.093017578125, "learning_rate": 9.995221833527873e-07, "loss": 0.0096, "reward": 1.9684556722640991, "reward_std": 0.14412052184343338, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.05411389470100403, "rewards/tag_count_reward": 0.9947916865348816, "step": 155 }, { "clip_ratio": 0.0, "completion_length": 551.2500305175781, "epoch": 0.11436950146627566, "grad_norm": 2.3123417188932414, "kl": 0.1025390625, "learning_rate": 9.994676286328118e-07, "loss": 0.0526, "reward": 1.957583487033844, "reward_std": 0.08784053660929203, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.06324994564056396, "rewards/tag_count_reward": 1.0, "step": 156 }, { "clip_ratio": 0.0, "completion_length": 555.4375305175781, "epoch": 0.11510263929618768, "grad_norm": 2.364580200558665, "kl": 0.10546875, "learning_rate": 9.994101273968526e-07, "loss": -0.0002, "reward": 1.9639785885810852, "reward_std": 0.2663791626691818, "rewards/accuracy_reward": 0.06250000186264515, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.07247984036803246, "rewards/tag_count_reward": 0.9739583432674408, "step": 157 }, { "clip_ratio": 0.0, "completion_length": 499.6458435058594, "epoch": 0.1158357771260997, "grad_norm": 2.314171354857538, "kl": 0.109130859375, "learning_rate": 9.99349680021863e-07, "loss": 0.0229, "reward": 2.168446123600006, "reward_std": 0.11961105465888977, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.05030396394431591, "rewards/tag_count_reward": 0.9895833432674408, "step": 158 }, { "clip_ratio": 0.0, "completion_length": 566.8750305175781, "epoch": 0.11656891495601172, "grad_norm": 2.3309162622734263, "kl": 0.094970703125, "learning_rate": 9.992862869041102e-07, "loss": -0.0195, "reward": 2.275045156478882, "reward_std": 0.29623332619667053, "rewards/accuracy_reward": 0.375, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.05828823521733284, "rewards/tag_count_reward": 0.9791666865348816, "step": 159 }, { "clip_ratio": 0.0, "completion_length": 555.9791870117188, "epoch": 0.11730205278592376, "grad_norm": 2.4125752206797633, "kl": 0.099609375, "learning_rate": 9.992199484591717e-07, "loss": -0.0082, "reward": 2.190155267715454, "reward_std": 0.15859412401914597, "rewards/accuracy_reward": 0.25000000558793545, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.05984491854906082, "rewards/tag_count_reward": 1.0, "step": 160 }, { "clip_ratio": 0.0, "completion_length": 591.3333435058594, "epoch": 0.11803519061583578, "grad_norm": 2.492754544431609, "kl": 0.095703125, "learning_rate": 9.991506651219344e-07, "loss": 0.0522, "reward": 1.9383143186569214, "reward_std": 0.043892914429306984, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.04779680632054806, "rewards/tag_count_reward": 1.0, "step": 161 }, { "clip_ratio": 0.0, "completion_length": 516.0, "epoch": 0.1187683284457478, "grad_norm": 2.5627904900534975, "kl": 0.099609375, "learning_rate": 9.990784373465895e-07, "loss": 0.023, "reward": 2.162955939769745, "reward_std": 0.23758363723754883, "rewards/accuracy_reward": 0.25000000558793545, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.05579404905438423, "rewards/tag_count_reward": 0.9895833432674408, "step": 162 }, { "clip_ratio": 0.0, "completion_length": 541.3958435058594, "epoch": 0.11950146627565983, "grad_norm": 2.5031797258210755, "kl": 0.097412109375, "learning_rate": 9.99003265606631e-07, "loss": 0.0472, "reward": 2.071454405784607, "reward_std": 0.14920172840356827, "rewards/accuracy_reward": 0.125, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.053545668721199036, "rewards/tag_count_reward": 1.0, "step": 163 }, { "clip_ratio": 0.0, "completion_length": 551.3958740234375, "epoch": 0.12023460410557185, "grad_norm": 2.547042259721191, "kl": 0.09033203125, "learning_rate": 9.989251503948531e-07, "loss": 0.0935, "reward": 2.1519184708595276, "reward_std": 0.1169380396604538, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.049470532685518265, "rewards/tag_count_reward": 1.0, "step": 164 }, { "clip_ratio": 0.0, "completion_length": 612.5208740234375, "epoch": 0.12096774193548387, "grad_norm": 2.2645297165562805, "kl": 0.093994140625, "learning_rate": 9.988440922233447e-07, "loss": 0.0799, "reward": 2.1653225421905518, "reward_std": 0.2170577123761177, "rewards/accuracy_reward": 0.2291666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.06384427845478058, "rewards/tag_count_reward": 1.0, "step": 165 }, { "clip_ratio": 0.0, "completion_length": 592.1666870117188, "epoch": 0.1217008797653959, "grad_norm": 2.431650538562101, "kl": 0.095458984375, "learning_rate": 9.987600916234887e-07, "loss": 0.1683, "reward": 1.938806176185608, "reward_std": 0.045028104446828365, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.05424942448735237, "rewards/tag_count_reward": 1.0, "step": 166 }, { "clip_ratio": 0.0, "completion_length": 603.6875305175781, "epoch": 0.12243401759530792, "grad_norm": 2.335284594764373, "kl": 0.08935546875, "learning_rate": 9.986731491459567e-07, "loss": 0.0245, "reward": 2.197871446609497, "reward_std": 0.15528446715325117, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.052128592506051064, "rewards/tag_count_reward": 1.0, "step": 167 }, { "clip_ratio": 0.0, "completion_length": 569.7500305175781, "epoch": 0.12316715542521994, "grad_norm": 2.3828481630789384, "kl": 0.093017578125, "learning_rate": 9.985832653607058e-07, "loss": 0.0957, "reward": 2.192627727985382, "reward_std": 0.03842144086956978, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.050427746027708054, "rewards/tag_count_reward": 1.0, "step": 168 }, { "clip_ratio": 0.0, "completion_length": 578.1666870117188, "epoch": 0.12390029325513197, "grad_norm": 2.160386876446579, "kl": 0.092529296875, "learning_rate": 9.984904408569757e-07, "loss": -0.0218, "reward": 1.9351824522018433, "reward_std": 0.14489805325865746, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.06481760554015636, "rewards/tag_count_reward": 1.0, "step": 169 }, { "clip_ratio": 0.0, "completion_length": 591.9791870117188, "epoch": 0.12463343108504399, "grad_norm": 2.582494728505769, "kl": 0.099853515625, "learning_rate": 9.98394676243284e-07, "loss": 0.1153, "reward": 2.173978328704834, "reward_std": 0.23065098375082016, "rewards/accuracy_reward": 0.2708333432674408, "rewards/reasoning_steps_reward": 0.9583333730697632, "rewards/repetition_penalty_reward": -0.04997998662292957, "rewards/tag_count_reward": 0.9947916865348816, "step": 170 }, { "clip_ratio": 0.0, "completion_length": 525.9791870117188, "epoch": 0.125366568914956, "grad_norm": 2.360291931663355, "kl": 0.1015625, "learning_rate": 9.982959721474219e-07, "loss": -0.0145, "reward": 2.192633032798767, "reward_std": 0.03827578388154507, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.050422552973032, "rewards/tag_count_reward": 1.0, "step": 171 }, { "clip_ratio": 0.0, "completion_length": 497.54168701171875, "epoch": 0.12609970674486803, "grad_norm": 2.644000756748961, "kl": 0.107666015625, "learning_rate": 9.981943292164507e-07, "loss": -0.0266, "reward": 2.1891872882843018, "reward_std": 0.16495954617857933, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.05560445971786976, "rewards/tag_count_reward": 0.9947916865348816, "step": 172 }, { "clip_ratio": 0.0, "completion_length": 498.79168701171875, "epoch": 0.12683284457478006, "grad_norm": 2.4636582856331315, "kl": 0.104248046875, "learning_rate": 9.980897481166977e-07, "loss": 0.0109, "reward": 1.9538467526435852, "reward_std": 0.18937725573778152, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.056570012122392654, "rewards/tag_count_reward": 0.9895833730697632, "step": 173 }, { "clip_ratio": 0.0, "completion_length": 535.8750305175781, "epoch": 0.12756598240469208, "grad_norm": 2.3291019775434783, "kl": 0.10302734375, "learning_rate": 9.979822295337521e-07, "loss": -0.0373, "reward": 1.9403542280197144, "reward_std": 0.08776792883872986, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.03534029796719551, "rewards/tag_count_reward": 0.9895833432674408, "step": 174 }, { "clip_ratio": 0.0, "completion_length": 615.0000305175781, "epoch": 0.1282991202346041, "grad_norm": 2.105843364049476, "kl": 0.08984375, "learning_rate": 9.978717741724588e-07, "loss": 0.0256, "reward": 1.9570952653884888, "reward_std": 0.03407964622601867, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.035960348322987556, "rewards/tag_count_reward": 1.0, "step": 175 }, { "clip_ratio": 0.0, "completion_length": 594.4791870117188, "epoch": 0.12903225806451613, "grad_norm": 2.585212670700021, "kl": 0.09912109375, "learning_rate": 9.977583827569156e-07, "loss": 0.1018, "reward": 2.166927397251129, "reward_std": 0.08626264333724976, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.05529484711587429, "rewards/tag_count_reward": 0.9791666865348816, "step": 176 }, { "clip_ratio": 0.0, "completion_length": 574.4375, "epoch": 0.12976539589442815, "grad_norm": 2.3815955363709747, "kl": 0.105712890625, "learning_rate": 9.976420560304679e-07, "loss": -0.0233, "reward": 2.071603298187256, "reward_std": 0.26225684583187103, "rewards/accuracy_reward": 0.14583333395421505, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.04992457665503025, "rewards/tag_count_reward": 0.9895833432674408, "step": 177 }, { "clip_ratio": 0.0, "completion_length": 593.3958435058594, "epoch": 0.13049853372434017, "grad_norm": 2.243479758643072, "kl": 0.100341796875, "learning_rate": 9.975227947557036e-07, "loss": 0.0422, "reward": 1.9449068903923035, "reward_std": 0.04067743383347988, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.04814865067601204, "rewards/tag_count_reward": 1.0, "step": 178 }, { "clip_ratio": 0.0, "completion_length": 546.5208435058594, "epoch": 0.1312316715542522, "grad_norm": 2.3950868693631735, "kl": 0.103759765625, "learning_rate": 9.974005997144479e-07, "loss": 0.0154, "reward": 2.0247842669487, "reward_std": 0.1975030368193984, "rewards/accuracy_reward": 0.0833333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0585490707308054, "rewards/tag_count_reward": 1.0, "step": 179 }, { "clip_ratio": 0.0, "completion_length": 606.6458435058594, "epoch": 0.13196480938416422, "grad_norm": 2.2458633692097654, "kl": 0.1025390625, "learning_rate": 9.97275471707759e-07, "loss": 0.015, "reward": 1.9811076521873474, "reward_std": 0.08166738552972674, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03972569480538368, "rewards/tag_count_reward": 1.0, "step": 180 }, { "clip_ratio": 0.0, "completion_length": 597.5416870117188, "epoch": 0.13269794721407624, "grad_norm": 2.208686653320979, "kl": 0.103515625, "learning_rate": 9.97147411555922e-07, "loss": -0.027, "reward": 2.1563020944595337, "reward_std": 0.2810695618391037, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.05376743897795677, "rewards/tag_count_reward": 0.9947916865348816, "step": 181 }, { "clip_ratio": 0.0, "completion_length": 615.4583740234375, "epoch": 0.13343108504398826, "grad_norm": 2.223637211770504, "kl": 0.098876953125, "learning_rate": 9.970164200984443e-07, "loss": 0.1221, "reward": 1.9517627954483032, "reward_std": 0.02038181759417057, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04823717288672924, "rewards/tag_count_reward": 1.0, "step": 182 }, { "clip_ratio": 0.0, "completion_length": 539.5416870117188, "epoch": 0.13416422287390029, "grad_norm": 2.185131937450779, "kl": 0.112060546875, "learning_rate": 9.968824981940493e-07, "loss": 0.0155, "reward": 2.033494234085083, "reward_std": 0.2103577759116888, "rewards/accuracy_reward": 0.125, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.056783486157655716, "rewards/tag_count_reward": 0.9791666865348816, "step": 183 }, { "clip_ratio": 0.0, "completion_length": 549.4166870117188, "epoch": 0.1348973607038123, "grad_norm": 2.2887490821871546, "kl": 0.103759765625, "learning_rate": 9.967456467206712e-07, "loss": 0.0392, "reward": 1.9644048810005188, "reward_std": 0.016866141464561224, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03559521585702896, "rewards/tag_count_reward": 1.0, "step": 184 }, { "clip_ratio": 0.0, "completion_length": 553.9583740234375, "epoch": 0.13563049853372433, "grad_norm": 2.075544050145113, "kl": 0.10791015625, "learning_rate": 9.966058665754494e-07, "loss": 0.0403, "reward": 2.1910440325737, "reward_std": 0.20203115046024323, "rewards/accuracy_reward": 0.25000000558793545, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04333098791539669, "rewards/tag_count_reward": 0.984375, "step": 185 }, { "clip_ratio": 0.0, "completion_length": 575.8125, "epoch": 0.13636363636363635, "grad_norm": 2.444422689957302, "kl": 0.107666015625, "learning_rate": 9.964631586747222e-07, "loss": -0.0033, "reward": 2.5688068866729736, "reward_std": 0.3440853953361511, "rewards/accuracy_reward": 0.625, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.056193241849541664, "rewards/tag_count_reward": 1.0, "step": 186 }, { "clip_ratio": 0.0, "completion_length": 553.1458435058594, "epoch": 0.13709677419354838, "grad_norm": 2.3343020822873486, "kl": 0.10595703125, "learning_rate": 9.96317523954021e-07, "loss": 0.0113, "reward": 1.9665733575820923, "reward_std": 0.1026981808245182, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.0473155677318573, "rewards/tag_count_reward": 1.0, "step": 187 }, { "clip_ratio": 0.0, "completion_length": 527.1458435058594, "epoch": 0.1378299120234604, "grad_norm": 2.329781106869945, "kl": 0.11572265625, "learning_rate": 9.96168963368064e-07, "loss": 0.0768, "reward": 1.9751185774803162, "reward_std": 0.09284409787505865, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.035298120230436325, "rewards/tag_count_reward": 0.9895833432674408, "step": 188 }, { "clip_ratio": 0.0, "completion_length": 568.8125305175781, "epoch": 0.13856304985337242, "grad_norm": 2.482124615192751, "kl": 0.1162109375, "learning_rate": 9.960174778907511e-07, "loss": 0.0832, "reward": 1.9523666501045227, "reward_std": 0.051067665219306946, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03548061661422253, "rewards/tag_count_reward": 0.9947916865348816, "step": 189 }, { "clip_ratio": 0.0, "completion_length": 502.2083435058594, "epoch": 0.13929618768328444, "grad_norm": 2.7539326807930724, "kl": 0.128173828125, "learning_rate": 9.958630685151552e-07, "loss": -0.1067, "reward": 1.9934042692184448, "reward_std": 0.14851272106170654, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04305417276918888, "rewards/tag_count_reward": 0.9947916865348816, "step": 190 }, { "clip_ratio": 0.0, "completion_length": 507.1458435058594, "epoch": 0.14002932551319647, "grad_norm": 2.4239115093163774, "kl": 0.124755859375, "learning_rate": 9.957057362535175e-07, "loss": -0.0111, "reward": 1.9687672853469849, "reward_std": 0.11539381369948387, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04164946265518665, "rewards/tag_count_reward": 0.9895833432674408, "step": 191 }, { "clip_ratio": 0.0, "completion_length": 498.97918701171875, "epoch": 0.14076246334310852, "grad_norm": 2.519049551235326, "kl": 0.12939453125, "learning_rate": 9.955454821372408e-07, "loss": -0.0006, "reward": 1.9685339331626892, "reward_std": 0.20130838453769684, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.045354995876550674, "rewards/tag_count_reward": 0.9791666865348816, "step": 192 }, { "clip_ratio": 0.0, "completion_length": 544.125, "epoch": 0.14149560117302054, "grad_norm": 2.2656746914975896, "kl": 0.12841796875, "learning_rate": 9.953823072168818e-07, "loss": 0.0209, "reward": 2.5191445350646973, "reward_std": 0.3530745655298233, "rewards/accuracy_reward": 0.5833333432674408, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.039883313700556755, "rewards/tag_count_reward": 0.9895833432674408, "step": 193 }, { "clip_ratio": 0.0, "completion_length": 482.1666717529297, "epoch": 0.14222873900293256, "grad_norm": 2.4378035107662335, "kl": 0.13720703125, "learning_rate": 9.95216212562145e-07, "loss": 0.0679, "reward": 1.995971918106079, "reward_std": 0.1278656329959631, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03354204259812832, "rewards/tag_count_reward": 0.9947916865348816, "step": 194 }, { "clip_ratio": 0.0, "completion_length": 512.6666870117188, "epoch": 0.1429618768328446, "grad_norm": 2.2820945478569987, "kl": 0.125732421875, "learning_rate": 9.950471992618755e-07, "loss": 0.0234, "reward": 1.9476872682571411, "reward_std": 0.05502081662416458, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.03842394798994064, "rewards/tag_count_reward": 1.0, "step": 195 }, { "clip_ratio": 0.0, "completion_length": 588.9166870117188, "epoch": 0.1436950146627566, "grad_norm": 2.5028248973645755, "kl": 0.12353515625, "learning_rate": 9.948752684240514e-07, "loss": 0.0617, "reward": 1.9217350482940674, "reward_std": 0.07230636849999428, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.05569552071392536, "rewards/tag_count_reward": 0.984375, "step": 196 }, { "clip_ratio": 0.0, "completion_length": 537.6250305175781, "epoch": 0.14442815249266863, "grad_norm": 2.428726069054531, "kl": 0.115234375, "learning_rate": 9.94700421175777e-07, "loss": 0.0636, "reward": 2.27834951877594, "reward_std": 0.2576148062944412, "rewards/accuracy_reward": 0.3333333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.054983872920274734, "rewards/tag_count_reward": 1.0, "step": 197 }, { "clip_ratio": 0.0, "completion_length": 531.4166717529297, "epoch": 0.14516129032258066, "grad_norm": 2.0921870960930993, "kl": 0.119873046875, "learning_rate": 9.945226586632757e-07, "loss": 0.0528, "reward": 2.110503077507019, "reward_std": 0.16293007880449295, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.04921908490359783, "rewards/tag_count_reward": 1.0, "step": 198 }, { "clip_ratio": 0.0, "completion_length": 533.7500305175781, "epoch": 0.14589442815249268, "grad_norm": 2.5544009433480603, "kl": 0.12841796875, "learning_rate": 9.94341982051882e-07, "loss": 0.1531, "reward": 2.2020418643951416, "reward_std": 0.04367715120315552, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.04101373255252838, "rewards/tag_count_reward": 1.0, "step": 199 }, { "clip_ratio": 0.0, "completion_length": 510.3958435058594, "epoch": 0.1466275659824047, "grad_norm": 2.2761467748200888, "kl": 0.1337890625, "learning_rate": 9.94158392526034e-07, "loss": 0.0584, "reward": 1.9831884503364563, "reward_std": 0.0833294466137886, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03243660740554333, "rewards/tag_count_reward": 0.9947916865348816, "step": 200 }, { "clip_ratio": 0.0, "completion_length": 521.9791870117188, "epoch": 0.14736070381231672, "grad_norm": 2.676798078696678, "kl": 0.124267578125, "learning_rate": 9.939718912892649e-07, "loss": 0.1067, "reward": 2.4419782757759094, "reward_std": 0.08668806962668896, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03718854580074549, "rewards/tag_count_reward": 1.0, "step": 201 }, { "clip_ratio": 0.0, "completion_length": 528.8541870117188, "epoch": 0.14809384164222875, "grad_norm": 2.201643714021596, "kl": 0.128662109375, "learning_rate": 9.937824795641966e-07, "loss": 0.0402, "reward": 2.1860954761505127, "reward_std": 0.05664380919188261, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.044807299971580505, "rewards/tag_count_reward": 0.9947916865348816, "step": 202 }, { "clip_ratio": 0.0, "completion_length": 536.0625305175781, "epoch": 0.14882697947214077, "grad_norm": 2.3529632476015143, "kl": 0.121826171875, "learning_rate": 9.935901585925309e-07, "loss": 0.0326, "reward": 1.9818111658096313, "reward_std": 0.07902031671255827, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.04596656560897827, "rewards/tag_count_reward": 1.0, "step": 203 }, { "clip_ratio": 0.0, "completion_length": 506.6250305175781, "epoch": 0.1495601173020528, "grad_norm": 2.216555100598739, "kl": 0.1240234375, "learning_rate": 9.933949296350412e-07, "loss": 0.0381, "reward": 1.9562581181526184, "reward_std": 0.034087372943758965, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.036797499284148216, "rewards/tag_count_reward": 1.0, "step": 204 }, { "clip_ratio": 0.0, "completion_length": 532.0833587646484, "epoch": 0.15029325513196481, "grad_norm": 2.252511834459066, "kl": 0.125732421875, "learning_rate": 9.931967939715641e-07, "loss": 0.0761, "reward": 1.960132360458374, "reward_std": 0.02215556614100933, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.039867645129561424, "rewards/tag_count_reward": 1.0, "step": 205 }, { "clip_ratio": 0.0, "completion_length": 515.1041870117188, "epoch": 0.15102639296187684, "grad_norm": 2.1090934911188004, "kl": 0.127197265625, "learning_rate": 9.929957529009918e-07, "loss": 0.0337, "reward": 1.995642364025116, "reward_std": 0.16707224398851395, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03560762293636799, "rewards/tag_count_reward": 0.9895833432674408, "step": 206 }, { "clip_ratio": 0.0, "completion_length": 481.4166717529297, "epoch": 0.15175953079178886, "grad_norm": 2.051423391569217, "kl": 0.13623046875, "learning_rate": 9.927918077412628e-07, "loss": -0.0048, "reward": 2.2103371620178223, "reward_std": 0.03615569323301315, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.03271842747926712, "rewards/tag_count_reward": 1.0, "step": 207 }, { "clip_ratio": 0.0, "completion_length": 471.3541717529297, "epoch": 0.15249266862170088, "grad_norm": 2.664772912161007, "kl": 0.15576171875, "learning_rate": 9.925849598293537e-07, "loss": 0.032, "reward": 2.417176127433777, "reward_std": 0.247524194419384, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03594890981912613, "rewards/tag_count_reward": 0.9947916865348816, "step": 208 }, { "clip_ratio": 0.0, "completion_length": 527.3958435058594, "epoch": 0.1532258064516129, "grad_norm": 2.7816240748478847, "kl": 0.14501953125, "learning_rate": 9.9237521052127e-07, "loss": -0.0241, "reward": 2.0370668172836304, "reward_std": 0.22554985573515296, "rewards/accuracy_reward": 0.1041666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.0393221378326416, "rewards/tag_count_reward": 0.9791666865348816, "step": 209 }, { "clip_ratio": 0.0, "completion_length": 507.7708435058594, "epoch": 0.15395894428152493, "grad_norm": 2.4349590137522963, "kl": 0.13427734375, "learning_rate": 9.921625611920383e-07, "loss": 0.0232, "reward": 1.9492262601852417, "reward_std": 0.13168392330408096, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.047301506623625755, "rewards/tag_count_reward": 0.9895833432674408, "step": 210 }, { "clip_ratio": 0.0, "completion_length": 469.5625, "epoch": 0.15469208211143695, "grad_norm": 2.288220045433554, "kl": 0.1591796875, "learning_rate": 9.919470132356951e-07, "loss": 0.0722, "reward": 2.18506121635437, "reward_std": 0.11852885410189629, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03368875943124294, "rewards/tag_count_reward": 0.9895833432674408, "step": 211 }, { "clip_ratio": 0.0, "completion_length": 505.52085876464844, "epoch": 0.15542521994134897, "grad_norm": 2.4198341602493847, "kl": 0.14501953125, "learning_rate": 9.917285680652805e-07, "loss": 0.0407, "reward": 1.9904600381851196, "reward_std": 0.08161180093884468, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03037323523312807, "rewards/tag_count_reward": 1.0, "step": 212 }, { "clip_ratio": 0.0, "completion_length": 512.8333587646484, "epoch": 0.156158357771261, "grad_norm": 2.4090799066503377, "kl": 0.1572265625, "learning_rate": 9.915072271128267e-07, "loss": 0.0703, "reward": 2.2472126483917236, "reward_std": 0.13479421101510525, "rewards/accuracy_reward": 0.2916666679084301, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.030565201304852962, "rewards/tag_count_reward": 1.0, "step": 213 }, { "clip_ratio": 0.0, "completion_length": 475.8541717529297, "epoch": 0.15689149560117302, "grad_norm": 2.41265188004121, "kl": 0.15380859375, "learning_rate": 9.912829918293494e-07, "loss": -0.0356, "reward": 2.0343655347824097, "reward_std": 0.12357844784855843, "rewards/accuracy_reward": 0.0625, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02813446894288063, "rewards/tag_count_reward": 1.0, "step": 214 }, { "clip_ratio": 0.0, "completion_length": 509.97918701171875, "epoch": 0.15762463343108504, "grad_norm": 2.5135257964207063, "kl": 0.1591796875, "learning_rate": 9.910558636848384e-07, "loss": 0.0185, "reward": 2.0641019344329834, "reward_std": 0.20102613419294357, "rewards/accuracy_reward": 0.10416666977107525, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04006483219563961, "rewards/tag_count_reward": 1.0, "step": 215 }, { "clip_ratio": 0.0, "completion_length": 554.1250305175781, "epoch": 0.15835777126099707, "grad_norm": 2.4705818727306736, "kl": 0.15234375, "learning_rate": 9.908258441682483e-07, "loss": 0.1556, "reward": 1.9976014494895935, "reward_std": 0.11414317414164543, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04406532645225525, "rewards/tag_count_reward": 1.0, "step": 216 }, { "clip_ratio": 0.0, "completion_length": 447.06251525878906, "epoch": 0.1590909090909091, "grad_norm": 2.3388114870879972, "kl": 0.17236328125, "learning_rate": 9.905929347874875e-07, "loss": 0.1217, "reward": 2.140606701374054, "reward_std": 0.1381812645122409, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026059916242957115, "rewards/tag_count_reward": 1.0, "step": 217 }, { "clip_ratio": 0.0, "completion_length": 459.6458435058594, "epoch": 0.1598240469208211, "grad_norm": 2.4861533154135693, "kl": 0.16357421875, "learning_rate": 9.903571370694094e-07, "loss": 0.0441, "reward": 1.9717371463775635, "reward_std": 0.04522157460451126, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01784633845090866, "rewards/tag_count_reward": 0.9895833432674408, "step": 218 }, { "clip_ratio": 0.0, "completion_length": 455.9583435058594, "epoch": 0.16055718475073313, "grad_norm": 2.337842900368109, "kl": 0.17333984375, "learning_rate": 9.901184525598027e-07, "loss": 0.0557, "reward": 2.0282862186431885, "reward_std": 0.1811966523528099, "rewards/accuracy_reward": 0.0625, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.029005596414208412, "rewards/tag_count_reward": 0.9947916865348816, "step": 219 }, { "clip_ratio": 0.0, "completion_length": 483.7708435058594, "epoch": 0.16129032258064516, "grad_norm": 2.855863111930354, "kl": 0.18115234375, "learning_rate": 9.898768828233794e-07, "loss": 0.0382, "reward": 2.1459413170814514, "reward_std": 0.22265248745679855, "rewards/accuracy_reward": 0.18750000558793545, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03461429849267006, "rewards/tag_count_reward": 1.0, "step": 220 }, { "clip_ratio": 0.0, "completion_length": 510.7083435058594, "epoch": 0.16202346041055718, "grad_norm": 2.512452871604219, "kl": 0.1767578125, "learning_rate": 9.896324294437672e-07, "loss": -0.055, "reward": 2.0672988891601562, "reward_std": 0.27433933317661285, "rewards/accuracy_reward": 0.10416666977107525, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03165953606367111, "rewards/tag_count_reward": 0.9947916865348816, "step": 221 }, { "clip_ratio": 0.0, "completion_length": 482.37501525878906, "epoch": 0.1627565982404692, "grad_norm": 2.497431270952545, "kl": 0.185546875, "learning_rate": 9.893850940234968e-07, "loss": 0.0612, "reward": 2.032313048839569, "reward_std": 0.1233673607930541, "rewards/accuracy_reward": 0.0625, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03018704056739807, "rewards/tag_count_reward": 1.0, "step": 222 }, { "clip_ratio": 0.0, "completion_length": 434.0208435058594, "epoch": 0.16348973607038123, "grad_norm": 2.2171293698344763, "kl": 0.197265625, "learning_rate": 9.891348781839924e-07, "loss": 0.0565, "reward": 1.9839209914207458, "reward_std": 0.08671730384230614, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.036912497133016586, "rewards/tag_count_reward": 1.0, "step": 223 }, { "clip_ratio": 0.0, "completion_length": 458.68751525878906, "epoch": 0.16422287390029325, "grad_norm": 2.4157632367769506, "kl": 0.203125, "learning_rate": 9.888817835655614e-07, "loss": 0.0593, "reward": 2.203449249267578, "reward_std": 0.17519402969628572, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025717386044561863, "rewards/tag_count_reward": 1.0, "step": 224 }, { "clip_ratio": 0.0, "completion_length": 457.2708435058594, "epoch": 0.16495601173020527, "grad_norm": 2.1811901804547307, "kl": 0.203125, "learning_rate": 9.88625811827383e-07, "loss": 0.0335, "reward": 1.9971312880516052, "reward_std": 0.09780013933777809, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018493805546313524, "rewards/tag_count_reward": 0.9947916865348816, "step": 225 }, { "clip_ratio": 0.0, "completion_length": 451.3333435058594, "epoch": 0.1656891495601173, "grad_norm": 2.6019578181839544, "kl": 0.1923828125, "learning_rate": 9.883669646474973e-07, "loss": 0.0033, "reward": 2.0132646560668945, "reward_std": 0.11891119182109833, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023193655535578728, "rewards/tag_count_reward": 0.9947916865348816, "step": 226 }, { "clip_ratio": 0.0, "completion_length": 415.9791717529297, "epoch": 0.16642228739002932, "grad_norm": 2.3555410793667133, "kl": 0.20703125, "learning_rate": 9.881052437227952e-07, "loss": 0.034, "reward": 2.7692244052886963, "reward_std": 0.2729061245918274, "rewards/accuracy_reward": 0.8125000298023224, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.025914611294865608, "rewards/tag_count_reward": 0.9895833432674408, "step": 227 }, { "clip_ratio": 0.0, "completion_length": 497.02085876464844, "epoch": 0.16715542521994134, "grad_norm": 2.3616091059425903, "kl": 0.185546875, "learning_rate": 9.878406507690057e-07, "loss": 0.0621, "reward": 1.9636911153793335, "reward_std": 0.03294796636328101, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.029364525340497494, "rewards/tag_count_reward": 1.0, "step": 228 }, { "clip_ratio": 0.0, "completion_length": 432.9791717529297, "epoch": 0.16788856304985336, "grad_norm": 2.6391305475663263, "kl": 0.1962890625, "learning_rate": 9.875731875206867e-07, "loss": -0.0174, "reward": 1.9969558715820312, "reward_std": 0.16866409545764327, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.023877506144344807, "rewards/tag_count_reward": 1.0, "step": 229 }, { "clip_ratio": 0.0, "completion_length": 430.18751525878906, "epoch": 0.16862170087976538, "grad_norm": 2.4059793072528244, "kl": 0.20166015625, "learning_rate": 9.873028557312117e-07, "loss": 0.0309, "reward": 2.2606165409088135, "reward_std": 0.4185255467891693, "rewards/accuracy_reward": 0.2916666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.031050268560647964, "rewards/tag_count_reward": 1.0, "step": 230 }, { "clip_ratio": 0.0, "completion_length": 487.9583435058594, "epoch": 0.1693548387096774, "grad_norm": 2.3075648633918915, "kl": 0.19482421875, "learning_rate": 9.870296571727593e-07, "loss": 0.0661, "reward": 2.091531753540039, "reward_std": 0.15529845468699932, "rewards/accuracy_reward": 0.125, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02825986687093973, "rewards/tag_count_reward": 0.9947916865348816, "step": 231 }, { "clip_ratio": 0.0, "completion_length": 472.2083435058594, "epoch": 0.17008797653958943, "grad_norm": 2.322103071917664, "kl": 0.2080078125, "learning_rate": 9.867535936363014e-07, "loss": -0.0263, "reward": 2.481574058532715, "reward_std": 0.3974439948797226, "rewards/accuracy_reward": 0.5416666865348816, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.035787204280495644, "rewards/tag_count_reward": 0.9895833432674408, "step": 232 }, { "clip_ratio": 0.0, "completion_length": 481.8750305175781, "epoch": 0.17082111436950145, "grad_norm": 2.512718529548141, "kl": 0.19189453125, "learning_rate": 9.864746669315918e-07, "loss": 0.0368, "reward": 2.0009429454803467, "reward_std": 0.08158356510102749, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01989055424928665, "rewards/tag_count_reward": 1.0, "step": 233 }, { "clip_ratio": 0.0, "completion_length": 565.75, "epoch": 0.17155425219941348, "grad_norm": 2.4334008578543735, "kl": 0.17529296875, "learning_rate": 9.861928788871535e-07, "loss": 0.0988, "reward": 2.0528666973114014, "reward_std": 0.2537204041145742, "rewards/accuracy_reward": 0.1041666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02873060293495655, "rewards/tag_count_reward": 0.984375, "step": 234 }, { "clip_ratio": 0.0, "completion_length": 469.4583435058594, "epoch": 0.17228739002932553, "grad_norm": 2.3972353284423304, "kl": 0.2041015625, "learning_rate": 9.859082313502675e-07, "loss": 0.0168, "reward": 2.2656466960906982, "reward_std": 0.33335772156715393, "rewards/accuracy_reward": 0.2916666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026020064018666744, "rewards/tag_count_reward": 1.0, "step": 235 }, { "clip_ratio": 0.0, "completion_length": 463.16668701171875, "epoch": 0.17302052785923755, "grad_norm": 2.575842711191836, "kl": 0.19970703125, "learning_rate": 9.8562072618696e-07, "loss": 0.0457, "reward": 2.0120570063591003, "reward_std": 0.17007002606987953, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.022665289230644703, "rewards/tag_count_reward": 1.0, "step": 236 }, { "clip_ratio": 0.0, "completion_length": 443.7083435058594, "epoch": 0.17375366568914957, "grad_norm": 2.545760915484643, "kl": 0.201171875, "learning_rate": 9.853303652819915e-07, "loss": 0.0767, "reward": 2.1953667402267456, "reward_std": 0.10493875481188297, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.026855461299419403, "rewards/tag_count_reward": 1.0, "step": 237 }, { "clip_ratio": 0.0, "completion_length": 444.2708435058594, "epoch": 0.1744868035190616, "grad_norm": 2.7557749541944268, "kl": 0.2119140625, "learning_rate": 9.850371505388423e-07, "loss": 0.094, "reward": 1.9746447801589966, "reward_std": 0.013843707740306854, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025355206802487373, "rewards/tag_count_reward": 1.0, "step": 238 }, { "clip_ratio": 0.0, "completion_length": 483.6250305175781, "epoch": 0.17521994134897362, "grad_norm": 2.299442195057228, "kl": 0.19873046875, "learning_rate": 9.847410838797023e-07, "loss": -0.005, "reward": 2.4455658197402954, "reward_std": 0.34960001707077026, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028392494656145573, "rewards/tag_count_reward": 0.9947916865348816, "step": 239 }, { "clip_ratio": 0.0, "completion_length": 466.81251525878906, "epoch": 0.17595307917888564, "grad_norm": 2.4207285910028347, "kl": 0.21337890625, "learning_rate": 9.844421672454568e-07, "loss": 0.0215, "reward": 2.165701985359192, "reward_std": 0.2287914901971817, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.032214757055044174, "rewards/tag_count_reward": 0.9895833432674408, "step": 240 }, { "clip_ratio": 0.0, "completion_length": 561.4166870117188, "epoch": 0.17668621700879766, "grad_norm": 2.606638173149529, "kl": 0.2119140625, "learning_rate": 9.84140402595674e-07, "loss": -0.0635, "reward": 2.1458557844161987, "reward_std": 0.42205628752708435, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.0399082750082016, "rewards/tag_count_reward": 0.9843750298023224, "step": 241 }, { "clip_ratio": 0.0, "completion_length": 488.2083435058594, "epoch": 0.1774193548387097, "grad_norm": 2.3562918788818914, "kl": 0.1962890625, "learning_rate": 9.838357919085933e-07, "loss": -0.0014, "reward": 2.0086446404457092, "reward_std": 0.1550162397325039, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.026077693328261375, "rewards/tag_count_reward": 1.0, "step": 242 }, { "clip_ratio": 0.0, "completion_length": 465.68751525878906, "epoch": 0.1781524926686217, "grad_norm": 3.015046047499862, "kl": 0.19775390625, "learning_rate": 9.835283371811109e-07, "loss": -0.0831, "reward": 2.0528554916381836, "reward_std": 0.28607890009880066, "rewards/accuracy_reward": 0.0833333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025269586592912674, "rewards/tag_count_reward": 0.9947916865348816, "step": 243 }, { "clip_ratio": 0.0, "completion_length": 556.4166870117188, "epoch": 0.17888563049853373, "grad_norm": 2.324053410132789, "kl": 0.18603515625, "learning_rate": 9.832180404287672e-07, "loss": 0.0356, "reward": 2.3999125957489014, "reward_std": 0.34801794588565826, "rewards/accuracy_reward": 0.4375000298023224, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03064296767115593, "rewards/tag_count_reward": 1.0, "step": 244 }, { "clip_ratio": 0.0, "completion_length": 547.3125305175781, "epoch": 0.17961876832844575, "grad_norm": 2.483675172609574, "kl": 0.18701171875, "learning_rate": 9.829049036857338e-07, "loss": -0.0115, "reward": 2.0943979024887085, "reward_std": 0.2273593619465828, "rewards/accuracy_reward": 0.125, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030602077022194862, "rewards/tag_count_reward": 1.0, "step": 245 }, { "clip_ratio": 0.0, "completion_length": 516.2500305175781, "epoch": 0.18035190615835778, "grad_norm": 2.261230507474765, "kl": 0.18115234375, "learning_rate": 9.825889290048009e-07, "loss": -0.0115, "reward": 1.9686490893363953, "reward_std": 0.03400381188839674, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.024406529031693935, "rewards/tag_count_reward": 1.0, "step": 246 }, { "clip_ratio": 0.0, "completion_length": 578.375, "epoch": 0.1810850439882698, "grad_norm": 2.4081317889508536, "kl": 0.1767578125, "learning_rate": 9.822701184573617e-07, "loss": -0.0415, "reward": 2.3041937351226807, "reward_std": 0.4814845621585846, "rewards/accuracy_reward": 0.3333333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.029139596037566662, "rewards/tag_count_reward": 1.0, "step": 247 }, { "clip_ratio": 0.0, "completion_length": 695.2916870117188, "epoch": 0.18181818181818182, "grad_norm": 2.013219887318077, "kl": 0.16162109375, "learning_rate": 9.819484741334009e-07, "loss": -0.0087, "reward": 2.386115550994873, "reward_std": 0.4011628329753876, "rewards/accuracy_reward": 0.458333358168602, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.04444010742008686, "rewards/tag_count_reward": 0.9791666865348816, "step": 248 }, { "clip_ratio": 0.0, "completion_length": 662.6041870117188, "epoch": 0.18255131964809385, "grad_norm": 2.121922144164688, "kl": 0.1767578125, "learning_rate": 9.816239981414798e-07, "loss": 0.1018, "reward": 2.027174472808838, "reward_std": 0.22916549444198608, "rewards/accuracy_reward": 0.0833333358168602, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.04227009415626526, "rewards/tag_count_reward": 1.0, "step": 249 }, { "clip_ratio": 0.0, "completion_length": 616.3750305175781, "epoch": 0.18328445747800587, "grad_norm": 2.323030202609986, "kl": 0.1826171875, "learning_rate": 9.812966926087233e-07, "loss": -0.0441, "reward": 2.2639695405960083, "reward_std": 0.4275299608707428, "rewards/accuracy_reward": 0.3333333358168602, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.04158623516559601, "rewards/tag_count_reward": 0.9791666865348816, "step": 250 }, { "clip_ratio": 0.0, "completion_length": 634.3541870117188, "epoch": 0.1840175953079179, "grad_norm": 2.2692788356949185, "kl": 0.18994140625, "learning_rate": 9.809665596808052e-07, "loss": -0.0243, "reward": 2.1400359869003296, "reward_std": 0.311031524091959, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.03704751655459404, "rewards/tag_count_reward": 0.9895833432674408, "step": 251 }, { "clip_ratio": 0.0, "completion_length": 642.2916870117188, "epoch": 0.18475073313782991, "grad_norm": 1.9963336498318256, "kl": 0.1689453125, "learning_rate": 9.80633601521935e-07, "loss": -0.0222, "reward": 2.2108322381973267, "reward_std": 0.38528619706630707, "rewards/accuracy_reward": 0.3125000149011612, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.044376133009791374, "rewards/tag_count_reward": 0.9635416865348816, "step": 252 }, { "clip_ratio": 0.0, "completion_length": 571.9375305175781, "epoch": 0.18548387096774194, "grad_norm": 2.1395612032304605, "kl": 0.1923828125, "learning_rate": 9.802978203148422e-07, "loss": 0.0066, "reward": 2.5844688415527344, "reward_std": 0.35509173572063446, "rewards/accuracy_reward": 0.6458333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.05094785802066326, "rewards/tag_count_reward": 0.9895833432674408, "step": 253 }, { "clip_ratio": 0.0, "completion_length": 617.6458435058594, "epoch": 0.18621700879765396, "grad_norm": 1.895221846754078, "kl": 0.185546875, "learning_rate": 9.799592182607642e-07, "loss": 0.0354, "reward": 2.242966055870056, "reward_std": 0.3878481537103653, "rewards/accuracy_reward": 0.3125, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.05217302590608597, "rewards/tag_count_reward": 0.9895833432674408, "step": 254 }, { "clip_ratio": 0.0, "completion_length": 625.3541870117188, "epoch": 0.18695014662756598, "grad_norm": 2.3377888789478147, "kl": 0.17138671875, "learning_rate": 9.796177975794298e-07, "loss": 0.0756, "reward": 2.0248435735702515, "reward_std": 0.2661816030740738, "rewards/accuracy_reward": 0.10416666977107525, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.041128676384687424, "rewards/tag_count_reward": 0.9687500298023224, "step": 255 }, { "clip_ratio": 0.0, "completion_length": 680.1041870117188, "epoch": 0.187683284457478, "grad_norm": 2.1512517683797894, "kl": 0.1572265625, "learning_rate": 9.79273560509046e-07, "loss": -0.0595, "reward": 2.554927945137024, "reward_std": 0.4941348135471344, "rewards/accuracy_reward": 0.625, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.04576661065220833, "rewards/tag_count_reward": 0.9895833432674408, "step": 256 }, { "clip_ratio": 0.0, "completion_length": 633.8125305175781, "epoch": 0.18841642228739003, "grad_norm": 2.7630684380572603, "kl": 0.18310546875, "learning_rate": 9.789265093062822e-07, "loss": 0.0441, "reward": 1.9179131984710693, "reward_std": 0.12576640397310257, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0404201652854681, "rewards/tag_count_reward": 0.9583333432674408, "step": 257 }, { "clip_ratio": 0.0, "completion_length": 573.6875305175781, "epoch": 0.18914956011730205, "grad_norm": 2.396423250364506, "kl": 0.18017578125, "learning_rate": 9.785766462462569e-07, "loss": -0.012, "reward": 2.2367652654647827, "reward_std": 0.28754863142967224, "rewards/accuracy_reward": 0.2916666716337204, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.030595812946558, "rewards/tag_count_reward": 0.9895833432674408, "step": 258 }, { "clip_ratio": 0.0, "completion_length": 590.2708435058594, "epoch": 0.18988269794721407, "grad_norm": 2.450809574492532, "kl": 0.18896484375, "learning_rate": 9.782239736225212e-07, "loss": 0.0219, "reward": 2.128044545650482, "reward_std": 0.2238977551460266, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03862226940691471, "rewards/tag_count_reward": 0.9583333432674408, "step": 259 }, { "clip_ratio": 0.0, "completion_length": 609.7708740234375, "epoch": 0.1906158357771261, "grad_norm": 2.472322699922063, "kl": 0.1943359375, "learning_rate": 9.778684937470449e-07, "loss": -0.0137, "reward": 2.080292046070099, "reward_std": 0.2516675442457199, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.03602752089500427, "rewards/tag_count_reward": 0.9218750298023224, "step": 260 }, { "clip_ratio": 0.0, "completion_length": 601.0416870117188, "epoch": 0.19134897360703812, "grad_norm": 2.371141310592578, "kl": 0.189453125, "learning_rate": 9.775102089502006e-07, "loss": 0.0055, "reward": 2.4550044536590576, "reward_std": 0.4750853478908539, "rewards/accuracy_reward": 0.5833333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.041523367166519165, "rewards/tag_count_reward": 0.9270833432674408, "step": 261 }, { "clip_ratio": 0.0, "completion_length": 671.1875305175781, "epoch": 0.19208211143695014, "grad_norm": 2.2457744714397823, "kl": 0.16943359375, "learning_rate": 9.771491215807494e-07, "loss": 0.036, "reward": 2.567304491996765, "reward_std": 0.27622977923601866, "rewards/accuracy_reward": 0.6458333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.057695645838975906, "rewards/tag_count_reward": 0.9791666865348816, "step": 262 }, { "clip_ratio": 0.0, "completion_length": 619.9166870117188, "epoch": 0.19281524926686217, "grad_norm": 2.2302173486507244, "kl": 0.1708984375, "learning_rate": 9.76785234005824e-07, "loss": 0.0567, "reward": 2.4142779111862183, "reward_std": 0.15193303674459457, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04405543580651283, "rewards/tag_count_reward": 0.9791666865348816, "step": 263 }, { "clip_ratio": 0.0, "completion_length": 642.8333435058594, "epoch": 0.1935483870967742, "grad_norm": 2.0145102947205626, "kl": 0.169921875, "learning_rate": 9.764185486109145e-07, "loss": 0.0162, "reward": 2.172469139099121, "reward_std": 0.23066802322864532, "rewards/accuracy_reward": 0.25000000558793545, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04628100246191025, "rewards/tag_count_reward": 0.96875, "step": 264 }, { "clip_ratio": 0.0, "completion_length": 629.7500305175781, "epoch": 0.1942815249266862, "grad_norm": 2.0038074970978776, "kl": 0.1630859375, "learning_rate": 9.760490677998522e-07, "loss": 0.0212, "reward": 2.1879026293754578, "reward_std": 0.05894716642796993, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.04820855334401131, "rewards/tag_count_reward": 1.0, "step": 265 }, { "clip_ratio": 0.0, "completion_length": 698.6875305175781, "epoch": 0.19501466275659823, "grad_norm": 2.2584906259961723, "kl": 0.14599609375, "learning_rate": 9.756767939947943e-07, "loss": 0.0732, "reward": 2.205389618873596, "reward_std": 0.015887603163719177, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.044610561802983284, "rewards/tag_count_reward": 1.0, "step": 266 }, { "clip_ratio": 0.0, "completion_length": 643.2291870117188, "epoch": 0.19574780058651026, "grad_norm": 2.342739537155942, "kl": 0.16650390625, "learning_rate": 9.753017296362071e-07, "loss": 0.0894, "reward": 2.6615763902664185, "reward_std": 0.15336718410253525, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.043284930288791656, "rewards/tag_count_reward": 0.9895833432674408, "step": 267 }, { "clip_ratio": 0.0, "completion_length": 628.5, "epoch": 0.19648093841642228, "grad_norm": 2.2474890401882432, "kl": 0.15087890625, "learning_rate": 9.749238771828508e-07, "loss": 0.0413, "reward": 2.3291032314300537, "reward_std": 0.29626287519931793, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.05631352588534355, "rewards/tag_count_reward": 0.9687500298023224, "step": 268 }, { "clip_ratio": 0.0, "completion_length": 607.5208435058594, "epoch": 0.1972140762463343, "grad_norm": 2.1688862261955264, "kl": 0.17529296875, "learning_rate": 9.745432391117634e-07, "loss": 0.0159, "reward": 2.17527574300766, "reward_std": 0.11154358834028244, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.0400020070374012, "rewards/tag_count_reward": 1.0, "step": 269 }, { "clip_ratio": 0.0, "completion_length": 616.2291870117188, "epoch": 0.19794721407624633, "grad_norm": 2.272593328423986, "kl": 0.18359375, "learning_rate": 9.741598179182438e-07, "loss": 0.0672, "reward": 2.16366446018219, "reward_std": 0.218302384018898, "rewards/accuracy_reward": 0.22916667722165585, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.037724535912275314, "rewards/tag_count_reward": 0.9791666865348816, "step": 270 }, { "clip_ratio": 0.0, "completion_length": 586.5416870117188, "epoch": 0.19868035190615835, "grad_norm": 2.145976535817121, "kl": 0.16748046875, "learning_rate": 9.73773616115836e-07, "loss": 0.0378, "reward": 2.7103559970855713, "reward_std": 0.017032308503985405, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0396440215408802, "rewards/tag_count_reward": 1.0, "step": 271 }, { "clip_ratio": 0.0, "completion_length": 587.3541870117188, "epoch": 0.19941348973607037, "grad_norm": 2.3472461743913517, "kl": 0.18408203125, "learning_rate": 9.733846362363127e-07, "loss": 0.0767, "reward": 2.156644105911255, "reward_std": 0.13528572767972946, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.037800345569849014, "rewards/tag_count_reward": 1.0, "step": 272 }, { "clip_ratio": 0.0, "completion_length": 604.0208435058594, "epoch": 0.2001466275659824, "grad_norm": 2.087923346496973, "kl": 0.162109375, "learning_rate": 9.729928808296582e-07, "loss": 0.0308, "reward": 2.173057436943054, "reward_std": 0.15152714774012566, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03527583181858063, "rewards/tag_count_reward": 0.9791666865348816, "step": 273 }, { "clip_ratio": 0.0, "completion_length": 501.2083435058594, "epoch": 0.20087976539589442, "grad_norm": 2.1601887207000066, "kl": 0.19677734375, "learning_rate": 9.725983524640518e-07, "loss": 0.0332, "reward": 1.9604487419128418, "reward_std": 0.03581496700644493, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03260680288076401, "rewards/tag_count_reward": 1.0, "step": 274 }, { "clip_ratio": 0.0, "completion_length": 496.00001525878906, "epoch": 0.20161290322580644, "grad_norm": 2.393209229139749, "kl": 0.18994140625, "learning_rate": 9.722010537258516e-07, "loss": 0.0625, "reward": 2.202053427696228, "reward_std": 0.04029459087178111, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.034057735465466976, "rewards/tag_count_reward": 1.0, "step": 275 }, { "clip_ratio": 0.0, "completion_length": 487.875, "epoch": 0.20234604105571846, "grad_norm": 2.382514107507378, "kl": 0.21142578125, "learning_rate": 9.71800987219577e-07, "loss": 0.0567, "reward": 2.2166248559951782, "reward_std": 0.036631904542446136, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02643065620213747, "rewards/tag_count_reward": 1.0, "step": 276 }, { "clip_ratio": 0.0, "completion_length": 486.29168701171875, "epoch": 0.20307917888563048, "grad_norm": 2.2044873376719267, "kl": 0.2001953125, "learning_rate": 9.713981555678912e-07, "loss": 0.0244, "reward": 2.4693304300308228, "reward_std": 0.03412939794361591, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.023725169710814953, "rewards/tag_count_reward": 1.0, "step": 277 }, { "clip_ratio": 0.0, "completion_length": 469.0416717529297, "epoch": 0.20381231671554254, "grad_norm": 2.3670075341233603, "kl": 0.20458984375, "learning_rate": 9.709925614115849e-07, "loss": 0.0399, "reward": 2.7084211111068726, "reward_std": 0.0664563006721437, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.024217759259045124, "rewards/tag_count_reward": 0.9895833432674408, "step": 278 }, { "clip_ratio": 0.0, "completion_length": 486.5, "epoch": 0.20454545454545456, "grad_norm": 2.191730991957038, "kl": 0.22216796875, "learning_rate": 9.705842074095585e-07, "loss": 0.0132, "reward": 2.4243065118789673, "reward_std": 0.15112738870084286, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.02881855797022581, "rewards/tag_count_reward": 0.9947916865348816, "step": 279 }, { "clip_ratio": 0.0, "completion_length": 451.35418701171875, "epoch": 0.20527859237536658, "grad_norm": 2.4081949008810892, "kl": 0.20947265625, "learning_rate": 9.70173096238805e-07, "loss": 0.0093, "reward": 2.2207422256469727, "reward_std": 0.02591150812804699, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024049400817602873, "rewards/tag_count_reward": 0.9947916865348816, "step": 280 }, { "clip_ratio": 0.0, "completion_length": 515.75, "epoch": 0.2060117302052786, "grad_norm": 2.1888626480370226, "kl": 0.19970703125, "learning_rate": 9.697592305943917e-07, "loss": 0.059, "reward": 2.210089087486267, "reward_std": 0.05643948353827, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.02602214924991131, "rewards/tag_count_reward": 1.0, "step": 281 }, { "clip_ratio": 0.0, "completion_length": 430.06251525878906, "epoch": 0.20674486803519063, "grad_norm": 2.306900726886975, "kl": 0.234375, "learning_rate": 9.693426131894432e-07, "loss": 0.0345, "reward": 2.2312620878219604, "reward_std": 0.009207626804709435, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018738062120974064, "rewards/tag_count_reward": 1.0, "step": 282 }, { "clip_ratio": 0.0, "completion_length": 468.8125, "epoch": 0.20747800586510265, "grad_norm": 2.50092671528508, "kl": 0.2138671875, "learning_rate": 9.689232467551244e-07, "loss": 0.0516, "reward": 2.225408911705017, "reward_std": 0.02732059359550476, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019382827915251255, "rewards/tag_count_reward": 0.9947916865348816, "step": 283 }, { "clip_ratio": 0.0, "completion_length": 484.00001525878906, "epoch": 0.20821114369501467, "grad_norm": 2.242448357158379, "kl": 0.21533203125, "learning_rate": 9.6850113404062e-07, "loss": 0.0484, "reward": 2.4914671182632446, "reward_std": 0.08239620085805655, "rewards/accuracy_reward": 0.5208333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.029366333037614822, "rewards/tag_count_reward": 1.0, "step": 284 }, { "clip_ratio": 0.0, "completion_length": 529.5416870117188, "epoch": 0.2089442815249267, "grad_norm": 2.1535790290088297, "kl": 0.1884765625, "learning_rate": 9.68076277813119e-07, "loss": 0.0288, "reward": 2.2297321557998657, "reward_std": 0.009526743553578854, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02026796853169799, "rewards/tag_count_reward": 1.0, "step": 285 }, { "clip_ratio": 0.0, "completion_length": 474.4583435058594, "epoch": 0.20967741935483872, "grad_norm": 2.2587788742542476, "kl": 0.23486328125, "learning_rate": 9.67648680857796e-07, "loss": 0.0467, "reward": 2.2198795676231384, "reward_std": 0.011471754871308804, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030120414681732655, "rewards/tag_count_reward": 1.0, "step": 286 }, { "clip_ratio": 0.0, "completion_length": 503.2708435058594, "epoch": 0.21041055718475074, "grad_norm": 2.2164587123984467, "kl": 0.22021484375, "learning_rate": 9.672183459777922e-07, "loss": 0.0474, "reward": 2.4430179595947266, "reward_std": 0.08577717002481222, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025732184760272503, "rewards/tag_count_reward": 0.9895833432674408, "step": 287 }, { "clip_ratio": 0.0, "completion_length": 469.5833435058594, "epoch": 0.21114369501466276, "grad_norm": 2.362743221503858, "kl": 0.224609375, "learning_rate": 9.667852759941971e-07, "loss": 0.0202, "reward": 2.4414188861846924, "reward_std": 0.10892442986369133, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.02733116876333952, "rewards/tag_count_reward": 0.9895833432674408, "step": 288 }, { "clip_ratio": 0.0, "completion_length": 431.3333435058594, "epoch": 0.2118768328445748, "grad_norm": 2.394523367281667, "kl": 0.24853515625, "learning_rate": 9.663494737460316e-07, "loss": 0.03, "reward": 2.4509284496307373, "reward_std": 0.11794408224523067, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.01608549989759922, "rewards/tag_count_reward": 0.9947916865348816, "step": 289 }, { "clip_ratio": 0.0, "completion_length": 415.35418701171875, "epoch": 0.2126099706744868, "grad_norm": 2.247143974513609, "kl": 0.23779296875, "learning_rate": 9.659109420902268e-07, "loss": 0.0313, "reward": 2.225088596343994, "reward_std": 0.02959524537436664, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.017966951709240675, "rewards/tag_count_reward": 1.0, "step": 290 }, { "clip_ratio": 0.0, "completion_length": 443.2916717529297, "epoch": 0.21334310850439883, "grad_norm": 2.2347183819263403, "kl": 0.244140625, "learning_rate": 9.654696839016074e-07, "loss": 0.0357, "reward": 2.465317726135254, "reward_std": 0.046680874191224575, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024265680462121964, "rewards/tag_count_reward": 0.9895833432674408, "step": 291 }, { "clip_ratio": 0.0, "completion_length": 470.47918701171875, "epoch": 0.21407624633431085, "grad_norm": 2.2968281472400487, "kl": 0.24072265625, "learning_rate": 9.650257020728714e-07, "loss": 0.0412, "reward": 2.6881933212280273, "reward_std": 0.11760031245648861, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03055677004158497, "rewards/tag_count_reward": 0.9895833432674408, "step": 292 }, { "clip_ratio": 0.0, "completion_length": 428.9583435058594, "epoch": 0.21480938416422288, "grad_norm": 2.6544213087854915, "kl": 0.24755859375, "learning_rate": 9.645789995145727e-07, "loss": 0.0737, "reward": 2.4477990865707397, "reward_std": 0.11806142888963223, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.01921477634459734, "rewards/tag_count_reward": 0.9947916865348816, "step": 293 }, { "clip_ratio": 0.0, "completion_length": 465.2291717529297, "epoch": 0.2155425219941349, "grad_norm": 2.4050553831347856, "kl": 0.23291015625, "learning_rate": 9.641295791551005e-07, "loss": 0.0226, "reward": 2.7129329442977905, "reward_std": 0.07855140301398933, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.016233783215284348, "rewards/tag_count_reward": 1.0, "step": 294 }, { "clip_ratio": 0.0, "completion_length": 434.875, "epoch": 0.21627565982404692, "grad_norm": 2.469488150327451, "kl": 0.2431640625, "learning_rate": 9.63677443940661e-07, "loss": 0.0516, "reward": 2.4651942253112793, "reward_std": 0.03956274874508381, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.022653001360595226, "rewards/tag_count_reward": 0.9947916865348816, "step": 295 }, { "clip_ratio": 0.0, "completion_length": 485.06251525878906, "epoch": 0.21700879765395895, "grad_norm": 2.2897669783775063, "kl": 0.23681640625, "learning_rate": 9.632225968352577e-07, "loss": 0.0192, "reward": 2.234976053237915, "reward_std": 0.010166795924305916, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.015024024993181229, "rewards/tag_count_reward": 1.0, "step": 296 }, { "clip_ratio": 0.0, "completion_length": 480.7708435058594, "epoch": 0.21774193548387097, "grad_norm": 2.3075335900529943, "kl": 0.26123046875, "learning_rate": 9.62765040820672e-07, "loss": -0.0061, "reward": 2.465970039367676, "reward_std": 0.06250624358654022, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018405052833259106, "rewards/tag_count_reward": 0.9843750298023224, "step": 297 }, { "clip_ratio": 0.0, "completion_length": 415.2708435058594, "epoch": 0.218475073313783, "grad_norm": 2.2643235265355184, "kl": 0.24951171875, "learning_rate": 9.623047788964438e-07, "loss": 0.0291, "reward": 1.9690996408462524, "reward_std": 0.06486085802316666, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.013539265841245651, "rewards/tag_count_reward": 0.9895833432674408, "step": 298 }, { "clip_ratio": 0.0, "completion_length": 468.87501525878906, "epoch": 0.21920821114369501, "grad_norm": 2.3259557842622804, "kl": 0.2607421875, "learning_rate": 9.61841814079852e-07, "loss": 0.0513, "reward": 2.464868426322937, "reward_std": 0.07912581693381071, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01429840736091137, "rewards/tag_count_reward": 1.0, "step": 299 }, { "clip_ratio": 0.0, "completion_length": 500.9583435058594, "epoch": 0.21994134897360704, "grad_norm": 2.298594281909164, "kl": 0.2568359375, "learning_rate": 9.613761494058944e-07, "loss": 0.0419, "reward": 2.2143933176994324, "reward_std": 0.0780528886243701, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01477324403822422, "rewards/tag_count_reward": 1.0, "step": 300 }, { "clip_ratio": 0.0, "completion_length": 494.1666717529297, "epoch": 0.22067448680351906, "grad_norm": 2.473598352431548, "kl": 0.2529296875, "learning_rate": 9.609077879272677e-07, "loss": 0.0927, "reward": 2.474399447441101, "reward_std": 0.025512659922242165, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.020392313599586487, "rewards/tag_count_reward": 0.9947916865348816, "step": 301 }, { "clip_ratio": 0.0, "completion_length": 432.85418701171875, "epoch": 0.22140762463343108, "grad_norm": 2.4769798485568435, "kl": 0.2724609375, "learning_rate": 9.604367327143478e-07, "loss": 0.0744, "reward": 2.7345763444900513, "reward_std": 0.009569189045578241, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.015423623379319906, "rewards/tag_count_reward": 1.0, "step": 302 }, { "clip_ratio": 0.0, "completion_length": 485.9583435058594, "epoch": 0.2221407624633431, "grad_norm": 2.059232163669745, "kl": 0.271484375, "learning_rate": 9.599629868551694e-07, "loss": 0.0208, "reward": 2.714403748512268, "reward_std": 0.07619825517758727, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.014763115905225277, "rewards/tag_count_reward": 1.0, "step": 303 }, { "clip_ratio": 0.0, "completion_length": 451.75001525878906, "epoch": 0.22287390029325513, "grad_norm": 2.284473798213216, "kl": 0.2861328125, "learning_rate": 9.594865534554065e-07, "loss": 0.0445, "reward": 2.2089085578918457, "reward_std": 0.10246565705165267, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.0133137465454638, "rewards/tag_count_reward": 1.0, "step": 304 }, { "clip_ratio": 0.0, "completion_length": 447.125, "epoch": 0.22360703812316715, "grad_norm": 2.340649408671738, "kl": 0.2978515625, "learning_rate": 9.59007435638351e-07, "loss": 0.0569, "reward": 2.461604356765747, "reward_std": 0.0789910196326673, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01756238378584385, "rewards/tag_count_reward": 1.0, "step": 305 }, { "clip_ratio": 0.0, "completion_length": 440.1875, "epoch": 0.22434017595307917, "grad_norm": 2.3064380591396088, "kl": 0.3193359375, "learning_rate": 9.585256365448923e-07, "loss": 0.0539, "reward": 1.9698742032051086, "reward_std": 0.04649292537942529, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01450089504942298, "rewards/tag_count_reward": 0.984375, "step": 306 }, { "clip_ratio": 0.0, "completion_length": 457.0833435058594, "epoch": 0.2250733137829912, "grad_norm": 2.648188486696646, "kl": 0.33203125, "learning_rate": 9.580411593334981e-07, "loss": 0.1127, "reward": 2.217950165271759, "reward_std": 0.06731168460100889, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.014688808470964432, "rewards/tag_count_reward": 0.9895833432674408, "step": 307 }, { "clip_ratio": 0.0, "completion_length": 441.37501525878906, "epoch": 0.22580645161290322, "grad_norm": 2.309228727065534, "kl": 0.3447265625, "learning_rate": 9.575540071801917e-07, "loss": 0.0566, "reward": 2.4283636808395386, "reward_std": 0.15362592600286007, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.012608693446964025, "rewards/tag_count_reward": 0.9895833432674408, "step": 308 }, { "clip_ratio": 0.0, "completion_length": 465.18751525878906, "epoch": 0.22653958944281524, "grad_norm": 2.614179249326471, "kl": 0.384765625, "learning_rate": 9.570641832785327e-07, "loss": 0.094, "reward": 2.6822060346603394, "reward_std": 0.13260857947170734, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01571073057129979, "rewards/tag_count_reward": 0.9895833432674408, "step": 309 }, { "clip_ratio": 0.0, "completion_length": 450.7708435058594, "epoch": 0.22727272727272727, "grad_norm": 2.417871183482608, "kl": 0.4619140625, "learning_rate": 9.565716908395955e-07, "loss": 0.0293, "reward": 2.8668872117996216, "reward_std": 0.2632347270846367, "rewards/accuracy_reward": 0.8958333730697632, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018529420718550682, "rewards/tag_count_reward": 0.9895833432674408, "step": 310 }, { "clip_ratio": 0.0, "completion_length": 432.56251525878906, "epoch": 0.2280058651026393, "grad_norm": 3.0765232905747797, "kl": 0.4814453125, "learning_rate": 9.56076533091948e-07, "loss": -0.0079, "reward": 2.4041491746902466, "reward_std": 0.23771148175001144, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.012517540715634823, "rewards/tag_count_reward": 0.9791666865348816, "step": 311 }, { "clip_ratio": 0.0, "completion_length": 469.8958435058594, "epoch": 0.2287390029325513, "grad_norm": 2.240908886249793, "kl": 0.4716796875, "learning_rate": 9.55578713281631e-07, "loss": 0.0439, "reward": 2.4784440994262695, "reward_std": 0.030433733016252518, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.014611462131142616, "rewards/tag_count_reward": 1.0, "step": 312 }, { "clip_ratio": 0.0, "completion_length": 455.1666717529297, "epoch": 0.22947214076246333, "grad_norm": 2.520573028942589, "kl": 0.4326171875, "learning_rate": 9.550782346721367e-07, "loss": 0.0075, "reward": 2.711169958114624, "reward_std": 0.07687797211110592, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.011052338872104883, "rewards/tag_count_reward": 0.9791666865348816, "step": 313 }, { "clip_ratio": 0.0, "completion_length": 428.8541717529297, "epoch": 0.23020527859237536, "grad_norm": 2.3656359919816317, "kl": 0.357421875, "learning_rate": 9.545751005443868e-07, "loss": 0.038, "reward": 2.6742382049560547, "reward_std": 0.11472383048385382, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 0.9722222089767456, "rewards/repetition_penalty_reward": -0.016734230797737837, "rewards/tag_count_reward": 0.9895833432674408, "step": 314 }, { "clip_ratio": 0.0, "completion_length": 444.85418701171875, "epoch": 0.23093841642228738, "grad_norm": 2.3225730368171638, "kl": 0.3291015625, "learning_rate": 9.54069314196712e-07, "loss": 0.0479, "reward": 2.733163356781006, "reward_std": 0.029640105552971363, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.009892286732792854, "rewards/tag_count_reward": 1.0, "step": 315 }, { "clip_ratio": 0.0, "completion_length": 441.16668701171875, "epoch": 0.2316715542521994, "grad_norm": 2.0945568663598833, "kl": 0.330078125, "learning_rate": 9.535608789448292e-07, "loss": -0.0054, "reward": 2.2132298946380615, "reward_std": 0.074584124609828, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.015936759300529957, "rewards/tag_count_reward": 1.0, "step": 316 }, { "clip_ratio": 0.0, "completion_length": 445.9166717529297, "epoch": 0.23240469208211142, "grad_norm": 2.2173137518118557, "kl": 0.31640625, "learning_rate": 9.53049798121821e-07, "loss": 0.0195, "reward": 2.6972047090530396, "reward_std": 0.1472337245941162, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.011128632817417383, "rewards/tag_count_reward": 0.9791666865348816, "step": 317 }, { "clip_ratio": 0.0, "completion_length": 434.0208435058594, "epoch": 0.23313782991202345, "grad_norm": 2.2714359196119216, "kl": 0.298828125, "learning_rate": 9.525360750781127e-07, "loss": 0.023, "reward": 2.46536922454834, "reward_std": 0.05973909609019756, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.013797647785395384, "rewards/tag_count_reward": 1.0, "step": 318 }, { "clip_ratio": 0.0, "completion_length": 472.50001525878906, "epoch": 0.23387096774193547, "grad_norm": 2.3225641603627203, "kl": 0.310546875, "learning_rate": 9.520197131814513e-07, "loss": -0.0068, "reward": 2.4521666765213013, "reward_std": 0.10633763670921326, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.013111229985952377, "rewards/tag_count_reward": 0.9791666865348816, "step": 319 }, { "clip_ratio": 0.0, "completion_length": 471.54168701171875, "epoch": 0.23460410557184752, "grad_norm": 2.7993035724053668, "kl": 0.2958984375, "learning_rate": 9.515007158168826e-07, "loss": 0.0932, "reward": 2.487420082092285, "reward_std": 0.01015830459073186, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.012580114882439375, "rewards/tag_count_reward": 1.0, "step": 320 }, { "clip_ratio": 0.0, "completion_length": 462.37501525878906, "epoch": 0.23533724340175954, "grad_norm": 2.299642746627153, "kl": 0.3056640625, "learning_rate": 9.509790863867295e-07, "loss": 0.0191, "reward": 2.6968696117401123, "reward_std": 0.13661886751651764, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.014936027117073536, "rewards/tag_count_reward": 0.9895833432674408, "step": 321 }, { "clip_ratio": 0.0, "completion_length": 464.31251525878906, "epoch": 0.23607038123167157, "grad_norm": 2.302831523396359, "kl": 0.3037109375, "learning_rate": 9.504548283105699e-07, "loss": 0.0451, "reward": 2.207852602005005, "reward_std": 0.10325483605265617, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.01436968706548214, "rewards/tag_count_reward": 1.0, "step": 322 }, { "clip_ratio": 0.0, "completion_length": 450.7916717529297, "epoch": 0.2368035190615836, "grad_norm": 2.5294492963009243, "kl": 0.3466796875, "learning_rate": 9.499279450252134e-07, "loss": 0.0435, "reward": 2.4603819847106934, "reward_std": 0.07487065717577934, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.01531249936670065, "rewards/tag_count_reward": 0.9895833432674408, "step": 323 }, { "clip_ratio": 0.0, "completion_length": 480.0625, "epoch": 0.2375366568914956, "grad_norm": 2.895694121410533, "kl": 0.369140625, "learning_rate": 9.493984399846797e-07, "loss": 0.0487, "reward": 2.2224709391593933, "reward_std": 0.04628606699407101, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.010167931206524372, "rewards/tag_count_reward": 0.9895833432674408, "step": 324 }, { "clip_ratio": 0.0, "completion_length": 439.1875, "epoch": 0.23826979472140764, "grad_norm": 2.8875499359334356, "kl": 0.3447265625, "learning_rate": 9.488663166601758e-07, "loss": 0.0641, "reward": 2.215409219264984, "reward_std": 0.07832126133143902, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.01375747099518776, "rewards/tag_count_reward": 1.0, "step": 325 }, { "clip_ratio": 0.0, "completion_length": 504.1458435058594, "epoch": 0.23900293255131966, "grad_norm": 3.1100488401558453, "kl": 0.40234375, "learning_rate": 9.483315785400726e-07, "loss": 0.1025, "reward": 2.4860798120498657, "reward_std": 0.008559350855648518, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013920174911618233, "rewards/tag_count_reward": 1.0, "step": 326 }, { "clip_ratio": 0.0, "completion_length": 478.77085876464844, "epoch": 0.23973607038123168, "grad_norm": 4.380042599170408, "kl": 0.435546875, "learning_rate": 9.477942291298827e-07, "loss": 0.0878, "reward": 2.7181226015090942, "reward_std": 0.08395953010767698, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.009308004518970847, "rewards/tag_count_reward": 0.984375, "step": 327 }, { "clip_ratio": 0.0, "completion_length": 486.16668701171875, "epoch": 0.2404692082111437, "grad_norm": 5.1350823066691245, "kl": 0.4521484375, "learning_rate": 9.472542719522371e-07, "loss": 0.0624, "reward": 2.9456522464752197, "reward_std": 0.14936763793230057, "rewards/accuracy_reward": 0.9791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.012681130785495043, "rewards/tag_count_reward": 0.9791666865348816, "step": 328 }, { "clip_ratio": 0.0, "completion_length": 477.2708435058594, "epoch": 0.24120234604105573, "grad_norm": 24.607110481841282, "kl": 1.037109375, "learning_rate": 9.467117105468623e-07, "loss": 0.1827, "reward": 2.4695119857788086, "reward_std": 0.07141697756014764, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.013127019628882408, "rewards/tag_count_reward": 0.9895833432674408, "step": 329 }, { "clip_ratio": 0.0, "completion_length": 452.8333435058594, "epoch": 0.24193548387096775, "grad_norm": 65.55752383537028, "kl": 1.91015625, "learning_rate": 9.461665484705571e-07, "loss": 0.1293, "reward": 2.2242066860198975, "reward_std": 0.04578477330505848, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010168405715376139, "rewards/tag_count_reward": 0.984375, "step": 330 }, { "clip_ratio": 0.0, "completion_length": 460.18751525878906, "epoch": 0.24266862170087977, "grad_norm": 27.520381811057142, "kl": 1.36328125, "learning_rate": 9.456187892971687e-07, "loss": 0.1238, "reward": 2.4319708347320557, "reward_std": 0.148020438849926, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.009001555386930704, "rewards/tag_count_reward": 0.9687500298023224, "step": 331 }, { "clip_ratio": 0.0, "completion_length": 470.9166717529297, "epoch": 0.2434017595307918, "grad_norm": 5.897429847753733, "kl": 0.580078125, "learning_rate": 9.450684366175703e-07, "loss": 0.0808, "reward": 2.2141419649124146, "reward_std": 0.08268354088068008, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.013288679532706738, "rewards/tag_count_reward": 0.9843750298023224, "step": 332 }, { "clip_ratio": 0.0, "completion_length": 459.56251525878906, "epoch": 0.24413489736070382, "grad_norm": 4.821721286003609, "kl": 0.48828125, "learning_rate": 9.445154940396367e-07, "loss": 0.0482, "reward": 2.23613703250885, "reward_std": 0.008423366351053119, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01386305270716548, "rewards/tag_count_reward": 1.0, "step": 333 }, { "clip_ratio": 0.0, "completion_length": 461.72918701171875, "epoch": 0.24486803519061584, "grad_norm": 4.5305477335749424, "kl": 0.39453125, "learning_rate": 9.439599651882211e-07, "loss": 0.0641, "reward": 1.9890110492706299, "reward_std": 0.12262681126594543, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010989056900143623, "rewards/tag_count_reward": 0.9791666865348816, "step": 334 }, { "clip_ratio": 0.0, "completion_length": 429.4166717529297, "epoch": 0.24560117302052786, "grad_norm": 3.0480664219928526, "kl": 0.3857421875, "learning_rate": 9.43401853705131e-07, "loss": 0.0616, "reward": 2.2384225726127625, "reward_std": 0.008934499695897102, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.011577559635043144, "rewards/tag_count_reward": 1.0, "step": 335 }, { "clip_ratio": 0.0, "completion_length": 443.37501525878906, "epoch": 0.24633431085043989, "grad_norm": 2.9009848462349264, "kl": 0.443359375, "learning_rate": 9.428411632491048e-07, "loss": 0.054, "reward": 2.1951489448547363, "reward_std": 0.10253655072301626, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013184436596930027, "rewards/tag_count_reward": 1.0, "step": 336 }, { "clip_ratio": 0.0, "completion_length": 432.5833435058594, "epoch": 0.2470674486803519, "grad_norm": 7.098171576738261, "kl": 0.4609375, "learning_rate": 9.422778974957873e-07, "loss": -0.0096, "reward": 2.2252637147903442, "reward_std": 0.20562057942152023, "rewards/accuracy_reward": 0.25000000558793545, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.007375375367701054, "rewards/tag_count_reward": 0.9895833432674408, "step": 337 }, { "clip_ratio": 0.0, "completion_length": 456.5833435058594, "epoch": 0.24780058651026393, "grad_norm": 4.45344368844307, "kl": 0.4345703125, "learning_rate": 9.41712060137706e-07, "loss": 0.0747, "reward": 2.2418224811553955, "reward_std": 0.006912755314260721, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.008177584502846003, "rewards/tag_count_reward": 1.0, "step": 338 }, { "clip_ratio": 0.0, "completion_length": 496.27085876464844, "epoch": 0.24853372434017595, "grad_norm": 6.6523792877479675, "kl": 0.51953125, "learning_rate": 9.411436548842465e-07, "loss": 0.1281, "reward": 2.1983273029327393, "reward_std": 0.1371548529714346, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.013478230219334364, "rewards/tag_count_reward": 0.9895833432674408, "step": 339 }, { "clip_ratio": 0.0, "completion_length": 474.12501525878906, "epoch": 0.24926686217008798, "grad_norm": 3.555591622399211, "kl": 0.599609375, "learning_rate": 9.405726854616288e-07, "loss": 0.0606, "reward": 2.445413589477539, "reward_std": 0.10354876844212413, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.012919682543724775, "rewards/tag_count_reward": 1.0, "step": 340 }, { "clip_ratio": 0.0, "completion_length": 459.8333435058594, "epoch": 0.25, "grad_norm": 4.3052672307178526, "kl": 0.67578125, "learning_rate": 9.399991556128821e-07, "loss": 0.0259, "reward": 2.4495670795440674, "reward_std": 0.14659255044534802, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.00876629538834095, "rewards/tag_count_reward": 1.0, "step": 341 }, { "clip_ratio": 0.0, "completion_length": 447.7708435058594, "epoch": 0.250733137829912, "grad_norm": 7.772010913258851, "kl": 0.564453125, "learning_rate": 9.394230690978209e-07, "loss": 0.0445, "reward": 2.4938722252845764, "reward_std": 0.0052741896361112595, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.00612785667181015, "rewards/tag_count_reward": 1.0, "step": 342 }, { "clip_ratio": 0.0, "completion_length": 432.29168701171875, "epoch": 0.25146627565982405, "grad_norm": 5.815094134114697, "kl": 0.5634765625, "learning_rate": 9.388444296930198e-07, "loss": 0.0456, "reward": 2.4859039783477783, "reward_std": 0.022549844114109874, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.00888778269290924, "rewards/tag_count_reward": 0.9947916865348816, "step": 343 }, { "clip_ratio": 0.0, "completion_length": 478.0208435058594, "epoch": 0.25219941348973607, "grad_norm": 9.393448474685972, "kl": 0.58203125, "learning_rate": 9.382632411917896e-07, "loss": 0.0719, "reward": 2.7372546195983887, "reward_std": 0.007755943341180682, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.012745357118546963, "rewards/tag_count_reward": 1.0, "step": 344 }, { "clip_ratio": 0.0, "completion_length": 434.4791717529297, "epoch": 0.2529325513196481, "grad_norm": 5.15109218597285, "kl": 0.611328125, "learning_rate": 9.376795074041511e-07, "loss": 0.0736, "reward": 1.992047905921936, "reward_std": 0.007327872794121504, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0079521588049829, "rewards/tag_count_reward": 1.0, "step": 345 }, { "clip_ratio": 0.0, "completion_length": 523.4375152587891, "epoch": 0.2536656891495601, "grad_norm": 32.341677849946365, "kl": 1.18359375, "learning_rate": 9.370932321568116e-07, "loss": 0.2629, "reward": 2.4761067628860474, "reward_std": 0.04537955252453685, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01347667258232832, "rewards/tag_count_reward": 0.9895833432674408, "step": 346 }, { "clip_ratio": 0.0, "completion_length": 522.5, "epoch": 0.25439882697947214, "grad_norm": 28.958564691751725, "kl": 2.06640625, "learning_rate": 9.36504419293139e-07, "loss": 0.2469, "reward": 2.23344486951828, "reward_std": 0.02600767184048891, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01134685194119811, "rewards/tag_count_reward": 0.9947916865348816, "step": 347 }, { "clip_ratio": 0.0, "completion_length": 511.7083435058594, "epoch": 0.25513196480938416, "grad_norm": 50.041207601652744, "kl": 2.23046875, "learning_rate": 9.359130726731364e-07, "loss": 0.2488, "reward": 2.407168984413147, "reward_std": 0.16616932256147265, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.016442209482192993, "rewards/tag_count_reward": 0.9791666865348816, "step": 348 }, { "clip_ratio": 0.0, "completion_length": 468.9583435058594, "epoch": 0.2558651026392962, "grad_norm": 8.90761662789548, "kl": 1.38671875, "learning_rate": 9.353191961734172e-07, "loss": 0.151, "reward": 2.44021075963974, "reward_std": 0.12433899194002151, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.011178134009242058, "rewards/tag_count_reward": 1.0, "step": 349 }, { "clip_ratio": 0.0, "completion_length": 452.3333435058594, "epoch": 0.2565982404692082, "grad_norm": 27.671954739000196, "kl": 0.587890625, "learning_rate": 9.347227936871798e-07, "loss": 0.153, "reward": 2.46974778175354, "reward_std": 0.07871474791318178, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.009418966248631477, "rewards/tag_count_reward": 1.0, "step": 350 }, { "clip_ratio": 0.0, "completion_length": 426.06251525878906, "epoch": 0.25733137829912023, "grad_norm": 4.774439255958196, "kl": 0.546875, "learning_rate": 9.341238691241817e-07, "loss": 0.0509, "reward": 2.4622308015823364, "reward_std": 0.1002035103738308, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.00999134499579668, "rewards/tag_count_reward": 1.0, "step": 351 }, { "clip_ratio": 0.0, "completion_length": 454.1041717529297, "epoch": 0.25806451612903225, "grad_norm": 7.830072090889241, "kl": 0.47265625, "learning_rate": 9.33522426410714e-07, "loss": 0.0825, "reward": 2.218370020389557, "reward_std": 0.07793176430277526, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010796691058203578, "rewards/tag_count_reward": 1.0, "step": 352 }, { "clip_ratio": 0.0, "completion_length": 443.3125, "epoch": 0.2587976539589443, "grad_norm": 5.924916968606259, "kl": 0.966796875, "learning_rate": 9.329184694895761e-07, "loss": 0.0849, "reward": 2.240042746067047, "reward_std": 0.006934693781659007, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.00995726091787219, "rewards/tag_count_reward": 1.0, "step": 353 }, { "clip_ratio": 0.0, "completion_length": 480.62501525878906, "epoch": 0.2595307917888563, "grad_norm": 17.74912709208456, "kl": 1.1953125, "learning_rate": 9.323120023200489e-07, "loss": 0.0712, "reward": 2.201182782649994, "reward_std": 0.10400985646992922, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.00715067470446229, "rewards/tag_count_reward": 1.0, "step": 354 }, { "clip_ratio": 0.0, "completion_length": 430.87501525878906, "epoch": 0.2602639296187683, "grad_norm": 5.313270709600229, "kl": 0.552734375, "learning_rate": 9.317030288778698e-07, "loss": 0.0269, "reward": 2.7366254329681396, "reward_std": 0.009935126174241304, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013374709524214268, "rewards/tag_count_reward": 1.0, "step": 355 }, { "clip_ratio": 0.0, "completion_length": 443.50001525878906, "epoch": 0.26099706744868034, "grad_norm": 8.580062044235554, "kl": 0.4072265625, "learning_rate": 9.31091553155206e-07, "loss": -0.008, "reward": 2.261789321899414, "reward_std": 0.07716286275535822, "rewards/accuracy_reward": 0.27083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.009044124279171228, "rewards/tag_count_reward": 1.0, "step": 356 }, { "clip_ratio": 0.0, "completion_length": 439.87501525878906, "epoch": 0.26173020527859236, "grad_norm": 3.3894194064813576, "kl": 0.3427734375, "learning_rate": 9.30477579160629e-07, "loss": 0.0268, "reward": 2.9032580852508545, "reward_std": 0.2916463017463684, "rewards/accuracy_reward": 0.9166666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013408492784947157, "rewards/tag_count_reward": 1.0, "step": 357 }, { "clip_ratio": 0.0, "completion_length": 444.87501525878906, "epoch": 0.2624633431085044, "grad_norm": 3.2273474029147002, "kl": 0.333984375, "learning_rate": 9.298611109190873e-07, "loss": 0.0711, "reward": 2.4881826639175415, "reward_std": 0.010318218730390072, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.011817488819360733, "rewards/tag_count_reward": 1.0, "step": 358 }, { "clip_ratio": 0.0, "completion_length": 411.5416717529297, "epoch": 0.2631964809384164, "grad_norm": 2.7387770155536293, "kl": 0.330078125, "learning_rate": 9.29242152471881e-07, "loss": 0.0634, "reward": 2.7406634092330933, "reward_std": 0.006734333001077175, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.009336648043245077, "rewards/tag_count_reward": 1.0, "step": 359 }, { "clip_ratio": 0.0, "completion_length": 453.85418701171875, "epoch": 0.26392961876832843, "grad_norm": 2.420440222664315, "kl": 0.322265625, "learning_rate": 9.286207078766348e-07, "loss": -0.0136, "reward": 2.707916021347046, "reward_std": 0.11361571587622166, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010834058281034231, "rewards/tag_count_reward": 0.9895833432674408, "step": 360 }, { "clip_ratio": 0.0, "completion_length": 456.29168701171875, "epoch": 0.26466275659824046, "grad_norm": 2.2704308739154233, "kl": 0.322265625, "learning_rate": 9.279967812072717e-07, "loss": 0.0326, "reward": 2.7088087797164917, "reward_std": 0.1134856294374913, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.009941152296960354, "rewards/tag_count_reward": 0.9895833432674408, "step": 361 }, { "clip_ratio": 0.0, "completion_length": 449.54168701171875, "epoch": 0.2653958944281525, "grad_norm": 2.6765119623946028, "kl": 0.3115234375, "learning_rate": 9.273703765539856e-07, "loss": 0.0532, "reward": 2.6793943643569946, "reward_std": 0.14904301008209586, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.008105819579213858, "rewards/tag_count_reward": 0.9791666865348816, "step": 362 }, { "clip_ratio": 0.0, "completion_length": 472.6458435058594, "epoch": 0.2661290322580645, "grad_norm": 2.334226266301674, "kl": 0.322265625, "learning_rate": 9.267414980232156e-07, "loss": 0.019, "reward": 2.674462676048279, "reward_std": 0.17157460749149323, "rewards/accuracy_reward": 0.6875000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013037360273301601, "rewards/tag_count_reward": 1.0, "step": 363 }, { "clip_ratio": 0.0, "completion_length": 462.2916717529297, "epoch": 0.2668621700879765, "grad_norm": 2.3774041304092175, "kl": 0.328125, "learning_rate": 9.261101497376183e-07, "loss": 0.045, "reward": 2.487114191055298, "reward_std": 0.008130959002301097, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.012885776348412037, "rewards/tag_count_reward": 1.0, "step": 364 }, { "clip_ratio": 0.0, "completion_length": 446.4166717529297, "epoch": 0.26759530791788855, "grad_norm": 2.2848004947697653, "kl": 0.296875, "learning_rate": 9.254763358360404e-07, "loss": 0.0543, "reward": 2.480908155441284, "reward_std": 0.039752448443323374, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.008675360586494207, "rewards/tag_count_reward": 0.9895833432674408, "step": 365 }, { "clip_ratio": 0.0, "completion_length": 429.8958435058594, "epoch": 0.26832844574780057, "grad_norm": 4.808091297344941, "kl": 0.337890625, "learning_rate": 9.248400604734929e-07, "loss": 0.0747, "reward": 2.193628668785095, "reward_std": 0.16610116511583328, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.009496402461081743, "rewards/tag_count_reward": 0.9739583432674408, "step": 366 }, { "clip_ratio": 0.0, "completion_length": 441.66668701171875, "epoch": 0.2690615835777126, "grad_norm": 2.720687783905429, "kl": 0.333984375, "learning_rate": 9.242013278211226e-07, "loss": 0.046, "reward": 2.4815183877944946, "reward_std": 0.02494870498776436, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013273344840854406, "rewards/tag_count_reward": 0.9947916865348816, "step": 367 }, { "clip_ratio": 0.0, "completion_length": 479.6458435058594, "epoch": 0.2697947214076246, "grad_norm": 2.362309533450861, "kl": 0.3330078125, "learning_rate": 9.235601420661854e-07, "loss": 0.0187, "reward": 2.4759992361068726, "reward_std": 0.04343183804303408, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013584255706518888, "rewards/tag_count_reward": 0.9895833432674408, "step": 368 }, { "clip_ratio": 0.0, "completion_length": 416.43751525878906, "epoch": 0.27052785923753664, "grad_norm": 2.899002651460847, "kl": 0.359375, "learning_rate": 9.22916507412019e-07, "loss": 0.0293, "reward": 2.932681679725647, "reward_std": 0.19901539757847786, "rewards/accuracy_reward": 0.9583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010026867035776377, "rewards/tag_count_reward": 0.9843750298023224, "step": 369 }, { "clip_ratio": 0.0, "completion_length": 445.7291717529297, "epoch": 0.27126099706744866, "grad_norm": 2.5066447293184693, "kl": 0.4091796875, "learning_rate": 9.222704280780141e-07, "loss": 0.0151, "reward": 2.1932421922683716, "reward_std": 0.12493787333369255, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.008146827574819326, "rewards/tag_count_reward": 1.0, "step": 370 }, { "clip_ratio": 0.0, "completion_length": 458.97918701171875, "epoch": 0.2719941348973607, "grad_norm": 2.5236209781956798, "kl": 0.431640625, "learning_rate": 9.216219082995888e-07, "loss": 0.0406, "reward": 2.231236457824707, "reward_std": 0.029424049891531467, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.011819119565188885, "rewards/tag_count_reward": 1.0, "step": 371 }, { "clip_ratio": 0.0, "completion_length": 463.9583435058594, "epoch": 0.2727272727272727, "grad_norm": 2.617827825257977, "kl": 0.52734375, "learning_rate": 9.209709523281587e-07, "loss": 0.0485, "reward": 2.238226532936096, "reward_std": 0.00962640717625618, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.011773544363677502, "rewards/tag_count_reward": 1.0, "step": 372 }, { "clip_ratio": 0.0, "completion_length": 514.4166870117188, "epoch": 0.27346041055718473, "grad_norm": 19.050904643830204, "kl": 0.6015625, "learning_rate": 9.20317564431111e-07, "loss": 0.2268, "reward": 2.730633020401001, "reward_std": 0.025867179967463017, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.014158804900944233, "rewards/tag_count_reward": 0.9947916865348816, "step": 373 }, { "clip_ratio": 0.0, "completion_length": 472.00001525878906, "epoch": 0.27419354838709675, "grad_norm": 5.424273414099897, "kl": 0.9375, "learning_rate": 9.196617488917744e-07, "loss": 0.1183, "reward": 2.421747326850891, "reward_std": 0.22977232933044434, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.008808402344584465, "rewards/tag_count_reward": 0.9791666865348816, "step": 374 }, { "clip_ratio": 0.0, "completion_length": 433.2291717529297, "epoch": 0.2749266862170088, "grad_norm": 6.51976669479584, "kl": 1.08203125, "learning_rate": 9.190035100093934e-07, "loss": 0.0622, "reward": 2.4945380687713623, "reward_std": 0.005227633053436875, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.005462077213451266, "rewards/tag_count_reward": 1.0, "step": 375 }, { "clip_ratio": 0.0, "completion_length": 464.89585876464844, "epoch": 0.2756598240469208, "grad_norm": 12.551923228481087, "kl": 2.0859375, "learning_rate": 9.183428520990979e-07, "loss": 0.2216, "reward": 2.7031989097595215, "reward_std": 0.13187365233898163, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010342813329771161, "rewards/tag_count_reward": 0.984375, "step": 376 }, { "clip_ratio": 0.0, "completion_length": 482.9375305175781, "epoch": 0.2763929618768328, "grad_norm": 16.830833423074687, "kl": 2.2734375, "learning_rate": 9.176797794918766e-07, "loss": 0.3048, "reward": 2.2347198724746704, "reward_std": 0.024037683382630348, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010071944678202271, "rewards/tag_count_reward": 0.9947916865348816, "step": 377 }, { "clip_ratio": 0.0, "completion_length": 451.875, "epoch": 0.27712609970674484, "grad_norm": 5.46041924584287, "kl": 1.14453125, "learning_rate": 9.170142965345474e-07, "loss": 0.072, "reward": 2.2421175837516785, "reward_std": 0.005587070249021053, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.00788251287303865, "rewards/tag_count_reward": 1.0, "step": 378 }, { "clip_ratio": 0.0, "completion_length": 460.6458435058594, "epoch": 0.27785923753665687, "grad_norm": 7.285649773138782, "kl": 1.068359375, "learning_rate": 9.163464075897298e-07, "loss": 0.0662, "reward": 2.491139054298401, "reward_std": 0.0070568344090133905, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.008861158974468708, "rewards/tag_count_reward": 1.0, "step": 379 }, { "clip_ratio": 0.0, "completion_length": 480.29168701171875, "epoch": 0.2785923753665689, "grad_norm": 3.9106992141973937, "kl": 0.5595703125, "learning_rate": 9.156761170358157e-07, "loss": 0.0285, "reward": 2.4897278547286987, "reward_std": 0.007526291534304619, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010272293351590633, "rewards/tag_count_reward": 1.0, "step": 380 }, { "clip_ratio": 0.0, "completion_length": 417.0, "epoch": 0.2793255131964809, "grad_norm": 3.1702715469625407, "kl": 0.3798828125, "learning_rate": 9.15003429266941e-07, "loss": 0.0529, "reward": 2.733282208442688, "reward_std": 0.029079753905534744, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.009773501195013523, "rewards/tag_count_reward": 1.0, "step": 381 }, { "clip_ratio": 0.0, "completion_length": 448.0833435058594, "epoch": 0.28005865102639294, "grad_norm": 4.944163236819044, "kl": 0.3662109375, "learning_rate": 9.143283486929567e-07, "loss": 0.0531, "reward": 2.7383430004119873, "reward_std": 0.007190744858235121, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.011657027993351221, "rewards/tag_count_reward": 1.0, "step": 382 }, { "clip_ratio": 0.0, "completion_length": 437.1458435058594, "epoch": 0.28079178885630496, "grad_norm": 3.016754740608089, "kl": 0.38671875, "learning_rate": 9.136508797393999e-07, "loss": 0.0325, "reward": 2.474005341529846, "reward_std": 0.0372047598939389, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.012105878442525864, "rewards/tag_count_reward": 1.0, "step": 383 }, { "clip_ratio": 0.0, "completion_length": 416.75001525878906, "epoch": 0.28152492668621704, "grad_norm": 2.902012592999401, "kl": 0.314453125, "learning_rate": 9.129710268474652e-07, "loss": 0.0353, "reward": 2.4901161789894104, "reward_std": 0.007005092455074191, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.009883841034024954, "rewards/tag_count_reward": 1.0, "step": 384 }, { "clip_ratio": 0.0, "completion_length": 422.2291717529297, "epoch": 0.28225806451612906, "grad_norm": 4.99704224289973, "kl": 0.40234375, "learning_rate": 9.122887944739747e-07, "loss": 0.0667, "reward": 2.4585758447647095, "reward_std": 0.11278056120499969, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01017436245456338, "rewards/tag_count_reward": 0.9895833432674408, "step": 385 }, { "clip_ratio": 0.0, "completion_length": 417.93751525878906, "epoch": 0.2829912023460411, "grad_norm": 3.207572366812953, "kl": 0.3779296875, "learning_rate": 9.116041870913498e-07, "loss": 0.0314, "reward": 2.7009975910186768, "reward_std": 0.12883948348462582, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.012544109020382166, "rewards/tag_count_reward": 0.9843750298023224, "step": 386 }, { "clip_ratio": 0.0, "completion_length": 411.7916717529297, "epoch": 0.2837243401759531, "grad_norm": 3.455020971556175, "kl": 0.4365234375, "learning_rate": 9.109172091875813e-07, "loss": 0.0533, "reward": 2.4524088501930237, "reward_std": 0.11647403798997402, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.016341304872184992, "rewards/tag_count_reward": 0.9895833432674408, "step": 387 }, { "clip_ratio": 0.0, "completion_length": 452.56251525878906, "epoch": 0.2844574780058651, "grad_norm": 5.021925533382141, "kl": 0.724609375, "learning_rate": 9.102278652662004e-07, "loss": 0.1052, "reward": 2.488985061645508, "reward_std": 0.006767880171537399, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01101499842479825, "rewards/tag_count_reward": 1.0, "step": 388 }, { "clip_ratio": 0.0, "completion_length": 444.9166717529297, "epoch": 0.28519061583577715, "grad_norm": 4.237467170835808, "kl": 0.658203125, "learning_rate": 9.095361598462483e-07, "loss": 0.0624, "reward": 2.239302396774292, "reward_std": 0.008108881767839193, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010697626508772373, "rewards/tag_count_reward": 1.0, "step": 389 }, { "clip_ratio": 0.0, "completion_length": 530.1875, "epoch": 0.2859237536656892, "grad_norm": 28.582624592710438, "kl": 1.0234375, "learning_rate": 9.088420974622477e-07, "loss": 0.3534, "reward": 2.409157693386078, "reward_std": 0.1983821764588356, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.009245133958756924, "rewards/tag_count_reward": 0.9531250298023224, "step": 390 }, { "clip_ratio": 0.0, "completion_length": 436.5, "epoch": 0.2866568914956012, "grad_norm": 16.488952128647092, "kl": 1.00390625, "learning_rate": 9.081456826641725e-07, "loss": 0.2223, "reward": 2.2143354415893555, "reward_std": 0.09019448328763247, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.011359069496393204, "rewards/tag_count_reward": 0.9895833432674408, "step": 391 }, { "clip_ratio": 0.0, "completion_length": 658.8125305175781, "epoch": 0.2873900293255132, "grad_norm": 39.80719794913905, "kl": 4.6328125, "learning_rate": 9.074469200174174e-07, "loss": 0.4892, "reward": 2.1642203330993652, "reward_std": 0.3343081623315811, "rewards/accuracy_reward": 0.2708333395421505, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.0059186555445194244, "rewards/tag_count_reward": 0.9062500298023224, "step": 392 }, { "clip_ratio": 0.0, "completion_length": 592.4375, "epoch": 0.28812316715542524, "grad_norm": 18.491008323344726, "kl": 3.28125, "learning_rate": 9.067458141027694e-07, "loss": 0.3781, "reward": 2.3546512126922607, "reward_std": 0.2867864668369293, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.008195979055017233, "rewards/tag_count_reward": 0.9531250298023224, "step": 393 }, { "clip_ratio": 0.0, "completion_length": 464.0625, "epoch": 0.28885630498533726, "grad_norm": 9.410508125378785, "kl": 1.158203125, "learning_rate": 9.060423695163764e-07, "loss": 0.1123, "reward": 2.433341920375824, "reward_std": 0.14605514332652092, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.014574835076928139, "rewards/tag_count_reward": 0.9687500298023224, "step": 394 }, { "clip_ratio": 0.0, "completion_length": 453.18751525878906, "epoch": 0.2895894428152493, "grad_norm": 4.260076100230029, "kl": 0.810546875, "learning_rate": 9.053365908697181e-07, "loss": 0.0791, "reward": 2.447802424430847, "reward_std": 0.13183430582284927, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010531116742640734, "rewards/tag_count_reward": 0.9791666865348816, "step": 395 }, { "clip_ratio": 0.0, "completion_length": 511.1875, "epoch": 0.2903225806451613, "grad_norm": 31.980218688565657, "kl": 0.724609375, "learning_rate": 9.046284827895746e-07, "loss": 0.3394, "reward": 2.4297736287117004, "reward_std": 0.2000921592116356, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.012934757862240076, "rewards/tag_count_reward": 0.9635416865348816, "step": 396 }, { "clip_ratio": 0.0, "completion_length": 420.37501525878906, "epoch": 0.29105571847507333, "grad_norm": 3.3699821366968212, "kl": 0.501953125, "learning_rate": 9.039180499179973e-07, "loss": 0.0411, "reward": 2.673474907875061, "reward_std": 0.11912347562611103, "rewards/accuracy_reward": 0.6875, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.014025256969034672, "rewards/tag_count_reward": 1.0, "step": 397 }, { "clip_ratio": 0.0, "completion_length": 571.7708435058594, "epoch": 0.29178885630498536, "grad_norm": 19.872465311776544, "kl": 1.212890625, "learning_rate": 9.032052969122781e-07, "loss": 0.35, "reward": 2.1702736616134644, "reward_std": 0.17028960585594177, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017226402647793293, "rewards/tag_count_reward": 0.9583333432674408, "step": 398 }, { "clip_ratio": 0.0, "completion_length": 441.68751525878906, "epoch": 0.2925219941348974, "grad_norm": 14.8359150453612, "kl": 0.9453125, "learning_rate": 9.024902284449181e-07, "loss": 0.2402, "reward": 2.6931817531585693, "reward_std": 0.14234615676105022, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.013415634166449308, "rewards/tag_count_reward": 0.9843750298023224, "step": 399 }, { "clip_ratio": 0.0, "completion_length": 448.6666717529297, "epoch": 0.2932551319648094, "grad_norm": 12.406429499182185, "kl": 0.853515625, "learning_rate": 9.017728492035978e-07, "loss": 0.0062, "reward": 2.2172019481658936, "reward_std": 0.04633156303316355, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.015437084715813398, "rewards/tag_count_reward": 0.9895833432674408, "step": 400 }, { "clip_ratio": 0.0, "completion_length": 481.2083435058594, "epoch": 0.2939882697947214, "grad_norm": 13.871825287384443, "kl": 1.33984375, "learning_rate": 9.010531638911465e-07, "loss": 0.0515, "reward": 2.1981945037841797, "reward_std": 0.14712171256542206, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010138920042663813, "rewards/tag_count_reward": 0.9791666865348816, "step": 401 }, { "clip_ratio": 0.0, "completion_length": 414.9583435058594, "epoch": 0.29472140762463345, "grad_norm": 3.378020607722317, "kl": 0.4638671875, "learning_rate": 9.003311772255106e-07, "loss": 0.0614, "reward": 1.9696611762046814, "reward_std": 0.047247909009456635, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.018186070024967194, "rewards/tag_count_reward": 0.9947916865348816, "step": 402 }, { "clip_ratio": 0.0, "completion_length": 429.54168701171875, "epoch": 0.29545454545454547, "grad_norm": 5.901425872850612, "kl": 0.4560546875, "learning_rate": 8.996068939397233e-07, "loss": -0.0004, "reward": 2.6546950340270996, "reward_std": 0.1604913668707013, "rewards/accuracy_reward": 0.6875, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.011971795465797186, "rewards/tag_count_reward": 0.9791666865348816, "step": 403 }, { "clip_ratio": 0.0, "completion_length": 421.43751525878906, "epoch": 0.2961876832844575, "grad_norm": 3.2421459777125703, "kl": 0.4638671875, "learning_rate": 8.98880318781874e-07, "loss": 0.0314, "reward": 2.4531060457229614, "reward_std": 0.07071313261985779, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.01564402598887682, "rewards/tag_count_reward": 0.9895833730697632, "step": 404 }, { "clip_ratio": 0.0, "completion_length": 498.1458435058594, "epoch": 0.2969208211143695, "grad_norm": 8.347266976997215, "kl": 1.50390625, "learning_rate": 8.981514565150758e-07, "loss": 0.2145, "reward": 2.5850037336349487, "reward_std": 0.260950468480587, "rewards/accuracy_reward": 0.6666666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01916305348277092, "rewards/tag_count_reward": 0.9375000298023224, "step": 405 }, { "clip_ratio": 0.0, "completion_length": 441.37501525878906, "epoch": 0.29765395894428154, "grad_norm": 12.481716504683666, "kl": 0.7548828125, "learning_rate": 8.974203119174354e-07, "loss": 0.2422, "reward": 2.203757107257843, "reward_std": 0.09926661476492882, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.013256918638944626, "rewards/tag_count_reward": 0.9739583432674408, "step": 406 }, { "clip_ratio": 0.0, "completion_length": 449.9583435058594, "epoch": 0.29838709677419356, "grad_norm": 20.555500703198362, "kl": 0.615234375, "learning_rate": 8.966868897820217e-07, "loss": -0.0515, "reward": 2.4867031574249268, "reward_std": 0.007832993287593126, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013296912424266338, "rewards/tag_count_reward": 1.0, "step": 407 }, { "clip_ratio": 0.0, "completion_length": 405.93751525878906, "epoch": 0.2991202346041056, "grad_norm": 4.055403353159463, "kl": 0.3583984375, "learning_rate": 8.95951194916834e-07, "loss": 0.0832, "reward": 2.4320058822631836, "reward_std": 0.18413013964891434, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.015910861548036337, "rewards/tag_count_reward": 0.9895833432674408, "step": 408 }, { "clip_ratio": 0.0, "completion_length": 438.5833435058594, "epoch": 0.2998533724340176, "grad_norm": 21.309615660717455, "kl": 0.556640625, "learning_rate": 8.952132321447707e-07, "loss": 0.2093, "reward": 2.435889959335327, "reward_std": 0.1703370437026024, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.010290783829987049, "rewards/tag_count_reward": 0.9739583432674408, "step": 409 }, { "clip_ratio": 0.0, "completion_length": 396.7916717529297, "epoch": 0.30058651026392963, "grad_norm": 4.617605546670348, "kl": 0.3681640625, "learning_rate": 8.944730063035973e-07, "loss": 0.0619, "reward": 2.2141241431236267, "reward_std": 0.0688196187838912, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.013306473847478628, "rewards/tag_count_reward": 0.984375, "step": 410 }, { "clip_ratio": 0.0, "completion_length": 433.31251525878906, "epoch": 0.30131964809384165, "grad_norm": 3.1262505616665806, "kl": 0.4921875, "learning_rate": 8.937305222459156e-07, "loss": 0.0334, "reward": 2.4272764921188354, "reward_std": 0.2919757291674614, "rewards/accuracy_reward": 0.45833333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010223534889519215, "rewards/tag_count_reward": 0.9791666865348816, "step": 411 }, { "clip_ratio": 0.0, "completion_length": 566.5, "epoch": 0.3020527859237537, "grad_norm": 14.530492359657364, "kl": 3.458984375, "learning_rate": 8.929857848391308e-07, "loss": 0.4275, "reward": 2.0301916003227234, "reward_std": 0.32898490130901337, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.009738977532833815, "rewards/tag_count_reward": 0.8802083432674408, "step": 412 }, { "clip_ratio": 0.0, "completion_length": 546.8125305175781, "epoch": 0.3027859237536657, "grad_norm": 16.103298225019277, "kl": 1.69921875, "learning_rate": 8.922387989654202e-07, "loss": 0.5684, "reward": 2.507889747619629, "reward_std": 0.48949065804481506, "rewards/accuracy_reward": 0.6250000298023224, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.011207538656890392, "rewards/tag_count_reward": 0.9010416865348816, "step": 413 }, { "clip_ratio": 0.0, "completion_length": 497.56251525878906, "epoch": 0.3035190615835777, "grad_norm": 8.332687659618854, "kl": 2.203125, "learning_rate": 8.914895695217011e-07, "loss": 0.513, "reward": 2.341689109802246, "reward_std": 0.316712461411953, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.014213801361620426, "rewards/tag_count_reward": 0.9322916865348816, "step": 414 }, { "clip_ratio": 0.0, "completion_length": 835.7500305175781, "epoch": 0.30425219941348974, "grad_norm": 39.2963962140608, "kl": 5.984375, "learning_rate": 8.907381014195989e-07, "loss": 0.7308, "reward": 2.089294672012329, "reward_std": 0.4655766934156418, "rewards/accuracy_reward": 0.3125000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.004455308895558119, "rewards/tag_count_reward": 0.78125, "step": 415 }, { "clip_ratio": 0.0, "completion_length": 724.8750305175781, "epoch": 0.30498533724340177, "grad_norm": 14.61779528445159, "kl": 4.7890625, "learning_rate": 8.899843995854142e-07, "loss": 1.091, "reward": 2.286299228668213, "reward_std": 0.3357388377189636, "rewards/accuracy_reward": 0.4166666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0157841881737113, "rewards/tag_count_reward": 0.8854166865348816, "step": 416 }, { "clip_ratio": 0.0, "completion_length": 600.0625305175781, "epoch": 0.3057184750733138, "grad_norm": 14.48002086498947, "kl": 3.3359375, "learning_rate": 8.892284689600917e-07, "loss": 0.4826, "reward": 2.311539053916931, "reward_std": 0.3173745572566986, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.013113727793097496, "rewards/tag_count_reward": 0.9218750298023224, "step": 417 }, { "clip_ratio": 0.0, "completion_length": 536.7916870117188, "epoch": 0.3064516129032258, "grad_norm": 12.734623287810342, "kl": 3.015625, "learning_rate": 8.88470314499186e-07, "loss": 0.5337, "reward": 2.267106056213379, "reward_std": 0.41217756271362305, "rewards/accuracy_reward": 0.375, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.012407994829118252, "rewards/tag_count_reward": 0.9114583432674408, "step": 418 }, { "clip_ratio": 0.0, "completion_length": 568.4791870117188, "epoch": 0.30718475073313783, "grad_norm": 16.767038731550095, "kl": 3.59375, "learning_rate": 8.877099411728314e-07, "loss": 0.4868, "reward": 2.0982948541641235, "reward_std": 0.2590210735797882, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.009344066493213177, "rewards/tag_count_reward": 0.9270833432674408, "step": 419 }, { "clip_ratio": 0.0, "completion_length": 474.7083435058594, "epoch": 0.30791788856304986, "grad_norm": 6.045360274295084, "kl": 1.833984375, "learning_rate": 8.869473539657074e-07, "loss": 0.2533, "reward": 2.378178358078003, "reward_std": 0.27451788214966655, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01244673365727067, "rewards/tag_count_reward": 0.953125, "step": 420 }, { "clip_ratio": 0.0, "completion_length": 454.0208435058594, "epoch": 0.3086510263929619, "grad_norm": 5.635897679272559, "kl": 0.875, "learning_rate": 8.861825578770069e-07, "loss": 0.0649, "reward": 2.3644415140151978, "reward_std": 0.2842712849378586, "rewards/accuracy_reward": 0.4166666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.012294710613787174, "rewards/tag_count_reward": 0.9739583432674408, "step": 421 }, { "clip_ratio": 0.0, "completion_length": 422.37501525878906, "epoch": 0.3093841642228739, "grad_norm": 3.9672336398242374, "kl": 0.5673828125, "learning_rate": 8.854155579204036e-07, "loss": 0.0106, "reward": 2.935629725456238, "reward_std": 0.12748303078114986, "rewards/accuracy_reward": 0.9791666865348816, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.017495382577180862, "rewards/tag_count_reward": 0.9947916865348816, "step": 422 }, { "clip_ratio": 0.0, "completion_length": 449.50001525878906, "epoch": 0.3101173020527859, "grad_norm": 4.249069913225125, "kl": 0.5078125, "learning_rate": 8.84646359124018e-07, "loss": 0.0199, "reward": 2.429058015346527, "reward_std": 0.17704454949125648, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.015386473387479782, "rewards/tag_count_reward": 1.0, "step": 423 }, { "clip_ratio": 0.0, "completion_length": 444.6458435058594, "epoch": 0.31085043988269795, "grad_norm": 11.718746465515885, "kl": 0.59375, "learning_rate": 8.83874966530386e-07, "loss": -0.0174, "reward": 2.216548442840576, "reward_std": 0.11675330437719822, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017826574854552746, "rewards/tag_count_reward": 0.984375, "step": 424 }, { "clip_ratio": 0.0, "completion_length": 428.7291717529297, "epoch": 0.31158357771260997, "grad_norm": 3.99754517757241, "kl": 0.3359375, "learning_rate": 8.831013851964253e-07, "loss": -0.0056, "reward": 2.6643564701080322, "reward_std": 0.20628847926855087, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.01967142429202795, "rewards/tag_count_reward": 0.9895833432674408, "step": 425 }, { "clip_ratio": 0.0, "completion_length": 407.68751525878906, "epoch": 0.312316715542522, "grad_norm": 3.0243285560646918, "kl": 0.3134765625, "learning_rate": 8.823256201934009e-07, "loss": -0.0137, "reward": 2.472022771835327, "reward_std": 0.03336298279464245, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.021032861433923244, "rewards/tag_count_reward": 1.0, "step": 426 }, { "clip_ratio": 0.0, "completion_length": 468.1666717529297, "epoch": 0.313049853372434, "grad_norm": 6.739707368856835, "kl": 0.541015625, "learning_rate": 8.81547676606894e-07, "loss": 0.0628, "reward": 2.461642861366272, "reward_std": 0.08788503333926201, "rewards/accuracy_reward": 0.5208333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.01926009263843298, "rewards/tag_count_reward": 0.9739583432674408, "step": 427 }, { "clip_ratio": 0.0, "completion_length": 425.2708435058594, "epoch": 0.31378299120234604, "grad_norm": 6.748145314971122, "kl": 0.421875, "learning_rate": 8.807675595367674e-07, "loss": 0.1198, "reward": 2.1878883838653564, "reward_std": 0.13607415929436684, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025653420016169548, "rewards/tag_count_reward": 0.984375, "step": 428 }, { "clip_ratio": 0.0, "completion_length": 430.31251525878906, "epoch": 0.31451612903225806, "grad_norm": 7.1793774023164145, "kl": 0.5166015625, "learning_rate": 8.799852740971323e-07, "loss": 0.0118, "reward": 2.4083354473114014, "reward_std": 0.20769641920924187, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9722222089767456, "rewards/repetition_penalty_reward": -0.027428350411355495, "rewards/tag_count_reward": 0.984375, "step": 429 }, { "clip_ratio": 0.0, "completion_length": 415.6458435058594, "epoch": 0.3152492668621701, "grad_norm": 4.178510346631957, "kl": 0.353515625, "learning_rate": 8.792008254163145e-07, "loss": 0.0192, "reward": 2.216019093990326, "reward_std": 0.05499478569254279, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02009205799549818, "rewards/tag_count_reward": 1.0, "step": 430 }, { "clip_ratio": 0.0, "completion_length": 547.1875152587891, "epoch": 0.3159824046920821, "grad_norm": 12.228179652584759, "kl": 1.5625, "learning_rate": 8.784142186368214e-07, "loss": 0.2563, "reward": 2.4159883856773376, "reward_std": 0.11997727863490582, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026720006950199604, "rewards/tag_count_reward": 0.9427083730697632, "step": 431 }, { "clip_ratio": 0.0, "completion_length": 506.875, "epoch": 0.31671554252199413, "grad_norm": 8.456013908011222, "kl": 1.7109375, "learning_rate": 8.776254589153081e-07, "loss": 0.2094, "reward": 2.1472460627555847, "reward_std": 0.1969931460916996, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.012476155068725348, "rewards/tag_count_reward": 0.9583333730697632, "step": 432 }, { "clip_ratio": 0.0, "completion_length": 560.7083587646484, "epoch": 0.31744868035190615, "grad_norm": 20.44702979218028, "kl": 2.8828125, "learning_rate": 8.768345514225431e-07, "loss": 0.3417, "reward": 2.5165975093841553, "reward_std": 0.4180612862110138, "rewards/accuracy_reward": 0.645833358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019861001521348953, "rewards/tag_count_reward": 0.8906250298023224, "step": 433 }, { "clip_ratio": 0.0, "completion_length": 627.4375305175781, "epoch": 0.3181818181818182, "grad_norm": 13.430874608120844, "kl": 4.34375, "learning_rate": 8.76041501343375e-07, "loss": 0.7845, "reward": 2.2989736795425415, "reward_std": 0.6288765966892242, "rewards/accuracy_reward": 0.5208333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.022207094356417656, "rewards/tag_count_reward": 0.8072916865348816, "step": 434 }, { "clip_ratio": 0.0, "completion_length": 590.9791870117188, "epoch": 0.3189149560117302, "grad_norm": 29.560773552065303, "kl": 4.59375, "learning_rate": 8.752463138766981e-07, "loss": 0.5636, "reward": 2.0621426105499268, "reward_std": 0.28306297212839127, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.01771850883960724, "rewards/tag_count_reward": 0.8645833432674408, "step": 435 }, { "clip_ratio": 0.0, "completion_length": 619.7916870117188, "epoch": 0.3196480938416422, "grad_norm": 22.441957468784093, "kl": 3.7109375, "learning_rate": 8.744489942354187e-07, "loss": 0.675, "reward": 2.1259257793426514, "reward_std": 0.3373235762119293, "rewards/accuracy_reward": 0.25000000558793545, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019907758571207523, "rewards/tag_count_reward": 0.8958333432674408, "step": 436 }, { "clip_ratio": 0.0, "completion_length": 685.5625305175781, "epoch": 0.32038123167155425, "grad_norm": 12.825077973059518, "kl": 4.0, "learning_rate": 8.736495476464205e-07, "loss": 0.68, "reward": 2.335544228553772, "reward_std": 0.2173374593257904, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02383077424019575, "rewards/tag_count_reward": 0.8802083432674408, "step": 437 }, { "clip_ratio": 0.0, "completion_length": 541.1666717529297, "epoch": 0.32111436950146627, "grad_norm": 20.43850556963088, "kl": 2.25, "learning_rate": 8.728479793505308e-07, "loss": 0.4834, "reward": 1.9290331602096558, "reward_std": 0.1057676412165165, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024091887287795544, "rewards/tag_count_reward": 0.9531250298023224, "step": 438 }, { "clip_ratio": 0.0, "completion_length": 578.0416870117188, "epoch": 0.3218475073313783, "grad_norm": 16.888244534119845, "kl": 2.65625, "learning_rate": 8.720442946024855e-07, "loss": 0.3783, "reward": 2.275824010372162, "reward_std": 0.40398430824279785, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.024523337371647358, "rewards/tag_count_reward": 0.9114583730697632, "step": 439 }, { "clip_ratio": 0.0, "completion_length": 530.0416717529297, "epoch": 0.3225806451612903, "grad_norm": 26.375467728384585, "kl": 1.7265625, "learning_rate": 8.712384986708953e-07, "loss": 0.4054, "reward": 2.3330430388450623, "reward_std": 0.3513893634080887, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.024595932103693485, "rewards/tag_count_reward": 0.9270833432674408, "step": 440 }, { "clip_ratio": 0.0, "completion_length": 566.2083435058594, "epoch": 0.32331378299120234, "grad_norm": 24.099276246580605, "kl": 2.86328125, "learning_rate": 8.704305968382112e-07, "loss": 0.7114, "reward": 2.1199684143066406, "reward_std": 0.2733127325773239, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.020656662993133068, "rewards/tag_count_reward": 0.9322916865348816, "step": 441 }, { "clip_ratio": 0.0, "completion_length": 596.9166870117188, "epoch": 0.32404692082111436, "grad_norm": 24.107301778659405, "kl": 4.1875, "learning_rate": 8.696205944006891e-07, "loss": 0.7505, "reward": 2.107983946800232, "reward_std": 0.268925741314888, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017016036435961723, "rewards/tag_count_reward": 0.8958333432674408, "step": 442 }, { "clip_ratio": 0.0, "completion_length": 613.2916870117188, "epoch": 0.3247800586510264, "grad_norm": 49.39061448190325, "kl": 6.421875, "learning_rate": 8.688084966683557e-07, "loss": 0.5934, "reward": 2.1915996074676514, "reward_std": 0.4412628412246704, "rewards/accuracy_reward": 0.3541666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.016733783297240734, "rewards/tag_count_reward": 0.8541666865348816, "step": 443 }, { "clip_ratio": 0.0, "completion_length": 836.8750305175781, "epoch": 0.3255131964809384, "grad_norm": 77.73635730936726, "kl": 11.03125, "learning_rate": 8.679943089649737e-07, "loss": 0.951, "reward": 1.8464696407318115, "reward_std": 0.46440982818603516, "rewards/accuracy_reward": 0.12500000558793545, "rewards/reasoning_steps_reward": 0.9791667461395264, "rewards/repetition_penalty_reward": -0.018113664351403713, "rewards/tag_count_reward": 0.7604166865348816, "step": 444 }, { "clip_ratio": 0.0, "completion_length": 684.8125305175781, "epoch": 0.3262463343108504, "grad_norm": 43.20397648755034, "kl": 6.484375, "learning_rate": 8.671780366280067e-07, "loss": 0.9399, "reward": 2.0885465145111084, "reward_std": 0.508235827088356, "rewards/accuracy_reward": 0.2916666865348816, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.02082844078540802, "rewards/tag_count_reward": 0.8385416865348816, "step": 445 }, { "clip_ratio": 0.0, "completion_length": 457.4583435058594, "epoch": 0.32697947214076245, "grad_norm": 13.78943058368852, "kl": 2.203125, "learning_rate": 8.66359685008584e-07, "loss": 0.2261, "reward": 2.1724031567573547, "reward_std": 0.14021247252821922, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.01856919191777706, "rewards/tag_count_reward": 0.9479166865348816, "step": 446 }, { "clip_ratio": 0.0, "completion_length": 464.9375, "epoch": 0.3277126099706745, "grad_norm": 18.641202440944074, "kl": 1.48046875, "learning_rate": 8.655392594714661e-07, "loss": 0.3041, "reward": 2.3693387508392334, "reward_std": 0.2319539673626423, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.016078063752502203, "rewards/tag_count_reward": 0.9479166865348816, "step": 447 }, { "clip_ratio": 0.0, "completion_length": 511.50001525878906, "epoch": 0.3284457478005865, "grad_norm": 10.100849867977697, "kl": 1.19140625, "learning_rate": 8.647167653950092e-07, "loss": 0.2502, "reward": 1.9037845134735107, "reward_std": 0.1535169705748558, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.01982666738331318, "rewards/tag_count_reward": 0.9375, "step": 448 }, { "clip_ratio": 0.0, "completion_length": 415.3958435058594, "epoch": 0.3291788856304985, "grad_norm": 6.39755512266102, "kl": 0.931640625, "learning_rate": 8.638922081711295e-07, "loss": 0.0919, "reward": 2.8837406635284424, "reward_std": 0.193980960175395, "rewards/accuracy_reward": 0.9375, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.015564900357276201, "rewards/tag_count_reward": 0.96875, "step": 449 }, { "clip_ratio": 0.0, "completion_length": 504.8958435058594, "epoch": 0.32991202346041054, "grad_norm": 11.85855042798546, "kl": 1.451171875, "learning_rate": 8.63065593205269e-07, "loss": 0.366, "reward": 2.794464588165283, "reward_std": 0.44993289560079575, "rewards/accuracy_reward": 0.8958333730697632, "rewards/reasoning_steps_reward": 0.9791666567325592, "rewards/repetition_penalty_reward": -0.01803573127835989, "rewards/tag_count_reward": 0.9375, "step": 450 }, { "clip_ratio": 0.0, "completion_length": 443.0833435058594, "epoch": 0.33064516129032256, "grad_norm": 18.349554724503466, "kl": 1.203125, "learning_rate": 8.622369259163588e-07, "loss": 0.3583, "reward": 2.379144310951233, "reward_std": 0.22539783269166946, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.011480757035315037, "rewards/tag_count_reward": 0.953125, "step": 451 }, { "clip_ratio": 0.0, "completion_length": 543.8125305175781, "epoch": 0.3313782991202346, "grad_norm": 42.831501126634414, "kl": 1.2265625, "learning_rate": 8.614062117367846e-07, "loss": 0.6447, "reward": 2.5604196786880493, "reward_std": 0.5051280111074448, "rewards/accuracy_reward": 0.6666666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.014233164954930544, "rewards/tag_count_reward": 0.9218750298023224, "step": 452 }, { "clip_ratio": 0.0, "completion_length": 563.6458435058594, "epoch": 0.3321114369501466, "grad_norm": 8.498854597259745, "kl": 1.859375, "learning_rate": 8.6057345611235e-07, "loss": 0.3801, "reward": 2.386213481426239, "reward_std": 0.2767922133207321, "rewards/accuracy_reward": 0.5000000204890966, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.0096200630068779, "rewards/tag_count_reward": 0.9166666865348816, "step": 453 }, { "clip_ratio": 0.0, "completion_length": 467.68751525878906, "epoch": 0.33284457478005863, "grad_norm": 10.655046074477987, "kl": 1.2392578125, "learning_rate": 8.597386645022423e-07, "loss": 0.2894, "reward": 2.347404420375824, "reward_std": 0.3219123715534806, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.01023465720936656, "rewards/tag_count_reward": 0.9479166865348816, "step": 454 }, { "clip_ratio": 0.0, "completion_length": 387.2083435058594, "epoch": 0.33357771260997066, "grad_norm": 13.16671124138141, "kl": 1.1796875, "learning_rate": 8.589018423789951e-07, "loss": 0.2794, "reward": 2.66903293132782, "reward_std": 0.16836241632699966, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.014994919300079346, "rewards/tag_count_reward": 0.96875, "step": 455 }, { "clip_ratio": 0.0, "completion_length": 594.8958435058594, "epoch": 0.3343108504398827, "grad_norm": 13.161941875430305, "kl": 3.9375, "learning_rate": 8.580629952284534e-07, "loss": 1.3136, "reward": 2.3228167295455933, "reward_std": 0.4149315506219864, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.010516760405153036, "rewards/tag_count_reward": 0.8958333432674408, "step": 456 }, { "clip_ratio": 0.0, "completion_length": 530.5208435058594, "epoch": 0.3350439882697947, "grad_norm": 26.618723827545875, "kl": 4.53125, "learning_rate": 8.572221285497378e-07, "loss": 0.8534, "reward": 1.884135127067566, "reward_std": 0.19925276935100555, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.01517043774947524, "rewards/tag_count_reward": 0.90625, "step": 457 }, { "clip_ratio": 0.0, "completion_length": 550.4166870117188, "epoch": 0.3357771260997067, "grad_norm": 28.703091822389183, "kl": 3.875, "learning_rate": 8.563792478552071e-07, "loss": 0.7495, "reward": 2.278742492198944, "reward_std": 0.38903089612722397, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.012924263253808022, "rewards/tag_count_reward": 0.9166666865348816, "step": 458 }, { "clip_ratio": 0.0, "completion_length": 611.0416870117188, "epoch": 0.33651026392961875, "grad_norm": 27.90049267018379, "kl": 4.4296875, "learning_rate": 8.555343586704242e-07, "loss": 0.8235, "reward": 2.1318034529685974, "reward_std": 0.558455765247345, "rewards/accuracy_reward": 0.3125, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.012293811421841383, "rewards/tag_count_reward": 0.859375, "step": 459 }, { "clip_ratio": 0.0, "completion_length": 533.5416717529297, "epoch": 0.33724340175953077, "grad_norm": 22.430193881631553, "kl": 3.5234375, "learning_rate": 8.546874665341181e-07, "loss": 0.4792, "reward": 2.256991147994995, "reward_std": 0.4296814352273941, "rewards/accuracy_reward": 0.4166666716337204, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.012106035370379686, "rewards/tag_count_reward": 0.8802083432674408, "step": 460 }, { "clip_ratio": 0.0, "completion_length": 562.0416870117188, "epoch": 0.3379765395894428, "grad_norm": 20.303533243217544, "kl": 2.9609375, "learning_rate": 8.538385769981488e-07, "loss": 0.4452, "reward": 2.025463283061981, "reward_std": 0.41344672441482544, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.012731055729091167, "rewards/tag_count_reward": 0.8645833730697632, "step": 461 }, { "clip_ratio": 0.0, "completion_length": 611.2500152587891, "epoch": 0.3387096774193548, "grad_norm": 11.57285794591874, "kl": 2.9765625, "learning_rate": 8.529876956274698e-07, "loss": 0.6475, "reward": 2.311686098575592, "reward_std": 0.29963135719299316, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.018175031058490276, "rewards/tag_count_reward": 0.8645833730697632, "step": 462 }, { "clip_ratio": 0.0, "completion_length": 452.56251525878906, "epoch": 0.33944281524926684, "grad_norm": 6.220846068922538, "kl": 1.05859375, "learning_rate": 8.521348280000924e-07, "loss": 0.2887, "reward": 2.1803677678108215, "reward_std": 0.08419075515121222, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017548908479511738, "rewards/tag_count_reward": 0.9479166865348816, "step": 463 }, { "clip_ratio": 0.0, "completion_length": 493.85418701171875, "epoch": 0.34017595307917886, "grad_norm": 24.101035227375732, "kl": 1.578125, "learning_rate": 8.512799797070492e-07, "loss": 0.6789, "reward": 2.3085252046585083, "reward_std": 0.4135167822241783, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9791666567325592, "rewards/repetition_penalty_reward": -0.019599976018071175, "rewards/tag_count_reward": 0.9114583432674408, "step": 464 }, { "clip_ratio": 0.0, "completion_length": 468.1666717529297, "epoch": 0.3409090909090909, "grad_norm": 23.665461110852053, "kl": 1.0625, "learning_rate": 8.504231563523573e-07, "loss": 0.5876, "reward": 2.1471781730651855, "reward_std": 0.3046387583017349, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.014280236326158047, "rewards/tag_count_reward": 0.9531250298023224, "step": 465 }, { "clip_ratio": 0.0, "completion_length": 466.5000305175781, "epoch": 0.3416422287390029, "grad_norm": 8.232517144541893, "kl": 1.29296875, "learning_rate": 8.495643635529809e-07, "loss": 0.2298, "reward": 2.6147353649139404, "reward_std": 0.2032028939574957, "rewards/accuracy_reward": 0.6875, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.015473092906177044, "rewards/tag_count_reward": 0.9427083432674408, "step": 466 }, { "clip_ratio": 0.0, "completion_length": 348.4166717529297, "epoch": 0.34237536656891493, "grad_norm": 10.338324346543883, "kl": 0.7734375, "learning_rate": 8.487036069387953e-07, "loss": 0.0658, "reward": 2.450138568878174, "reward_std": 0.1330912890844047, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013403074350208044, "rewards/tag_count_reward": 0.984375, "step": 467 }, { "clip_ratio": 0.0, "completion_length": 430.16668701171875, "epoch": 0.34310850439882695, "grad_norm": 15.816345871806671, "kl": 0.84765625, "learning_rate": 8.4784089215255e-07, "loss": 0.1334, "reward": 2.4448307752609253, "reward_std": 0.14922725409269333, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013502615503966808, "rewards/tag_count_reward": 0.9791666865348816, "step": 468 }, { "clip_ratio": 0.0, "completion_length": 399.9791717529297, "epoch": 0.34384164222873903, "grad_norm": 25.38232544197375, "kl": 0.8203125, "learning_rate": 8.469762248498307e-07, "loss": 0.1797, "reward": 1.9597020745277405, "reward_std": 0.06006414443254471, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0194645868614316, "rewards/tag_count_reward": 0.9791666865348816, "step": 469 }, { "clip_ratio": 0.0, "completion_length": 456.68751525878906, "epoch": 0.34457478005865105, "grad_norm": 11.384146881799431, "kl": 1.71875, "learning_rate": 8.461096106990241e-07, "loss": 0.4859, "reward": 2.122660994529724, "reward_std": 0.3244031220674515, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.01622793171554804, "rewards/tag_count_reward": 0.9375000298023224, "step": 470 }, { "clip_ratio": 0.0, "completion_length": 427.5208435058594, "epoch": 0.3453079178885631, "grad_norm": 7.64042070760251, "kl": 1.35546875, "learning_rate": 8.452410553812784e-07, "loss": 0.334, "reward": 2.4085933566093445, "reward_std": 0.2034970000386238, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.015017746016383171, "rewards/tag_count_reward": 0.9583333730697632, "step": 471 }, { "clip_ratio": 0.0, "completion_length": 599.1666870117188, "epoch": 0.3460410557184751, "grad_norm": 12.197273976275588, "kl": 2.9609375, "learning_rate": 8.443705645904678e-07, "loss": 0.7164, "reward": 2.0220767855644226, "reward_std": 0.40696029365062714, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.010909311473369598, "rewards/tag_count_reward": 0.8802083432674408, "step": 472 }, { "clip_ratio": 0.0, "completion_length": 459.1458435058594, "epoch": 0.3467741935483871, "grad_norm": 26.782940873029442, "kl": 2.94140625, "learning_rate": 8.434981440331549e-07, "loss": 0.4876, "reward": 1.8921802043914795, "reward_std": 0.2121252343058586, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9791666567325592, "rewards/repetition_penalty_reward": -0.014069781638681889, "rewards/tag_count_reward": 0.9270833432674408, "step": 473 }, { "clip_ratio": 0.0, "completion_length": 535.5, "epoch": 0.34750733137829914, "grad_norm": 18.31310978578589, "kl": 2.7265625, "learning_rate": 8.426237994285525e-07, "loss": 0.7942, "reward": 2.537790536880493, "reward_std": 0.4235878735780716, "rewards/accuracy_reward": 0.6458333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.014292994979768991, "rewards/tag_count_reward": 0.90625, "step": 474 }, { "clip_ratio": 0.0, "completion_length": 572.1458435058594, "epoch": 0.34824046920821117, "grad_norm": 11.85507041751325, "kl": 2.875, "learning_rate": 8.417475365084871e-07, "loss": 0.8618, "reward": 2.508628726005554, "reward_std": 0.5431976318359375, "rewards/accuracy_reward": 0.6458333432674408, "rewards/reasoning_steps_reward": 0.9722222089767456, "rewards/repetition_penalty_reward": -0.015676844865083694, "rewards/tag_count_reward": 0.9062500298023224, "step": 475 }, { "clip_ratio": 0.0, "completion_length": 610.3125305175781, "epoch": 0.3489736070381232, "grad_norm": 28.639038429417123, "kl": 3.234375, "learning_rate": 8.408693610173603e-07, "loss": 0.5474, "reward": 2.3011252880096436, "reward_std": 0.40642261505126953, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9583333730697632, "rewards/repetition_penalty_reward": -0.011374846566468477, "rewards/tag_count_reward": 0.875, "step": 476 }, { "clip_ratio": 0.0, "completion_length": 408.4166717529297, "epoch": 0.3497067448680352, "grad_norm": 18.88985204699429, "kl": 1.48046875, "learning_rate": 8.39989278712112e-07, "loss": 0.5386, "reward": 2.6290500164031982, "reward_std": 0.32656364142894745, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.015047314576804638, "rewards/tag_count_reward": 0.9635416865348816, "step": 477 }, { "clip_ratio": 0.0, "completion_length": 578.3333587646484, "epoch": 0.35043988269794724, "grad_norm": 11.451206533941894, "kl": 2.65234375, "learning_rate": 8.391072953621826e-07, "loss": 0.5397, "reward": 2.279217839241028, "reward_std": 0.34354250878095627, "rewards/accuracy_reward": 0.4166666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.014185154810547829, "rewards/tag_count_reward": 0.890625, "step": 478 }, { "clip_ratio": 0.0, "completion_length": 410.6458435058594, "epoch": 0.35117302052785926, "grad_norm": 7.27832965984057, "kl": 1.1796875, "learning_rate": 8.382234167494747e-07, "loss": 0.3302, "reward": 2.147601544857025, "reward_std": 0.2115662805736065, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.01732906699180603, "rewards/tag_count_reward": 0.9635416865348816, "step": 479 }, { "clip_ratio": 0.0, "completion_length": 371.93751525878906, "epoch": 0.3519061583577713, "grad_norm": 10.45608432872573, "kl": 1.12109375, "learning_rate": 8.373376486683149e-07, "loss": 0.2231, "reward": 2.1965653896331787, "reward_std": 0.09365869220346212, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022184578701853752, "rewards/tag_count_reward": 0.96875, "step": 480 }, { "clip_ratio": 0.0, "completion_length": 357.7083435058594, "epoch": 0.3526392961876833, "grad_norm": 9.253957750096733, "kl": 0.95703125, "learning_rate": 8.364499969254174e-07, "loss": 0.1801, "reward": 2.3384850025177, "reward_std": 0.3187159597873688, "rewards/accuracy_reward": 0.4166666716337204, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.022626149468123913, "rewards/tag_count_reward": 0.9583333432674408, "step": 481 }, { "clip_ratio": 0.0, "completion_length": 353.00001525878906, "epoch": 0.3533724340175953, "grad_norm": 7.045055176755826, "kl": 1.1748046875, "learning_rate": 8.35560467339844e-07, "loss": 0.1558, "reward": 2.6686532497406006, "reward_std": 0.20646232273429632, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.017110710497945547, "rewards/tag_count_reward": 0.9635416865348816, "step": 482 }, { "clip_ratio": 0.0, "completion_length": 308.91668701171875, "epoch": 0.35410557184750735, "grad_norm": 4.762432815291462, "kl": 0.3720703125, "learning_rate": 8.346690657429672e-07, "loss": 0.0247, "reward": 2.2265055179595947, "reward_std": 0.031516775488853455, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.016550065483897924, "rewards/tag_count_reward": 1.0, "step": 483 }, { "clip_ratio": 0.0, "completion_length": 333.00001525878906, "epoch": 0.3548387096774194, "grad_norm": 8.645951391731817, "kl": 0.46484375, "learning_rate": 8.337757979784315e-07, "loss": 0.0418, "reward": 2.2375470995903015, "reward_std": 0.0075986802112311125, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.012452858500182629, "rewards/tag_count_reward": 1.0, "step": 484 }, { "clip_ratio": 0.0, "completion_length": 397.7083435058594, "epoch": 0.3555718475073314, "grad_norm": 7.441907733438074, "kl": 0.646484375, "learning_rate": 8.328806699021155e-07, "loss": 0.203, "reward": 2.6804587841033936, "reward_std": 0.17669676430523396, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.020930366590619087, "rewards/tag_count_reward": 0.9791666865348816, "step": 485 }, { "clip_ratio": 0.0, "completion_length": 374.25001525878906, "epoch": 0.3563049853372434, "grad_norm": 6.326596783751009, "kl": 0.44140625, "learning_rate": 8.319836873820926e-07, "loss": 0.0394, "reward": 2.6759073734283447, "reward_std": 0.12399306986480951, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.013328731525689363, "rewards/tag_count_reward": 0.953125, "step": 486 }, { "clip_ratio": 0.0, "completion_length": 411.9166717529297, "epoch": 0.35703812316715544, "grad_norm": 17.88113563714185, "kl": 0.6357421875, "learning_rate": 8.310848562985935e-07, "loss": 0.0157, "reward": 2.671896457672119, "reward_std": 0.22538332641124725, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.013867528643459082, "rewards/tag_count_reward": 0.9843750298023224, "step": 487 }, { "clip_ratio": 0.0, "completion_length": 419.56251525878906, "epoch": 0.35777126099706746, "grad_norm": 8.614526501689246, "kl": 0.9169921875, "learning_rate": 8.301841825439674e-07, "loss": 0.328, "reward": 2.169337749481201, "reward_std": 0.2654203623533249, "rewards/accuracy_reward": 0.22916667722165585, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.021634459495544434, "rewards/tag_count_reward": 0.96875, "step": 488 }, { "clip_ratio": 0.0, "completion_length": 434.7708435058594, "epoch": 0.3585043988269795, "grad_norm": 9.126566406167589, "kl": 0.71875, "learning_rate": 8.292816720226429e-07, "loss": 0.3833, "reward": 2.3876397609710693, "reward_std": 0.2834826409816742, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013401838950812817, "rewards/tag_count_reward": 0.9635416865348816, "step": 489 }, { "clip_ratio": 0.0, "completion_length": 384.4583435058594, "epoch": 0.3592375366568915, "grad_norm": 4.9502265845193145, "kl": 0.5224609375, "learning_rate": 8.283773306510903e-07, "loss": 0.0681, "reward": 2.374760627746582, "reward_std": 0.37007441371679306, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01586443232372403, "rewards/tag_count_reward": 0.9947916865348816, "step": 490 }, { "clip_ratio": 0.0, "completion_length": 419.9791717529297, "epoch": 0.35997067448680353, "grad_norm": 6.705292937977975, "kl": 0.7734375, "learning_rate": 8.274711643577812e-07, "loss": 0.1881, "reward": 2.1267879605293274, "reward_std": 0.17756187077611685, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01904546469449997, "rewards/tag_count_reward": 0.9791666865348816, "step": 491 }, { "clip_ratio": 0.0, "completion_length": 436.37501525878906, "epoch": 0.36070381231671556, "grad_norm": 6.170093231897857, "kl": 0.794921875, "learning_rate": 8.265631790831513e-07, "loss": 0.2968, "reward": 1.9459967613220215, "reward_std": 0.09664355963468552, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0227531548589468, "rewards/tag_count_reward": 0.96875, "step": 492 }, { "clip_ratio": 0.0, "completion_length": 454.5833435058594, "epoch": 0.3614369501466276, "grad_norm": 11.129735561164555, "kl": 1.158203125, "learning_rate": 8.256533807795607e-07, "loss": 0.366, "reward": 1.9230691194534302, "reward_std": 0.2331734150648117, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.026583708822727203, "rewards/tag_count_reward": 0.9427083432674408, "step": 493 }, { "clip_ratio": 0.0, "completion_length": 491.64585876464844, "epoch": 0.3621700879765396, "grad_norm": 11.115235107040252, "kl": 1.203125, "learning_rate": 8.247417754112548e-07, "loss": 0.5382, "reward": 1.9273149967193604, "reward_std": 0.23959805816411972, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.02233781013637781, "rewards/tag_count_reward": 0.9427083432674408, "step": 494 }, { "clip_ratio": 0.0, "completion_length": 509.43751525878906, "epoch": 0.3629032258064516, "grad_norm": 15.68299311864471, "kl": 1.400390625, "learning_rate": 8.238283689543252e-07, "loss": 0.4682, "reward": 1.9225846529006958, "reward_std": 0.3457977622747421, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.018387647345662117, "rewards/tag_count_reward": 0.9270833432674408, "step": 495 }, { "clip_ratio": 0.0, "completion_length": 547.4166870117188, "epoch": 0.36363636363636365, "grad_norm": 6.773746675176151, "kl": 1.05078125, "learning_rate": 8.229131673966708e-07, "loss": 0.3773, "reward": 1.9133232831954956, "reward_std": 0.26244693249464035, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.018968377262353897, "rewards/tag_count_reward": 0.9322916865348816, "step": 496 }, { "clip_ratio": 0.0, "completion_length": 502.04168701171875, "epoch": 0.36436950146627567, "grad_norm": 8.290342632836095, "kl": 0.978515625, "learning_rate": 8.219961767379586e-07, "loss": 0.4217, "reward": 2.1250281929969788, "reward_std": 0.25319022685289383, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.024277442134916782, "rewards/tag_count_reward": 0.9479166865348816, "step": 497 }, { "clip_ratio": 0.0, "completion_length": 452.8958435058594, "epoch": 0.3651026392961877, "grad_norm": 10.77527531050857, "kl": 0.8828125, "learning_rate": 8.21077402989584e-07, "loss": 0.3326, "reward": 2.1549705266952515, "reward_std": 0.19552180310711265, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.025585101917386055, "rewards/tag_count_reward": 0.9791666865348816, "step": 498 }, { "clip_ratio": 0.0, "completion_length": 551.8125152587891, "epoch": 0.3658357771260997, "grad_norm": 63.70158740638487, "kl": 2.2021484375, "learning_rate": 8.201568521746314e-07, "loss": 0.6263, "reward": 2.017093539237976, "reward_std": 0.38255129382014275, "rewards/accuracy_reward": 0.16666667722165585, "rewards/reasoning_steps_reward": 0.965277761220932, "rewards/repetition_penalty_reward": -0.021101072430610657, "rewards/tag_count_reward": 0.9062500298023224, "step": 499 }, { "clip_ratio": 0.0, "completion_length": 413.25001525878906, "epoch": 0.36656891495601174, "grad_norm": 14.422611186019756, "kl": 0.6953125, "learning_rate": 8.192345303278351e-07, "loss": 0.2335, "reward": 1.9486969113349915, "reward_std": 0.1088674496859312, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.021789253689348698, "rewards/tag_count_reward": 0.984375, "step": 500 }, { "clip_ratio": 0.0, "completion_length": 446.3333435058594, "epoch": 0.36730205278592376, "grad_norm": 43.60045387371852, "kl": 1.4921875, "learning_rate": 8.183104434955395e-07, "loss": 0.5019, "reward": 1.9558414220809937, "reward_std": 0.16421754658222198, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.026797562837600708, "rewards/tag_count_reward": 0.9687500298023224, "step": 501 }, { "clip_ratio": 0.0, "completion_length": 435.18751525878906, "epoch": 0.3680351906158358, "grad_norm": 139.1743325769162, "kl": 1.78515625, "learning_rate": 8.173845977356594e-07, "loss": 0.468, "reward": 1.9216685891151428, "reward_std": 0.14234358817338943, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.027984178625047207, "rewards/tag_count_reward": 0.9635416865348816, "step": 502 }, { "clip_ratio": 0.0, "completion_length": 430.7916717529297, "epoch": 0.3687683284457478, "grad_norm": 145.4472171807426, "kl": 3.671875, "learning_rate": 8.164569991176405e-07, "loss": 0.7199, "reward": 2.0667614936828613, "reward_std": 0.3162480816245079, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.023516294546425343, "rewards/tag_count_reward": 0.9375000298023224, "step": 503 }, { "clip_ratio": 0.0, "completion_length": 395.4791717529297, "epoch": 0.36950146627565983, "grad_norm": 25.89876003065372, "kl": 1.1796875, "learning_rate": 8.155276537224194e-07, "loss": 0.4086, "reward": 1.917527198791504, "reward_std": 0.22057230025529861, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.01997281052172184, "rewards/tag_count_reward": 0.9583333730697632, "step": 504 }, { "clip_ratio": 0.0, "completion_length": 472.83335876464844, "epoch": 0.37023460410557185, "grad_norm": 150.8827097979688, "kl": 5.7138671875, "learning_rate": 8.145965676423837e-07, "loss": 0.5975, "reward": 1.8190749883651733, "reward_std": 0.19131757970899343, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9305555820465088, "rewards/repetition_penalty_reward": -0.017730488441884518, "rewards/tag_count_reward": 0.90625, "step": 505 }, { "clip_ratio": 0.0, "completion_length": 443.9166717529297, "epoch": 0.3709677419354839, "grad_norm": 22.8478805496014, "kl": 1.41015625, "learning_rate": 8.136637469813322e-07, "loss": 0.2955, "reward": 1.9509990811347961, "reward_std": 0.09807348623871803, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.02122316136956215, "rewards/tag_count_reward": 0.9791666865348816, "step": 506 }, { "clip_ratio": 0.0, "completion_length": 468.70835876464844, "epoch": 0.3717008797653959, "grad_norm": 45.306018801969856, "kl": 1.9794921875, "learning_rate": 8.127291978544354e-07, "loss": 0.4052, "reward": 1.8993171453475952, "reward_std": 0.19246254302561283, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.020821738056838512, "rewards/tag_count_reward": 0.9479166865348816, "step": 507 }, { "clip_ratio": 0.0, "completion_length": 483.9375305175781, "epoch": 0.3724340175953079, "grad_norm": 40.87142241010334, "kl": 1.71875, "learning_rate": 8.117929263881942e-07, "loss": 0.577, "reward": 1.9061757326126099, "reward_std": 0.18592821806669235, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.022643746808171272, "rewards/tag_count_reward": 0.9635416865348816, "step": 508 }, { "clip_ratio": 0.0, "completion_length": 659.2916870117188, "epoch": 0.37316715542521994, "grad_norm": 115.61873644021891, "kl": 5.5703125, "learning_rate": 8.108549387204003e-07, "loss": 1.1436, "reward": 1.7973338961601257, "reward_std": 0.3085802122950554, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9513889253139496, "rewards/repetition_penalty_reward": -0.013430023565888405, "rewards/tag_count_reward": 0.8593750298023224, "step": 509 }, { "clip_ratio": 0.0, "completion_length": 567.9583435058594, "epoch": 0.37390029325513197, "grad_norm": 26.50908771676089, "kl": 2.62890625, "learning_rate": 8.09915241000097e-07, "loss": 0.4909, "reward": 1.8572303652763367, "reward_std": 0.30207861959934235, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.958333432674408, "rewards/repetition_penalty_reward": -0.017769566271454096, "rewards/tag_count_reward": 0.8958333432674408, "step": 510 }, { "clip_ratio": 0.0, "completion_length": 386.5208435058594, "epoch": 0.374633431085044, "grad_norm": 11.75236991564634, "kl": 0.5400390625, "learning_rate": 8.089738393875371e-07, "loss": 0.075, "reward": 1.9674814939498901, "reward_std": 0.0658029681071639, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.015157437417656183, "rewards/tag_count_reward": 0.9895833432674408, "step": 511 }, { "clip_ratio": 0.0, "completion_length": 474.83335876464844, "epoch": 0.375366568914956, "grad_norm": 42.108031734612005, "kl": 1.453125, "learning_rate": 8.080307400541438e-07, "loss": 0.3499, "reward": 1.920167326927185, "reward_std": 0.1508631743490696, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.015596664976328611, "rewards/tag_count_reward": 0.9427083432674408, "step": 512 }, { "clip_ratio": 0.0, "completion_length": 685.3958740234375, "epoch": 0.37609970674486803, "grad_norm": 34.90049805526321, "kl": 2.3828125, "learning_rate": 8.070859491824697e-07, "loss": 0.6664, "reward": 1.6676543951034546, "reward_std": 0.4811856746673584, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9097222685813904, "rewards/repetition_penalty_reward": -0.01810950506478548, "rewards/tag_count_reward": 0.7760416865348816, "step": 513 }, { "clip_ratio": 0.0, "completion_length": 657.5416870117188, "epoch": 0.37683284457478006, "grad_norm": 39.735639055218364, "kl": 4.6875, "learning_rate": 8.061394729661564e-07, "loss": 0.6612, "reward": 1.9260591268539429, "reward_std": 0.5082510709762573, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9305555820465088, "rewards/repetition_penalty_reward": -0.020121485460549593, "rewards/tag_count_reward": 0.8281250298023224, "step": 514 }, { "clip_ratio": 0.0, "completion_length": 730.8333435058594, "epoch": 0.3775659824046921, "grad_norm": 29.337370968228235, "kl": 4.078125, "learning_rate": 8.051913176098937e-07, "loss": 0.5808, "reward": 1.7654681205749512, "reward_std": 0.5887759923934937, "rewards/accuracy_reward": 0.1041666716337204, "rewards/reasoning_steps_reward": 0.9027777910232544, "rewards/repetition_penalty_reward": -0.012309623882174492, "rewards/tag_count_reward": 0.7708333432674408, "step": 515 }, { "clip_ratio": 0.0, "completion_length": 674.8541870117188, "epoch": 0.3782991202346041, "grad_norm": 56.98097498554146, "kl": 4.3203125, "learning_rate": 8.042414893293796e-07, "loss": 0.7913, "reward": 1.6792184114456177, "reward_std": 0.48759625852108, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9097221791744232, "rewards/repetition_penalty_reward": -0.011753852013498545, "rewards/tag_count_reward": 0.7812500298023224, "step": 516 }, { "clip_ratio": 0.0, "completion_length": 614.2708740234375, "epoch": 0.3790322580645161, "grad_norm": 64.88172032217534, "kl": 4.671875, "learning_rate": 8.032899943512785e-07, "loss": 0.8505, "reward": 1.762292504310608, "reward_std": 0.3404271602630615, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9305556118488312, "rewards/repetition_penalty_reward": -0.012013083323836327, "rewards/tag_count_reward": 0.84375, "step": 517 }, { "clip_ratio": 0.0, "completion_length": 538.2500305175781, "epoch": 0.37976539589442815, "grad_norm": 57.22402136535535, "kl": 5.5, "learning_rate": 8.023368389131815e-07, "loss": 0.7508, "reward": 1.769698143005371, "reward_std": 0.43798917531967163, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9305555522441864, "rewards/repetition_penalty_reward": -0.015024195425212383, "rewards/tag_count_reward": 0.8541666865348816, "step": 518 }, { "clip_ratio": 0.0, "completion_length": 615.625, "epoch": 0.38049853372434017, "grad_norm": 33.77773388568201, "kl": 5.6875, "learning_rate": 8.013820292635645e-07, "loss": 0.6833, "reward": 1.5902305245399475, "reward_std": 0.46116843819618225, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8611111044883728, "rewards/repetition_penalty_reward": -0.0156723465770483, "rewards/tag_count_reward": 0.7447916865348816, "step": 519 }, { "clip_ratio": 0.0, "completion_length": 590.2291870117188, "epoch": 0.3812316715542522, "grad_norm": 39.50621531257809, "kl": 4.1484375, "learning_rate": 8.00425571661748e-07, "loss": 0.5575, "reward": 1.657406985759735, "reward_std": 0.4130554646253586, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8888888955116272, "rewards/repetition_penalty_reward": -0.012731941416859627, "rewards/tag_count_reward": 0.7812500298023224, "step": 520 }, { "clip_ratio": 0.0, "completion_length": 395.4791717529297, "epoch": 0.3819648093841642, "grad_norm": 12.986688921654492, "kl": 0.744140625, "learning_rate": 7.994674723778559e-07, "loss": 0.1766, "reward": 1.9165648221969604, "reward_std": 0.13351602852344513, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.01572699472308159, "rewards/tag_count_reward": 0.9531250298023224, "step": 521 }, { "clip_ratio": 0.0, "completion_length": 533.0833435058594, "epoch": 0.38269794721407624, "grad_norm": 15.833753392603134, "kl": 2.90625, "learning_rate": 7.985077376927734e-07, "loss": 0.3802, "reward": 1.7569490671157837, "reward_std": 0.6436266899108887, "rewards/accuracy_reward": 0.14583333395421505, "rewards/reasoning_steps_reward": 0.8750000298023224, "rewards/repetition_penalty_reward": -0.01909259706735611, "rewards/tag_count_reward": 0.7552083432674408, "step": 522 }, { "clip_ratio": 0.0, "completion_length": 441.12501525878906, "epoch": 0.38343108504398826, "grad_norm": 28.762701865778784, "kl": 1.380859375, "learning_rate": 7.975463738981078e-07, "loss": 0.1971, "reward": 1.8199856877326965, "reward_std": 0.37226808071136475, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.944444477558136, "rewards/repetition_penalty_reward": -0.015083822421729565, "rewards/tag_count_reward": 0.890625, "step": 523 }, { "clip_ratio": 0.0, "completion_length": 466.1041717529297, "epoch": 0.3841642228739003, "grad_norm": 19.52603801272797, "kl": 1.326171875, "learning_rate": 7.965833872961455e-07, "loss": 0.2245, "reward": 2.1146620512008667, "reward_std": 0.39242707192897797, "rewards/accuracy_reward": 0.2500000149011612, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.01901856018230319, "rewards/tag_count_reward": 0.9114583432674408, "step": 524 }, { "clip_ratio": 0.0, "completion_length": 430.43751525878906, "epoch": 0.3848973607038123, "grad_norm": 55.45231535878981, "kl": 1.9140625, "learning_rate": 7.956187841998115e-07, "loss": 0.1649, "reward": 1.8645141124725342, "reward_std": 0.23954802006483078, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.015694267582148314, "rewards/tag_count_reward": 0.9010416865348816, "step": 525 }, { "clip_ratio": 0.0, "completion_length": 401.93751525878906, "epoch": 0.38563049853372433, "grad_norm": 377.2247094447684, "kl": 9.21875, "learning_rate": 7.946525709326278e-07, "loss": 0.8067, "reward": 2.0004115104675293, "reward_std": 0.5633389353752136, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9513889253139496, "rewards/repetition_penalty_reward": -0.01868572738021612, "rewards/tag_count_reward": 0.8385416865348816, "step": 526 }, { "clip_ratio": 0.0, "completion_length": 367.72918701171875, "epoch": 0.38636363636363635, "grad_norm": 161.61232633219788, "kl": 6.125, "learning_rate": 7.936847538286718e-07, "loss": 0.3533, "reward": 1.9359618425369263, "reward_std": 0.4424874410033226, "rewards/accuracy_reward": 0.125, "rewards/reasoning_steps_reward": 0.9236111044883728, "rewards/repetition_penalty_reward": -0.01889927126467228, "rewards/tag_count_reward": 0.9062500298023224, "step": 527 }, { "clip_ratio": 0.0, "completion_length": 423.3958435058594, "epoch": 0.3870967741935484, "grad_norm": 527.3727986310253, "kl": 22.8125, "learning_rate": 7.927153392325354e-07, "loss": 1.1885, "reward": 1.6565093398094177, "reward_std": 0.403203621506691, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 0.8472223281860352, "rewards/repetition_penalty_reward": -0.018837854266166687, "rewards/tag_count_reward": 0.7864583432674408, "step": 528 }, { "clip_ratio": 0.0, "completion_length": 456.6458435058594, "epoch": 0.3878299120234604, "grad_norm": 440.63322937171586, "kl": 15.8125, "learning_rate": 7.917443334992828e-07, "loss": 0.7378, "reward": 1.6151684522628784, "reward_std": 0.482510581612587, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8541666865348816, "rewards/repetition_penalty_reward": -0.015039919875562191, "rewards/tag_count_reward": 0.7760416865348816, "step": 529 }, { "clip_ratio": 0.0, "completion_length": 424.0416717529297, "epoch": 0.3885630498533724, "grad_norm": 26.175191456831758, "kl": 2.45703125, "learning_rate": 7.907717429944086e-07, "loss": 0.2639, "reward": 1.9304482340812683, "reward_std": 0.43061530590057373, "rewards/accuracy_reward": 0.1041666716337204, "rewards/reasoning_steps_reward": 0.951388955116272, "rewards/repetition_penalty_reward": -0.01573240989819169, "rewards/tag_count_reward": 0.8906250298023224, "step": 530 }, { "clip_ratio": 0.0, "completion_length": 392.54168701171875, "epoch": 0.38929618768328444, "grad_norm": 108.93805300951178, "kl": 3.6484375, "learning_rate": 7.897975740937971e-07, "loss": 0.2442, "reward": 1.883158266544342, "reward_std": 0.4635079950094223, "rewards/accuracy_reward": 0.1458333432674408, "rewards/reasoning_steps_reward": 0.9166667461395264, "rewards/repetition_penalty_reward": -0.017883423250168562, "rewards/tag_count_reward": 0.8385416865348816, "step": 531 }, { "clip_ratio": 0.0, "completion_length": 366.8958435058594, "epoch": 0.39002932551319647, "grad_norm": 16.48994140669, "kl": 2.4921875, "learning_rate": 7.888218331836796e-07, "loss": 0.1526, "reward": 1.784899353981018, "reward_std": 0.46506692469120026, "rewards/accuracy_reward": 0.0833333358168602, "rewards/reasoning_steps_reward": 0.8611111640930176, "rewards/repetition_penalty_reward": -0.0137117775157094, "rewards/tag_count_reward": 0.8541666865348816, "step": 532 }, { "clip_ratio": 0.0, "completion_length": 400.12501525878906, "epoch": 0.3907624633431085, "grad_norm": 33.79509036125726, "kl": 1.51171875, "learning_rate": 7.878445266605926e-07, "loss": 0.3311, "reward": 1.8051475286483765, "reward_std": 0.43154197931289673, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9305555522441864, "rewards/repetition_penalty_reward": -0.021241379901766777, "rewards/tag_count_reward": 0.8750000298023224, "step": 533 }, { "clip_ratio": 0.0, "completion_length": 420.43751525878906, "epoch": 0.3914956011730205, "grad_norm": 23.678075563625335, "kl": 1.59375, "learning_rate": 7.868656609313362e-07, "loss": 0.2481, "reward": 2.0650064945220947, "reward_std": 0.543351411819458, "rewards/accuracy_reward": 0.3125000149011612, "rewards/reasoning_steps_reward": 0.9236111342906952, "rewards/repetition_penalty_reward": -0.025271371938288212, "rewards/tag_count_reward": 0.8541666865348816, "step": 534 }, { "clip_ratio": 0.0, "completion_length": 392.37501525878906, "epoch": 0.39222873900293254, "grad_norm": 22.282050616576207, "kl": 2.3359375, "learning_rate": 7.858852424129324e-07, "loss": 0.2428, "reward": 1.957649052143097, "reward_std": 0.49290692806243896, "rewards/accuracy_reward": 0.16666667722165585, "rewards/reasoning_steps_reward": 0.9375, "rewards/repetition_penalty_reward": -0.016309399623423815, "rewards/tag_count_reward": 0.8697916865348816, "step": 535 }, { "clip_ratio": 0.0, "completion_length": 396.0208435058594, "epoch": 0.39296187683284456, "grad_norm": 30.90817986477352, "kl": 1.71875, "learning_rate": 7.849032775325824e-07, "loss": 0.2754, "reward": 1.8743115067481995, "reward_std": 0.2350020781159401, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.014577506110072136, "rewards/tag_count_reward": 0.9166666865348816, "step": 536 }, { "clip_ratio": 0.0, "completion_length": 319.04168701171875, "epoch": 0.3936950146627566, "grad_norm": 78.90804951173097, "kl": 1.923828125, "learning_rate": 7.839197727276241e-07, "loss": 0.0928, "reward": 1.9332107305526733, "reward_std": 0.16725022345781326, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.018178191035985947, "rewards/tag_count_reward": 0.9583333730697632, "step": 537 }, { "clip_ratio": 0.0, "completion_length": 372.4583435058594, "epoch": 0.3944281524926686, "grad_norm": 21.225171374291154, "kl": 1.98828125, "learning_rate": 7.829347344454912e-07, "loss": 0.2118, "reward": 1.8861700892448425, "reward_std": 0.2927362248301506, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.028760461136698723, "rewards/tag_count_reward": 0.9218750298023224, "step": 538 }, { "clip_ratio": 0.0, "completion_length": 377.5833435058594, "epoch": 0.3951612903225806, "grad_norm": 317.0519620942445, "kl": 8.40625, "learning_rate": 7.819481691436702e-07, "loss": 0.7884, "reward": 1.985667645931244, "reward_std": 0.473098486661911, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 0.9375000596046448, "rewards/repetition_penalty_reward": -0.02995745651423931, "rewards/tag_count_reward": 0.9114583432674408, "step": 539 }, { "clip_ratio": 0.0, "completion_length": 382.18751525878906, "epoch": 0.39589442815249265, "grad_norm": 267.21929696428475, "kl": 9.0625, "learning_rate": 7.809600832896575e-07, "loss": 0.8933, "reward": 1.7952839732170105, "reward_std": 0.3674754351377487, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9305556118488312, "rewards/repetition_penalty_reward": -0.025896546430885792, "rewards/tag_count_reward": 0.890625, "step": 540 }, { "clip_ratio": 0.0, "completion_length": 330.7708435058594, "epoch": 0.3966275659824047, "grad_norm": 76.5820325975885, "kl": 3.671875, "learning_rate": 7.799704833609181e-07, "loss": 0.178, "reward": 1.8716963529586792, "reward_std": 0.5188014209270477, "rewards/accuracy_reward": 0.1041666716337204, "rewards/reasoning_steps_reward": 0.9166666865348816, "rewards/repetition_penalty_reward": -0.018928625620901585, "rewards/tag_count_reward": 0.8697916865348816, "step": 541 }, { "clip_ratio": 0.0, "completion_length": 353.3958435058594, "epoch": 0.3973607038123167, "grad_norm": 16.599400277332023, "kl": 2.689453125, "learning_rate": 7.789793758448425e-07, "loss": 0.1534, "reward": 1.7878945469856262, "reward_std": 0.3570132479071617, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9375000298023224, "rewards/repetition_penalty_reward": -0.02460549771785736, "rewards/tag_count_reward": 0.8750000298023224, "step": 542 }, { "clip_ratio": 0.0, "completion_length": 345.3958435058594, "epoch": 0.3980938416422287, "grad_norm": 11.634824337183156, "kl": 1.5234375, "learning_rate": 7.779867672387041e-07, "loss": 0.0672, "reward": 1.8370883464813232, "reward_std": 0.2631266638636589, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.944444477558136, "rewards/repetition_penalty_reward": -0.029231124557554722, "rewards/tag_count_reward": 0.9218750298023224, "step": 543 }, { "clip_ratio": 0.0, "completion_length": 357.7083435058594, "epoch": 0.39882697947214074, "grad_norm": 19.639453122914002, "kl": 1.65234375, "learning_rate": 7.769926640496172e-07, "loss": 0.0726, "reward": 1.8252497911453247, "reward_std": 0.3410525470972061, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9375000298023224, "rewards/repetition_penalty_reward": -0.023708656430244446, "rewards/tag_count_reward": 0.9114583730697632, "step": 544 }, { "clip_ratio": 0.0, "completion_length": 361.79168701171875, "epoch": 0.39956011730205276, "grad_norm": 12.084490762032976, "kl": 2.21875, "learning_rate": 7.759970727944935e-07, "loss": 0.0599, "reward": 1.9157754182815552, "reward_std": 0.457296222448349, "rewards/accuracy_reward": 0.125, "rewards/reasoning_steps_reward": 0.9444445371627808, "rewards/repetition_penalty_reward": -0.023460770957171917, "rewards/tag_count_reward": 0.8697916865348816, "step": 545 }, { "clip_ratio": 0.0, "completion_length": 393.00001525878906, "epoch": 0.4002932551319648, "grad_norm": 6.558264956898931, "kl": 0.908203125, "learning_rate": 7.75e-07, "loss": 0.0845, "reward": 2.127650022506714, "reward_std": 0.2033278215676546, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.03207237180322409, "rewards/tag_count_reward": 0.9583333432674408, "step": 546 }, { "clip_ratio": 0.0, "completion_length": 342.4375, "epoch": 0.4010263929618768, "grad_norm": 19.898261789890924, "kl": 1.8046875, "learning_rate": 7.740014522025161e-07, "loss": -0.008, "reward": 1.6998693346977234, "reward_std": 0.4570258557796478, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.881944477558136, "rewards/repetition_penalty_reward": -0.02061682753264904, "rewards/tag_count_reward": 0.8385416865348816, "step": 547 }, { "clip_ratio": 0.0, "completion_length": 334.4583435058594, "epoch": 0.40175953079178883, "grad_norm": 18.537407189384908, "kl": 3.6171875, "learning_rate": 7.730014359480907e-07, "loss": 0.2439, "reward": 1.6735277771949768, "reward_std": 0.42917297780513763, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9027778506278992, "rewards/repetition_penalty_reward": -0.02091677486896515, "rewards/tag_count_reward": 0.7916666865348816, "step": 548 }, { "clip_ratio": 0.0, "completion_length": 352.25, "epoch": 0.40249266862170086, "grad_norm": 18.14928759631804, "kl": 2.76171875, "learning_rate": 7.719999577923992e-07, "loss": 0.2292, "reward": 1.7686306834220886, "reward_std": 0.3580906316637993, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9375, "rewards/repetition_penalty_reward": -0.028244351968169212, "rewards/tag_count_reward": 0.8593750298023224, "step": 549 }, { "clip_ratio": 0.0, "completion_length": 496.2083435058594, "epoch": 0.4032258064516129, "grad_norm": 78.70452703784652, "kl": 4.25390625, "learning_rate": 7.709970243007005e-07, "loss": 0.3906, "reward": 1.721606194972992, "reward_std": 0.36094823479652405, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.045755001716315746, "rewards/tag_count_reward": 0.8020833432674408, "step": 550 }, { "clip_ratio": 0.0, "completion_length": 363.7291717529297, "epoch": 0.4039589442815249, "grad_norm": 17.339896170616278, "kl": 3.609375, "learning_rate": 7.699926420477944e-07, "loss": 0.0842, "reward": 1.8914108872413635, "reward_std": 0.4989718496799469, "rewards/accuracy_reward": 0.1458333432674408, "rewards/reasoning_steps_reward": 0.9166666865348816, "rewards/repetition_penalty_reward": -0.02525589056313038, "rewards/tag_count_reward": 0.8541666865348816, "step": 551 }, { "clip_ratio": 0.0, "completion_length": 388.8333435058594, "epoch": 0.4046920821114369, "grad_norm": 92.32063302575182, "kl": 9.015625, "learning_rate": 7.689868176179775e-07, "loss": 0.3559, "reward": 1.6695754528045654, "reward_std": 0.4239191710948944, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8749999701976776, "rewards/repetition_penalty_reward": -0.03354958910495043, "rewards/tag_count_reward": 0.828125, "step": 552 }, { "clip_ratio": 0.0, "completion_length": 382.00001525878906, "epoch": 0.40542521994134895, "grad_norm": 23.4586762829187, "kl": 2.8359375, "learning_rate": 7.679795576050013e-07, "loss": 0.0767, "reward": 1.771458923816681, "reward_std": 0.41519203782081604, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9305556118488312, "rewards/repetition_penalty_reward": -0.03409669268876314, "rewards/tag_count_reward": 0.875, "step": 553 }, { "clip_ratio": 0.0, "completion_length": 417.7916717529297, "epoch": 0.40615835777126097, "grad_norm": 13.509457748782284, "kl": 3.4453125, "learning_rate": 7.669708686120282e-07, "loss": 0.2013, "reward": 1.7756158113479614, "reward_std": 0.5011553764343262, "rewards/accuracy_reward": 0.0833333358168602, "rewards/reasoning_steps_reward": 0.902777761220932, "rewards/repetition_penalty_reward": -0.03341205231845379, "rewards/tag_count_reward": 0.8229166865348816, "step": 554 }, { "clip_ratio": 0.0, "completion_length": 417.2708435058594, "epoch": 0.40689149560117305, "grad_norm": 10.674269085308994, "kl": 0.70703125, "learning_rate": 7.659607572515884e-07, "loss": 0.0192, "reward": 1.9148439764976501, "reward_std": 0.18313531577587128, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.022656043991446495, "rewards/tag_count_reward": 0.9583333730697632, "step": 555 }, { "clip_ratio": 0.0, "completion_length": 368.87501525878906, "epoch": 0.40762463343108507, "grad_norm": 8.82430760579767, "kl": 0.8994140625, "learning_rate": 7.649492301455363e-07, "loss": 0.0822, "reward": 2.157172203063965, "reward_std": 0.19496209174394608, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.028591575101017952, "rewards/tag_count_reward": 0.9635416865348816, "step": 556 }, { "clip_ratio": 0.0, "completion_length": 429.2291717529297, "epoch": 0.4083577712609971, "grad_norm": 23.122266769477772, "kl": 1.99609375, "learning_rate": 7.639362939250076e-07, "loss": 0.3555, "reward": 1.9366953372955322, "reward_std": 0.46149471402168274, "rewards/accuracy_reward": 0.1458333432674408, "rewards/reasoning_steps_reward": 0.9444444477558136, "rewards/repetition_penalty_reward": -0.04420755058526993, "rewards/tag_count_reward": 0.8906250298023224, "step": 557 }, { "clip_ratio": 0.0, "completion_length": 435.8333435058594, "epoch": 0.4090909090909091, "grad_norm": 41.38830099154527, "kl": 4.744140625, "learning_rate": 7.629219552303754e-07, "loss": 0.5603, "reward": 1.8085232377052307, "reward_std": 0.2977941185235977, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.03175466135144234, "rewards/tag_count_reward": 0.875, "step": 558 }, { "clip_ratio": 0.0, "completion_length": 379.1666717529297, "epoch": 0.40982404692082114, "grad_norm": 20.29442532919148, "kl": 1.107421875, "learning_rate": 7.619062207112068e-07, "loss": 0.212, "reward": 1.9000099897384644, "reward_std": 0.2104347199201584, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.028809464536607265, "rewards/tag_count_reward": 0.9427083432674408, "step": 559 }, { "clip_ratio": 0.0, "completion_length": 366.54168701171875, "epoch": 0.41055718475073316, "grad_norm": 8.727738513089783, "kl": 1.0615234375, "learning_rate": 7.608890970262194e-07, "loss": 0.0707, "reward": 1.9224351644515991, "reward_std": 0.123790193349123, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9722222089767456, "rewards/repetition_penalty_reward": -0.028953732922673225, "rewards/tag_count_reward": 0.9791666865348816, "step": 560 }, { "clip_ratio": 0.0, "completion_length": 379.12501525878906, "epoch": 0.4112903225806452, "grad_norm": 14.723980230907916, "kl": 1.166015625, "learning_rate": 7.598705908432375e-07, "loss": 0.2379, "reward": 2.101100206375122, "reward_std": 0.3793186992406845, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.04299703519791365, "rewards/tag_count_reward": 0.9427083432674408, "step": 561 }, { "clip_ratio": 0.0, "completion_length": 416.3125, "epoch": 0.4120234604105572, "grad_norm": 9.505002145135581, "kl": 1.28515625, "learning_rate": 7.588507088391487e-07, "loss": 0.2369, "reward": 1.803601086139679, "reward_std": 0.3192812353372574, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.05924614891409874, "rewards/tag_count_reward": 0.8906250298023224, "step": 562 }, { "clip_ratio": 0.0, "completion_length": 477.29168701171875, "epoch": 0.41275659824046923, "grad_norm": 14.43161494631361, "kl": 1.603515625, "learning_rate": 7.578294576998594e-07, "loss": 0.2643, "reward": 2.1314350366592407, "reward_std": 0.541794627904892, "rewards/accuracy_reward": 0.3333333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.07689837738871574, "rewards/tag_count_reward": 0.8958333432674408, "step": 563 }, { "clip_ratio": 0.0, "completion_length": 491.0208435058594, "epoch": 0.41348973607038125, "grad_norm": 14.848679504928713, "kl": 1.63671875, "learning_rate": 7.568068441202519e-07, "loss": 0.544, "reward": 1.7557438015937805, "reward_std": 0.41191862523555756, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9444444179534912, "rewards/repetition_penalty_reward": -0.10536744445562363, "rewards/tag_count_reward": 0.9166666865348816, "step": 564 }, { "clip_ratio": 0.0, "completion_length": 559.1250152587891, "epoch": 0.4142228739002933, "grad_norm": 49.12981105771819, "kl": 5.0625, "learning_rate": 7.557828748041399e-07, "loss": 0.7422, "reward": 1.5565171241760254, "reward_std": 0.9230948686599731, "rewards/accuracy_reward": 0.125, "rewards/reasoning_steps_reward": 0.8125, "rewards/repetition_penalty_reward": -0.1674412302672863, "rewards/tag_count_reward": 0.7864583432674408, "step": 565 }, { "clip_ratio": 0.0, "completion_length": 439.00001525878906, "epoch": 0.4149560117302053, "grad_norm": 29.421892340472176, "kl": 3.5234375, "learning_rate": 7.547575564642248e-07, "loss": 0.4018, "reward": 1.9395569562911987, "reward_std": 0.5897162556648254, "rewards/accuracy_reward": 0.18750000558793545, "rewards/reasoning_steps_reward": 0.951388955116272, "rewards/repetition_penalty_reward": -0.0795401930809021, "rewards/tag_count_reward": 0.8802083432674408, "step": 566 }, { "clip_ratio": 0.0, "completion_length": 649.0833435058594, "epoch": 0.4156891495601173, "grad_norm": 16.04782963498486, "kl": 2.515625, "learning_rate": 7.537308958220516e-07, "loss": 0.5091, "reward": 1.5547910928726196, "reward_std": 0.8353258967399597, "rewards/accuracy_reward": 0.1875000111758709, "rewards/reasoning_steps_reward": 0.8263889253139496, "rewards/repetition_penalty_reward": -0.22472276538610458, "rewards/tag_count_reward": 0.7656250298023224, "step": 567 }, { "clip_ratio": 0.0, "completion_length": 625.6666870117188, "epoch": 0.41642228739002934, "grad_norm": 17.758582322238592, "kl": 2.8671875, "learning_rate": 7.527028996079647e-07, "loss": 0.451, "reward": 1.5067104697227478, "reward_std": 0.675364226102829, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8958333432674408, "rewards/repetition_penalty_reward": -0.17558125406503677, "rewards/tag_count_reward": 0.7864583730697632, "step": 568 }, { "clip_ratio": 0.0, "completion_length": 558.6250152587891, "epoch": 0.41715542521994137, "grad_norm": 15.295317041112742, "kl": 1.40234375, "learning_rate": 7.516735745610641e-07, "loss": 0.5649, "reward": 1.7219347953796387, "reward_std": 0.4677434712648392, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.951388955116272, "rewards/repetition_penalty_reward": -0.12007906287908554, "rewards/tag_count_reward": 0.8906250298023224, "step": 569 }, { "clip_ratio": 0.0, "completion_length": 585.5625305175781, "epoch": 0.4178885630498534, "grad_norm": 56.80424556977392, "kl": 0.93359375, "learning_rate": 7.50642927429161e-07, "loss": 0.5743, "reward": 1.7323921918869019, "reward_std": 0.47409604489803314, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.10788561776280403, "rewards/tag_count_reward": 0.8541666865348816, "step": 570 }, { "clip_ratio": 0.0, "completion_length": 460.7708435058594, "epoch": 0.4186217008797654, "grad_norm": 143.1280548384441, "kl": 8.46875, "learning_rate": 7.496109649687336e-07, "loss": 0.5721, "reward": 1.6834684610366821, "reward_std": 0.6084943413734436, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.916666716337204, "rewards/repetition_penalty_reward": -0.10298987478017807, "rewards/tag_count_reward": 0.8697916865348816, "step": 571 }, { "clip_ratio": 0.0, "completion_length": 471.1666717529297, "epoch": 0.41935483870967744, "grad_norm": 129677.80062291445, "kl": 768.5, "learning_rate": 7.485776939448831e-07, "loss": 64.3934, "reward": 1.7917594909667969, "reward_std": 0.3990024924278259, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9444444179534912, "rewards/repetition_penalty_reward": -0.06935160979628563, "rewards/tag_count_reward": 0.9166666865348816, "step": 572 }, { "clip_ratio": 0.0, "completion_length": 489.29168701171875, "epoch": 0.42008797653958946, "grad_norm": 3810326.3768206835, "kl": 14339.5, "learning_rate": 7.475431211312886e-07, "loss": 1697.3242, "reward": 1.9151965379714966, "reward_std": 0.544783428311348, "rewards/accuracy_reward": 0.16666667722165585, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.09174791350960732, "rewards/tag_count_reward": 0.8750000298023224, "step": 573 }, { "clip_ratio": 0.0, "completion_length": 506.3750305175781, "epoch": 0.4208211143695015, "grad_norm": 37994.57020687281, "kl": 282.0, "learning_rate": 7.465072533101633e-07, "loss": 20.5828, "reward": 2.073698043823242, "reward_std": 0.3669005036354065, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.07039912790060043, "rewards/tag_count_reward": 0.9427083432674408, "step": 574 }, { "clip_ratio": 0.0, "completion_length": 465.125, "epoch": 0.4215542521994135, "grad_norm": 1231.8851669305193, "kl": 9.5, "learning_rate": 7.454700972722102e-07, "loss": 1.2964, "reward": 1.826287567615509, "reward_std": 0.4451441168785095, "rewards/accuracy_reward": 0.0625, "rewards/reasoning_steps_reward": 0.951388955116272, "rewards/repetition_penalty_reward": -0.046976424753665924, "rewards/tag_count_reward": 0.8593750298023224, "step": 575 }, { "clip_ratio": 0.0, "completion_length": 440.2916717529297, "epoch": 0.4222873900293255, "grad_norm": 13.78789946205212, "kl": 1.40625, "learning_rate": 7.444316598165771e-07, "loss": 0.1817, "reward": 1.935817837715149, "reward_std": 0.4649259001016617, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.944444477558136, "rewards/repetition_penalty_reward": -0.05550179071724415, "rewards/tag_count_reward": 0.8177083730697632, "step": 576 }, { "clip_ratio": 0.0, "completion_length": 469.10418701171875, "epoch": 0.42302052785923755, "grad_norm": 12.69432344405188, "kl": 0.638671875, "learning_rate": 7.433919477508122e-07, "loss": 0.2687, "reward": 1.9238691926002502, "reward_std": 0.20775415003299713, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03446421958506107, "rewards/tag_count_reward": 0.9375, "step": 577 }, { "clip_ratio": 0.0, "completion_length": 434.25001525878906, "epoch": 0.4237536656891496, "grad_norm": 12.156642111389008, "kl": 0.55859375, "learning_rate": 7.423509678908197e-07, "loss": 0.2053, "reward": 1.9572932720184326, "reward_std": 0.23687171936035156, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.027081767097115517, "rewards/tag_count_reward": 0.9427083432674408, "step": 578 }, { "clip_ratio": 0.0, "completion_length": 491.9166717529297, "epoch": 0.4244868035190616, "grad_norm": 12.901366210323863, "kl": 1.06640625, "learning_rate": 7.413087270608142e-07, "loss": 0.3456, "reward": 1.9319626688957214, "reward_std": 0.44203020632267, "rewards/accuracy_reward": 0.1458333432674408, "rewards/reasoning_steps_reward": 0.9583334028720856, "rewards/repetition_penalty_reward": -0.031579023227095604, "rewards/tag_count_reward": 0.8593750298023224, "step": 579 }, { "clip_ratio": 0.0, "completion_length": 425.0, "epoch": 0.4252199413489736, "grad_norm": 5.351454860184803, "kl": 0.92578125, "learning_rate": 7.402652320932778e-07, "loss": 0.0296, "reward": 1.9846732020378113, "reward_std": 0.47941556572914124, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9305555820465088, "rewards/repetition_penalty_reward": -0.02400730736553669, "rewards/tag_count_reward": 0.8906250298023224, "step": 580 }, { "clip_ratio": 0.0, "completion_length": 531.4375305175781, "epoch": 0.42595307917888564, "grad_norm": 7.963531966396582, "kl": 0.79296875, "learning_rate": 7.392204898289134e-07, "loss": 0.4304, "reward": 1.854543387889862, "reward_std": 0.24265119433403015, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.029137184843420982, "rewards/tag_count_reward": 0.8906250298023224, "step": 581 }, { "clip_ratio": 0.0, "completion_length": 436.7083435058594, "epoch": 0.42668621700879766, "grad_norm": 5.484500184954792, "kl": 1.0458984375, "learning_rate": 7.381745071166008e-07, "loss": 0.2121, "reward": 1.8330828547477722, "reward_std": 0.2432198002934456, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9375, "rewards/repetition_penalty_reward": -0.03670879080891609, "rewards/tag_count_reward": 0.9322916865348816, "step": 582 }, { "clip_ratio": 0.0, "completion_length": 689.2291717529297, "epoch": 0.4274193548387097, "grad_norm": 4.163101070137608, "kl": 1.4345703125, "learning_rate": 7.371272908133517e-07, "loss": 0.2019, "reward": 1.8426572680473328, "reward_std": 0.3393200449645519, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.840277761220932, "rewards/repetition_penalty_reward": -0.0288705974817276, "rewards/tag_count_reward": 0.78125, "step": 583 }, { "clip_ratio": 0.0, "completion_length": 523.0000305175781, "epoch": 0.4281524926686217, "grad_norm": 5.91565183853525, "kl": 0.826171875, "learning_rate": 7.360788477842648e-07, "loss": 0.4024, "reward": 2.025831699371338, "reward_std": 0.3373628966510296, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.033196162432432175, "rewards/tag_count_reward": 0.9062500298023224, "step": 584 }, { "clip_ratio": 0.0, "completion_length": 658.75, "epoch": 0.42888563049853373, "grad_norm": 7.01606828210777, "kl": 0.93359375, "learning_rate": 7.350291849024802e-07, "loss": 0.6511, "reward": 2.2463026642799377, "reward_std": 0.3866378962993622, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.958333283662796, "rewards/repetition_penalty_reward": -0.040155697613954544, "rewards/tag_count_reward": 0.8906250298023224, "step": 585 }, { "clip_ratio": 0.0, "completion_length": 631.1041870117188, "epoch": 0.42961876832844575, "grad_norm": 5.140346553565747, "kl": 1.259765625, "learning_rate": 7.339783090491357e-07, "loss": 0.4162, "reward": 1.7601860761642456, "reward_std": 0.364535853266716, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9236111640930176, "rewards/repetition_penalty_reward": -0.02280005533248186, "rewards/tag_count_reward": 0.8593750298023224, "step": 586 }, { "clip_ratio": 0.0, "completion_length": 842.2083740234375, "epoch": 0.4303519061583578, "grad_norm": 8.722800979257899, "kl": 2.41796875, "learning_rate": 7.329262271133198e-07, "loss": 0.5189, "reward": 1.3562277555465698, "reward_std": 0.685504287481308, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.7430556118488312, "rewards/repetition_penalty_reward": -0.027452876791357994, "rewards/tag_count_reward": 0.6406250298023224, "step": 587 }, { "clip_ratio": 0.0, "completion_length": 617.9583587646484, "epoch": 0.4310850439882698, "grad_norm": 5.737199758652915, "kl": 1.12890625, "learning_rate": 7.318729459920285e-07, "loss": 0.3464, "reward": 1.767576813697815, "reward_std": 0.36326418817043304, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9236111342906952, "rewards/repetition_penalty_reward": -0.025825947523117065, "rewards/tag_count_reward": 0.8697916865348816, "step": 588 }, { "clip_ratio": 0.0, "completion_length": 667.5416870117188, "epoch": 0.4318181818181818, "grad_norm": 9.981678337385196, "kl": 1.375, "learning_rate": 7.308184725901184e-07, "loss": 0.4301, "reward": 1.7179012894630432, "reward_std": 0.4680147022008896, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8819444477558136, "rewards/repetition_penalty_reward": -0.018209854140877724, "rewards/tag_count_reward": 0.8541666865348816, "step": 589 }, { "clip_ratio": 0.0, "completion_length": 527.3958435058594, "epoch": 0.43255131964809385, "grad_norm": 5.970499864844468, "kl": 0.75390625, "learning_rate": 7.297628138202627e-07, "loss": 0.6867, "reward": 1.855049967765808, "reward_std": 0.2735976576805115, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.026894627138972282, "rewards/tag_count_reward": 0.9166666865348816, "step": 590 }, { "clip_ratio": 0.0, "completion_length": 423.5625, "epoch": 0.43328445747800587, "grad_norm": 6.423946368381474, "kl": 0.734375, "learning_rate": 7.287059766029048e-07, "loss": 0.3071, "reward": 2.242375373840332, "reward_std": 0.3972780704498291, "rewards/accuracy_reward": 0.3750000149011612, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.023249639198184013, "rewards/tag_count_reward": 0.9322916865348816, "step": 591 }, { "clip_ratio": 0.0, "completion_length": 595.7291870117188, "epoch": 0.4340175953079179, "grad_norm": 4.887451300319632, "kl": 0.9296875, "learning_rate": 7.276479678662142e-07, "loss": 0.694, "reward": 2.0136942863464355, "reward_std": 0.4511658251285553, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9305555522441864, "rewards/repetition_penalty_reward": -0.021027985960245132, "rewards/tag_count_reward": 0.8958333730697632, "step": 592 }, { "clip_ratio": 0.0, "completion_length": 523.5416870117188, "epoch": 0.4347507331378299, "grad_norm": 7.661491807669517, "kl": 0.939453125, "learning_rate": 7.265887945460399e-07, "loss": 0.4451, "reward": 1.8619901537895203, "reward_std": 0.2486792877316475, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.025162680074572563, "rewards/tag_count_reward": 0.9218750298023224, "step": 593 }, { "clip_ratio": 0.0, "completion_length": 451.5833435058594, "epoch": 0.43548387096774194, "grad_norm": 9.321412543739491, "kl": 0.796875, "learning_rate": 7.255284635858654e-07, "loss": 0.4428, "reward": 1.8684191703796387, "reward_std": 0.2588324770331383, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.03609476983547211, "rewards/tag_count_reward": 0.9322916865348816, "step": 594 }, { "clip_ratio": 0.0, "completion_length": 606.2708435058594, "epoch": 0.43621700879765396, "grad_norm": 5.574219705760587, "kl": 1.2734375, "learning_rate": 7.244669819367635e-07, "loss": 0.8792, "reward": 1.7151203751564026, "reward_std": 0.5295811295509338, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.8888889253139496, "rewards/repetition_penalty_reward": -0.022726922295987606, "rewards/tag_count_reward": 0.8489583432674408, "step": 595 }, { "clip_ratio": 0.0, "completion_length": 483.6041717529297, "epoch": 0.436950146627566, "grad_norm": 7.654459613602763, "kl": 0.75390625, "learning_rate": 7.2340435655735e-07, "loss": 0.7024, "reward": 1.8384817838668823, "reward_std": 0.36958810687065125, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9374999701976776, "rewards/repetition_penalty_reward": -0.020893272012472153, "rewards/tag_count_reward": 0.921875, "step": 596 }, { "clip_ratio": 0.0, "completion_length": 465.8958435058594, "epoch": 0.437683284457478, "grad_norm": 7.854088448222693, "kl": 0.619140625, "learning_rate": 7.223405944137391e-07, "loss": 0.6323, "reward": 1.9075292348861694, "reward_std": 0.23356913030147552, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.024762499146163464, "rewards/tag_count_reward": 0.953125, "step": 597 }, { "clip_ratio": 0.0, "completion_length": 538.9375152587891, "epoch": 0.43841642228739003, "grad_norm": 10.82627900467281, "kl": 1.205078125, "learning_rate": 7.212757024794963e-07, "loss": 0.6552, "reward": 1.823222577571869, "reward_std": 0.34460779651999474, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9375, "rewards/repetition_penalty_reward": -0.025735745206475258, "rewards/tag_count_reward": 0.9114583730697632, "step": 598 }, { "clip_ratio": 0.0, "completion_length": 434.50001525878906, "epoch": 0.43914956011730205, "grad_norm": 17.304697371168608, "kl": 0.70703125, "learning_rate": 7.202096877355943e-07, "loss": 0.4098, "reward": 1.9486218690872192, "reward_std": 0.09528243541717529, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02533646486699581, "rewards/tag_count_reward": 0.9739583432674408, "step": 599 }, { "clip_ratio": 0.0, "completion_length": 521.3750152587891, "epoch": 0.4398826979472141, "grad_norm": 10.712298357559916, "kl": 0.5244140625, "learning_rate": 7.191425571703661e-07, "loss": 0.3627, "reward": 1.8139049410820007, "reward_std": 0.31058455258607864, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9097222685813904, "rewards/repetition_penalty_reward": -0.017692371271550655, "rewards/tag_count_reward": 0.921875, "step": 600 }, { "clip_ratio": 0.0, "completion_length": 430.1041717529297, "epoch": 0.4406158357771261, "grad_norm": 7.6757110936792765, "kl": 0.626953125, "learning_rate": 7.180743177794596e-07, "loss": 0.2585, "reward": 1.9423617124557495, "reward_std": 0.10574298352003098, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021179988980293274, "rewards/tag_count_reward": 0.9635416865348816, "step": 601 }, { "clip_ratio": 0.0, "completion_length": 387.9583435058594, "epoch": 0.4413489736070381, "grad_norm": 4.489868769575807, "kl": 0.4794921875, "learning_rate": 7.170049765657915e-07, "loss": 0.1947, "reward": 1.9460084438323975, "reward_std": 0.1343548847362399, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.017533178441226482, "rewards/tag_count_reward": 0.984375, "step": 602 }, { "clip_ratio": 0.0, "completion_length": 469.5625305175781, "epoch": 0.44208211143695014, "grad_norm": 7.150423783435916, "kl": 1.00390625, "learning_rate": 7.159345405395021e-07, "loss": 0.4532, "reward": 2.0186060070991516, "reward_std": 0.43175019323825836, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9097222089767456, "rewards/repetition_penalty_reward": -0.026532936841249466, "rewards/tag_count_reward": 0.9270833432674408, "step": 603 }, { "clip_ratio": 0.0, "completion_length": 484.0833435058594, "epoch": 0.44281524926686217, "grad_norm": 14.289522543503129, "kl": 0.783203125, "learning_rate": 7.148630167179082e-07, "loss": 0.7236, "reward": 1.852538287639618, "reward_std": 0.35337747633457184, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.944444477558136, "rewards/repetition_penalty_reward": -0.024197802878916264, "rewards/tag_count_reward": 0.9322916865348816, "step": 604 }, { "clip_ratio": 0.0, "completion_length": 601.2291717529297, "epoch": 0.4435483870967742, "grad_norm": 11.952594597793968, "kl": 0.5205078125, "learning_rate": 7.137904121254578e-07, "loss": 0.1753, "reward": 1.763185977935791, "reward_std": 0.2357357507571578, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9097222685813904, "rewards/repetition_penalty_reward": -0.026744673028588295, "rewards/tag_count_reward": 0.8802083432674408, "step": 605 }, { "clip_ratio": 0.0, "completion_length": 375.25001525878906, "epoch": 0.4442815249266862, "grad_norm": 3.703107528625227, "kl": 0.408203125, "learning_rate": 7.127167337936845e-07, "loss": 0.0196, "reward": 1.982218861579895, "reward_std": 0.011761181056499481, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017781143076717854, "rewards/tag_count_reward": 1.0, "step": 606 }, { "clip_ratio": 0.0, "completion_length": 429.3958435058594, "epoch": 0.44501466275659823, "grad_norm": 20.03177568564077, "kl": 0.65625, "learning_rate": 7.116419887611605e-07, "loss": 0.4329, "reward": 2.1548226475715637, "reward_std": 0.25495945662260056, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.022260690107941628, "rewards/tag_count_reward": 0.9687500298023224, "step": 607 }, { "clip_ratio": 0.0, "completion_length": 320.6666717529297, "epoch": 0.44574780058651026, "grad_norm": 3.804969767713346, "kl": 0.4248046875, "learning_rate": 7.105661840734506e-07, "loss": 0.017, "reward": 2.4734570384025574, "reward_std": 0.01555747864767909, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02654305938631296, "rewards/tag_count_reward": 1.0, "step": 608 }, { "clip_ratio": 0.0, "completion_length": 360.8333435058594, "epoch": 0.4464809384164223, "grad_norm": 3.580449205687791, "kl": 0.390625, "learning_rate": 7.094893267830666e-07, "loss": 0.0074, "reward": 2.1633920669555664, "reward_std": 0.1211917675100267, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024107900448143482, "rewards/tag_count_reward": 1.0, "step": 609 }, { "clip_ratio": 0.0, "completion_length": 338.8541717529297, "epoch": 0.4472140762463343, "grad_norm": 3.775779917602866, "kl": 0.42578125, "learning_rate": 7.084114239494209e-07, "loss": 0.0204, "reward": 1.9812775254249573, "reward_std": 0.01065837498754263, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018722583539783955, "rewards/tag_count_reward": 1.0, "step": 610 }, { "clip_ratio": 0.0, "completion_length": 323.62501525878906, "epoch": 0.4479472140762463, "grad_norm": 3.4928929408017324, "kl": 0.39453125, "learning_rate": 7.073324826387792e-07, "loss": 0.0202, "reward": 2.223867654800415, "reward_std": 0.014575928449630737, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026132473722100258, "rewards/tag_count_reward": 1.0, "step": 611 }, { "clip_ratio": 0.0, "completion_length": 417.50001525878906, "epoch": 0.44868035190615835, "grad_norm": 40.78856978225071, "kl": 0.494140625, "learning_rate": 7.062525099242161e-07, "loss": 0.2242, "reward": 2.185540795326233, "reward_std": 0.16539057483896613, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.019320469349622726, "rewards/tag_count_reward": 0.9895833432674408, "step": 612 }, { "clip_ratio": 0.0, "completion_length": 374.31251525878906, "epoch": 0.44941348973607037, "grad_norm": 3.2067514129166974, "kl": 0.34375, "learning_rate": 7.051715128855665e-07, "loss": 0.037, "reward": 1.9826087951660156, "reward_std": 0.010490587446838617, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017391293309628963, "rewards/tag_count_reward": 1.0, "step": 613 }, { "clip_ratio": 0.0, "completion_length": 389.8958435058594, "epoch": 0.4501466275659824, "grad_norm": 85.78107516846322, "kl": 1.4580078125, "learning_rate": 7.040894986093814e-07, "loss": 0.2902, "reward": 1.982279896736145, "reward_std": 0.09320326708257198, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028136792592704296, "rewards/tag_count_reward": 0.9895833432674408, "step": 614 }, { "clip_ratio": 0.0, "completion_length": 385.7708435058594, "epoch": 0.4508797653958944, "grad_norm": 2.815951559024344, "kl": 0.330078125, "learning_rate": 7.030064741888798e-07, "loss": 0.0066, "reward": 1.9806150794029236, "reward_std": 0.012543195858597755, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019384922459721565, "rewards/tag_count_reward": 1.0, "step": 615 }, { "clip_ratio": 0.0, "completion_length": 401.87501525878906, "epoch": 0.45161290322580644, "grad_norm": 52.55065929758277, "kl": 1.103515625, "learning_rate": 7.019224467239025e-07, "loss": 0.2908, "reward": 2.6719231605529785, "reward_std": 0.2000698889605701, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.022521505132317543, "rewards/tag_count_reward": 0.9791666865348816, "step": 616 }, { "clip_ratio": 0.0, "completion_length": 392.2083435058594, "epoch": 0.45234604105571846, "grad_norm": 63.52951187805423, "kl": 1.818359375, "learning_rate": 7.008374233208667e-07, "loss": 0.4064, "reward": 1.9457017183303833, "reward_std": 0.10482199117541313, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.0265205055475235, "rewards/tag_count_reward": 0.9791666865348816, "step": 617 }, { "clip_ratio": 0.0, "completion_length": 386.1666717529297, "epoch": 0.4530791788856305, "grad_norm": 2.7120803027810174, "kl": 0.306640625, "learning_rate": 6.997514110927176e-07, "loss": 0.0391, "reward": 1.9820279479026794, "reward_std": 0.009422503411769867, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017972068395465612, "rewards/tag_count_reward": 1.0, "step": 618 }, { "clip_ratio": 0.0, "completion_length": 386.0833435058594, "epoch": 0.4538123167155425, "grad_norm": 2.632747872063834, "kl": 0.3076171875, "learning_rate": 6.986644171588833e-07, "loss": 0.026, "reward": 2.224295914173126, "reward_std": 0.013977315742522478, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02570413239300251, "rewards/tag_count_reward": 1.0, "step": 619 }, { "clip_ratio": 0.0, "completion_length": 385.6041717529297, "epoch": 0.45454545454545453, "grad_norm": 2.8309147137871613, "kl": 0.3193359375, "learning_rate": 6.975764486452274e-07, "loss": 0.0364, "reward": 2.4791120290756226, "reward_std": 0.010151438880711794, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.020888016559183598, "rewards/tag_count_reward": 1.0, "step": 620 }, { "clip_ratio": 0.0, "completion_length": 517.3333435058594, "epoch": 0.45527859237536655, "grad_norm": 35.189103192855995, "kl": 1.181640625, "learning_rate": 6.964875126840023e-07, "loss": 0.332, "reward": 1.9061416387557983, "reward_std": 0.1184447268024087, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.019205675926059484, "rewards/tag_count_reward": 0.9322916865348816, "step": 621 }, { "clip_ratio": 0.0, "completion_length": 483.41668701171875, "epoch": 0.4560117302052786, "grad_norm": 66.83089282659809, "kl": 0.80859375, "learning_rate": 6.953976164138027e-07, "loss": 0.3831, "reward": 1.9612208604812622, "reward_std": 0.18723435327410698, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017945820465683937, "rewards/tag_count_reward": 0.9583333432674408, "step": 622 }, { "clip_ratio": 0.0, "completion_length": 405.3125, "epoch": 0.4567448680351906, "grad_norm": 24.93646819069722, "kl": 0.423828125, "learning_rate": 6.943067669795185e-07, "loss": 0.2185, "reward": 2.20524525642395, "reward_std": 0.06307589076459408, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02912973240017891, "rewards/tag_count_reward": 0.984375, "step": 623 }, { "clip_ratio": 0.0, "completion_length": 475.3333435058594, "epoch": 0.4574780058651026, "grad_norm": 43.16415673427742, "kl": 1.5078125, "learning_rate": 6.932149715322886e-07, "loss": 0.4779, "reward": 2.1597553491592407, "reward_std": 0.28502483665943146, "rewards/accuracy_reward": 0.2708333432674408, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.024272335693240166, "rewards/tag_count_reward": 0.9270833432674408, "step": 624 }, { "clip_ratio": 0.0, "completion_length": 385.5208435058594, "epoch": 0.45821114369501464, "grad_norm": 2.6099512459968337, "kl": 0.3330078125, "learning_rate": 6.921222372294529e-07, "loss": 0.0136, "reward": 1.9811992049217224, "reward_std": 0.011327789165079594, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018800782039761543, "rewards/tag_count_reward": 1.0, "step": 625 }, { "clip_ratio": 0.0, "completion_length": 458.8958435058594, "epoch": 0.45894428152492667, "grad_norm": 11.277984902992305, "kl": 0.5712890625, "learning_rate": 6.910285712345065e-07, "loss": 0.2365, "reward": 2.384067118167877, "reward_std": 0.24414547625929117, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022182924207299948, "rewards/tag_count_reward": 0.96875, "step": 626 }, { "clip_ratio": 0.0, "completion_length": 410.06251525878906, "epoch": 0.4596774193548387, "grad_norm": 5718.023712216834, "kl": 57.65234375, "learning_rate": 6.899339807170521e-07, "loss": 8.7021, "reward": 1.962087333202362, "reward_std": 0.06420454941689968, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02228778973221779, "rewards/tag_count_reward": 0.984375, "step": 627 }, { "clip_ratio": 0.0, "completion_length": 488.43751525878906, "epoch": 0.4604105571847507, "grad_norm": 7150.668791263632, "kl": 115.890625, "learning_rate": 6.888384728527532e-07, "loss": 12.9975, "reward": 2.137051224708557, "reward_std": 0.3158814609050751, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.020934815518558025, "rewards/tag_count_reward": 0.9427083432674408, "step": 628 }, { "clip_ratio": 0.0, "completion_length": 516.4583587646484, "epoch": 0.46114369501466274, "grad_norm": 8244.611512935808, "kl": 215.15234375, "learning_rate": 6.877420548232871e-07, "loss": 22.5839, "reward": 2.3565913438796997, "reward_std": 0.23759527131915092, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.020144745707511902, "rewards/tag_count_reward": 0.953125, "step": 629 }, { "clip_ratio": 0.0, "completion_length": 412.50001525878906, "epoch": 0.46187683284457476, "grad_norm": 10.67765423637928, "kl": 0.3955078125, "learning_rate": 6.866447338162975e-07, "loss": 0.2181, "reward": 1.9459137916564941, "reward_std": 0.10560346394777298, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.02110019139945507, "rewards/tag_count_reward": 0.9739583432674408, "step": 630 }, { "clip_ratio": 0.0, "completion_length": 437.9583435058594, "epoch": 0.4626099706744868, "grad_norm": 11.397428720346749, "kl": 0.3564453125, "learning_rate": 6.855465170253479e-07, "loss": 0.2139, "reward": 2.1977256536483765, "reward_std": 0.11736413929611444, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021024403162300587, "rewards/tag_count_reward": 0.9895833432674408, "step": 631 }, { "clip_ratio": 0.0, "completion_length": 566.6250152587891, "epoch": 0.4633431085043988, "grad_norm": 594219.6460833271, "kl": 8192.1513671875, "learning_rate": 6.844474116498742e-07, "loss": 662.2282, "reward": 1.8581966161727905, "reward_std": 0.1760244253091514, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.965277761220932, "rewards/repetition_penalty_reward": -0.018539583310484886, "rewards/tag_count_reward": 0.9114583432674408, "step": 632 }, { "clip_ratio": 0.0, "completion_length": 495.41668701171875, "epoch": 0.4640762463343108, "grad_norm": 36.19090425617034, "kl": 0.5078125, "learning_rate": 6.833474248951372e-07, "loss": 0.5879, "reward": 1.9268569946289062, "reward_std": 0.15591158717870712, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02105969563126564, "rewards/tag_count_reward": 0.9479166865348816, "step": 633 }, { "clip_ratio": 0.0, "completion_length": 500.2708435058594, "epoch": 0.46480938416422285, "grad_norm": 35.71072908814854, "kl": 0.626953125, "learning_rate": 6.822465639721759e-07, "loss": 0.6124, "reward": 1.8967062830924988, "reward_std": 0.24515408277511597, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.018224384635686874, "rewards/tag_count_reward": 0.9218750298023224, "step": 634 }, { "clip_ratio": 0.0, "completion_length": 552.2916870117188, "epoch": 0.4655425219941349, "grad_norm": 95.19295300214017, "kl": 2.70703125, "learning_rate": 6.811448360977596e-07, "loss": 0.6447, "reward": 2.1306493878364563, "reward_std": 0.2458694726228714, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025600734166800976, "rewards/tag_count_reward": 0.9270833730697632, "step": 635 }, { "clip_ratio": 0.0, "completion_length": 497.875, "epoch": 0.4662756598240469, "grad_norm": 44.15333708156457, "kl": 3.00390625, "learning_rate": 6.800422484943417e-07, "loss": 0.6232, "reward": 2.059449315071106, "reward_std": 0.43308182805776596, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.009995209518820047, "rewards/tag_count_reward": 0.8958333432674408, "step": 636 }, { "clip_ratio": 0.0, "completion_length": 537.4583435058594, "epoch": 0.4670087976539589, "grad_norm": 18.55847948178239, "kl": 1.6640625, "learning_rate": 6.789388083900107e-07, "loss": 0.6202, "reward": 1.8917770385742188, "reward_std": 0.2116881087422371, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.017945221159607172, "rewards/tag_count_reward": 0.9166666865348816, "step": 637 }, { "clip_ratio": 0.0, "completion_length": 520.1041870117188, "epoch": 0.46774193548387094, "grad_norm": 89.39468645925008, "kl": 3.1953125, "learning_rate": 6.778345230184444e-07, "loss": 0.7171, "reward": 2.0839056372642517, "reward_std": 0.33054201304912567, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.018525044433772564, "rewards/tag_count_reward": 0.9218750298023224, "step": 638 }, { "clip_ratio": 0.0, "completion_length": 549.2708587646484, "epoch": 0.46847507331378296, "grad_norm": 174.26079161224064, "kl": 4.8349609375, "learning_rate": 6.767293996188617e-07, "loss": 0.8189, "reward": 1.8636788129806519, "reward_std": 0.21642551571130753, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9583333730697632, "rewards/repetition_penalty_reward": -0.016529573127627373, "rewards/tag_count_reward": 0.921875, "step": 639 }, { "clip_ratio": 0.0, "completion_length": 534.9166870117188, "epoch": 0.46920821114369504, "grad_norm": 8.999925285401675, "kl": 0.55078125, "learning_rate": 6.756234454359747e-07, "loss": 0.6351, "reward": 2.163374960422516, "reward_std": 0.22264650464057922, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013708437792956829, "rewards/tag_count_reward": 0.9479166865348816, "step": 640 }, { "clip_ratio": 0.0, "completion_length": 432.8958435058594, "epoch": 0.46994134897360706, "grad_norm": 32.04944493041125, "kl": 1.7216796875, "learning_rate": 6.745166677199426e-07, "loss": 0.4936, "reward": 1.9193060994148254, "reward_std": 0.12536632781848311, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.01993004372343421, "rewards/tag_count_reward": 0.953125, "step": 641 }, { "clip_ratio": 0.0, "completion_length": 530.9166870117188, "epoch": 0.4706744868035191, "grad_norm": 4.858789332267112, "kl": 0.7841796875, "learning_rate": 6.734090737263228e-07, "loss": 0.4012, "reward": 2.09183669090271, "reward_std": 0.37891362607479095, "rewards/accuracy_reward": 0.22916667722165585, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.015802277252078056, "rewards/tag_count_reward": 0.90625, "step": 642 }, { "clip_ratio": 0.0, "completion_length": 495.20835876464844, "epoch": 0.4714076246334311, "grad_norm": 9.696078299064906, "kl": 1.1181640625, "learning_rate": 6.72300670716024e-07, "loss": 0.4131, "reward": 2.2706050872802734, "reward_std": 0.4469531998038292, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.02106165699660778, "rewards/tag_count_reward": 0.9375, "step": 643 }, { "clip_ratio": 0.0, "completion_length": 437.0416717529297, "epoch": 0.47214076246334313, "grad_norm": 56.129441337094114, "kl": 0.8037109375, "learning_rate": 6.711914659552582e-07, "loss": 0.2667, "reward": 1.9539560675621033, "reward_std": 0.11034089652821422, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.016530086286365986, "rewards/tag_count_reward": 0.984375, "step": 644 }, { "clip_ratio": 0.0, "completion_length": 479.0625, "epoch": 0.47287390029325516, "grad_norm": 6.554459896651348, "kl": 1.6884765625, "learning_rate": 6.70081466715494e-07, "loss": 0.7371, "reward": 2.142140805721283, "reward_std": 0.2443421185016632, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.02278986293822527, "rewards/tag_count_reward": 0.9427083432674408, "step": 645 }, { "clip_ratio": 0.0, "completion_length": 387.9375, "epoch": 0.4736070381231672, "grad_norm": 3.081702889935719, "kl": 0.328125, "learning_rate": 6.689706802734074e-07, "loss": 0.0562, "reward": 1.9774357676506042, "reward_std": 0.011296110693365335, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0225642342120409, "rewards/tag_count_reward": 1.0, "step": 646 }, { "clip_ratio": 0.0, "completion_length": 390.5416717529297, "epoch": 0.4743401759530792, "grad_norm": 2.3819237237230233, "kl": 0.3095703125, "learning_rate": 6.678591139108357e-07, "loss": 0.0264, "reward": 1.98467218875885, "reward_std": 0.00978016760200262, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.015327838715165854, "rewards/tag_count_reward": 1.0, "step": 647 }, { "clip_ratio": 0.0, "completion_length": 398.7291717529297, "epoch": 0.4750733137829912, "grad_norm": 189.14687079985282, "kl": 5.7578125, "learning_rate": 6.667467749147286e-07, "loss": 0.9668, "reward": 2.2045015692710876, "reward_std": 0.23807373270392418, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01945687923580408, "rewards/tag_count_reward": 0.9739583730697632, "step": 648 }, { "clip_ratio": 0.0, "completion_length": 388.8958435058594, "epoch": 0.47580645161290325, "grad_norm": 2.57434822582919, "kl": 0.33203125, "learning_rate": 6.656336705771011e-07, "loss": 0.0186, "reward": 2.2355542182922363, "reward_std": 0.008534513413906097, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.014445837587118149, "rewards/tag_count_reward": 1.0, "step": 649 }, { "clip_ratio": 0.0, "completion_length": 393.37501525878906, "epoch": 0.47653958944281527, "grad_norm": 3.0952793766641205, "kl": 0.396484375, "learning_rate": 6.64519808194985e-07, "loss": 0.2247, "reward": 1.9666798114776611, "reward_std": 0.060795275028795004, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01769520714879036, "rewards/tag_count_reward": 0.984375, "step": 650 }, { "clip_ratio": 0.0, "completion_length": 417.4583435058594, "epoch": 0.4772727272727273, "grad_norm": 3.0392356131499154, "kl": 0.3896484375, "learning_rate": 6.634051950703822e-07, "loss": 0.1672, "reward": 1.970879852771759, "reward_std": 0.06113596260547638, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013495232909917831, "rewards/tag_count_reward": 0.984375, "step": 651 }, { "clip_ratio": 0.0, "completion_length": 502.5208435058594, "epoch": 0.4780058651026393, "grad_norm": 5.45870950409704, "kl": 0.52734375, "learning_rate": 6.622898385102155e-07, "loss": 0.6202, "reward": 2.131287693977356, "reward_std": 0.2859160155057907, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.018017924390733242, "rewards/tag_count_reward": 0.9270833432674408, "step": 652 }, { "clip_ratio": 0.0, "completion_length": 441.2291717529297, "epoch": 0.47873900293255134, "grad_norm": 5.882125844096172, "kl": 0.4833984375, "learning_rate": 6.611737458262817e-07, "loss": 0.3992, "reward": 1.9523559212684631, "reward_std": 0.12868256820365787, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.011185822542756796, "rewards/tag_count_reward": 0.9635416865348816, "step": 653 }, { "clip_ratio": 0.0, "completion_length": 362.87501525878906, "epoch": 0.47947214076246336, "grad_norm": 2.5561243606757795, "kl": 0.3603515625, "learning_rate": 6.600569243352032e-07, "loss": 0.0136, "reward": 2.214748799800873, "reward_std": 0.08119441103190184, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01441799383610487, "rewards/tag_count_reward": 1.0, "step": 654 }, { "clip_ratio": 0.0, "completion_length": 430.93751525878906, "epoch": 0.4802052785923754, "grad_norm": 3.516134461894916, "kl": 0.40234375, "learning_rate": 6.589393813583804e-07, "loss": 0.2105, "reward": 1.9548333883285522, "reward_std": 0.11708549410104752, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.012180462013930082, "rewards/tag_count_reward": 0.9739583432674408, "step": 655 }, { "clip_ratio": 0.0, "completion_length": 499.43751525878906, "epoch": 0.4809384164222874, "grad_norm": 7.738154644635577, "kl": 0.7421875, "learning_rate": 6.578211242219429e-07, "loss": 0.5974, "reward": 2.122410535812378, "reward_std": 0.37096796184778214, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.019950571469962597, "rewards/tag_count_reward": 0.9479166865348816, "step": 656 }, { "clip_ratio": 0.0, "completion_length": 456.91668701171875, "epoch": 0.48167155425219943, "grad_norm": 3.665848451385049, "kl": 0.505859375, "learning_rate": 6.567021602567025e-07, "loss": 0.3077, "reward": 1.9053021669387817, "reward_std": 0.22535160183906555, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.01657282141968608, "rewards/tag_count_reward": 0.9427083730697632, "step": 657 }, { "clip_ratio": 0.0, "completion_length": 481.6875305175781, "epoch": 0.48240469208211145, "grad_norm": 93.57278211870775, "kl": 1.8466796875, "learning_rate": 6.555824967981047e-07, "loss": 0.3318, "reward": 1.928690791130066, "reward_std": 0.13919126242399216, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.014017567038536072, "rewards/tag_count_reward": 0.9635416865348816, "step": 658 }, { "clip_ratio": 0.0, "completion_length": 645.0833587646484, "epoch": 0.4831378299120235, "grad_norm": 64.99941439648312, "kl": 3.0859375, "learning_rate": 6.5446214118618e-07, "loss": 0.7544, "reward": 1.7792481184005737, "reward_std": 0.43339620530605316, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.909722238779068, "rewards/repetition_penalty_reward": -0.010682477615773678, "rewards/tag_count_reward": 0.8593750298023224, "step": 659 }, { "clip_ratio": 0.0, "completion_length": 526.9166717529297, "epoch": 0.4838709677419355, "grad_norm": 10.62723614596486, "kl": 0.767578125, "learning_rate": 6.533411007654971e-07, "loss": 0.5132, "reward": 2.2929834127426147, "reward_std": 0.40529730916023254, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.022988826036453247, "rewards/tag_count_reward": 0.9270833432674408, "step": 660 }, { "clip_ratio": 0.0, "completion_length": 482.50001525878906, "epoch": 0.4846041055718475, "grad_norm": 4.804017350233964, "kl": 0.4326171875, "learning_rate": 6.522193828851133e-07, "loss": 0.3632, "reward": 1.9454867243766785, "reward_std": 0.12859196960926056, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018054993357509375, "rewards/tag_count_reward": 0.9635416865348816, "step": 661 }, { "clip_ratio": 0.0, "completion_length": 468.5833435058594, "epoch": 0.48533724340175954, "grad_norm": 3.08066934948154, "kl": 0.419921875, "learning_rate": 6.510969948985278e-07, "loss": 0.1574, "reward": 1.9654483199119568, "reward_std": 0.0782310045324266, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013718396425247192, "rewards/tag_count_reward": 0.9791666865348816, "step": 662 }, { "clip_ratio": 0.0, "completion_length": 539.7291717529297, "epoch": 0.48607038123167157, "grad_norm": 37.92285966334552, "kl": 0.845703125, "learning_rate": 6.49973944163632e-07, "loss": 0.2618, "reward": 1.9112797379493713, "reward_std": 0.23218698799610138, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.01927582174539566, "rewards/tag_count_reward": 0.9375, "step": 663 }, { "clip_ratio": 0.0, "completion_length": 501.62501525878906, "epoch": 0.4868035190615836, "grad_norm": 3.1051785069904727, "kl": 0.525390625, "learning_rate": 6.488502380426625e-07, "loss": 0.305, "reward": 1.9035282135009766, "reward_std": 0.2898576706647873, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.023555109277367592, "rewards/tag_count_reward": 0.9270833432674408, "step": 664 }, { "clip_ratio": 0.0, "completion_length": 525.4791870117188, "epoch": 0.4875366568914956, "grad_norm": 4.4288209453427205, "kl": 0.55859375, "learning_rate": 6.477258839021519e-07, "loss": 0.4062, "reward": 2.354710102081299, "reward_std": 0.31429797410964966, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.013345747254788876, "rewards/tag_count_reward": 0.9583333432674408, "step": 665 }, { "clip_ratio": 0.0, "completion_length": 642.9375, "epoch": 0.48826979472140764, "grad_norm": 19.16682548072327, "kl": 1.3427734375, "learning_rate": 6.466008891128814e-07, "loss": 0.6273, "reward": 1.7963217496871948, "reward_std": 0.35257695615291595, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9444444179534912, "rewards/repetition_penalty_reward": -0.012706090230494738, "rewards/tag_count_reward": 0.8645833432674408, "step": 666 }, { "clip_ratio": 0.0, "completion_length": 858.1041870117188, "epoch": 0.48900293255131966, "grad_norm": 7.789096598920068, "kl": 1.79296875, "learning_rate": 6.454752610498317e-07, "loss": 0.9041, "reward": 1.8778371810913086, "reward_std": 0.5289941728115082, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9375000298023224, "rewards/repetition_penalty_reward": -0.012787907849997282, "rewards/tag_count_reward": 0.7656250298023224, "step": 667 }, { "clip_ratio": 0.0, "completion_length": 671.4583740234375, "epoch": 0.4897360703812317, "grad_norm": 14.917358872470148, "kl": 1.033203125, "learning_rate": 6.44349007092135e-07, "loss": 0.7233, "reward": 1.8810378909111023, "reward_std": 0.295869916677475, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02000374812632799, "rewards/tag_count_reward": 0.8802083432674408, "step": 668 }, { "clip_ratio": 0.0, "completion_length": 681.5208435058594, "epoch": 0.4904692082111437, "grad_norm": 6.570905634332384, "kl": 1.029296875, "learning_rate": 6.432221346230269e-07, "loss": 0.6374, "reward": 1.8366022109985352, "reward_std": 0.24832479655742645, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.014092199504375458, "rewards/tag_count_reward": 0.8645833432674408, "step": 669 }, { "clip_ratio": 0.0, "completion_length": 813.3333435058594, "epoch": 0.4912023460410557, "grad_norm": 7.4287126906422305, "kl": 1.06640625, "learning_rate": 6.420946510297972e-07, "loss": 0.6177, "reward": 1.6895498037338257, "reward_std": 0.4474469721317291, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9375000596046448, "rewards/repetition_penalty_reward": -0.013575221411883831, "rewards/tag_count_reward": 0.765625, "step": 670 }, { "clip_ratio": 0.0, "completion_length": 547.0208587646484, "epoch": 0.49193548387096775, "grad_norm": 8.927590711913593, "kl": 0.697265625, "learning_rate": 6.409665637037424e-07, "loss": 0.57, "reward": 1.8452393412590027, "reward_std": 0.31727316975593567, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.014135753270238638, "rewards/tag_count_reward": 0.9010416865348816, "step": 671 }, { "clip_ratio": 0.0, "completion_length": 705.3958740234375, "epoch": 0.49266862170087977, "grad_norm": 163.2675376427096, "kl": 2.66796875, "learning_rate": 6.398378800401165e-07, "loss": 0.6324, "reward": 2.011242628097534, "reward_std": 0.4974839687347412, "rewards/accuracy_reward": 0.25000000558793545, "rewards/reasoning_steps_reward": 0.9513889253139496, "rewards/repetition_penalty_reward": -0.0182713121175766, "rewards/tag_count_reward": 0.8281250298023224, "step": 672 }, { "clip_ratio": 0.0, "completion_length": 488.85418701171875, "epoch": 0.4934017595307918, "grad_norm": 6.11656633433207, "kl": 0.65234375, "learning_rate": 6.38708607438083e-07, "loss": 0.3118, "reward": 2.2023180723190308, "reward_std": 0.18355269357562065, "rewards/accuracy_reward": 0.27083333395421505, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.014695859514176846, "rewards/tag_count_reward": 0.9531250298023224, "step": 673 }, { "clip_ratio": 0.0, "completion_length": 770.5, "epoch": 0.4941348973607038, "grad_norm": 15.273687304662792, "kl": 0.986328125, "learning_rate": 6.375787533006663e-07, "loss": 0.7539, "reward": 1.7749381065368652, "reward_std": 0.43552330136299133, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.018464769702404737, "rewards/tag_count_reward": 0.8281250298023224, "step": 674 }, { "clip_ratio": 0.0, "completion_length": 678.0208435058594, "epoch": 0.49486803519061584, "grad_norm": 33.6137719864803, "kl": 0.86328125, "learning_rate": 6.364483250347026e-07, "loss": 0.7555, "reward": 1.79839426279068, "reward_std": 0.37121328711509705, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9583333730697632, "rewards/repetition_penalty_reward": -0.014105740003287792, "rewards/tag_count_reward": 0.8541666865348816, "step": 675 }, { "clip_ratio": 0.0, "completion_length": 457.7083435058594, "epoch": 0.49560117302052786, "grad_norm": 35.389228624701744, "kl": 0.537109375, "learning_rate": 6.353173300507927e-07, "loss": 0.2845, "reward": 1.9582674503326416, "reward_std": 0.06902057770639658, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.015690917149186134, "rewards/tag_count_reward": 0.9739583432674408, "step": 676 }, { "clip_ratio": 0.0, "completion_length": 567.4166870117188, "epoch": 0.4963343108504399, "grad_norm": 134.9821081723637, "kl": 0.685546875, "learning_rate": 6.341857757632519e-07, "loss": 0.4546, "reward": 2.1232765913009644, "reward_std": 0.26673005893826485, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.010404008906334639, "rewards/tag_count_reward": 0.9322916865348816, "step": 677 }, { "clip_ratio": 0.0, "completion_length": 516.5, "epoch": 0.4970674486803519, "grad_norm": 2037.170056297148, "kl": 13.15625, "learning_rate": 6.330536695900622e-07, "loss": 2.2526, "reward": 1.8933287858963013, "reward_std": 0.2740514874458313, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018129662610590458, "rewards/tag_count_reward": 0.9114583432674408, "step": 678 }, { "clip_ratio": 0.0, "completion_length": 445.2708435058594, "epoch": 0.49780058651026393, "grad_norm": 76.54198788837205, "kl": 0.474609375, "learning_rate": 6.319210189528236e-07, "loss": 0.2031, "reward": 2.4265220165252686, "reward_std": 0.14526664419099689, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01618638401851058, "rewards/tag_count_reward": 0.984375, "step": 679 }, { "clip_ratio": 0.0, "completion_length": 387.66668701171875, "epoch": 0.49853372434017595, "grad_norm": 5.588199951956003, "kl": 0.4443359375, "learning_rate": 6.307878312767053e-07, "loss": 0.0165, "reward": 1.9998939037322998, "reward_std": 0.08089298009872437, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.020939476788043976, "rewards/tag_count_reward": 1.0, "step": 680 }, { "clip_ratio": 0.0, "completion_length": 397.68751525878906, "epoch": 0.499266862170088, "grad_norm": 3.4731976371563276, "kl": 0.3974609375, "learning_rate": 6.296541139903971e-07, "loss": 0.0327, "reward": 1.9804431796073914, "reward_std": 0.011249262373894453, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019556839019060135, "rewards/tag_count_reward": 1.0, "step": 681 }, { "clip_ratio": 0.0, "completion_length": 456.2916717529297, "epoch": 0.5, "grad_norm": 59.927063692740646, "kl": 1.0146484375, "learning_rate": 6.285198745260607e-07, "loss": 0.2747, "reward": 2.4471405744552612, "reward_std": 0.13266543252393603, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.016401106491684914, "rewards/tag_count_reward": 0.984375, "step": 682 }, { "clip_ratio": 0.0, "completion_length": 439.85418701171875, "epoch": 0.500733137829912, "grad_norm": 21.88560764955481, "kl": 0.4296875, "learning_rate": 6.273851203192812e-07, "loss": 0.1805, "reward": 2.208536744117737, "reward_std": 0.07908271765336394, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.020630055107176304, "rewards/tag_count_reward": 0.9791666865348816, "step": 683 }, { "clip_ratio": 0.0, "completion_length": 485.58335876464844, "epoch": 0.501466275659824, "grad_norm": 42.20078371159907, "kl": 0.9228515625, "learning_rate": 6.262498588090177e-07, "loss": 0.3676, "reward": 1.9697940945625305, "reward_std": 0.1708728987723589, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.014581031166017056, "rewards/tag_count_reward": 0.9635416865348816, "step": 684 }, { "clip_ratio": 0.0, "completion_length": 389.31251525878906, "epoch": 0.5021994134897361, "grad_norm": 31.580308454709506, "kl": 1.0966796875, "learning_rate": 6.251140974375553e-07, "loss": 0.0433, "reward": 1.9906669855117798, "reward_std": 0.08381603378802538, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02495804987847805, "rewards/tag_count_reward": 0.9947916865348816, "step": 685 }, { "clip_ratio": 0.0, "completion_length": 396.5208435058594, "epoch": 0.5029325513196481, "grad_norm": 2.872906331597035, "kl": 0.376953125, "learning_rate": 6.239778436504557e-07, "loss": 0.0133, "reward": 2.2325448989868164, "reward_std": 0.009967302437871695, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01745530776679516, "rewards/tag_count_reward": 1.0, "step": 686 }, { "clip_ratio": 0.0, "completion_length": 433.93751525878906, "epoch": 0.5036656891495601, "grad_norm": 4.24523828143186, "kl": 0.3828125, "learning_rate": 6.228411048965091e-07, "loss": 0.0373, "reward": 2.255759596824646, "reward_std": 0.07914062775671482, "rewards/accuracy_reward": 0.27083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.015073771122843027, "rewards/tag_count_reward": 1.0, "step": 687 }, { "clip_ratio": 0.0, "completion_length": 449.7916717529297, "epoch": 0.5043988269794721, "grad_norm": 1346.6192060577287, "kl": 11.50390625, "learning_rate": 6.217038886276845e-07, "loss": 1.8327, "reward": 2.191466748714447, "reward_std": 0.08447868470102549, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02728334814310074, "rewards/tag_count_reward": 0.96875, "step": 688 }, { "clip_ratio": 0.0, "completion_length": 400.3333435058594, "epoch": 0.5051319648093842, "grad_norm": 3.0240299756883062, "kl": 0.3564453125, "learning_rate": 6.205662022990814e-07, "loss": 0.0086, "reward": 2.2511375546455383, "reward_std": 0.1696284885983914, "rewards/accuracy_reward": 0.2708333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01969574484974146, "rewards/tag_count_reward": 1.0, "step": 689 }, { "clip_ratio": 0.0, "completion_length": 385.9166717529297, "epoch": 0.5058651026392962, "grad_norm": 3.2645934417801348, "kl": 0.3955078125, "learning_rate": 6.194280533688809e-07, "loss": -0.0101, "reward": 2.0662108659744263, "reward_std": 0.24287840723991394, "rewards/accuracy_reward": 0.08333333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017122572287917137, "rewards/tag_count_reward": 1.0, "step": 690 }, { "clip_ratio": 0.0, "completion_length": 424.81251525878906, "epoch": 0.5065982404692082, "grad_norm": 3.119032374376464, "kl": 0.3515625, "learning_rate": 6.182894492982969e-07, "loss": 0.0356, "reward": 2.2503284215927124, "reward_std": 0.07950186124071479, "rewards/accuracy_reward": 0.27083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.020504925400018692, "rewards/tag_count_reward": 1.0, "step": 691 }, { "clip_ratio": 0.0, "completion_length": 506.97918701171875, "epoch": 0.5073313782991202, "grad_norm": 24.267710233732213, "kl": 1.06640625, "learning_rate": 6.171503975515265e-07, "loss": 0.2784, "reward": 1.993408441543579, "reward_std": 0.2670893892645836, "rewards/accuracy_reward": 0.06250000186264515, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.027424994856119156, "rewards/tag_count_reward": 0.9583333432674408, "step": 692 }, { "clip_ratio": 0.0, "completion_length": 492.1458435058594, "epoch": 0.5080645161290323, "grad_norm": 70.97774177635779, "kl": 1.5078125, "learning_rate": 6.160109055957018e-07, "loss": 0.4593, "reward": 2.029819965362549, "reward_std": 0.3379950374364853, "rewards/accuracy_reward": 0.1041666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.027471921406686306, "rewards/tag_count_reward": 0.9531250298023224, "step": 693 }, { "clip_ratio": 0.0, "completion_length": 541.4166717529297, "epoch": 0.5087976539589443, "grad_norm": 24.216422460189005, "kl": 1.705078125, "learning_rate": 6.148709809008409e-07, "loss": 0.468, "reward": 1.959551990032196, "reward_std": 0.3680807575583458, "rewards/accuracy_reward": 0.08333333395421505, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.024823017418384552, "rewards/tag_count_reward": 0.9218750298023224, "step": 694 }, { "clip_ratio": 0.0, "completion_length": 643.4166870117188, "epoch": 0.5095307917888563, "grad_norm": 36.347168163912684, "kl": 3.21875, "learning_rate": 6.137306309397986e-07, "loss": 0.8839, "reward": 2.162288546562195, "reward_std": 0.46551817655563354, "rewards/accuracy_reward": 0.3125000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03042002022266388, "rewards/tag_count_reward": 0.8802083432674408, "step": 695 }, { "clip_ratio": 0.0, "completion_length": 757.7916870117188, "epoch": 0.5102639296187683, "grad_norm": 51.19725173793361, "kl": 4.4453125, "learning_rate": 6.125898631882174e-07, "loss": 0.8215, "reward": 2.0202183723449707, "reward_std": 0.491269052028656, "rewards/accuracy_reward": 0.1875000074505806, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021448221988976002, "rewards/tag_count_reward": 0.8541666865348816, "step": 696 }, { "clip_ratio": 0.0, "completion_length": 723.6666870117188, "epoch": 0.5109970674486803, "grad_norm": 32.11893685151525, "kl": 3.703125, "learning_rate": 6.114486851244793e-07, "loss": 0.8158, "reward": 2.060223937034607, "reward_std": 0.4704440087080002, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028317784890532494, "rewards/tag_count_reward": 0.8802083730697632, "step": 697 }, { "clip_ratio": 0.0, "completion_length": 857.7916870117188, "epoch": 0.5117302052785924, "grad_norm": 18.784707199764576, "kl": 3.7109375, "learning_rate": 6.103071042296551e-07, "loss": 0.7494, "reward": 1.9297525882720947, "reward_std": 0.38079968094825745, "rewards/accuracy_reward": 0.12500000558793545, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023372489027678967, "rewards/tag_count_reward": 0.828125, "step": 698 }, { "clip_ratio": 0.0, "completion_length": 870.1875305175781, "epoch": 0.5124633431085044, "grad_norm": 14.677437827096641, "kl": 2.234375, "learning_rate": 6.091651279874574e-07, "loss": 0.5725, "reward": 2.004097819328308, "reward_std": 0.5094822645187378, "rewards/accuracy_reward": 0.2708333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.02888854220509529, "rewards/tag_count_reward": 0.7760416865348816, "step": 699 }, { "clip_ratio": 0.0, "completion_length": 828.7708740234375, "epoch": 0.5131964809384164, "grad_norm": 11.821380844349749, "kl": 1.96484375, "learning_rate": 6.0802276388419e-07, "loss": 0.459, "reward": 2.201478123664856, "reward_std": 0.5592115372419357, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.027688627131283283, "rewards/tag_count_reward": 0.8125000298023224, "step": 700 }, { "clip_ratio": 0.0, "completion_length": 871.2500305175781, "epoch": 0.5139296187683284, "grad_norm": 22.568142496900485, "kl": 2.109375, "learning_rate": 6.068800194086989e-07, "loss": 0.7207, "reward": 1.8745365738868713, "reward_std": 0.4898378700017929, "rewards/accuracy_reward": 0.1458333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0265052430331707, "rewards/tag_count_reward": 0.7552083432674408, "step": 701 }, { "clip_ratio": 0.0, "completion_length": 1003.5416870117188, "epoch": 0.5146627565982405, "grad_norm": 28.302531166420113, "kl": 3.484375, "learning_rate": 6.057369020523249e-07, "loss": 0.5456, "reward": 1.9380370378494263, "reward_std": 0.4967404901981354, "rewards/accuracy_reward": 0.3125, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.028976941481232643, "rewards/tag_count_reward": 0.6614583432674408, "step": 702 }, { "clip_ratio": 0.0, "completion_length": 816.2291870117188, "epoch": 0.5153958944281525, "grad_norm": 10.672850559175945, "kl": 2.453125, "learning_rate": 6.045934193088521e-07, "loss": 0.5804, "reward": 2.036511540412903, "reward_std": 0.5523365437984467, "rewards/accuracy_reward": 0.2916666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025988491252064705, "rewards/tag_count_reward": 0.7708333432674408, "step": 703 }, { "clip_ratio": 0.0, "completion_length": 764.3125, "epoch": 0.5161290322580645, "grad_norm": 39.735651963128014, "kl": 2.3515625, "learning_rate": 6.034495786744605e-07, "loss": 0.7015, "reward": 1.9637788534164429, "reward_std": 0.5475030541419983, "rewards/accuracy_reward": 0.2500000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02580451127141714, "rewards/tag_count_reward": 0.7395833432674408, "step": 704 }, { "clip_ratio": 0.0, "completion_length": 895.4166870117188, "epoch": 0.5168621700879765, "grad_norm": 27.423399312980752, "kl": 1.82421875, "learning_rate": 6.023053876476761e-07, "loss": 0.7014, "reward": 1.7406333684921265, "reward_std": 0.5075753778219223, "rewards/accuracy_reward": 0.0833333358168602, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.026727747172117233, "rewards/tag_count_reward": 0.6979166865348816, "step": 705 }, { "clip_ratio": 0.0, "completion_length": 887.8125, "epoch": 0.5175953079178885, "grad_norm": 21.410467041793627, "kl": 2.12109375, "learning_rate": 6.01160853729322e-07, "loss": 0.6424, "reward": 2.2163281440734863, "reward_std": 0.7610662877559662, "rewards/accuracy_reward": 0.5416666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02325525600463152, "rewards/tag_count_reward": 0.6979166865348816, "step": 706 }, { "clip_ratio": 0.0, "completion_length": 783.6041870117188, "epoch": 0.5183284457478006, "grad_norm": 27.22995745005586, "kl": 1.9765625, "learning_rate": 6.000159844224693e-07, "loss": 0.6752, "reward": 1.9180487990379333, "reward_std": 0.5266700834035873, "rewards/accuracy_reward": 0.229166679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.029867863282561302, "rewards/tag_count_reward": 0.7187500298023224, "step": 707 }, { "clip_ratio": 0.0, "completion_length": 804.5000305175781, "epoch": 0.5190615835777126, "grad_norm": 35.4164868878839, "kl": 2.60546875, "learning_rate": 5.988707872323869e-07, "loss": 0.6535, "reward": 2.0066672563552856, "reward_std": 0.5585867166519165, "rewards/accuracy_reward": 0.3541666865348816, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.031527137383818626, "rewards/tag_count_reward": 0.6979166865348816, "step": 708 }, { "clip_ratio": 0.0, "completion_length": 681.2500305175781, "epoch": 0.5197947214076246, "grad_norm": 42.58931418473554, "kl": 2.08203125, "learning_rate": 5.977252696664944e-07, "loss": 0.4843, "reward": 2.1872243881225586, "reward_std": 0.6637312173843384, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03152578230947256, "rewards/tag_count_reward": 0.7812500298023224, "step": 709 }, { "clip_ratio": 0.0, "completion_length": 786.3541870117188, "epoch": 0.5205278592375366, "grad_norm": 935.9645040339218, "kl": 6.93359375, "learning_rate": 5.965794392343109e-07, "loss": 0.8653, "reward": 1.8304345607757568, "reward_std": 0.45126180350780487, "rewards/accuracy_reward": 0.125, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023732122033834457, "rewards/tag_count_reward": 0.7291666865348816, "step": 710 }, { "clip_ratio": 0.0, "completion_length": 814.2083740234375, "epoch": 0.5212609970674487, "grad_norm": 175.06802015741212, "kl": 5.234375, "learning_rate": 5.954333034474068e-07, "loss": 0.9546, "reward": 1.9417012929916382, "reward_std": 0.5177436172962189, "rewards/accuracy_reward": 0.229166679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02704876847565174, "rewards/tag_count_reward": 0.7395833730697632, "step": 711 }, { "clip_ratio": 0.0, "completion_length": 907.2708740234375, "epoch": 0.5219941348973607, "grad_norm": 48.81417891282692, "kl": 3.7734375, "learning_rate": 5.942868698193538e-07, "loss": 0.7592, "reward": 1.8927278518676758, "reward_std": 0.5524491965770721, "rewards/accuracy_reward": 0.2500000074505806, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.029147211462259293, "rewards/tag_count_reward": 0.671875, "step": 712 }, { "clip_ratio": 0.0, "completion_length": 995.7292175292969, "epoch": 0.5227272727272727, "grad_norm": 139.6123070474988, "kl": 6.234375, "learning_rate": 5.931401458656767e-07, "loss": 0.8624, "reward": 1.710960566997528, "reward_std": 0.45395517349243164, "rewards/accuracy_reward": 0.0833333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028622757643461227, "rewards/tag_count_reward": 0.65625, "step": 713 }, { "clip_ratio": 0.0, "completion_length": 937.7291870117188, "epoch": 0.5234604105571847, "grad_norm": 71.2076844274725, "kl": 3.390625, "learning_rate": 5.919931391038036e-07, "loss": 0.6294, "reward": 1.9709495306015015, "reward_std": 0.7317883968353271, "rewards/accuracy_reward": 0.3125000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023842180147767067, "rewards/tag_count_reward": 0.6822916865348816, "step": 714 }, { "clip_ratio": 0.0, "completion_length": 904.5208435058594, "epoch": 0.5241935483870968, "grad_norm": 41.114481536946386, "kl": 3.4375, "learning_rate": 5.908458570530159e-07, "loss": 0.7369, "reward": 2.023408830165863, "reward_std": 0.7161271274089813, "rewards/accuracy_reward": 0.3750000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0338828694075346, "rewards/tag_count_reward": 0.6822916865348816, "step": 715 }, { "clip_ratio": 0.0, "completion_length": 909.625, "epoch": 0.5249266862170088, "grad_norm": 84.79032997074457, "kl": 4.078125, "learning_rate": 5.896983072344002e-07, "loss": 0.8869, "reward": 1.843784511089325, "reward_std": 0.5026722997426987, "rewards/accuracy_reward": 0.2083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026007202453911304, "rewards/tag_count_reward": 0.6614583432674408, "step": 716 }, { "clip_ratio": 0.0, "completion_length": 1022.0416870117188, "epoch": 0.5256598240469208, "grad_norm": 192.89777198011595, "kl": 4.53125, "learning_rate": 5.885504971707984e-07, "loss": 0.892, "reward": 1.8601387739181519, "reward_std": 0.5931964814662933, "rewards/accuracy_reward": 0.2916666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030486296862363815, "rewards/tag_count_reward": 0.5989583432674408, "step": 717 }, { "clip_ratio": 0.0, "completion_length": 1011.5625305175781, "epoch": 0.5263929618768328, "grad_norm": 84.5201492882185, "kl": 3.578125, "learning_rate": 5.874024343867583e-07, "loss": 0.8796, "reward": 1.7078853249549866, "reward_std": 0.5262158215045929, "rewards/accuracy_reward": 0.1041666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021281354129314423, "rewards/tag_count_reward": 0.625, "step": 718 }, { "clip_ratio": 0.0, "completion_length": 866.7291870117188, "epoch": 0.5271260997067448, "grad_norm": 43.25500266243787, "kl": 2.0234375, "learning_rate": 5.862541264084846e-07, "loss": 0.7905, "reward": 1.9441114664077759, "reward_std": 0.5518654584884644, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.029846942983567715, "rewards/tag_count_reward": 0.7239583730697632, "step": 719 }, { "clip_ratio": 0.0, "completion_length": 1271.8333740234375, "epoch": 0.5278592375366569, "grad_norm": 70.03986882968586, "kl": 4.6640625, "learning_rate": 5.851055807637892e-07, "loss": 0.7332, "reward": 1.7508472800254822, "reward_std": 0.6809434294700623, "rewards/accuracy_reward": 0.2916666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030402755364775658, "rewards/tag_count_reward": 0.4895833432674408, "step": 720 }, { "clip_ratio": 0.0, "completion_length": 887.2708740234375, "epoch": 0.5285923753665689, "grad_norm": 58.680154588359265, "kl": 1.609375, "learning_rate": 5.839568049820423e-07, "loss": 0.5473, "reward": 2.2397154569625854, "reward_std": 0.7437396347522736, "rewards/accuracy_reward": 0.5833333730697632, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025909623131155968, "rewards/tag_count_reward": 0.6822916865348816, "step": 721 }, { "clip_ratio": 0.0, "completion_length": 953.0417175292969, "epoch": 0.5293255131964809, "grad_norm": 36.23136798556369, "kl": 2.73046875, "learning_rate": 5.828078065941225e-07, "loss": 0.7843, "reward": 1.9861117005348206, "reward_std": 0.7317814528942108, "rewards/accuracy_reward": 0.3541666865348816, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02777726948261261, "rewards/tag_count_reward": 0.6666666865348816, "step": 722 }, { "clip_ratio": 0.0, "completion_length": 1127.5208435058594, "epoch": 0.5300586510263929, "grad_norm": 152.22428686952142, "kl": 8.75, "learning_rate": 5.816585931323679e-07, "loss": 1.1706, "reward": 1.7878764867782593, "reward_std": 0.623430460691452, "rewards/accuracy_reward": 0.2708333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02462371066212654, "rewards/tag_count_reward": 0.5416666716337204, "step": 723 }, { "clip_ratio": 0.0, "completion_length": 1075.0416870117188, "epoch": 0.530791788856305, "grad_norm": 199.94524867359831, "kl": 6.140625, "learning_rate": 5.805091721305264e-07, "loss": 0.941, "reward": 1.703671395778656, "reward_std": 0.5384046882390976, "rewards/accuracy_reward": 0.1458333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03070361725986004, "rewards/tag_count_reward": 0.5885416865348816, "step": 724 }, { "clip_ratio": 0.0, "completion_length": 1210.6250610351562, "epoch": 0.531524926686217, "grad_norm": 31.78135460130777, "kl": 3.453125, "learning_rate": 5.793595511237067e-07, "loss": 0.7312, "reward": 1.6445240378379822, "reward_std": 0.531583845615387, "rewards/accuracy_reward": 0.1458333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022142627276480198, "rewards/tag_count_reward": 0.5208333432674408, "step": 725 }, { "clip_ratio": 0.0, "completion_length": 985.4166870117188, "epoch": 0.532258064516129, "grad_norm": 28.979448973880633, "kl": 2.671875, "learning_rate": 5.782097376483281e-07, "loss": 0.6303, "reward": 2.1739208698272705, "reward_std": 0.731708288192749, "rewards/accuracy_reward": 0.5625000298023224, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02920421026647091, "rewards/tag_count_reward": 0.6406250298023224, "step": 726 }, { "clip_ratio": 0.0, "completion_length": 873.7083435058594, "epoch": 0.532991202346041, "grad_norm": 17.02643874548067, "kl": 2.875, "learning_rate": 5.770597392420724e-07, "loss": 0.7263, "reward": 1.9745637774467468, "reward_std": 0.5388579368591309, "rewards/accuracy_reward": 0.2916666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030644604936242104, "rewards/tag_count_reward": 0.7135416865348816, "step": 727 }, { "clip_ratio": 0.0, "completion_length": 1042.1458740234375, "epoch": 0.533724340175953, "grad_norm": 69.68061124599399, "kl": 5.828125, "learning_rate": 5.759095634438331e-07, "loss": 0.7917, "reward": 1.869464635848999, "reward_std": 0.6467162668704987, "rewards/accuracy_reward": 0.3333333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026368800550699234, "rewards/tag_count_reward": 0.5625000298023224, "step": 728 }, { "clip_ratio": 0.0, "completion_length": 1039.625, "epoch": 0.5344574780058651, "grad_norm": 28.905645179972318, "kl": 4.609375, "learning_rate": 5.747592177936669e-07, "loss": 0.7375, "reward": 1.7492436170578003, "reward_std": 0.6043170094490051, "rewards/accuracy_reward": 0.1875000074505806, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026798022910952568, "rewards/tag_count_reward": 0.5885416716337204, "step": 729 }, { "clip_ratio": 0.0, "completion_length": 927.3958435058594, "epoch": 0.5351906158357771, "grad_norm": 35.675576206435075, "kl": 1.775390625, "learning_rate": 5.736087098327438e-07, "loss": 0.5483, "reward": 2.0073986649513245, "reward_std": 0.4903212785720825, "rewards/accuracy_reward": 0.375, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023851441219449043, "rewards/tag_count_reward": 0.6562500149011612, "step": 730 }, { "clip_ratio": 0.0, "completion_length": 972.8958435058594, "epoch": 0.5359237536656891, "grad_norm": 44.292846595818396, "kl": 1.5859375, "learning_rate": 5.72458047103298e-07, "loss": 0.5784, "reward": 1.8557514548301697, "reward_std": 0.5746296048164368, "rewards/accuracy_reward": 0.2708333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.026192973367869854, "rewards/tag_count_reward": 0.6250000298023224, "step": 731 }, { "clip_ratio": 0.0, "completion_length": 1216.7291870117188, "epoch": 0.5366568914956011, "grad_norm": 31.19157029985504, "kl": 2.7265625, "learning_rate": 5.713072371485783e-07, "loss": 0.6276, "reward": 1.7138049602508545, "reward_std": 0.5935538113117218, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02577844075858593, "rewards/tag_count_reward": 0.4895833432674408, "step": 732 }, { "clip_ratio": 0.0, "completion_length": 1203.354248046875, "epoch": 0.5373900293255132, "grad_norm": 23.572066288373804, "kl": 4.59375, "learning_rate": 5.701562875127986e-07, "loss": 0.7303, "reward": 1.4909549355506897, "reward_std": 0.4130130261182785, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0246700718998909, "rewards/tag_count_reward": 0.4947916865348816, "step": 733 }, { "clip_ratio": 0.0, "completion_length": 1072.75, "epoch": 0.5381231671554252, "grad_norm": 58.6203440113047, "kl": 5.65625, "learning_rate": 5.690052057410885e-07, "loss": 0.8741, "reward": 2.004941701889038, "reward_std": 0.7281341552734375, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03151674382388592, "rewards/tag_count_reward": 0.5781250298023224, "step": 734 }, { "clip_ratio": 0.0, "completion_length": 960.6458740234375, "epoch": 0.5388563049853372, "grad_norm": 15.332845312739705, "kl": 3.2421875, "learning_rate": 5.678539993794435e-07, "loss": 0.6949, "reward": 1.7642380595207214, "reward_std": 0.5163361728191376, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03263694699853659, "rewards/tag_count_reward": 0.6302083432674408, "step": 735 }, { "clip_ratio": 0.0, "completion_length": 861.8958435058594, "epoch": 0.5395894428152492, "grad_norm": 44.73059916771767, "kl": 2.1171875, "learning_rate": 5.667026759746766e-07, "loss": 0.7097, "reward": 2.103455603122711, "reward_std": 0.6455656886100769, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.031961165368556976, "rewards/tag_count_reward": 0.65625, "step": 736 }, { "clip_ratio": 0.0, "completion_length": 839.375, "epoch": 0.5403225806451613, "grad_norm": 41.77144504367488, "kl": 2.1171875, "learning_rate": 5.655512430743674e-07, "loss": 0.6587, "reward": 2.0384910702705383, "reward_std": 0.6447021067142487, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.029217291623353958, "rewards/tag_count_reward": 0.6718750298023224, "step": 737 }, { "clip_ratio": 0.0, "completion_length": 837.0833435058594, "epoch": 0.5410557184750733, "grad_norm": 127.98277053064835, "kl": 5.734375, "learning_rate": 5.643997082268135e-07, "loss": 0.8749, "reward": 1.7961390018463135, "reward_std": 0.5022356361150742, "rewards/accuracy_reward": 0.18750000558793545, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0319860465824604, "rewards/tag_count_reward": 0.6406250298023224, "step": 738 }, { "clip_ratio": 0.0, "completion_length": 792.3958740234375, "epoch": 0.5417888563049853, "grad_norm": 662.9580863707783, "kl": 9.25, "learning_rate": 5.63248078980981e-07, "loss": 1.3764, "reward": 1.779019296169281, "reward_std": 0.45541003346443176, "rewards/accuracy_reward": 0.125, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028272357769310474, "rewards/tag_count_reward": 0.6822916865348816, "step": 739 }, { "clip_ratio": 0.0, "completion_length": 865.1250305175781, "epoch": 0.5425219941348973, "grad_norm": 53.129583456399715, "kl": 5.125, "learning_rate": 5.620963628864547e-07, "loss": 0.8805, "reward": 2.087476968765259, "reward_std": 0.6577587723731995, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02710629813373089, "rewards/tag_count_reward": 0.6770833432674408, "step": 740 }, { "clip_ratio": 0.0, "completion_length": 799.125, "epoch": 0.5432551319648093, "grad_norm": 47.00311282870559, "kl": 2.5234375, "learning_rate": 5.609445674933886e-07, "loss": 0.7153, "reward": 1.8923691511154175, "reward_std": 0.5534811615943909, "rewards/accuracy_reward": 0.2500000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.029505793005228043, "rewards/tag_count_reward": 0.671875, "step": 741 }, { "clip_ratio": 0.0, "completion_length": 1089.6458740234375, "epoch": 0.5439882697947214, "grad_norm": 25.779958061997704, "kl": 3.6015625, "learning_rate": 5.59792700352457e-07, "loss": 0.747, "reward": 1.580857276916504, "reward_std": 0.4260368049144745, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028517894446849823, "rewards/tag_count_reward": 0.567708358168602, "step": 742 }, { "clip_ratio": 0.0, "completion_length": 885.6666870117188, "epoch": 0.5447214076246334, "grad_norm": 26.450864433872862, "kl": 4.078125, "learning_rate": 5.586407690148037e-07, "loss": 0.6532, "reward": 1.9195644855499268, "reward_std": 0.5701983422040939, "rewards/accuracy_reward": 0.3125000111758709, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028352378867566586, "rewards/tag_count_reward": 0.6354166865348816, "step": 743 }, { "clip_ratio": 0.0, "completion_length": 776.0833740234375, "epoch": 0.5454545454545454, "grad_norm": 24.009947817737903, "kl": 3.109375, "learning_rate": 5.574887810319946e-07, "loss": 0.6748, "reward": 1.9898723363876343, "reward_std": 0.6006492376327515, "rewards/accuracy_reward": 0.3541666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03096117079257965, "rewards/tag_count_reward": 0.6666666865348816, "step": 744 }, { "clip_ratio": 0.0, "completion_length": 960.0208740234375, "epoch": 0.5461876832844574, "grad_norm": 33.79328160042349, "kl": 5.625, "learning_rate": 5.563367439559657e-07, "loss": 0.9255, "reward": 1.6592678427696228, "reward_std": 0.4428327679634094, "rewards/accuracy_reward": 0.0833333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028232182376086712, "rewards/tag_count_reward": 0.6041666865348816, "step": 745 }, { "clip_ratio": 0.0, "completion_length": 610.1875305175781, "epoch": 0.5469208211143695, "grad_norm": 25.32495424125435, "kl": 2.6484375, "learning_rate": 5.551846653389754e-07, "loss": 0.6263, "reward": 2.261523962020874, "reward_std": 0.5468275845050812, "rewards/accuracy_reward": 0.5416666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03014293871819973, "rewards/tag_count_reward": 0.75, "step": 746 }, { "clip_ratio": 0.0, "completion_length": 681.7708740234375, "epoch": 0.5476539589442815, "grad_norm": 73.78278585282082, "kl": 5.328125, "learning_rate": 5.540325527335547e-07, "loss": 0.9039, "reward": 1.9347513914108276, "reward_std": 0.47221839427948, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.033998677507042885, "rewards/tag_count_reward": 0.7395833730697632, "step": 747 }, { "clip_ratio": 0.0, "completion_length": 791.4791870117188, "epoch": 0.5483870967741935, "grad_norm": 90.42777038947605, "kl": 6.1640625, "learning_rate": 5.528804136924571e-07, "loss": 0.9191, "reward": 1.9225658774375916, "reward_std": 0.4990091919898987, "rewards/accuracy_reward": 0.2083333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025350838899612427, "rewards/tag_count_reward": 0.7395833730697632, "step": 748 }, { "clip_ratio": 0.0, "completion_length": 874.2291870117188, "epoch": 0.5491202346041055, "grad_norm": 18.450252877965003, "kl": 4.484375, "learning_rate": 5.517282557686091e-07, "loss": 0.696, "reward": 1.7531982064247131, "reward_std": 0.5390603840351105, "rewards/accuracy_reward": 0.1875000111758709, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028051897883415222, "rewards/tag_count_reward": 0.5937500298023224, "step": 749 }, { "clip_ratio": 0.0, "completion_length": 653.0208435058594, "epoch": 0.5498533724340176, "grad_norm": 52.140637292566204, "kl": 1.453125, "learning_rate": 5.505760865150613e-07, "loss": 0.4664, "reward": 2.0003074407577515, "reward_std": 0.5500705242156982, "rewards/accuracy_reward": 0.2916666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03094259649515152, "rewards/tag_count_reward": 0.7395833432674408, "step": 750 }, { "clip_ratio": 0.0, "completion_length": 571.9791870117188, "epoch": 0.5505865102639296, "grad_norm": 38.39356021633937, "kl": 0.94921875, "learning_rate": 5.494239134849388e-07, "loss": 0.3249, "reward": 2.200824737548828, "reward_std": 0.644676610827446, "rewards/accuracy_reward": 0.458333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03355027176439762, "rewards/tag_count_reward": 0.7760416865348816, "step": 751 }, { "clip_ratio": 0.0, "completion_length": 780.4375305175781, "epoch": 0.5513196480938416, "grad_norm": 57.63946506829144, "kl": 1.86328125, "learning_rate": 5.48271744231391e-07, "loss": 0.6424, "reward": 1.7193049192428589, "reward_std": 0.3903712034225464, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030695032328367233, "rewards/tag_count_reward": 0.7083333730697632, "step": 752 }, { "clip_ratio": 0.0, "completion_length": 773.2708435058594, "epoch": 0.5520527859237536, "grad_norm": 46.738838518138174, "kl": 3.81640625, "learning_rate": 5.471195863075431e-07, "loss": 0.7577, "reward": 1.955974280834198, "reward_std": 0.6104044318199158, "rewards/accuracy_reward": 0.3125000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028400634415447712, "rewards/tag_count_reward": 0.671875, "step": 753 }, { "clip_ratio": 0.0, "completion_length": 757.1458435058594, "epoch": 0.5527859237536656, "grad_norm": 70.68302962276599, "kl": 5.609375, "learning_rate": 5.459674472664451e-07, "loss": 0.7409, "reward": 1.9917755722999573, "reward_std": 0.5736424177885056, "rewards/accuracy_reward": 0.3333333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02905783336609602, "rewards/tag_count_reward": 0.6875, "step": 754 }, { "clip_ratio": 0.0, "completion_length": 824.3333740234375, "epoch": 0.5535190615835777, "grad_norm": 57.39574382552266, "kl": 5.40625, "learning_rate": 5.448153346610246e-07, "loss": 0.8608, "reward": 1.7963295578956604, "reward_std": 0.5229825377464294, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0265871724113822, "rewards/tag_count_reward": 0.6354166865348816, "step": 755 }, { "clip_ratio": 0.0, "completion_length": 594.6250305175781, "epoch": 0.5542521994134897, "grad_norm": 16.480205960183262, "kl": 2.0546875, "learning_rate": 5.436632560440343e-07, "loss": 0.2668, "reward": 1.8957591652870178, "reward_std": 0.5015785098075867, "rewards/accuracy_reward": 0.2083333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03132414259016514, "rewards/tag_count_reward": 0.7187500298023224, "step": 756 }, { "clip_ratio": 0.0, "completion_length": 718.8333435058594, "epoch": 0.5549853372434017, "grad_norm": 35.6619139947997, "kl": 2.5, "learning_rate": 5.425112189680056e-07, "loss": 0.6965, "reward": 1.8117918968200684, "reward_std": 0.4706086814403534, "rewards/accuracy_reward": 0.1458333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026749805547297, "rewards/tag_count_reward": 0.6927083730697632, "step": 757 }, { "clip_ratio": 0.0, "completion_length": 725.7500305175781, "epoch": 0.5557184750733137, "grad_norm": 23.569771815446728, "kl": 3.6953125, "learning_rate": 5.413592309851963e-07, "loss": 0.6908, "reward": 1.898435115814209, "reward_std": 0.5756514668464661, "rewards/accuracy_reward": 0.2708333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03906491957604885, "rewards/tag_count_reward": 0.6666666865348816, "step": 758 }, { "clip_ratio": 0.0, "completion_length": 788.1875, "epoch": 0.5564516129032258, "grad_norm": 10.635869330106727, "kl": 3.6171875, "learning_rate": 5.402072996475432e-07, "loss": 0.687, "reward": 2.104902744293213, "reward_std": 0.6718497276306152, "rewards/accuracy_reward": 0.4375000298023224, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03051395993679762, "rewards/tag_count_reward": 0.6979166865348816, "step": 759 }, { "clip_ratio": 0.0, "completion_length": 599.7916870117188, "epoch": 0.5571847507331378, "grad_norm": 7.951630673232543, "kl": 1.93359375, "learning_rate": 5.390554325066114e-07, "loss": 0.3941, "reward": 1.937668800354004, "reward_std": 0.4101848602294922, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.031081256456673145, "rewards/tag_count_reward": 0.8020833432674408, "step": 760 }, { "clip_ratio": 0.0, "completion_length": 652.4791870117188, "epoch": 0.5579178885630498, "grad_norm": 18.25749794578066, "kl": 3.859375, "learning_rate": 5.379036371135453e-07, "loss": 0.6928, "reward": 1.7474979162216187, "reward_std": 0.34113722294569016, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02854372374713421, "rewards/tag_count_reward": 0.7760416865348816, "step": 761 }, { "clip_ratio": 0.0, "completion_length": 562.3750305175781, "epoch": 0.5586510263929618, "grad_norm": 11.93049781932339, "kl": 2.515625, "learning_rate": 5.367519210190191e-07, "loss": 0.5786, "reward": 2.156864047050476, "reward_std": 0.6391881704330444, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030635923147201538, "rewards/tag_count_reward": 0.7708333730697632, "step": 762 }, { "clip_ratio": 0.0, "completion_length": 582.7708435058594, "epoch": 0.5593841642228738, "grad_norm": 24.399553709161914, "kl": 2.58203125, "learning_rate": 5.356002917731864e-07, "loss": 0.5849, "reward": 1.8041017055511475, "reward_std": 0.3177919387817383, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.034439937211573124, "rewards/tag_count_reward": 0.7968750298023224, "step": 763 }, { "clip_ratio": 0.0, "completion_length": 643.2083435058594, "epoch": 0.5601173020527859, "grad_norm": 9.855696397433391, "kl": 2.3203125, "learning_rate": 5.344487569256327e-07, "loss": 0.4902, "reward": 2.1870139837265015, "reward_std": 0.6778040528297424, "rewards/accuracy_reward": 0.4375000298023224, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03694435581564903, "rewards/tag_count_reward": 0.7864583432674408, "step": 764 }, { "clip_ratio": 0.0, "completion_length": 735.0833740234375, "epoch": 0.5608504398826979, "grad_norm": 9.493175966711302, "kl": 3.28125, "learning_rate": 5.332973240253234e-07, "loss": 0.617, "reward": 2.359973430633545, "reward_std": 0.752300500869751, "rewards/accuracy_reward": 0.6458333730697632, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025443391874432564, "rewards/tag_count_reward": 0.7395833432674408, "step": 765 }, { "clip_ratio": 0.0, "completion_length": 662.4791870117188, "epoch": 0.5615835777126099, "grad_norm": 9.25052938293034, "kl": 3.3359375, "learning_rate": 5.321460006205565e-07, "loss": 0.6242, "reward": 1.9359582662582397, "reward_std": 0.5609914809465408, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03800009563565254, "rewards/tag_count_reward": 0.7239583432674408, "step": 766 }, { "clip_ratio": 0.0, "completion_length": 582.6250305175781, "epoch": 0.5623167155425219, "grad_norm": 19.547089040558244, "kl": 1.88671875, "learning_rate": 5.309947942589114e-07, "loss": 0.5295, "reward": 2.224452018737793, "reward_std": 0.6053149402141571, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0359647162258625, "rewards/tag_count_reward": 0.8229166865348816, "step": 767 }, { "clip_ratio": 0.0, "completion_length": 707.3125305175781, "epoch": 0.5630498533724341, "grad_norm": 49.41788880303197, "kl": 5.1953125, "learning_rate": 5.298437124872013e-07, "loss": 0.8048, "reward": 1.7083222270011902, "reward_std": 0.3275119215250015, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04167785868048668, "rewards/tag_count_reward": 0.75, "step": 768 }, { "clip_ratio": 0.0, "completion_length": 491.16668701171875, "epoch": 0.5637829912023461, "grad_norm": 10.00552596434989, "kl": 1.140625, "learning_rate": 5.286927628514217e-07, "loss": 0.178, "reward": 2.171727418899536, "reward_std": 0.43328505754470825, "rewards/accuracy_reward": 0.2916666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03139766305685043, "rewards/tag_count_reward": 0.9114583730697632, "step": 769 }, { "clip_ratio": 0.0, "completion_length": 474.1458435058594, "epoch": 0.5645161290322581, "grad_norm": 15.644106947033833, "kl": 1.78515625, "learning_rate": 5.27541952896702e-07, "loss": 0.3509, "reward": 2.035298228263855, "reward_std": 0.40297262370586395, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0324101448059082, "rewards/tag_count_reward": 0.8802083730697632, "step": 770 }, { "clip_ratio": 0.0, "completion_length": 612.3125, "epoch": 0.5652492668621701, "grad_norm": 74.13652540498771, "kl": 4.0859375, "learning_rate": 5.263912901672561e-07, "loss": 0.7203, "reward": 2.2360010147094727, "reward_std": 0.5515131950378418, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.034832315519452095, "rewards/tag_count_reward": 0.8541666865348816, "step": 771 }, { "clip_ratio": 0.0, "completion_length": 493.5833435058594, "epoch": 0.5659824046920822, "grad_norm": 8.6360252044363, "kl": 1.291015625, "learning_rate": 5.252407822063332e-07, "loss": 0.199, "reward": 1.9898422360420227, "reward_std": 0.3721628934144974, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025782877579331398, "rewards/tag_count_reward": 0.8281250298023224, "step": 772 }, { "clip_ratio": 0.0, "completion_length": 510.79168701171875, "epoch": 0.5667155425219942, "grad_norm": 22.245313397026667, "kl": 0.763671875, "learning_rate": 5.240904365561669e-07, "loss": 0.1329, "reward": 2.130028009414673, "reward_std": 0.4469859153032303, "rewards/accuracy_reward": 0.3125, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03143035527318716, "rewards/tag_count_reward": 0.8489583730697632, "step": 773 }, { "clip_ratio": 0.0, "completion_length": 542.0833435058594, "epoch": 0.5674486803519062, "grad_norm": 16.58377376399728, "kl": 0.9140625, "learning_rate": 5.229402607579277e-07, "loss": 0.1466, "reward": 1.8467799425125122, "reward_std": 0.4250985085964203, "rewards/accuracy_reward": 0.1041666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.03863669000566006, "rewards/tag_count_reward": 0.8020833730697632, "step": 774 }, { "clip_ratio": 0.0, "completion_length": 500.2708435058594, "epoch": 0.5681818181818182, "grad_norm": 6.383368310310399, "kl": 0.599609375, "learning_rate": 5.217902623516717e-07, "loss": 0.0636, "reward": 2.2422069311141968, "reward_std": 0.6459289789199829, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03383495286107063, "rewards/tag_count_reward": 0.8385416865348816, "step": 775 }, { "clip_ratio": 0.0, "completion_length": 436.93751525878906, "epoch": 0.5689149560117303, "grad_norm": 6.21109163951882, "kl": 0.765625, "learning_rate": 5.206404488762933e-07, "loss": 0.1408, "reward": 2.2062381505966187, "reward_std": 0.4520144984126091, "rewards/accuracy_reward": 0.35416667722165585, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04376194626092911, "rewards/tag_count_reward": 0.8958333432674408, "step": 776 }, { "clip_ratio": 0.0, "completion_length": 470.5, "epoch": 0.5696480938416423, "grad_norm": 11.089411677156567, "kl": 0.81640625, "learning_rate": 5.194908278694736e-07, "loss": 0.0762, "reward": 2.4853463172912598, "reward_std": 0.4210961163043976, "rewards/accuracy_reward": 0.6041666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.035487134009599686, "rewards/tag_count_reward": 0.9166666865348816, "step": 777 }, { "clip_ratio": 0.0, "completion_length": 489.50001525878906, "epoch": 0.5703812316715543, "grad_norm": 5.649480356582214, "kl": 1.078125, "learning_rate": 5.183414068676321e-07, "loss": 0.1901, "reward": 2.011263608932495, "reward_std": 0.28793443739414215, "rewards/accuracy_reward": 0.125, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03561144880950451, "rewards/tag_count_reward": 0.9218750298023224, "step": 778 }, { "clip_ratio": 0.0, "completion_length": 584.0833435058594, "epoch": 0.5711143695014663, "grad_norm": 45.52300400424727, "kl": 3.40625, "learning_rate": 5.171921934058775e-07, "loss": 0.2789, "reward": 2.1760915517807007, "reward_std": 0.6365633606910706, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.05307510495185852, "rewards/tag_count_reward": 0.8333333432674408, "step": 779 }, { "clip_ratio": 0.0, "completion_length": 585.8750305175781, "epoch": 0.5718475073313783, "grad_norm": 18.610581583071962, "kl": 2.71875, "learning_rate": 5.160431950179577e-07, "loss": 0.5859, "reward": 2.0504915714263916, "reward_std": 0.4479319751262665, "rewards/accuracy_reward": 0.2291666679084301, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.055411264300346375, "rewards/tag_count_reward": 0.8906250298023224, "step": 780 }, { "clip_ratio": 0.0, "completion_length": 434.5833435058594, "epoch": 0.5725806451612904, "grad_norm": 11.952290754347626, "kl": 1.0283203125, "learning_rate": 5.148944192362108e-07, "loss": 0.1858, "reward": 2.307385563850403, "reward_std": 0.5844225138425827, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03636452369391918, "rewards/tag_count_reward": 0.8854166865348816, "step": 781 }, { "clip_ratio": 0.0, "completion_length": 552.1041870117188, "epoch": 0.5733137829912024, "grad_norm": 35.347459858014, "kl": 0.90234375, "learning_rate": 5.137458735915154e-07, "loss": 0.3501, "reward": 2.0038501024246216, "reward_std": 0.3343339115381241, "rewards/accuracy_reward": 0.1458333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.05864992365241051, "rewards/tag_count_reward": 0.9166666865348816, "step": 782 }, { "clip_ratio": 0.0, "completion_length": 568.875, "epoch": 0.5740469208211144, "grad_norm": 12.297175402440889, "kl": 1.1396484375, "learning_rate": 5.125975656132417e-07, "loss": 0.2126, "reward": 2.012514591217041, "reward_std": 0.3432723581790924, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04998534545302391, "rewards/tag_count_reward": 0.8958333730697632, "step": 783 }, { "clip_ratio": 0.0, "completion_length": 481.79168701171875, "epoch": 0.5747800586510264, "grad_norm": 10.558960266569809, "kl": 0.802734375, "learning_rate": 5.114495028292016e-07, "loss": 0.2242, "reward": 2.0655742287635803, "reward_std": 0.28712810575962067, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03859252668917179, "rewards/tag_count_reward": 0.9375000298023224, "step": 784 }, { "clip_ratio": 0.0, "completion_length": 488.2708435058594, "epoch": 0.5755131964809385, "grad_norm": 2.6200652288087567, "kl": 0.3642578125, "learning_rate": 5.103016927655997e-07, "loss": 0.0087, "reward": 1.981192708015442, "reward_std": 0.18765612691640854, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024015757255256176, "rewards/tag_count_reward": 0.9635416865348816, "step": 785 }, { "clip_ratio": 0.0, "completion_length": 499.41668701171875, "epoch": 0.5762463343108505, "grad_norm": 3.1631841093704973, "kl": 0.3896484375, "learning_rate": 5.09154142946984e-07, "loss": 0.0069, "reward": 2.364975333213806, "reward_std": 0.23586179316043854, "rewards/accuracy_reward": 0.4166666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025649672374129295, "rewards/tag_count_reward": 0.9739583432674408, "step": 786 }, { "clip_ratio": 0.0, "completion_length": 505.2083435058594, "epoch": 0.5769794721407625, "grad_norm": 143.08815161211146, "kl": 3.056640625, "learning_rate": 5.080068608961965e-07, "loss": 0.3752, "reward": 2.5566176176071167, "reward_std": 0.5731297731399536, "rewards/accuracy_reward": 0.6875000298023224, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.047549253329634666, "rewards/tag_count_reward": 0.9166666865348816, "step": 787 }, { "clip_ratio": 0.0, "completion_length": 460.0416717529297, "epoch": 0.5777126099706745, "grad_norm": 4.206661576025536, "kl": 0.4052734375, "learning_rate": 5.068598541343234e-07, "loss": 0.0059, "reward": 2.296063780784607, "reward_std": 0.33597442507743835, "rewards/accuracy_reward": 0.3750000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.032061291858553886, "rewards/tag_count_reward": 0.9531250298023224, "step": 788 }, { "clip_ratio": 0.0, "completion_length": 546.4791870117188, "epoch": 0.5784457478005866, "grad_norm": 120.55140185532244, "kl": 3.2978515625, "learning_rate": 5.057131301806463e-07, "loss": 0.4419, "reward": 2.227821409702301, "reward_std": 0.4409473240375519, "rewards/accuracy_reward": 0.3333333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.037803635001182556, "rewards/tag_count_reward": 0.9322916865348816, "step": 789 }, { "clip_ratio": 0.0, "completion_length": 483.54168701171875, "epoch": 0.5791788856304986, "grad_norm": 104.00715130995349, "kl": 3.8349609375, "learning_rate": 5.045666965525934e-07, "loss": 0.4773, "reward": 1.9311189651489258, "reward_std": 0.2578328549861908, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.053256092593073845, "rewards/tag_count_reward": 0.9427083730697632, "step": 790 }, { "clip_ratio": 0.0, "completion_length": 516.5416870117188, "epoch": 0.5799120234604106, "grad_norm": 63.31064118479974, "kl": 2.330078125, "learning_rate": 5.034205607656892e-07, "loss": 0.4197, "reward": 2.354369640350342, "reward_std": 0.4720756560564041, "rewards/accuracy_reward": 0.458333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04667205922305584, "rewards/tag_count_reward": 0.9427083730697632, "step": 791 }, { "clip_ratio": 0.0, "completion_length": 460.7291717529297, "epoch": 0.5806451612903226, "grad_norm": 2.3908286014334084, "kl": 0.3271484375, "learning_rate": 5.022747303335056e-07, "loss": -0.0011, "reward": 2.4147515296936035, "reward_std": 0.2904389202594757, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04358202964067459, "rewards/tag_count_reward": 0.9791666865348816, "step": 792 }, { "clip_ratio": 0.0, "completion_length": 441.93751525878906, "epoch": 0.5813782991202346, "grad_norm": 3.227299361726288, "kl": 0.3525390625, "learning_rate": 5.01129212767613e-07, "loss": 0.0012, "reward": 2.1776065826416016, "reward_std": 0.25459666550159454, "rewards/accuracy_reward": 0.2291666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03593523986637592, "rewards/tag_count_reward": 0.9843750298023224, "step": 793 }, { "clip_ratio": 0.0, "completion_length": 483.5208435058594, "epoch": 0.5821114369501467, "grad_norm": 14.106380781519254, "kl": 0.423828125, "learning_rate": 4.999840155775309e-07, "loss": 0.1497, "reward": 2.129902422428131, "reward_std": 0.27987994998693466, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03676432091742754, "rewards/tag_count_reward": 0.9791666865348816, "step": 794 }, { "clip_ratio": 0.0, "completion_length": 462.1875, "epoch": 0.5828445747800587, "grad_norm": 2.6873106735549945, "kl": 0.333984375, "learning_rate": 4.988391462706778e-07, "loss": -0.0256, "reward": 2.383494257926941, "reward_std": 0.4004615694284439, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04358913190662861, "rewards/tag_count_reward": 0.96875, "step": 795 }, { "clip_ratio": 0.0, "completion_length": 512.0833587646484, "epoch": 0.5835777126099707, "grad_norm": 61.545096621591924, "kl": 0.833984375, "learning_rate": 4.976946123523238e-07, "loss": 0.3688, "reward": 1.9309577345848083, "reward_std": 0.21935289353132248, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03779237158596516, "rewards/tag_count_reward": 0.9479166865348816, "step": 796 }, { "clip_ratio": 0.0, "completion_length": 525.0833587646484, "epoch": 0.5843108504398827, "grad_norm": 22.68360336288523, "kl": 0.583984375, "learning_rate": 4.965504213255394e-07, "loss": 0.1888, "reward": 2.5947617292404175, "reward_std": 0.2748841196298599, "rewards/accuracy_reward": 0.6458333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0302383815869689, "rewards/tag_count_reward": 0.9791666865348816, "step": 797 }, { "clip_ratio": 0.0, "completion_length": 513.7916717529297, "epoch": 0.5850439882697948, "grad_norm": 10.564895085544316, "kl": 0.615234375, "learning_rate": 4.954065806911479e-07, "loss": 0.1524, "reward": 2.284646213054657, "reward_std": 0.31503356993198395, "rewards/accuracy_reward": 0.3541666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04347882606089115, "rewards/tag_count_reward": 0.9739583432674408, "step": 798 }, { "clip_ratio": 0.0, "completion_length": 531.8333435058594, "epoch": 0.5857771260997068, "grad_norm": 2.2790864512290345, "kl": 0.2880859375, "learning_rate": 4.942630979476751e-07, "loss": -0.002, "reward": 2.2442930936813354, "reward_std": 0.29676416516304016, "rewards/accuracy_reward": 0.2916666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03695691470056772, "rewards/tag_count_reward": 0.9895833432674408, "step": 799 }, { "clip_ratio": 0.0, "completion_length": 511.06251525878906, "epoch": 0.5865102639296188, "grad_norm": 6.579082854864219, "kl": 0.8193359375, "learning_rate": 4.931199805913011e-07, "loss": 0.1724, "reward": 2.698206305503845, "reward_std": 0.3876982778310776, "rewards/accuracy_reward": 0.7500000298023224, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.036168841645121574, "rewards/tag_count_reward": 0.984375, "step": 800 }, { "clip_ratio": 0.0, "completion_length": 472.1458435058594, "epoch": 0.5872434017595308, "grad_norm": 2.41829385323641, "kl": 0.3115234375, "learning_rate": 4.919772361158101e-07, "loss": 0.0349, "reward": 2.018336534500122, "reward_std": 0.15515825897455215, "rewards/accuracy_reward": 0.0625, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03374684602022171, "rewards/tag_count_reward": 0.9895833432674408, "step": 801 }, { "clip_ratio": 0.0, "completion_length": 553.0625305175781, "epoch": 0.5879765395894428, "grad_norm": 19.929897025421052, "kl": 0.794921875, "learning_rate": 4.908348720125425e-07, "loss": 0.387, "reward": 2.131339430809021, "reward_std": 0.3124546855688095, "rewards/accuracy_reward": 0.22916667722165585, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.05963288061320782, "rewards/tag_count_reward": 0.96875, "step": 802 }, { "clip_ratio": 0.0, "completion_length": 510.9166717529297, "epoch": 0.5887096774193549, "grad_norm": 2.518606038282831, "kl": 0.3056640625, "learning_rate": 4.896928957703449e-07, "loss": -0.0062, "reward": 2.529616117477417, "reward_std": 0.34028075635433197, "rewards/accuracy_reward": 0.5625000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.027675677090883255, "rewards/tag_count_reward": 0.9947916865348816, "step": 803 }, { "clip_ratio": 0.0, "completion_length": 542.8125152587891, "epoch": 0.5894428152492669, "grad_norm": 2.3826860767082434, "kl": 0.296875, "learning_rate": 4.885513148755207e-07, "loss": 0.0078, "reward": 2.4222702980041504, "reward_std": 0.10937895812094212, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0360631225630641, "rewards/tag_count_reward": 1.0, "step": 804 }, { "clip_ratio": 0.0, "completion_length": 461.9583435058594, "epoch": 0.5901759530791789, "grad_norm": 2.994895720204756, "kl": 0.3349609375, "learning_rate": 4.874101368117825e-07, "loss": 0.0334, "reward": 2.157669723033905, "reward_std": 0.14807479828596115, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03851080499589443, "rewards/tag_count_reward": 0.9947916865348816, "step": 805 }, { "clip_ratio": 0.0, "completion_length": 705.8125, "epoch": 0.5909090909090909, "grad_norm": 6.6598445557207775, "kl": 1.083984375, "learning_rate": 4.862693690602015e-07, "loss": 0.2519, "reward": 1.9881237149238586, "reward_std": 0.6753402352333069, "rewards/accuracy_reward": 0.3125000149011612, "rewards/reasoning_steps_reward": 0.9305556416511536, "rewards/repetition_penalty_reward": -0.1455569975078106, "rewards/tag_count_reward": 0.890625, "step": 806 }, { "clip_ratio": 0.0, "completion_length": 462.1875305175781, "epoch": 0.591642228739003, "grad_norm": 2.535021258369295, "kl": 0.3037109375, "learning_rate": 4.851290190991592e-07, "loss": 0.0369, "reward": 2.1697142124176025, "reward_std": 0.22860877960920334, "rewards/accuracy_reward": 0.2083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03861923888325691, "rewards/tag_count_reward": 1.0, "step": 807 }, { "clip_ratio": 0.0, "completion_length": 528.8541717529297, "epoch": 0.592375366568915, "grad_norm": 8.442938463472068, "kl": 0.548828125, "learning_rate": 4.839890944042982e-07, "loss": 0.4029, "reward": 2.536786675453186, "reward_std": 0.5133970677852631, "rewards/accuracy_reward": 0.6666666865348816, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.07085224241018295, "rewards/tag_count_reward": 0.96875, "step": 808 }, { "clip_ratio": 0.0, "completion_length": 602.8750305175781, "epoch": 0.593108504398827, "grad_norm": 9.598470869048404, "kl": 0.681640625, "learning_rate": 4.828496024484737e-07, "loss": 0.563, "reward": 2.4663429260253906, "reward_std": 0.6941099762916565, "rewards/accuracy_reward": 0.6666666865348816, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.11351823806762695, "rewards/tag_count_reward": 0.9479166865348816, "step": 809 }, { "clip_ratio": 0.0, "completion_length": 578.7916870117188, "epoch": 0.593841642228739, "grad_norm": 6.7043292538402355, "kl": 0.5751953125, "learning_rate": 4.817105507017031e-07, "loss": 0.3073, "reward": 2.1016518473625183, "reward_std": 0.38975587766617537, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.06848715990781784, "rewards/tag_count_reward": 0.96875, "step": 810 }, { "clip_ratio": 0.0, "completion_length": 561.8541870117188, "epoch": 0.594574780058651, "grad_norm": 5.364924944020763, "kl": 0.5087890625, "learning_rate": 4.80571946631119e-07, "loss": 0.2888, "reward": 2.335654377937317, "reward_std": 0.3818696141242981, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.07754014804959297, "rewards/tag_count_reward": 0.9687500298023224, "step": 811 }, { "clip_ratio": 0.0, "completion_length": 629.6041870117188, "epoch": 0.5953079178885631, "grad_norm": 17.22316737485544, "kl": 0.83203125, "learning_rate": 4.794337977009185e-07, "loss": 0.4649, "reward": 2.1872715950012207, "reward_std": 0.6461509764194489, "rewards/accuracy_reward": 0.4166666716337204, "rewards/reasoning_steps_reward": 0.9513889253139496, "rewards/repetition_penalty_reward": -0.11307556927204132, "rewards/tag_count_reward": 0.9322916865348816, "step": 812 }, { "clip_ratio": 0.0, "completion_length": 497.0833435058594, "epoch": 0.5960410557184751, "grad_norm": 4.506049614552083, "kl": 0.392578125, "learning_rate": 4.782961113723155e-07, "loss": 0.0013, "reward": 2.4307708740234375, "reward_std": 0.21797269582748413, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03277082648128271, "rewards/tag_count_reward": 0.9843750298023224, "step": 813 }, { "clip_ratio": 0.0, "completion_length": 501.04168701171875, "epoch": 0.5967741935483871, "grad_norm": 5.653050177842232, "kl": 0.359375, "learning_rate": 4.771588951034908e-07, "loss": 0.0841, "reward": 2.136378526687622, "reward_std": 0.18476275354623795, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04070477373898029, "rewards/tag_count_reward": 0.9895833432674408, "step": 814 }, { "clip_ratio": 0.0, "completion_length": 592.3958435058594, "epoch": 0.5975073313782991, "grad_norm": 8.107659270152181, "kl": 0.677734375, "learning_rate": 4.760221563495442e-07, "loss": 0.4121, "reward": 1.859411895275116, "reward_std": 0.30691710114479065, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.0728798434138298, "rewards/tag_count_reward": 0.953125, "step": 815 }, { "clip_ratio": 0.0, "completion_length": 603.2083587646484, "epoch": 0.5982404692082112, "grad_norm": 8.61441795310822, "kl": 0.8076171875, "learning_rate": 4.7488590256244464e-07, "loss": 0.27, "reward": 2.0655227303504944, "reward_std": 0.2962368158623576, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.06815778650343418, "rewards/tag_count_reward": 0.953125, "step": 816 }, { "clip_ratio": 0.0, "completion_length": 573.8750305175781, "epoch": 0.5989736070381232, "grad_norm": 5.521496682273347, "kl": 0.4189453125, "learning_rate": 4.737501411909823e-07, "loss": 0.1851, "reward": 1.9504121541976929, "reward_std": 0.07043480267748237, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04437953233718872, "rewards/tag_count_reward": 0.9947916865348816, "step": 817 }, { "clip_ratio": 0.0, "completion_length": 634.0208435058594, "epoch": 0.5997067448680352, "grad_norm": 48.31708229348907, "kl": 0.8837890625, "learning_rate": 4.726148796807189e-07, "loss": 0.4391, "reward": 2.511850595474243, "reward_std": 0.4401156660169363, "rewards/accuracy_reward": 0.6666666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.09752444550395012, "rewards/tag_count_reward": 0.9427083432674408, "step": 818 }, { "clip_ratio": 0.0, "completion_length": 600.6041870117188, "epoch": 0.6004398826979472, "grad_norm": 18.48522146272581, "kl": 0.591796875, "learning_rate": 4.7148012547393925e-07, "loss": 0.3583, "reward": 2.0937827825546265, "reward_std": 0.4004402905702591, "rewards/accuracy_reward": 0.22916667722165585, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.07462002336978912, "rewards/tag_count_reward": 0.953125, "step": 819 }, { "clip_ratio": 0.0, "completion_length": 534.0625, "epoch": 0.6011730205278593, "grad_norm": 25.1154484599897, "kl": 0.5087890625, "learning_rate": 4.703458860096029e-07, "loss": 0.3437, "reward": 2.3650914430618286, "reward_std": 0.36587995290756226, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.056783514097332954, "rewards/tag_count_reward": 0.9635416865348816, "step": 820 }, { "clip_ratio": 0.0, "completion_length": 601.9583435058594, "epoch": 0.6019061583577713, "grad_norm": 56.1391698496597, "kl": 0.55859375, "learning_rate": 4.692121687232947e-07, "loss": 0.4509, "reward": 2.358267903327942, "reward_std": 0.37750261276960373, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.07228780351579189, "rewards/tag_count_reward": 0.9583333432674408, "step": 821 }, { "clip_ratio": 0.0, "completion_length": 690.8541870117188, "epoch": 0.6026392961876833, "grad_norm": 77.0977681349766, "kl": 0.80078125, "learning_rate": 4.6807898104717646e-07, "loss": 0.4843, "reward": 2.49636173248291, "reward_std": 0.5223219692707062, "rewards/accuracy_reward": 0.6875, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.09912440925836563, "rewards/tag_count_reward": 0.9218750298023224, "step": 822 }, { "clip_ratio": 0.0, "completion_length": 659.8333740234375, "epoch": 0.6033724340175953, "grad_norm": 90.83371935104837, "kl": 0.78125, "learning_rate": 4.669463304099378e-07, "loss": 0.4246, "reward": 2.3206570148468018, "reward_std": 0.3920210897922516, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.06475986167788506, "rewards/tag_count_reward": 0.9479166865348816, "step": 823 }, { "clip_ratio": 0.0, "completion_length": 557.8958435058594, "epoch": 0.6041055718475073, "grad_norm": 83.46788794993483, "kl": 0.486328125, "learning_rate": 4.658142242367481e-07, "loss": 0.2377, "reward": 1.9662054181098938, "reward_std": 0.24156315624713898, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03900298476219177, "rewards/tag_count_reward": 0.9635416865348816, "step": 824 }, { "clip_ratio": 0.0, "completion_length": 594.4166870117188, "epoch": 0.6048387096774194, "grad_norm": 99.53659085190704, "kl": 0.501953125, "learning_rate": 4.6468266994920725e-07, "loss": 0.3287, "reward": 2.660859227180481, "reward_std": 0.20507730171084404, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.04052967391908169, "rewards/tag_count_reward": 0.9791666865348816, "step": 825 }, { "clip_ratio": 0.0, "completion_length": 628.0625, "epoch": 0.6055718475073314, "grad_norm": 107.50085509256628, "kl": 1.2265625, "learning_rate": 4.6355167496529736e-07, "loss": 0.4649, "reward": 2.8116308450698853, "reward_std": 0.42871353030204773, "rewards/accuracy_reward": 0.8958333730697632, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.037327712401747704, "rewards/tag_count_reward": 0.953125, "step": 826 }, { "clip_ratio": 0.0, "completion_length": 563.1041870117188, "epoch": 0.6063049853372434, "grad_norm": 2.173496981310859, "kl": 0.2783203125, "learning_rate": 4.624212466993338e-07, "loss": 0.0277, "reward": 2.7007949352264404, "reward_std": 0.0959350299090147, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023163380101323128, "rewards/tag_count_reward": 0.9947916865348816, "step": 827 }, { "clip_ratio": 0.0, "completion_length": 584.0833435058594, "epoch": 0.6070381231671554, "grad_norm": 103.29060562834302, "kl": 0.4951171875, "learning_rate": 4.6129139256191694e-07, "loss": 0.221, "reward": 2.6394749879837036, "reward_std": 0.2795245796442032, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03240011818706989, "rewards/tag_count_reward": 0.9635416865348816, "step": 828 }, { "clip_ratio": 0.0, "completion_length": 502.91668701171875, "epoch": 0.6077712609970675, "grad_norm": 39.15109601141074, "kl": 0.3662109375, "learning_rate": 4.6016211995988354e-07, "loss": 0.1692, "reward": 2.188190758228302, "reward_std": 0.11946690082550049, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030559342354536057, "rewards/tag_count_reward": 0.9687500298023224, "step": 829 }, { "clip_ratio": 0.0, "completion_length": 564.4791870117188, "epoch": 0.6085043988269795, "grad_norm": 38.62440550306976, "kl": 0.5986328125, "learning_rate": 4.590334362962576e-07, "loss": 0.1895, "reward": 2.1769083738327026, "reward_std": 0.1629810556769371, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.027952796779572964, "rewards/tag_count_reward": 0.96875, "step": 830 }, { "clip_ratio": 0.0, "completion_length": 565.9791870117188, "epoch": 0.6092375366568915, "grad_norm": 58.04515466935745, "kl": 0.92578125, "learning_rate": 4.579053489702027e-07, "loss": 0.3217, "reward": 1.8493124842643738, "reward_std": 0.23448748886585236, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.04304862767457962, "rewards/tag_count_reward": 0.90625, "step": 831 }, { "clip_ratio": 0.0, "completion_length": 531.1041870117188, "epoch": 0.6099706744868035, "grad_norm": 38.13203675394943, "kl": 0.623046875, "learning_rate": 4.5677786537697305e-07, "loss": 0.1495, "reward": 2.393915832042694, "reward_std": 0.2697073072195053, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.029695474542677402, "rewards/tag_count_reward": 0.9791666865348816, "step": 832 }, { "clip_ratio": 0.0, "completion_length": 600.7916870117188, "epoch": 0.6107038123167156, "grad_norm": 49.182375386745534, "kl": 0.73828125, "learning_rate": 4.5565099290786503e-07, "loss": 0.2963, "reward": 2.4932984113693237, "reward_std": 0.4969503581523895, "rewards/accuracy_reward": 0.6250000298023224, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.032743205316364765, "rewards/tag_count_reward": 0.9010416865348816, "step": 833 }, { "clip_ratio": 0.0, "completion_length": 580.0416870117188, "epoch": 0.6114369501466276, "grad_norm": 41.51590481283137, "kl": 1.5390625, "learning_rate": 4.545247389501683e-07, "loss": 0.2424, "reward": 2.3008153438568115, "reward_std": 0.44770413637161255, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.02904599905014038, "rewards/tag_count_reward": 0.90625, "step": 834 }, { "clip_ratio": 0.0, "completion_length": 542.4791870117188, "epoch": 0.6121700879765396, "grad_norm": 8.246011895143058, "kl": 0.498046875, "learning_rate": 4.533991108871187e-07, "loss": 0.0343, "reward": 2.6834018230438232, "reward_std": 0.13353276997804642, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030139826238155365, "rewards/tag_count_reward": 0.984375, "step": 835 }, { "clip_ratio": 0.0, "completion_length": 482.0, "epoch": 0.6129032258064516, "grad_norm": 6.039794360121051, "kl": 0.6142578125, "learning_rate": 4.5227411609784815e-07, "loss": -0.0113, "reward": 1.8693063855171204, "reward_std": 0.24499638378620148, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.958333432674408, "rewards/repetition_penalty_reward": -0.031735366210341454, "rewards/tag_count_reward": 0.9427083730697632, "step": 836 }, { "clip_ratio": 0.0, "completion_length": 510.4375305175781, "epoch": 0.6136363636363636, "grad_norm": 8.66036724884791, "kl": 0.568359375, "learning_rate": 4.5114976195733765e-07, "loss": 0.0996, "reward": 2.1761062145233154, "reward_std": 0.3079090416431427, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03743559028953314, "rewards/tag_count_reward": 0.9635416865348816, "step": 837 }, { "clip_ratio": 0.0, "completion_length": 580.7083435058594, "epoch": 0.6143695014662757, "grad_norm": 18.181326014032848, "kl": 1.322265625, "learning_rate": 4.50026055836368e-07, "loss": 0.2942, "reward": 2.2245534658432007, "reward_std": 0.4769989103078842, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.025446675717830658, "rewards/tag_count_reward": 0.8750000298023224, "step": 838 }, { "clip_ratio": 0.0, "completion_length": 559.7916870117188, "epoch": 0.6151026392961877, "grad_norm": 19.95718172965576, "kl": 2.0390625, "learning_rate": 4.4890300510147227e-07, "loss": 0.346, "reward": 2.209823966026306, "reward_std": 0.5654549300670624, "rewards/accuracy_reward": 0.41666667722165585, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.034967634826898575, "rewards/tag_count_reward": 0.8489583432674408, "step": 839 }, { "clip_ratio": 0.0, "completion_length": 536.8750305175781, "epoch": 0.6158357771260997, "grad_norm": 17.575865868504714, "kl": 2.046875, "learning_rate": 4.477806171148866e-07, "loss": 0.2612, "reward": 2.2379040718078613, "reward_std": 0.5341173410415649, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.027721069753170013, "rewards/tag_count_reward": 0.8906250298023224, "step": 840 }, { "clip_ratio": 0.0, "completion_length": 578.4791870117188, "epoch": 0.6165689149560117, "grad_norm": 17.666580487382166, "kl": 1.64453125, "learning_rate": 4.4665889923450296e-07, "loss": 0.3535, "reward": 2.071997821331024, "reward_std": 0.34587332606315613, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.0252243559807539, "rewards/tag_count_reward": 0.9166666865348816, "step": 841 }, { "clip_ratio": 0.0, "completion_length": 644.1875305175781, "epoch": 0.6173020527859238, "grad_norm": 20.63550526329548, "kl": 1.85546875, "learning_rate": 4.455378588138199e-07, "loss": 0.3912, "reward": 2.285509705543518, "reward_std": 0.3511117547750473, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.023518214002251625, "rewards/tag_count_reward": 0.8854166865348816, "step": 842 }, { "clip_ratio": 0.0, "completion_length": 666.2916870117188, "epoch": 0.6180351906158358, "grad_norm": 36.605797933132436, "kl": 2.40234375, "learning_rate": 4.444175032018953e-07, "loss": 0.4583, "reward": 2.0598318576812744, "reward_std": 0.4984627813100815, "rewards/accuracy_reward": 0.2500000111758709, "rewards/reasoning_steps_reward": 0.9722222089767456, "rewards/repetition_penalty_reward": -0.026973859407007694, "rewards/tag_count_reward": 0.8645833730697632, "step": 843 }, { "clip_ratio": 0.0, "completion_length": 513.4791870117188, "epoch": 0.6187683284457478, "grad_norm": 24.959025179156438, "kl": 1.7265625, "learning_rate": 4.4329783974329725e-07, "loss": 0.2129, "reward": 2.3399888277053833, "reward_std": 0.34585636854171753, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.024594685062766075, "rewards/tag_count_reward": 0.9479166865348816, "step": 844 }, { "clip_ratio": 0.0, "completion_length": 592.8333740234375, "epoch": 0.6195014662756598, "grad_norm": 69.5951941998575, "kl": 4.34375, "learning_rate": 4.4217887577805715e-07, "loss": 0.2437, "reward": 2.0339152812957764, "reward_std": 0.4007200002670288, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.944444477558136, "rewards/repetition_penalty_reward": -0.025112648494541645, "rewards/tag_count_reward": 0.9270833432674408, "step": 845 }, { "clip_ratio": 0.0, "completion_length": 515.6041870117188, "epoch": 0.6202346041055719, "grad_norm": 5.858765632804924, "kl": 0.4033203125, "learning_rate": 4.410606186416196e-07, "loss": 0.0125, "reward": 2.3807249665260315, "reward_std": 0.3000316321849823, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.018580769188702106, "rewards/tag_count_reward": 0.9687500298023224, "step": 846 }, { "clip_ratio": 0.0, "completion_length": 521.2291717529297, "epoch": 0.6209677419354839, "grad_norm": 4.655844732233269, "kl": 0.619140625, "learning_rate": 4.399430756647967e-07, "loss": -0.0073, "reward": 2.2223470211029053, "reward_std": 0.12450448423624039, "rewards/accuracy_reward": 0.27083333395421505, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.024180795066058636, "rewards/tag_count_reward": 0.9895833730697632, "step": 847 }, { "clip_ratio": 0.0, "completion_length": 549.2083740234375, "epoch": 0.6217008797653959, "grad_norm": 52.630109847835854, "kl": 1.5185546875, "learning_rate": 4.3882625417371845e-07, "loss": 0.2387, "reward": 2.3624809980392456, "reward_std": 0.3020096570253372, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03335241973400116, "rewards/tag_count_reward": 0.9583333730697632, "step": 848 }, { "clip_ratio": 0.0, "completion_length": 605.5833435058594, "epoch": 0.6224340175953079, "grad_norm": 28.526409416857014, "kl": 2.85546875, "learning_rate": 4.377101614897846e-07, "loss": 0.352, "reward": 2.195107579231262, "reward_std": 0.5218371152877808, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9652778506278992, "rewards/repetition_penalty_reward": -0.03579512611031532, "rewards/tag_count_reward": 0.8697916865348816, "step": 849 }, { "clip_ratio": 0.0, "completion_length": 577.4166870117188, "epoch": 0.6231671554252199, "grad_norm": 37.260587944631546, "kl": 0.650390625, "learning_rate": 4.36594804929618e-07, "loss": 0.2685, "reward": 2.1056848764419556, "reward_std": 0.3163499981164932, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.034940145909786224, "rewards/tag_count_reward": 0.9322916865348816, "step": 850 }, { "clip_ratio": 0.0, "completion_length": 567.8125305175781, "epoch": 0.623900293255132, "grad_norm": 33.15755307793014, "kl": 0.625, "learning_rate": 4.3548019180501507e-07, "loss": 0.2817, "reward": 2.364286184310913, "reward_std": 0.4741703271865845, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.033283334225416183, "rewards/tag_count_reward": 0.9322916865348816, "step": 851 }, { "clip_ratio": 0.0, "completion_length": 598.2708435058594, "epoch": 0.624633431085044, "grad_norm": 36.042445713246025, "kl": 1.0625, "learning_rate": 4.3436632942289906e-07, "loss": 0.3447, "reward": 2.3856871128082275, "reward_std": 0.2271340936422348, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02577120065689087, "rewards/tag_count_reward": 0.9322916865348816, "step": 852 }, { "clip_ratio": 0.0, "completion_length": 653.6250305175781, "epoch": 0.625366568914956, "grad_norm": 23.030611745917525, "kl": 1.63671875, "learning_rate": 4.332532250852714e-07, "loss": 0.2798, "reward": 2.5467569828033447, "reward_std": 0.2622019178234041, "rewards/accuracy_reward": 0.6666666865348816, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.029631933197379112, "rewards/tag_count_reward": 0.9166666865348816, "step": 853 }, { "clip_ratio": 0.0, "completion_length": 552.1250152587891, "epoch": 0.626099706744868, "grad_norm": 45.62070307646606, "kl": 1.939453125, "learning_rate": 4.3214088608916434e-07, "loss": 0.3685, "reward": 2.087291955947876, "reward_std": 0.2783538084477186, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.029027479700744152, "rewards/tag_count_reward": 0.9427083432674408, "step": 854 }, { "clip_ratio": 0.0, "completion_length": 482.3541717529297, "epoch": 0.6268328445747801, "grad_norm": 2.7971623845184217, "kl": 0.3408203125, "learning_rate": 4.310293197265925e-07, "loss": 0.028, "reward": 2.187386155128479, "reward_std": 0.08395049534738064, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.04178059101104736, "rewards/tag_count_reward": 0.9791666865348816, "step": 855 }, { "clip_ratio": 0.0, "completion_length": 597.6875305175781, "epoch": 0.6275659824046921, "grad_norm": 27.38210025221624, "kl": 1.9375, "learning_rate": 4.2991853328450614e-07, "loss": 0.4026, "reward": 2.115206718444824, "reward_std": 0.2261986956000328, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03409888409078121, "rewards/tag_count_reward": 0.9062500298023224, "step": 856 }, { "clip_ratio": 0.0, "completion_length": 473.4791717529297, "epoch": 0.6282991202346041, "grad_norm": 2.6566176387216998, "kl": 0.322265625, "learning_rate": 4.288085340447416e-07, "loss": 0.0651, "reward": 2.2131537199020386, "reward_std": 0.01848464645445347, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0368463434278965, "rewards/tag_count_reward": 1.0, "step": 857 }, { "clip_ratio": 0.0, "completion_length": 565.0833435058594, "epoch": 0.6290322580645161, "grad_norm": 79.93200932269248, "kl": 4.1328125, "learning_rate": 4.276993292839761e-07, "loss": 0.468, "reward": 2.1203547716140747, "reward_std": 0.47893695533275604, "rewards/accuracy_reward": 0.3541666716337204, "rewards/reasoning_steps_reward": 0.9305555522441864, "rewards/repetition_penalty_reward": -0.02374246809631586, "rewards/tag_count_reward": 0.8593750298023224, "step": 858 }, { "clip_ratio": 0.0, "completion_length": 481.58335876464844, "epoch": 0.6297653958944281, "grad_norm": 26.589906300736143, "kl": 1.98046875, "learning_rate": 4.265909262736771e-07, "loss": 0.1595, "reward": 2.593329668045044, "reward_std": 0.3750048503279686, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.03514259681105614, "rewards/tag_count_reward": 0.9479166865348816, "step": 859 }, { "clip_ratio": 0.0, "completion_length": 503.62501525878906, "epoch": 0.6304985337243402, "grad_norm": 37.42836620968271, "kl": 3.140625, "learning_rate": 4.254833322800574e-07, "loss": 0.308, "reward": 2.2453845143318176, "reward_std": 0.41341114044189453, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.030657252296805382, "rewards/tag_count_reward": 0.9010416865348816, "step": 860 }, { "clip_ratio": 0.0, "completion_length": 572.6666870117188, "epoch": 0.6312316715542522, "grad_norm": 28.722487794274045, "kl": 2.09375, "learning_rate": 4.243765545640251e-07, "loss": 0.383, "reward": 2.0623372197151184, "reward_std": 0.3823268860578537, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.031412893906235695, "rewards/tag_count_reward": 0.8854166865348816, "step": 861 }, { "clip_ratio": 0.0, "completion_length": 557.7708740234375, "epoch": 0.6319648093841642, "grad_norm": 111.3986361153325, "kl": 2.76953125, "learning_rate": 4.2327060038113837e-07, "loss": 0.5271, "reward": 2.1418474912643433, "reward_std": 0.25660426169633865, "rewards/accuracy_reward": 0.27083333395421505, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.033499814569950104, "rewards/tag_count_reward": 0.9322916865348816, "step": 862 }, { "clip_ratio": 0.0, "completion_length": 563.25, "epoch": 0.6326979472140762, "grad_norm": 10.055558041580015, "kl": 1.318359375, "learning_rate": 4.221654769815557e-07, "loss": 0.1609, "reward": 2.0751833319664, "reward_std": 0.3235623463988304, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.030719462782144547, "rewards/tag_count_reward": 0.9322916865348816, "step": 863 }, { "clip_ratio": 0.0, "completion_length": 676.1666870117188, "epoch": 0.6334310850439883, "grad_norm": 19.25924102043533, "kl": 2.25, "learning_rate": 4.210611916099893e-07, "loss": 0.5061, "reward": 2.6473987102508545, "reward_std": 0.6665554046630859, "rewards/accuracy_reward": 0.8125000298023224, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024476418271660805, "rewards/tag_count_reward": 0.8593750298023224, "step": 864 }, { "clip_ratio": 0.0, "completion_length": 508.43751525878906, "epoch": 0.6341642228739003, "grad_norm": 28.250092775048678, "kl": 1.51953125, "learning_rate": 4.199577515056585e-07, "loss": 0.3389, "reward": 2.5335538387298584, "reward_std": 0.5366398096084595, "rewards/accuracy_reward": 0.6458333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028946143575012684, "rewards/tag_count_reward": 0.9166666865348816, "step": 865 }, { "clip_ratio": 0.0, "completion_length": 594.7708435058594, "epoch": 0.6348973607038123, "grad_norm": 26.750223462361088, "kl": 3.7421875, "learning_rate": 4.188551639022403e-07, "loss": 0.6715, "reward": 2.3956226110458374, "reward_std": 0.6546805948019028, "rewards/accuracy_reward": 0.6250000149011612, "rewards/reasoning_steps_reward": 0.9652778506278992, "rewards/repetition_penalty_reward": -0.02278023213148117, "rewards/tag_count_reward": 0.8281250298023224, "step": 866 }, { "clip_ratio": 0.0, "completion_length": 469.95835876464844, "epoch": 0.6356304985337243, "grad_norm": 29.408610351733603, "kl": 2.384765625, "learning_rate": 4.177534360278243e-07, "loss": 0.1665, "reward": 2.386035919189453, "reward_std": 0.1823833561502397, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.965277761220932, "rewards/repetition_penalty_reward": -0.021950202994048595, "rewards/tag_count_reward": 0.9427083432674408, "step": 867 }, { "clip_ratio": 0.0, "completion_length": 521.2083435058594, "epoch": 0.6363636363636364, "grad_norm": 110.36280472120994, "kl": 3.8681640625, "learning_rate": 4.166525751048627e-07, "loss": 0.2498, "reward": 2.2553958892822266, "reward_std": 0.3366401940584183, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.025854194536805153, "rewards/tag_count_reward": 0.9270833432674408, "step": 868 }, { "clip_ratio": 0.0, "completion_length": 476.18751525878906, "epoch": 0.6370967741935484, "grad_norm": 13.18550031145197, "kl": 1.59375, "learning_rate": 4.1555258835012584e-07, "loss": 0.1452, "reward": 2.1661083102226257, "reward_std": 0.2015923522412777, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02486394625157118, "rewards/tag_count_reward": 0.9687500298023224, "step": 869 }, { "clip_ratio": 0.0, "completion_length": 517.2500305175781, "epoch": 0.6378299120234604, "grad_norm": 40.70040525572899, "kl": 2.9375, "learning_rate": 4.14453482974652e-07, "loss": 0.6383, "reward": 2.278277277946472, "reward_std": 0.4681243896484375, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.032486576586961746, "rewards/tag_count_reward": 0.9010416865348816, "step": 870 }, { "clip_ratio": 0.0, "completion_length": 467.29168701171875, "epoch": 0.6385630498533724, "grad_norm": 17.795600069868737, "kl": 3.828125, "learning_rate": 4.133552661837025e-07, "loss": 0.2062, "reward": 2.1841301918029785, "reward_std": 0.5829594433307648, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9444444179534912, "rewards/repetition_penalty_reward": -0.020731104537844658, "rewards/tag_count_reward": 0.8645833432674408, "step": 871 }, { "clip_ratio": 0.0, "completion_length": 542.3125152587891, "epoch": 0.6392961876832844, "grad_norm": 43.26945725564012, "kl": 4.08984375, "learning_rate": 4.1225794517671285e-07, "loss": 0.197, "reward": 2.4382702112197876, "reward_std": 0.2991560027003288, "rewards/accuracy_reward": 0.6041666865348816, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.027007670141756535, "rewards/tag_count_reward": 0.8750000298023224, "step": 872 }, { "clip_ratio": 0.0, "completion_length": 483.6875, "epoch": 0.6400293255131965, "grad_norm": 17.639295704811964, "kl": 2.29296875, "learning_rate": 4.1116152714724674e-07, "loss": 0.1169, "reward": 2.432923674583435, "reward_std": 0.628165602684021, "rewards/accuracy_reward": 0.625, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.028881912119686604, "rewards/tag_count_reward": 0.8645833432674408, "step": 873 }, { "clip_ratio": 0.0, "completion_length": 463.2083435058594, "epoch": 0.6407624633431085, "grad_norm": 29.68998907263042, "kl": 1.32421875, "learning_rate": 4.1006601928294784e-07, "loss": 0.2301, "reward": 2.397642493247986, "reward_std": 0.24895144253969193, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02944089937955141, "rewards/tag_count_reward": 0.9687500298023224, "step": 874 }, { "clip_ratio": 0.0, "completion_length": 566.6250305175781, "epoch": 0.6414956011730205, "grad_norm": 27.368778550673742, "kl": 1.6611328125, "learning_rate": 4.089714287654935e-07, "loss": 0.4332, "reward": 1.908960998058319, "reward_std": 0.1527821570634842, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028539063408970833, "rewards/tag_count_reward": 0.9375, "step": 875 }, { "clip_ratio": 0.0, "completion_length": 501.54168701171875, "epoch": 0.6422287390029325, "grad_norm": 39.697759908513824, "kl": 4.875, "learning_rate": 4.078777627705471e-07, "loss": 0.238, "reward": 2.047454297542572, "reward_std": 0.3919724375009537, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.023726369254291058, "rewards/tag_count_reward": 0.8906250298023224, "step": 876 }, { "clip_ratio": 0.0, "completion_length": 688.1875, "epoch": 0.6429618768328446, "grad_norm": 169.83376657930464, "kl": 7.9609375, "learning_rate": 4.0678502846771144e-07, "loss": 0.8107, "reward": 2.132183790206909, "reward_std": 0.7917422652244568, "rewards/accuracy_reward": 0.479166679084301, "rewards/reasoning_steps_reward": 0.9444445073604584, "rewards/repetition_penalty_reward": -0.02059413492679596, "rewards/tag_count_reward": 0.7291666865348816, "step": 877 }, { "clip_ratio": 0.0, "completion_length": 425.31251525878906, "epoch": 0.6436950146627566, "grad_norm": 12.794387813445342, "kl": 2.21875, "learning_rate": 4.056932330204814e-07, "loss": -0.0479, "reward": 2.2172069549560547, "reward_std": 0.5325516164302826, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9305555820465088, "rewards/repetition_penalty_reward": -0.031056921929121017, "rewards/tag_count_reward": 0.9218750298023224, "step": 878 }, { "clip_ratio": 0.0, "completion_length": 455.5208435058594, "epoch": 0.6444281524926686, "grad_norm": 14.235078049102897, "kl": 2.2890625, "learning_rate": 4.046023835861974e-07, "loss": -0.0001, "reward": 2.0200721621513367, "reward_std": 0.48293229937553406, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9444445371627808, "rewards/repetition_penalty_reward": -0.023330658674240112, "rewards/tag_count_reward": 0.9114583432674408, "step": 879 }, { "clip_ratio": 0.0, "completion_length": 432.5208435058594, "epoch": 0.6451612903225806, "grad_norm": 11.144196187589753, "kl": 2.453125, "learning_rate": 4.0351248731599783e-07, "loss": 0.0756, "reward": 2.54099702835083, "reward_std": 0.31450748443603516, "rewards/accuracy_reward": 0.6666666865348816, "rewards/reasoning_steps_reward": 0.9652778506278992, "rewards/repetition_penalty_reward": -0.02844744734466076, "rewards/tag_count_reward": 0.9375000298023224, "step": 880 }, { "clip_ratio": 0.0, "completion_length": 449.7083435058594, "epoch": 0.6458944281524927, "grad_norm": 131.37077538412728, "kl": 6.78125, "learning_rate": 4.024235513547727e-07, "loss": 0.2691, "reward": 2.0147141218185425, "reward_std": 0.3997432589530945, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9583333134651184, "rewards/repetition_penalty_reward": -0.026952626183629036, "rewards/tag_count_reward": 0.8958333432674408, "step": 881 }, { "clip_ratio": 0.0, "completion_length": 495.33335876464844, "epoch": 0.6466275659824047, "grad_norm": 18.476069092789036, "kl": 1.51953125, "learning_rate": 4.0133558284111667e-07, "loss": 0.3195, "reward": 2.448356866836548, "reward_std": 0.07861550897359848, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030809794552624226, "rewards/tag_count_reward": 0.9791666865348816, "step": 882 }, { "clip_ratio": 0.0, "completion_length": 489.1875, "epoch": 0.6473607038123167, "grad_norm": 69.41799580183432, "kl": 5.953125, "learning_rate": 4.002485889072824e-07, "loss": 0.4949, "reward": 2.1838159561157227, "reward_std": 0.4903464764356613, "rewards/accuracy_reward": 0.3750000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02972586080431938, "rewards/tag_count_reward": 0.8385416865348816, "step": 883 }, { "clip_ratio": 0.0, "completion_length": 465.5833435058594, "epoch": 0.6480938416422287, "grad_norm": 29.130292646666057, "kl": 4.484375, "learning_rate": 3.9916257667913334e-07, "loss": 0.3188, "reward": 2.184706151485443, "reward_std": 0.5841209590435028, "rewards/accuracy_reward": 0.41666667722165585, "rewards/reasoning_steps_reward": 0.9236111342906952, "rewards/repetition_penalty_reward": -0.03057180345058441, "rewards/tag_count_reward": 0.8750000298023224, "step": 884 }, { "clip_ratio": 0.0, "completion_length": 456.8958435058594, "epoch": 0.6488269794721407, "grad_norm": 17.0524891536949, "kl": 2.40234375, "learning_rate": 3.9807755327609726e-07, "loss": 0.045, "reward": 2.5808953046798706, "reward_std": 0.4498814642429352, "rewards/accuracy_reward": 0.6875000149011612, "rewards/reasoning_steps_reward": 0.9722222089767456, "rewards/repetition_penalty_reward": -0.02153550460934639, "rewards/tag_count_reward": 0.9427083730697632, "step": 885 }, { "clip_ratio": 0.0, "completion_length": 457.62501525878906, "epoch": 0.6495601173020528, "grad_norm": 8.964102948355622, "kl": 0.87890625, "learning_rate": 3.969935258111202e-07, "loss": 0.0768, "reward": 2.437723398208618, "reward_std": 0.12438751524314284, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.029290529899299145, "rewards/tag_count_reward": 0.9947916865348816, "step": 886 }, { "clip_ratio": 0.0, "completion_length": 451.10418701171875, "epoch": 0.6502932551319648, "grad_norm": 10.937132516667535, "kl": 2.02734375, "learning_rate": 3.959105013906184e-07, "loss": -0.0086, "reward": 2.470559239387512, "reward_std": 0.5151764750480652, "rewards/accuracy_reward": 0.625, "rewards/reasoning_steps_reward": 0.9583333730697632, "rewards/repetition_penalty_reward": -0.02944083884358406, "rewards/tag_count_reward": 0.9166666865348816, "step": 887 }, { "clip_ratio": 0.0, "completion_length": 449.37501525878906, "epoch": 0.6510263929618768, "grad_norm": 367.04646478880846, "kl": 13.46875, "learning_rate": 3.9482848711443337e-07, "loss": 0.5179, "reward": 2.417901396751404, "reward_std": 0.6228555738925934, "rewards/accuracy_reward": 0.6250000298023224, "rewards/reasoning_steps_reward": 0.944444477558136, "rewards/repetition_penalty_reward": -0.03175152279436588, "rewards/tag_count_reward": 0.8802083730697632, "step": 888 }, { "clip_ratio": 0.0, "completion_length": 485.8125, "epoch": 0.6517595307917888, "grad_norm": 10.951809989922257, "kl": 2.0263671875, "learning_rate": 3.9374749007578387e-07, "loss": 0.1927, "reward": 2.1799603700637817, "reward_std": 0.10181637620553374, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02837303653359413, "rewards/tag_count_reward": 0.9583333432674408, "step": 889 }, { "clip_ratio": 0.0, "completion_length": 482.97918701171875, "epoch": 0.6524926686217009, "grad_norm": 20.50596982482667, "kl": 1.58203125, "learning_rate": 3.926675173612207e-07, "loss": 0.3129, "reward": 2.075802803039551, "reward_std": 0.3111816346645355, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.028363876976072788, "rewards/tag_count_reward": 0.9375000298023224, "step": 890 }, { "clip_ratio": 0.0, "completion_length": 501.0, "epoch": 0.6532258064516129, "grad_norm": 15.292377341650585, "kl": 1.27734375, "learning_rate": 3.9158857605057925e-07, "loss": 0.2264, "reward": 2.1522390842437744, "reward_std": 0.22247625887393951, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.03526082541793585, "rewards/tag_count_reward": 0.9791666865348816, "step": 891 }, { "clip_ratio": 0.0, "completion_length": 433.9583435058594, "epoch": 0.6539589442815249, "grad_norm": 9.261789926032487, "kl": 1.4775390625, "learning_rate": 3.905106732169333e-07, "loss": 0.1089, "reward": 1.9519801139831543, "reward_std": 0.06554302759468555, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.032394807785749435, "rewards/tag_count_reward": 0.984375, "step": 892 }, { "clip_ratio": 0.0, "completion_length": 441.1041717529297, "epoch": 0.6546920821114369, "grad_norm": 30.881304534061154, "kl": 3.4609375, "learning_rate": 3.894338159265496e-07, "loss": -0.0058, "reward": 2.3012085556983948, "reward_std": 0.3849090188741684, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.033860914409160614, "rewards/tag_count_reward": 0.9531250298023224, "step": 893 }, { "clip_ratio": 0.0, "completion_length": 494.2083435058594, "epoch": 0.655425219941349, "grad_norm": 42.785081016691045, "kl": 3.8515625, "learning_rate": 3.883580112388396e-07, "loss": 0.2296, "reward": 2.1021071076393127, "reward_std": 0.2982870936393738, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9583333730697632, "rewards/repetition_penalty_reward": -0.02810133434832096, "rewards/tag_count_reward": 0.9635416865348816, "step": 894 }, { "clip_ratio": 0.0, "completion_length": 478.1458435058594, "epoch": 0.656158357771261, "grad_norm": 21.129811313268846, "kl": 2.3798828125, "learning_rate": 3.872832662063156e-07, "loss": 0.1403, "reward": 2.5664910078048706, "reward_std": 0.23936194740235806, "rewards/accuracy_reward": 0.6666666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02725913841277361, "rewards/tag_count_reward": 0.9270833432674408, "step": 895 }, { "clip_ratio": 0.0, "completion_length": 468.7083435058594, "epoch": 0.656891495601173, "grad_norm": 12.376034144224413, "kl": 1.5751953125, "learning_rate": 3.8620958787454214e-07, "loss": 0.0745, "reward": 2.406980276107788, "reward_std": 0.23445120081305504, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025311593897640705, "rewards/tag_count_reward": 0.9739583432674408, "step": 896 }, { "clip_ratio": 0.0, "completion_length": 531.3541870117188, "epoch": 0.657624633431085, "grad_norm": 480.63424417984, "kl": 2.890625, "learning_rate": 3.8513698328209197e-07, "loss": 0.1273, "reward": 1.9812852144241333, "reward_std": 0.01012721424922347, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018714841455221176, "rewards/tag_count_reward": 1.0, "step": 897 }, { "clip_ratio": 0.0, "completion_length": 435.2291717529297, "epoch": 0.658357771260997, "grad_norm": 11.679079028241473, "kl": 1.40625, "learning_rate": 3.8406545946049795e-07, "loss": 0.059, "reward": 2.402749180793762, "reward_std": 0.22888202592730522, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.03301479481160641, "rewards/tag_count_reward": 0.9635416865348816, "step": 898 }, { "clip_ratio": 0.0, "completion_length": 439.6875, "epoch": 0.6590909090909091, "grad_norm": 8.277870464945842, "kl": 0.8046875, "learning_rate": 3.8299502343420844e-07, "loss": -0.0369, "reward": 2.406039595603943, "reward_std": 0.24393422901630402, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.024516088888049126, "rewards/tag_count_reward": 0.9791666865348816, "step": 899 }, { "clip_ratio": 0.0, "completion_length": 491.6875, "epoch": 0.6598240469208211, "grad_norm": 9.435685885433191, "kl": 0.4560546875, "learning_rate": 3.819256822205403e-07, "loss": 0.0238, "reward": 2.6473140716552734, "reward_std": 0.1842022268101573, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.033241670578718185, "rewards/tag_count_reward": 0.9791666865348816, "step": 900 }, { "clip_ratio": 0.0, "completion_length": 484.7708435058594, "epoch": 0.6605571847507331, "grad_norm": 19.098315989449652, "kl": 0.65625, "learning_rate": 3.8085744282963385e-07, "loss": 0.234, "reward": 2.652103304862976, "reward_std": 0.23995533771812916, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03018837235867977, "rewards/tag_count_reward": 0.9739583432674408, "step": 901 }, { "clip_ratio": 0.0, "completion_length": 440.0833435058594, "epoch": 0.6612903225806451, "grad_norm": 6.5723034415157455, "kl": 0.8974609375, "learning_rate": 3.7979031226440563e-07, "loss": -0.0024, "reward": 2.4431477785110474, "reward_std": 0.09110517054796219, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.032546691596508026, "rewards/tag_count_reward": 0.9895833432674408, "step": 902 }, { "clip_ratio": 0.0, "completion_length": 531.875, "epoch": 0.6620234604105572, "grad_norm": 23.22294687828924, "kl": 1.6357421875, "learning_rate": 3.787242975205037e-07, "loss": 0.3049, "reward": 2.138901650905609, "reward_std": 0.24177566915750504, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022556803189218044, "rewards/tag_count_reward": 0.9531250298023224, "step": 903 }, { "clip_ratio": 0.0, "completion_length": 506.8958435058594, "epoch": 0.6627565982404692, "grad_norm": 11.32486166375894, "kl": 0.791015625, "learning_rate": 3.7765940558626096e-07, "loss": 0.1872, "reward": 2.700243353843689, "reward_std": 0.08089052699506283, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0237151263281703, "rewards/tag_count_reward": 0.9739583432674408, "step": 904 }, { "clip_ratio": 0.0, "completion_length": 454.8541717529297, "epoch": 0.6634897360703812, "grad_norm": 27.191439861059074, "kl": 2.7958984375, "learning_rate": 3.7659564344264994e-07, "loss": 0.1591, "reward": 2.1992244124412537, "reward_std": 0.08250689785927534, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.029942267574369907, "rewards/tag_count_reward": 0.9791666865348816, "step": 905 }, { "clip_ratio": 0.0, "completion_length": 469.35418701171875, "epoch": 0.6642228739002932, "grad_norm": 15.490785850217982, "kl": 1.1484375, "learning_rate": 3.755330180632367e-07, "loss": 0.2352, "reward": 2.706392288208008, "reward_std": 0.18476082384586334, "rewards/accuracy_reward": 0.7500000298023224, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.021038434468209743, "rewards/tag_count_reward": 0.984375, "step": 906 }, { "clip_ratio": 0.0, "completion_length": 428.6458435058594, "epoch": 0.6649560117302052, "grad_norm": 3.090235873839556, "kl": 0.423828125, "learning_rate": 3.7447153641413463e-07, "loss": 0.0419, "reward": 2.4653698205947876, "reward_std": 0.0442873565480113, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024213584139943123, "rewards/tag_count_reward": 0.9895833432674408, "step": 907 }, { "clip_ratio": 0.0, "completion_length": 439.75001525878906, "epoch": 0.6656891495601173, "grad_norm": 3.7331624609730536, "kl": 0.46875, "learning_rate": 3.734112054539602e-07, "loss": 0.0311, "reward": 2.458762228488922, "reward_std": 0.049120476469397545, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03082120604813099, "rewards/tag_count_reward": 0.9895833432674408, "step": 908 }, { "clip_ratio": 0.0, "completion_length": 510.3958435058594, "epoch": 0.6664222873900293, "grad_norm": 63.54354843221626, "kl": 5.3203125, "learning_rate": 3.7235203213378577e-07, "loss": 0.7462, "reward": 2.206181049346924, "reward_std": 0.49717573821544647, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.03861072286963463, "rewards/tag_count_reward": 0.8697916865348816, "step": 909 }, { "clip_ratio": 0.0, "completion_length": 486.4583435058594, "epoch": 0.6671554252199413, "grad_norm": 5.683418641201284, "kl": 1.1435546875, "learning_rate": 3.7129402339709526e-07, "loss": 0.1774, "reward": 2.4271193742752075, "reward_std": 0.11568078305572271, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.02600565180182457, "rewards/tag_count_reward": 0.9739583432674408, "step": 910 }, { "clip_ratio": 0.0, "completion_length": 468.6458435058594, "epoch": 0.6678885630498533, "grad_norm": 7.617088538664095, "kl": 0.6669921875, "learning_rate": 3.7023718617973745e-07, "loss": 0.2202, "reward": 2.146929442882538, "reward_std": 0.25023030769079924, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.019737189635634422, "rewards/tag_count_reward": 0.9791666865348816, "step": 911 }, { "clip_ratio": 0.0, "completion_length": 497.1458435058594, "epoch": 0.6686217008797654, "grad_norm": 14.494392463460322, "kl": 1.21484375, "learning_rate": 3.6918152740988163e-07, "loss": 0.2576, "reward": 2.3849753737449646, "reward_std": 0.29698850959539413, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02821914665400982, "rewards/tag_count_reward": 0.9479166865348816, "step": 912 }, { "clip_ratio": 0.0, "completion_length": 463.06251525878906, "epoch": 0.6693548387096774, "grad_norm": 7.684407419722121, "kl": 1.83984375, "learning_rate": 3.681270540079715e-07, "loss": 0.0467, "reward": 2.5368086099624634, "reward_std": 0.45409172028303146, "rewards/accuracy_reward": 0.645833358168602, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02742769941687584, "rewards/tag_count_reward": 0.9322916865348816, "step": 913 }, { "clip_ratio": 0.0, "completion_length": 417.5416717529297, "epoch": 0.6700879765395894, "grad_norm": 5.5507149790575365, "kl": 1.115234375, "learning_rate": 3.6707377288668017e-07, "loss": 0.0354, "reward": 2.65383780002594, "reward_std": 0.1826682216487825, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.026717686094343662, "rewards/tag_count_reward": 0.9791666865348816, "step": 914 }, { "clip_ratio": 0.0, "completion_length": 434.04168701171875, "epoch": 0.6708211143695014, "grad_norm": 4.575828310673458, "kl": 0.4970703125, "learning_rate": 3.6602169095086434e-07, "loss": -0.0017, "reward": 2.404066324234009, "reward_std": 0.25298498570919037, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023017059080302715, "rewards/tag_count_reward": 0.9687500298023224, "step": 915 }, { "clip_ratio": 0.0, "completion_length": 521.0625, "epoch": 0.6715542521994134, "grad_norm": 35.23107554083441, "kl": 1.98046875, "learning_rate": 3.649708150975197e-07, "loss": 0.4676, "reward": 1.9426430463790894, "reward_std": 0.26459813117980957, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02610696479678154, "rewards/tag_count_reward": 0.9270833730697632, "step": 916 }, { "clip_ratio": 0.0, "completion_length": 460.37501525878906, "epoch": 0.6722873900293255, "grad_norm": 34.980059238686245, "kl": 1.9990234375, "learning_rate": 3.6392115221573515e-07, "loss": 0.1372, "reward": 2.4030675888061523, "reward_std": 0.1829522354528308, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024015815928578377, "rewards/tag_count_reward": 0.96875, "step": 917 }, { "clip_ratio": 0.0, "completion_length": 463.87501525878906, "epoch": 0.6730205278592375, "grad_norm": 3.538742597821187, "kl": 0.796875, "learning_rate": 3.6287270918664837e-07, "loss": 0.0092, "reward": 2.6298184394836426, "reward_std": 0.27230154722929, "rewards/accuracy_reward": 0.6875000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021223259158432484, "rewards/tag_count_reward": 0.9635416865348816, "step": 918 }, { "clip_ratio": 0.0, "completion_length": 471.8958435058594, "epoch": 0.6737536656891495, "grad_norm": 12.629477020085433, "kl": 0.779296875, "learning_rate": 3.6182549288339905e-07, "loss": 0.1492, "reward": 2.2270041704177856, "reward_std": 0.23701730370521545, "rewards/accuracy_reward": 0.27083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02820419240742922, "rewards/tag_count_reward": 0.984375, "step": 919 }, { "clip_ratio": 0.0, "completion_length": 426.18751525878906, "epoch": 0.6744868035190615, "grad_norm": 3.4352358495668835, "kl": 0.392578125, "learning_rate": 3.607795101710865e-07, "loss": 0.0245, "reward": 2.4957412481307983, "reward_std": 0.07916717370972037, "rewards/accuracy_reward": 0.5208333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025092164054512978, "rewards/tag_count_reward": 1.0, "step": 920 }, { "clip_ratio": 0.0, "completion_length": 457.2916717529297, "epoch": 0.6752199413489736, "grad_norm": 3.1335327742824615, "kl": 0.37109375, "learning_rate": 3.597347679067222e-07, "loss": 0.0056, "reward": 2.2258177399635315, "reward_std": 0.011170658748596907, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024182401597499847, "rewards/tag_count_reward": 1.0, "step": 921 }, { "clip_ratio": 0.0, "completion_length": 541.1458435058594, "epoch": 0.6759530791788856, "grad_norm": 26.459636599607972, "kl": 3.4921875, "learning_rate": 3.586912729391857e-07, "loss": 0.3193, "reward": 2.3519192934036255, "reward_std": 0.32803425192832947, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02308077458292246, "rewards/tag_count_reward": 0.9166666865348816, "step": 922 }, { "clip_ratio": 0.0, "completion_length": 500.9583435058594, "epoch": 0.6766862170087976, "grad_norm": 3.626780250846866, "kl": 0.3642578125, "learning_rate": 3.576490321091805e-07, "loss": 0.0544, "reward": 2.2058114409446716, "reward_std": 0.08280891180038452, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023355277255177498, "rewards/tag_count_reward": 1.0, "step": 923 }, { "clip_ratio": 0.0, "completion_length": 450.0833435058594, "epoch": 0.6774193548387096, "grad_norm": 6.7576974288115785, "kl": 0.7353515625, "learning_rate": 3.5660805224918766e-07, "loss": 0.0833, "reward": 2.2170188426971436, "reward_std": 0.060911763459444046, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01735620480030775, "rewards/tag_count_reward": 0.984375, "step": 924 }, { "clip_ratio": 0.0, "completion_length": 468.7291717529297, "epoch": 0.6781524926686217, "grad_norm": 54.06107893396688, "kl": 0.638671875, "learning_rate": 3.555683401834229e-07, "loss": 0.1777, "reward": 2.3949154019355774, "reward_std": 0.20315419882535934, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.030431894585490227, "rewards/tag_count_reward": 0.9739583432674408, "step": 925 }, { "clip_ratio": 0.0, "completion_length": 459.0, "epoch": 0.6788856304985337, "grad_norm": 3.7304310474391342, "kl": 0.3349609375, "learning_rate": 3.545299027277897e-07, "loss": 0.0282, "reward": 2.420232832431793, "reward_std": 0.18446220457553864, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.027683909982442856, "rewards/tag_count_reward": 0.9895833432674408, "step": 926 }, { "clip_ratio": 0.0, "completion_length": 491.0625305175781, "epoch": 0.6796187683284457, "grad_norm": 2.8505337887237907, "kl": 0.3388671875, "learning_rate": 3.5349274668983674e-07, "loss": 0.0159, "reward": 2.2121185064315796, "reward_std": 0.07882623188197613, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01704825460910797, "rewards/tag_count_reward": 1.0, "step": 927 }, { "clip_ratio": 0.0, "completion_length": 459.81251525878906, "epoch": 0.6803519061583577, "grad_norm": 5.799691271582148, "kl": 0.447265625, "learning_rate": 3.5245687886871146e-07, "loss": -0.0252, "reward": 2.419118046760559, "reward_std": 0.1844143122434616, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02879861183464527, "rewards/tag_count_reward": 0.96875, "step": 928 }, { "clip_ratio": 0.0, "completion_length": 466.0, "epoch": 0.6810850439882697, "grad_norm": 3.071210500109716, "kl": 0.3623046875, "learning_rate": 3.514223060551169e-07, "loss": 0.0314, "reward": 2.18208384513855, "reward_std": 0.1516893282532692, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026249597780406475, "rewards/tag_count_reward": 0.9791666865348816, "step": 929 }, { "clip_ratio": 0.0, "completion_length": 459.81251525878906, "epoch": 0.6818181818181818, "grad_norm": 20.45901417722426, "kl": 1.2587890625, "learning_rate": 3.5038903503126615e-07, "loss": 0.2261, "reward": 2.42332124710083, "reward_std": 0.12113591935485601, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.028067844919860363, "rewards/tag_count_reward": 0.9583333432674408, "step": 930 }, { "clip_ratio": 0.0, "completion_length": 484.31251525878906, "epoch": 0.6825513196480938, "grad_norm": 3204.5978551725684, "kl": 18.5478515625, "learning_rate": 3.493570725708389e-07, "loss": 2.0223, "reward": 2.3534035682678223, "reward_std": 0.24623488262295723, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026804929599165916, "rewards/tag_count_reward": 0.9427083432674408, "step": 931 }, { "clip_ratio": 0.0, "completion_length": 534.5833435058594, "epoch": 0.6832844574780058, "grad_norm": 86.07803816104924, "kl": 4.0537109375, "learning_rate": 3.483264254389358e-07, "loss": 0.4928, "reward": 2.188483476638794, "reward_std": 0.2733900472521782, "rewards/accuracy_reward": 0.25000000558793545, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025058263912796974, "rewards/tag_count_reward": 0.9635416865348816, "step": 932 }, { "clip_ratio": 0.0, "completion_length": 459.0208435058594, "epoch": 0.6840175953079178, "grad_norm": 22.812614433514874, "kl": 1.59375, "learning_rate": 3.472971003920353e-07, "loss": 0.1919, "reward": 2.3389848470687866, "reward_std": 0.393401101231575, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.030806850641965866, "rewards/tag_count_reward": 0.9531250298023224, "step": 933 }, { "clip_ratio": 0.0, "completion_length": 536.6250152587891, "epoch": 0.6847507331378299, "grad_norm": 2.672732005418521, "kl": 0.3359375, "learning_rate": 3.462691041779483e-07, "loss": 0.0192, "reward": 2.4783459901809692, "reward_std": 0.011177142150700092, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02165412623435259, "rewards/tag_count_reward": 1.0, "step": 934 }, { "clip_ratio": 0.0, "completion_length": 448.8958435058594, "epoch": 0.6854838709677419, "grad_norm": 5.349593962272704, "kl": 0.416015625, "learning_rate": 3.4524244353577513e-07, "loss": 0.0268, "reward": 2.214493155479431, "reward_std": 0.051492489874362946, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02335413731634617, "rewards/tag_count_reward": 0.9947916865348816, "step": 935 }, { "clip_ratio": 0.0, "completion_length": 523.7708435058594, "epoch": 0.6862170087976539, "grad_norm": 14.297145277012756, "kl": 1.078125, "learning_rate": 3.4421712519586016e-07, "loss": 0.2576, "reward": 2.4406609535217285, "reward_std": 0.1489308439195156, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017672521993517876, "rewards/tag_count_reward": 0.9791666865348816, "step": 936 }, { "clip_ratio": 0.0, "completion_length": 493.5208435058594, "epoch": 0.6869501466275659, "grad_norm": 32.92184403209849, "kl": 1.6328125, "learning_rate": 3.4319315587974806e-07, "loss": 0.3352, "reward": 2.660341739654541, "reward_std": 0.22230494022369385, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.027158407494425774, "rewards/tag_count_reward": 0.9791666865348816, "step": 937 }, { "clip_ratio": 0.0, "completion_length": 466.8541717529297, "epoch": 0.6876832844574781, "grad_norm": 8.17285504048901, "kl": 0.4970703125, "learning_rate": 3.4217054230014073e-07, "loss": 0.0179, "reward": 2.4424521923065186, "reward_std": 0.1104821190237999, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02629770338535309, "rewards/tag_count_reward": 0.9895833432674408, "step": 938 }, { "clip_ratio": 0.0, "completion_length": 449.3125, "epoch": 0.6884164222873901, "grad_norm": 8.937956638965543, "kl": 0.55859375, "learning_rate": 3.411492911608513e-07, "loss": 0.0235, "reward": 1.9536827206611633, "reward_std": 0.07988723739981651, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.02548403199762106, "rewards/tag_count_reward": 1.0, "step": 939 }, { "clip_ratio": 0.0, "completion_length": 591.7708435058594, "epoch": 0.6891495601173021, "grad_norm": 83.16419128491232, "kl": 5.015625, "learning_rate": 3.401294091567626e-07, "loss": 0.7022, "reward": 1.9109691977500916, "reward_std": 0.28589335083961487, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.04041970055550337, "rewards/tag_count_reward": 0.9375000298023224, "step": 940 }, { "clip_ratio": 0.0, "completion_length": 468.4375, "epoch": 0.6898826979472141, "grad_norm": 57.040934440046975, "kl": 3.2890625, "learning_rate": 3.3911090297378054e-07, "loss": 0.3252, "reward": 2.1530415415763855, "reward_std": 0.22559036314487457, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.027514135465025902, "rewards/tag_count_reward": 0.9583333730697632, "step": 941 }, { "clip_ratio": 0.0, "completion_length": 456.00001525878906, "epoch": 0.6906158357771262, "grad_norm": 2.8055744806214746, "kl": 0.349609375, "learning_rate": 3.3809377928879327e-07, "loss": 0.0145, "reward": 2.45681369304657, "reward_std": 0.08130190032534301, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02235298790037632, "rewards/tag_count_reward": 1.0, "step": 942 }, { "clip_ratio": 0.0, "completion_length": 487.31251525878906, "epoch": 0.6913489736070382, "grad_norm": 4.906686721090249, "kl": 0.345703125, "learning_rate": 3.370780447696246e-07, "loss": 0.0158, "reward": 2.449528932571411, "reward_std": 0.09997610654681921, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02442941814661026, "rewards/tag_count_reward": 0.9947916865348816, "step": 943 }, { "clip_ratio": 0.0, "completion_length": 466.50001525878906, "epoch": 0.6920821114369502, "grad_norm": 7.045604029987775, "kl": 1.572265625, "learning_rate": 3.360637060749924e-07, "loss": 0.0785, "reward": 1.9422118663787842, "reward_std": 0.07563984906300902, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02306581661105156, "rewards/tag_count_reward": 0.9791666865348816, "step": 944 }, { "clip_ratio": 0.0, "completion_length": 472.3333435058594, "epoch": 0.6928152492668622, "grad_norm": 4.579383798845335, "kl": 0.3828125, "learning_rate": 3.350507698544636e-07, "loss": 0.0729, "reward": 2.200911045074463, "reward_std": 0.07387769222259521, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02304728701710701, "rewards/tag_count_reward": 0.9739583432674408, "step": 945 }, { "clip_ratio": 0.0, "completion_length": 478.29168701171875, "epoch": 0.6935483870967742, "grad_norm": 11.882042524248234, "kl": 1.51171875, "learning_rate": 3.340392427484117e-07, "loss": 0.1229, "reward": 2.3835513591766357, "reward_std": 0.24781902134418488, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022698688320815563, "rewards/tag_count_reward": 0.9479166865348816, "step": 946 }, { "clip_ratio": 0.0, "completion_length": 460.2708435058594, "epoch": 0.6942815249266863, "grad_norm": 16.42715424721198, "kl": 1.53515625, "learning_rate": 3.330291313879716e-07, "loss": 0.2588, "reward": 2.4029751420021057, "reward_std": 0.23220430314540863, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.0310527877882123, "rewards/tag_count_reward": 0.9687500298023224, "step": 947 }, { "clip_ratio": 0.0, "completion_length": 460.75001525878906, "epoch": 0.6950146627565983, "grad_norm": 22.424491823020983, "kl": 2.09765625, "learning_rate": 3.3202044239499867e-07, "loss": 0.3906, "reward": 2.326223611831665, "reward_std": 0.40988829731941223, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.024470987729728222, "rewards/tag_count_reward": 0.9270833432674408, "step": 948 }, { "clip_ratio": 0.0, "completion_length": 560.9166870117188, "epoch": 0.6957478005865103, "grad_norm": 84.3739723227209, "kl": 7.5234375, "learning_rate": 3.3101318238202245e-07, "loss": 0.415, "reward": 2.309966802597046, "reward_std": 0.2813423126935959, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.019894277676939964, "rewards/tag_count_reward": 0.8854166865348816, "step": 949 }, { "clip_ratio": 0.0, "completion_length": 498.7708435058594, "epoch": 0.6964809384164223, "grad_norm": 34.148485717679705, "kl": 3.671875, "learning_rate": 3.3000735795220566e-07, "loss": 0.4339, "reward": 2.2733023166656494, "reward_std": 0.4924462139606476, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.023572625126689672, "rewards/tag_count_reward": 0.8802083432674408, "step": 950 }, { "clip_ratio": 0.0, "completion_length": 540.9166717529297, "epoch": 0.6972140762463344, "grad_norm": 41.73333734505236, "kl": 4.234375, "learning_rate": 3.290029756992995e-07, "loss": 0.4413, "reward": 2.3373730182647705, "reward_std": 0.3256710283458233, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.027210503350943327, "rewards/tag_count_reward": 0.9062500298023224, "step": 951 }, { "clip_ratio": 0.0, "completion_length": 519.5625, "epoch": 0.6979472140762464, "grad_norm": 76.8637407662865, "kl": 6.9921875, "learning_rate": 3.280000422076007e-07, "loss": 0.6329, "reward": 2.4091644287109375, "reward_std": 0.26125724613666534, "rewards/accuracy_reward": 0.5416666865348816, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.024863524362444878, "rewards/tag_count_reward": 0.9062500298023224, "step": 952 }, { "clip_ratio": 0.0, "completion_length": 460.1041717529297, "epoch": 0.6986803519061584, "grad_norm": 5.933236012161051, "kl": 0.72265625, "learning_rate": 3.269985640519093e-07, "loss": 0.043, "reward": 2.4667599201202393, "reward_std": 0.04559937119483948, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022823692299425602, "rewards/tag_count_reward": 0.9895833432674408, "step": 953 }, { "clip_ratio": 0.0, "completion_length": 446.2291717529297, "epoch": 0.6994134897360704, "grad_norm": 11.219462628947934, "kl": 0.67578125, "learning_rate": 3.259985477974838e-07, "loss": 0.0423, "reward": 2.6256914138793945, "reward_std": 0.2616752292960882, "rewards/accuracy_reward": 0.6875, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030558819882571697, "rewards/tag_count_reward": 0.96875, "step": 954 }, { "clip_ratio": 0.0, "completion_length": 508.47918701171875, "epoch": 0.7001466275659824, "grad_norm": 43.155555290044155, "kl": 1.7216796875, "learning_rate": 3.250000000000001e-07, "loss": 0.3124, "reward": 2.1255834102630615, "reward_std": 0.252389095723629, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.021986079402267933, "rewards/tag_count_reward": 0.9322916865348816, "step": 955 }, { "clip_ratio": 0.0, "completion_length": 489.18751525878906, "epoch": 0.7008797653958945, "grad_norm": 25.849133555001085, "kl": 3.09375, "learning_rate": 3.240029272055065e-07, "loss": 0.3879, "reward": 2.3594762086868286, "reward_std": 0.5459988713264465, "rewards/accuracy_reward": 0.5625000149011612, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.025940511375665665, "rewards/tag_count_reward": 0.8645833730697632, "step": 956 }, { "clip_ratio": 0.0, "completion_length": 492.4375, "epoch": 0.7016129032258065, "grad_norm": 8.258539904474, "kl": 0.76171875, "learning_rate": 3.230073359503829e-07, "loss": 0.0552, "reward": 2.192040979862213, "reward_std": 0.13235630840063095, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021500680595636368, "rewards/tag_count_reward": 0.984375, "step": 957 }, { "clip_ratio": 0.0, "completion_length": 464.3333435058594, "epoch": 0.7023460410557185, "grad_norm": 8.679112971623974, "kl": 1.78125, "learning_rate": 3.220132327612959e-07, "loss": 0.0354, "reward": 2.4318766593933105, "reward_std": 0.42502492666244507, "rewards/accuracy_reward": 0.5833333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.02124821860343218, "rewards/tag_count_reward": 0.890625, "step": 958 }, { "clip_ratio": 0.0, "completion_length": 515.6458435058594, "epoch": 0.7030791788856305, "grad_norm": 27.796035153234367, "kl": 1.87890625, "learning_rate": 3.210206241551576e-07, "loss": 0.1472, "reward": 2.4991250038146973, "reward_std": 0.5063153505325317, "rewards/accuracy_reward": 0.6250000149011612, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.028652765788137913, "rewards/tag_count_reward": 0.9166666865348816, "step": 959 }, { "clip_ratio": 0.0, "completion_length": 562.5208435058594, "epoch": 0.7038123167155426, "grad_norm": 16.67018322666244, "kl": 3.1796875, "learning_rate": 3.2002951663908184e-07, "loss": 0.3967, "reward": 2.2723554372787476, "reward_std": 0.4776581823825836, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.02451958041638136, "rewards/tag_count_reward": 0.9010416865348816, "step": 960 }, { "clip_ratio": 0.0, "completion_length": 559.6250305175781, "epoch": 0.7045454545454546, "grad_norm": 18.351794895624867, "kl": 2.2109375, "learning_rate": 3.1903991671034257e-07, "loss": 0.3861, "reward": 2.369241237640381, "reward_std": 0.22771289199590683, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026591990143060684, "rewards/tag_count_reward": 0.9166666865348816, "step": 961 }, { "clip_ratio": 0.0, "completion_length": 485.47918701171875, "epoch": 0.7052785923753666, "grad_norm": 13.326292369321743, "kl": 2.2421875, "learning_rate": 3.180518308563298e-07, "loss": 0.2725, "reward": 2.3383569717407227, "reward_std": 0.3190757483243942, "rewards/accuracy_reward": 0.4166666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02101796120405197, "rewards/tag_count_reward": 0.9427083432674408, "step": 962 }, { "clip_ratio": 0.0, "completion_length": 503.2291717529297, "epoch": 0.7060117302052786, "grad_norm": 38.51770861382194, "kl": 4.75, "learning_rate": 3.1706526555450874e-07, "loss": 0.5938, "reward": 2.50447940826416, "reward_std": 0.4233640283346176, "rewards/accuracy_reward": 0.6458333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.03197902161628008, "rewards/tag_count_reward": 0.9114583432674408, "step": 963 }, { "clip_ratio": 0.0, "completion_length": 483.4583435058594, "epoch": 0.7067448680351907, "grad_norm": 23.230178570798735, "kl": 4.46875, "learning_rate": 3.160802272723758e-07, "loss": 0.2725, "reward": 2.4244531393051147, "reward_std": 0.4745712876319885, "rewards/accuracy_reward": 0.5833333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.030408103950321674, "rewards/tag_count_reward": 0.8854166865348816, "step": 964 }, { "clip_ratio": 0.0, "completion_length": 505.6041717529297, "epoch": 0.7074780058651027, "grad_norm": 88.08557499027471, "kl": 7.671875, "learning_rate": 3.150967224674177e-07, "loss": 0.5289, "reward": 1.9547891020774841, "reward_std": 0.3700524717569351, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.02437763847410679, "rewards/tag_count_reward": 0.8541666865348816, "step": 965 }, { "clip_ratio": 0.0, "completion_length": 504.31251525878906, "epoch": 0.7082111436950147, "grad_norm": 15.371719850558602, "kl": 2.17578125, "learning_rate": 3.141147575870676e-07, "loss": 0.3489, "reward": 2.378512144088745, "reward_std": 0.308450847864151, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.019057356752455235, "rewards/tag_count_reward": 0.9322916865348816, "step": 966 }, { "clip_ratio": 0.0, "completion_length": 497.60418701171875, "epoch": 0.7089442815249267, "grad_norm": 11.79394222832606, "kl": 2.16015625, "learning_rate": 3.1313433906866374e-07, "loss": 0.2853, "reward": 2.5798145532608032, "reward_std": 0.33551979064941406, "rewards/accuracy_reward": 0.6666666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024352273903787136, "rewards/tag_count_reward": 0.9375000298023224, "step": 967 }, { "clip_ratio": 0.0, "completion_length": 448.29168701171875, "epoch": 0.7096774193548387, "grad_norm": 14.817819935865485, "kl": 3.67578125, "learning_rate": 3.1215547333940764e-07, "loss": 0.1242, "reward": 2.103150486946106, "reward_std": 0.185621440410614, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9722222089767456, "rewards/repetition_penalty_reward": -0.020113405771553516, "rewards/tag_count_reward": 0.9010416865348816, "step": 968 }, { "clip_ratio": 0.0, "completion_length": 452.12501525878906, "epoch": 0.7104105571847508, "grad_norm": 9.799356075669513, "kl": 2.66015625, "learning_rate": 3.1117816681632046e-07, "loss": 0.1152, "reward": 2.278549313545227, "reward_std": 0.29779060930013657, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.023534007370471954, "rewards/tag_count_reward": 0.9270833432674408, "step": 969 }, { "clip_ratio": 0.0, "completion_length": 490.7291717529297, "epoch": 0.7111436950146628, "grad_norm": 23.529385447949288, "kl": 2.2421875, "learning_rate": 3.1020242590620293e-07, "loss": 0.2476, "reward": 1.8638713955879211, "reward_std": 0.26981137692928314, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9652778506278992, "rewards/repetition_penalty_reward": -0.023281488567590714, "rewards/tag_count_reward": 0.921875, "step": 970 }, { "clip_ratio": 0.0, "completion_length": 577.4166870117188, "epoch": 0.7118768328445748, "grad_norm": 17.290844738388767, "kl": 0.8828125, "learning_rate": 3.092282570055913e-07, "loss": 0.0675, "reward": 2.3447808027267456, "reward_std": 0.29597213864326477, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.012858240399509668, "rewards/tag_count_reward": 0.9270833432674408, "step": 971 }, { "clip_ratio": 0.0, "completion_length": 448.4583435058594, "epoch": 0.7126099706744868, "grad_norm": 25.22529983173914, "kl": 2.34375, "learning_rate": 3.0825566650071726e-07, "loss": 0.2648, "reward": 2.5462182760238647, "reward_std": 0.37398337107151747, "rewards/accuracy_reward": 0.6666666865348816, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.026698616333305836, "rewards/tag_count_reward": 0.9270833432674408, "step": 972 }, { "clip_ratio": 0.0, "completion_length": 556.3125, "epoch": 0.7133431085043989, "grad_norm": 57.26538846475611, "kl": 2.48828125, "learning_rate": 3.072846607674645e-07, "loss": 0.6625, "reward": 2.342366099357605, "reward_std": 0.38562868535518646, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0222172187641263, "rewards/tag_count_reward": 0.9270833730697632, "step": 973 }, { "clip_ratio": 0.0, "completion_length": 504.4791717529297, "epoch": 0.7140762463343109, "grad_norm": 28.943299265226212, "kl": 4.5546875, "learning_rate": 3.063152461713282e-07, "loss": 0.3193, "reward": 2.0221683979034424, "reward_std": 0.4104950726032257, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.02123451605439186, "rewards/tag_count_reward": 0.8906250298023224, "step": 974 }, { "clip_ratio": 0.0, "completion_length": 545.0000152587891, "epoch": 0.7148093841642229, "grad_norm": 117.23597424908264, "kl": 11.8203125, "learning_rate": 3.053474290673722e-07, "loss": 0.9305, "reward": 2.131690263748169, "reward_std": 0.6166838854551315, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9513889253139496, "rewards/repetition_penalty_reward": -0.017615425400435925, "rewards/tag_count_reward": 0.71875, "step": 975 }, { "clip_ratio": 0.0, "completion_length": 508.16668701171875, "epoch": 0.7155425219941349, "grad_norm": 26.99768837152522, "kl": 4.7421875, "learning_rate": 3.0438121580018863e-07, "loss": 0.5739, "reward": 2.092137038707733, "reward_std": 0.2849106341600418, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9722222089767456, "rewards/repetition_penalty_reward": -0.020710238255560398, "rewards/tag_count_reward": 0.9322916865348816, "step": 976 }, { "clip_ratio": 0.0, "completion_length": 576.7083435058594, "epoch": 0.716275659824047, "grad_norm": 30.385974891375128, "kl": 3.8359375, "learning_rate": 3.0341661270385446e-07, "loss": 0.4812, "reward": 2.785071015357971, "reward_std": 0.561683714389801, "rewards/accuracy_reward": 0.9166666865348816, "rewards/reasoning_steps_reward": 0.965277761220932, "rewards/repetition_penalty_reward": -0.023956785909831524, "rewards/tag_count_reward": 0.9270833730697632, "step": 977 }, { "clip_ratio": 0.0, "completion_length": 405.2291717529297, "epoch": 0.717008797653959, "grad_norm": 10.81014359610337, "kl": 1.70703125, "learning_rate": 3.024536261018923e-07, "loss": 0.0371, "reward": 2.183934807777405, "reward_std": 0.1534016877412796, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02266244124621153, "rewards/tag_count_reward": 0.9635416865348816, "step": 978 }, { "clip_ratio": 0.0, "completion_length": 505.4166717529297, "epoch": 0.717741935483871, "grad_norm": 20.583436231099498, "kl": 4.859375, "learning_rate": 3.0149226230722656e-07, "loss": 0.5499, "reward": 2.495741605758667, "reward_std": 0.45375633239746094, "rewards/accuracy_reward": 0.645833358168602, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.021619529463350773, "rewards/tag_count_reward": 0.8854166865348816, "step": 979 }, { "clip_ratio": 0.0, "completion_length": 563.1875305175781, "epoch": 0.718475073313783, "grad_norm": 21.523757743477038, "kl": 3.1484375, "learning_rate": 3.0053252762214434e-07, "loss": 0.5152, "reward": 2.2795934081077576, "reward_std": 0.5194336771965027, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02075397362932563, "rewards/tag_count_reward": 0.8697916865348816, "step": 980 }, { "clip_ratio": 0.0, "completion_length": 462.6458435058594, "epoch": 0.719208211143695, "grad_norm": 9.328779600109762, "kl": 2.67578125, "learning_rate": 2.9957442833825207e-07, "loss": 0.023, "reward": 1.9250345826148987, "reward_std": 0.28224488347768784, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.02461833320558071, "rewards/tag_count_reward": 0.9635416865348816, "step": 981 }, { "clip_ratio": 0.0, "completion_length": 517.7083435058594, "epoch": 0.7199413489736071, "grad_norm": 68.08501111143843, "kl": 7.2265625, "learning_rate": 2.986179707364355e-07, "loss": 0.6412, "reward": 2.2177438735961914, "reward_std": 0.499715194106102, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.025311839766800404, "rewards/tag_count_reward": 0.875, "step": 982 }, { "clip_ratio": 0.0, "completion_length": 544.9166870117188, "epoch": 0.7206744868035191, "grad_norm": 53.114941492476, "kl": 6.5390625, "learning_rate": 2.9766316108681867e-07, "loss": 0.3512, "reward": 2.465793192386627, "reward_std": 0.4085993431508541, "rewards/accuracy_reward": 0.6666666716337204, "rewards/reasoning_steps_reward": 0.944444477558136, "rewards/repetition_penalty_reward": -0.020318015478551388, "rewards/tag_count_reward": 0.875, "step": 983 }, { "clip_ratio": 0.0, "completion_length": 470.87501525878906, "epoch": 0.7214076246334311, "grad_norm": 21.15094205176047, "kl": 5.328125, "learning_rate": 2.9671000564872155e-07, "loss": 0.2666, "reward": 2.513883590698242, "reward_std": 0.42151640355587006, "rewards/accuracy_reward": 0.645833358168602, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.017366615124046803, "rewards/tag_count_reward": 0.9062500298023224, "step": 984 }, { "clip_ratio": 0.0, "completion_length": 524.2916717529297, "epoch": 0.7221407624633431, "grad_norm": 29.752132464122592, "kl": 4.78125, "learning_rate": 2.9575851067062065e-07, "loss": 0.5037, "reward": 2.2734715938568115, "reward_std": 0.5585350394248962, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9513888955116272, "rewards/repetition_penalty_reward": -0.02166739758104086, "rewards/tag_count_reward": 0.8854166865348816, "step": 985 }, { "clip_ratio": 0.0, "completion_length": 507.8333435058594, "epoch": 0.7228739002932552, "grad_norm": 14.723027004397416, "kl": 4.6484375, "learning_rate": 2.948086823901064e-07, "loss": 0.4912, "reward": 2.0892439484596252, "reward_std": 0.30918999761343, "rewards/accuracy_reward": 0.22916667722165585, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.025339476764202118, "rewards/tag_count_reward": 0.9062500298023224, "step": 986 }, { "clip_ratio": 0.0, "completion_length": 487.375, "epoch": 0.7236070381231672, "grad_norm": 21.42719494284422, "kl": 3.1953125, "learning_rate": 2.938605270338438e-07, "loss": 0.305, "reward": 2.3562979698181152, "reward_std": 0.33968590199947357, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.022174209356307983, "rewards/tag_count_reward": 0.9270833730697632, "step": 987 }, { "clip_ratio": 0.0, "completion_length": 581.4166870117188, "epoch": 0.7243401759530792, "grad_norm": 57.98934250609035, "kl": 2.4609375, "learning_rate": 2.929140508175303e-07, "loss": 0.4714, "reward": 2.6463661193847656, "reward_std": 0.5489610433578491, "rewards/accuracy_reward": 0.8333333730697632, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.01856454461812973, "rewards/tag_count_reward": 0.8385416865348816, "step": 988 }, { "clip_ratio": 0.0, "completion_length": 504.12501525878906, "epoch": 0.7250733137829912, "grad_norm": 5.472385083004388, "kl": 0.865234375, "learning_rate": 2.919692599458562e-07, "loss": -0.0366, "reward": 2.373246669769287, "reward_std": 0.2569347317330539, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.019114472903311253, "rewards/tag_count_reward": 0.96875, "step": 989 }, { "clip_ratio": 0.0, "completion_length": 519.3541717529297, "epoch": 0.7258064516129032, "grad_norm": 39.08643489144547, "kl": 1.72265625, "learning_rate": 2.910261606124628e-07, "loss": 0.3925, "reward": 2.668207883834839, "reward_std": 0.20043479651212692, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019292150624096394, "rewards/tag_count_reward": 0.9583333730697632, "step": 990 }, { "clip_ratio": 0.0, "completion_length": 477.75, "epoch": 0.7265395894428153, "grad_norm": 10.397393837487297, "kl": 3.3984375, "learning_rate": 2.9008475899990303e-07, "loss": 0.3512, "reward": 1.907959520816803, "reward_std": 0.17317739874124527, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.0225960835814476, "rewards/tag_count_reward": 0.9375000298023224, "step": 991 }, { "clip_ratio": 0.0, "completion_length": 423.0, "epoch": 0.7272727272727273, "grad_norm": 6.425647931776092, "kl": 1.0693359375, "learning_rate": 2.891450612795996e-07, "loss": 0.0657, "reward": 2.434900164604187, "reward_std": 0.11677450966089964, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.033849772065877914, "rewards/tag_count_reward": 0.9895833432674408, "step": 992 }, { "clip_ratio": 0.0, "completion_length": 430.8958435058594, "epoch": 0.7280058651026393, "grad_norm": 7.681856829230564, "kl": 0.69140625, "learning_rate": 2.88207073611806e-07, "loss": 0.0605, "reward": 2.4544625282287598, "reward_std": 0.09726346097886562, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019495833665132523, "rewards/tag_count_reward": 0.9947916865348816, "step": 993 }, { "clip_ratio": 0.0, "completion_length": 536.3750152587891, "epoch": 0.7287390029325513, "grad_norm": 32.06588173332881, "kl": 3.70703125, "learning_rate": 2.8727080214556457e-07, "loss": 0.5747, "reward": 2.737803339958191, "reward_std": 0.5189033299684525, "rewards/accuracy_reward": 0.8750000298023224, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.024349420331418514, "rewards/tag_count_reward": 0.9010416865348816, "step": 994 }, { "clip_ratio": 0.0, "completion_length": 475.4166717529297, "epoch": 0.7294721407624634, "grad_norm": 34.63378364115736, "kl": 3.9453125, "learning_rate": 2.863362530186678e-07, "loss": 0.2747, "reward": 2.359430193901062, "reward_std": 0.3197652115486562, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02772271167486906, "rewards/tag_count_reward": 0.9635416865348816, "step": 995 }, { "clip_ratio": 0.0, "completion_length": 442.2916717529297, "epoch": 0.7302052785923754, "grad_norm": 11.427440648824225, "kl": 1.951171875, "learning_rate": 2.8540343235761647e-07, "loss": 0.1325, "reward": 2.1879454851150513, "reward_std": 0.10632134415209293, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02038785256445408, "rewards/tag_count_reward": 0.9583333432674408, "step": 996 }, { "clip_ratio": 0.0, "completion_length": 514.9375152587891, "epoch": 0.7309384164222874, "grad_norm": 40.24736372449541, "kl": 5.890625, "learning_rate": 2.844723462775808e-07, "loss": 0.6704, "reward": 2.442981481552124, "reward_std": 0.531048059463501, "rewards/accuracy_reward": 0.6250000298023224, "rewards/reasoning_steps_reward": 0.9374999701976776, "rewards/repetition_penalty_reward": -0.025768570601940155, "rewards/tag_count_reward": 0.90625, "step": 997 }, { "clip_ratio": 0.0, "completion_length": 502.7708435058594, "epoch": 0.7316715542521994, "grad_norm": 5.792758545167139, "kl": 0.421875, "learning_rate": 2.835430008823595e-07, "loss": 0.0022, "reward": 2.200049042701721, "reward_std": 0.11532247066497803, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018701023422181606, "rewards/tag_count_reward": 0.9895833432674408, "step": 998 }, { "clip_ratio": 0.0, "completion_length": 443.22918701171875, "epoch": 0.7324046920821115, "grad_norm": 4.830574121704087, "kl": 0.392578125, "learning_rate": 2.8261540226434053e-07, "loss": 0.0439, "reward": 2.2262839674949646, "reward_std": 0.012909410055726767, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023716239258646965, "rewards/tag_count_reward": 1.0, "step": 999 }, { "clip_ratio": 0.0, "completion_length": 456.06251525878906, "epoch": 0.7331378299120235, "grad_norm": 15.554349111823829, "kl": 1.70703125, "learning_rate": 2.816895565044605e-07, "loss": 0.175, "reward": 2.610460877418518, "reward_std": 0.27700604498386383, "rewards/accuracy_reward": 0.6875000298023224, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03537251055240631, "rewards/tag_count_reward": 0.9583333432674408, "step": 1000 }, { "clip_ratio": 0.0, "completion_length": 444.31251525878906, "epoch": 0.7338709677419355, "grad_norm": 2.7793634910585094, "kl": 0.390625, "learning_rate": 2.8076546967216487e-07, "loss": 0.0144, "reward": 2.483184516429901, "reward_std": 0.011150819715112448, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.016815499402582645, "rewards/tag_count_reward": 1.0, "step": 1001 }, { "clip_ratio": 0.0, "completion_length": 484.10418701171875, "epoch": 0.7346041055718475, "grad_norm": 2.713549340063795, "kl": 0.34375, "learning_rate": 2.798431478253686e-07, "loss": 0.0461, "reward": 2.456187903881073, "reward_std": 0.07819372694939375, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02297879382967949, "rewards/tag_count_reward": 1.0, "step": 1002 }, { "clip_ratio": 0.0, "completion_length": 452.2083435058594, "epoch": 0.7353372434017595, "grad_norm": 50.419548757796655, "kl": 1.0673828125, "learning_rate": 2.78922597010416e-07, "loss": -0.0155, "reward": 2.23778760433197, "reward_std": 0.04544468317180872, "rewards/accuracy_reward": 0.27083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022629093378782272, "rewards/tag_count_reward": 0.9895833432674408, "step": 1003 }, { "clip_ratio": 0.0, "completion_length": 474.79168701171875, "epoch": 0.7360703812316716, "grad_norm": 19.19048360727899, "kl": 0.7900390625, "learning_rate": 2.7800382326204126e-07, "loss": 0.1705, "reward": 2.125224530696869, "reward_std": 0.3402523510158062, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.020608842372894287, "rewards/tag_count_reward": 0.9375000298023224, "step": 1004 }, { "clip_ratio": 0.0, "completion_length": 425.25001525878906, "epoch": 0.7368035190615836, "grad_norm": 6.278542797034713, "kl": 0.5361328125, "learning_rate": 2.770868326033291e-07, "loss": 0.0111, "reward": 2.221871554851532, "reward_std": 0.028360038995742798, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02292016614228487, "rewards/tag_count_reward": 0.9947916865348816, "step": 1005 }, { "clip_ratio": 0.0, "completion_length": 480.1875, "epoch": 0.7375366568914956, "grad_norm": 29.55654512712693, "kl": 0.8466796875, "learning_rate": 2.761716310456748e-07, "loss": 0.2498, "reward": 2.4252238273620605, "reward_std": 0.16729998588562012, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02790138777345419, "rewards/tag_count_reward": 0.9739583432674408, "step": 1006 }, { "clip_ratio": 0.0, "completion_length": 425.3958435058594, "epoch": 0.7382697947214076, "grad_norm": 8.38447348775669, "kl": 0.83203125, "learning_rate": 2.7525822458874524e-07, "loss": 0.0214, "reward": 2.186690092086792, "reward_std": 0.12868688255548477, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.021643266081809998, "rewards/tag_count_reward": 0.9791666865348816, "step": 1007 }, { "clip_ratio": 0.0, "completion_length": 472.41668701171875, "epoch": 0.7390029325513197, "grad_norm": 3.1114573861463257, "kl": 0.349609375, "learning_rate": 2.7434661922043914e-07, "loss": 0.011, "reward": 2.4778274297714233, "reward_std": 0.0110463029704988, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0221726531162858, "rewards/tag_count_reward": 1.0, "step": 1008 }, { "clip_ratio": 0.0, "completion_length": 517.3125152587891, "epoch": 0.7397360703812317, "grad_norm": 5.211723665878622, "kl": 1.21875, "learning_rate": 2.7343682091684863e-07, "loss": 0.0202, "reward": 2.127828001976013, "reward_std": 0.2590429000556469, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.019741501659154892, "rewards/tag_count_reward": 0.953125, "step": 1009 }, { "clip_ratio": 0.0, "completion_length": 465.3958435058594, "epoch": 0.7404692082111437, "grad_norm": 51.62344516592479, "kl": 3.9140625, "learning_rate": 2.725288356422187e-07, "loss": 0.3732, "reward": 2.1338714361190796, "reward_std": 0.23930185288190842, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01717032492160797, "rewards/tag_count_reward": 0.9427083432674408, "step": 1010 }, { "clip_ratio": 0.0, "completion_length": 491.6458435058594, "epoch": 0.7412023460410557, "grad_norm": 8.872840652024472, "kl": 0.521484375, "learning_rate": 2.7162266934890975e-07, "loss": 0.0284, "reward": 2.226145088672638, "reward_std": 0.008340620435774326, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023855021223425865, "rewards/tag_count_reward": 1.0, "step": 1011 }, { "clip_ratio": 0.0, "completion_length": 477.5833435058594, "epoch": 0.7419354838709677, "grad_norm": 12.230811361407058, "kl": 0.763671875, "learning_rate": 2.707183279773569e-07, "loss": 0.0509, "reward": 1.968254268169403, "reward_std": 0.043425518088042736, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021329134702682495, "rewards/tag_count_reward": 0.9895833432674408, "step": 1012 }, { "clip_ratio": 0.0, "completion_length": 555.4583435058594, "epoch": 0.7426686217008798, "grad_norm": 29.90738464709304, "kl": 3.765625, "learning_rate": 2.698158174560326e-07, "loss": 0.838, "reward": 2.151626408100128, "reward_std": 0.2698955833911896, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02024863287806511, "rewards/tag_count_reward": 0.9427083730697632, "step": 1013 }, { "clip_ratio": 0.0, "completion_length": 436.8333435058594, "epoch": 0.7434017595307918, "grad_norm": 13.228223995097775, "kl": 1.28125, "learning_rate": 2.6891514370140645e-07, "loss": 0.1214, "reward": 1.9460716843605042, "reward_std": 0.11848503537476063, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.024414503015577793, "rewards/tag_count_reward": 0.984375, "step": 1014 }, { "clip_ratio": 0.0, "completion_length": 488.75001525878906, "epoch": 0.7441348973607038, "grad_norm": 10.713414373967986, "kl": 1.26953125, "learning_rate": 2.6801631261790747e-07, "loss": 0.2509, "reward": 2.429144859313965, "reward_std": 0.17435809317976236, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.022244098596274853, "rewards/tag_count_reward": 0.9791666865348816, "step": 1015 }, { "clip_ratio": 0.0, "completion_length": 481.56251525878906, "epoch": 0.7448680351906158, "grad_norm": 35.03331056974067, "kl": 3.734375, "learning_rate": 2.6711933009788443e-07, "loss": 0.2651, "reward": 2.333442807197571, "reward_std": 0.3207738474011421, "rewards/accuracy_reward": 0.4166666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.024196190759539604, "rewards/tag_count_reward": 0.9479166865348816, "step": 1016 }, { "clip_ratio": 0.0, "completion_length": 505.41668701171875, "epoch": 0.7456011730205279, "grad_norm": 11.324692674530787, "kl": 3.4765625, "learning_rate": 2.6622420202156843e-07, "loss": 0.4659, "reward": 2.356720805168152, "reward_std": 0.36550724506378174, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02522371429949999, "rewards/tag_count_reward": 0.9166666865348816, "step": 1017 }, { "clip_ratio": 0.0, "completion_length": 460.25001525878906, "epoch": 0.7463343108504399, "grad_norm": 15.163042787217018, "kl": 1.9248046875, "learning_rate": 2.653309342570327e-07, "loss": 0.2913, "reward": 2.355245351791382, "reward_std": 0.31433626264333725, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02843527588993311, "rewards/tag_count_reward": 0.9531250298023224, "step": 1018 }, { "clip_ratio": 0.0, "completion_length": 464.9166717529297, "epoch": 0.7470674486803519, "grad_norm": 11.573884174204364, "kl": 1.68359375, "learning_rate": 2.64439532660156e-07, "loss": 0.1937, "reward": 2.401885986328125, "reward_std": 0.2557380646467209, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.02519738022238016, "rewards/tag_count_reward": 0.9687500298023224, "step": 1019 }, { "clip_ratio": 0.0, "completion_length": 499.10418701171875, "epoch": 0.7478005865102639, "grad_norm": 3.365385207956817, "kl": 0.3740234375, "learning_rate": 2.6355000307458253e-07, "loss": 0.0359, "reward": 2.2115240693092346, "reward_std": 0.07929915469139814, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017642635852098465, "rewards/tag_count_reward": 1.0, "step": 1020 }, { "clip_ratio": 0.0, "completion_length": 443.5416717529297, "epoch": 0.748533724340176, "grad_norm": 5.113590914631466, "kl": 1.4794921875, "learning_rate": 2.6266235133168503e-07, "loss": 0.0197, "reward": 2.3895903825759888, "reward_std": 0.21832578629255295, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.021868057548999786, "rewards/tag_count_reward": 0.9531250298023224, "step": 1021 }, { "clip_ratio": 0.0, "completion_length": 520.0625152587891, "epoch": 0.749266862170088, "grad_norm": 11.69931205695156, "kl": 2.16796875, "learning_rate": 2.617765832505253e-07, "loss": 0.3627, "reward": 2.5873245000839233, "reward_std": 0.2989402636885643, "rewards/accuracy_reward": 0.6875, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02378668077290058, "rewards/tag_count_reward": 0.9375000298023224, "step": 1022 }, { "clip_ratio": 0.0, "completion_length": 450.3541717529297, "epoch": 0.75, "grad_norm": 4.454764934413698, "kl": 0.4326171875, "learning_rate": 2.608927046378173e-07, "loss": 0.061, "reward": 2.2236390113830566, "reward_std": 0.01102450443431735, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02636106964200735, "rewards/tag_count_reward": 1.0, "step": 1023 }, { "clip_ratio": 0.0, "completion_length": 465.41668701171875, "epoch": 0.750733137829912, "grad_norm": 6.766283640737552, "kl": 1.3330078125, "learning_rate": 2.6001072128788795e-07, "loss": 0.0486, "reward": 2.1870197653770447, "reward_std": 0.11883579567074776, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.017841434106230736, "rewards/tag_count_reward": 0.9687500298023224, "step": 1024 }, { "clip_ratio": 0.0, "completion_length": 500.75001525878906, "epoch": 0.751466275659824, "grad_norm": 11.468805888423775, "kl": 2.26953125, "learning_rate": 2.5913063898263975e-07, "loss": 0.4147, "reward": 2.1911988258361816, "reward_std": 0.11320665059611201, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.015398475807160139, "rewards/tag_count_reward": 0.9635416865348816, "step": 1025 }, { "clip_ratio": 0.0, "completion_length": 546.2291870117188, "epoch": 0.7521994134897361, "grad_norm": 62.06601803645074, "kl": 4.2333984375, "learning_rate": 2.5825246349151296e-07, "loss": 0.3466, "reward": 2.1409623622894287, "reward_std": 0.19636456668376923, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.020496048033237457, "rewards/tag_count_reward": 0.9322916865348816, "step": 1026 }, { "clip_ratio": 0.0, "completion_length": 469.8958435058594, "epoch": 0.7529325513196481, "grad_norm": 6.915256602758481, "kl": 1.46484375, "learning_rate": 2.573762005714474e-07, "loss": 0.084, "reward": 1.950873076915741, "reward_std": 0.1016916565131396, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.0213491627946496, "rewards/tag_count_reward": 0.9791666865348816, "step": 1027 }, { "clip_ratio": 0.0, "completion_length": 526.2500152587891, "epoch": 0.7536656891495601, "grad_norm": 15.895006895352855, "kl": 1.791015625, "learning_rate": 2.5650185596684513e-07, "loss": 0.1532, "reward": 1.9461398124694824, "reward_std": 0.21893376857042313, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.020874238573014736, "rewards/tag_count_reward": 0.953125, "step": 1028 }, { "clip_ratio": 0.0, "completion_length": 481.3541717529297, "epoch": 0.7543988269794721, "grad_norm": 10.156286209074864, "kl": 1.3154296875, "learning_rate": 2.556294354095321e-07, "loss": 0.154, "reward": 2.8607027530670166, "reward_std": 0.29952816385775805, "rewards/accuracy_reward": 0.9583333432674408, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.02992249419912696, "rewards/tag_count_reward": 0.9739583432674408, "step": 1029 }, { "clip_ratio": 0.0, "completion_length": 435.7083435058594, "epoch": 0.7551319648093842, "grad_norm": 11.155712646658968, "kl": 0.6943359375, "learning_rate": 2.547589446187217e-07, "loss": 0.0538, "reward": 2.7281389236450195, "reward_std": 0.01100248470902443, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021861168555915356, "rewards/tag_count_reward": 1.0, "step": 1030 }, { "clip_ratio": 0.0, "completion_length": 505.25001525878906, "epoch": 0.7558651026392962, "grad_norm": 20.534972400969437, "kl": 3.484375, "learning_rate": 2.5389038930097594e-07, "loss": 0.44, "reward": 2.0344144701957703, "reward_std": 0.44953444600105286, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.028085462749004364, "rewards/tag_count_reward": 0.8750000298023224, "step": 1031 }, { "clip_ratio": 0.0, "completion_length": 408.0208435058594, "epoch": 0.7565982404692082, "grad_norm": 6.69655925623942, "kl": 0.875, "learning_rate": 2.5302377515016916e-07, "loss": 0.0028, "reward": 2.4262704849243164, "reward_std": 0.16276609152555466, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.028590744361281395, "rewards/tag_count_reward": 0.9895833432674408, "step": 1032 }, { "clip_ratio": 0.0, "completion_length": 479.62501525878906, "epoch": 0.7573313782991202, "grad_norm": 23.560206724875798, "kl": 1.322265625, "learning_rate": 2.5215910784745e-07, "loss": 0.327, "reward": 2.4021449089050293, "reward_std": 0.27173662185668945, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019729983061552048, "rewards/tag_count_reward": 0.9635416865348816, "step": 1033 }, { "clip_ratio": 0.0, "completion_length": 493.91668701171875, "epoch": 0.7580645161290323, "grad_norm": 8.901515825557974, "kl": 0.66796875, "learning_rate": 2.5129639306120467e-07, "loss": 0.0747, "reward": 2.471045136451721, "reward_std": 0.03157370677217841, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018538357689976692, "rewards/tag_count_reward": 0.9895833432674408, "step": 1034 }, { "clip_ratio": 0.0, "completion_length": 487.93751525878906, "epoch": 0.7587976539589443, "grad_norm": 12.897795206877362, "kl": 2.890625, "learning_rate": 2.5043563644701905e-07, "loss": 0.4459, "reward": 2.2605895400047302, "reward_std": 0.49446502327919006, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 0.9583334028720856, "rewards/repetition_penalty_reward": -0.02586888149380684, "rewards/tag_count_reward": 0.9114583432674408, "step": 1035 }, { "clip_ratio": 0.0, "completion_length": 491.75001525878906, "epoch": 0.7595307917888563, "grad_norm": 17.104268270251925, "kl": 1.79296875, "learning_rate": 2.495768436476427e-07, "loss": 0.3061, "reward": 2.1887518167495728, "reward_std": 0.13222945481538773, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02478979155421257, "rewards/tag_count_reward": 0.9635416865348816, "step": 1036 }, { "clip_ratio": 0.0, "completion_length": 448.81251525878906, "epoch": 0.7602639296187683, "grad_norm": 13.687511958831577, "kl": 1.87890625, "learning_rate": 2.4872002029295057e-07, "loss": 0.2406, "reward": 2.3408429622650146, "reward_std": 0.30387626588344574, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.020268219523131847, "rewards/tag_count_reward": 0.9375000298023224, "step": 1037 }, { "clip_ratio": 0.0, "completion_length": 424.16668701171875, "epoch": 0.7609970674486803, "grad_norm": 11.657813766508225, "kl": 2.53125, "learning_rate": 2.4786517199990764e-07, "loss": 0.1052, "reward": 2.448213815689087, "reward_std": 0.0886231642216444, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.018800227902829647, "rewards/tag_count_reward": 0.9739583432674408, "step": 1038 }, { "clip_ratio": 0.0, "completion_length": 429.43751525878906, "epoch": 0.7617302052785924, "grad_norm": 27.142165668107076, "kl": 2.048828125, "learning_rate": 2.470123043725304e-07, "loss": 0.1753, "reward": 2.197335362434387, "reward_std": 0.10227534547448158, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02662304975092411, "rewards/tag_count_reward": 0.9739583432674408, "step": 1039 }, { "clip_ratio": 0.0, "completion_length": 479.7291717529297, "epoch": 0.7624633431085044, "grad_norm": 13.619344186767858, "kl": 0.86328125, "learning_rate": 2.4616142300185125e-07, "loss": 0.0615, "reward": 2.9825761318206787, "reward_std": 0.01011534919962287, "rewards/accuracy_reward": 1.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017423998564481735, "rewards/tag_count_reward": 1.0, "step": 1040 }, { "clip_ratio": 0.0, "completion_length": 566.9583740234375, "epoch": 0.7631964809384164, "grad_norm": 11.118122539502124, "kl": 2.19921875, "learning_rate": 2.453125334658819e-07, "loss": 0.4298, "reward": 2.101525664329529, "reward_std": 0.36887478455901146, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9583333730697632, "rewards/repetition_penalty_reward": -0.018265982158482075, "rewards/tag_count_reward": 0.9322916865348816, "step": 1041 }, { "clip_ratio": 0.0, "completion_length": 519.5625305175781, "epoch": 0.7639296187683284, "grad_norm": 40.200766946980536, "kl": 6.421875, "learning_rate": 2.444656413295758e-07, "loss": 0.6816, "reward": 2.028169095516205, "reward_std": 0.4279635548591614, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9375, "rewards/repetition_penalty_reward": -0.01870589703321457, "rewards/tag_count_reward": 0.8802083730697632, "step": 1042 }, { "clip_ratio": 0.0, "completion_length": 440.1666717529297, "epoch": 0.7646627565982405, "grad_norm": 8.382550351469364, "kl": 0.5185546875, "learning_rate": 2.4362075214479296e-07, "loss": 0.0405, "reward": 2.469712793827057, "reward_std": 0.014355432707816362, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030287256464362144, "rewards/tag_count_reward": 1.0, "step": 1043 }, { "clip_ratio": 0.0, "completion_length": 474.8333435058594, "epoch": 0.7653958944281525, "grad_norm": 23.33023618104358, "kl": 1.25, "learning_rate": 2.4277787145026235e-07, "loss": 0.2837, "reward": 2.182273507118225, "reward_std": 0.1675884760916233, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.020851566456258297, "rewards/tag_count_reward": 0.9739583730697632, "step": 1044 }, { "clip_ratio": 0.0, "completion_length": 423.25001525878906, "epoch": 0.7661290322580645, "grad_norm": 8.571737590563789, "kl": 0.4443359375, "learning_rate": 2.4193700477154653e-07, "loss": 0.0521, "reward": 2.2212164402008057, "reward_std": 0.027374879457056522, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02357533946633339, "rewards/tag_count_reward": 0.9947916865348816, "step": 1045 }, { "clip_ratio": 0.0, "completion_length": 467.0416717529297, "epoch": 0.7668621700879765, "grad_norm": 14.989284883435417, "kl": 1.318359375, "learning_rate": 2.4109815762100485e-07, "loss": 0.2389, "reward": 2.384683132171631, "reward_std": 0.3186868131160736, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.023303166031837463, "rewards/tag_count_reward": 0.9635416865348816, "step": 1046 }, { "clip_ratio": 0.0, "completion_length": 460.7083435058594, "epoch": 0.7675953079178885, "grad_norm": 13.662504493852214, "kl": 1.90625, "learning_rate": 2.402613354977577e-07, "loss": 0.2727, "reward": 2.3878710865974426, "reward_std": 0.22365204244852066, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.032267773523926735, "rewards/tag_count_reward": 0.9479166865348816, "step": 1047 }, { "clip_ratio": 0.0, "completion_length": 466.8541717529297, "epoch": 0.7683284457478006, "grad_norm": 13.403401572039542, "kl": 1.337890625, "learning_rate": 2.3942654388764985e-07, "loss": 0.0999, "reward": 2.4051430225372314, "reward_std": 0.1881369361653924, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.028884700499475002, "rewards/tag_count_reward": 0.96875, "step": 1048 }, { "clip_ratio": 0.0, "completion_length": 443.2291717529297, "epoch": 0.7690615835777126, "grad_norm": 7.891190935869175, "kl": 0.5068359375, "learning_rate": 2.385937882632155e-07, "loss": 0.0426, "reward": 2.22906756401062, "reward_std": 0.009660831652581692, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.020932521671056747, "rewards/tag_count_reward": 1.0, "step": 1049 }, { "clip_ratio": 0.0, "completion_length": 490.97918701171875, "epoch": 0.7697947214076246, "grad_norm": 13.667764691578203, "kl": 2.5078125, "learning_rate": 2.377630740836411e-07, "loss": 0.5406, "reward": 2.324502468109131, "reward_std": 0.36973246932029724, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02792811580002308, "rewards/tag_count_reward": 0.9427083730697632, "step": 1050 }, { "clip_ratio": 0.0, "completion_length": 527.2500152587891, "epoch": 0.7705278592375366, "grad_norm": 17.845348751851517, "kl": 2.1953125, "learning_rate": 2.3693440679473106e-07, "loss": 0.3505, "reward": 2.4364585876464844, "reward_std": 0.15930373966693878, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.020138615276664495, "rewards/tag_count_reward": 0.9635416865348816, "step": 1051 }, { "clip_ratio": 0.0, "completion_length": 447.3333435058594, "epoch": 0.7712609970674487, "grad_norm": 15.068120067737828, "kl": 1.03125, "learning_rate": 2.3610779182887044e-07, "loss": 0.0492, "reward": 2.2252615690231323, "reward_std": 0.012322348076850176, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024738566018640995, "rewards/tag_count_reward": 1.0, "step": 1052 }, { "clip_ratio": 0.0, "completion_length": 423.87501525878906, "epoch": 0.7719941348973607, "grad_norm": 23.95294488179518, "kl": 4.4453125, "learning_rate": 2.352832346049909e-07, "loss": 0.2367, "reward": 2.359814405441284, "reward_std": 0.38373930752277374, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02907445840537548, "rewards/tag_count_reward": 0.9375000298023224, "step": 1053 }, { "clip_ratio": 0.0, "completion_length": 530.6041717529297, "epoch": 0.7727272727272727, "grad_norm": 21.090369592035625, "kl": 2.4375, "learning_rate": 2.3446074052853398e-07, "loss": 0.2974, "reward": 2.8810449838638306, "reward_std": 0.25332972407341003, "rewards/accuracy_reward": 0.9375000298023224, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02520513255149126, "rewards/tag_count_reward": 0.96875, "step": 1054 }, { "clip_ratio": 0.0, "completion_length": 545.1041870117188, "epoch": 0.7734604105571847, "grad_norm": 18.541416537922142, "kl": 1.6875, "learning_rate": 2.33640314991416e-07, "loss": 0.3785, "reward": 2.4116270542144775, "reward_std": 0.18292175233364105, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.022400776855647564, "rewards/tag_count_reward": 0.9479166865348816, "step": 1055 }, { "clip_ratio": 0.0, "completion_length": 473.9791717529297, "epoch": 0.7741935483870968, "grad_norm": 7.841288579770671, "kl": 1.44140625, "learning_rate": 2.3282196337199336e-07, "loss": 0.0469, "reward": 2.1994876861572266, "reward_std": 0.0929469121620059, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.02620673179626465, "rewards/tag_count_reward": 0.9895833432674408, "step": 1056 }, { "clip_ratio": 0.0, "completion_length": 446.5208435058594, "epoch": 0.7749266862170088, "grad_norm": 6.869915300476055, "kl": 0.5234375, "learning_rate": 2.3200569103502623e-07, "loss": 0.0616, "reward": 2.4464250206947327, "reward_std": 0.1164440019056201, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022325048223137856, "rewards/tag_count_reward": 0.9895833432674408, "step": 1057 }, { "clip_ratio": 0.0, "completion_length": 511.8958435058594, "epoch": 0.7756598240469208, "grad_norm": 27.238285089186576, "kl": 3.625, "learning_rate": 2.311915033316443e-07, "loss": 0.5597, "reward": 2.585835337638855, "reward_std": 0.3875824511051178, "rewards/accuracy_reward": 0.6875000149011612, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.027012161910533905, "rewards/tag_count_reward": 0.9322916865348816, "step": 1058 }, { "clip_ratio": 0.0, "completion_length": 435.5833435058594, "epoch": 0.7763929618768328, "grad_norm": 15.130806571886888, "kl": 2.33203125, "learning_rate": 2.3037940559931084e-07, "loss": 0.2147, "reward": 2.1360930800437927, "reward_std": 0.21016471646726131, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.028837502002716064, "rewards/tag_count_reward": 0.9635416865348816, "step": 1059 }, { "clip_ratio": 0.0, "completion_length": 472.8333435058594, "epoch": 0.7771260997067448, "grad_norm": 13.573733175581262, "kl": 1.35546875, "learning_rate": 2.2956940316178884e-07, "loss": 0.2755, "reward": 2.1681582927703857, "reward_std": 0.1469826102256775, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.033230654895305634, "rewards/tag_count_reward": 0.9583333730697632, "step": 1060 }, { "clip_ratio": 0.0, "completion_length": 429.9583435058594, "epoch": 0.7778592375366569, "grad_norm": 8.50090498665031, "kl": 1.953125, "learning_rate": 2.2876150132910466e-07, "loss": 0.0629, "reward": 2.407193899154663, "reward_std": 0.23686270415782928, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.02509789913892746, "rewards/tag_count_reward": 0.9739583432674408, "step": 1061 }, { "clip_ratio": 0.0, "completion_length": 465.7708435058594, "epoch": 0.7785923753665689, "grad_norm": 15.772638637934058, "kl": 1.3466796875, "learning_rate": 2.2795570539751462e-07, "loss": 0.275, "reward": 2.4364675879478455, "reward_std": 0.1519809989258647, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02186590526252985, "rewards/tag_count_reward": 0.9791666865348816, "step": 1062 }, { "clip_ratio": 0.0, "completion_length": 489.54168701171875, "epoch": 0.7793255131964809, "grad_norm": 14.481423143656839, "kl": 1.7978515625, "learning_rate": 2.2715202064946925e-07, "loss": 0.3213, "reward": 1.9716586470603943, "reward_std": 0.13082672283053398, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02313299011439085, "rewards/tag_count_reward": 0.9739583432674408, "step": 1063 }, { "clip_ratio": 0.0, "completion_length": 509.83335876464844, "epoch": 0.7800586510263929, "grad_norm": 27.515519930915364, "kl": 2.30859375, "learning_rate": 2.263504523535795e-07, "loss": 0.6025, "reward": 2.1453371047973633, "reward_std": 0.28774622082710266, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.02827406395226717, "rewards/tag_count_reward": 0.9583333730697632, "step": 1064 }, { "clip_ratio": 0.0, "completion_length": 484.87501525878906, "epoch": 0.780791788856305, "grad_norm": 14.499965581307233, "kl": 3.7109375, "learning_rate": 2.255510057645812e-07, "loss": 0.6072, "reward": 2.1432166695594788, "reward_std": 0.28085365891456604, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.026922253891825676, "rewards/tag_count_reward": 0.9479166865348816, "step": 1065 }, { "clip_ratio": 0.0, "completion_length": 458.7291717529297, "epoch": 0.781524926686217, "grad_norm": 52.53300723069954, "kl": 3.0625, "learning_rate": 2.2475368612330182e-07, "loss": 0.2262, "reward": 2.1494187712669373, "reward_std": 0.20825094729661942, "rewards/accuracy_reward": 0.22916667722165585, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02592852246016264, "rewards/tag_count_reward": 0.953125, "step": 1066 }, { "clip_ratio": 0.0, "completion_length": 405.9166717529297, "epoch": 0.782258064516129, "grad_norm": 30.234824161584932, "kl": 3.40625, "learning_rate": 2.2395849865662487e-07, "loss": 0.2387, "reward": 2.6589914560317993, "reward_std": 0.22190771251916885, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.0354531928896904, "rewards/tag_count_reward": 0.9791666865348816, "step": 1067 }, { "clip_ratio": 0.0, "completion_length": 495.91668701171875, "epoch": 0.782991202346041, "grad_norm": 22.759908198553244, "kl": 2.1484375, "learning_rate": 2.2316544857745678e-07, "loss": 0.2113, "reward": 2.4049072265625, "reward_std": 0.19384130463004112, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02217612974345684, "rewards/tag_count_reward": 0.9687500298023224, "step": 1068 }, { "clip_ratio": 0.0, "completion_length": 464.75001525878906, "epoch": 0.783724340175953, "grad_norm": 36.380039815706255, "kl": 3.642578125, "learning_rate": 2.223745410846919e-07, "loss": 0.3783, "reward": 2.373390316963196, "reward_std": 0.24706235527992249, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.025915381498634815, "rewards/tag_count_reward": 0.9479166865348816, "step": 1069 }, { "clip_ratio": 0.0, "completion_length": 437.91668701171875, "epoch": 0.7844574780058651, "grad_norm": 6.394595193309836, "kl": 0.50390625, "learning_rate": 2.2158578136317851e-07, "loss": 0.042, "reward": 2.244118392467499, "reward_std": 0.0806655939668417, "rewards/accuracy_reward": 0.2708333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026714895851910114, "rewards/tag_count_reward": 1.0, "step": 1070 }, { "clip_ratio": 0.0, "completion_length": 489.9583435058594, "epoch": 0.7851906158357771, "grad_norm": 14.072098484553294, "kl": 1.5390625, "learning_rate": 2.2079917458368563e-07, "loss": 0.2618, "reward": 2.6522449254989624, "reward_std": 0.17534173047170043, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024838595651090145, "rewards/tag_count_reward": 0.96875, "step": 1071 }, { "clip_ratio": 0.0, "completion_length": 469.7083435058594, "epoch": 0.7859237536656891, "grad_norm": 15.845583496718644, "kl": 1.81640625, "learning_rate": 2.2001472590286778e-07, "loss": 0.338, "reward": 2.3699461221694946, "reward_std": 0.28435058146715164, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.03283172659575939, "rewards/tag_count_reward": 0.9583333432674408, "step": 1072 }, { "clip_ratio": 0.0, "completion_length": 473.9166717529297, "epoch": 0.7866568914956011, "grad_norm": 28.450144567340768, "kl": 0.958984375, "learning_rate": 2.1923244046323263e-07, "loss": 0.2799, "reward": 2.455487847328186, "reward_std": 0.07996404357254505, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023678927682340145, "rewards/tag_count_reward": 0.9791666865348816, "step": 1073 }, { "clip_ratio": 0.0, "completion_length": 483.3333435058594, "epoch": 0.7873900293255132, "grad_norm": 9.489962973191355, "kl": 0.45703125, "learning_rate": 2.1845232339310594e-07, "loss": 0.0416, "reward": 2.480448365211487, "reward_std": 0.1484880894422531, "rewards/accuracy_reward": 0.5000000204890966, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019551726058125496, "rewards/tag_count_reward": 1.0, "step": 1074 }, { "clip_ratio": 0.0, "completion_length": 451.12501525878906, "epoch": 0.7881231671554252, "grad_norm": 10.244882208914222, "kl": 0.84765625, "learning_rate": 2.1767437980659913e-07, "loss": 0.1509, "reward": 1.9186533689498901, "reward_std": 0.13470640173181891, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.030999451875686646, "rewards/tag_count_reward": 0.9635416865348816, "step": 1075 }, { "clip_ratio": 0.0, "completion_length": 549.3333435058594, "epoch": 0.7888563049853372, "grad_norm": 40.24012373692978, "kl": 2.4765625, "learning_rate": 2.168986148035748e-07, "loss": 0.8641, "reward": 2.33972430229187, "reward_std": 0.34671784937381744, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.033539606258273125, "rewards/tag_count_reward": 0.9218750298023224, "step": 1076 }, { "clip_ratio": 0.0, "completion_length": 437.50001525878906, "epoch": 0.7895894428152492, "grad_norm": 8.330876049338293, "kl": 0.884765625, "learning_rate": 2.161250334696139e-07, "loss": 0.1215, "reward": 2.4110584259033203, "reward_std": 0.1561183100566268, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026441698893904686, "rewards/tag_count_reward": 0.9791666865348816, "step": 1077 }, { "clip_ratio": 0.0, "completion_length": 502.0, "epoch": 0.7903225806451613, "grad_norm": 14.720403395941624, "kl": 1.7216796875, "learning_rate": 2.15353640875982e-07, "loss": 0.3272, "reward": 2.7998095750808716, "reward_std": 0.3225977895781398, "rewards/accuracy_reward": 0.9375, "rewards/reasoning_steps_reward": 0.9513888955116272, "rewards/repetition_penalty_reward": -0.03699600324034691, "rewards/tag_count_reward": 0.9479166865348816, "step": 1078 }, { "clip_ratio": 0.0, "completion_length": 487.6875305175781, "epoch": 0.7910557184750733, "grad_norm": 13.10411383821781, "kl": 1.7822265625, "learning_rate": 2.1458444207959653e-07, "loss": 0.4046, "reward": 2.6844232082366943, "reward_std": 0.10148819629102945, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02911848109215498, "rewards/tag_count_reward": 0.9635416865348816, "step": 1079 }, { "clip_ratio": 0.0, "completion_length": 463.66668701171875, "epoch": 0.7917888563049853, "grad_norm": 12.822024813188964, "kl": 1.4140625, "learning_rate": 2.1381744212299303e-07, "loss": 0.0617, "reward": 2.677112340927124, "reward_std": 0.17961880564689636, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.022540478967130184, "rewards/tag_count_reward": 0.9843750298023224, "step": 1080 }, { "clip_ratio": 0.0, "completion_length": 452.2708435058594, "epoch": 0.7925219941348973, "grad_norm": 8.946363278262771, "kl": 1.1591796875, "learning_rate": 2.1305264603429261e-07, "loss": 0.2552, "reward": 2.203592896461487, "reward_std": 0.09071096312254667, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025573882274329662, "rewards/tag_count_reward": 0.9791666865348816, "step": 1081 }, { "clip_ratio": 0.0, "completion_length": 452.5208435058594, "epoch": 0.7932551319648093, "grad_norm": 20.42295106802756, "kl": 1.91796875, "learning_rate": 2.1229005882716854e-07, "loss": 0.2016, "reward": 2.4216538667678833, "reward_std": 0.1843453124165535, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02626277320086956, "rewards/tag_count_reward": 0.9687500298023224, "step": 1082 }, { "clip_ratio": 0.0, "completion_length": 510.66668701171875, "epoch": 0.7939882697947214, "grad_norm": 12.095799408086036, "kl": 2.515625, "learning_rate": 2.11529685500814e-07, "loss": 0.6492, "reward": 2.6630301475524902, "reward_std": 0.2383480668067932, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.022733934223651886, "rewards/tag_count_reward": 0.9635416865348816, "step": 1083 }, { "clip_ratio": 0.0, "completion_length": 472.3125, "epoch": 0.7947214076246334, "grad_norm": 14.367290877648022, "kl": 1.642578125, "learning_rate": 2.1077153103990854e-07, "loss": 0.3468, "reward": 2.667872428894043, "reward_std": 0.2050323337316513, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.024836037307977676, "rewards/tag_count_reward": 0.984375, "step": 1084 }, { "clip_ratio": 0.0, "completion_length": 446.2708435058594, "epoch": 0.7954545454545454, "grad_norm": 5.5197384607329845, "kl": 0.5810546875, "learning_rate": 2.1001560041458573e-07, "loss": 0.106, "reward": 2.179283857345581, "reward_std": 0.16980974515900016, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.027313435450196266, "rewards/tag_count_reward": 0.984375, "step": 1085 }, { "clip_ratio": 0.0, "completion_length": 469.5833435058594, "epoch": 0.7961876832844574, "grad_norm": 7.0820397306841825, "kl": 0.5576171875, "learning_rate": 2.0926189858040118e-07, "loss": 0.0277, "reward": 2.4565619230270386, "reward_std": 0.1886884942650795, "rewards/accuracy_reward": 0.5000000204890966, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03302140720188618, "rewards/tag_count_reward": 0.9895833432674408, "step": 1086 }, { "clip_ratio": 0.0, "completion_length": 441.9791717529297, "epoch": 0.7969208211143695, "grad_norm": 5.213802139763893, "kl": 0.80859375, "learning_rate": 2.0851043047829886e-07, "loss": 0.0479, "reward": 1.9766491651535034, "reward_std": 0.012979320250451565, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0233508488163352, "rewards/tag_count_reward": 1.0, "step": 1087 }, { "clip_ratio": 0.0, "completion_length": 427.2083435058594, "epoch": 0.7976539589442815, "grad_norm": 6.021152479365784, "kl": 1.365234375, "learning_rate": 2.0776120103457987e-07, "loss": 0.0423, "reward": 2.4225898385047913, "reward_std": 0.19223318248987198, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02359061688184738, "rewards/tag_count_reward": 0.9739583432674408, "step": 1088 }, { "clip_ratio": 0.0, "completion_length": 467.4166717529297, "epoch": 0.7983870967741935, "grad_norm": 13.154645588183993, "kl": 0.888671875, "learning_rate": 2.070142151608692e-07, "loss": 0.1902, "reward": 2.6917331218719482, "reward_std": 0.12653527967631817, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.02354466635733843, "rewards/tag_count_reward": 0.9791666865348816, "step": 1089 }, { "clip_ratio": 0.0, "completion_length": 518.7291717529297, "epoch": 0.7991202346041055, "grad_norm": 44.5268065319015, "kl": 2.80859375, "learning_rate": 2.0626947775408444e-07, "loss": 0.666, "reward": 2.565494179725647, "reward_std": 0.44115468859672546, "rewards/accuracy_reward": 0.6875000149011612, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.04214485548436642, "rewards/tag_count_reward": 0.9479166865348816, "step": 1090 }, { "clip_ratio": 0.0, "completion_length": 456.91668701171875, "epoch": 0.7998533724340176, "grad_norm": 10.596915261641815, "kl": 0.97265625, "learning_rate": 2.0552699369640263e-07, "loss": 0.0155, "reward": 2.1983089447021484, "reward_std": 0.07216011872515082, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025649478659033775, "rewards/tag_count_reward": 0.9739583432674408, "step": 1091 }, { "clip_ratio": 0.0, "completion_length": 498.5208435058594, "epoch": 0.8005865102639296, "grad_norm": 22.398728890888155, "kl": 2.08203125, "learning_rate": 2.0478676785522936e-07, "loss": 0.4324, "reward": 2.3511595726013184, "reward_std": 0.3135446608066559, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.03425721637904644, "rewards/tag_count_reward": 0.96875, "step": 1092 }, { "clip_ratio": 0.0, "completion_length": 406.56251525878906, "epoch": 0.8013196480938416, "grad_norm": 11.634427759894564, "kl": 1.40234375, "learning_rate": 2.0404880508316597e-07, "loss": 0.185, "reward": 2.701943874359131, "reward_std": 0.08190051838755608, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.027222798205912113, "rewards/tag_count_reward": 0.9791666865348816, "step": 1093 }, { "clip_ratio": 0.0, "completion_length": 460.47918701171875, "epoch": 0.8020527859237536, "grad_norm": 31.318190759662002, "kl": 1.11328125, "learning_rate": 2.0331311021797836e-07, "loss": 0.2413, "reward": 2.1923688650131226, "reward_std": 0.12720130756497383, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02290901355445385, "rewards/tag_count_reward": 0.9791666865348816, "step": 1094 }, { "clip_ratio": 0.0, "completion_length": 483.16668701171875, "epoch": 0.8027859237536656, "grad_norm": 16.85003870587138, "kl": 1.560546875, "learning_rate": 2.0257968808256465e-07, "loss": 0.2716, "reward": 2.6920918226242065, "reward_std": 0.1356341354548931, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021449715830385685, "rewards/tag_count_reward": 0.984375, "step": 1095 }, { "clip_ratio": 0.0, "completion_length": 450.8541717529297, "epoch": 0.8035190615835777, "grad_norm": 4.965107999898547, "kl": 0.4130859375, "learning_rate": 2.018485434849243e-07, "loss": 0.0301, "reward": 2.9743878841400146, "reward_std": 0.010763033293187618, "rewards/accuracy_reward": 1.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025612126104533672, "rewards/tag_count_reward": 1.0, "step": 1096 }, { "clip_ratio": 0.0, "completion_length": 444.25, "epoch": 0.8042521994134897, "grad_norm": 11.890604702844437, "kl": 0.79296875, "learning_rate": 2.01119681218126e-07, "loss": 0.1023, "reward": 2.1787749528884888, "reward_std": 0.1856333538889885, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019141847267746925, "rewards/tag_count_reward": 0.9687500298023224, "step": 1097 }, { "clip_ratio": 0.0, "completion_length": 568.7916870117188, "epoch": 0.8049853372434017, "grad_norm": 31.623012706962754, "kl": 5.75390625, "learning_rate": 2.0039310606027664e-07, "loss": 0.7709, "reward": 2.1479861736297607, "reward_std": 0.5531877502799034, "rewards/accuracy_reward": 0.3541666865348816, "rewards/reasoning_steps_reward": 0.9722222089767456, "rewards/repetition_penalty_reward": -0.037777796387672424, "rewards/tag_count_reward": 0.859375, "step": 1098 }, { "clip_ratio": 0.0, "completion_length": 454.5, "epoch": 0.8057184750733137, "grad_norm": 8.212504455950251, "kl": 1.162109375, "learning_rate": 1.996688227744894e-07, "loss": 0.0791, "reward": 2.2392346262931824, "reward_std": 0.1137156393378973, "rewards/accuracy_reward": 0.2708333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02118207886815071, "rewards/tag_count_reward": 0.9895833432674408, "step": 1099 }, { "clip_ratio": 0.0, "completion_length": 455.85418701171875, "epoch": 0.8064516129032258, "grad_norm": 5.727879818508118, "kl": 0.728515625, "learning_rate": 1.989468361088536e-07, "loss": 0.0209, "reward": 2.187276005744934, "reward_std": 0.14227101765573025, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.024529581889510155, "rewards/tag_count_reward": 0.9895833432674408, "step": 1100 }, { "clip_ratio": 0.0, "completion_length": 462.1041717529297, "epoch": 0.8071847507331378, "grad_norm": 17.32722660816832, "kl": 0.80078125, "learning_rate": 1.9822715079640217e-07, "loss": 0.2153, "reward": 2.3942633867263794, "reward_std": 0.2861350402235985, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03108394891023636, "rewards/tag_count_reward": 0.9739583730697632, "step": 1101 }, { "clip_ratio": 0.0, "completion_length": 459.75001525878906, "epoch": 0.8079178885630498, "grad_norm": 8.204192297581018, "kl": 1.5859375, "learning_rate": 1.9750977155508201e-07, "loss": 0.1769, "reward": 2.646689295768738, "reward_std": 0.29864780604839325, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019977413583546877, "rewards/tag_count_reward": 0.9583333730697632, "step": 1102 }, { "clip_ratio": 0.0, "completion_length": 507.83335876464844, "epoch": 0.8086510263929618, "grad_norm": 29.183446868684296, "kl": 2.73046875, "learning_rate": 1.9679470308772193e-07, "loss": 0.5248, "reward": 2.162355661392212, "reward_std": 0.15475745499134064, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.03208877891302109, "rewards/tag_count_reward": 0.9583333730697632, "step": 1103 }, { "clip_ratio": 0.0, "completion_length": 452.1041717529297, "epoch": 0.8093841642228738, "grad_norm": 9.188573023239696, "kl": 1.494140625, "learning_rate": 1.960819500820027e-07, "loss": 0.3051, "reward": 2.201483130455017, "reward_std": 0.06417120806872845, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03289197292178869, "rewards/tag_count_reward": 0.984375, "step": 1104 }, { "clip_ratio": 0.0, "completion_length": 468.66668701171875, "epoch": 0.8101173020527859, "grad_norm": 7.686165194216877, "kl": 1.32421875, "learning_rate": 1.9537151721042544e-07, "loss": 0.1647, "reward": 2.193081557750702, "reward_std": 0.11296498030424118, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02046018559485674, "rewards/tag_count_reward": 0.9635416865348816, "step": 1105 }, { "clip_ratio": 0.0, "completion_length": 480.06251525878906, "epoch": 0.8108504398826979, "grad_norm": 16.66437068456755, "kl": 1.755859375, "learning_rate": 1.94663409130282e-07, "loss": 0.1331, "reward": 2.607347011566162, "reward_std": 0.26773516833782196, "rewards/accuracy_reward": 0.6875000149011612, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.029805734753608704, "rewards/tag_count_reward": 0.9635416865348816, "step": 1106 }, { "clip_ratio": 0.0, "completion_length": 428.1458435058594, "epoch": 0.8115835777126099, "grad_norm": 17.084109852146433, "kl": 0.8701171875, "learning_rate": 1.939576304836235e-07, "loss": 0.25, "reward": 2.6833220720291138, "reward_std": 0.1571330651640892, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.023275235667824745, "rewards/tag_count_reward": 0.984375, "step": 1107 }, { "clip_ratio": 0.0, "completion_length": 438.1458435058594, "epoch": 0.8123167155425219, "grad_norm": 8.723073183829872, "kl": 0.986328125, "learning_rate": 1.9325418589723063e-07, "loss": 0.056, "reward": 2.417007803916931, "reward_std": 0.19728430174291134, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.02743679191917181, "rewards/tag_count_reward": 0.9791666865348816, "step": 1108 }, { "clip_ratio": 0.0, "completion_length": 406.6666717529297, "epoch": 0.8130498533724341, "grad_norm": 6.426501532620271, "kl": 0.560546875, "learning_rate": 1.9255307998258253e-07, "loss": 0.063, "reward": 2.729717493057251, "reward_std": 0.010303615126758814, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.020282641053199768, "rewards/tag_count_reward": 1.0, "step": 1109 }, { "clip_ratio": 0.0, "completion_length": 544.3541717529297, "epoch": 0.8137829912023461, "grad_norm": 14.856144237306745, "kl": 1.06640625, "learning_rate": 1.9185431733582755e-07, "loss": 0.3038, "reward": 1.9218279719352722, "reward_std": 0.19134631752967834, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.02782483585178852, "rewards/tag_count_reward": 0.9635416865348816, "step": 1110 }, { "clip_ratio": 0.0, "completion_length": 474.41668701171875, "epoch": 0.8145161290322581, "grad_norm": 10.786777205070308, "kl": 0.71484375, "learning_rate": 1.9115790253775211e-07, "loss": 0.0898, "reward": 2.22803795337677, "reward_std": 0.012332984246313572, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02196214161813259, "rewards/tag_count_reward": 1.0, "step": 1111 }, { "clip_ratio": 0.0, "completion_length": 444.81251525878906, "epoch": 0.8152492668621701, "grad_norm": 28.863582269415154, "kl": 1.203125, "learning_rate": 1.9046384015375157e-07, "loss": 0.1168, "reward": 2.685865640640259, "reward_std": 0.11857039388269186, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.032884408719837666, "rewards/tag_count_reward": 0.9895833432674408, "step": 1112 }, { "clip_ratio": 0.0, "completion_length": 481.58335876464844, "epoch": 0.8159824046920822, "grad_norm": 17.369555980254702, "kl": 2.251953125, "learning_rate": 1.897721347337995e-07, "loss": 0.35, "reward": 2.158285677433014, "reward_std": 0.2446889877319336, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.022269919514656067, "rewards/tag_count_reward": 0.9583333432674408, "step": 1113 }, { "clip_ratio": 0.0, "completion_length": 469.1041717529297, "epoch": 0.8167155425219942, "grad_norm": 24.713866753619417, "kl": 2.609375, "learning_rate": 1.8908279081241863e-07, "loss": 0.2155, "reward": 2.4129501581192017, "reward_std": 0.19241027534008026, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024549901485443115, "rewards/tag_count_reward": 0.9583333432674408, "step": 1114 }, { "clip_ratio": 0.0, "completion_length": 455.8333435058594, "epoch": 0.8174486803519062, "grad_norm": 5.506746123637378, "kl": 0.396484375, "learning_rate": 1.8839581290865014e-07, "loss": 0.0246, "reward": 2.4741445779800415, "reward_std": 0.012463107705116272, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025855595245957375, "rewards/tag_count_reward": 1.0, "step": 1115 }, { "clip_ratio": 0.0, "completion_length": 474.9791717529297, "epoch": 0.8181818181818182, "grad_norm": 15.558503462858535, "kl": 1.7021484375, "learning_rate": 1.877112055260253e-07, "loss": 0.4739, "reward": 2.3901326656341553, "reward_std": 0.29310342809185386, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.036950779147446156, "rewards/tag_count_reward": 0.96875, "step": 1116 }, { "clip_ratio": 0.0, "completion_length": 495.4791717529297, "epoch": 0.8189149560117303, "grad_norm": 6.5218599318564765, "kl": 1.484375, "learning_rate": 1.8702897315253473e-07, "loss": 0.2184, "reward": 2.606796145439148, "reward_std": 0.29400739073753357, "rewards/accuracy_reward": 0.6875000298023224, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03209282737225294, "rewards/tag_count_reward": 0.9583333432674408, "step": 1117 }, { "clip_ratio": 0.0, "completion_length": 492.0625, "epoch": 0.8196480938416423, "grad_norm": 6.440030855286785, "kl": 0.716796875, "learning_rate": 1.8634912026059996e-07, "loss": 0.0511, "reward": 2.47199022769928, "reward_std": 0.012247168458998203, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028009764850139618, "rewards/tag_count_reward": 1.0, "step": 1118 }, { "clip_ratio": 0.0, "completion_length": 501.3958435058594, "epoch": 0.8203812316715543, "grad_norm": 14.602053842200213, "kl": 2.78515625, "learning_rate": 1.8567165130704314e-07, "loss": 0.4461, "reward": 2.591174006462097, "reward_std": 0.3254464864730835, "rewards/accuracy_reward": 0.6666666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023409449495375156, "rewards/tag_count_reward": 0.9479166865348816, "step": 1119 }, { "clip_ratio": 0.0, "completion_length": 477.1250305175781, "epoch": 0.8211143695014663, "grad_norm": 13.113340234764465, "kl": 0.970703125, "learning_rate": 1.849965707330589e-07, "loss": 0.2927, "reward": 2.440806031227112, "reward_std": 0.12088531628251076, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03141636308282614, "rewards/tag_count_reward": 0.9791666865348816, "step": 1120 }, { "clip_ratio": 0.0, "completion_length": 451.1875, "epoch": 0.8218475073313783, "grad_norm": 9.162423347156384, "kl": 0.515625, "learning_rate": 1.8432388296418411e-07, "loss": 0.0294, "reward": 2.6752790212631226, "reward_std": 0.13515725173056126, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0226375968195498, "rewards/tag_count_reward": 0.9895833432674408, "step": 1121 }, { "clip_ratio": 0.0, "completion_length": 489.1041717529297, "epoch": 0.8225806451612904, "grad_norm": 17.42462161111345, "kl": 3.28125, "learning_rate": 1.836535924102701e-07, "loss": 0.5659, "reward": 2.626253128051758, "reward_std": 0.2911444902420044, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.03173297643661499, "rewards/tag_count_reward": 0.9635416865348816, "step": 1122 }, { "clip_ratio": 0.0, "completion_length": 473.6041717529297, "epoch": 0.8233137829912024, "grad_norm": 7.8820590601670055, "kl": 0.8828125, "learning_rate": 1.8298570346545247e-07, "loss": 0.2549, "reward": 1.948028564453125, "reward_std": 0.10797865828499198, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.031137985177338123, "rewards/tag_count_reward": 0.9791666865348816, "step": 1123 }, { "clip_ratio": 0.0, "completion_length": 422.91668701171875, "epoch": 0.8240469208211144, "grad_norm": 4.156410874608448, "kl": 0.3779296875, "learning_rate": 1.823202205081234e-07, "loss": 0.0052, "reward": 2.8767616748809814, "reward_std": 0.20274555077776313, "rewards/accuracy_reward": 0.9375, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.020807928405702114, "rewards/tag_count_reward": 0.9739583432674408, "step": 1124 }, { "clip_ratio": 0.0, "completion_length": 436.62501525878906, "epoch": 0.8247800586510264, "grad_norm": 6.798128573391826, "kl": 1.0078125, "learning_rate": 1.81657147900902e-07, "loss": 0.2609, "reward": 2.441956400871277, "reward_std": 0.24651040136814117, "rewards/accuracy_reward": 0.5000000204890966, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0372103713452816, "rewards/tag_count_reward": 0.9791666865348816, "step": 1125 }, { "clip_ratio": 0.0, "completion_length": 521.0625305175781, "epoch": 0.8255131964809385, "grad_norm": 18.921212456734796, "kl": 2.3857421875, "learning_rate": 1.8099648999060658e-07, "loss": 0.3091, "reward": 2.5876940488815308, "reward_std": 0.265332761220634, "rewards/accuracy_reward": 0.6875, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.0320977782830596, "rewards/tag_count_reward": 0.953125, "step": 1126 }, { "clip_ratio": 0.0, "completion_length": 435.4791717529297, "epoch": 0.8262463343108505, "grad_norm": 11.468478482896765, "kl": 0.8310546875, "learning_rate": 1.8033825110822542e-07, "loss": 0.1217, "reward": 2.2099719047546387, "reward_std": 0.07140774186700583, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024402983486652374, "rewards/tag_count_reward": 0.984375, "step": 1127 }, { "clip_ratio": 0.0, "completion_length": 561.7708587646484, "epoch": 0.8269794721407625, "grad_norm": 20.46432385173581, "kl": 2.1875, "learning_rate": 1.7968243556888903e-07, "loss": 0.4326, "reward": 1.8412283658981323, "reward_std": 0.23730798065662384, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.056341132149100304, "rewards/tag_count_reward": 0.9322916865348816, "step": 1128 }, { "clip_ratio": 0.0, "completion_length": 422.4791717529297, "epoch": 0.8277126099706745, "grad_norm": 7.042971822737819, "kl": 0.49609375, "learning_rate": 1.7902904767184125e-07, "loss": 0.0169, "reward": 2.967703938484192, "reward_std": 0.013247686438262463, "rewards/accuracy_reward": 1.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03229622729122639, "rewards/tag_count_reward": 1.0, "step": 1129 }, { "clip_ratio": 0.0, "completion_length": 485.95835876464844, "epoch": 0.8284457478005866, "grad_norm": 12.848811794663536, "kl": 1.41015625, "learning_rate": 1.7837809170041112e-07, "loss": 0.1916, "reward": 2.4559115171432495, "reward_std": 0.0645798435434699, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02325524017214775, "rewards/tag_count_reward": 0.9791666865348816, "step": 1130 }, { "clip_ratio": 0.0, "completion_length": 431.8958435058594, "epoch": 0.8291788856304986, "grad_norm": 10.31962203088763, "kl": 1.35546875, "learning_rate": 1.7772957192198578e-07, "loss": 0.1319, "reward": 2.4423160552978516, "reward_std": 0.12352583277970552, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.022961710579693317, "rewards/tag_count_reward": 0.9791666865348816, "step": 1131 }, { "clip_ratio": 0.0, "completion_length": 645.1250305175781, "epoch": 0.8299120234604106, "grad_norm": 25.376081021002964, "kl": 2.890625, "learning_rate": 1.77083492587981e-07, "loss": 0.704, "reward": 2.2698137760162354, "reward_std": 0.4809025526046753, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.049630703404545784, "rewards/tag_count_reward": 0.8958333730697632, "step": 1132 }, { "clip_ratio": 0.0, "completion_length": 480.9166717529297, "epoch": 0.8306451612903226, "grad_norm": 19.581286588231446, "kl": 2.84375, "learning_rate": 1.7643985793381443e-07, "loss": 0.5377, "reward": 2.567038655281067, "reward_std": 0.5014015734195709, "rewards/accuracy_reward": 0.6875000149011612, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.03886430896818638, "rewards/tag_count_reward": 0.953125, "step": 1133 }, { "clip_ratio": 0.0, "completion_length": 437.8333435058594, "epoch": 0.8313782991202346, "grad_norm": 6.4055923974991735, "kl": 0.9765625, "learning_rate": 1.7579867217887734e-07, "loss": 0.0771, "reward": 2.2104244232177734, "reward_std": 0.044318219646811485, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.029159046709537506, "rewards/tag_count_reward": 0.9895833432674408, "step": 1134 }, { "clip_ratio": 0.0, "completion_length": 480.25001525878906, "epoch": 0.8321114369501467, "grad_norm": 16.04432271166462, "kl": 2.5546875, "learning_rate": 1.7515993952650716e-07, "loss": 0.3801, "reward": 2.606268286705017, "reward_std": 0.38931509852409363, "rewards/accuracy_reward": 0.6875000149011612, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.032620541751384735, "rewards/tag_count_reward": 0.9583333730697632, "step": 1135 }, { "clip_ratio": 0.0, "completion_length": 536.2916870117188, "epoch": 0.8328445747800587, "grad_norm": 18.81150662725805, "kl": 3.421875, "learning_rate": 1.745236641639596e-07, "loss": 0.4975, "reward": 2.0818097591400146, "reward_std": 0.31087224930524826, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.024093182757496834, "rewards/tag_count_reward": 0.9322916865348816, "step": 1136 }, { "clip_ratio": 0.0, "completion_length": 529.5625, "epoch": 0.8335777126099707, "grad_norm": 72.73956524353684, "kl": 4.28515625, "learning_rate": 1.7388985026238179e-07, "loss": 0.6823, "reward": 2.3693403005599976, "reward_std": 0.247810211032629, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.0369098074734211, "rewards/tag_count_reward": 0.9479166865348816, "step": 1137 }, { "clip_ratio": 0.0, "completion_length": 464.375, "epoch": 0.8343108504398827, "grad_norm": 4.609364870013891, "kl": 0.486328125, "learning_rate": 1.7325850197678425e-07, "loss": 0.0561, "reward": 2.4805774688720703, "reward_std": 0.009202238637953997, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01942244078963995, "rewards/tag_count_reward": 1.0, "step": 1138 }, { "clip_ratio": 0.0, "completion_length": 436.91668701171875, "epoch": 0.8350439882697948, "grad_norm": 7.809025581477368, "kl": 0.7265625, "learning_rate": 1.726296234460143e-07, "loss": 0.0077, "reward": 2.4543532133102417, "reward_std": 0.10273497458547354, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.017868993803858757, "rewards/tag_count_reward": 1.0, "step": 1139 }, { "clip_ratio": 0.0, "completion_length": 426.6666717529297, "epoch": 0.8357771260997068, "grad_norm": 6.403930601129145, "kl": 0.5244140625, "learning_rate": 1.7200321879272821e-07, "loss": 0.0159, "reward": 2.459131121635437, "reward_std": 0.08256103657186031, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.020035718102008104, "rewards/tag_count_reward": 1.0, "step": 1140 }, { "clip_ratio": 0.0, "completion_length": 458.10418701171875, "epoch": 0.8365102639296188, "grad_norm": 3.918762150296133, "kl": 0.5322265625, "learning_rate": 1.7137929212336504e-07, "loss": 0.038, "reward": 2.2210280895233154, "reward_std": 0.010795718058943748, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028971992433071136, "rewards/tag_count_reward": 1.0, "step": 1141 }, { "clip_ratio": 0.0, "completion_length": 413.56251525878906, "epoch": 0.8372434017595308, "grad_norm": 8.59347643879693, "kl": 0.3896484375, "learning_rate": 1.7075784752811885e-07, "loss": 0.0366, "reward": 2.4820199012756348, "reward_std": 0.0108394306153059, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.017980179749429226, "rewards/tag_count_reward": 1.0, "step": 1142 }, { "clip_ratio": 0.0, "completion_length": 421.97918701171875, "epoch": 0.8379765395894428, "grad_norm": 14.774468827305876, "kl": 1.431640625, "learning_rate": 1.7013888908091262e-07, "loss": 0.1838, "reward": 1.9469353556632996, "reward_std": 0.08503856789320707, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.02355077490210533, "rewards/tag_count_reward": 0.984375, "step": 1143 }, { "clip_ratio": 0.0, "completion_length": 452.43751525878906, "epoch": 0.8387096774193549, "grad_norm": 58.13871324922508, "kl": 2.9189453125, "learning_rate": 1.6952242083937097e-07, "loss": 0.1956, "reward": 2.162394165992737, "reward_std": 0.1664935452863574, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.023369694128632545, "rewards/tag_count_reward": 0.9635416865348816, "step": 1144 }, { "clip_ratio": 0.0, "completion_length": 459.2708435058594, "epoch": 0.8394428152492669, "grad_norm": 8.424296203656573, "kl": 1.720703125, "learning_rate": 1.6890844684479386e-07, "loss": 0.1905, "reward": 2.372523546218872, "reward_std": 0.37988629192113876, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9722222089767456, "rewards/repetition_penalty_reward": -0.037198787555098534, "rewards/tag_count_reward": 0.9791666865348816, "step": 1145 }, { "clip_ratio": 0.0, "completion_length": 493.2708435058594, "epoch": 0.8401759530791789, "grad_norm": 22.816794127705354, "kl": 3.078125, "learning_rate": 1.6829697112213025e-07, "loss": 0.3806, "reward": 2.0439147353172302, "reward_std": 0.3693936914205551, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.025529890321195126, "rewards/tag_count_reward": 0.8958333432674408, "step": 1146 }, { "clip_ratio": 0.0, "completion_length": 508.8958435058594, "epoch": 0.8409090909090909, "grad_norm": 12.890399440726029, "kl": 3.3203125, "learning_rate": 1.676879976799511e-07, "loss": 0.462, "reward": 2.3762617111206055, "reward_std": 0.2129766345024109, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.02478000707924366, "rewards/tag_count_reward": 0.921875, "step": 1147 }, { "clip_ratio": 0.0, "completion_length": 478.85418701171875, "epoch": 0.841642228739003, "grad_norm": 19.300394028055656, "kl": 1.95703125, "learning_rate": 1.6708153051042396e-07, "loss": 0.3214, "reward": 2.646037220954895, "reward_std": 0.2798616886138916, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.024101640097796917, "rewards/tag_count_reward": 0.9687500298023224, "step": 1148 }, { "clip_ratio": 0.0, "completion_length": 464.0208435058594, "epoch": 0.842375366568915, "grad_norm": 37.132576967657364, "kl": 1.580078125, "learning_rate": 1.6647757358928588e-07, "loss": 0.3186, "reward": 2.4220274090766907, "reward_std": 0.19371453439816833, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.027625439688563347, "rewards/tag_count_reward": 0.9635416865348816, "step": 1149 }, { "clip_ratio": 0.0, "completion_length": 486.12501525878906, "epoch": 0.843108504398827, "grad_norm": 15.356178665631093, "kl": 1.7890625, "learning_rate": 1.6587613087581832e-07, "loss": 0.2237, "reward": 2.3487614393234253, "reward_std": 0.4307193011045456, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.036655270494520664, "rewards/tag_count_reward": 0.9479166865348816, "step": 1150 }, { "clip_ratio": 0.0, "completion_length": 461.18751525878906, "epoch": 0.843841642228739, "grad_norm": 15.508356091158527, "kl": 2.2734375, "learning_rate": 1.652772063128201e-07, "loss": 0.5044, "reward": 2.67854905128479, "reward_std": 0.16638881852850318, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0245761526748538, "rewards/tag_count_reward": 0.9739583432674408, "step": 1151 }, { "clip_ratio": 0.0, "completion_length": 557.3125305175781, "epoch": 0.844574780058651, "grad_norm": 11.973782162720196, "kl": 2.30078125, "learning_rate": 1.6468080382658273e-07, "loss": 0.5168, "reward": 2.1508249044418335, "reward_std": 0.2192765325307846, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0210501654073596, "rewards/tag_count_reward": 0.9635416865348816, "step": 1152 }, { "clip_ratio": 0.0, "completion_length": 468.2083435058594, "epoch": 0.8453079178885631, "grad_norm": 8.18992294263394, "kl": 0.724609375, "learning_rate": 1.6408692732686346e-07, "loss": 0.0572, "reward": 2.428075909614563, "reward_std": 0.14463306940160692, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.014632577076554298, "rewards/tag_count_reward": 0.984375, "step": 1153 }, { "clip_ratio": 0.0, "completion_length": 436.7083435058594, "epoch": 0.8460410557184751, "grad_norm": 8.203490331986044, "kl": 1.4375, "learning_rate": 1.634955807068609e-07, "loss": 0.2517, "reward": 2.4469761848449707, "reward_std": 0.1210167994722724, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.0304544810205698, "rewards/tag_count_reward": 0.984375, "step": 1154 }, { "clip_ratio": 0.0, "completion_length": 425.9583435058594, "epoch": 0.8467741935483871, "grad_norm": 14.864662992722334, "kl": 0.638671875, "learning_rate": 1.6290676784318818e-07, "loss": 0.0969, "reward": 2.7268673181533813, "reward_std": 0.011194499209523201, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02313273958861828, "rewards/tag_count_reward": 1.0, "step": 1155 }, { "clip_ratio": 0.0, "completion_length": 498.60418701171875, "epoch": 0.8475073313782991, "grad_norm": 86.480628341073, "kl": 5.9453125, "learning_rate": 1.6232049259584885e-07, "loss": 0.4999, "reward": 2.3408678770065308, "reward_std": 0.2514855843037367, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.027187674306333065, "rewards/tag_count_reward": 0.9375, "step": 1156 }, { "clip_ratio": 0.0, "completion_length": 444.50001525878906, "epoch": 0.8482404692082112, "grad_norm": 10.416530700300191, "kl": 1.046875, "learning_rate": 1.6173675880821045e-07, "loss": 0.0737, "reward": 2.2128702998161316, "reward_std": 0.05877932719886303, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02150482777506113, "rewards/tag_count_reward": 0.9843750298023224, "step": 1157 }, { "clip_ratio": 0.0, "completion_length": 502.81251525878906, "epoch": 0.8489736070381232, "grad_norm": 31.385957646206187, "kl": 3.5703125, "learning_rate": 1.6115557030698022e-07, "loss": 0.6985, "reward": 1.8645765781402588, "reward_std": 0.3231193572282791, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.057298444211483, "rewards/tag_count_reward": 0.9427083432674408, "step": 1158 }, { "clip_ratio": 0.0, "completion_length": 482.45835876464844, "epoch": 0.8497067448680352, "grad_norm": 34.994146803580094, "kl": 3.578125, "learning_rate": 1.6057693090217925e-07, "loss": 0.3718, "reward": 2.401385545730591, "reward_std": 0.24192781373858452, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.03437837865203619, "rewards/tag_count_reward": 0.9635416865348816, "step": 1159 }, { "clip_ratio": 0.0, "completion_length": 449.75, "epoch": 0.8504398826979472, "grad_norm": 6.489234968620317, "kl": 0.603515625, "learning_rate": 1.6000084438711793e-07, "loss": 0.0343, "reward": 1.9903408288955688, "reward_std": 0.09549414180219173, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025284183211624622, "rewards/tag_count_reward": 0.9947916865348816, "step": 1160 }, { "clip_ratio": 0.0, "completion_length": 468.4791717529297, "epoch": 0.8511730205278593, "grad_norm": 10.677103082846811, "kl": 1.9345703125, "learning_rate": 1.5942731453837122e-07, "loss": 0.3723, "reward": 2.1833943724632263, "reward_std": 0.17482627276331186, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.017994558438658714, "rewards/tag_count_reward": 0.9791666865348816, "step": 1161 }, { "clip_ratio": 0.0, "completion_length": 465.37501525878906, "epoch": 0.8519061583577713, "grad_norm": 6.559132148219684, "kl": 0.4189453125, "learning_rate": 1.5885634511575347e-07, "loss": 0.0566, "reward": 2.4819835424423218, "reward_std": 0.01208042073994875, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01801648736000061, "rewards/tag_count_reward": 1.0, "step": 1162 }, { "clip_ratio": 0.0, "completion_length": 513.2083435058594, "epoch": 0.8526392961876833, "grad_norm": 16.350981255664664, "kl": 2.6640625, "learning_rate": 1.5828793986229406e-07, "loss": 0.8073, "reward": 2.545455813407898, "reward_std": 0.5733033120632172, "rewards/accuracy_reward": 0.6666666865348816, "rewards/reasoning_steps_reward": 0.9652778506278992, "rewards/repetition_penalty_reward": -0.034405382350087166, "rewards/tag_count_reward": 0.9479166865348816, "step": 1163 }, { "clip_ratio": 0.0, "completion_length": 552.0000305175781, "epoch": 0.8533724340175953, "grad_norm": 16.641164201933258, "kl": 2.689453125, "learning_rate": 1.5772210250421265e-07, "loss": 0.3053, "reward": 2.5371824502944946, "reward_std": 0.40333325043320656, "rewards/accuracy_reward": 0.6458333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.025317718740552664, "rewards/tag_count_reward": 0.9375, "step": 1164 }, { "clip_ratio": 0.0, "completion_length": 434.3541717529297, "epoch": 0.8541055718475073, "grad_norm": 8.346505801510467, "kl": 0.53515625, "learning_rate": 1.5715883675089524e-07, "loss": 0.0462, "reward": 2.4680683612823486, "reward_std": 0.017134539783000946, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.031931765377521515, "rewards/tag_count_reward": 1.0, "step": 1165 }, { "clip_ratio": 0.0, "completion_length": 487.125, "epoch": 0.8548387096774194, "grad_norm": 17.46249875334204, "kl": 1.2744140625, "learning_rate": 1.5659814629486896e-07, "loss": 0.2772, "reward": 1.9583578705787659, "reward_std": 0.07822929974645376, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02080879732966423, "rewards/tag_count_reward": 0.9791666865348816, "step": 1166 }, { "clip_ratio": 0.0, "completion_length": 453.79168701171875, "epoch": 0.8555718475073314, "grad_norm": 15.376964004491224, "kl": 1.38671875, "learning_rate": 1.56040034811779e-07, "loss": 0.2414, "reward": 2.2069084644317627, "reward_std": 0.08541684225201607, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.0205221064388752, "rewards/tag_count_reward": 0.984375, "step": 1167 }, { "clip_ratio": 0.0, "completion_length": 449.31251525878906, "epoch": 0.8563049853372434, "grad_norm": 9.078560179362256, "kl": 1.28125, "learning_rate": 1.5548450596036328e-07, "loss": 0.0469, "reward": 2.198473036289215, "reward_std": 0.08828200493007898, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.022013090550899506, "rewards/tag_count_reward": 0.984375, "step": 1168 }, { "clip_ratio": 0.0, "completion_length": 455.87501525878906, "epoch": 0.8570381231671554, "grad_norm": 13.56245035458174, "kl": 0.8818359375, "learning_rate": 1.5493156338242971e-07, "loss": 0.1281, "reward": 2.4598007202148438, "reward_std": 0.07951221195980906, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019365964457392693, "rewards/tag_count_reward": 0.9791666865348816, "step": 1169 }, { "clip_ratio": 0.0, "completion_length": 616.9791870117188, "epoch": 0.8577712609970675, "grad_norm": 26.531238059381693, "kl": 3.109375, "learning_rate": 1.5438121070283128e-07, "loss": 0.5462, "reward": 2.1935884952545166, "reward_std": 0.5035333782434464, "rewards/accuracy_reward": 0.3750000149011612, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.05814778432250023, "rewards/tag_count_reward": 0.9114583432674408, "step": 1170 }, { "clip_ratio": 0.0, "completion_length": 439.0833435058594, "epoch": 0.8585043988269795, "grad_norm": 14.122971665456385, "kl": 1.74609375, "learning_rate": 1.5383345152944294e-07, "loss": 0.2753, "reward": 2.3982787132263184, "reward_std": 0.27669892460107803, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.023596427403390408, "rewards/tag_count_reward": 0.9635416865348816, "step": 1171 }, { "clip_ratio": 0.0, "completion_length": 473.4583435058594, "epoch": 0.8592375366568915, "grad_norm": 13.759082348025947, "kl": 1.82421875, "learning_rate": 1.5328828945313764e-07, "loss": 0.3198, "reward": 2.3727020025253296, "reward_std": 0.26478735730051994, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.03354795090854168, "rewards/tag_count_reward": 0.9687500298023224, "step": 1172 }, { "clip_ratio": 0.0, "completion_length": 471.3125, "epoch": 0.8599706744868035, "grad_norm": 12.864618269615681, "kl": 1.615234375, "learning_rate": 1.5274572804776284e-07, "loss": 0.3322, "reward": 2.208186984062195, "reward_std": 0.10108592826873064, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.014035395812243223, "rewards/tag_count_reward": 0.9791666865348816, "step": 1173 }, { "clip_ratio": 0.0, "completion_length": 446.7291717529297, "epoch": 0.8607038123167156, "grad_norm": 8.624897435289583, "kl": 1.2255859375, "learning_rate": 1.5220577087011731e-07, "loss": 0.2663, "reward": 2.4378814697265625, "reward_std": 0.14415617939084768, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.027396278455853462, "rewards/tag_count_reward": 0.9791666865348816, "step": 1174 }, { "clip_ratio": 0.0, "completion_length": 463.8125, "epoch": 0.8614369501466276, "grad_norm": 11.736172117731995, "kl": 1.0546875, "learning_rate": 1.5166842145992737e-07, "loss": 0.2542, "reward": 1.9319466352462769, "reward_std": 0.1647976189851761, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.028122933581471443, "rewards/tag_count_reward": 0.9739583432674408, "step": 1175 }, { "clip_ratio": 0.0, "completion_length": 427.2708435058594, "epoch": 0.8621700879765396, "grad_norm": 14.550521023717632, "kl": 2.2421875, "learning_rate": 1.5113368333982424e-07, "loss": 0.172, "reward": 2.188947319984436, "reward_std": 0.09993797168135643, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02112211100757122, "rewards/tag_count_reward": 0.9739583432674408, "step": 1176 }, { "clip_ratio": 0.0, "completion_length": 451.50001525878906, "epoch": 0.8629032258064516, "grad_norm": 17.998782549691533, "kl": 1.345703125, "learning_rate": 1.5060156001532024e-07, "loss": 0.2064, "reward": 2.1768603920936584, "reward_std": 0.16409555356949568, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.028000716120004654, "rewards/tag_count_reward": 0.96875, "step": 1177 }, { "clip_ratio": 0.0, "completion_length": 377.7916717529297, "epoch": 0.8636363636363636, "grad_norm": 7.291251293870026, "kl": 0.85546875, "learning_rate": 1.5007205497478672e-07, "loss": 0.0573, "reward": 2.477933406829834, "reward_std": 0.014990990981459618, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022066613659262657, "rewards/tag_count_reward": 1.0, "step": 1178 }, { "clip_ratio": 0.0, "completion_length": 479.8541717529297, "epoch": 0.8643695014662757, "grad_norm": 12.249922267015474, "kl": 1.8388671875, "learning_rate": 1.4954517168943015e-07, "loss": 0.3705, "reward": 2.452030062675476, "reward_std": 0.0967979496344924, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021928315982222557, "rewards/tag_count_reward": 0.9739583432674408, "step": 1179 }, { "clip_ratio": 0.0, "completion_length": 446.8958435058594, "epoch": 0.8651026392961877, "grad_norm": 16.52706856842083, "kl": 1.943359375, "learning_rate": 1.4902091361327049e-07, "loss": 0.113, "reward": 2.450742721557617, "reward_std": 0.10176345612853765, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.02147953025996685, "rewards/tag_count_reward": 0.9791666865348816, "step": 1180 }, { "clip_ratio": 0.0, "completion_length": 458.8541717529297, "epoch": 0.8658357771260997, "grad_norm": 16.404992846234286, "kl": 0.71875, "learning_rate": 1.4849928418311742e-07, "loss": 0.0433, "reward": 2.4826748371124268, "reward_std": 0.009747894946485758, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01732515450567007, "rewards/tag_count_reward": 1.0, "step": 1181 }, { "clip_ratio": 0.0, "completion_length": 545.0208435058594, "epoch": 0.8665689149560117, "grad_norm": 101.74778357413155, "kl": 5.078125, "learning_rate": 1.4798028681854875e-07, "loss": 0.9233, "reward": 2.569236397743225, "reward_std": 0.4390984922647476, "rewards/accuracy_reward": 0.6875000149011612, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.04187488742172718, "rewards/tag_count_reward": 0.9583333432674408, "step": 1182 }, { "clip_ratio": 0.0, "completion_length": 492.25001525878906, "epoch": 0.8673020527859238, "grad_norm": 26.246204917626002, "kl": 3.287109375, "learning_rate": 1.4746392492188717e-07, "loss": 0.552, "reward": 2.1077802181243896, "reward_std": 0.3405429106205702, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9583333730697632, "rewards/repetition_penalty_reward": -0.03805319964885712, "rewards/tag_count_reward": 0.9583333432674408, "step": 1183 }, { "clip_ratio": 0.0, "completion_length": 499.62501525878906, "epoch": 0.8680351906158358, "grad_norm": 35.521939075442454, "kl": 3.9921875, "learning_rate": 1.46950201878179e-07, "loss": 0.4332, "reward": 1.9154297709465027, "reward_std": 0.1296032927930355, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.023806375451385975, "rewards/tag_count_reward": 0.953125, "step": 1184 }, { "clip_ratio": 0.0, "completion_length": 411.7083435058594, "epoch": 0.8687683284457478, "grad_norm": 34.922739418576704, "kl": 2.0234375, "learning_rate": 1.464391210551707e-07, "loss": 0.0956, "reward": 2.4347234964370728, "reward_std": 0.23893173784017563, "rewards/accuracy_reward": 0.47916667722165585, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02881821058690548, "rewards/tag_count_reward": 0.9843750298023224, "step": 1185 }, { "clip_ratio": 0.0, "completion_length": 535.5833435058594, "epoch": 0.8695014662756598, "grad_norm": 14.431104694894044, "kl": 2.265625, "learning_rate": 1.4593068580328804e-07, "loss": 0.4029, "reward": 2.140149235725403, "reward_std": 0.2738456204533577, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.026517387479543686, "rewards/tag_count_reward": 0.9375, "step": 1186 }, { "clip_ratio": 0.0, "completion_length": 431.0, "epoch": 0.8702346041055719, "grad_norm": 13.622638724234001, "kl": 0.95703125, "learning_rate": 1.4542489945561314e-07, "loss": 0.0834, "reward": 2.226495683193207, "reward_std": 0.012074500788003206, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023504262790083885, "rewards/tag_count_reward": 1.0, "step": 1187 }, { "clip_ratio": 0.0, "completion_length": 567.4375305175781, "epoch": 0.8709677419354839, "grad_norm": 45.186053187429835, "kl": 3.1171875, "learning_rate": 1.4492176532786332e-07, "loss": 0.9847, "reward": 2.02320396900177, "reward_std": 0.5472326725721359, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.05144892632961273, "rewards/tag_count_reward": 0.9010416865348816, "step": 1188 }, { "clip_ratio": 0.0, "completion_length": 525.75, "epoch": 0.8717008797653959, "grad_norm": 25.416128836027184, "kl": 2.2890625, "learning_rate": 1.4442128671836896e-07, "loss": 0.5288, "reward": 2.13466614484787, "reward_std": 0.25415295362472534, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021583949215710163, "rewards/tag_count_reward": 0.9479166865348816, "step": 1189 }, { "clip_ratio": 0.0, "completion_length": 410.0833435058594, "epoch": 0.8724340175953079, "grad_norm": 9.144807724609006, "kl": 0.6318359375, "learning_rate": 1.4392346690805202e-07, "loss": 0.0231, "reward": 1.9753663539886475, "reward_std": 0.011926409788429737, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024633726105093956, "rewards/tag_count_reward": 1.0, "step": 1190 }, { "clip_ratio": 0.0, "completion_length": 419.8958435058594, "epoch": 0.8731671554252199, "grad_norm": 14.712869736478394, "kl": 0.4169921875, "learning_rate": 1.4342830916040455e-07, "loss": 0.0499, "reward": 2.230422258377075, "reward_std": 0.010051010176539421, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01957788784056902, "rewards/tag_count_reward": 1.0, "step": 1191 }, { "clip_ratio": 0.0, "completion_length": 413.66668701171875, "epoch": 0.873900293255132, "grad_norm": 11.265022293824497, "kl": 1.419921875, "learning_rate": 1.4293581672146727e-07, "loss": 0.1369, "reward": 2.434700131416321, "reward_std": 0.15641514584422112, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023633372969925404, "rewards/tag_count_reward": 0.9791666865348816, "step": 1192 }, { "clip_ratio": 0.0, "completion_length": 520.7500305175781, "epoch": 0.874633431085044, "grad_norm": 11.514329955466966, "kl": 3.640625, "learning_rate": 1.424459928198083e-07, "loss": 0.5466, "reward": 2.527653455734253, "reward_std": 0.499290406703949, "rewards/accuracy_reward": 0.6458333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.02095778752118349, "rewards/tag_count_reward": 0.9166666865348816, "step": 1193 }, { "clip_ratio": 0.0, "completion_length": 485.75001525878906, "epoch": 0.875366568914956, "grad_norm": 13.6591202283712, "kl": 1.0458984375, "learning_rate": 1.419588406665019e-07, "loss": 0.2251, "reward": 2.1814990043640137, "reward_std": 0.16585272178053856, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.032042741775512695, "rewards/tag_count_reward": 0.984375, "step": 1194 }, { "clip_ratio": 0.0, "completion_length": 477.8333435058594, "epoch": 0.876099706744868, "grad_norm": 22.93844108305487, "kl": 2.21875, "learning_rate": 1.4147436345510758e-07, "loss": 0.5344, "reward": 2.395401358604431, "reward_std": 0.26352133601903915, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.028209981508553028, "rewards/tag_count_reward": 0.9583333730697632, "step": 1195 }, { "clip_ratio": 0.0, "completion_length": 475.875, "epoch": 0.8768328445747801, "grad_norm": 22.306040255922916, "kl": 1.7109375, "learning_rate": 1.4099256436164898e-07, "loss": 0.5019, "reward": 2.3607842922210693, "reward_std": 0.33058515191078186, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02636854164302349, "rewards/tag_count_reward": 0.9635416865348816, "step": 1196 }, { "clip_ratio": 0.0, "completion_length": 546.6458435058594, "epoch": 0.8775659824046921, "grad_norm": 18.68583800339866, "kl": 3.7109375, "learning_rate": 1.405134465445934e-07, "loss": 0.5144, "reward": 1.8736881017684937, "reward_std": 0.26206301152706146, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.029089704155921936, "rewards/tag_count_reward": 0.9166666865348816, "step": 1197 }, { "clip_ratio": 0.0, "completion_length": 514.0, "epoch": 0.8782991202346041, "grad_norm": 18.250109922553403, "kl": 4.234375, "learning_rate": 1.4003701314483042e-07, "loss": 0.6348, "reward": 1.910728394985199, "reward_std": 0.18229395896196365, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02156328223645687, "rewards/tag_count_reward": 0.9322916865348816, "step": 1198 }, { "clip_ratio": 0.0, "completion_length": 522.1041870117188, "epoch": 0.8790322580645161, "grad_norm": 35.35953038811825, "kl": 3.84375, "learning_rate": 1.395632672856523e-07, "loss": 0.948, "reward": 2.56496000289917, "reward_std": 0.4408094882965088, "rewards/accuracy_reward": 0.6875000298023224, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.04094269871711731, "rewards/tag_count_reward": 0.9322916865348816, "step": 1199 }, { "clip_ratio": 0.0, "completion_length": 481.25001525878906, "epoch": 0.8797653958944281, "grad_norm": 18.7935164342564, "kl": 3.859375, "learning_rate": 1.3909221207273232e-07, "loss": 0.5714, "reward": 2.5062928199768066, "reward_std": 0.5336205363273621, "rewards/accuracy_reward": 0.625, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.03537383023649454, "rewards/tag_count_reward": 0.9375, "step": 1200 }, { "clip_ratio": 0.0, "completion_length": 454.87501525878906, "epoch": 0.8804985337243402, "grad_norm": 22.88119881713903, "kl": 2.09765625, "learning_rate": 1.3862385059410565e-07, "loss": 0.0926, "reward": 1.9407550692558289, "reward_std": 0.2067294642329216, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.026258857920765877, "rewards/tag_count_reward": 0.9531250298023224, "step": 1201 }, { "clip_ratio": 0.0, "completion_length": 425.4166717529297, "epoch": 0.8812316715542522, "grad_norm": 14.47777815558714, "kl": 0.8076171875, "learning_rate": 1.381581859201479e-07, "loss": 0.2542, "reward": 2.196794271469116, "reward_std": 0.1006504325196147, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.03063626028597355, "rewards/tag_count_reward": 0.984375, "step": 1202 }, { "clip_ratio": 0.0, "completion_length": 469.43751525878906, "epoch": 0.8819648093841642, "grad_norm": 9.167261630009445, "kl": 1.208984375, "learning_rate": 1.376952211035562e-07, "loss": 0.0365, "reward": 2.668929934501648, "reward_std": 0.22134077921509743, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.018570033833384514, "rewards/tag_count_reward": 0.9791666865348816, "step": 1203 }, { "clip_ratio": 0.0, "completion_length": 419.9375, "epoch": 0.8826979472140762, "grad_norm": 8.414115405334524, "kl": 0.986328125, "learning_rate": 1.3723495917932803e-07, "loss": 0.0615, "reward": 2.1917551159858704, "reward_std": 0.11872765189036727, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02699489053338766, "rewards/tag_count_reward": 0.9895833432674408, "step": 1204 }, { "clip_ratio": 0.0, "completion_length": 408.2083435058594, "epoch": 0.8834310850439883, "grad_norm": 11.814801184529381, "kl": 0.869140625, "learning_rate": 1.3677740316474236e-07, "loss": 0.0662, "reward": 2.4721962213516235, "reward_std": 0.015518466010689735, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.027803857810795307, "rewards/tag_count_reward": 1.0, "step": 1205 }, { "clip_ratio": 0.0, "completion_length": 487.41668701171875, "epoch": 0.8841642228739003, "grad_norm": 26.396037778778368, "kl": 3.11328125, "learning_rate": 1.3632255605933892e-07, "loss": 0.4828, "reward": 2.3294543027877808, "reward_std": 0.3807095689699054, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.03339302074164152, "rewards/tag_count_reward": 0.9322916865348816, "step": 1206 }, { "clip_ratio": 0.0, "completion_length": 524.1458435058594, "epoch": 0.8848973607038123, "grad_norm": 22.881106867606714, "kl": 2.7265625, "learning_rate": 1.358704208448994e-07, "loss": 0.3669, "reward": 2.072553813457489, "reward_std": 0.34631527960300446, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.029876701533794403, "rewards/tag_count_reward": 0.921875, "step": 1207 }, { "clip_ratio": 0.0, "completion_length": 560.7500152587891, "epoch": 0.8856304985337243, "grad_norm": 24.40594031253714, "kl": 4.421875, "learning_rate": 1.3542100048542723e-07, "loss": 0.7167, "reward": 2.2936359643936157, "reward_std": 0.47601407766342163, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.03796131908893585, "rewards/tag_count_reward": 0.9010416865348816, "step": 1208 }, { "clip_ratio": 0.0, "completion_length": 452.62501525878906, "epoch": 0.8863636363636364, "grad_norm": 9.913065932344221, "kl": 2.974609375, "learning_rate": 1.349742979271286e-07, "loss": 0.4171, "reward": 2.3427236080169678, "reward_std": 0.2851731404662132, "rewards/accuracy_reward": 0.4375, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.028804211877286434, "rewards/tag_count_reward": 0.9479166865348816, "step": 1209 }, { "clip_ratio": 0.0, "completion_length": 472.72918701171875, "epoch": 0.8870967741935484, "grad_norm": 9.604198032999998, "kl": 1.984375, "learning_rate": 1.3453031609839265e-07, "loss": 0.2398, "reward": 2.3267154693603516, "reward_std": 0.3730589300394058, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.0291875172406435, "rewards/tag_count_reward": 0.9322916865348816, "step": 1210 }, { "clip_ratio": 0.0, "completion_length": 553.6666870117188, "epoch": 0.8878299120234604, "grad_norm": 18.106414337727866, "kl": 2.814453125, "learning_rate": 1.3408905790977318e-07, "loss": 0.4803, "reward": 2.5070048570632935, "reward_std": 0.44919482129625976, "rewards/accuracy_reward": 0.6458333432674408, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.02598138339817524, "rewards/tag_count_reward": 0.921875, "step": 1211 }, { "clip_ratio": 0.0, "completion_length": 496.16668701171875, "epoch": 0.8885630498533724, "grad_norm": 36.20233541493654, "kl": 2.24609375, "learning_rate": 1.3365052625396833e-07, "loss": 0.6054, "reward": 2.33927983045578, "reward_std": 0.382673054933548, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.03224811889231205, "rewards/tag_count_reward": 0.9270833432674408, "step": 1212 }, { "clip_ratio": 0.0, "completion_length": 434.8958435058594, "epoch": 0.8892961876832844, "grad_norm": 16.41131988450452, "kl": 0.74609375, "learning_rate": 1.3321472400580273e-07, "loss": 0.2423, "reward": 1.95037442445755, "reward_std": 0.11447333311662078, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02011175360530615, "rewards/tag_count_reward": 0.984375, "step": 1213 }, { "clip_ratio": 0.0, "completion_length": 422.31251525878906, "epoch": 0.8900293255131965, "grad_norm": 3.9170854028531417, "kl": 0.3603515625, "learning_rate": 1.3278165402220787e-07, "loss": 0.0441, "reward": 2.4774237275123596, "reward_std": 0.009436836931854486, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022576178424060345, "rewards/tag_count_reward": 1.0, "step": 1214 }, { "clip_ratio": 0.0, "completion_length": 448.9583435058594, "epoch": 0.8907624633431085, "grad_norm": 9.61162686291127, "kl": 0.6162109375, "learning_rate": 1.3235131914220406e-07, "loss": 0.0524, "reward": 2.1963998675346375, "reward_std": 0.11624268954619765, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022350233048200607, "rewards/tag_count_reward": 0.9895833432674408, "step": 1215 }, { "clip_ratio": 0.0, "completion_length": 516.0208435058594, "epoch": 0.8914956011730205, "grad_norm": 12.978659682793714, "kl": 0.75390625, "learning_rate": 1.3192372218688105e-07, "loss": 0.1972, "reward": 2.4325318336486816, "reward_std": 0.31300088763237, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02753778174519539, "rewards/tag_count_reward": 0.9739583730697632, "step": 1216 }, { "clip_ratio": 0.0, "completion_length": 512.9791870117188, "epoch": 0.8922287390029325, "grad_norm": 19.904083399665815, "kl": 3.35546875, "learning_rate": 1.3149886595938024e-07, "loss": 0.6293, "reward": 1.8745170831680298, "reward_std": 0.2614711821079254, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.028260784223675728, "rewards/tag_count_reward": 0.9166666865348816, "step": 1217 }, { "clip_ratio": 0.0, "completion_length": 529.3125305175781, "epoch": 0.8929618768328446, "grad_norm": 16.529405747089786, "kl": 2.06640625, "learning_rate": 1.3107675324487567e-07, "loss": 0.3418, "reward": 2.143596053123474, "reward_std": 0.2585507184267044, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.026542955078184605, "rewards/tag_count_reward": 0.9479166865348816, "step": 1218 }, { "clip_ratio": 0.0, "completion_length": 425.3958435058594, "epoch": 0.8936950146627566, "grad_norm": 7.720965596393824, "kl": 0.650390625, "learning_rate": 1.3065738681055659e-07, "loss": 0.0344, "reward": 2.6714645624160767, "reward_std": 0.19277621805667877, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.024716145358979702, "rewards/tag_count_reward": 0.9947916865348816, "step": 1219 }, { "clip_ratio": 0.0, "completion_length": 436.91668701171875, "epoch": 0.8944281524926686, "grad_norm": 15.006457009295644, "kl": 1.58984375, "learning_rate": 1.302407694056083e-07, "loss": 0.3509, "reward": 2.410833716392517, "reward_std": 0.24607142433524132, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.028402483090758324, "rewards/tag_count_reward": 0.9739583432674408, "step": 1220 }, { "clip_ratio": 0.0, "completion_length": 476.5833435058594, "epoch": 0.8951612903225806, "grad_norm": 12.304779027075734, "kl": 3.09375, "learning_rate": 1.2982690376119503e-07, "loss": 0.4719, "reward": 2.622697591781616, "reward_std": 0.33321088552474976, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.026607911102473736, "rewards/tag_count_reward": 0.9479166865348816, "step": 1221 }, { "clip_ratio": 0.0, "completion_length": 463.41668701171875, "epoch": 0.8958944281524927, "grad_norm": 9.34249431097831, "kl": 1.8505859375, "learning_rate": 1.2941579259044145e-07, "loss": 0.3572, "reward": 2.2139004468917847, "reward_std": 0.05917600984685123, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.020474628545343876, "rewards/tag_count_reward": 0.984375, "step": 1222 }, { "clip_ratio": 0.0, "completion_length": 568.0, "epoch": 0.8966275659824047, "grad_norm": 32.507561910320035, "kl": 3.80859375, "learning_rate": 1.290074385884151e-07, "loss": 0.8871, "reward": 2.7785022258758545, "reward_std": 0.6063933372497559, "rewards/accuracy_reward": 0.9166666865348816, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.03573389910161495, "rewards/tag_count_reward": 0.9322916865348816, "step": 1223 }, { "clip_ratio": 0.0, "completion_length": 419.3958435058594, "epoch": 0.8973607038123167, "grad_norm": 13.8733638371739, "kl": 1.4609375, "learning_rate": 1.2860184443210882e-07, "loss": 0.0547, "reward": 2.6921792030334473, "reward_std": 0.13939360855147243, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.019626601599156857, "rewards/tag_count_reward": 0.9895833432674408, "step": 1224 }, { "clip_ratio": 0.0, "completion_length": 468.00001525878906, "epoch": 0.8980938416422287, "grad_norm": 43.64985284284774, "kl": 1.40234375, "learning_rate": 1.28199012780423e-07, "loss": 0.3605, "reward": 2.430221438407898, "reward_std": 0.18982463865540922, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02116764523088932, "rewards/tag_count_reward": 0.9791666865348816, "step": 1225 }, { "clip_ratio": 0.0, "completion_length": 578.0208435058594, "epoch": 0.8988269794721407, "grad_norm": 96.47944289282677, "kl": 5.515625, "learning_rate": 1.277989462741482e-07, "loss": 0.7847, "reward": 2.0465909242630005, "reward_std": 0.5166294425725937, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.069728534668684, "rewards/tag_count_reward": 0.9218750298023224, "step": 1226 }, { "clip_ratio": 0.0, "completion_length": 395.0, "epoch": 0.8995601173020528, "grad_norm": 11.508237038705388, "kl": 2.203125, "learning_rate": 1.2740164753594811e-07, "loss": 0.1107, "reward": 2.3587182760238647, "reward_std": 0.3529049903154373, "rewards/accuracy_reward": 0.4375000149011612, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.021490052342414856, "rewards/tag_count_reward": 0.9635416865348816, "step": 1227 }, { "clip_ratio": 0.0, "completion_length": 444.9791717529297, "epoch": 0.9002932551319648, "grad_norm": 9.91761080011214, "kl": 1.1171875, "learning_rate": 1.2700711917034177e-07, "loss": 0.0532, "reward": 2.222747564315796, "reward_std": 0.025613236241042614, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022044179029762745, "rewards/tag_count_reward": 0.9947916865348816, "step": 1228 }, { "clip_ratio": 0.0, "completion_length": 479.7708435058594, "epoch": 0.9010263929618768, "grad_norm": 20.729917098151063, "kl": 3.1640625, "learning_rate": 1.2661536376368723e-07, "loss": 0.39, "reward": 2.633534550666809, "reward_std": 0.3318822309374809, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02445169072598219, "rewards/tag_count_reward": 0.9635416865348816, "step": 1229 }, { "clip_ratio": 0.0, "completion_length": 491.04168701171875, "epoch": 0.9017595307917888, "grad_norm": 18.64841756924077, "kl": 1.595703125, "learning_rate": 1.2622638388416385e-07, "loss": 0.3109, "reward": 2.4270366430282593, "reward_std": 0.1843274012207985, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02088008914142847, "rewards/tag_count_reward": 0.9687500298023224, "step": 1230 }, { "clip_ratio": 0.0, "completion_length": 460.68751525878906, "epoch": 0.9024926686217009, "grad_norm": 14.523184732588259, "kl": 0.984375, "learning_rate": 1.2584018208175614e-07, "loss": 0.0594, "reward": 2.475397825241089, "reward_std": 0.025617304258048534, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01939393486827612, "rewards/tag_count_reward": 0.9947916865348816, "step": 1231 }, { "clip_ratio": 0.0, "completion_length": 443.2291717529297, "epoch": 0.9032258064516129, "grad_norm": 16.281236006778695, "kl": 1.4375, "learning_rate": 1.2545676088823656e-07, "loss": 0.0787, "reward": 2.6920576095581055, "reward_std": 0.1349460631608963, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.019748042337596416, "rewards/tag_count_reward": 0.9895833730697632, "step": 1232 }, { "clip_ratio": 0.0, "completion_length": 626.0625305175781, "epoch": 0.9039589442815249, "grad_norm": 24.419234328744878, "kl": 2.33203125, "learning_rate": 1.2507612281714912e-07, "loss": 0.5811, "reward": 1.8098527789115906, "reward_std": 0.42128223180770874, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9513888657093048, "rewards/repetition_penalty_reward": -0.05820286273956299, "rewards/tag_count_reward": 0.9166666865348816, "step": 1233 }, { "clip_ratio": 0.0, "completion_length": 497.0833435058594, "epoch": 0.9046920821114369, "grad_norm": 9.210214401503478, "kl": 1.6591796875, "learning_rate": 1.2469827036379294e-07, "loss": 0.3667, "reward": 2.687574028968811, "reward_std": 0.1076923101209104, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02596777305006981, "rewards/tag_count_reward": 0.9635416865348816, "step": 1234 }, { "clip_ratio": 0.0, "completion_length": 479.89585876464844, "epoch": 0.905425219941349, "grad_norm": 21.720247309897992, "kl": 2.3671875, "learning_rate": 1.2432320600520568e-07, "loss": 0.3584, "reward": 2.1699541807174683, "reward_std": 0.18353354185819626, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02275434322655201, "rewards/tag_count_reward": 0.9635416865348816, "step": 1235 }, { "clip_ratio": 0.0, "completion_length": 586.7291717529297, "epoch": 0.906158357771261, "grad_norm": 29.757104937366716, "kl": 2.986328125, "learning_rate": 1.2395093220014776e-07, "loss": 0.3691, "reward": 2.2928508520126343, "reward_std": 0.2880670018494129, "rewards/accuracy_reward": 0.4166666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.03701025526970625, "rewards/tag_count_reward": 0.9270833432674408, "step": 1236 }, { "clip_ratio": 0.0, "completion_length": 507.75, "epoch": 0.906891495601173, "grad_norm": 12.415793376854689, "kl": 2.44140625, "learning_rate": 1.235814513890855e-07, "loss": 0.34, "reward": 2.7981667518615723, "reward_std": 0.5196200609207153, "rewards/accuracy_reward": 0.9166666865348816, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.03690283186733723, "rewards/tag_count_reward": 0.9531250298023224, "step": 1237 }, { "clip_ratio": 0.0, "completion_length": 442.625, "epoch": 0.907624633431085, "grad_norm": 9.400049645745804, "kl": 1.259765625, "learning_rate": 1.232147659941761e-07, "loss": 0.3226, "reward": 2.4332507848739624, "reward_std": 0.15348897967487574, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.030291039496660233, "rewards/tag_count_reward": 0.984375, "step": 1238 }, { "clip_ratio": 0.0, "completion_length": 495.3125, "epoch": 0.908357771260997, "grad_norm": 7.834262887636369, "kl": 0.9462890625, "learning_rate": 1.228508784192506e-07, "loss": 0.2261, "reward": 2.2043164372444153, "reward_std": 0.08925489895045757, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024850289337337017, "rewards/tag_count_reward": 0.9791666865348816, "step": 1239 }, { "clip_ratio": 0.0, "completion_length": 391.125, "epoch": 0.9090909090909091, "grad_norm": 6.238896486111178, "kl": 0.40625, "learning_rate": 1.2248979104979925e-07, "loss": 0.0608, "reward": 2.9420199394226074, "reward_std": 0.12390666967257857, "rewards/accuracy_reward": 0.9791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02672996185719967, "rewards/tag_count_reward": 0.9895833432674408, "step": 1240 }, { "clip_ratio": 0.0, "completion_length": 460.43751525878906, "epoch": 0.9098240469208211, "grad_norm": 18.532777826662272, "kl": 1.0888671875, "learning_rate": 1.2213150625295507e-07, "loss": 0.2335, "reward": 2.4339605569839478, "reward_std": 0.16963256802409887, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.026108983904123306, "rewards/tag_count_reward": 0.9947916865348816, "step": 1241 }, { "clip_ratio": 0.0, "completion_length": 476.4791717529297, "epoch": 0.9105571847507331, "grad_norm": 11.675756206099358, "kl": 2.9296875, "learning_rate": 1.2177602637747878e-07, "loss": 0.3675, "reward": 2.145154595375061, "reward_std": 0.18334370106458664, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.033664777874946594, "rewards/tag_count_reward": 0.9427083432674408, "step": 1242 }, { "clip_ratio": 0.0, "completion_length": 466.8333435058594, "epoch": 0.9112903225806451, "grad_norm": 16.09050924699834, "kl": 0.79296875, "learning_rate": 1.2142335375374307e-07, "loss": 0.2617, "reward": 2.4195971488952637, "reward_std": 0.22157121449708939, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.026583565399050713, "rewards/tag_count_reward": 0.9739583432674408, "step": 1243 }, { "clip_ratio": 0.0, "completion_length": 569.1458587646484, "epoch": 0.9120234604105572, "grad_norm": 78.30173265234124, "kl": 1.275390625, "learning_rate": 1.2107349069371778e-07, "loss": 0.4808, "reward": 2.301529884338379, "reward_std": 0.3600015640258789, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.021386777982115746, "rewards/tag_count_reward": 0.9270833432674408, "step": 1244 }, { "clip_ratio": 0.0, "completion_length": 424.91668701171875, "epoch": 0.9127565982404692, "grad_norm": 8.469807341370322, "kl": 1.4306640625, "learning_rate": 1.2072643949095406e-07, "loss": 0.1403, "reward": 2.1825441122055054, "reward_std": 0.16613122075796127, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02058098092675209, "rewards/tag_count_reward": 0.9739583730697632, "step": 1245 }, { "clip_ratio": 0.0, "completion_length": 437.4791717529297, "epoch": 0.9134897360703812, "grad_norm": 5.404996461007736, "kl": 0.5380859375, "learning_rate": 1.2038220242057014e-07, "loss": 0.029, "reward": 2.972726821899414, "reward_std": 0.010692416690289974, "rewards/accuracy_reward": 1.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02727325912564993, "rewards/tag_count_reward": 1.0, "step": 1246 }, { "clip_ratio": 0.0, "completion_length": 501.2708435058594, "epoch": 0.9142228739002932, "grad_norm": 23.52216583858044, "kl": 1.69140625, "learning_rate": 1.2004078173923584e-07, "loss": 0.2544, "reward": 2.575079560279846, "reward_std": 0.2885263650678098, "rewards/accuracy_reward": 0.6875, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.04471219517290592, "rewards/tag_count_reward": 0.953125, "step": 1247 }, { "clip_ratio": 0.0, "completion_length": 445.4583435058594, "epoch": 0.9149560117302052, "grad_norm": 26.89486686665491, "kl": 1.7890625, "learning_rate": 1.1970217968515773e-07, "loss": 0.2036, "reward": 1.9872766733169556, "reward_std": 0.16751797497272491, "rewards/accuracy_reward": 0.0416666679084301, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.028348451480269432, "rewards/tag_count_reward": 0.9739583432674408, "step": 1248 }, { "clip_ratio": 0.0, "completion_length": 433.5, "epoch": 0.9156891495601173, "grad_norm": 10.715869001628393, "kl": 1.19140625, "learning_rate": 1.1936639847806516e-07, "loss": 0.27, "reward": 2.4046987295150757, "reward_std": 0.2617655247449875, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.02932926919311285, "rewards/tag_count_reward": 0.9895833432674408, "step": 1249 }, { "clip_ratio": 0.0, "completion_length": 452.8333435058594, "epoch": 0.9164222873900293, "grad_norm": 33.601663618254626, "kl": 5.296875, "learning_rate": 1.1903344031919474e-07, "loss": 0.3832, "reward": 2.310217022895813, "reward_std": 0.30332429707050323, "rewards/accuracy_reward": 0.4166666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.031796906143426895, "rewards/tag_count_reward": 0.9322916865348816, "step": 1250 }, { "clip_ratio": 0.0, "completion_length": 523.125, "epoch": 0.9171554252199413, "grad_norm": 40.609113917444844, "kl": 5.0, "learning_rate": 1.1870330739127676e-07, "loss": 0.6408, "reward": 2.7230184078216553, "reward_std": 0.4918576031923294, "rewards/accuracy_reward": 0.8958333432674408, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.0356622040271759, "rewards/tag_count_reward": 0.890625, "step": 1251 }, { "clip_ratio": 0.0, "completion_length": 472.33335876464844, "epoch": 0.9178885630498533, "grad_norm": 8.756571405355077, "kl": 1.392578125, "learning_rate": 1.1837600185852021e-07, "loss": 0.254, "reward": 2.383722424507141, "reward_std": 0.35482998192310333, "rewards/accuracy_reward": 0.47916667722165585, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.027735993266105652, "rewards/tag_count_reward": 0.9531250298023224, "step": 1252 }, { "clip_ratio": 0.0, "completion_length": 514.125, "epoch": 0.9186217008797654, "grad_norm": 12.083545586661096, "kl": 1.7421875, "learning_rate": 1.1805152586659917e-07, "loss": 0.544, "reward": 2.0876930952072144, "reward_std": 0.39283469319343567, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9583334028720856, "rewards/repetition_penalty_reward": -0.05293210782110691, "rewards/tag_count_reward": 0.953125, "step": 1253 }, { "clip_ratio": 0.0, "completion_length": 404.81251525878906, "epoch": 0.9193548387096774, "grad_norm": 19.594726969083506, "kl": 1.513671875, "learning_rate": 1.1772988154263827e-07, "loss": 0.0739, "reward": 2.4594032764434814, "reward_std": 0.08329129219055176, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019763246178627014, "rewards/tag_count_reward": 1.0, "step": 1254 }, { "clip_ratio": 0.0, "completion_length": 468.89585876464844, "epoch": 0.9200879765395894, "grad_norm": 13.53668123908568, "kl": 1.2998046875, "learning_rate": 1.1741107099519908e-07, "loss": 0.4167, "reward": 2.84389591217041, "reward_std": 0.40458007249981165, "rewards/accuracy_reward": 0.9375, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.03457646816968918, "rewards/tag_count_reward": 0.96875, "step": 1255 }, { "clip_ratio": 0.0, "completion_length": 463.8333435058594, "epoch": 0.9208211143695014, "grad_norm": 13.28386012669474, "kl": 1.55859375, "learning_rate": 1.1709509631426599e-07, "loss": 0.1879, "reward": 2.147110402584076, "reward_std": 0.2086901506409049, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02129247970879078, "rewards/tag_count_reward": 0.9739583432674408, "step": 1256 }, { "clip_ratio": 0.0, "completion_length": 520.2291870117188, "epoch": 0.9215542521994134, "grad_norm": 18.597312059452115, "kl": 1.416015625, "learning_rate": 1.167819595712329e-07, "loss": 0.2873, "reward": 2.6505656242370605, "reward_std": 0.19670572225004435, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.024781646206974983, "rewards/tag_count_reward": 0.9739583432674408, "step": 1257 }, { "clip_ratio": 0.0, "completion_length": 470.16668701171875, "epoch": 0.9222873900293255, "grad_norm": 27.28829956069516, "kl": 1.01171875, "learning_rate": 1.1647166281888922e-07, "loss": 0.2236, "reward": 2.4381165504455566, "reward_std": 0.13398257829248905, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025425168685615063, "rewards/tag_count_reward": 0.984375, "step": 1258 }, { "clip_ratio": 0.0, "completion_length": 528.4166870117188, "epoch": 0.9230205278592375, "grad_norm": 25.886415618898045, "kl": 1.421875, "learning_rate": 1.1616420809140668e-07, "loss": 0.4292, "reward": 2.6378209590911865, "reward_std": 0.33803558349609375, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.021901373751461506, "rewards/tag_count_reward": 0.9583333730697632, "step": 1259 }, { "clip_ratio": 0.0, "completion_length": 448.6666717529297, "epoch": 0.9237536656891495, "grad_norm": 7.422823393613719, "kl": 1.1396484375, "learning_rate": 1.1585959740432584e-07, "loss": 0.0856, "reward": 2.69480299949646, "reward_std": 0.14079985255375504, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.015266527887433767, "rewards/tag_count_reward": 0.9947916865348816, "step": 1260 }, { "clip_ratio": 0.0, "completion_length": 484.2916717529297, "epoch": 0.9244868035190615, "grad_norm": 14.2945025870546, "kl": 0.798828125, "learning_rate": 1.1555783275454323e-07, "loss": 0.1845, "reward": 2.6793763637542725, "reward_std": 0.18702445551753044, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023748809471726418, "rewards/tag_count_reward": 0.9739583432674408, "step": 1261 }, { "clip_ratio": 0.0, "completion_length": 480.00001525878906, "epoch": 0.9252199413489736, "grad_norm": 35.82067585260527, "kl": 1.615234375, "learning_rate": 1.1525891612029762e-07, "loss": 0.5305, "reward": 2.1269590258598328, "reward_std": 0.3459872379899025, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.04144384525716305, "rewards/tag_count_reward": 0.9531250298023224, "step": 1262 }, { "clip_ratio": 0.0, "completion_length": 416.8125, "epoch": 0.9259530791788856, "grad_norm": 18.34196969705659, "kl": 0.880859375, "learning_rate": 1.1496284946115756e-07, "loss": 0.3035, "reward": 2.1744728088378906, "reward_std": 0.18630531942471862, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03212452307343483, "rewards/tag_count_reward": 0.984375, "step": 1263 }, { "clip_ratio": 0.0, "completion_length": 477.7083435058594, "epoch": 0.9266862170087976, "grad_norm": 13.058537172952395, "kl": 1.1953125, "learning_rate": 1.1466963471800852e-07, "loss": 0.2567, "reward": 1.9618767499923706, "reward_std": 0.06609710678458214, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022498343139886856, "rewards/tag_count_reward": 0.984375, "step": 1264 }, { "clip_ratio": 0.0, "completion_length": 450.4375, "epoch": 0.9274193548387096, "grad_norm": 93.37185157819849, "kl": 4.87890625, "learning_rate": 1.1437927381303987e-07, "loss": 0.5945, "reward": 2.346126675605774, "reward_std": 0.3644867241382599, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.03755396045744419, "rewards/tag_count_reward": 0.9531250298023224, "step": 1265 }, { "clip_ratio": 0.0, "completion_length": 584.6458587646484, "epoch": 0.9281524926686217, "grad_norm": 33.76460029432066, "kl": 4.23046875, "learning_rate": 1.140917686497326e-07, "loss": 0.2746, "reward": 2.094321370124817, "reward_std": 0.17763768695294857, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.965277761220932, "rewards/repetition_penalty_reward": -0.042831544764339924, "rewards/tag_count_reward": 0.9218750298023224, "step": 1266 }, { "clip_ratio": 0.0, "completion_length": 385.0208435058594, "epoch": 0.9288856304985337, "grad_norm": 8.536556995623087, "kl": 1.265625, "learning_rate": 1.1380712111284648e-07, "loss": 0.0854, "reward": 2.4350061416625977, "reward_std": 0.1251449566334486, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.03200777154415846, "rewards/tag_count_reward": 0.9947916865348816, "step": 1267 }, { "clip_ratio": 0.0, "completion_length": 535.875, "epoch": 0.9296187683284457, "grad_norm": 40.76353309315977, "kl": 4.78125, "learning_rate": 1.1352533306840818e-07, "loss": 0.5238, "reward": 2.105325937271118, "reward_std": 0.17815731838345528, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.03877122886478901, "rewards/tag_count_reward": 0.9010416865348816, "step": 1268 }, { "clip_ratio": 0.0, "completion_length": 473.75, "epoch": 0.9303519061583577, "grad_norm": 18.746529200671343, "kl": 2.71875, "learning_rate": 1.1324640636369844e-07, "loss": 0.3549, "reward": 2.0650742650032043, "reward_std": 0.34250083193182945, "rewards/accuracy_reward": 0.1875, "rewards/reasoning_steps_reward": 0.9652778506278992, "rewards/repetition_penalty_reward": -0.030411873012781143, "rewards/tag_count_reward": 0.9427083730697632, "step": 1269 }, { "clip_ratio": 0.0, "completion_length": 519.1666870117188, "epoch": 0.9310850439882697, "grad_norm": 19.92539027857833, "kl": 1.70703125, "learning_rate": 1.129703428272407e-07, "loss": 0.5492, "reward": 2.290251612663269, "reward_std": 0.4478663057088852, "rewards/accuracy_reward": 0.3958333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.03960954770445824, "rewards/tag_count_reward": 0.9479166865348816, "step": 1270 }, { "clip_ratio": 0.0, "completion_length": 446.1666717529297, "epoch": 0.9318181818181818, "grad_norm": 9.895164235157509, "kl": 0.76171875, "learning_rate": 1.1269714426878821e-07, "loss": 0.0382, "reward": 1.9925453066825867, "reward_std": 0.09840421006083488, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023079711012542248, "rewards/tag_count_reward": 0.9947916865348816, "step": 1271 }, { "clip_ratio": 0.0, "completion_length": 518.0416870117188, "epoch": 0.9325513196480938, "grad_norm": 15.315615201974119, "kl": 1.39453125, "learning_rate": 1.1242681247931319e-07, "loss": 0.4336, "reward": 2.3977036476135254, "reward_std": 0.25321169197559357, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.04326878301799297, "rewards/tag_count_reward": 0.9687500298023224, "step": 1272 }, { "clip_ratio": 0.0, "completion_length": 514.9583587646484, "epoch": 0.9332844574780058, "grad_norm": 10.106874745508811, "kl": 1.427734375, "learning_rate": 1.1215934923099414e-07, "loss": 0.2627, "reward": 2.1677005290985107, "reward_std": 0.1815025839023292, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.03021625615656376, "rewards/tag_count_reward": 0.96875, "step": 1273 }, { "clip_ratio": 0.0, "completion_length": 411.87501525878906, "epoch": 0.9340175953079178, "grad_norm": 9.054120736545237, "kl": 0.982421875, "learning_rate": 1.1189475627720488e-07, "loss": 0.0483, "reward": 2.7023842334747314, "reward_std": 0.32853075861930847, "rewards/accuracy_reward": 0.7500000298023224, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.02331024780869484, "rewards/tag_count_reward": 0.9895833432674408, "step": 1274 }, { "clip_ratio": 0.0, "completion_length": 478.1875, "epoch": 0.9347507331378299, "grad_norm": 14.23126659602911, "kl": 0.462890625, "learning_rate": 1.1163303535250265e-07, "loss": 0.0494, "reward": 2.2366175055503845, "reward_std": 0.009620955679565668, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013382541947066784, "rewards/tag_count_reward": 1.0, "step": 1275 }, { "clip_ratio": 0.0, "completion_length": 546.8541870117188, "epoch": 0.9354838709677419, "grad_norm": 16.41714592228837, "kl": 2.0625, "learning_rate": 1.1137418817261706e-07, "loss": 0.6162, "reward": 1.894201636314392, "reward_std": 0.26965878158807755, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.03982617612928152, "rewards/tag_count_reward": 0.9479166865348816, "step": 1276 }, { "clip_ratio": 0.0, "completion_length": 595.7083435058594, "epoch": 0.9362170087976539, "grad_norm": 26.283905458815948, "kl": 3.166015625, "learning_rate": 1.1111821643443861e-07, "loss": 0.6573, "reward": 2.7341933250427246, "reward_std": 0.5056948512792587, "rewards/accuracy_reward": 0.8958333730697632, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.04358469694852829, "rewards/tag_count_reward": 0.9166666865348816, "step": 1277 }, { "clip_ratio": 0.0, "completion_length": 573.1666870117188, "epoch": 0.9369501466275659, "grad_norm": 37.98541597450915, "kl": 3.1796875, "learning_rate": 1.1086512181600757e-07, "loss": 0.4427, "reward": 2.0666526556015015, "reward_std": 0.3224688321352005, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9375000298023224, "rewards/repetition_penalty_reward": -0.042722418904304504, "rewards/tag_count_reward": 0.921875, "step": 1278 }, { "clip_ratio": 0.0, "completion_length": 489.93751525878906, "epoch": 0.9376832844574781, "grad_norm": 13.211520187016381, "kl": 1.947265625, "learning_rate": 1.1061490597650326e-07, "loss": 0.509, "reward": 2.851483106613159, "reward_std": 0.40315455198287964, "rewards/accuracy_reward": 0.9166666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023517010267823935, "rewards/tag_count_reward": 0.9583333432674408, "step": 1279 }, { "clip_ratio": 0.0, "completion_length": 462.3958435058594, "epoch": 0.9384164222873901, "grad_norm": 13.920681467096035, "kl": 1.82421875, "learning_rate": 1.1036757055623277e-07, "loss": 0.2295, "reward": 2.3927890062332153, "reward_std": 0.21528612356632948, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.029086150228977203, "rewards/tag_count_reward": 0.984375, "step": 1280 }, { "clip_ratio": 0.0, "completion_length": 432.0208435058594, "epoch": 0.9391495601173021, "grad_norm": 19.243232638505983, "kl": 0.9921875, "learning_rate": 1.1012311717662053e-07, "loss": 0.248, "reward": 2.65679132938385, "reward_std": 0.18610515631735325, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.032444894313812256, "rewards/tag_count_reward": 0.9947916865348816, "step": 1281 }, { "clip_ratio": 0.0, "completion_length": 445.7291717529297, "epoch": 0.9398826979472141, "grad_norm": 12.50332420462573, "kl": 1.984375, "learning_rate": 1.0988154744019741e-07, "loss": 0.3032, "reward": 2.6296029090881348, "reward_std": 0.33814698457717896, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.028383249416947365, "rewards/tag_count_reward": 0.9635416865348816, "step": 1282 }, { "clip_ratio": 0.0, "completion_length": 450.91668701171875, "epoch": 0.9406158357771262, "grad_norm": 25.595049083415667, "kl": 2.2109375, "learning_rate": 1.096428629305905e-07, "loss": 0.2472, "reward": 2.625510334968567, "reward_std": 0.2489984016865492, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.03594798780977726, "rewards/tag_count_reward": 0.9739583432674408, "step": 1283 }, { "clip_ratio": 0.0, "completion_length": 532.2500305175781, "epoch": 0.9413489736070382, "grad_norm": 25.12369374073104, "kl": 3.4921875, "learning_rate": 1.0940706521251251e-07, "loss": 0.3859, "reward": 2.824937343597412, "reward_std": 0.3094763532280922, "rewards/accuracy_reward": 0.8958333432674408, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.022284891456365585, "rewards/tag_count_reward": 0.9583333432674408, "step": 1284 }, { "clip_ratio": 0.0, "completion_length": 487.6458435058594, "epoch": 0.9420821114369502, "grad_norm": 9.331338158859646, "kl": 1.044921875, "learning_rate": 1.091741558317517e-07, "loss": 0.4031, "reward": 2.372399091720581, "reward_std": 0.3811139464378357, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.03385098837316036, "rewards/tag_count_reward": 0.9687500298023224, "step": 1285 }, { "clip_ratio": 0.0, "completion_length": 560.3125305175781, "epoch": 0.9428152492668622, "grad_norm": 27.446421938812083, "kl": 3.1953125, "learning_rate": 1.0894413631516144e-07, "loss": 0.5668, "reward": 1.978183925151825, "reward_std": 0.4256092458963394, "rewards/accuracy_reward": 0.1666666716337204, "rewards/reasoning_steps_reward": 0.951388955116272, "rewards/repetition_penalty_reward": -0.05133005045354366, "rewards/tag_count_reward": 0.9114583730697632, "step": 1286 }, { "clip_ratio": 0.0, "completion_length": 498.79168701171875, "epoch": 0.9435483870967742, "grad_norm": 19.534270584207096, "kl": 1.30859375, "learning_rate": 1.087170081706506e-07, "loss": 0.1945, "reward": 2.385071635246277, "reward_std": 0.34212101995944977, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.024650713428854942, "rewards/tag_count_reward": 0.9583333432674408, "step": 1287 }, { "clip_ratio": 0.0, "completion_length": 510.29168701171875, "epoch": 0.9442815249266863, "grad_norm": 10.502662630078762, "kl": 1.193359375, "learning_rate": 1.084927728871733e-07, "loss": 0.2169, "reward": 2.1598105430603027, "reward_std": 0.26580096036195755, "rewards/accuracy_reward": 0.22916667722165585, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.029425671324133873, "rewards/tag_count_reward": 0.9739583730697632, "step": 1288 }, { "clip_ratio": 0.0, "completion_length": 533.1250152587891, "epoch": 0.9450146627565983, "grad_norm": 13.347926029552143, "kl": 2.462890625, "learning_rate": 1.0827143193471942e-07, "loss": 0.5006, "reward": 2.7681411504745483, "reward_std": 0.380189748480916, "rewards/accuracy_reward": 0.8958333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.03915059007704258, "rewards/tag_count_reward": 0.9322916865348816, "step": 1289 }, { "clip_ratio": 0.0, "completion_length": 476.1875, "epoch": 0.9457478005865103, "grad_norm": 9.069726802816593, "kl": 0.8916015625, "learning_rate": 1.0805298676430476e-07, "loss": 0.2142, "reward": 2.1827036142349243, "reward_std": 0.17475424334406853, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.02562977932393551, "rewards/tag_count_reward": 0.9791666865348816, "step": 1290 }, { "clip_ratio": 0.0, "completion_length": 453.04168701171875, "epoch": 0.9464809384164223, "grad_norm": 19.11237742206637, "kl": 1.53515625, "learning_rate": 1.0783743880796175e-07, "loss": 0.4172, "reward": 2.1592122316360474, "reward_std": 0.33645743131637573, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.03523235023021698, "rewards/tag_count_reward": 0.9583333432674408, "step": 1291 }, { "clip_ratio": 0.0, "completion_length": 408.7083435058594, "epoch": 0.9472140762463344, "grad_norm": 8.358380076378197, "kl": 1.1708984375, "learning_rate": 1.0762478947872983e-07, "loss": 0.0129, "reward": 2.4570391178131104, "reward_std": 0.08680337062105536, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9930555522441864, "rewards/repetition_penalty_reward": -0.02039137203246355, "rewards/tag_count_reward": 0.984375, "step": 1292 }, { "clip_ratio": 0.0, "completion_length": 531.0000305175781, "epoch": 0.9479472140762464, "grad_norm": 19.714536710646975, "kl": 2.4140625, "learning_rate": 1.0741504017064627e-07, "loss": 0.4306, "reward": 2.0773446559906006, "reward_std": 0.4128805547952652, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9583333432674408, "rewards/repetition_penalty_reward": -0.02161379996687174, "rewards/tag_count_reward": 0.9114583432674408, "step": 1293 }, { "clip_ratio": 0.0, "completion_length": 552.2083435058594, "epoch": 0.9486803519061584, "grad_norm": 16.131634788364266, "kl": 1.17578125, "learning_rate": 1.0720819225873714e-07, "loss": 0.5132, "reward": 2.111280918121338, "reward_std": 0.29894164204597473, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9652778804302216, "rewards/repetition_penalty_reward": -0.04149695113301277, "rewards/tag_count_reward": 0.9375000298023224, "step": 1294 }, { "clip_ratio": 0.0, "completion_length": 630.2916870117188, "epoch": 0.9494134897360704, "grad_norm": 18.93383557304818, "kl": 1.9453125, "learning_rate": 1.0700424709900813e-07, "loss": 0.4087, "reward": 2.7276880741119385, "reward_std": 0.521675705909729, "rewards/accuracy_reward": 0.8750000298023224, "rewards/reasoning_steps_reward": 0.958333432674408, "rewards/repetition_penalty_reward": -0.04314529709517956, "rewards/tag_count_reward": 0.9375, "step": 1295 }, { "clip_ratio": 0.0, "completion_length": 421.2083435058594, "epoch": 0.9501466275659824, "grad_norm": 11.487695498130753, "kl": 0.4814453125, "learning_rate": 1.0680320602843588e-07, "loss": 0.0643, "reward": 1.978863000869751, "reward_std": 0.008737085154280066, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.0211369963362813, "rewards/tag_count_reward": 1.0, "step": 1296 }, { "clip_ratio": 0.0, "completion_length": 410.9583435058594, "epoch": 0.9508797653958945, "grad_norm": 9.690609279712593, "kl": 0.546875, "learning_rate": 1.0660507036495882e-07, "loss": 0.1067, "reward": 2.6800166368484497, "reward_std": 0.18910772260278463, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.01963624032214284, "rewards/tag_count_reward": 0.984375, "step": 1297 }, { "clip_ratio": 0.0, "completion_length": 407.93751525878906, "epoch": 0.9516129032258065, "grad_norm": 7.924204047826416, "kl": 0.59375, "learning_rate": 1.0640984140746907e-07, "loss": 0.0238, "reward": 2.7067856788635254, "reward_std": 0.0974614720325917, "rewards/accuracy_reward": 0.7291666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01717284694314003, "rewards/tag_count_reward": 0.9947916865348816, "step": 1298 }, { "clip_ratio": 0.0, "completion_length": 468.54168701171875, "epoch": 0.9523460410557185, "grad_norm": 17.894742340360352, "kl": 0.787109375, "learning_rate": 1.0621752043580337e-07, "loss": 0.151, "reward": 2.209893822669983, "reward_std": 0.06893100775778294, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02448125835508108, "rewards/tag_count_reward": 0.984375, "step": 1299 }, { "clip_ratio": 0.0, "completion_length": 468.31251525878906, "epoch": 0.9530791788856305, "grad_norm": 33.42783345561901, "kl": 1.515625, "learning_rate": 1.0602810871073524e-07, "loss": 0.4529, "reward": 2.6334890127182007, "reward_std": 0.3280341923236847, "rewards/accuracy_reward": 0.6875000149011612, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.031441716477274895, "rewards/tag_count_reward": 0.984375, "step": 1300 }, { "clip_ratio": 0.0, "completion_length": 500.6458435058594, "epoch": 0.9538123167155426, "grad_norm": 13.936654154401653, "kl": 2.8125, "learning_rate": 1.0584160747396613e-07, "loss": 0.6693, "reward": 2.0740148425102234, "reward_std": 0.44181716442108154, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.04230464622378349, "rewards/tag_count_reward": 0.9427083432674408, "step": 1301 }, { "clip_ratio": 0.0, "completion_length": 367.6041717529297, "epoch": 0.9545454545454546, "grad_norm": 9.041598291124687, "kl": 0.658203125, "learning_rate": 1.0565801794811783e-07, "loss": 0.0691, "reward": 2.9332125186920166, "reward_std": 0.1549149751663208, "rewards/accuracy_reward": 0.9583333730697632, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02512078545987606, "rewards/tag_count_reward": 1.0, "step": 1302 }, { "clip_ratio": 0.0, "completion_length": 453.00001525878906, "epoch": 0.9552785923753666, "grad_norm": 20.85958631389929, "kl": 0.888671875, "learning_rate": 1.0547734133672415e-07, "loss": 0.2331, "reward": 2.6373571157455444, "reward_std": 0.314850315451622, "rewards/accuracy_reward": 0.708333358168602, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.03104559425264597, "rewards/tag_count_reward": 0.9739583432674408, "step": 1303 }, { "clip_ratio": 0.0, "completion_length": 422.4375, "epoch": 0.9560117302052786, "grad_norm": 5.871158172325361, "kl": 0.728515625, "learning_rate": 1.0529957882422293e-07, "loss": 0.0489, "reward": 2.234027624130249, "reward_std": 0.007591208443045616, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.015972347930073738, "rewards/tag_count_reward": 1.0, "step": 1304 }, { "clip_ratio": 0.0, "completion_length": 431.7916717529297, "epoch": 0.9567448680351907, "grad_norm": 9.173786941390691, "kl": 1.23046875, "learning_rate": 1.0512473157594867e-07, "loss": 0.2378, "reward": 1.9633594751358032, "reward_std": 0.07479000370949507, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02101556584239006, "rewards/tag_count_reward": 0.984375, "step": 1305 }, { "clip_ratio": 0.0, "completion_length": 501.87501525878906, "epoch": 0.9574780058651027, "grad_norm": 24.293729166843942, "kl": 1.845703125, "learning_rate": 1.0495280073812452e-07, "loss": 0.2547, "reward": 2.377333641052246, "reward_std": 0.2659803181886673, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.020235841162502766, "rewards/tag_count_reward": 0.9322916865348816, "step": 1306 }, { "clip_ratio": 0.0, "completion_length": 428.75, "epoch": 0.9582111436950147, "grad_norm": 8.978838251394835, "kl": 1.2578125, "learning_rate": 1.0478378743785488e-07, "loss": 0.0763, "reward": 2.881321907043457, "reward_std": 0.2719968855381012, "rewards/accuracy_reward": 0.9166666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02492798399180174, "rewards/tag_count_reward": 0.9895833432674408, "step": 1307 }, { "clip_ratio": 0.0, "completion_length": 438.4583435058594, "epoch": 0.9589442815249267, "grad_norm": 11.981739176953312, "kl": 0.87109375, "learning_rate": 1.0461769278311814e-07, "loss": 0.0627, "reward": 2.2008538246154785, "reward_std": 0.11611452978104353, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01789623498916626, "rewards/tag_count_reward": 0.9895833432674408, "step": 1308 }, { "clip_ratio": 0.0, "completion_length": 397.3958435058594, "epoch": 0.9596774193548387, "grad_norm": 17.04000369636799, "kl": 2.03515625, "learning_rate": 1.0445451786275917e-07, "loss": 0.0789, "reward": 2.6977003812789917, "reward_std": 0.24708323180675507, "rewards/accuracy_reward": 0.7500000298023224, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.022785856388509274, "rewards/tag_count_reward": 0.984375, "step": 1309 }, { "clip_ratio": 0.0, "completion_length": 401.9583435058594, "epoch": 0.9604105571847508, "grad_norm": 15.440879775642438, "kl": 1.908203125, "learning_rate": 1.0429426374648246e-07, "loss": 0.0467, "reward": 2.4519360065460205, "reward_std": 0.11012615030631423, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.01855024415999651, "rewards/tag_count_reward": 0.984375, "step": 1310 }, { "clip_ratio": 0.0, "completion_length": 474.8333435058594, "epoch": 0.9611436950146628, "grad_norm": 63.444012755191835, "kl": 5.65625, "learning_rate": 1.0413693148484486e-07, "loss": 0.4911, "reward": 2.6069284677505493, "reward_std": 0.26500143855810165, "rewards/accuracy_reward": 0.6875, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.01980782952159643, "rewards/tag_count_reward": 0.9531250298023224, "step": 1311 }, { "clip_ratio": 0.0, "completion_length": 477.2083435058594, "epoch": 0.9618768328445748, "grad_norm": 16.284465114202774, "kl": 1.77734375, "learning_rate": 1.0398252210924889e-07, "loss": 0.4078, "reward": 2.1387650966644287, "reward_std": 0.26892125606536865, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9861111640930176, "rewards/repetition_penalty_reward": -0.03484594263136387, "rewards/tag_count_reward": 0.9583333432674408, "step": 1312 }, { "clip_ratio": 0.0, "completion_length": 612.6666870117188, "epoch": 0.9626099706744868, "grad_norm": 57.44133112793273, "kl": 5.65625, "learning_rate": 1.0383103663193576e-07, "loss": 1.1269, "reward": 2.1499454975128174, "reward_std": 0.6474978625774384, "rewards/accuracy_reward": 0.375, "rewards/reasoning_steps_reward": 0.9583333730697632, "rewards/repetition_penalty_reward": -0.05838798172771931, "rewards/tag_count_reward": 0.8750000298023224, "step": 1313 }, { "clip_ratio": 0.0, "completion_length": 432.6041717529297, "epoch": 0.9633431085043989, "grad_norm": 11.55554991626459, "kl": 1.234375, "learning_rate": 1.0368247604597903e-07, "loss": 0.2578, "reward": 2.634058356285095, "reward_std": 0.2719671204686165, "rewards/accuracy_reward": 0.6875000149011612, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022191734984517097, "rewards/tag_count_reward": 0.9687500298023224, "step": 1314 }, { "clip_ratio": 0.0, "completion_length": 564.7291717529297, "epoch": 0.9640762463343109, "grad_norm": 35.946046250598904, "kl": 2.578125, "learning_rate": 1.0353684132527778e-07, "loss": 0.5811, "reward": 2.1190367937088013, "reward_std": 0.24303436279296875, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.05283825471997261, "rewards/tag_count_reward": 0.9427083432674408, "step": 1315 }, { "clip_ratio": 0.0, "completion_length": 486.5833435058594, "epoch": 0.9648093841642229, "grad_norm": 13.375233254984911, "kl": 1.296875, "learning_rate": 1.0339413342455055e-07, "loss": 0.4354, "reward": 2.1532533764839172, "reward_std": 0.19535775389522314, "rewards/accuracy_reward": 0.27083333395421505, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.03251057770103216, "rewards/tag_count_reward": 0.9427083730697632, "step": 1316 }, { "clip_ratio": 0.0, "completion_length": 481.8541717529297, "epoch": 0.9655425219941349, "grad_norm": 82.3335585692892, "kl": 1.66796875, "learning_rate": 1.0325435327932868e-07, "loss": 0.3158, "reward": 2.188454806804657, "reward_std": 0.14728715340606868, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.019878730177879333, "rewards/tag_count_reward": 0.9791666865348816, "step": 1317 }, { "clip_ratio": 0.0, "completion_length": 515.0208435058594, "epoch": 0.966275659824047, "grad_norm": 8.842073408230645, "kl": 1.07421875, "learning_rate": 1.0311750180595063e-07, "loss": 0.3922, "reward": 2.0934900045394897, "reward_std": 0.36611997336149216, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.972222238779068, "rewards/repetition_penalty_reward": -0.034982199780642986, "rewards/tag_count_reward": 0.9479166865348816, "step": 1318 }, { "clip_ratio": 0.0, "completion_length": 467.0416717529297, "epoch": 0.967008797653959, "grad_norm": 14.852296919095123, "kl": 1.8115234375, "learning_rate": 1.0298357990155564e-07, "loss": 0.1998, "reward": 2.190988779067993, "reward_std": 0.24773601721972227, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.022552933543920517, "rewards/tag_count_reward": 0.9635416865348816, "step": 1319 }, { "clip_ratio": 0.0, "completion_length": 404.7708435058594, "epoch": 0.967741935483871, "grad_norm": 6.1793249158960855, "kl": 0.384765625, "learning_rate": 1.0285258844407792e-07, "loss": 0.033, "reward": 2.462868571281433, "reward_std": 0.08148548658937216, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01629808358848095, "rewards/tag_count_reward": 1.0, "step": 1320 }, { "clip_ratio": 0.0, "completion_length": 461.1458435058594, "epoch": 0.968475073313783, "grad_norm": 9.003767777901873, "kl": 0.8037109375, "learning_rate": 1.0272452829224101e-07, "loss": 0.2078, "reward": 2.176120936870575, "reward_std": 0.1836662571877241, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.030476349405944347, "rewards/tag_count_reward": 0.984375, "step": 1321 }, { "clip_ratio": 0.0, "completion_length": 433.4583435058594, "epoch": 0.969208211143695, "grad_norm": 13.03765997652474, "kl": 1.515625, "learning_rate": 1.025994002855521e-07, "loss": 0.2286, "reward": 2.6289268732070923, "reward_std": 0.23047233279794455, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.03253147657960653, "rewards/tag_count_reward": 0.9739583432674408, "step": 1322 }, { "clip_ratio": 0.0, "completion_length": 482.7708435058594, "epoch": 0.9699413489736071, "grad_norm": 20.321561708883095, "kl": 2.59375, "learning_rate": 1.024772052442964e-07, "loss": 0.3407, "reward": 2.7547218799591064, "reward_std": 0.471389040350914, "rewards/accuracy_reward": 0.8958333730697632, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.0438892375677824, "rewards/tag_count_reward": 0.9375000298023224, "step": 1323 }, { "clip_ratio": 0.0, "completion_length": 455.4375, "epoch": 0.9706744868035191, "grad_norm": 4.241229172746372, "kl": 0.3759765625, "learning_rate": 1.02357943969532e-07, "loss": 0.0252, "reward": 2.7372305393218994, "reward_std": 0.00828889012336731, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.012769339140504599, "rewards/tag_count_reward": 1.0, "step": 1324 }, { "clip_ratio": 0.0, "completion_length": 449.79168701171875, "epoch": 0.9714076246334311, "grad_norm": 8.59881732463318, "kl": 1.044921875, "learning_rate": 1.0224161724308424e-07, "loss": 0.1959, "reward": 2.18610417842865, "reward_std": 0.1191612258553505, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.023965245112776756, "rewards/tag_count_reward": 0.9739583432674408, "step": 1325 }, { "clip_ratio": 0.0, "completion_length": 444.5208435058594, "epoch": 0.9721407624633431, "grad_norm": 15.822434043127465, "kl": 0.62109375, "learning_rate": 1.0212822582754111e-07, "loss": 0.2747, "reward": 2.4358739852905273, "reward_std": 0.147409213706851, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.034612227231264114, "rewards/tag_count_reward": 0.984375, "step": 1326 }, { "clip_ratio": 0.0, "completion_length": 397.375, "epoch": 0.9728739002932552, "grad_norm": 10.88147486521058, "kl": 0.9609375, "learning_rate": 1.0201777046624782e-07, "loss": 0.0638, "reward": 2.4328325986862183, "reward_std": 0.12669185176491737, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.025500833988189697, "rewards/tag_count_reward": 0.9791666865348816, "step": 1327 }, { "clip_ratio": 0.0, "completion_length": 390.2083435058594, "epoch": 0.9736070381231672, "grad_norm": 16.115241449324763, "kl": 0.4658203125, "learning_rate": 1.0191025188330209e-07, "loss": 0.0356, "reward": 2.231991171836853, "reward_std": 0.011185419745743275, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018008952029049397, "rewards/tag_count_reward": 1.0, "step": 1328 }, { "clip_ratio": 0.0, "completion_length": 478.31251525878906, "epoch": 0.9743401759530792, "grad_norm": 18.949099202801406, "kl": 1.4453125, "learning_rate": 1.0180567078354935e-07, "loss": 0.5752, "reward": 2.1427964568138123, "reward_std": 0.2880570739507675, "rewards/accuracy_reward": 0.2291666716337204, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.04123134259134531, "rewards/tag_count_reward": 0.9687500298023224, "step": 1329 }, { "clip_ratio": 0.0, "completion_length": 428.3958435058594, "epoch": 0.9750733137829912, "grad_norm": 7.7299459378810385, "kl": 0.556640625, "learning_rate": 1.0170402785257827e-07, "loss": 0.0326, "reward": 2.7296062707901, "reward_std": 0.010482048150151968, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02039369009435177, "rewards/tag_count_reward": 1.0, "step": 1330 }, { "clip_ratio": 0.0, "completion_length": 373.47918701171875, "epoch": 0.9758064516129032, "grad_norm": 7.036182974367325, "kl": 0.798828125, "learning_rate": 1.0160532375671602e-07, "loss": 0.0637, "reward": 2.454910397529602, "reward_std": 0.08339553605765104, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.024256338365375996, "rewards/tag_count_reward": 1.0, "step": 1331 }, { "clip_ratio": 0.0, "completion_length": 616.3125305175781, "epoch": 0.9765395894428153, "grad_norm": 23.087064957780417, "kl": 1.5859375, "learning_rate": 1.0150955914302412e-07, "loss": 0.2972, "reward": 2.4457297325134277, "reward_std": 0.4613430858589709, "rewards/accuracy_reward": 0.625, "rewards/reasoning_steps_reward": 0.9583333730697632, "rewards/repetition_penalty_reward": -0.04906199872493744, "rewards/tag_count_reward": 0.9114583432674408, "step": 1332 }, { "clip_ratio": 0.0, "completion_length": 441.3541717529297, "epoch": 0.9772727272727273, "grad_norm": 27.7813185862066, "kl": 1.37109375, "learning_rate": 1.0141673463929408e-07, "loss": 0.2585, "reward": 2.4148507118225098, "reward_std": 0.22696854081004858, "rewards/accuracy_reward": 0.4791666716337204, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.027857608161866665, "rewards/tag_count_reward": 0.984375, "step": 1333 }, { "clip_ratio": 0.0, "completion_length": 584.9791717529297, "epoch": 0.9780058651026393, "grad_norm": 14.123296378129437, "kl": 1.74609375, "learning_rate": 1.0132685085404337e-07, "loss": 0.5404, "reward": 2.511002779006958, "reward_std": 0.4390052007511258, "rewards/accuracy_reward": 0.6458333432674408, "rewards/reasoning_steps_reward": 0.9652777910232544, "rewards/repetition_penalty_reward": -0.04281671531498432, "rewards/tag_count_reward": 0.9427083432674408, "step": 1334 }, { "clip_ratio": 0.0, "completion_length": 448.1458435058594, "epoch": 0.9787390029325513, "grad_norm": 15.585193483352148, "kl": 1.09375, "learning_rate": 1.0123990837651128e-07, "loss": 0.2101, "reward": 2.183935582637787, "reward_std": 0.13025381043553352, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.979166716337204, "rewards/repetition_penalty_reward": -0.029606159776449203, "rewards/tag_count_reward": 0.984375, "step": 1335 }, { "clip_ratio": 0.0, "completion_length": 446.81251525878906, "epoch": 0.9794721407624634, "grad_norm": 18.778431032287738, "kl": 1.517578125, "learning_rate": 1.0115590777665522e-07, "loss": 0.3054, "reward": 2.205144762992859, "reward_std": 0.09740199614316225, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01881374605000019, "rewards/tag_count_reward": 0.9739583432674408, "step": 1336 }, { "clip_ratio": 0.0, "completion_length": 507.60418701171875, "epoch": 0.9802052785923754, "grad_norm": 11.377846780811359, "kl": 2.37109375, "learning_rate": 1.0107484960514692e-07, "loss": 0.5538, "reward": 2.415365695953369, "reward_std": 0.2122260332107544, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.03428706340491772, "rewards/tag_count_reward": 0.9635416865348816, "step": 1337 }, { "clip_ratio": 0.0, "completion_length": 441.125, "epoch": 0.9809384164222874, "grad_norm": 6.413201892131247, "kl": 0.517578125, "learning_rate": 1.009967343933688e-07, "loss": 0.0587, "reward": 2.2234453558921814, "reward_std": 0.04166480875574052, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.016138038132339716, "rewards/tag_count_reward": 0.9895833432674408, "step": 1338 }, { "clip_ratio": 0.0, "completion_length": 432.1666717529297, "epoch": 0.9816715542521994, "grad_norm": 13.366893199956893, "kl": 0.5498046875, "learning_rate": 1.0092156265341051e-07, "loss": 0.0339, "reward": 2.7342896461486816, "reward_std": 0.010294946376234293, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.01571043487638235, "rewards/tag_count_reward": 1.0, "step": 1339 }, { "clip_ratio": 0.0, "completion_length": 528.2083435058594, "epoch": 0.9824046920821115, "grad_norm": 26.73321311740256, "kl": 2.1796875, "learning_rate": 1.0084933487806555e-07, "loss": 0.132, "reward": 2.178415536880493, "reward_std": 0.24035960249602795, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.036862210370600224, "rewards/tag_count_reward": 0.9791666865348816, "step": 1340 }, { "clip_ratio": 0.0, "completion_length": 491.04168701171875, "epoch": 0.9831378299120235, "grad_norm": 17.211256661631793, "kl": 1.6953125, "learning_rate": 1.0078005154082807e-07, "loss": 0.5392, "reward": 2.8317149877548218, "reward_std": 0.3898848742246628, "rewards/accuracy_reward": 0.9375000298023224, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.04849330708384514, "rewards/tag_count_reward": 0.9635416865348816, "step": 1341 }, { "clip_ratio": 0.0, "completion_length": 463.3125, "epoch": 0.9838709677419355, "grad_norm": 17.9294618469925, "kl": 0.806640625, "learning_rate": 1.0071371309588976e-07, "loss": 0.2448, "reward": 2.6838122606277466, "reward_std": 0.1747398329898715, "rewards/accuracy_reward": 0.7291666716337204, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02278487477451563, "rewards/tag_count_reward": 0.984375, "step": 1342 }, { "clip_ratio": 0.0, "completion_length": 542.2291870117188, "epoch": 0.9846041055718475, "grad_norm": 14.081741076763183, "kl": 2.796875, "learning_rate": 1.006503199781369e-07, "loss": 0.4648, "reward": 2.298216700553894, "reward_std": 0.4470781087875366, "rewards/accuracy_reward": 0.4166666865348816, "rewards/reasoning_steps_reward": 0.9722222685813904, "rewards/repetition_penalty_reward": -0.03338063694536686, "rewards/tag_count_reward": 0.9427083730697632, "step": 1343 }, { "clip_ratio": 0.0, "completion_length": 544.0416870117188, "epoch": 0.9853372434017595, "grad_norm": 10.81673080758501, "kl": 2.078125, "learning_rate": 1.0058987260314736e-07, "loss": 0.2835, "reward": 2.3138773441314697, "reward_std": 0.31794362515211105, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.042025452479720116, "rewards/tag_count_reward": 0.9114583432674408, "step": 1344 }, { "clip_ratio": 0.0, "completion_length": 562.1458435058594, "epoch": 0.9860703812316716, "grad_norm": 20.449252275235782, "kl": 2.265625, "learning_rate": 1.0053237136718819e-07, "loss": 0.4601, "reward": 1.795183539390564, "reward_std": 0.45035097002983093, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9305556416511536, "rewards/repetition_penalty_reward": -0.05724710505455732, "rewards/tag_count_reward": 0.9010416865348816, "step": 1345 }, { "clip_ratio": 0.0, "completion_length": 463.3125, "epoch": 0.9868035190615836, "grad_norm": 12.875001769568685, "kl": 1.662109375, "learning_rate": 1.0047781664721258e-07, "loss": 0.252, "reward": 2.38944935798645, "reward_std": 0.32722827047109604, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.023745193146169186, "rewards/tag_count_reward": 0.9687500298023224, "step": 1346 }, { "clip_ratio": 0.0, "completion_length": 442.1875, "epoch": 0.9875366568914956, "grad_norm": 9.587097766988357, "kl": 0.685546875, "learning_rate": 1.004262088008578e-07, "loss": 0.0343, "reward": 1.9816942811012268, "reward_std": 0.010914890561252832, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.018305765464901924, "rewards/tag_count_reward": 1.0, "step": 1347 }, { "clip_ratio": 0.0, "completion_length": 439.47918701171875, "epoch": 0.9882697947214076, "grad_norm": 7.081412711269489, "kl": 0.919921875, "learning_rate": 1.0037754816644265e-07, "loss": 0.189, "reward": 2.4098252058029175, "reward_std": 0.17200952302664518, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.027674859389662743, "rewards/tag_count_reward": 0.9791666865348816, "step": 1348 }, { "clip_ratio": 0.0, "completion_length": 449.3750305175781, "epoch": 0.9890029325513197, "grad_norm": 8.416775160633, "kl": 0.701171875, "learning_rate": 1.003318350629653e-07, "loss": 0.1907, "reward": 2.4872329235076904, "reward_std": 0.1283271312713623, "rewards/accuracy_reward": 0.5208333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.023183775134384632, "rewards/tag_count_reward": 0.9895833432674408, "step": 1349 }, { "clip_ratio": 0.0, "completion_length": 486.7916717529297, "epoch": 0.9897360703812317, "grad_norm": 9.157980739902086, "kl": 0.70703125, "learning_rate": 1.002890697901011e-07, "loss": 0.3773, "reward": 1.9337705969810486, "reward_std": 0.18084391951560974, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02282659476622939, "rewards/tag_count_reward": 0.9635416865348816, "step": 1350 }, { "clip_ratio": 0.0, "completion_length": 470.33335876464844, "epoch": 0.9904692082111437, "grad_norm": 63.26862900641875, "kl": 0.46484375, "learning_rate": 1.0024925262820074e-07, "loss": 0.1247, "reward": 2.6610316038131714, "reward_std": 0.16902299923822284, "rewards/accuracy_reward": 0.7083333432674408, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02646848652511835, "rewards/tag_count_reward": 0.9791666865348816, "step": 1351 }, { "clip_ratio": 0.0, "completion_length": 430.7916717529297, "epoch": 0.9912023460410557, "grad_norm": 7.50035566920925, "kl": 0.78515625, "learning_rate": 1.0021238383828834e-07, "loss": 0.2537, "reward": 2.181891918182373, "reward_std": 0.17814004700630903, "rewards/accuracy_reward": 0.25, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.026441490277647972, "rewards/tag_count_reward": 0.9791666865348816, "step": 1352 }, { "clip_ratio": 0.0, "completion_length": 487.2500305175781, "epoch": 0.9919354838709677, "grad_norm": 16.736728648194653, "kl": 2.486328125, "learning_rate": 1.0017846366205987e-07, "loss": 0.4959, "reward": 2.3572378158569336, "reward_std": 0.33005889039486647, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9791666865348816, "rewards/repetition_penalty_reward": -0.028178916312754154, "rewards/tag_count_reward": 0.9479166865348816, "step": 1353 }, { "clip_ratio": 0.0, "completion_length": 499.375, "epoch": 0.9926686217008798, "grad_norm": 11.655766844116405, "kl": 1.046875, "learning_rate": 1.0014749232188132e-07, "loss": 0.2876, "reward": 1.95725417137146, "reward_std": 0.0797755979001522, "rewards/accuracy_reward": 0.0, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02712089940905571, "rewards/tag_count_reward": 0.984375, "step": 1354 }, { "clip_ratio": 0.0, "completion_length": 499.35418701171875, "epoch": 0.9934017595307918, "grad_norm": 8.054933338089828, "kl": 1.009765625, "learning_rate": 1.0011947002078743e-07, "loss": 0.432, "reward": 2.3805216550827026, "reward_std": 0.3503710813820362, "rewards/accuracy_reward": 0.4583333432674408, "rewards/reasoning_steps_reward": 0.9861111342906952, "rewards/repetition_penalty_reward": -0.03267285693436861, "rewards/tag_count_reward": 0.96875, "step": 1355 }, { "clip_ratio": 0.0, "completion_length": 457.0833435058594, "epoch": 0.9941348973607038, "grad_norm": 14.538818455565952, "kl": 1.00390625, "learning_rate": 1.000943969424804e-07, "loss": 0.2383, "reward": 2.6256481409072876, "reward_std": 0.301572248339653, "rewards/accuracy_reward": 0.6875000298023224, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.03233811724931002, "rewards/tag_count_reward": 0.984375, "step": 1356 }, { "clip_ratio": 0.0, "completion_length": 565.1666870117188, "epoch": 0.9948680351906158, "grad_norm": 22.732330191271593, "kl": 2.26171875, "learning_rate": 1.0007227325132845e-07, "loss": 0.3831, "reward": 2.1746811270713806, "reward_std": 0.5294123976491392, "rewards/accuracy_reward": 0.375, "rewards/reasoning_steps_reward": 0.9375, "rewards/repetition_penalty_reward": -0.049277255311608315, "rewards/tag_count_reward": 0.9114583432674408, "step": 1357 }, { "clip_ratio": 0.0, "completion_length": 470.5416717529297, "epoch": 0.9956011730205279, "grad_norm": 16.24419619781198, "kl": 0.76953125, "learning_rate": 1.00053099092365e-07, "loss": 0.199, "reward": 1.9626244902610779, "reward_std": 0.10940095037221909, "rewards/accuracy_reward": 0.02083333395421505, "rewards/reasoning_steps_reward": 0.9861111044883728, "rewards/repetition_penalty_reward": -0.023486592806875706, "rewards/tag_count_reward": 0.9791666865348816, "step": 1358 }, { "clip_ratio": 0.0, "completion_length": 547.3125305175781, "epoch": 0.9963343108504399, "grad_norm": 17.350196853804658, "kl": 0.9150390625, "learning_rate": 1.0003687459128753e-07, "loss": 0.2186, "reward": 2.4492987990379333, "reward_std": 0.07192742079496384, "rewards/accuracy_reward": 0.5, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.029867922887206078, "rewards/tag_count_reward": 0.9791666865348816, "step": 1359 }, { "clip_ratio": 0.0, "completion_length": 580.7083435058594, "epoch": 0.9970674486803519, "grad_norm": 19.53300669702914, "kl": 2.453125, "learning_rate": 1.0002359985445691e-07, "loss": 0.6501, "reward": 2.0493897199630737, "reward_std": 0.4308444410562515, "rewards/accuracy_reward": 0.2083333432674408, "rewards/reasoning_steps_reward": 0.9652778208255768, "rewards/repetition_penalty_reward": -0.04088811669498682, "rewards/tag_count_reward": 0.9166666865348816, "step": 1360 }, { "clip_ratio": 0.0, "completion_length": 470.02085876464844, "epoch": 0.9978005865102639, "grad_norm": 13.39377168275178, "kl": 0.537109375, "learning_rate": 1.0001327496889658e-07, "loss": 0.2151, "reward": 2.9133403301239014, "reward_std": 0.23964431881904602, "rewards/accuracy_reward": 0.9583333730697632, "rewards/reasoning_steps_reward": 0.9930555820465088, "rewards/repetition_penalty_reward": -0.02242350485175848, "rewards/tag_count_reward": 0.9843750298023224, "step": 1361 }, { "clip_ratio": 0.0, "completion_length": 433.8958435058594, "epoch": 0.998533724340176, "grad_norm": 136.78654651795395, "kl": 0.955078125, "learning_rate": 1.0000590000229203e-07, "loss": 0.2355, "reward": 2.4190428256988525, "reward_std": 0.2147923707962036, "rewards/accuracy_reward": 0.4791666865348816, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.02366560883820057, "rewards/tag_count_reward": 0.9635416865348816, "step": 1362 }, { "clip_ratio": 0.0, "completion_length": 563.8571624755859, "epoch": 0.999266862170088, "grad_norm": 19.87535727669407, "kl": 1.158203125, "learning_rate": 1.000014750029904e-07, "loss": 0.2502, "reward": 2.7152678966522217, "reward_std": 0.07797089219093323, "rewards/accuracy_reward": 0.75, "rewards/reasoning_steps_reward": 1.0, "rewards/repetition_penalty_reward": -0.013898835051804781, "rewards/tag_count_reward": 0.9791666865348816, "step": 1363 }, { "epoch": 0.999266862170088, "step": 1363, "total_flos": 0.0, "train_loss": 2.0727225511886656, "train_runtime": 64632.8349, "train_samples_per_second": 0.084, "train_steps_per_second": 0.021 } ], "logging_steps": 1, "max_steps": 1364, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 200, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }