{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.662721893491124, "eval_steps": 200.0, "global_step": 3600, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 193.0, "completions/mean_length": 76.0703125, "completions/min_length": 18.0, "epoch": 0.0007396449704142012, "frac_reward_zero_std": 0.0, "grad_norm": 3.671875, "kl": 0.0, "learning_rate": 2.9585798816568045e-09, "loss": -3.5390257835388184e-08, "reward": 1.4256246089935303, "reward_std": 0.3674412667751312, "rewards/FidelityReward/mean": 0.6892863512039185, "rewards/FidelityReward/std": 0.20130027830600739, "rewards/JudgeFidelityReward/mean": 0.6054890155792236, "rewards/JudgeFidelityReward/std": 0.30885493755340576, "rewards/SelfEvolvingFormatReward/mean": 0.8671875, "rewards/SelfEvolvingFormatReward/std": 0.3400367796421051, "step": 1 }, { "clip_ratio/high_max": 0.0003869968932121992, "clip_ratio/high_mean": 7.72563143982552e-05, "clip_ratio/low_mean": 3.048780490644276e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010774411930469796, "completions/clipped_ratio": 0.0, "completions/max_length": 231.5, "completions/mean_length": 74.9140625, "completions/min_length": 18.0, "epoch": 0.0036982248520710057, "frac_reward_zero_std": 0.0, "grad_norm": 3.78125, "kl": 0.00043141010974068195, "learning_rate": 1.4792899408284022e-08, "loss": -8.471141336485744e-05, "reward": 1.3640888929367065, "reward_std": 0.3614848405122757, "rewards/FidelityReward/mean": 0.6683639287948608, "rewards/FidelityReward/std": 0.20970967411994934, "rewards/JudgeFidelityReward/mean": 0.5594186186790466, "rewards/JudgeFidelityReward/std": 0.30836305022239685, "rewards/SelfEvolvingFormatReward/mean": 0.83203125, "rewards/SelfEvolvingFormatReward/std": 0.37442445755004883, "step": 5 }, { "clip_ratio/high_max": 0.0009308077860623598, "clip_ratio/high_mean": 0.0002875228296034038, "clip_ratio/low_mean": 3.985902876593173e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00032738185836933553, "completions/clipped_ratio": 0.0, "completions/max_length": 210.5, "completions/mean_length": 78.28125, "completions/min_length": 18.0, "epoch": 0.0073964497041420114, "frac_reward_zero_std": 0.0, "grad_norm": 5.0, "kl": 0.0006112438510172069, "learning_rate": 2.9585798816568044e-08, "loss": 0.00023797880858182906, "reward": 1.4635030031204224, "reward_std": 0.37530525028705597, "rewards/FidelityReward/mean": 0.7540956437587738, "rewards/FidelityReward/std": 0.19652967900037766, "rewards/JudgeFidelityReward/mean": 0.6102209091186523, "rewards/JudgeFidelityReward/std": 0.3318042457103729, "rewards/SelfEvolvingFormatReward/mean": 0.80859375, "rewards/SelfEvolvingFormatReward/std": 0.39215683937072754, "step": 10 }, { "clip_ratio/high_max": 0.0006585306487977505, "clip_ratio/high_mean": 0.00023432199959643184, "clip_ratio/low_mean": 5.996526451781392e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00029428726993501185, "completions/clipped_ratio": 0.0, "completions/max_length": 176.33333333333334, "completions/mean_length": 78.26822916666667, "completions/min_length": 18.0, "epoch": 0.011094674556213017, "frac_reward_zero_std": 0.0, "grad_norm": 3.515625, "kl": 0.0006082158535718918, "learning_rate": 4.437869822485207e-08, "loss": 7.99284316599369e-05, "reward": 1.4291691382726033, "reward_std": 0.36614983280499774, "rewards/FidelityReward/mean": 0.7301369309425354, "rewards/FidelityReward/std": 0.20399118959903717, "rewards/JudgeFidelityReward/mean": 0.5803560415903727, "rewards/JudgeFidelityReward/std": 0.34638898571332294, "rewards/SelfEvolvingFormatReward/mean": 0.8177083333333334, "rewards/SelfEvolvingFormatReward/std": 0.38587407271067303, "step": 15 }, { "clip_ratio/high_max": 0.0004901968408375978, "clip_ratio/high_mean": 0.00013072149886284024, "clip_ratio/low_mean": 0.0001148064315202646, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002455279347486794, "completions/clipped_ratio": 0.0, "completions/max_length": 228.0, "completions/mean_length": 74.982421875, "completions/min_length": 18.0, "epoch": 0.014792899408284023, "frac_reward_zero_std": 0.0, "grad_norm": 4.40625, "kl": 0.0006263210554607213, "learning_rate": 5.917159763313609e-08, "loss": 1.3579614460468291e-06, "reward": 1.4216499328613281, "reward_std": 0.33751754462718964, "rewards/FidelityReward/mean": 0.690717875957489, "rewards/FidelityReward/std": 0.21761607378721237, "rewards/JudgeFidelityReward/mean": 0.5966298580169678, "rewards/JudgeFidelityReward/std": 0.30593594908714294, "rewards/SelfEvolvingFormatReward/mean": 0.865234375, "rewards/SelfEvolvingFormatReward/std": 0.34184157848358154, "step": 20 }, { "clip_ratio/high_max": 0.00041467188857495784, "clip_ratio/high_mean": 0.00013362044119276107, "clip_ratio/low_mean": 9.85076228971593e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023212806554511188, "completions/clipped_ratio": 0.0, "completions/max_length": 209.33333333333334, "completions/mean_length": 78.25390625, "completions/min_length": 18.0, "epoch": 0.01849112426035503, "frac_reward_zero_std": 0.0, "grad_norm": 3.703125, "kl": 0.0005871019558981061, "learning_rate": 7.396449704142011e-08, "loss": 0.000105749792419374, "reward": 1.4413857062657673, "reward_std": 0.34226512908935547, "rewards/FidelityReward/mean": 0.7129021485646566, "rewards/FidelityReward/std": 0.21335932612419128, "rewards/JudgeFidelityReward/mean": 0.6067067782084147, "rewards/JudgeFidelityReward/std": 0.3204115529855092, "rewards/SelfEvolvingFormatReward/mean": 0.8502604166666666, "rewards/SelfEvolvingFormatReward/std": 0.3559674024581909, "step": 25 }, { "clip_ratio/high_max": 0.0008985038031823934, "clip_ratio/high_mean": 0.00029577945242635905, "clip_ratio/low_mean": 0.0001300744363106787, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004258538829162717, "completions/clipped_ratio": 0.0, "completions/max_length": 174.0, "completions/mean_length": 77.603515625, "completions/min_length": 18.0, "epoch": 0.022189349112426034, "frac_reward_zero_std": 0.0, "grad_norm": 4.53125, "kl": 0.0005966695491224528, "learning_rate": 8.875739644970414e-08, "loss": 0.0001702972687780857, "reward": 1.4576577544212341, "reward_std": 0.36230581998825073, "rewards/FidelityReward/mean": 0.728948175907135, "rewards/FidelityReward/std": 0.19791119545698166, "rewards/JudgeFidelityReward/mean": 0.6214816570281982, "rewards/JudgeFidelityReward/std": 0.3158075362443924, "rewards/SelfEvolvingFormatReward/mean": 0.8359375, "rewards/SelfEvolvingFormatReward/std": 0.36969611048698425, "step": 30 }, { "clip_ratio/high_max": 0.00038500839145854114, "clip_ratio/high_mean": 0.00010735052346717566, "clip_ratio/low_mean": 6.112714181654156e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016847766237333418, "completions/clipped_ratio": 0.0, "completions/max_length": 179.66666666666666, "completions/mean_length": 77.6328125, "completions/min_length": 18.0, "epoch": 0.025887573964497042, "frac_reward_zero_std": 0.0, "grad_norm": 3.765625, "kl": 0.0005852889502421021, "learning_rate": 1.0355029585798816e-07, "loss": 0.0001043007243424654, "reward": 1.4882162809371948, "reward_std": 0.33131828904151917, "rewards/FidelityReward/mean": 0.74555637439092, "rewards/FidelityReward/std": 0.19814972579479218, "rewards/JudgeFidelityReward/mean": 0.6272468566894531, "rewards/JudgeFidelityReward/std": 0.31311507026354474, "rewards/SelfEvolvingFormatReward/mean": 0.8580729166666666, "rewards/SelfEvolvingFormatReward/std": 0.3470127085844676, "step": 35 }, { "clip_ratio/high_max": 0.0008879650617018342, "clip_ratio/high_mean": 0.0002974147733766586, "clip_ratio/low_mean": 0.00011814615718321874, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004155609291046858, "completions/clipped_ratio": 0.0, "completions/max_length": 205.0, "completions/mean_length": 75.609375, "completions/min_length": 18.0, "epoch": 0.029585798816568046, "frac_reward_zero_std": 0.0, "grad_norm": 4.84375, "kl": 0.0006030320539139212, "learning_rate": 1.1834319526627217e-07, "loss": 2.7182046324014665e-06, "reward": 1.4277613162994385, "reward_std": 0.36199377477169037, "rewards/FidelityReward/mean": 0.7107739448547363, "rewards/FidelityReward/std": 0.2109760344028473, "rewards/JudgeFidelityReward/mean": 0.5960841178894043, "rewards/JudgeFidelityReward/std": 0.3255346864461899, "rewards/SelfEvolvingFormatReward/mean": 0.837890625, "rewards/SelfEvolvingFormatReward/std": 0.3676500618457794, "step": 40 }, { "clip_ratio/high_max": 0.0005873812828212976, "clip_ratio/high_mean": 0.0002048984752036631, "clip_ratio/low_mean": 4.2198103619739415e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002470965730026364, "completions/clipped_ratio": 0.0, "completions/max_length": 197.66666666666666, "completions/mean_length": 78.80859375, "completions/min_length": 18.0, "epoch": 0.03328402366863906, "frac_reward_zero_std": 0.0, "grad_norm": 4.84375, "kl": 0.0006142814294435084, "learning_rate": 1.331360946745562e-07, "loss": -8.607950294390321e-05, "reward": 1.4356428782145183, "reward_std": 0.3711263835430145, "rewards/FidelityReward/mean": 0.7383571863174438, "rewards/FidelityReward/std": 0.2049396683772405, "rewards/JudgeFidelityReward/mean": 0.591186006863912, "rewards/JudgeFidelityReward/std": 0.3356272280216217, "rewards/SelfEvolvingFormatReward/mean": 0.8033854166666666, "rewards/SelfEvolvingFormatReward/std": 0.3929189344247182, "step": 45 }, { "clip_ratio/high_max": 0.0011678082635626196, "clip_ratio/high_mean": 0.0003107733558863401, "clip_ratio/low_mean": 8.101938583422452e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003917927504517138, "completions/clipped_ratio": 0.0, "completions/max_length": 191.0, "completions/mean_length": 77.58984375, "completions/min_length": 18.0, "epoch": 0.03698224852071006, "frac_reward_zero_std": 0.0, "grad_norm": 3.109375, "kl": 0.0006190464366227389, "learning_rate": 1.4792899408284022e-07, "loss": 3.992556594312191e-06, "reward": 1.4567926526069641, "reward_std": 0.3507862985134125, "rewards/FidelityReward/mean": 0.7188418805599213, "rewards/FidelityReward/std": 0.21651969850063324, "rewards/JudgeFidelityReward/mean": 0.6243390440940857, "rewards/JudgeFidelityReward/std": 0.3272186815738678, "rewards/SelfEvolvingFormatReward/mean": 0.8515625, "rewards/SelfEvolvingFormatReward/std": 0.3541274219751358, "step": 50 }, { "clip_ratio/high_max": 0.00045795710757374765, "clip_ratio/high_mean": 0.00014709476381540297, "clip_ratio/low_mean": 0.0001755973498802632, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003226921195164323, "completions/clipped_ratio": 0.0, "completions/max_length": 246.0, "completions/mean_length": 79.66927083333333, "completions/min_length": 18.0, "epoch": 0.040680473372781065, "frac_reward_zero_std": 0.0, "grad_norm": 6.1875, "kl": 0.0006364648463204503, "learning_rate": 1.6272189349112426e-07, "loss": -0.00015164623036980628, "reward": 1.4810880025227864, "reward_std": 0.36538488666216534, "rewards/FidelityReward/mean": 0.7392007112503052, "rewards/FidelityReward/std": 0.20198150972525278, "rewards/JudgeFidelityReward/mean": 0.6374204754829407, "rewards/JudgeFidelityReward/std": 0.329936941464742, "rewards/SelfEvolvingFormatReward/mean": 0.8463541666666666, "rewards/SelfEvolvingFormatReward/std": 0.36091674367586773, "step": 55 }, { "clip_ratio/high_max": 0.0016810505418106913, "clip_ratio/high_mean": 0.0004927146132104099, "clip_ratio/low_mean": 0.00011921913392143324, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0006119337282143533, "completions/clipped_ratio": 0.0, "completions/max_length": 220.0, "completions/mean_length": 77.142578125, "completions/min_length": 18.0, "epoch": 0.04437869822485207, "frac_reward_zero_std": 0.0, "grad_norm": 3.734375, "kl": 0.0006852958584204316, "learning_rate": 1.7751479289940827e-07, "loss": -0.00018674861639738082, "reward": 1.479350745677948, "reward_std": 0.32729165256023407, "rewards/FidelityReward/mean": 0.7229913771152496, "rewards/FidelityReward/std": 0.21209533512592316, "rewards/JudgeFidelityReward/mean": 0.6338124871253967, "rewards/JudgeFidelityReward/std": 0.29375705122947693, "rewards/SelfEvolvingFormatReward/mean": 0.87890625, "rewards/SelfEvolvingFormatReward/std": 0.326654851436615, "step": 60 }, { "clip_ratio/high_max": 0.00031302616698667405, "clip_ratio/high_mean": 6.004877504892647e-05, "clip_ratio/low_mean": 6.60366567899473e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012608543038368226, "completions/clipped_ratio": 0.0, "completions/max_length": 174.66666666666666, "completions/mean_length": 75.90364583333333, "completions/min_length": 18.0, "epoch": 0.04807692307692308, "frac_reward_zero_std": 0.0, "grad_norm": 3.890625, "kl": 0.0007252910640090704, "learning_rate": 1.9230769230769231e-07, "loss": -9.892939124256372e-05, "reward": 1.4579964876174927, "reward_std": 0.37911028663317364, "rewards/FidelityReward/mean": 0.7221436301867167, "rewards/FidelityReward/std": 0.21909584601720175, "rewards/JudgeFidelityReward/mean": 0.6201432148615519, "rewards/JudgeFidelityReward/std": 0.32359103361765545, "rewards/SelfEvolvingFormatReward/mean": 0.8515625, "rewards/SelfEvolvingFormatReward/std": 0.35589608550071716, "step": 65 }, { "clip_ratio/high_max": 0.0005156282568350434, "clip_ratio/high_mean": 0.00022456933511421083, "clip_ratio/low_mean": 8.012482576305046e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003046941536013037, "completions/clipped_ratio": 0.0, "completions/max_length": 161.5, "completions/mean_length": 77.025390625, "completions/min_length": 18.0, "epoch": 0.051775147928994084, "frac_reward_zero_std": 0.0, "grad_norm": 3.359375, "kl": 0.0007379044312983751, "learning_rate": 2.0710059171597633e-07, "loss": -0.00011538814287632704, "reward": 1.5485591292381287, "reward_std": 0.31653086841106415, "rewards/FidelityReward/mean": 0.7626369595527649, "rewards/FidelityReward/std": 0.2016252800822258, "rewards/JudgeFidelityReward/mean": 0.6773132085800171, "rewards/JudgeFidelityReward/std": 0.29787328839302063, "rewards/SelfEvolvingFormatReward/mean": 0.89453125, "rewards/SelfEvolvingFormatReward/std": 0.3077581524848938, "step": 70 }, { "clip_ratio/high_max": 0.0006739318370819092, "clip_ratio/high_mean": 0.0002765225013718009, "clip_ratio/low_mean": 4.532605380518362e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00032184855081140995, "completions/clipped_ratio": 0.0, "completions/max_length": 182.66666666666666, "completions/mean_length": 76.59635416666667, "completions/min_length": 18.0, "epoch": 0.05547337278106509, "frac_reward_zero_std": 0.0, "grad_norm": 5.40625, "kl": 0.0008889023563824594, "learning_rate": 2.2189349112426034e-07, "loss": 2.839486114680767e-07, "reward": 1.5371168851852417, "reward_std": 0.322932630777359, "rewards/FidelityReward/mean": 0.7711425622304281, "rewards/FidelityReward/std": 0.1799267679452896, "rewards/JudgeFidelityReward/mean": 0.6582507093747457, "rewards/JudgeFidelityReward/std": 0.30686138073603314, "rewards/SelfEvolvingFormatReward/mean": 0.8736979166666666, "rewards/SelfEvolvingFormatReward/std": 0.3320201138655345, "step": 75 }, { "clip_ratio/high_max": 0.0008634536527097226, "clip_ratio/high_mean": 0.00036664897343143823, "clip_ratio/low_mean": 0.0001285171718336642, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004951661685481668, "completions/clipped_ratio": 0.0, "completions/max_length": 209.0, "completions/mean_length": 74.0234375, "completions/min_length": 18.0, "epoch": 0.05917159763313609, "frac_reward_zero_std": 0.0, "grad_norm": 4.46875, "kl": 0.0009641199023462832, "learning_rate": 2.3668639053254435e-07, "loss": 4.4167973101139066e-08, "reward": 1.403045892715454, "reward_std": 0.35805201530456543, "rewards/FidelityReward/mean": 0.6812058389186859, "rewards/FidelityReward/std": 0.21219906955957413, "rewards/JudgeFidelityReward/mean": 0.5784457921981812, "rewards/JudgeFidelityReward/std": 0.31330087780952454, "rewards/SelfEvolvingFormatReward/mean": 0.865234375, "rewards/SelfEvolvingFormatReward/std": 0.3411656767129898, "step": 80 }, { "clip_ratio/high_max": 0.0006637629354372621, "clip_ratio/high_mean": 0.00017012786120176315, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017012786120176315, "completions/clipped_ratio": 0.0, "completions/max_length": 215.33333333333334, "completions/mean_length": 74.52994791666667, "completions/min_length": 18.0, "epoch": 0.0628698224852071, "frac_reward_zero_std": 0.0, "grad_norm": 4.375, "kl": 0.0010848390636965632, "learning_rate": 2.514792899408284e-07, "loss": 6.076069548726082e-05, "reward": 1.5213927030563354, "reward_std": 0.3201733926932017, "rewards/FidelityReward/mean": 0.750914732615153, "rewards/FidelityReward/std": 0.2059360792239507, "rewards/JudgeFidelityReward/mean": 0.6594454447428385, "rewards/JudgeFidelityReward/std": 0.3134966393311818, "rewards/SelfEvolvingFormatReward/mean": 0.8815104166666666, "rewards/SelfEvolvingFormatReward/std": 0.3229447702566783, "step": 85 }, { "clip_ratio/high_max": 0.0006265180069021881, "clip_ratio/high_mean": 0.00011194620456080883, "clip_ratio/low_mean": 8.717153395991773e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019911773269996048, "completions/clipped_ratio": 0.0, "completions/max_length": 196.0, "completions/mean_length": 80.421875, "completions/min_length": 18.0, "epoch": 0.06656804733727811, "frac_reward_zero_std": 0.0, "grad_norm": 3.171875, "kl": 0.0012909778626635671, "learning_rate": 2.662721893491124e-07, "loss": 6.761741824448108e-05, "reward": 1.542263150215149, "reward_std": 0.30513499677181244, "rewards/FidelityReward/mean": 0.7501616775989532, "rewards/FidelityReward/std": 0.20223476737737656, "rewards/JudgeFidelityReward/mean": 0.6896717548370361, "rewards/JudgeFidelityReward/std": 0.2823178917169571, "rewards/SelfEvolvingFormatReward/mean": 0.89453125, "rewards/SelfEvolvingFormatReward/std": 0.3066966086626053, "step": 90 }, { "clip_ratio/high_max": 0.0005196209880523383, "clip_ratio/high_mean": 0.00015159674949245528, "clip_ratio/low_mean": 2.095557429129258e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017255232232855634, "completions/clipped_ratio": 0.0, "completions/max_length": 166.0, "completions/mean_length": 74.48046875, "completions/min_length": 18.0, "epoch": 0.07026627218934911, "frac_reward_zero_std": 0.0, "grad_norm": 4.25, "kl": 0.0016055097104981542, "learning_rate": 2.810650887573964e-07, "loss": 0.000196324591524899, "reward": 1.5246152480443318, "reward_std": 0.29590947429339093, "rewards/FidelityReward/mean": 0.7337117592493693, "rewards/FidelityReward/std": 0.22528647383054098, "rewards/JudgeFidelityReward/mean": 0.666442354520162, "rewards/JudgeFidelityReward/std": 0.3001788755257924, "rewards/SelfEvolvingFormatReward/mean": 0.9153645833333334, "rewards/SelfEvolvingFormatReward/std": 0.26931790510813397, "step": 95 }, { "clip_ratio/high_max": 0.0009233135147951544, "clip_ratio/high_mean": 0.00021089569781906903, "clip_ratio/low_mean": 0.00015629476547474042, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000367190467659384, "completions/clipped_ratio": 0.0, "completions/max_length": 187.5, "completions/mean_length": 76.751953125, "completions/min_length": 18.0, "epoch": 0.07396449704142012, "frac_reward_zero_std": 0.0, "grad_norm": 4.0625, "kl": 0.0017562908586114645, "learning_rate": 2.9585798816568045e-07, "loss": 0.00032348157837986947, "reward": 1.4658676385879517, "reward_std": 0.280498206615448, "rewards/FidelityReward/mean": 0.6774753332138062, "rewards/FidelityReward/std": 0.22155186533927917, "rewards/JudgeFidelityReward/mean": 0.6353783309459686, "rewards/JudgeFidelityReward/std": 0.27386045455932617, "rewards/SelfEvolvingFormatReward/mean": 0.94140625, "rewards/SelfEvolvingFormatReward/std": 0.23532284796237946, "step": 100 }, { "clip_ratio/high_max": 0.0004778528003953397, "clip_ratio/high_mean": 0.00013738811248913406, "clip_ratio/low_mean": 8.551625942345708e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000222904363181442, "completions/clipped_ratio": 0.0, "completions/max_length": 173.33333333333334, "completions/mean_length": 76.8828125, "completions/min_length": 18.0, "epoch": 0.07766272189349112, "frac_reward_zero_std": 0.0, "grad_norm": 2.765625, "kl": 0.0022867013700306415, "learning_rate": 3.1065088757396443e-07, "loss": 3.662020899355412e-05, "reward": 1.5086064338684082, "reward_std": 0.28400980432828266, "rewards/FidelityReward/mean": 0.7085907459259033, "rewards/FidelityReward/std": 0.20379472275575003, "rewards/JudgeFidelityReward/mean": 0.65471883614858, "rewards/JudgeFidelityReward/std": 0.26751768092314404, "rewards/SelfEvolvingFormatReward/mean": 0.9453125, "rewards/SelfEvolvingFormatReward/std": 0.22603721916675568, "step": 105 }, { "clip_ratio/high_max": 0.0009092223132029176, "clip_ratio/high_mean": 0.000234163214918226, "clip_ratio/low_mean": 0.00010074989113491028, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003349131206050515, "completions/clipped_ratio": 0.0, "completions/max_length": 203.0, "completions/mean_length": 74.650390625, "completions/min_length": 18.0, "epoch": 0.08136094674556213, "frac_reward_zero_std": 0.0, "grad_norm": 3.28125, "kl": 0.0021838200045749545, "learning_rate": 3.254437869822485e-07, "loss": 8.953195065259933e-05, "reward": 1.5663861632347107, "reward_std": 0.25924045592546463, "rewards/FidelityReward/mean": 0.7391667366027832, "rewards/FidelityReward/std": 0.1978147029876709, "rewards/JudgeFidelityReward/mean": 0.6993606984615326, "rewards/JudgeFidelityReward/std": 0.23920167237520218, "rewards/SelfEvolvingFormatReward/mean": 0.955078125, "rewards/SelfEvolvingFormatReward/std": 0.20705200731754303, "step": 110 }, { "clip_ratio/high_max": 0.00035654244711622594, "clip_ratio/high_mean": 8.581429137848318e-05, "clip_ratio/low_mean": 2.0177561964374036e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010599185479804873, "completions/clipped_ratio": 0.0, "completions/max_length": 210.33333333333334, "completions/mean_length": 76.328125, "completions/min_length": 18.0, "epoch": 0.08505917159763314, "frac_reward_zero_std": 0.0, "grad_norm": 2.703125, "kl": 0.0025876018684357406, "learning_rate": 3.4023668639053256e-07, "loss": 0.00015779398381710052, "reward": 1.6441294352213542, "reward_std": 0.2458642621835073, "rewards/FidelityReward/mean": 0.791078507900238, "rewards/FidelityReward/std": 0.17843289176623026, "rewards/JudgeFidelityReward/mean": 0.7490705053011576, "rewards/JudgeFidelityReward/std": 0.24861251811186472, "rewards/SelfEvolvingFormatReward/mean": 0.95703125, "rewards/SelfEvolvingFormatReward/std": 0.202033132314682, "step": 115 }, { "clip_ratio/high_max": 0.0008097812533378601, "clip_ratio/high_mean": 0.00025663914857432246, "clip_ratio/low_mean": 0.00016841468750499188, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00042505383025854827, "completions/clipped_ratio": 0.0, "completions/max_length": 237.0, "completions/mean_length": 76.19921875, "completions/min_length": 18.0, "epoch": 0.08875739644970414, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.003342992113903165, "learning_rate": 3.5502958579881655e-07, "loss": 6.172244902700186e-05, "reward": 1.5805195569992065, "reward_std": 0.24818717688322067, "rewards/FidelityReward/mean": 0.7410861849784851, "rewards/FidelityReward/std": 0.20152831077575684, "rewards/JudgeFidelityReward/mean": 0.7179290652275085, "rewards/JudgeFidelityReward/std": 0.23686103522777557, "rewards/SelfEvolvingFormatReward/mean": 0.9609375, "rewards/SelfEvolvingFormatReward/std": 0.19385965168476105, "step": 120 }, { "clip_ratio/high_max": 0.0005899485433474184, "clip_ratio/high_mean": 0.00016556976770516486, "clip_ratio/low_mean": 4.45947764092125e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021016455139033496, "completions/clipped_ratio": 0.0, "completions/max_length": 190.33333333333334, "completions/mean_length": 75.9140625, "completions/min_length": 18.0, "epoch": 0.09245562130177515, "frac_reward_zero_std": 0.0, "grad_norm": 3.484375, "kl": 0.0037140291649848224, "learning_rate": 3.698224852071006e-07, "loss": 0.00021550285164266824, "reward": 1.6331121524175007, "reward_std": 0.2311947246392568, "rewards/FidelityReward/mean": 0.7684813737869263, "rewards/FidelityReward/std": 0.19635804990927377, "rewards/JudgeFidelityReward/mean": 0.7474907437960306, "rewards/JudgeFidelityReward/std": 0.23119314511617026, "rewards/SelfEvolvingFormatReward/mean": 0.9817708333333334, "rewards/SelfEvolvingFormatReward/std": 0.12893695632616678, "step": 125 }, { "clip_ratio/high_max": 0.0006695815012790262, "clip_ratio/high_mean": 0.00018743461987469345, "clip_ratio/low_mean": 3.933192347176373e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022676654334645717, "completions/clipped_ratio": 0.0, "completions/max_length": 232.5, "completions/mean_length": 78.01171875, "completions/min_length": 22.5, "epoch": 0.09615384615384616, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.003960429551079869, "learning_rate": 3.8461538461538463e-07, "loss": 0.00011658230796456337, "reward": 1.5157253742218018, "reward_std": 0.275722473859787, "rewards/FidelityReward/mean": 0.6951830387115479, "rewards/FidelityReward/std": 0.1996193304657936, "rewards/JudgeFidelityReward/mean": 0.6703815758228302, "rewards/JudgeFidelityReward/std": 0.23143912106752396, "rewards/SelfEvolvingFormatReward/mean": 0.970703125, "rewards/SelfEvolvingFormatReward/std": 0.16638851910829544, "step": 130 }, { "clip_ratio/high_max": 0.0003530600690282881, "clip_ratio/high_mean": 0.00010106583358719945, "clip_ratio/low_mean": 1.9561815133783966e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012062764726579189, "completions/clipped_ratio": 0.0, "completions/max_length": 184.0, "completions/mean_length": 75.37109375, "completions/min_length": 30.0, "epoch": 0.09985207100591716, "frac_reward_zero_std": 0.0, "grad_norm": 3.15625, "kl": 0.0037666080985218286, "learning_rate": 3.994082840236686e-07, "loss": 0.00016040131449699402, "reward": 1.5316251913706462, "reward_std": 0.25091712673505145, "rewards/FidelityReward/mean": 0.698408285776774, "rewards/FidelityReward/std": 0.20659279823303223, "rewards/JudgeFidelityReward/mean": 0.6807567079861959, "rewards/JudgeFidelityReward/std": 0.2332766850789388, "rewards/SelfEvolvingFormatReward/mean": 0.9856770833333334, "rewards/SelfEvolvingFormatReward/std": 0.11704373856385548, "step": 135 }, { "clip_ratio/high_max": 0.000656886026263237, "clip_ratio/high_mean": 0.00012930409284308552, "clip_ratio/low_mean": 1.855976297520101e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014786385581828655, "completions/clipped_ratio": 0.0, "completions/max_length": 182.0, "completions/mean_length": 74.873046875, "completions/min_length": 34.0, "epoch": 0.10355029585798817, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.0035390025936067105, "learning_rate": 4.1420118343195265e-07, "loss": 0.0001448759576305747, "reward": 1.5512990355491638, "reward_std": 0.25688354671001434, "rewards/FidelityReward/mean": 0.7086969614028931, "rewards/FidelityReward/std": 0.1906806007027626, "rewards/JudgeFidelityReward/mean": 0.6949697434902191, "rewards/JudgeFidelityReward/std": 0.21443936973810196, "rewards/SelfEvolvingFormatReward/mean": 0.990234375, "rewards/SelfEvolvingFormatReward/std": 0.09338126704096794, "step": 140 }, { "clip_ratio/high_max": 0.00034949140390381215, "clip_ratio/high_mean": 0.00012190653942525386, "clip_ratio/low_mean": 1.8811135669238864e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014071767800487577, "completions/clipped_ratio": 0.0, "completions/max_length": 179.66666666666666, "completions/mean_length": 76.03255208333333, "completions/min_length": 21.333333333333332, "epoch": 0.10724852071005918, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.004208648251369596, "learning_rate": 4.289940828402367e-07, "loss": 0.0001065621618181467, "reward": 1.621166507403056, "reward_std": 0.23352229595184326, "rewards/FidelityReward/mean": 0.7568122744560242, "rewards/FidelityReward/std": 0.19883511463801065, "rewards/JudgeFidelityReward/mean": 0.7417292992273966, "rewards/JudgeFidelityReward/std": 0.2171912888685862, "rewards/SelfEvolvingFormatReward/mean": 0.9869791666666666, "rewards/SelfEvolvingFormatReward/std": 0.11330605794986089, "step": 145 }, { "clip_ratio/high_max": 0.0008546121651306748, "clip_ratio/high_mean": 0.00026121330156456677, "clip_ratio/low_mean": 2.072968491120264e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028194298211019485, "completions/clipped_ratio": 0.0, "completions/max_length": 180.5, "completions/mean_length": 76.30859375, "completions/min_length": 23.5, "epoch": 0.11094674556213018, "frac_reward_zero_std": 0.0, "grad_norm": 2.859375, "kl": 0.0046937459148466585, "learning_rate": 4.437869822485207e-07, "loss": 0.00022468063980340958, "reward": 1.6230171918869019, "reward_std": 0.2270878478884697, "rewards/FidelityReward/mean": 0.761558473110199, "rewards/FidelityReward/std": 0.1993263065814972, "rewards/JudgeFidelityReward/mean": 0.7404955625534058, "rewards/JudgeFidelityReward/std": 0.22186553478240967, "rewards/SelfEvolvingFormatReward/mean": 0.982421875, "rewards/SelfEvolvingFormatReward/std": 0.11841673403978348, "step": 150 }, { "clip_ratio/high_max": 0.0005609891028143466, "clip_ratio/high_mean": 0.000152421131497249, "clip_ratio/low_mean": 7.102480012690648e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002234459272585809, "completions/clipped_ratio": 0.0, "completions/max_length": 167.66666666666666, "completions/mean_length": 74.74739583333333, "completions/min_length": 18.0, "epoch": 0.11464497041420119, "frac_reward_zero_std": 0.0, "grad_norm": 3.109375, "kl": 0.0038529610261321067, "learning_rate": 4.585798816568047e-07, "loss": 0.0001837142277508974, "reward": 1.6152094999949138, "reward_std": 0.2258508950471878, "rewards/FidelityReward/mean": 0.7475018898646036, "rewards/FidelityReward/std": 0.20464515189329782, "rewards/JudgeFidelityReward/mean": 0.7445299625396729, "rewards/JudgeFidelityReward/std": 0.21239998439947763, "rewards/SelfEvolvingFormatReward/mean": 0.9908854166666666, "rewards/SelfEvolvingFormatReward/std": 0.09475251783927281, "step": 155 }, { "clip_ratio/high_max": 0.0009129192447289824, "clip_ratio/high_mean": 0.00019480377668514847, "clip_ratio/low_mean": 9.941382450051606e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00029421760118566456, "completions/clipped_ratio": 0.0, "completions/max_length": 236.5, "completions/mean_length": 77.837890625, "completions/min_length": 35.5, "epoch": 0.11834319526627218, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.00335396328009665, "learning_rate": 4.733727810650887e-07, "loss": 0.00014170492067933083, "reward": 1.5811875462532043, "reward_std": 0.21947431564331055, "rewards/FidelityReward/mean": 0.7218867540359497, "rewards/FidelityReward/std": 0.2275790497660637, "rewards/JudgeFidelityReward/mean": 0.7264140248298645, "rewards/JudgeFidelityReward/std": 0.23837125301361084, "rewards/SelfEvolvingFormatReward/mean": 0.9921875, "rewards/SelfEvolvingFormatReward/std": 0.08821486681699753, "step": 160 }, { "clip_ratio/high_max": 0.0003391335834749043, "clip_ratio/high_mean": 0.00010858398745767772, "clip_ratio/low_mean": 4.682278085965663e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001554067770484835, "completions/clipped_ratio": 0.0, "completions/max_length": 181.33333333333334, "completions/mean_length": 74.09895833333333, "completions/min_length": 27.333333333333332, "epoch": 0.1220414201183432, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.004568177415058017, "learning_rate": 4.881656804733728e-07, "loss": 0.0003632883541285992, "reward": 1.5487396319707234, "reward_std": 0.24543859561284384, "rewards/FidelityReward/mean": 0.7058204015096029, "rewards/FidelityReward/std": 0.1950870007276535, "rewards/JudgeFidelityReward/mean": 0.697557250658671, "rewards/JudgeFidelityReward/std": 0.20752751330534616, "rewards/SelfEvolvingFormatReward/mean": 0.98828125, "rewards/SelfEvolvingFormatReward/std": 0.08763878295818965, "step": 165 }, { "clip_ratio/high_max": 0.000723751331679523, "clip_ratio/high_mean": 0.00015111079846974462, "clip_ratio/low_mean": 7.954409811645746e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023065489367581903, "completions/clipped_ratio": 0.0, "completions/max_length": 182.0, "completions/mean_length": 76.32421875, "completions/min_length": 28.5, "epoch": 0.1257396449704142, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.004835613584145903, "learning_rate": 5.029585798816568e-07, "loss": 0.00027148635126650333, "reward": 1.5066508650779724, "reward_std": 0.25154024362564087, "rewards/FidelityReward/mean": 0.6782102584838867, "rewards/FidelityReward/std": 0.2135109156370163, "rewards/JudgeFidelityReward/mean": 0.6607874929904938, "rewards/JudgeFidelityReward/std": 0.229012593626976, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 170 }, { "clip_ratio/high_max": 0.0004487697500735521, "clip_ratio/high_mean": 9.188174444716424e-05, "clip_ratio/low_mean": 1.9485580560285597e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00011136732355225832, "completions/clipped_ratio": 0.0, "completions/max_length": 195.66666666666666, "completions/mean_length": 76.35416666666667, "completions/min_length": 30.0, "epoch": 0.1294378698224852, "frac_reward_zero_std": 0.0, "grad_norm": 2.5625, "kl": 0.00403714356943965, "learning_rate": 5.177514792899408e-07, "loss": 0.00032718414440751075, "reward": 1.6116866270701091, "reward_std": 0.22560793658097586, "rewards/FidelityReward/mean": 0.7444245219230652, "rewards/FidelityReward/std": 0.19309627513090769, "rewards/JudgeFidelityReward/mean": 0.7397325237592062, "rewards/JudgeFidelityReward/std": 0.19634168843428293, "rewards/SelfEvolvingFormatReward/mean": 0.9947916666666666, "rewards/SelfEvolvingFormatReward/std": 0.05677593996127447, "step": 175 }, { "clip_ratio/high_max": 0.0006952151074074209, "clip_ratio/high_mean": 0.00018168584647355602, "clip_ratio/low_mean": 0.00010482099023647606, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002865068381652236, "completions/clipped_ratio": 0.0, "completions/max_length": 200.5, "completions/mean_length": 78.23046875, "completions/min_length": 29.5, "epoch": 0.13313609467455623, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.004172865673899651, "learning_rate": 5.325443786982249e-07, "loss": 0.00020625563338398932, "reward": 1.579735279083252, "reward_std": 0.23965521901845932, "rewards/FidelityReward/mean": 0.7275005280971527, "rewards/FidelityReward/std": 0.21266812831163406, "rewards/JudgeFidelityReward/mean": 0.7103288769721985, "rewards/JudgeFidelityReward/std": 0.22931399941444397, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.0539139099419117, "step": 180 }, { "clip_ratio/high_max": 0.00048597900895401834, "clip_ratio/high_mean": 0.0001467286085244268, "clip_ratio/low_mean": 1.9379844889044762e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016610845341347158, "completions/clipped_ratio": 0.0, "completions/max_length": 207.33333333333334, "completions/mean_length": 74.88411458333333, "completions/min_length": 32.0, "epoch": 0.13683431952662722, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.004265676578506827, "learning_rate": 5.473372781065088e-07, "loss": 0.00029984773136675357, "reward": 1.67446768283844, "reward_std": 0.20402832329273224, "rewards/FidelityReward/mean": 0.7889950474103292, "rewards/FidelityReward/std": 0.17871600886185965, "rewards/JudgeFidelityReward/mean": 0.7774557669957479, "rewards/JudgeFidelityReward/std": 0.19036372502644858, "rewards/SelfEvolvingFormatReward/mean": 0.9934895833333334, "rewards/SelfEvolvingFormatReward/std": 0.07964324454466502, "step": 185 }, { "clip_ratio/high_max": 0.0005640282644890249, "clip_ratio/high_mean": 0.00021692929440177977, "clip_ratio/low_mean": 4.3073005508631466e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026000231155194343, "completions/clipped_ratio": 0.0, "completions/max_length": 236.5, "completions/mean_length": 74.51171875, "completions/min_length": 36.0, "epoch": 0.14053254437869822, "frac_reward_zero_std": 0.0, "grad_norm": 2.703125, "kl": 0.004636182775720954, "learning_rate": 5.621301775147928e-07, "loss": 0.00020830929279327392, "reward": 1.5857666730880737, "reward_std": 0.23736579716205597, "rewards/FidelityReward/mean": 0.7221141457557678, "rewards/FidelityReward/std": 0.1894068792462349, "rewards/JudgeFidelityReward/mean": 0.7331644594669342, "rewards/JudgeFidelityReward/std": 0.19472990185022354, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 190 }, { "clip_ratio/high_max": 0.000543058686889708, "clip_ratio/high_mean": 0.00014274955028668047, "clip_ratio/low_mean": 3.714605700224638e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017989560728892684, "completions/clipped_ratio": 0.0, "completions/max_length": 171.66666666666666, "completions/mean_length": 74.74869791666667, "completions/min_length": 32.666666666666664, "epoch": 0.14423076923076922, "frac_reward_zero_std": 0.0, "grad_norm": 2.3125, "kl": 0.004157255636528134, "learning_rate": 5.769230769230768e-07, "loss": 0.00016258005052804947, "reward": 1.5798383156458538, "reward_std": 0.24192858735720316, "rewards/FidelityReward/mean": 0.7249130805333456, "rewards/FidelityReward/std": 0.1957223912080129, "rewards/JudgeFidelityReward/mean": 0.7137567400932312, "rewards/JudgeFidelityReward/std": 0.20534354448318481, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 195 }, { "clip_ratio/high_max": 0.0007136451778933406, "clip_ratio/high_mean": 0.00013016032753512263, "clip_ratio/low_mean": 0.00013822944747516886, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002683897735551, "completions/clipped_ratio": 0.0, "completions/max_length": 182.0, "completions/mean_length": 72.427734375, "completions/min_length": 24.5, "epoch": 0.14792899408284024, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.006779930274933577, "learning_rate": 5.917159763313609e-07, "loss": 0.00045772669836878775, "reward": 1.5800213813781738, "reward_std": 0.2315586805343628, "rewards/FidelityReward/mean": 0.7223304212093353, "rewards/FidelityReward/std": 0.21326880902051926, "rewards/JudgeFidelityReward/mean": 0.721241295337677, "rewards/JudgeFidelityReward/std": 0.20918673276901245, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 200 }, { "clip_ratio/high_max": 0.00035588777391240003, "clip_ratio/high_mean": 9.926515631377696e-05, "clip_ratio/low_mean": 9.087424841709435e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019013939891010523, "completions/clipped_ratio": 0.0, "completions/max_length": 220.66666666666666, "completions/mean_length": 76.03776041666667, "completions/min_length": 33.333333333333336, "epoch": 0.15162721893491124, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.004310575081035495, "learning_rate": 6.065088757396449e-07, "loss": 8.594116661697626e-05, "reward": 1.5248489379882812, "reward_std": 0.24307898183663687, "rewards/FidelityReward/mean": 0.6884802579879761, "rewards/FidelityReward/std": 0.21808533867200217, "rewards/JudgeFidelityReward/mean": 0.6844561696052551, "rewards/JudgeFidelityReward/std": 0.2324487417936325, "rewards/SelfEvolvingFormatReward/mean": 0.98828125, "rewards/SelfEvolvingFormatReward/std": 0.08387244989474614, "step": 205 }, { "clip_ratio/high_max": 0.0008027448784559965, "clip_ratio/high_mean": 0.0001569948362885043, "clip_ratio/low_mean": 2.218278677901253e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017917762161232532, "completions/clipped_ratio": 0.0, "completions/max_length": 181.0, "completions/mean_length": 72.478515625, "completions/min_length": 25.0, "epoch": 0.15532544378698224, "frac_reward_zero_std": 0.015625, "grad_norm": 2.234375, "kl": 0.0057964538224041465, "learning_rate": 6.213017751479289e-07, "loss": 0.00035618292167782784, "reward": 1.5802724957466125, "reward_std": 0.2259870395064354, "rewards/FidelityReward/mean": 0.7233002781867981, "rewards/FidelityReward/std": 0.1956050619482994, "rewards/JudgeFidelityReward/mean": 0.7198038101196289, "rewards/JudgeFidelityReward/std": 0.2129446044564247, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 210 }, { "clip_ratio/high_max": 0.0004973528673872351, "clip_ratio/high_mean": 7.932765292935073e-05, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 7.932765292935073e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 188.66666666666666, "completions/mean_length": 73.29166666666667, "completions/min_length": 33.333333333333336, "epoch": 0.15902366863905326, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.0034227475989609957, "learning_rate": 6.360946745562131e-07, "loss": 0.00016083396039903163, "reward": 1.6271089712778728, "reward_std": 0.2132364163796107, "rewards/FidelityReward/mean": 0.7515268127123514, "rewards/FidelityReward/std": 0.1950797438621521, "rewards/JudgeFidelityReward/mean": 0.755070428053538, "rewards/JudgeFidelityReward/std": 0.19536665578683218, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 215 }, { "clip_ratio/high_max": 0.0006260376190766692, "clip_ratio/high_mean": 0.00015343615523306654, "clip_ratio/low_mean": 3.7354975938796996e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019079113117186352, "completions/clipped_ratio": 0.0, "completions/max_length": 225.0, "completions/mean_length": 76.646484375, "completions/min_length": 29.5, "epoch": 0.16272189349112426, "frac_reward_zero_std": 0.0, "grad_norm": 2.296875, "kl": 0.004407513746991753, "learning_rate": 6.50887573964497e-07, "loss": 0.00035992427729070187, "reward": 1.6081228256225586, "reward_std": 0.22748324275016785, "rewards/FidelityReward/mean": 0.7396865487098694, "rewards/FidelityReward/std": 0.19264580309391022, "rewards/JudgeFidelityReward/mean": 0.736872524023056, "rewards/JudgeFidelityReward/std": 0.19723959267139435, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 220 }, { "clip_ratio/high_max": 0.00035893141757696866, "clip_ratio/high_mean": 0.0001066614466253668, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001066614466253668, "completions/clipped_ratio": 0.0, "completions/max_length": 202.66666666666666, "completions/mean_length": 75.95963541666667, "completions/min_length": 32.333333333333336, "epoch": 0.16642011834319526, "frac_reward_zero_std": 0.0, "grad_norm": 2.734375, "kl": 0.0037112430203706027, "learning_rate": 6.65680473372781e-07, "loss": 0.0002257226500660181, "reward": 1.6131808757781982, "reward_std": 0.2212344358364741, "rewards/FidelityReward/mean": 0.7407519221305847, "rewards/FidelityReward/std": 0.19331168135007223, "rewards/JudgeFidelityReward/mean": 0.746159831682841, "rewards/JudgeFidelityReward/std": 0.19670240581035614, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 225 }, { "clip_ratio/high_max": 0.0008240769733674824, "clip_ratio/high_mean": 0.00020038667134940623, "clip_ratio/low_mean": 1.982553512789309e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022021221229806542, "completions/clipped_ratio": 0.0, "completions/max_length": 182.5, "completions/mean_length": 73.85546875, "completions/min_length": 22.5, "epoch": 0.17011834319526628, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.004782723682001233, "learning_rate": 6.804733727810651e-07, "loss": 0.00028980993665754794, "reward": 1.6581194996833801, "reward_std": 0.19545793533325195, "rewards/FidelityReward/mean": 0.7730163931846619, "rewards/FidelityReward/std": 0.1944940835237503, "rewards/JudgeFidelityReward/mean": 0.7721592783927917, "rewards/JudgeFidelityReward/std": 0.20489365607500076, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 230 }, { "clip_ratio/high_max": 0.00032628598855808375, "clip_ratio/high_mean": 7.802409236319363e-05, "clip_ratio/low_mean": 8.635731064714492e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016438140883110464, "completions/clipped_ratio": 0.0, "completions/max_length": 186.0, "completions/mean_length": 76.73567708333333, "completions/min_length": 29.666666666666668, "epoch": 0.17381656804733728, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.005551622714847326, "learning_rate": 6.952662721893491e-07, "loss": 0.0002630680799484253, "reward": 1.627314805984497, "reward_std": 0.21191666026910147, "rewards/FidelityReward/mean": 0.75544806321462, "rewards/FidelityReward/std": 0.19514132539431253, "rewards/JudgeFidelityReward/mean": 0.752848227818807, "rewards/JudgeFidelityReward/std": 0.20027470091978708, "rewards/SelfEvolvingFormatReward/mean": 0.9908854166666666, "rewards/SelfEvolvingFormatReward/std": 0.0916591336329778, "step": 235 }, { "clip_ratio/high_max": 0.00035574467619881034, "clip_ratio/high_mean": 0.00010321754962205887, "clip_ratio/low_mean": 5.995259707560763e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001631701466976665, "completions/clipped_ratio": 0.0, "completions/max_length": 171.5, "completions/mean_length": 72.548828125, "completions/min_length": 19.0, "epoch": 0.17751479289940827, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.005328218638896942, "learning_rate": 7.100591715976331e-07, "loss": 0.0002182585187256336, "reward": 1.5876181721687317, "reward_std": 0.2141198292374611, "rewards/FidelityReward/mean": 0.7200265228748322, "rewards/FidelityReward/std": 0.2103036716580391, "rewards/JudgeFidelityReward/mean": 0.7371362149715424, "rewards/JudgeFidelityReward/std": 0.19870862364768982, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 240 }, { "clip_ratio/high_max": 0.0003726798691786826, "clip_ratio/high_mean": 0.00015563723864033818, "clip_ratio/low_mean": 4.1390728438273075e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019702797289937734, "completions/clipped_ratio": 0.0, "completions/max_length": 181.0, "completions/mean_length": 75.80729166666667, "completions/min_length": 25.666666666666668, "epoch": 0.1812130177514793, "frac_reward_zero_std": 0.0, "grad_norm": 2.34375, "kl": 0.007615132350474596, "learning_rate": 7.248520710059172e-07, "loss": 0.000332780834287405, "reward": 1.5743279854456584, "reward_std": 0.23914594451586405, "rewards/FidelityReward/mean": 0.7155332763989767, "rewards/FidelityReward/std": 0.20616556207338968, "rewards/JudgeFidelityReward/mean": 0.7214958270390829, "rewards/JudgeFidelityReward/std": 0.21354534228642783, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.03594260662794113, "step": 245 }, { "clip_ratio/high_max": 0.00035379381151869895, "clip_ratio/high_mean": 0.00010572594474069774, "clip_ratio/low_mean": 7.088873680913821e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017661468300502747, "completions/clipped_ratio": 0.0, "completions/max_length": 176.5, "completions/mean_length": 76.83203125, "completions/min_length": 36.5, "epoch": 0.1849112426035503, "frac_reward_zero_std": 0.0, "grad_norm": 2.265625, "kl": 0.0030050556641072037, "learning_rate": 7.396449704142012e-07, "loss": 0.0002614946337416768, "reward": 1.6486032009124756, "reward_std": 0.21043846011161804, "rewards/FidelityReward/mean": 0.7681878507137299, "rewards/FidelityReward/std": 0.18269339948892593, "rewards/JudgeFidelityReward/mean": 0.7627837955951691, "rewards/JudgeFidelityReward/std": 0.18522503972053528, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 250 }, { "clip_ratio/high_max": 0.00038065746193751695, "clip_ratio/high_mean": 0.00010704348678700626, "clip_ratio/low_mean": 4.215351946186274e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014919700915925205, "completions/clipped_ratio": 0.0, "completions/max_length": 181.0, "completions/mean_length": 76.37630208333333, "completions/min_length": 33.666666666666664, "epoch": 0.1886094674556213, "frac_reward_zero_std": 0.0, "grad_norm": 2.375, "kl": 0.0032995222602039574, "learning_rate": 7.544378698224852e-07, "loss": -8.875536732375622e-05, "reward": 1.6245471636454265, "reward_std": 0.21516391138235727, "rewards/FidelityReward/mean": 0.7535518606503805, "rewards/FidelityReward/std": 0.1991711457570394, "rewards/JudgeFidelityReward/mean": 0.7471988201141357, "rewards/JudgeFidelityReward/std": 0.20976770917574564, "rewards/SelfEvolvingFormatReward/mean": 0.9947916666666666, "rewards/SelfEvolvingFormatReward/std": 0.0710716222723325, "step": 255 }, { "clip_ratio/high_max": 0.0006571945734322072, "clip_ratio/high_mean": 0.00019772133964579552, "clip_ratio/low_mean": 3.974336868850514e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023746470978949218, "completions/clipped_ratio": 0.0, "completions/max_length": 177.5, "completions/mean_length": 75.64453125, "completions/min_length": 26.5, "epoch": 0.19230769230769232, "frac_reward_zero_std": 0.0, "grad_norm": 2.5625, "kl": 0.0052186662796884775, "learning_rate": 7.692307692307693e-07, "loss": 0.00027465191669762135, "reward": 1.5849738717079163, "reward_std": 0.22446510940790176, "rewards/FidelityReward/mean": 0.7305529415607452, "rewards/FidelityReward/std": 0.2132212221622467, "rewards/JudgeFidelityReward/mean": 0.714701235294342, "rewards/JudgeFidelityReward/std": 0.2324729710817337, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 260 }, { "clip_ratio/high_max": 0.00036939941346645353, "clip_ratio/high_mean": 0.0001234920637216419, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001234920637216419, "completions/clipped_ratio": 0.0, "completions/max_length": 192.66666666666666, "completions/mean_length": 77.58333333333333, "completions/min_length": 36.666666666666664, "epoch": 0.19600591715976332, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.004935395764186978, "learning_rate": 7.840236686390532e-07, "loss": 0.00024987058714032173, "reward": 1.6102235317230225, "reward_std": 0.21853365004062653, "rewards/FidelityReward/mean": 0.7427294452985128, "rewards/FidelityReward/std": 0.1931116133928299, "rewards/JudgeFidelityReward/mean": 0.7362902959187826, "rewards/JudgeFidelityReward/std": 0.19835140804449716, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 265 }, { "clip_ratio/high_max": 0.0007901903125457465, "clip_ratio/high_mean": 0.00019910504925064744, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019910504925064744, "completions/clipped_ratio": 0.0, "completions/max_length": 168.0, "completions/mean_length": 73.2421875, "completions/min_length": 31.0, "epoch": 0.1997041420118343, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.0045546690933406355, "learning_rate": 7.988165680473372e-07, "loss": 0.0001952233724296093, "reward": 1.6400188207626343, "reward_std": 0.20542198419570923, "rewards/FidelityReward/mean": 0.755267322063446, "rewards/FidelityReward/std": 0.19624163955450058, "rewards/JudgeFidelityReward/mean": 0.7714560627937317, "rewards/JudgeFidelityReward/std": 0.19428520649671555, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 270 }, { "clip_ratio/high_max": 0.00036779921501874926, "clip_ratio/high_mean": 0.00012132222182117402, "clip_ratio/low_mean": 1.598465460119769e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013730687787756325, "completions/clipped_ratio": 0.0, "completions/max_length": 201.33333333333334, "completions/mean_length": 76.35416666666667, "completions/min_length": 32.333333333333336, "epoch": 0.20340236686390534, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.003683377569541335, "learning_rate": 8.136094674556213e-07, "loss": 0.00011148808989673853, "reward": 1.5771976709365845, "reward_std": 0.23086512088775635, "rewards/FidelityReward/mean": 0.7193671266237894, "rewards/FidelityReward/std": 0.20169283946355185, "rewards/JudgeFidelityReward/mean": 0.7195673187573751, "rewards/JudgeFidelityReward/std": 0.1983887900908788, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 275 }, { "clip_ratio/high_max": 0.0008402469102293253, "clip_ratio/high_mean": 0.0002467891317792237, "clip_ratio/low_mean": 1.8939393339678644e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026572853093966844, "completions/clipped_ratio": 0.0, "completions/max_length": 185.5, "completions/mean_length": 74.111328125, "completions/min_length": 30.5, "epoch": 0.20710059171597633, "frac_reward_zero_std": 0.0, "grad_norm": 2.9375, "kl": 0.0034834238700568675, "learning_rate": 8.284023668639053e-07, "loss": 4.328354261815548e-05, "reward": 1.567642867565155, "reward_std": 0.22853394597768784, "rewards/FidelityReward/mean": 0.7201949059963226, "rewards/FidelityReward/std": 0.20932739973068237, "rewards/JudgeFidelityReward/mean": 0.6988022029399872, "rewards/JudgeFidelityReward/std": 0.22166401892900467, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 280 }, { "clip_ratio/high_max": 0.0001687763724476099, "clip_ratio/high_mean": 5.965878954157233e-05, "clip_ratio/low_mean": 6.887802155688404e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001285368110984564, "completions/clipped_ratio": 0.0, "completions/max_length": 162.66666666666666, "completions/mean_length": 76.49739583333333, "completions/min_length": 25.666666666666668, "epoch": 0.21079881656804733, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.004954201076179743, "learning_rate": 8.431952662721893e-07, "loss": 0.00025182641111314296, "reward": 1.5898105303446453, "reward_std": 0.23093755543231964, "rewards/FidelityReward/mean": 0.7319896419843038, "rewards/FidelityReward/std": 0.21901371081670126, "rewards/JudgeFidelityReward/mean": 0.7208501696586609, "rewards/JudgeFidelityReward/std": 0.23517810304959616, "rewards/SelfEvolvingFormatReward/mean": 0.9947916666666666, "rewards/SelfEvolvingFormatReward/std": 0.05880991121133169, "step": 285 }, { "clip_ratio/high_max": 0.0008808000595308841, "clip_ratio/high_mean": 0.00025915529840858653, "clip_ratio/low_mean": 5.628522194456309e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003154405174427666, "completions/clipped_ratio": 0.0, "completions/max_length": 205.0, "completions/mean_length": 77.427734375, "completions/min_length": 36.0, "epoch": 0.21449704142011836, "frac_reward_zero_std": 0.0, "grad_norm": 2.359375, "kl": 0.004625330725684762, "learning_rate": 8.579881656804734e-07, "loss": 0.0002785273361951113, "reward": 1.5979488492012024, "reward_std": 0.23004650324583054, "rewards/FidelityReward/mean": 0.7291464805603027, "rewards/FidelityReward/std": 0.20566554367542267, "rewards/JudgeFidelityReward/mean": 0.7395579814910889, "rewards/JudgeFidelityReward/std": 0.20671414583921432, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 290 }, { "clip_ratio/high_max": 0.0003826834028586745, "clip_ratio/high_mean": 6.32484967354685e-05, "clip_ratio/low_mean": 1.87265919521451e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 8.197508868761361e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 198.33333333333334, "completions/mean_length": 77.06640625, "completions/min_length": 35.333333333333336, "epoch": 0.21819526627218935, "frac_reward_zero_std": 0.0, "grad_norm": 2.203125, "kl": 0.003534910036250949, "learning_rate": 8.727810650887574e-07, "loss": 7.067378610372543e-05, "reward": 1.5984150568644206, "reward_std": 0.2144506871700287, "rewards/FidelityReward/mean": 0.7316045761108398, "rewards/FidelityReward/std": 0.19501345853010812, "rewards/JudgeFidelityReward/mean": 0.7375273704528809, "rewards/JudgeFidelityReward/std": 0.20488702754179636, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 295 }, { "clip_ratio/high_max": 0.0009115420165471732, "clip_ratio/high_mean": 0.00017273457633564248, "clip_ratio/low_mean": 3.8667095941491426e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021140167082194238, "completions/clipped_ratio": 0.0, "completions/max_length": 215.5, "completions/mean_length": 76.9375, "completions/min_length": 32.0, "epoch": 0.22189349112426035, "frac_reward_zero_std": 0.0, "grad_norm": 2.296875, "kl": 0.004532565455883741, "learning_rate": 8.875739644970413e-07, "loss": 0.00019146681297570466, "reward": 1.5411484241485596, "reward_std": 0.24160338938236237, "rewards/FidelityReward/mean": 0.6968820095062256, "rewards/FidelityReward/std": 0.19798675179481506, "rewards/JudgeFidelityReward/mean": 0.6924392879009247, "rewards/JudgeFidelityReward/std": 0.21497420221567154, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 300 }, { "clip_ratio/high_max": 0.00034367303596809505, "clip_ratio/high_mean": 0.00010363155597588047, "clip_ratio/low_mean": 8.149046043399721e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018512201495468616, "completions/clipped_ratio": 0.0, "completions/max_length": 204.0, "completions/mean_length": 78.52604166666667, "completions/min_length": 29.333333333333332, "epoch": 0.22559171597633135, "frac_reward_zero_std": 0.0, "grad_norm": 2.3125, "kl": 0.004549254896119237, "learning_rate": 9.023668639053253e-07, "loss": 0.00020064811687916517, "reward": 1.5184295574824016, "reward_std": 0.25198249022165936, "rewards/FidelityReward/mean": 0.6866604884465536, "rewards/FidelityReward/std": 0.21220469971497855, "rewards/JudgeFidelityReward/mean": 0.6752569476763407, "rewards/JudgeFidelityReward/std": 0.2185013989607493, "rewards/SelfEvolvingFormatReward/mean": 0.98828125, "rewards/SelfEvolvingFormatReward/std": 0.10367502272129059, "step": 305 }, { "clip_ratio/high_max": 0.0005044631659984588, "clip_ratio/high_mean": 8.109541522571817e-05, "clip_ratio/low_mean": 5.898018862353638e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014007560239406302, "completions/clipped_ratio": 0.0, "completions/max_length": 181.0, "completions/mean_length": 76.81640625, "completions/min_length": 33.0, "epoch": 0.22928994082840237, "frac_reward_zero_std": 0.0, "grad_norm": 2.359375, "kl": 0.0035251381574198603, "learning_rate": 9.171597633136094e-07, "loss": 0.00010519572533667088, "reward": 1.5916662812232971, "reward_std": 0.23096944391727448, "rewards/FidelityReward/mean": 0.7313039004802704, "rewards/FidelityReward/std": 0.20641234517097473, "rewards/JudgeFidelityReward/mean": 0.7265841960906982, "rewards/JudgeFidelityReward/std": 0.2131044790148735, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 310 }, { "clip_ratio/high_max": 0.0003528075176291168, "clip_ratio/high_mean": 0.00010788599611259997, "clip_ratio/low_mean": 1.6035920998547228e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012392192147672176, "completions/clipped_ratio": 0.0, "completions/max_length": 206.66666666666666, "completions/mean_length": 76.44921875, "completions/min_length": 23.0, "epoch": 0.23298816568047337, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.004372293781489134, "learning_rate": 9.319526627218934e-07, "loss": 0.00046087619848549365, "reward": 1.580803672472636, "reward_std": 0.23323039710521698, "rewards/FidelityReward/mean": 0.7217737038930258, "rewards/FidelityReward/std": 0.20620229343573251, "rewards/JudgeFidelityReward/mean": 0.7258724172910055, "rewards/JudgeFidelityReward/std": 0.20296656092007956, "rewards/SelfEvolvingFormatReward/mean": 0.9921875, "rewards/SelfEvolvingFormatReward/std": 0.0830875113606453, "step": 315 }, { "clip_ratio/high_max": 0.0010882479720748961, "clip_ratio/high_mean": 0.0001964120179763995, "clip_ratio/low_mean": 4.3816943070851265e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024022895959205926, "completions/clipped_ratio": 0.0, "completions/max_length": 173.0, "completions/mean_length": 77.15234375, "completions/min_length": 31.0, "epoch": 0.23668639053254437, "frac_reward_zero_std": 0.0, "grad_norm": 2.5625, "kl": 0.003684898978099227, "learning_rate": 9.467455621301774e-07, "loss": 0.00019501359201967716, "reward": 1.5699050426483154, "reward_std": 0.22025686502456665, "rewards/FidelityReward/mean": 0.716345876455307, "rewards/FidelityReward/std": 0.1991971731185913, "rewards/JudgeFidelityReward/mean": 0.7168841361999512, "rewards/JudgeFidelityReward/std": 0.19825129956007004, "rewards/SelfEvolvingFormatReward/mean": 0.990234375, "rewards/SelfEvolvingFormatReward/std": 0.09802134335041046, "step": 320 }, { "clip_ratio/high_max": 0.00032047703862190244, "clip_ratio/high_mean": 5.7542147987987843e-05, "clip_ratio/low_mean": 4.032258002553135e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 9.786472655832767e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 227.0, "completions/mean_length": 78.23828125, "completions/min_length": 29.666666666666668, "epoch": 0.2403846153846154, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.0037231978494673967, "learning_rate": 9.615384615384615e-07, "loss": 0.00023798225447535515, "reward": 1.5873865286509197, "reward_std": 0.216087207198143, "rewards/FidelityReward/mean": 0.7240996360778809, "rewards/FidelityReward/std": 0.20478838185469309, "rewards/JudgeFidelityReward/mean": 0.7304800152778625, "rewards/JudgeFidelityReward/std": 0.19590601325035095, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 325 }, { "clip_ratio/high_max": 0.001214018021710217, "clip_ratio/high_mean": 0.00022576071787625552, "clip_ratio/low_mean": 5.777376500191167e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028353448724374177, "completions/clipped_ratio": 0.0, "completions/max_length": 197.5, "completions/mean_length": 77.5859375, "completions/min_length": 34.0, "epoch": 0.2440828402366864, "frac_reward_zero_std": 0.0, "grad_norm": 2.75, "kl": 0.003020267700776458, "learning_rate": 9.763313609467456e-07, "loss": 0.0001926091150380671, "reward": 1.583326280117035, "reward_std": 0.215611070394516, "rewards/FidelityReward/mean": 0.7208338975906372, "rewards/FidelityReward/std": 0.21721940487623215, "rewards/JudgeFidelityReward/mean": 0.7288911640644073, "rewards/JudgeFidelityReward/std": 0.21514398604631424, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 330 }, { "clip_ratio/high_max": 0.0006473908899351955, "clip_ratio/high_mean": 0.0001990116579690948, "clip_ratio/low_mean": 7.965314580360427e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002786648110486567, "completions/clipped_ratio": 0.0, "completions/max_length": 215.33333333333334, "completions/mean_length": 78.453125, "completions/min_length": 27.666666666666668, "epoch": 0.24778106508875739, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.003736040508374572, "learning_rate": 9.911242603550295e-07, "loss": 0.00016432085540145635, "reward": 1.6120816469192505, "reward_std": 0.20730063319206238, "rewards/FidelityReward/mean": 0.7432451049486796, "rewards/FidelityReward/std": 0.19482400019963583, "rewards/JudgeFidelityReward/mean": 0.7415793339411417, "rewards/JudgeFidelityReward/std": 0.2052590548992157, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 335 }, { "clip_ratio/high_max": 0.0005129983881488442, "clip_ratio/high_mean": 0.00014088574098423124, "clip_ratio/low_mean": 0.0001001914293738082, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024107717326842247, "completions/clipped_ratio": 0.0, "completions/max_length": 228.0, "completions/mean_length": 82.447265625, "completions/min_length": 25.0, "epoch": 0.2514792899408284, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.00354620311409235, "learning_rate": 9.99999760690958e-07, "loss": 0.0001332641812041402, "reward": 1.6142170429229736, "reward_std": 0.24417465925216675, "rewards/FidelityReward/mean": 0.741961807012558, "rewards/FidelityReward/std": 0.19093072414398193, "rewards/JudgeFidelityReward/mean": 0.7562292218208313, "rewards/JudgeFidelityReward/std": 0.19742268323898315, "rewards/SelfEvolvingFormatReward/mean": 0.98828125, "rewards/SelfEvolvingFormatReward/std": 0.1078278198838234, "step": 340 }, { "clip_ratio/high_max": 0.0004455945803783834, "clip_ratio/high_mean": 5.569932254729792e-05, "clip_ratio/low_mean": 4.4342828914523125e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010004215291701258, "completions/clipped_ratio": 0.0, "completions/max_length": 202.0, "completions/mean_length": 79.15755208333333, "completions/min_length": 35.333333333333336, "epoch": 0.2551775147928994, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.0027213430032134054, "learning_rate": 9.999970684668656e-07, "loss": 0.00021023128647357225, "reward": 1.577234427134196, "reward_std": 0.24118857085704803, "rewards/FidelityReward/mean": 0.7168240547180176, "rewards/FidelityReward/std": 0.20993509391943613, "rewards/JudgeFidelityReward/mean": 0.7273310621579488, "rewards/JudgeFidelityReward/std": 0.2056044340133667, "rewards/SelfEvolvingFormatReward/mean": 0.9934895833333334, "rewards/SelfEvolvingFormatReward/std": 0.0776092732946078, "step": 345 }, { "clip_ratio/high_max": 0.0005272467737086117, "clip_ratio/high_mean": 0.00012875744578195737, "clip_ratio/low_mean": 0.00010668208706192673, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002354395212023519, "completions/clipped_ratio": 0.0, "completions/max_length": 194.0, "completions/mean_length": 76.75, "completions/min_length": 33.0, "epoch": 0.2588757396449704, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.00488456767052412, "learning_rate": 9.999913848985394e-07, "loss": 0.00011941661359742284, "reward": 1.6247674226760864, "reward_std": 0.22057413309812546, "rewards/FidelityReward/mean": 0.7520976066589355, "rewards/FidelityReward/std": 0.18548112362623215, "rewards/JudgeFidelityReward/mean": 0.7492459416389465, "rewards/JudgeFidelityReward/std": 0.18464912474155426, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 350 }, { "clip_ratio/high_max": 0.0005974117666482925, "clip_ratio/high_mean": 0.00013330449583008884, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013330449583008884, "completions/clipped_ratio": 0.0, "completions/max_length": 222.0, "completions/mean_length": 75.63932291666667, "completions/min_length": 22.666666666666668, "epoch": 0.2625739644970414, "frac_reward_zero_std": 0.0, "grad_norm": 2.484375, "kl": 0.0033731736475601792, "learning_rate": 9.99982710019982e-07, "loss": 0.00037330035120248796, "reward": 1.6035184860229492, "reward_std": 0.2325028826793035, "rewards/FidelityReward/mean": 0.7399032711982727, "rewards/FidelityReward/std": 0.1973745326201121, "rewards/JudgeFidelityReward/mean": 0.7298344969749451, "rewards/JudgeFidelityReward/std": 0.20017537474632263, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 355 }, { "clip_ratio/high_max": 0.0004937802557833492, "clip_ratio/high_mean": 0.0001397775486111641, "clip_ratio/low_mean": 0.00010083648521685973, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024061403964878992, "completions/clipped_ratio": 0.0, "completions/max_length": 235.0, "completions/mean_length": 77.744140625, "completions/min_length": 26.0, "epoch": 0.26627218934911245, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.0044661047868430614, "learning_rate": 9.99971043883093e-07, "loss": -7.902197539806366e-05, "reward": 1.6930675506591797, "reward_std": 0.20079079270362854, "rewards/FidelityReward/mean": 0.795357346534729, "rewards/FidelityReward/std": 0.18605702370405197, "rewards/JudgeFidelityReward/mean": 0.7993266582489014, "rewards/JudgeFidelityReward/std": 0.19381187856197357, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 360 }, { "clip_ratio/high_max": 0.0003519911901094019, "clip_ratio/high_mean": 0.00012461625738069415, "clip_ratio/low_mean": 8.633239776827395e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021094864932820202, "completions/clipped_ratio": 0.0, "completions/max_length": 200.0, "completions/mean_length": 77.60677083333333, "completions/min_length": 31.0, "epoch": 0.26997041420118345, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.005320477206259966, "learning_rate": 9.99956386557668e-07, "loss": 0.00038306815549731253, "reward": 1.582750956217448, "reward_std": 0.2314602086941401, "rewards/FidelityReward/mean": 0.725629468758901, "rewards/FidelityReward/std": 0.20285364985466003, "rewards/JudgeFidelityReward/mean": 0.7168471018473307, "rewards/JudgeFidelityReward/std": 0.20994759599367777, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 365 }, { "clip_ratio/high_max": 0.000911187892779708, "clip_ratio/high_mean": 0.0002778266556560993, "clip_ratio/low_mean": 2.3169601627159862e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003009962616488338, "completions/clipped_ratio": 0.0, "completions/max_length": 282.5, "completions/mean_length": 76.20703125, "completions/min_length": 31.5, "epoch": 0.27366863905325445, "frac_reward_zero_std": 0.0, "grad_norm": 2.859375, "kl": 0.0029462456237524747, "learning_rate": 9.999387381313971e-07, "loss": 7.117819041013718e-05, "reward": 1.6438066363334656, "reward_std": 0.20488698035478592, "rewards/FidelityReward/mean": 0.7633656859397888, "rewards/FidelityReward/std": 0.18229630589485168, "rewards/JudgeFidelityReward/mean": 0.7628349363803864, "rewards/JudgeFidelityReward/std": 0.18319884687662125, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 370 }, { "clip_ratio/high_max": 0.0004982089158147573, "clip_ratio/high_mean": 0.0001208358065923676, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001208358065923676, "completions/clipped_ratio": 0.0, "completions/max_length": 201.0, "completions/mean_length": 77.07421875, "completions/min_length": 29.333333333333332, "epoch": 0.27736686390532544, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.004105811379849911, "learning_rate": 9.999180987098663e-07, "loss": 0.0002257857471704483, "reward": 1.6031728188196819, "reward_std": 0.2213431845108668, "rewards/FidelityReward/mean": 0.7387064099311829, "rewards/FidelityReward/std": 0.20124627649784088, "rewards/JudgeFidelityReward/mean": 0.7302348216374716, "rewards/JudgeFidelityReward/std": 0.20426332453886667, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 375 }, { "clip_ratio/high_max": 0.000626299250870943, "clip_ratio/high_mean": 0.000124843989033252, "clip_ratio/low_mean": 6.135942385299131e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018620341434143483, "completions/clipped_ratio": 0.0, "completions/max_length": 204.5, "completions/mean_length": 80.203125, "completions/min_length": 38.5, "epoch": 0.28106508875739644, "frac_reward_zero_std": 0.0, "grad_norm": 2.34375, "kl": 0.004501394834369421, "learning_rate": 9.998944684165556e-07, "loss": 0.00017643426544964314, "reward": 1.5426501035690308, "reward_std": 0.23334498703479767, "rewards/FidelityReward/mean": 0.6949377357959747, "rewards/FidelityReward/std": 0.19855372607707977, "rewards/JudgeFidelityReward/mean": 0.6993308663368225, "rewards/JudgeFidelityReward/std": 0.2019430249929428, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 380 }, { "clip_ratio/high_max": 0.00047195333754643796, "clip_ratio/high_mean": 0.0001149619580246508, "clip_ratio/low_mean": 3.756541264010593e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00015252737211994826, "completions/clipped_ratio": 0.0, "completions/max_length": 208.33333333333334, "completions/mean_length": 77.72526041666667, "completions/min_length": 32.333333333333336, "epoch": 0.28476331360946744, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.004484582459554076, "learning_rate": 9.998678473928383e-07, "loss": 0.0003209258895367384, "reward": 1.6009987592697144, "reward_std": 0.21750461558500925, "rewards/FidelityReward/mean": 0.7290871739387512, "rewards/FidelityReward/std": 0.19694989422957102, "rewards/JudgeFidelityReward/mean": 0.7451253135999044, "rewards/JudgeFidelityReward/std": 0.18517773846785227, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 385 }, { "clip_ratio/high_max": 0.0006112761097028851, "clip_ratio/high_mean": 0.00018150312243960797, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018150312243960797, "completions/clipped_ratio": 0.0, "completions/max_length": 175.0, "completions/mean_length": 74.451171875, "completions/min_length": 30.5, "epoch": 0.28846153846153844, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.004569146735593677, "learning_rate": 9.998382357979809e-07, "loss": 7.147099822759628e-05, "reward": 1.6022274494171143, "reward_std": 0.22351990640163422, "rewards/FidelityReward/mean": 0.7309188544750214, "rewards/FidelityReward/std": 0.19946135580539703, "rewards/JudgeFidelityReward/mean": 0.7426170110702515, "rewards/JudgeFidelityReward/std": 0.18437837064266205, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 390 }, { "clip_ratio/high_max": 0.0003514938522130251, "clip_ratio/high_mean": 6.368396570906043e-05, "clip_ratio/low_mean": 7.99064408056438e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014359040651470422, "completions/clipped_ratio": 0.0, "completions/max_length": 205.0, "completions/mean_length": 77.72265625, "completions/min_length": 34.666666666666664, "epoch": 0.2921597633136095, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.640625, "kl": 0.004140248382464051, "learning_rate": 9.998056338091413e-07, "loss": 0.00021510145161300897, "reward": 1.5972025394439697, "reward_std": 0.22321287790934244, "rewards/FidelityReward/mean": 0.7304409941037496, "rewards/FidelityReward/std": 0.19903259972731271, "rewards/JudgeFidelityReward/mean": 0.7361272970835367, "rewards/JudgeFidelityReward/std": 0.20653599997361502, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 395 }, { "clip_ratio/high_max": 0.0006810780381783843, "clip_ratio/high_mean": 0.0001641606620978564, "clip_ratio/low_mean": 7.657924725208431e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024073990643955768, "completions/clipped_ratio": 0.0, "completions/max_length": 207.5, "completions/mean_length": 75.59765625, "completions/min_length": 30.0, "epoch": 0.2958579881656805, "frac_reward_zero_std": 0.0, "grad_norm": 2.3125, "kl": 0.004196207970380783, "learning_rate": 9.997700416213681e-07, "loss": 0.0001715986989438534, "reward": 1.581924319267273, "reward_std": 0.2254430204629898, "rewards/FidelityReward/mean": 0.7174841463565826, "rewards/FidelityReward/std": 0.203252412378788, "rewards/JudgeFidelityReward/mean": 0.7308335602283478, "rewards/JudgeFidelityReward/std": 0.19013725966215134, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 400 }, { "clip_ratio/high_max": 0.0006303128437139093, "clip_ratio/high_mean": 0.0001790351074305363, "clip_ratio/low_mean": 9.159505279967561e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027063015149906275, "completions/clipped_ratio": 0.0, "completions/max_length": 187.66666666666666, "completions/mean_length": 75.93359375, "completions/min_length": 33.666666666666664, "epoch": 0.2995562130177515, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.005377569049596786, "learning_rate": 9.997314594475993e-07, "loss": 0.00010402500629425049, "reward": 1.581634243329366, "reward_std": 0.2278990795214971, "rewards/FidelityReward/mean": 0.7175136605898539, "rewards/FidelityReward/std": 0.20120724538962045, "rewards/JudgeFidelityReward/mean": 0.7282411257425944, "rewards/JudgeFidelityReward/std": 0.1932964821656545, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 405 }, { "clip_ratio/high_max": 0.0008697797311469913, "clip_ratio/high_mean": 0.00017792533617466688, "clip_ratio/low_mean": 4.409243119880557e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022201777319423853, "completions/clipped_ratio": 0.0, "completions/max_length": 180.5, "completions/mean_length": 74.9609375, "completions/min_length": 36.0, "epoch": 0.3032544378698225, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.003906853450462222, "learning_rate": 9.99689887518662e-07, "loss": 0.0001311882399022579, "reward": 1.5853316187858582, "reward_std": 0.23123599588871002, "rewards/FidelityReward/mean": 0.726062685251236, "rewards/FidelityReward/std": 0.1976008266210556, "rewards/JudgeFidelityReward/mean": 0.7224441170692444, "rewards/JudgeFidelityReward/std": 0.18513276427984238, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 410 }, { "clip_ratio/high_max": 0.0003787393099628389, "clip_ratio/high_mean": 0.00011168808850925416, "clip_ratio/low_mean": 2.0308692182879894e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013199677923694252, "completions/clipped_ratio": 0.0, "completions/max_length": 224.66666666666666, "completions/mean_length": 75.72916666666667, "completions/min_length": 33.333333333333336, "epoch": 0.3069526627218935, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.0046688028611242775, "learning_rate": 9.99645326083269e-07, "loss": 0.00022146929986774922, "reward": 1.6212351719538372, "reward_std": 0.21892011165618896, "rewards/FidelityReward/mean": 0.7474627296129862, "rewards/FidelityReward/std": 0.20418972273667654, "rewards/JudgeFidelityReward/mean": 0.7527531186739603, "rewards/JudgeFidelityReward/std": 0.19217219452063242, "rewards/SelfEvolvingFormatReward/mean": 0.9947916666666666, "rewards/SelfEvolvingFormatReward/std": 0.05677593996127447, "step": 415 }, { "clip_ratio/high_max": 0.0008498297887854278, "clip_ratio/high_mean": 0.0001650583028094843, "clip_ratio/low_mean": 6.389168556779623e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022894999128766357, "completions/clipped_ratio": 0.0, "completions/max_length": 190.5, "completions/mean_length": 76.591796875, "completions/min_length": 34.5, "epoch": 0.3106508875739645, "frac_reward_zero_std": 0.015625, "grad_norm": 2.34375, "kl": 0.004350222367793322, "learning_rate": 9.995977754080194e-07, "loss": 0.00014423702377825977, "reward": 1.6055688261985779, "reward_std": 0.2212718054652214, "rewards/FidelityReward/mean": 0.7391683757305145, "rewards/FidelityReward/std": 0.20909542590379715, "rewards/JudgeFidelityReward/mean": 0.7328009009361267, "rewards/JudgeFidelityReward/std": 0.21856264770030975, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 420 }, { "clip_ratio/high_max": 0.0006773358909413218, "clip_ratio/high_mean": 0.00020413552410900593, "clip_ratio/low_mean": 0.00010613170161377639, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003102672169916332, "completions/clipped_ratio": 0.0, "completions/max_length": 186.0, "completions/mean_length": 75.19401041666667, "completions/min_length": 31.333333333333332, "epoch": 0.3143491124260355, "frac_reward_zero_std": 0.0, "grad_norm": 2.796875, "kl": 0.005468063382431864, "learning_rate": 9.995472357773955e-07, "loss": 0.0002481590025126934, "reward": 1.5933322509129841, "reward_std": 0.22873592376708984, "rewards/FidelityReward/mean": 0.7252380649248759, "rewards/FidelityReward/std": 0.19997332990169525, "rewards/JudgeFidelityReward/mean": 0.7374903559684753, "rewards/JudgeFidelityReward/std": 0.19462711612383524, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 425 }, { "clip_ratio/high_max": 0.000872133206576109, "clip_ratio/high_mean": 0.0002770154387690127, "clip_ratio/low_mean": 3.8357905577868226e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003153733399813063, "completions/clipped_ratio": 0.0, "completions/max_length": 176.5, "completions/mean_length": 73.48046875, "completions/min_length": 33.0, "epoch": 0.3180473372781065, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.004269814677536487, "learning_rate": 9.99493707493762e-07, "loss": 0.00011958237737417222, "reward": 1.6222223043441772, "reward_std": 0.20639118552207947, "rewards/FidelityReward/mean": 0.7434941828250885, "rewards/FidelityReward/std": 0.1987086534500122, "rewards/JudgeFidelityReward/mean": 0.7574561238288879, "rewards/JudgeFidelityReward/std": 0.19029084593057632, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 430 }, { "clip_ratio/high_max": 0.0003277215990237892, "clip_ratio/high_mean": 6.101331673562527e-05, "clip_ratio/low_mean": 6.429580389522016e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012530912063084542, "completions/clipped_ratio": 0.0, "completions/max_length": 165.33333333333334, "completions/mean_length": 76.046875, "completions/min_length": 29.666666666666668, "epoch": 0.3217455621301775, "frac_reward_zero_std": 0.0, "grad_norm": 2.75, "kl": 0.006023452710360288, "learning_rate": 9.994371908773639e-07, "loss": 0.00038827843964099883, "reward": 1.5885121424992878, "reward_std": 0.22648744781812033, "rewards/FidelityReward/mean": 0.7311528325080872, "rewards/FidelityReward/std": 0.19783128798007965, "rewards/JudgeFidelityReward/mean": 0.7173227866490682, "rewards/JudgeFidelityReward/std": 0.20645310978094736, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 435 }, { "clip_ratio/high_max": 0.0007642689044587315, "clip_ratio/high_mean": 0.0001780736041837372, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001780736041837372, "completions/clipped_ratio": 0.0, "completions/max_length": 180.0, "completions/mean_length": 73.376953125, "completions/min_length": 31.0, "epoch": 0.3254437869822485, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.00406330144032836, "learning_rate": 9.993776862663242e-07, "loss": 0.00015960405580699443, "reward": 1.6123402118682861, "reward_std": 0.21209101378917694, "rewards/FidelityReward/mean": 0.7431134879589081, "rewards/FidelityReward/std": 0.19300508499145508, "rewards/JudgeFidelityReward/mean": 0.7423597276210785, "rewards/JudgeFidelityReward/std": 0.19973712414503098, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 440 }, { "clip_ratio/high_max": 0.0004062744788825512, "clip_ratio/high_mean": 0.0001342109520919621, "clip_ratio/low_mean": 2.4177950399462135e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00015838890103623272, "completions/clipped_ratio": 0.0, "completions/max_length": 219.33333333333334, "completions/mean_length": 74.30208333333333, "completions/min_length": 31.333333333333332, "epoch": 0.3291420118343195, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.005300164502114057, "learning_rate": 9.993151940166429e-07, "loss": 0.00022084438242018223, "reward": 1.6131865978240967, "reward_std": 0.21638798713684082, "rewards/FidelityReward/mean": 0.7439119815826416, "rewards/FidelityReward/std": 0.20113109548886618, "rewards/JudgeFidelityReward/mean": 0.7411533991495768, "rewards/JudgeFidelityReward/std": 0.20087923606236777, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 445 }, { "clip_ratio/high_max": 0.001144710718654096, "clip_ratio/high_mean": 0.0001883053017081693, "clip_ratio/low_mean": 4.03529207687825e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002286582166561857, "completions/clipped_ratio": 0.0, "completions/max_length": 178.5, "completions/mean_length": 78.796875, "completions/min_length": 35.5, "epoch": 0.3328402366863905, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.004200386861339211, "learning_rate": 9.992497145021938e-07, "loss": 7.386580109596253e-05, "reward": 1.5970929861068726, "reward_std": 0.2244039922952652, "rewards/FidelityReward/mean": 0.7286294102668762, "rewards/FidelityReward/std": 0.20199375599622726, "rewards/JudgeFidelityReward/mean": 0.7408334016799927, "rewards/JudgeFidelityReward/std": 0.21094031631946564, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 450 }, { "clip_ratio/high_max": 0.00028011205140501263, "clip_ratio/high_mean": 3.501400642562658e-05, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 3.501400642562658e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 192.33333333333334, "completions/mean_length": 73.8046875, "completions/min_length": 32.666666666666664, "epoch": 0.33653846153846156, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.004656050121411681, "learning_rate": 9.991812481147228e-07, "loss": 0.00029065581038594245, "reward": 1.578345497449239, "reward_std": 0.21465438107649484, "rewards/FidelityReward/mean": 0.717643161614736, "rewards/FidelityReward/std": 0.2055301566918691, "rewards/JudgeFidelityReward/mean": 0.7240088582038879, "rewards/JudgeFidelityReward/std": 0.19882598519325256, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 455 }, { "clip_ratio/high_max": 0.0003181420499458909, "clip_ratio/high_mean": 0.0001107875577872619, "clip_ratio/low_mean": 0.00011828854330815375, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002290760981850326, "completions/clipped_ratio": 0.0, "completions/max_length": 184.0, "completions/mean_length": 78.173828125, "completions/min_length": 38.0, "epoch": 0.34023668639053256, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.005306034721434117, "learning_rate": 9.99109795263845e-07, "loss": 0.00012599548790603876, "reward": 1.5230527520179749, "reward_std": 0.2419554740190506, "rewards/FidelityReward/mean": 0.6844824552536011, "rewards/FidelityReward/std": 0.2178642302751541, "rewards/JudgeFidelityReward/mean": 0.6810468137264252, "rewards/JudgeFidelityReward/std": 0.21835779398679733, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 460 }, { "clip_ratio/high_max": 0.0004811496823094785, "clip_ratio/high_mean": 0.00014755556476302444, "clip_ratio/low_mean": 4.914205346722156e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019669762114062905, "completions/clipped_ratio": 0.0, "completions/max_length": 156.66666666666666, "completions/mean_length": 72.77994791666667, "completions/min_length": 30.333333333333332, "epoch": 0.34393491124260356, "frac_reward_zero_std": 0.0, "grad_norm": 2.21875, "kl": 0.0047093458008021115, "learning_rate": 9.990353563770434e-07, "loss": 0.0001271224580705166, "reward": 1.5683763424555461, "reward_std": 0.2250409722328186, "rewards/FidelityReward/mean": 0.7105253338813782, "rewards/FidelityReward/std": 0.21082917849222818, "rewards/JudgeFidelityReward/mean": 0.7183062036832174, "rewards/JudgeFidelityReward/std": 0.2045253018538157, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 465 }, { "clip_ratio/high_max": 0.00045858341036364436, "clip_ratio/high_mean": 7.394526764983311e-05, "clip_ratio/low_mean": 6.147692329250276e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013542218512156977, "completions/clipped_ratio": 0.0, "completions/max_length": 208.5, "completions/mean_length": 77.50390625, "completions/min_length": 37.5, "epoch": 0.34763313609467456, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.004784909496083856, "learning_rate": 9.98957931899665e-07, "loss": 0.00038713496178388596, "reward": 1.4819836616516113, "reward_std": 0.253999225795269, "rewards/FidelityReward/mean": 0.6562760472297668, "rewards/FidelityReward/std": 0.21925248205661774, "rewards/JudgeFidelityReward/mean": 0.6592278182506561, "rewards/JudgeFidelityReward/std": 0.2272598296403885, "rewards/SelfEvolvingFormatReward/mean": 0.9921875, "rewards/SelfEvolvingFormatReward/std": 0.0851639099419117, "step": 470 }, { "clip_ratio/high_max": 0.0005226247361861169, "clip_ratio/high_mean": 0.0001289721447392367, "clip_ratio/low_mean": 4.2560126166790725e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017153227236121893, "completions/clipped_ratio": 0.0, "completions/max_length": 191.33333333333334, "completions/mean_length": 75.62239583333333, "completions/min_length": 26.0, "epoch": 0.35133136094674555, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.004271224699914455, "learning_rate": 9.988775222949196e-07, "loss": 0.00011203060857951641, "reward": 1.583256204922994, "reward_std": 0.2262964447339376, "rewards/FidelityReward/mean": 0.7232652306556702, "rewards/FidelityReward/std": 0.20263943572839102, "rewards/JudgeFidelityReward/mean": 0.7212840716044108, "rewards/JudgeFidelityReward/std": 0.21408543487389883, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 475 }, { "clip_ratio/high_max": 0.0008471281500533224, "clip_ratio/high_mean": 0.0002723617129959166, "clip_ratio/low_mean": 5.818713980261236e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003305488615296781, "completions/clipped_ratio": 0.0, "completions/max_length": 201.5, "completions/mean_length": 75.3203125, "completions/min_length": 36.5, "epoch": 0.35502958579881655, "frac_reward_zero_std": 0.0, "grad_norm": 2.28125, "kl": 0.004977623652666807, "learning_rate": 9.987941280438748e-07, "loss": 0.0002562349196523428, "reward": 1.5691951513290405, "reward_std": 0.23039065301418304, "rewards/FidelityReward/mean": 0.7099032402038574, "rewards/FidelityReward/std": 0.2055836021900177, "rewards/JudgeFidelityReward/mean": 0.722490131855011, "rewards/JudgeFidelityReward/std": 0.19126499444246292, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 480 }, { "clip_ratio/high_max": 0.001039610104635358, "clip_ratio/high_mean": 0.0001715959282591939, "clip_ratio/low_mean": 2.0374897576402873e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001919708272907883, "completions/clipped_ratio": 0.0, "completions/max_length": 164.0, "completions/mean_length": 75.17838541666667, "completions/min_length": 33.0, "epoch": 0.35872781065088755, "frac_reward_zero_std": 0.0, "grad_norm": 2.484375, "kl": 0.003892081044614315, "learning_rate": 9.987077496454562e-07, "loss": 0.00027133552357554435, "reward": 1.641505201657613, "reward_std": 0.204638143380483, "rewards/FidelityReward/mean": 0.7603733738263448, "rewards/FidelityReward/std": 0.19646557172139487, "rewards/JudgeFidelityReward/mean": 0.7635656793912252, "rewards/JudgeFidelityReward/std": 0.1927777330080668, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 485 }, { "clip_ratio/high_max": 0.000680908642243594, "clip_ratio/high_mean": 0.00021664412051904946, "clip_ratio/low_mean": 1.8939393339678644e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002355835196794942, "completions/clipped_ratio": 0.0, "completions/max_length": 186.5, "completions/mean_length": 76.03125, "completions/min_length": 37.0, "epoch": 0.3624260355029586, "frac_reward_zero_std": 0.0, "grad_norm": 2.265625, "kl": 0.004572956031188369, "learning_rate": 9.98618387616441e-07, "loss": 0.00032343477942049504, "reward": 1.668691635131836, "reward_std": 0.20836133509874344, "rewards/FidelityReward/mean": 0.7820321321487427, "rewards/FidelityReward/std": 0.19326980412006378, "rewards/JudgeFidelityReward/mean": 0.7791783809661865, "rewards/JudgeFidelityReward/std": 0.20493730902671814, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 490 }, { "clip_ratio/high_max": 0.0005332522559911013, "clip_ratio/high_mean": 0.00013199985842220484, "clip_ratio/low_mean": 8.437936776317656e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002163792261853814, "completions/clipped_ratio": 0.0, "completions/max_length": 187.66666666666666, "completions/mean_length": 75.58723958333333, "completions/min_length": 35.333333333333336, "epoch": 0.3661242603550296, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.004817822389304638, "learning_rate": 9.985260424914583e-07, "loss": 0.0003316928166896105, "reward": 1.5592254002888997, "reward_std": 0.2241375595331192, "rewards/FidelityReward/mean": 0.7039845188458761, "rewards/FidelityReward/std": 0.2228608379761378, "rewards/JudgeFidelityReward/mean": 0.713085949420929, "rewards/JudgeFidelityReward/std": 0.22079327205816904, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 495 }, { "clip_ratio/high_max": 0.0005570569192059338, "clip_ratio/high_mean": 0.00011066206207033247, "clip_ratio/low_mean": 2.1496131375897674e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013215819199103863, "completions/clipped_ratio": 0.0, "completions/max_length": 208.5, "completions/mean_length": 76.1875, "completions/min_length": 32.0, "epoch": 0.3698224852071006, "frac_reward_zero_std": 0.0, "grad_norm": 2.28125, "kl": 0.005218829121440649, "learning_rate": 9.984307148229832e-07, "loss": 0.00017269542440772055, "reward": 1.5336284637451172, "reward_std": 0.23310653120279312, "rewards/FidelityReward/mean": 0.685589611530304, "rewards/FidelityReward/std": 0.21273424476385117, "rewards/JudgeFidelityReward/mean": 0.699984073638916, "rewards/JudgeFidelityReward/std": 0.2045726627111435, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 500 }, { "clip_ratio/high_max": 0.000465747935231775, "clip_ratio/high_mean": 0.0001327635021880269, "clip_ratio/low_mean": 7.961133378557861e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002123748417943716, "completions/clipped_ratio": 0.0, "completions/max_length": 168.33333333333334, "completions/mean_length": 74.02864583333333, "completions/min_length": 36.333333333333336, "epoch": 0.3735207100591716, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.0045192661229521034, "learning_rate": 9.983324051813345e-07, "loss": 0.000243771867826581, "reward": 1.6061004400253296, "reward_std": 0.2160571018854777, "rewards/FidelityReward/mean": 0.7408646543820699, "rewards/FidelityReward/std": 0.20111518104871115, "rewards/JudgeFidelityReward/mean": 0.7330756584803263, "rewards/JudgeFidelityReward/std": 0.20301035046577454, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 505 }, { "clip_ratio/high_max": 0.0008224799763411284, "clip_ratio/high_mean": 0.00020876895869150757, "clip_ratio/low_mean": 8.207779610529542e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002908467664383352, "completions/clipped_ratio": 0.0, "completions/max_length": 173.0, "completions/mean_length": 76.244140625, "completions/min_length": 33.5, "epoch": 0.3772189349112426, "frac_reward_zero_std": 0.0, "grad_norm": 2.796875, "kl": 0.00575639046728611, "learning_rate": 9.98231114154672e-07, "loss": 0.00018107066862285137, "reward": 1.5697448253631592, "reward_std": 0.24203072488307953, "rewards/FidelityReward/mean": 0.7179860174655914, "rewards/FidelityReward/std": 0.20253349095582962, "rewards/JudgeFidelityReward/mean": 0.7054708003997803, "rewards/JudgeFidelityReward/std": 0.2085747942328453, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 510 }, { "clip_ratio/high_max": 0.00037308370228856803, "clip_ratio/high_mean": 0.00016617493238300086, "clip_ratio/low_mean": 2.5150904548354447e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019132583984173834, "completions/clipped_ratio": 0.0, "completions/max_length": 203.33333333333334, "completions/mean_length": 75.19140625, "completions/min_length": 35.666666666666664, "epoch": 0.3809171597633136, "frac_reward_zero_std": 0.0, "grad_norm": 2.8125, "kl": 0.0061954913660883905, "learning_rate": 9.981268423489917e-07, "loss": 0.0002864347770810127, "reward": 1.5811909834543865, "reward_std": 0.22414042055606842, "rewards/FidelityReward/mean": 0.7183596690495809, "rewards/FidelityReward/std": 0.18975561360518137, "rewards/JudgeFidelityReward/mean": 0.7269646922747294, "rewards/JudgeFidelityReward/std": 0.17925227185090384, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 515 }, { "clip_ratio/high_max": 0.0007530835107900202, "clip_ratio/high_mean": 0.00021473063388839365, "clip_ratio/low_mean": 2.4154588754754513e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023888522409833967, "completions/clipped_ratio": 0.0, "completions/max_length": 230.5, "completions/mean_length": 73.36328125, "completions/min_length": 33.0, "epoch": 0.38461538461538464, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.005141549417749048, "learning_rate": 9.98019590388123e-07, "loss": 0.00044423486106097697, "reward": 1.5036233067512512, "reward_std": 0.23948071151971817, "rewards/FidelityReward/mean": 0.6704820990562439, "rewards/FidelityReward/std": 0.20140435546636581, "rewards/JudgeFidelityReward/mean": 0.6701886355876923, "rewards/JudgeFidelityReward/std": 0.19456730037927628, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 520 }, { "clip_ratio/high_max": 0.0005105372052639723, "clip_ratio/high_mean": 0.00012596883752848952, "clip_ratio/low_mean": 7.143101101974026e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001973998558241874, "completions/clipped_ratio": 0.0, "completions/max_length": 171.33333333333334, "completions/mean_length": 71.70052083333333, "completions/min_length": 32.666666666666664, "epoch": 0.38831360946745563, "frac_reward_zero_std": 0.0, "grad_norm": 2.484375, "kl": 0.005553788598626852, "learning_rate": 9.979093589137248e-07, "loss": 9.370541665703059e-05, "reward": 1.629715124766032, "reward_std": 0.21686954299608865, "rewards/FidelityReward/mean": 0.7490944862365723, "rewards/FidelityReward/std": 0.1808528850475947, "rewards/JudgeFidelityReward/mean": 0.7651473681131998, "rewards/JudgeFidelityReward/std": 0.1745821883281072, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 525 }, { "clip_ratio/high_max": 0.0007535572163760662, "clip_ratio/high_mean": 0.00021310555748641491, "clip_ratio/low_mean": 4.2295177991036326e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002554007398430258, "completions/clipped_ratio": 0.0, "completions/max_length": 157.5, "completions/mean_length": 72.267578125, "completions/min_length": 29.0, "epoch": 0.39201183431952663, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.0052893230691552166, "learning_rate": 9.977961485852816e-07, "loss": 0.000321463355794549, "reward": 1.6338945627212524, "reward_std": 0.18856710195541382, "rewards/FidelityReward/mean": 0.7574770748615265, "rewards/FidelityReward/std": 0.21237852424383163, "rewards/JudgeFidelityReward/mean": 0.7528350353240967, "rewards/JudgeFidelityReward/std": 0.19591570645570755, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 530 }, { "clip_ratio/high_max": 0.0006605676375329494, "clip_ratio/high_mean": 0.0001319200440775603, "clip_ratio/low_mean": 7.425837684422731e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002061784267425537, "completions/clipped_ratio": 0.0, "completions/max_length": 193.33333333333334, "completions/mean_length": 72.77864583333333, "completions/min_length": 32.333333333333336, "epoch": 0.39571005917159763, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.005590162891894579, "learning_rate": 9.976799600800992e-07, "loss": 0.0003315164241939783, "reward": 1.6124600172042847, "reward_std": 0.2257696489493052, "rewards/FidelityReward/mean": 0.7456648945808411, "rewards/FidelityReward/std": 0.19744583467642465, "rewards/JudgeFidelityReward/mean": 0.7414026459058126, "rewards/JudgeFidelityReward/std": 0.1980898082256317, "rewards/SelfEvolvingFormatReward/mean": 0.9921875, "rewards/SelfEvolvingFormatReward/std": 0.08618089556694031, "step": 535 }, { "clip_ratio/high_max": 0.0006767184007912874, "clip_ratio/high_mean": 0.00012066112976754085, "clip_ratio/low_mean": 6.545699725393205e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018611812556628137, "completions/clipped_ratio": 0.0, "completions/max_length": 168.0, "completions/mean_length": 74.95703125, "completions/min_length": 34.0, "epoch": 0.3994082840236686, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.0047003084793686865, "learning_rate": 9.975607940933016e-07, "loss": 9.61716752499342e-05, "reward": 1.5824631452560425, "reward_std": 0.22444194555282593, "rewards/FidelityReward/mean": 0.7240784764289856, "rewards/FidelityReward/std": 0.20574837923049927, "rewards/JudgeFidelityReward/mean": 0.7167693078517914, "rewards/JudgeFidelityReward/std": 0.20163577795028687, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 540 }, { "clip_ratio/high_max": 0.0007720194524154067, "clip_ratio/high_mean": 0.00020059113157913088, "clip_ratio/low_mean": 2.2461815387941896e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022305294405668973, "completions/clipped_ratio": 0.0, "completions/max_length": 166.0, "completions/mean_length": 71.09375, "completions/min_length": 29.666666666666668, "epoch": 0.4031065088757396, "frac_reward_zero_std": 0.0, "grad_norm": 2.375, "kl": 0.005164606403559447, "learning_rate": 9.974386513378256e-07, "loss": 4.9674510955810546e-05, "reward": 1.5805105765660603, "reward_std": 0.21991160015265146, "rewards/FidelityReward/mean": 0.7201693852742513, "rewards/FidelityReward/std": 0.21266967554887137, "rewards/JudgeFidelityReward/mean": 0.7245886325836182, "rewards/JudgeFidelityReward/std": 0.20358851055304208, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 545 }, { "clip_ratio/high_max": 0.0001912045874632895, "clip_ratio/high_mean": 0.0001038379967212677, "clip_ratio/low_mean": 1.979414082597941e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012363214045763016, "completions/clipped_ratio": 0.0, "completions/max_length": 223.0, "completions/mean_length": 75.919921875, "completions/min_length": 32.5, "epoch": 0.4068047337278107, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.004803359322249889, "learning_rate": 9.973135325444183e-07, "loss": 0.0003643834963440895, "reward": 1.5294098854064941, "reward_std": 0.23350167274475098, "rewards/FidelityReward/mean": 0.6834800541400909, "rewards/FidelityReward/std": 0.21197163313627243, "rewards/JudgeFidelityReward/mean": 0.6918596029281616, "rewards/JudgeFidelityReward/std": 0.20685400813817978, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 550 }, { "clip_ratio/high_max": 0.0003391536301933229, "clip_ratio/high_mean": 6.319287203950807e-05, "clip_ratio/low_mean": 4.3756357626989484e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010694922821130604, "completions/clipped_ratio": 0.0, "completions/max_length": 175.0, "completions/mean_length": 72.79166666666667, "completions/min_length": 31.666666666666668, "epoch": 0.41050295857988167, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.004528697952628136, "learning_rate": 9.971854384616299e-07, "loss": 0.0002466967329382896, "reward": 1.577980677286784, "reward_std": 0.22209947804609934, "rewards/FidelityReward/mean": 0.7198610305786133, "rewards/FidelityReward/std": 0.20966078837712607, "rewards/JudgeFidelityReward/mean": 0.7175413370132446, "rewards/JudgeFidelityReward/std": 0.2111555834611257, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 555 }, { "clip_ratio/high_max": 0.0006486186874099075, "clip_ratio/high_mean": 0.00018314977060072125, "clip_ratio/low_mean": 6.241696974029765e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002455667476169765, "completions/clipped_ratio": 0.0, "completions/max_length": 166.5, "completions/mean_length": 71.6796875, "completions/min_length": 35.5, "epoch": 0.41420118343195267, "frac_reward_zero_std": 0.0, "grad_norm": 2.25, "kl": 0.005289267469197512, "learning_rate": 9.970543698558126e-07, "loss": 0.00020542324054986238, "reward": 1.5367602705955505, "reward_std": 0.23760277777910233, "rewards/FidelityReward/mean": 0.6883118152618408, "rewards/FidelityReward/std": 0.20216688513755798, "rewards/JudgeFidelityReward/mean": 0.6968969404697418, "rewards/JudgeFidelityReward/std": 0.18837003409862518, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 560 }, { "clip_ratio/high_max": 0.00036377536598593, "clip_ratio/high_mean": 4.547192074824125e-05, "clip_ratio/low_mean": 4.274911480024457e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 8.822103554848582e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 158.33333333333334, "completions/mean_length": 72.9140625, "completions/min_length": 33.666666666666664, "epoch": 0.41789940828402367, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.005567907355725765, "learning_rate": 9.969203275111133e-07, "loss": 0.0003481769934296608, "reward": 1.586362640062968, "reward_std": 0.22411390642325082, "rewards/FidelityReward/mean": 0.7236157655715942, "rewards/FidelityReward/std": 0.19865581393241882, "rewards/JudgeFidelityReward/mean": 0.7294000585873922, "rewards/JudgeFidelityReward/std": 0.18821778893470764, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 565 }, { "clip_ratio/high_max": 0.0009140783920884132, "clip_ratio/high_mean": 0.0002833380654919893, "clip_ratio/low_mean": 8.825189579511061e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00037158996565267444, "completions/clipped_ratio": 0.0, "completions/max_length": 191.5, "completions/mean_length": 73.15625, "completions/min_length": 30.5, "epoch": 0.42159763313609466, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.0060081670992076395, "learning_rate": 9.967833122294703e-07, "loss": 0.00012509981170296668, "reward": 1.6168675422668457, "reward_std": 0.22259224951267242, "rewards/FidelityReward/mean": 0.742530882358551, "rewards/FidelityReward/std": 0.18932005017995834, "rewards/JudgeFidelityReward/mean": 0.7486733794212341, "rewards/JudgeFidelityReward/std": 0.18418226391077042, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 570 }, { "clip_ratio/high_max": 0.0007861514808610081, "clip_ratio/high_mean": 0.0001879980438388884, "clip_ratio/low_mean": 2.5227043079212307e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021322509273886682, "completions/clipped_ratio": 0.0, "completions/max_length": 191.33333333333334, "completions/mean_length": 72.23177083333333, "completions/min_length": 30.0, "epoch": 0.42529585798816566, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.53125, "kl": 0.006078614853322506, "learning_rate": 9.966433248306083e-07, "loss": 0.00024145201314240694, "reward": 1.5922782023747761, "reward_std": 0.21716703474521637, "rewards/FidelityReward/mean": 0.7252779603004456, "rewards/FidelityReward/std": 0.19530571500460306, "rewards/JudgeFidelityReward/mean": 0.7353025277455648, "rewards/JudgeFidelityReward/std": 0.18946941693623862, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 575 }, { "clip_ratio/high_max": 0.00019665684085339307, "clip_ratio/high_mean": 9.066893835552037e-05, "clip_ratio/low_mean": 0.0001085560506908223, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019922498322557657, "completions/clipped_ratio": 0.0, "completions/max_length": 199.0, "completions/mean_length": 71.310546875, "completions/min_length": 34.0, "epoch": 0.4289940828402367, "frac_reward_zero_std": 0.015625, "grad_norm": 2.546875, "kl": 0.005903937527909875, "learning_rate": 9.96500366152033e-07, "loss": 0.0003006262704730034, "reward": 1.6089168190956116, "reward_std": 0.21208610385656357, "rewards/FidelityReward/mean": 0.7444110214710236, "rewards/FidelityReward/std": 0.20026320964097977, "rewards/JudgeFidelityReward/mean": 0.732917845249176, "rewards/JudgeFidelityReward/std": 0.21775329113006592, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 580 }, { "clip_ratio/high_max": 0.0006746592000126839, "clip_ratio/high_mean": 0.00021694125607609748, "clip_ratio/low_mean": 8.94494034582749e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030639065662398937, "completions/clipped_ratio": 0.0, "completions/max_length": 161.66666666666666, "completions/mean_length": 71.5625, "completions/min_length": 34.333333333333336, "epoch": 0.4326923076923077, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.0066363726742565635, "learning_rate": 9.963544370490268e-07, "loss": 0.0002678203396499157, "reward": 1.6062829891840618, "reward_std": 0.21660259862740835, "rewards/FidelityReward/mean": 0.7349510788917542, "rewards/FidelityReward/std": 0.20004485547542572, "rewards/JudgeFidelityReward/mean": 0.7426639000574747, "rewards/JudgeFidelityReward/std": 0.19660129149754843, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 585 }, { "clip_ratio/high_max": 0.000564583926461637, "clip_ratio/high_mean": 0.00013318023411557078, "clip_ratio/low_mean": 7.076497568050399e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020394521416164935, "completions/clipped_ratio": 0.0, "completions/max_length": 183.5, "completions/mean_length": 71.615234375, "completions/min_length": 29.0, "epoch": 0.4363905325443787, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.007092813029885292, "learning_rate": 9.962055383946433e-07, "loss": 0.0005144505761563778, "reward": 1.5629644393920898, "reward_std": 0.22570040822029114, "rewards/FidelityReward/mean": 0.701526403427124, "rewards/FidelityReward/std": 0.22329634428024292, "rewards/JudgeFidelityReward/mean": 0.7287355065345764, "rewards/JudgeFidelityReward/std": 0.20491614192724228, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 590 }, { "clip_ratio/high_max": 0.0006850459380075335, "clip_ratio/high_mean": 0.0001287493825657293, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001287493825657293, "completions/clipped_ratio": 0.0, "completions/max_length": 175.0, "completions/mean_length": 72.20572916666667, "completions/min_length": 31.0, "epoch": 0.4400887573964497, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.0060988740995526316, "learning_rate": 9.960536710797018e-07, "loss": 0.00021004402078688144, "reward": 1.6301847298940022, "reward_std": 0.19023009637991586, "rewards/FidelityReward/mean": 0.7461740175882975, "rewards/FidelityReward/std": 0.19873250524202982, "rewards/JudgeFidelityReward/mean": 0.7680215040842692, "rewards/JudgeFidelityReward/std": 0.19027398029963175, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 595 }, { "clip_ratio/high_max": 0.0007448931224644184, "clip_ratio/high_mean": 0.00022014934802427888, "clip_ratio/low_mean": 3.8484857941512016e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00025863420451059936, "completions/clipped_ratio": 0.0, "completions/max_length": 199.0, "completions/mean_length": 71.609375, "completions/min_length": 32.0, "epoch": 0.4437869822485207, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.006103794323280454, "learning_rate": 9.958988360127824e-07, "loss": 0.0004165515303611755, "reward": 1.6216363906860352, "reward_std": 0.20299305021762848, "rewards/FidelityReward/mean": 0.7473165094852448, "rewards/FidelityReward/std": 0.19439513981342316, "rewards/JudgeFidelityReward/mean": 0.7486396729946136, "rewards/JudgeFidelityReward/std": 0.18912307173013687, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 600 }, { "clip_ratio/high_max": 0.0004447157960385084, "clip_ratio/high_mean": 9.297294309362769e-05, "clip_ratio/low_mean": 5.7775752793531865e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00015074869152158498, "completions/clipped_ratio": 0.0, "completions/max_length": 168.0, "completions/mean_length": 70.8671875, "completions/min_length": 30.666666666666668, "epoch": 0.4474852071005917, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.005839976854622364, "learning_rate": 9.957410341202207e-07, "loss": 0.0005338504910469055, "reward": 1.6190582116444905, "reward_std": 0.2013908475637436, "rewards/FidelityReward/mean": 0.7397825916608175, "rewards/FidelityReward/std": 0.20178098479906717, "rewards/JudgeFidelityReward/mean": 0.7585511803627014, "rewards/JudgeFidelityReward/std": 0.18173917631308237, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 605 }, { "clip_ratio/high_max": 0.0008941961801610887, "clip_ratio/high_mean": 0.00023507742589572443, "clip_ratio/low_mean": 0.00010604043200146407, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003411178535316139, "completions/clipped_ratio": 0.0, "completions/max_length": 153.0, "completions/mean_length": 71.333984375, "completions/min_length": 34.0, "epoch": 0.4511834319526627, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.006315773632377386, "learning_rate": 9.955802663461017e-07, "loss": 0.00027476907707750796, "reward": 1.5961090326309204, "reward_std": 0.20259761065244675, "rewards/FidelityReward/mean": 0.7303123772144318, "rewards/FidelityReward/std": 0.2053472250699997, "rewards/JudgeFidelityReward/mean": 0.7335463464260101, "rewards/JudgeFidelityReward/std": 0.1996864676475525, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 610 }, { "clip_ratio/high_max": 0.0007454977603629231, "clip_ratio/high_mean": 0.00018019478884525597, "clip_ratio/low_mean": 1.965408737305552e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019984887330792844, "completions/clipped_ratio": 0.0, "completions/max_length": 173.0, "completions/mean_length": 71.67578125, "completions/min_length": 35.333333333333336, "epoch": 0.45488165680473375, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.006054441630840302, "learning_rate": 9.954165336522545e-07, "loss": -5.8035273104906084e-05, "reward": 1.657427708307902, "reward_std": 0.21120497087637582, "rewards/FidelityReward/mean": 0.7716617385546366, "rewards/FidelityReward/std": 0.1905283977588018, "rewards/JudgeFidelityReward/mean": 0.7728339036305746, "rewards/JudgeFidelityReward/std": 0.19643322626749674, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 615 }, { "clip_ratio/high_max": 0.0005344014265574515, "clip_ratio/high_mean": 0.00013142796233296395, "clip_ratio/low_mean": 2.3854961909819394e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001552829227875918, "completions/clipped_ratio": 0.0, "completions/max_length": 188.5, "completions/mean_length": 72.7421875, "completions/min_length": 27.5, "epoch": 0.45857988165680474, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.0050896800123155115, "learning_rate": 9.952498370182464e-07, "loss": 0.00019999509677290917, "reward": 1.5939226150512695, "reward_std": 0.21549776941537857, "rewards/FidelityReward/mean": 0.7388612627983093, "rewards/FidelityReward/std": 0.19281917810440063, "rewards/JudgeFidelityReward/mean": 0.71207594871521, "rewards/JudgeFidelityReward/std": 0.21202928572893143, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 620 }, { "clip_ratio/high_max": 0.0003987803007476032, "clip_ratio/high_mean": 0.00014210809604264796, "clip_ratio/low_mean": 6.18588921497576e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000203966983826831, "completions/clipped_ratio": 0.0, "completions/max_length": 157.33333333333334, "completions/mean_length": 69.64453125, "completions/min_length": 31.666666666666668, "epoch": 0.46227810650887574, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.671875, "kl": 0.0063612829893827435, "learning_rate": 9.950801774413776e-07, "loss": 0.000221869139932096, "reward": 1.6842715342839558, "reward_std": 0.1970638632774353, "rewards/FidelityReward/mean": 0.7887694835662842, "rewards/FidelityReward/std": 0.1828629473845164, "rewards/JudgeFidelityReward/mean": 0.792306125164032, "rewards/JudgeFidelityReward/std": 0.1726495325565338, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 625 }, { "clip_ratio/high_max": 0.00141260230448097, "clip_ratio/high_mean": 0.0003376164997462183, "clip_ratio/low_mean": 4.6278483932837847e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003838949778582901, "completions/clipped_ratio": 0.001953125, "completions/max_length": 341.0, "completions/mean_length": 72.26953125, "completions/min_length": 37.0, "epoch": 0.46597633136094674, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.005668316781520843, "learning_rate": 9.949075559366744e-07, "loss": 0.0001330804545432329, "reward": 1.6394466757774353, "reward_std": 0.2050541788339615, "rewards/FidelityReward/mean": 0.7590413093566895, "rewards/FidelityReward/std": 0.2013796716928482, "rewards/JudgeFidelityReward/mean": 0.7627637386322021, "rewards/JudgeFidelityReward/std": 0.20320820063352585, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 630 }, { "clip_ratio/high_max": 0.0005785400746390224, "clip_ratio/high_mean": 9.644878446124494e-05, "clip_ratio/low_mean": 4.282322188373655e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013927199761383237, "completions/clipped_ratio": 0.0, "completions/max_length": 157.66666666666666, "completions/mean_length": 70.80208333333333, "completions/min_length": 27.0, "epoch": 0.46967455621301774, "frac_reward_zero_std": 0.0, "grad_norm": 3.03125, "kl": 0.005784693174064159, "learning_rate": 9.947319735368836e-07, "loss": 0.0003516314784064889, "reward": 1.5617891550064087, "reward_std": 0.23628100752830505, "rewards/FidelityReward/mean": 0.7075368960698446, "rewards/FidelityReward/std": 0.21146457890669504, "rewards/JudgeFidelityReward/mean": 0.7124108076095581, "rewards/JudgeFidelityReward/std": 0.1915887395540873, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.03594260662794113, "step": 635 }, { "clip_ratio/high_max": 0.0007243446190841496, "clip_ratio/high_mean": 0.00015349368477473034, "clip_ratio/low_mean": 8.967644535005093e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024317013012478128, "completions/clipped_ratio": 0.0, "completions/max_length": 232.0, "completions/mean_length": 71.533203125, "completions/min_length": 27.5, "epoch": 0.47337278106508873, "frac_reward_zero_std": 0.015625, "grad_norm": 2.53125, "kl": 0.007137319259345532, "learning_rate": 9.945534312924658e-07, "loss": 0.0005881566554307938, "reward": 1.6088988184928894, "reward_std": 0.21670196950435638, "rewards/FidelityReward/mean": 0.7449297606945038, "rewards/FidelityReward/std": 0.20173010230064392, "rewards/JudgeFidelityReward/mean": 0.7357507944107056, "rewards/JudgeFidelityReward/std": 0.2070879340171814, "rewards/SelfEvolvingFormatReward/mean": 0.9921875, "rewards/SelfEvolvingFormatReward/std": 0.08821486681699753, "step": 640 }, { "clip_ratio/high_max": 0.00028510335832834246, "clip_ratio/high_mean": 0.00010207785526290536, "clip_ratio/low_mean": 4.2850972386077045e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001449288334697485, "completions/clipped_ratio": 0.0, "completions/max_length": 174.0, "completions/mean_length": 72.90104166666667, "completions/min_length": 34.333333333333336, "epoch": 0.4770710059171598, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.006571745686233044, "learning_rate": 9.9437193027159e-07, "loss": 0.00024140598252415658, "reward": 1.6012916564941406, "reward_std": 0.21839267015457153, "rewards/FidelityReward/mean": 0.7321371038754781, "rewards/FidelityReward/std": 0.204448993007342, "rewards/JudgeFidelityReward/mean": 0.7383091052373251, "rewards/JudgeFidelityReward/std": 0.19310869773228964, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 645 }, { "clip_ratio/high_max": 0.000695967604406178, "clip_ratio/high_mean": 0.00025480559415882455, "clip_ratio/low_mean": 4.197137604933232e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002967769672977738, "completions/clipped_ratio": 0.0, "completions/max_length": 171.5, "completions/mean_length": 72.005859375, "completions/min_length": 30.5, "epoch": 0.4807692307692308, "frac_reward_zero_std": 0.0, "grad_norm": 2.765625, "kl": 0.0060329210013151165, "learning_rate": 9.94187471560127e-07, "loss": 0.0001928041223436594, "reward": 1.5998767018318176, "reward_std": 0.21486618369817734, "rewards/FidelityReward/mean": 0.7300117015838623, "rewards/FidelityReward/std": 0.18952462822198868, "rewards/JudgeFidelityReward/mean": 0.7397299706935883, "rewards/JudgeFidelityReward/std": 0.1907038688659668, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 650 }, { "clip_ratio/high_max": 0.00043910027015954257, "clip_ratio/high_mean": 0.00014433494652621448, "clip_ratio/low_mean": 4.415279836393893e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001884877448901534, "completions/clipped_ratio": 0.0, "completions/max_length": 180.0, "completions/mean_length": 71.78125, "completions/min_length": 32.333333333333336, "epoch": 0.4844674556213018, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.005963777750730514, "learning_rate": 9.940000562616418e-07, "loss": 0.00022868723608553408, "reward": 1.5500360329945881, "reward_std": 0.2312257041533788, "rewards/FidelityReward/mean": 0.6943297584851583, "rewards/FidelityReward/std": 0.20470226307710013, "rewards/JudgeFidelityReward/mean": 0.7140166759490967, "rewards/JudgeFidelityReward/std": 0.18882262706756592, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 655 }, { "clip_ratio/high_max": 0.0009046754566952586, "clip_ratio/high_mean": 0.00017238213040400298, "clip_ratio/low_mean": 4.031553835375235e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002126976673025638, "completions/clipped_ratio": 0.0, "completions/max_length": 179.5, "completions/mean_length": 74.931640625, "completions/min_length": 30.0, "epoch": 0.4881656804733728, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.005408292589709163, "learning_rate": 9.938096854973885e-07, "loss": 0.00020933817140758037, "reward": 1.58612060546875, "reward_std": 0.23440037667751312, "rewards/FidelityReward/mean": 0.7237842679023743, "rewards/FidelityReward/std": 0.19978391379117966, "rewards/JudgeFidelityReward/mean": 0.7266257703304291, "rewards/JudgeFidelityReward/std": 0.1985579878091812, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 660 }, { "clip_ratio/high_max": 0.0005538681172765792, "clip_ratio/high_mean": 8.507635357091203e-05, "clip_ratio/low_mean": 6.396024255082012e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014903660194249825, "completions/clipped_ratio": 0.0, "completions/max_length": 195.0, "completions/mean_length": 72.31380208333333, "completions/min_length": 27.333333333333332, "epoch": 0.4918639053254438, "frac_reward_zero_std": 0.0, "grad_norm": 2.734375, "kl": 0.00515822391025722, "learning_rate": 9.936163604063028e-07, "loss": 9.291311725974083e-05, "reward": 1.5997708241144817, "reward_std": 0.21310694018999735, "rewards/FidelityReward/mean": 0.7324663003285726, "rewards/FidelityReward/std": 0.19931689401467642, "rewards/JudgeFidelityReward/mean": 0.7372132539749146, "rewards/JudgeFidelityReward/std": 0.19269476334253946, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 665 }, { "clip_ratio/high_max": 0.0004623586079105735, "clip_ratio/high_mean": 0.00014152160729281605, "clip_ratio/low_mean": 6.2731777143199e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000204253381525632, "completions/clipped_ratio": 0.0, "completions/max_length": 170.5, "completions/mean_length": 74.466796875, "completions/min_length": 27.5, "epoch": 0.49556213017751477, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.00643755765631795, "learning_rate": 9.934200821449954e-07, "loss": 0.0004699853248894215, "reward": 1.6217013597488403, "reward_std": 0.2230663001537323, "rewards/FidelityReward/mean": 0.7523767948150635, "rewards/FidelityReward/std": 0.2089390531182289, "rewards/JudgeFidelityReward/mean": 0.7425554394721985, "rewards/JudgeFidelityReward/std": 0.20835676789283752, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 670 }, { "clip_ratio/high_max": 0.0006720155943185091, "clip_ratio/high_mean": 0.0001425338094122708, "clip_ratio/low_mean": 2.376425836700946e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001662980648688972, "completions/clipped_ratio": 0.0, "completions/max_length": 186.66666666666666, "completions/mean_length": 73.82552083333333, "completions/min_length": 35.333333333333336, "epoch": 0.4992603550295858, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.005238387454301119, "learning_rate": 9.932208518877445e-07, "loss": 0.00022551454603672027, "reward": 1.604408065478007, "reward_std": 0.21501963337262472, "rewards/FidelityReward/mean": 0.7359968622525533, "rewards/FidelityReward/std": 0.2131421665350596, "rewards/JudgeFidelityReward/mean": 0.7394265532493591, "rewards/JudgeFidelityReward/std": 0.21736423671245575, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 675 }, { "clip_ratio/high_max": 0.0005513536278158426, "clip_ratio/high_mean": 0.00022856914438307285, "clip_ratio/low_mean": 0.00015237722254823893, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003809463698416948, "completions/clipped_ratio": 0.001953125, "completions/max_length": 348.0, "completions/mean_length": 74.408203125, "completions/min_length": 30.0, "epoch": 0.5029585798816568, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.005579598434269428, "learning_rate": 9.9301867082649e-07, "loss": 0.00020234780386090278, "reward": 1.6544288396835327, "reward_std": 0.19427715986967087, "rewards/FidelityReward/mean": 0.767146646976471, "rewards/FidelityReward/std": 0.2079976201057434, "rewards/JudgeFidelityReward/mean": 0.778470516204834, "rewards/JudgeFidelityReward/std": 0.19694960862398148, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 680 }, { "clip_ratio/high_max": 0.0003767442423850298, "clip_ratio/high_mean": 0.00019489788683131338, "clip_ratio/low_mean": 1.7099863907787948e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002119977492839098, "completions/clipped_ratio": 0.0, "completions/max_length": 212.66666666666666, "completions/mean_length": 73.2734375, "completions/min_length": 28.333333333333332, "epoch": 0.5066568047337278, "frac_reward_zero_std": 0.0, "grad_norm": 2.703125, "kl": 0.0044481676071882244, "learning_rate": 9.928135401708254e-07, "loss": -7.6370220631361e-05, "reward": 1.626562198003133, "reward_std": 0.20109639565149942, "rewards/FidelityReward/mean": 0.7582831184069315, "rewards/FidelityReward/std": 0.17893587549527487, "rewards/JudgeFidelityReward/mean": 0.7378602226575216, "rewards/JudgeFidelityReward/std": 0.18195356925328574, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 685 }, { "clip_ratio/high_max": 0.0009293691138736904, "clip_ratio/high_mean": 0.000271870126016438, "clip_ratio/low_mean": 4.021598870167509e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00031208611326292155, "completions/clipped_ratio": 0.0, "completions/max_length": 199.5, "completions/mean_length": 73.88671875, "completions/min_length": 31.0, "epoch": 0.5103550295857988, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.005139791220426559, "learning_rate": 9.926054611479903e-07, "loss": 6.721317768096924e-05, "reward": 1.6314152479171753, "reward_std": 0.20062200725078583, "rewards/FidelityReward/mean": 0.7474513947963715, "rewards/FidelityReward/std": 0.19083551317453384, "rewards/JudgeFidelityReward/mean": 0.7679277062416077, "rewards/JudgeFidelityReward/std": 0.19080710411071777, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 690 }, { "clip_ratio/high_max": 0.0008312552236020565, "clip_ratio/high_mean": 0.00015912791714072226, "clip_ratio/low_mean": 8.238742739195005e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002415153503534384, "completions/clipped_ratio": 0.0, "completions/max_length": 205.33333333333334, "completions/mean_length": 74.5625, "completions/min_length": 30.666666666666668, "epoch": 0.5140532544378699, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.375, "kl": 0.004979859990999103, "learning_rate": 9.923944350028639e-07, "loss": 0.0003156815655529499, "reward": 1.6099307537078857, "reward_std": 0.20621439317862192, "rewards/FidelityReward/mean": 0.7399346828460693, "rewards/FidelityReward/std": 0.20232923328876495, "rewards/JudgeFidelityReward/mean": 0.7452002962430319, "rewards/JudgeFidelityReward/std": 0.1970912665128708, "rewards/SelfEvolvingFormatReward/mean": 0.9947916666666666, "rewards/SelfEvolvingFormatReward/std": 0.0710716222723325, "step": 695 }, { "clip_ratio/high_max": 0.0007822310319170356, "clip_ratio/high_mean": 0.000217958923894912, "clip_ratio/low_mean": 1.8796992662828416e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023675591801293195, "completions/clipped_ratio": 0.0, "completions/max_length": 188.0, "completions/mean_length": 75.072265625, "completions/min_length": 32.5, "epoch": 0.5177514792899408, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.006121774669736624, "learning_rate": 9.921804629979578e-07, "loss": 0.0002884154440835118, "reward": 1.5787036418914795, "reward_std": 0.2122170552611351, "rewards/FidelityReward/mean": 0.7151718735694885, "rewards/FidelityReward/std": 0.21381255984306335, "rewards/JudgeFidelityReward/mean": 0.7329229712486267, "rewards/JudgeFidelityReward/std": 0.20938808470964432, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 700 }, { "clip_ratio/high_max": 0.000328472547698766, "clip_ratio/high_mean": 9.801167761906982e-05, "clip_ratio/low_mean": 2.129471831722185e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001193063915707171, "completions/clipped_ratio": 0.0, "completions/max_length": 224.66666666666666, "completions/mean_length": 76.9609375, "completions/min_length": 27.0, "epoch": 0.5214497041420119, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.59375, "kl": 0.0068633265793323515, "learning_rate": 9.91963546413407e-07, "loss": 0.0003611671272665262, "reward": 1.549961010615031, "reward_std": 0.22021152079105377, "rewards/FidelityReward/mean": 0.7047725915908813, "rewards/FidelityReward/std": 0.20843865474065146, "rewards/JudgeFidelityReward/mean": 0.6929810245831808, "rewards/JudgeFidelityReward/std": 0.21604843934377035, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 705 }, { "clip_ratio/high_max": 0.0006697633885778487, "clip_ratio/high_mean": 0.00023971974733285605, "clip_ratio/low_mean": 7.630152540514246e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003160212654620409, "completions/clipped_ratio": 0.0, "completions/max_length": 165.0, "completions/mean_length": 76.453125, "completions/min_length": 39.5, "epoch": 0.5251479289940828, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.006149150989949703, "learning_rate": 9.917436865469635e-07, "loss": 0.0003717912128195167, "reward": 1.6227176189422607, "reward_std": 0.20432966947555542, "rewards/FidelityReward/mean": 0.7467849552631378, "rewards/FidelityReward/std": 0.1974077746272087, "rewards/JudgeFidelityReward/mean": 0.7538183629512787, "rewards/JudgeFidelityReward/std": 0.18925636261701584, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 710 }, { "clip_ratio/high_max": 0.00039043225115165114, "clip_ratio/high_mean": 0.00013413281412795187, "clip_ratio/low_mean": 3.843279264401644e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017256560386158526, "completions/clipped_ratio": 0.0, "completions/max_length": 195.0, "completions/mean_length": 76.67057291666667, "completions/min_length": 30.0, "epoch": 0.5288461538461539, "frac_reward_zero_std": 0.0, "grad_norm": 2.3125, "kl": 0.006146595068275929, "learning_rate": 9.915208847139881e-07, "loss": 0.0005229634698480368, "reward": 1.598864475886027, "reward_std": 0.21758118271827698, "rewards/FidelityReward/mean": 0.7301936546961466, "rewards/FidelityReward/std": 0.2040904462337494, "rewards/JudgeFidelityReward/mean": 0.7386436859766642, "rewards/JudgeFidelityReward/std": 0.19808891912301382, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 715 }, { "clip_ratio/high_max": 0.001308955904096365, "clip_ratio/high_mean": 0.0003477999940514565, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003477999940514565, "completions/clipped_ratio": 0.0, "completions/max_length": 260.0, "completions/mean_length": 76.671875, "completions/min_length": 32.0, "epoch": 0.5325443786982249, "frac_reward_zero_std": 0.0, "grad_norm": 2.375, "kl": 0.006043228507041931, "learning_rate": 9.91295142247443e-07, "loss": 0.00030688885599374773, "reward": 1.577130913734436, "reward_std": 0.20784278213977814, "rewards/FidelityReward/mean": 0.7184266448020935, "rewards/FidelityReward/std": 0.20604544132947922, "rewards/JudgeFidelityReward/mean": 0.7193616628646851, "rewards/JudgeFidelityReward/std": 0.20358354598283768, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 720 }, { "clip_ratio/high_max": 0.000521993450820446, "clip_ratio/high_mean": 0.00010389559902250767, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010389559902250767, "completions/clipped_ratio": 0.0, "completions/max_length": 205.33333333333334, "completions/mean_length": 76.44661458333333, "completions/min_length": 35.333333333333336, "epoch": 0.5362426035502958, "frac_reward_zero_std": 0.0, "grad_norm": 2.5625, "kl": 0.005042331945151091, "learning_rate": 9.910664604978822e-07, "loss": 0.00023713335394859314, "reward": 1.6138988335927327, "reward_std": 0.21222730974356332, "rewards/FidelityReward/mean": 0.7417107025782267, "rewards/FidelityReward/std": 0.20158163706461588, "rewards/JudgeFidelityReward/mean": 0.7456783850987753, "rewards/JudgeFidelityReward/std": 0.19640442232290903, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 725 }, { "clip_ratio/high_max": 0.0007870008936151862, "clip_ratio/high_mean": 0.0002485668286681175, "clip_ratio/low_mean": 3.751230542548001e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002860791340935975, "completions/clipped_ratio": 0.0, "completions/max_length": 255.0, "completions/mean_length": 78.3359375, "completions/min_length": 28.0, "epoch": 0.5399408284023669, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.006225531920790672, "learning_rate": 9.90834840833446e-07, "loss": 0.00024828570894896984, "reward": 1.6109023690223694, "reward_std": 0.21324671804904938, "rewards/FidelityReward/mean": 0.734077513217926, "rewards/FidelityReward/std": 0.2063208371400833, "rewards/JudgeFidelityReward/mean": 0.7575558722019196, "rewards/JudgeFidelityReward/std": 0.209877647459507, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 730 }, { "clip_ratio/high_max": 0.0007371589774265886, "clip_ratio/high_mean": 0.000230677321087569, "clip_ratio/low_mean": 4.405827785376459e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002747355960309505, "completions/clipped_ratio": 0.0, "completions/max_length": 181.33333333333334, "completions/mean_length": 76.24348958333333, "completions/min_length": 35.0, "epoch": 0.5436390532544378, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.00494412137195468, "learning_rate": 9.906002846398504e-07, "loss": 0.0001399720087647438, "reward": 1.576185663541158, "reward_std": 0.2198815792798996, "rewards/FidelityReward/mean": 0.7156232595443726, "rewards/FidelityReward/std": 0.21342013776302338, "rewards/JudgeFidelityReward/mean": 0.7211247881253561, "rewards/JudgeFidelityReward/std": 0.21292108297348022, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 735 }, { "clip_ratio/high_max": 0.00047237435355782507, "clip_ratio/high_mean": 0.00011676251306198537, "clip_ratio/low_mean": 0.00012881142320111393, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002455739420838654, "completions/clipped_ratio": 0.0, "completions/max_length": 217.0, "completions/mean_length": 76.681640625, "completions/min_length": 36.5, "epoch": 0.5473372781065089, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.006341333594173193, "learning_rate": 9.903627933203804e-07, "loss": 0.00040895426645874976, "reward": 1.5731014013290405, "reward_std": 0.22413166612386703, "rewards/FidelityReward/mean": 0.709240734577179, "rewards/FidelityReward/std": 0.2228947877883911, "rewards/JudgeFidelityReward/mean": 0.7355337142944336, "rewards/JudgeFidelityReward/std": 0.21072284132242203, "rewards/SelfEvolvingFormatReward/mean": 0.9921875, "rewards/SelfEvolvingFormatReward/std": 0.08821486681699753, "step": 740 }, { "clip_ratio/high_max": 0.0008029655553400516, "clip_ratio/high_mean": 0.00026207681512460115, "clip_ratio/low_mean": 1.860119082266465e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028067800449207427, "completions/clipped_ratio": 0.0, "completions/max_length": 189.33333333333334, "completions/mean_length": 76.67838541666667, "completions/min_length": 32.333333333333336, "epoch": 0.5510355029585798, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.0062568397261202335, "learning_rate": 9.901223682958812e-07, "loss": 0.0005213118623942137, "reward": 1.64272936185201, "reward_std": 0.20646803577740988, "rewards/FidelityReward/mean": 0.7633794943491617, "rewards/FidelityReward/std": 0.19549256066481271, "rewards/JudgeFidelityReward/mean": 0.7586995959281921, "rewards/JudgeFidelityReward/std": 0.20247668027877808, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 745 }, { "clip_ratio/high_max": 0.000855352554935962, "clip_ratio/high_mean": 0.00022937292815186083, "clip_ratio/low_mean": 6.073620461393148e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00029010913567617536, "completions/clipped_ratio": 0.0, "completions/max_length": 190.0, "completions/mean_length": 75.119140625, "completions/min_length": 34.5, "epoch": 0.5547337278106509, "frac_reward_zero_std": 0.0, "grad_norm": 2.484375, "kl": 0.0066133617423474785, "learning_rate": 9.898790110047487e-07, "loss": 0.0003524883184581995, "reward": 1.620184302330017, "reward_std": 0.20302222669124603, "rewards/FidelityReward/mean": 0.7497121691703796, "rewards/FidelityReward/std": 0.1864975243806839, "rewards/JudgeFidelityReward/mean": 0.7409440875053406, "rewards/JudgeFidelityReward/std": 0.1983468309044838, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 750 }, { "clip_ratio/high_max": 0.00048128791386261585, "clip_ratio/high_mean": 0.0001818574091885239, "clip_ratio/low_mean": 4.1041974327526987e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002228993864264339, "completions/clipped_ratio": 0.0, "completions/max_length": 191.33333333333334, "completions/mean_length": 77.0625, "completions/min_length": 29.0, "epoch": 0.5584319526627219, "frac_reward_zero_std": 0.0, "grad_norm": 2.5625, "kl": 0.0060218476690351965, "learning_rate": 9.89632722902923e-07, "loss": 0.000290187681093812, "reward": 1.6244206825892131, "reward_std": 0.19428101181983948, "rewards/FidelityReward/mean": 0.7485997279485067, "rewards/FidelityReward/std": 0.20297448337078094, "rewards/JudgeFidelityReward/mean": 0.7542461554209391, "rewards/JudgeFidelityReward/std": 0.20165124535560608, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 755 }, { "clip_ratio/high_max": 0.0007997292908839881, "clip_ratio/high_mean": 0.00023290744284167885, "clip_ratio/low_mean": 3.953271079808474e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002724401536397636, "completions/clipped_ratio": 0.0, "completions/max_length": 216.5, "completions/mean_length": 77.49609375, "completions/min_length": 34.5, "epoch": 0.5621301775147929, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.006000096909701824, "learning_rate": 9.893835054638772e-07, "loss": 0.0004094441421329975, "reward": 1.5942033529281616, "reward_std": 0.20754338800907135, "rewards/FidelityReward/mean": 0.730894148349762, "rewards/FidelityReward/std": 0.1948232725262642, "rewards/JudgeFidelityReward/mean": 0.7266183197498322, "rewards/JudgeFidelityReward/std": 0.19954975694417953, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 760 }, { "clip_ratio/high_max": 0.0003563633421435952, "clip_ratio/high_mean": 6.358578393701464e-05, "clip_ratio/low_mean": 3.8888737617526206e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010247452300973236, "completions/clipped_ratio": 0.0, "completions/max_length": 198.0, "completions/mean_length": 75.58463541666667, "completions/min_length": 36.0, "epoch": 0.5658284023668639, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.006447659712284804, "learning_rate": 9.891313601786105e-07, "loss": 0.00013560294173657893, "reward": 1.5976508458455403, "reward_std": 0.21834387381871542, "rewards/FidelityReward/mean": 0.7293724020322164, "rewards/FidelityReward/std": 0.20185677707195282, "rewards/JudgeFidelityReward/mean": 0.7391611735026041, "rewards/JudgeFidelityReward/std": 0.20101639131704965, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 765 }, { "clip_ratio/high_max": 0.0008768243016675115, "clip_ratio/high_mean": 0.00020880444499198346, "clip_ratio/low_mean": 5.691240367013961e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026571685448288916, "completions/clipped_ratio": 0.0, "completions/max_length": 187.5, "completions/mean_length": 77.072265625, "completions/min_length": 40.5, "epoch": 0.5695266272189349, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.007613780535757541, "learning_rate": 9.888762885556383e-07, "loss": 0.00038325986824929716, "reward": 1.6490915417671204, "reward_std": 0.20874547958374023, "rewards/FidelityReward/mean": 0.7703550457954407, "rewards/FidelityReward/std": 0.17690405994653702, "rewards/JudgeFidelityReward/mean": 0.7574730217456818, "rewards/JudgeFidelityReward/std": 0.18506287038326263, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 770 }, { "clip_ratio/high_max": 0.0007764343405142427, "clip_ratio/high_mean": 0.0001681651920080185, "clip_ratio/low_mean": 4.2834891064558175e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021100008161738514, "completions/clipped_ratio": 0.0, "completions/max_length": 201.0, "completions/mean_length": 79.46484375, "completions/min_length": 35.0, "epoch": 0.5732248520710059, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.006699245702475309, "learning_rate": 9.886182921209839e-07, "loss": 0.00035945745185017586, "reward": 1.677042285601298, "reward_std": 0.2039742370446523, "rewards/FidelityReward/mean": 0.7825523416201273, "rewards/FidelityReward/std": 0.18530964851379395, "rewards/JudgeFidelityReward/mean": 0.7928861776987711, "rewards/JudgeFidelityReward/std": 0.1743646115064621, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 775 }, { "clip_ratio/high_max": 0.001135384920053184, "clip_ratio/high_mean": 0.00030772154568694534, "clip_ratio/low_mean": 2.192982501583174e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00032965136924758556, "completions/clipped_ratio": 0.0, "completions/max_length": 201.0, "completions/mean_length": 75.3828125, "completions/min_length": 34.5, "epoch": 0.5769230769230769, "frac_reward_zero_std": 0.0, "grad_norm": 2.484375, "kl": 0.006617210432887077, "learning_rate": 9.883573724181682e-07, "loss": 0.00036316378973424437, "reward": 1.5546387434005737, "reward_std": 0.21611519902944565, "rewards/FidelityReward/mean": 0.7076002955436707, "rewards/FidelityReward/std": 0.22012800723314285, "rewards/JudgeFidelityReward/mean": 0.6979831755161285, "rewards/JudgeFidelityReward/std": 0.2173531875014305, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 780 }, { "clip_ratio/high_max": 0.0005025631049647928, "clip_ratio/high_mean": 0.00013755129766650497, "clip_ratio/low_mean": 5.6714133825153114e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019426543731242417, "completions/clipped_ratio": 0.0, "completions/max_length": 172.33333333333334, "completions/mean_length": 77.57421875, "completions/min_length": 33.0, "epoch": 0.5806213017751479, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.0072910023853182794, "learning_rate": 9.88093531008202e-07, "loss": 0.00037615546025335787, "reward": 1.5640195210774739, "reward_std": 0.22800476352373758, "rewards/FidelityReward/mean": 0.7105435331662496, "rewards/FidelityReward/std": 0.2092933108409246, "rewards/JudgeFidelityReward/mean": 0.7082541386286417, "rewards/JudgeFidelityReward/std": 0.20375020802021027, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 785 }, { "clip_ratio/high_max": 0.0005487236427143216, "clip_ratio/high_mean": 0.00013949119311291726, "clip_ratio/low_mean": 0.00010097268532263115, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024046387989073992, "completions/clipped_ratio": 0.0, "completions/max_length": 257.0, "completions/mean_length": 77.931640625, "completions/min_length": 39.0, "epoch": 0.584319526627219, "frac_reward_zero_std": 0.0, "grad_norm": 2.3125, "kl": 0.008372246194630862, "learning_rate": 9.878267694695757e-07, "loss": 0.0003032308537513018, "reward": 1.6130722761154175, "reward_std": 0.21621615439653397, "rewards/FidelityReward/mean": 0.7442932724952698, "rewards/FidelityReward/std": 0.21109506487846375, "rewards/JudgeFidelityReward/mean": 0.7434174716472626, "rewards/JudgeFidelityReward/std": 0.20939210802316666, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 790 }, { "clip_ratio/high_max": 0.0006030929740518332, "clip_ratio/high_mean": 0.00017470954917371274, "clip_ratio/low_mean": 6.542735791299492e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024013691581785678, "completions/clipped_ratio": 0.0, "completions/max_length": 183.33333333333334, "completions/mean_length": 78.54817708333333, "completions/min_length": 39.0, "epoch": 0.5880177514792899, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.007111062202602625, "learning_rate": 9.875570893982492e-07, "loss": 0.0002952625975012779, "reward": 1.6291248401006062, "reward_std": 0.2102301816145579, "rewards/FidelityReward/mean": 0.7526738047599792, "rewards/FidelityReward/std": 0.18910261988639832, "rewards/JudgeFidelityReward/mean": 0.7542041341463724, "rewards/JudgeFidelityReward/std": 0.1855981002251307, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 795 }, { "clip_ratio/high_max": 0.0011673670029267668, "clip_ratio/high_mean": 0.00031763374572619797, "clip_ratio/low_mean": 9.113819687627256e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00040877193678170445, "completions/clipped_ratio": 0.0, "completions/max_length": 214.5, "completions/mean_length": 76.083984375, "completions/min_length": 35.5, "epoch": 0.591715976331361, "frac_reward_zero_std": 0.0, "grad_norm": 2.78125, "kl": 0.007674399483948946, "learning_rate": 9.872844924076445e-07, "loss": 0.00047019091434776785, "reward": 1.6338527202606201, "reward_std": 0.20765593647956848, "rewards/FidelityReward/mean": 0.7534018456935883, "rewards/FidelityReward/std": 0.20375625789165497, "rewards/JudgeFidelityReward/mean": 0.7628550231456757, "rewards/JudgeFidelityReward/std": 0.19171332567930222, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 800 }, { "clip_ratio/high_max": 0.0006853373954072595, "clip_ratio/high_mean": 0.000122348967124708, "clip_ratio/low_mean": 4.292577505111694e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016527474508620799, "completions/clipped_ratio": 0.0, "completions/max_length": 169.33333333333334, "completions/mean_length": 75.37239583333333, "completions/min_length": 35.666666666666664, "epoch": 0.5954142011834319, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.006653666868805886, "learning_rate": 9.870089801286334e-07, "loss": 0.0001350505743175745, "reward": 1.5754904747009277, "reward_std": 0.22315271198749542, "rewards/FidelityReward/mean": 0.716998298962911, "rewards/FidelityReward/std": 0.19983708361784616, "rewards/JudgeFidelityReward/mean": 0.7195885380109152, "rewards/JudgeFidelityReward/std": 0.1928227742513021, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 805 }, { "clip_ratio/high_max": 0.0006157349795103073, "clip_ratio/high_mean": 0.00026414587046019733, "clip_ratio/low_mean": 4.817792942048982e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003123237984254956, "completions/clipped_ratio": 0.0, "completions/max_length": 176.0, "completions/mean_length": 75.431640625, "completions/min_length": 29.5, "epoch": 0.599112426035503, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.009010842721909284, "learning_rate": 9.867305542095302e-07, "loss": 0.0003471565432846546, "reward": 1.5644336342811584, "reward_std": 0.22022781521081924, "rewards/FidelityReward/mean": 0.7120645046234131, "rewards/FidelityReward/std": 0.21032097190618515, "rewards/JudgeFidelityReward/mean": 0.704738199710846, "rewards/JudgeFidelityReward/std": 0.2233249470591545, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 810 }, { "clip_ratio/high_max": 0.000564801786094904, "clip_ratio/high_mean": 0.00014699688763357698, "clip_ratio/low_mean": 3.9386170101352036e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018638305482454598, "completions/clipped_ratio": 0.0, "completions/max_length": 216.66666666666666, "completions/mean_length": 75.47135416666667, "completions/min_length": 34.0, "epoch": 0.602810650887574, "frac_reward_zero_std": 0.0, "grad_norm": 2.6875, "kl": 0.007973196543753147, "learning_rate": 9.8644921631608e-07, "loss": 0.00047316388227045534, "reward": 1.6186654965082805, "reward_std": 0.21414833764235178, "rewards/FidelityReward/mean": 0.7442921797434489, "rewards/FidelityReward/std": 0.19023569921652475, "rewards/JudgeFidelityReward/mean": 0.7513506809870402, "rewards/JudgeFidelityReward/std": 0.1960662305355072, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 815 }, { "clip_ratio/high_max": 0.000698703364469111, "clip_ratio/high_mean": 0.0002691949950531125, "clip_ratio/low_mean": 9.190987620968371e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00036110487999394534, "completions/clipped_ratio": 0.0, "completions/max_length": 165.5, "completions/mean_length": 76.349609375, "completions/min_length": 34.5, "epoch": 0.606508875739645, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.0069239710457623, "learning_rate": 9.861649681314497e-07, "loss": 0.0002308595459908247, "reward": 1.5997392535209656, "reward_std": 0.20219645649194717, "rewards/FidelityReward/mean": 0.7303381562232971, "rewards/FidelityReward/std": 0.22605765610933304, "rewards/JudgeFidelityReward/mean": 0.7427084147930145, "rewards/JudgeFidelityReward/std": 0.22530680149793625, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 820 }, { "clip_ratio/high_max": 0.00047020590864121914, "clip_ratio/high_mean": 0.00015636686002835632, "clip_ratio/low_mean": 0.00010813520348165184, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002645020722411573, "completions/clipped_ratio": 0.0, "completions/max_length": 216.0, "completions/mean_length": 77.66015625, "completions/min_length": 34.0, "epoch": 0.610207100591716, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.009316197037696839, "learning_rate": 9.858778113562174e-07, "loss": 0.00036242660135030744, "reward": 1.5634745756785076, "reward_std": 0.23912526667118073, "rewards/FidelityReward/mean": 0.7108033498128256, "rewards/FidelityReward/std": 0.2147970845301946, "rewards/JudgeFidelityReward/mean": 0.7131549517313639, "rewards/JudgeFidelityReward/std": 0.20235863824685416, "rewards/SelfEvolvingFormatReward/mean": 0.9921875, "rewards/SelfEvolvingFormatReward/std": 0.07082580029964447, "step": 825 }, { "clip_ratio/high_max": 0.0011055229580961168, "clip_ratio/high_mean": 0.000344687735196203, "clip_ratio/low_mean": 4.121943493373692e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003859071643091738, "completions/clipped_ratio": 0.0, "completions/max_length": 202.0, "completions/mean_length": 78.21484375, "completions/min_length": 30.5, "epoch": 0.613905325443787, "frac_reward_zero_std": 0.0, "grad_norm": 2.328125, "kl": 0.007284500636160374, "learning_rate": 9.855877477083631e-07, "loss": 0.0003501262050122023, "reward": 1.6009833216667175, "reward_std": 0.20200278609991074, "rewards/FidelityReward/mean": 0.728533148765564, "rewards/FidelityReward/std": 0.19882741570472717, "rewards/JudgeFidelityReward/mean": 0.7468535006046295, "rewards/JudgeFidelityReward/std": 0.18772178143262863, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 830 }, { "clip_ratio/high_max": 0.0005142310750670731, "clip_ratio/high_mean": 0.00014275758876465262, "clip_ratio/low_mean": 0.00011568458285182714, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00025844216579571364, "completions/clipped_ratio": 0.0, "completions/max_length": 172.33333333333334, "completions/mean_length": 75.8671875, "completions/min_length": 33.0, "epoch": 0.617603550295858, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.007333477679640055, "learning_rate": 9.85294778923257e-07, "loss": 0.00042571322992444036, "reward": 1.6090312004089355, "reward_std": 0.21055913964907327, "rewards/FidelityReward/mean": 0.7392082810401917, "rewards/FidelityReward/std": 0.20193726321061453, "rewards/JudgeFidelityReward/mean": 0.7409478227297465, "rewards/JudgeFidelityReward/std": 0.20308561623096466, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 835 }, { "clip_ratio/high_max": 0.0006796568515710533, "clip_ratio/high_mean": 0.0002712131245061755, "clip_ratio/low_mean": 6.82917088852264e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00033950484357774257, "completions/clipped_ratio": 0.0, "completions/max_length": 180.5, "completions/mean_length": 75.515625, "completions/min_length": 28.5, "epoch": 0.621301775147929, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.006410726346075535, "learning_rate": 9.849989067536507e-07, "loss": 4.7613726928830145e-05, "reward": 1.662231206893921, "reward_std": 0.19617703557014465, "rewards/FidelityReward/mean": 0.7756628692150116, "rewards/FidelityReward/std": 0.18430273979902267, "rewards/JudgeFidelityReward/mean": 0.7731368243694305, "rewards/JudgeFidelityReward/std": 0.18185824900865555, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 840 }, { "clip_ratio/high_max": 0.000666630279738456, "clip_ratio/high_mean": 0.0001284594021853991, "clip_ratio/low_mean": 1.8477457342669367e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014693686534883454, "completions/clipped_ratio": 0.0, "completions/max_length": 174.66666666666666, "completions/mean_length": 74.8125, "completions/min_length": 29.333333333333332, "epoch": 0.625, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.007177881337702275, "learning_rate": 9.847001329696652e-07, "loss": 0.0002669785637408495, "reward": 1.660553773244222, "reward_std": 0.19326584041118622, "rewards/FidelityReward/mean": 0.7789114713668823, "rewards/FidelityReward/std": 0.18812530736128488, "rewards/JudgeFidelityReward/mean": 0.7658888697624207, "rewards/JudgeFidelityReward/std": 0.19746166964372, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 845 }, { "clip_ratio/high_max": 0.0006065407767891883, "clip_ratio/high_mean": 0.00017542308196425438, "clip_ratio/low_mean": 2.0781380590051414e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001962044625543058, "completions/clipped_ratio": 0.0, "completions/max_length": 164.0, "completions/mean_length": 74.255859375, "completions/min_length": 31.5, "epoch": 0.628698224852071, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.007469234149903059, "learning_rate": 9.843984593587818e-07, "loss": 0.0004725962411612272, "reward": 1.6321430206298828, "reward_std": 0.20183520019054413, "rewards/FidelityReward/mean": 0.7538014650344849, "rewards/FidelityReward/std": 0.1958967223763466, "rewards/JudgeFidelityReward/mean": 0.7586361467838287, "rewards/JudgeFidelityReward/std": 0.19522550702095032, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 850 }, { "clip_ratio/high_max": 0.00038098860532045367, "clip_ratio/high_mean": 0.00018693693564273418, "clip_ratio/low_mean": 8.503712888341397e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002719740616157651, "completions/clipped_ratio": 0.0, "completions/max_length": 191.33333333333334, "completions/mean_length": 77.24348958333333, "completions/min_length": 33.333333333333336, "epoch": 0.632396449704142, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.006835707556456327, "learning_rate": 9.8409388772583e-07, "loss": 0.0002738918410614133, "reward": 1.5829636653264363, "reward_std": 0.21849354604880014, "rewards/FidelityReward/mean": 0.7160462141036987, "rewards/FidelityReward/std": 0.2219461351633072, "rewards/JudgeFidelityReward/mean": 0.7351369063059489, "rewards/JudgeFidelityReward/std": 0.21062604089577994, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 855 }, { "clip_ratio/high_max": 0.000748901255428791, "clip_ratio/high_mean": 0.0003011850465554744, "clip_ratio/low_mean": 7.637886446900665e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00037756392848677935, "completions/clipped_ratio": 0.0, "completions/max_length": 184.0, "completions/mean_length": 75.435546875, "completions/min_length": 36.0, "epoch": 0.636094674556213, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.00665720934048295, "learning_rate": 9.837864198929781e-07, "loss": 0.00017134281806647779, "reward": 1.5845158100128174, "reward_std": 0.2160099744796753, "rewards/FidelityReward/mean": 0.723716139793396, "rewards/FidelityReward/std": 0.20091509819030762, "rewards/JudgeFidelityReward/mean": 0.7235525548458099, "rewards/JudgeFidelityReward/std": 0.19612087309360504, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 860 }, { "clip_ratio/high_max": 0.00035094497725367545, "clip_ratio/high_mean": 8.238108566729351e-05, "clip_ratio/low_mean": 4.763748729601503e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013001857878407462, "completions/clipped_ratio": 0.0, "completions/max_length": 189.0, "completions/mean_length": 76.01953125, "completions/min_length": 38.333333333333336, "epoch": 0.639792899408284, "frac_reward_zero_std": 0.0, "grad_norm": 2.25, "kl": 0.007047826610505581, "learning_rate": 9.834760576997208e-07, "loss": 0.00037513538263738154, "reward": 1.6246096293131511, "reward_std": 0.20314790308475494, "rewards/FidelityReward/mean": 0.752312958240509, "rewards/FidelityReward/std": 0.1872710088888804, "rewards/JudgeFidelityReward/mean": 0.7471974889437357, "rewards/JudgeFidelityReward/std": 0.17355760435263315, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 865 }, { "clip_ratio/high_max": 0.0010696630808524788, "clip_ratio/high_mean": 0.0001918159716296941, "clip_ratio/low_mean": 4.300545551814139e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002348214271478355, "completions/clipped_ratio": 0.0, "completions/max_length": 184.0, "completions/mean_length": 76.013671875, "completions/min_length": 37.5, "epoch": 0.643491124260355, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.006967041827738285, "learning_rate": 9.831628030028696e-07, "loss": 0.0003243934828788042, "reward": 1.665343701839447, "reward_std": 0.1974812150001526, "rewards/FidelityReward/mean": 0.7775123715400696, "rewards/FidelityReward/std": 0.17904037237167358, "rewards/JudgeFidelityReward/mean": 0.7776155471801758, "rewards/JudgeFidelityReward/std": 0.17228947579860687, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 870 }, { "clip_ratio/high_max": 0.0006620194995775819, "clip_ratio/high_mean": 0.0002246265299618244, "clip_ratio/low_mean": 4.154117777943611e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026616770774126055, "completions/clipped_ratio": 0.0, "completions/max_length": 173.66666666666666, "completions/mean_length": 75.29557291666667, "completions/min_length": 33.666666666666664, "epoch": 0.647189349112426, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.008031763508915902, "learning_rate": 9.828466576765403e-07, "loss": 0.00022407553624361753, "reward": 1.6476030349731445, "reward_std": 0.18800397217273712, "rewards/FidelityReward/mean": 0.7637370824813843, "rewards/FidelityReward/std": 0.20707464714845022, "rewards/JudgeFidelityReward/mean": 0.7690338889757792, "rewards/JudgeFidelityReward/std": 0.20620483656724295, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 875 }, { "clip_ratio/high_max": 0.0007695504231378436, "clip_ratio/high_mean": 0.00019796034903265535, "clip_ratio/low_mean": 2.6399156195111572e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022435950813814998, "completions/clipped_ratio": 0.0, "completions/max_length": 193.0, "completions/mean_length": 75.57421875, "completions/min_length": 27.0, "epoch": 0.650887573964497, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.007655597850680351, "learning_rate": 9.825276236121434e-07, "loss": 0.0004556804429739714, "reward": 1.628957748413086, "reward_std": 0.20570534467697144, "rewards/FidelityReward/mean": 0.7537681460380554, "rewards/FidelityReward/std": 0.20845210552215576, "rewards/JudgeFidelityReward/mean": 0.7503792345523834, "rewards/JudgeFidelityReward/std": 0.2038949728012085, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 880 }, { "clip_ratio/high_max": 0.00048547302139922976, "clip_ratio/high_mean": 0.00016913856961764395, "clip_ratio/low_mean": 3.900156007148325e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020814012386836112, "completions/clipped_ratio": 0.0, "completions/max_length": 206.33333333333334, "completions/mean_length": 77.72005208333333, "completions/min_length": 33.333333333333336, "epoch": 0.6545857988165681, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.007581160310655833, "learning_rate": 9.822057027183714e-07, "loss": 0.00030965209007263186, "reward": 1.5899425347646077, "reward_std": 0.22981211046377817, "rewards/FidelityReward/mean": 0.729614237944285, "rewards/FidelityReward/std": 0.18747964004675546, "rewards/JudgeFidelityReward/mean": 0.7219586968421936, "rewards/JudgeFidelityReward/std": 0.19117925067742667, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 885 }, { "clip_ratio/high_max": 0.0009432363556697964, "clip_ratio/high_mean": 0.0002410289307590574, "clip_ratio/low_mean": 6.609153642784804e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003071204642765224, "completions/clipped_ratio": 0.0, "completions/max_length": 188.0, "completions/mean_length": 74.1796875, "completions/min_length": 36.0, "epoch": 0.658284023668639, "frac_reward_zero_std": 0.0, "grad_norm": 2.984375, "kl": 0.009068357199430466, "learning_rate": 9.818808969211877e-07, "loss": 0.0005230037495493889, "reward": 1.6364968419075012, "reward_std": 0.20403960347175598, "rewards/FidelityReward/mean": 0.7568115592002869, "rewards/FidelityReward/std": 0.19230209290981293, "rewards/JudgeFidelityReward/mean": 0.7613236606121063, "rewards/JudgeFidelityReward/std": 0.1824837028980255, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 890 }, { "clip_ratio/high_max": 0.0006653645308688283, "clip_ratio/high_mean": 0.00014354295271914452, "clip_ratio/low_mean": 2.1061499137431382e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016460444021504372, "completions/clipped_ratio": 0.0, "completions/max_length": 178.33333333333334, "completions/mean_length": 77.26953125, "completions/min_length": 33.666666666666664, "epoch": 0.6619822485207101, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.007567931152880192, "learning_rate": 9.815532081638163e-07, "loss": 0.0003873394802212715, "reward": 1.5944103002548218, "reward_std": 0.21255473295847574, "rewards/FidelityReward/mean": 0.7263364394505819, "rewards/FidelityReward/std": 0.2049984335899353, "rewards/JudgeFidelityReward/mean": 0.7374498248100281, "rewards/JudgeFidelityReward/std": 0.20268590251604715, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 895 }, { "clip_ratio/high_max": 0.0008377424441277981, "clip_ratio/high_mean": 0.00026236162520945074, "clip_ratio/low_mean": 2.3923444678075613e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002862850669771433, "completions/clipped_ratio": 0.0, "completions/max_length": 174.5, "completions/mean_length": 75.931640625, "completions/min_length": 36.5, "epoch": 0.665680473372781, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.007980367820709943, "learning_rate": 9.812226384067276e-07, "loss": 0.00040845982730388644, "reward": 1.6802208423614502, "reward_std": 0.17336410284042358, "rewards/FidelityReward/mean": 0.7810794413089752, "rewards/FidelityReward/std": 0.2015056386590004, "rewards/JudgeFidelityReward/mean": 0.8002357482910156, "rewards/JudgeFidelityReward/std": 0.19844864308834076, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 900 }, { "clip_ratio/high_max": 0.0005156411905772984, "clip_ratio/high_mean": 0.0001798563520424068, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001798563520424068, "completions/clipped_ratio": 0.0, "completions/max_length": 211.0, "completions/mean_length": 78.56510416666667, "completions/min_length": 38.0, "epoch": 0.6693786982248521, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.008538462780416011, "learning_rate": 9.808891896276297e-07, "loss": 0.00017935032956302167, "reward": 1.5977999369303386, "reward_std": 0.21426155666510263, "rewards/FidelityReward/mean": 0.7285613417625427, "rewards/FidelityReward/std": 0.201117143034935, "rewards/JudgeFidelityReward/mean": 0.7410813172658285, "rewards/JudgeFidelityReward/std": 0.1871932695309321, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 905 }, { "clip_ratio/high_max": 0.0006282229442149401, "clip_ratio/high_mean": 0.00011403907410567626, "clip_ratio/low_mean": 5.353096639737487e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016757004195824265, "completions/clipped_ratio": 0.0, "completions/max_length": 191.5, "completions/mean_length": 76.482421875, "completions/min_length": 36.0, "epoch": 0.6730769230769231, "frac_reward_zero_std": 0.0, "grad_norm": 2.34375, "kl": 0.009536414779722691, "learning_rate": 9.805528638214542e-07, "loss": 0.0005182779394090175, "reward": 1.6025909185409546, "reward_std": 0.21002235263586044, "rewards/FidelityReward/mean": 0.7330899238586426, "rewards/FidelityReward/std": 0.20444927364587784, "rewards/JudgeFidelityReward/mean": 0.740955114364624, "rewards/JudgeFidelityReward/std": 0.196851447224617, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 910 }, { "clip_ratio/high_max": 0.0005484477151185274, "clip_ratio/high_mean": 0.00017186185577884318, "clip_ratio/low_mean": 4.2494008084759116e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021435586968436836, "completions/clipped_ratio": 0.0, "completions/max_length": 184.66666666666666, "completions/mean_length": 76.515625, "completions/min_length": 30.666666666666668, "epoch": 0.6767751479289941, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.008743559941649437, "learning_rate": 9.802136630003454e-07, "loss": 0.000198856508359313, "reward": 1.616625150044759, "reward_std": 0.20388641953468323, "rewards/FidelityReward/mean": 0.746747632821401, "rewards/FidelityReward/std": 0.19720612466335297, "rewards/JudgeFidelityReward/mean": 0.7410572171211243, "rewards/JudgeFidelityReward/std": 0.2008254130681356, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 915 }, { "clip_ratio/high_max": 0.000983644521329552, "clip_ratio/high_mean": 0.00027796945360023526, "clip_ratio/low_mean": 5.522453138837591e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00033319397189188746, "completions/clipped_ratio": 0.0, "completions/max_length": 165.0, "completions/mean_length": 78.796875, "completions/min_length": 42.5, "epoch": 0.6804733727810651, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.00919725950807333, "learning_rate": 9.798715891936481e-07, "loss": 0.00044572283513844013, "reward": 1.6730541586875916, "reward_std": 0.20038362592458725, "rewards/FidelityReward/mean": 0.785513162612915, "rewards/FidelityReward/std": 0.18866869062185287, "rewards/JudgeFidelityReward/mean": 0.778988242149353, "rewards/JudgeFidelityReward/std": 0.19278603792190552, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 920 }, { "clip_ratio/high_max": 0.000466849550139159, "clip_ratio/high_mean": 0.00014965140144340694, "clip_ratio/low_mean": 6.552179984282702e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021517318673431874, "completions/clipped_ratio": 0.0, "completions/max_length": 190.33333333333334, "completions/mean_length": 75.71223958333333, "completions/min_length": 26.666666666666668, "epoch": 0.6841715976331361, "frac_reward_zero_std": 0.0, "grad_norm": 2.765625, "kl": 0.009553951770067215, "learning_rate": 9.795266444478953e-07, "loss": 0.0005463844165205956, "reward": 1.6196612914403279, "reward_std": 0.21349995831648508, "rewards/FidelityReward/mean": 0.7528519630432129, "rewards/FidelityReward/std": 0.18955649932225546, "rewards/JudgeFidelityReward/mean": 0.7388269901275635, "rewards/JudgeFidelityReward/std": 0.1914252738157908, "rewards/SelfEvolvingFormatReward/mean": 0.9947916666666666, "rewards/SelfEvolvingFormatReward/std": 0.05880991121133169, "step": 925 }, { "clip_ratio/high_max": 0.0007515575271099806, "clip_ratio/high_mean": 0.00015007682086434215, "clip_ratio/low_mean": 0.00012971619435120374, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002797930093947798, "completions/clipped_ratio": 0.0, "completions/max_length": 221.0, "completions/mean_length": 80.7265625, "completions/min_length": 35.5, "epoch": 0.6878698224852071, "frac_reward_zero_std": 0.0, "grad_norm": 2.328125, "kl": 0.010144796408712863, "learning_rate": 9.791788308267962e-07, "loss": 0.00039056022651493547, "reward": 1.6121777296066284, "reward_std": 0.21968581527471542, "rewards/FidelityReward/mean": 0.7451270520687103, "rewards/FidelityReward/std": 0.1969980075955391, "rewards/JudgeFidelityReward/mean": 0.7380076050758362, "rewards/JudgeFidelityReward/std": 0.20454847812652588, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 930 }, { "clip_ratio/high_max": 0.0005246573360636831, "clip_ratio/high_mean": 0.00016497320029884577, "clip_ratio/low_mean": 2.5227043079212307e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019020024337805807, "completions/clipped_ratio": 0.0, "completions/max_length": 223.66666666666666, "completions/mean_length": 77.20442708333333, "completions/min_length": 32.666666666666664, "epoch": 0.6915680473372781, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.375, "kl": 0.010300336219370366, "learning_rate": 9.788281504112232e-07, "loss": 0.00043566394597291944, "reward": 1.599662979443868, "reward_std": 0.21139054000377655, "rewards/FidelityReward/mean": 0.7398835817972819, "rewards/FidelityReward/std": 0.18745198845863342, "rewards/JudgeFidelityReward/mean": 0.7208608587582906, "rewards/JudgeFidelityReward/std": 0.20212671160697937, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 935 }, { "clip_ratio/high_max": 0.0007656911155208945, "clip_ratio/high_mean": 0.00019750687060877682, "clip_ratio/low_mean": 3.5919836955145e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002334267133846879, "completions/clipped_ratio": 0.0, "completions/max_length": 181.0, "completions/mean_length": 79.8359375, "completions/min_length": 35.0, "epoch": 0.6952662721893491, "frac_reward_zero_std": 0.0, "grad_norm": 2.859375, "kl": 0.008695637714117766, "learning_rate": 9.784746052992008e-07, "loss": 0.0005990506615489721, "reward": 1.632398247718811, "reward_std": 0.21080750972032547, "rewards/FidelityReward/mean": 0.7518240809440613, "rewards/FidelityReward/std": 0.20173879712820053, "rewards/JudgeFidelityReward/mean": 0.7631014287471771, "rewards/JudgeFidelityReward/std": 0.1924022138118744, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 940 }, { "clip_ratio/high_max": 0.00033271079882979393, "clip_ratio/high_mean": 0.00011417336645536125, "clip_ratio/low_mean": 1.7182130250148475e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001313554937951267, "completions/clipped_ratio": 0.0, "completions/max_length": 214.66666666666666, "completions/mean_length": 79.64322916666667, "completions/min_length": 33.666666666666664, "epoch": 0.6989644970414202, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.010127720050513744, "learning_rate": 9.781181976058912e-07, "loss": 0.00031165997497737406, "reward": 1.5436890522638957, "reward_std": 0.22916030387083688, "rewards/FidelityReward/mean": 0.699451764424642, "rewards/FidelityReward/std": 0.21879087885220846, "rewards/JudgeFidelityReward/mean": 0.6897766391436259, "rewards/JudgeFidelityReward/std": 0.21769803762435913, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 945 }, { "clip_ratio/high_max": 0.0008559303008951247, "clip_ratio/high_mean": 0.0002405242557870224, "clip_ratio/low_mean": 1.891074061859399e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00025943499349523336, "completions/clipped_ratio": 0.0, "completions/max_length": 195.0, "completions/mean_length": 78.892578125, "completions/min_length": 36.5, "epoch": 0.7026627218934911, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.00898878276348114, "learning_rate": 9.77758929463583e-07, "loss": 0.00016787583008408545, "reward": 1.585306704044342, "reward_std": 0.21339760720729828, "rewards/FidelityReward/mean": 0.7281883656978607, "rewards/FidelityReward/std": 0.2069738432765007, "rewards/JudgeFidelityReward/mean": 0.7161897122859955, "rewards/JudgeFidelityReward/std": 0.22955510765314102, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 950 }, { "clip_ratio/high_max": 0.0005102566443383694, "clip_ratio/high_mean": 0.00014613327221013606, "clip_ratio/low_mean": 2.120441058650613e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016733768279664218, "completions/clipped_ratio": 0.0, "completions/max_length": 181.33333333333334, "completions/mean_length": 75.52734375, "completions/min_length": 33.333333333333336, "epoch": 0.7063609467455622, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.008013569470494985, "learning_rate": 9.773968030216786e-07, "loss": 0.00035582613199949267, "reward": 1.5952293872833252, "reward_std": 0.2101673980553945, "rewards/FidelityReward/mean": 0.7357050577799479, "rewards/FidelityReward/std": 0.20056078831354776, "rewards/JudgeFidelityReward/mean": 0.7190487384796143, "rewards/JudgeFidelityReward/std": 0.21034741401672363, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 955 }, { "clip_ratio/high_max": 0.0008281227434054017, "clip_ratio/high_mean": 0.00016432859702035784, "clip_ratio/low_mean": 8.593605161877349e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002502646268112585, "completions/clipped_ratio": 0.0, "completions/max_length": 179.5, "completions/mean_length": 78.02734375, "completions/min_length": 38.5, "epoch": 0.7100591715976331, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.008681278955191375, "learning_rate": 9.770318204466802e-07, "loss": 0.0003728989278897643, "reward": 1.6017508506774902, "reward_std": 0.21824830770492554, "rewards/FidelityReward/mean": 0.7379050850868225, "rewards/FidelityReward/std": 0.21245436370372772, "rewards/JudgeFidelityReward/mean": 0.7335508763790131, "rewards/JudgeFidelityReward/std": 0.21298249810934067, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.0539139099419117, "step": 960 }, { "clip_ratio/high_max": 0.0004740736563690007, "clip_ratio/high_mean": 0.00014449056907324121, "clip_ratio/low_mean": 4.557064094115049e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019006120855920018, "completions/clipped_ratio": 0.0, "completions/max_length": 175.66666666666666, "completions/mean_length": 78.51953125, "completions/min_length": 34.0, "epoch": 0.7137573964497042, "frac_reward_zero_std": 0.0, "grad_norm": 2.34375, "kl": 0.009373440034687519, "learning_rate": 9.76663983922178e-07, "loss": 0.00041317339055240153, "reward": 1.6091177860895793, "reward_std": 0.20773491760094961, "rewards/FidelityReward/mean": 0.7395511468251547, "rewards/FidelityReward/std": 0.21139161785443625, "rewards/JudgeFidelityReward/mean": 0.7404353817303976, "rewards/JudgeFidelityReward/std": 0.2051190882921219, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 965 }, { "clip_ratio/high_max": 0.00100669686216861, "clip_ratio/high_mean": 0.00029536386718973515, "clip_ratio/low_mean": 4.4130129390396176e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00033949399366974833, "completions/clipped_ratio": 0.0, "completions/max_length": 179.5, "completions/mean_length": 77.595703125, "completions/min_length": 38.0, "epoch": 0.7174556213017751, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.010892840381711722, "learning_rate": 9.762932956488355e-07, "loss": 0.0004360329825431108, "reward": 1.6416314840316772, "reward_std": 0.20573857426643372, "rewards/FidelityReward/mean": 0.7582155764102936, "rewards/FidelityReward/std": 0.20210083574056625, "rewards/JudgeFidelityReward/mean": 0.7687848210334778, "rewards/JudgeFidelityReward/std": 0.2015414461493492, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 970 }, { "clip_ratio/high_max": 0.0007491207448765636, "clip_ratio/high_mean": 0.00024197938619181515, "clip_ratio/low_mean": 4.794548149220645e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002899248735047877, "completions/clipped_ratio": 0.0, "completions/max_length": 174.33333333333334, "completions/mean_length": 77.05729166666667, "completions/min_length": 38.0, "epoch": 0.7211538461538461, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.008953224588185548, "learning_rate": 9.759197578443786e-07, "loss": 0.00027884994633495805, "reward": 1.5883160432179768, "reward_std": 0.21989237268765768, "rewards/FidelityReward/mean": 0.7260799606641134, "rewards/FidelityReward/std": 0.22099070250988007, "rewards/JudgeFidelityReward/mean": 0.7296804388364156, "rewards/JudgeFidelityReward/std": 0.21684124072392783, "rewards/SelfEvolvingFormatReward/mean": 0.9947916666666666, "rewards/SelfEvolvingFormatReward/std": 0.0710716222723325, "step": 975 }, { "clip_ratio/high_max": 0.0005553581402637064, "clip_ratio/high_mean": 0.0002171030326280743, "clip_ratio/low_mean": 3.8349645910784604e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00025545267853885887, "completions/clipped_ratio": 0.0, "completions/max_length": 186.0, "completions/mean_length": 78.2421875, "completions/min_length": 34.5, "epoch": 0.7248520710059172, "frac_reward_zero_std": 0.0, "grad_norm": 2.296875, "kl": 0.008522516768425702, "learning_rate": 9.755433727435808e-07, "loss": 0.0004718140698969364, "reward": 1.5650818943977356, "reward_std": 0.23205962777137756, "rewards/FidelityReward/mean": 0.7119318842887878, "rewards/FidelityReward/std": 0.1999569833278656, "rewards/JudgeFidelityReward/mean": 0.7062998116016388, "rewards/JudgeFidelityReward/std": 0.1954384744167328, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 980 }, { "clip_ratio/high_max": 0.0003927145968191326, "clip_ratio/high_mean": 7.145068375393748e-05, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 7.145068375393748e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 215.66666666666666, "completions/mean_length": 77.6953125, "completions/min_length": 39.666666666666664, "epoch": 0.7285502958579881, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.71875, "kl": 0.01000453382730484, "learning_rate": 9.7516414259825e-07, "loss": 0.0005656406283378601, "reward": 1.6653670867284138, "reward_std": 0.19182939330736795, "rewards/FidelityReward/mean": 0.7732207576433817, "rewards/FidelityReward/std": 0.19536882638931274, "rewards/JudgeFidelityReward/mean": 0.7842927177747091, "rewards/JudgeFidelityReward/std": 0.19288448492685953, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 985 }, { "clip_ratio/high_max": 0.0008089524693787098, "clip_ratio/high_mean": 0.00019651731709018349, "clip_ratio/low_mean": 0.0001618316862732172, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00035834899754263463, "completions/clipped_ratio": 0.0, "completions/max_length": 189.0, "completions/mean_length": 77.296875, "completions/min_length": 34.0, "epoch": 0.7322485207100592, "frac_reward_zero_std": 0.0, "grad_norm": 2.25, "kl": 0.009089435264468194, "learning_rate": 9.747820696772152e-07, "loss": 0.0003466774709522724, "reward": 1.599022090435028, "reward_std": 0.20838038623332977, "rewards/FidelityReward/mean": 0.7372816205024719, "rewards/FidelityReward/std": 0.20864883065223694, "rewards/JudgeFidelityReward/mean": 0.7234810590744019, "rewards/JudgeFidelityReward/std": 0.21532336622476578, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 990 }, { "clip_ratio/high_max": 0.0001374570420011878, "clip_ratio/high_mean": 3.374939551576972e-05, "clip_ratio/low_mean": 4.4344679918140176e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 7.809407543390989e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 202.33333333333334, "completions/mean_length": 77.28125, "completions/min_length": 38.333333333333336, "epoch": 0.7359467455621301, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.40625, "kl": 0.007228501606732607, "learning_rate": 9.74397156266313e-07, "loss": 0.00034861185122281313, "reward": 1.650397817293803, "reward_std": 0.20299596587816873, "rewards/FidelityReward/mean": 0.7671893636385599, "rewards/FidelityReward/std": 0.1991123507420222, "rewards/JudgeFidelityReward/mean": 0.7677189509073893, "rewards/JudgeFidelityReward/std": 0.19992591937383017, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 995 }, { "clip_ratio/high_max": 0.0008185602724552155, "clip_ratio/high_mean": 0.0002663846709765494, "clip_ratio/low_mean": 0.0002151670923922211, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00048155173426494, "completions/clipped_ratio": 0.0, "completions/max_length": 224.0, "completions/mean_length": 77.0546875, "completions/min_length": 38.0, "epoch": 0.7396449704142012, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.009468212909996509, "learning_rate": 9.740094046683744e-07, "loss": 0.0003698504064232111, "reward": 1.6343551874160767, "reward_std": 0.20482155680656433, "rewards/FidelityReward/mean": 0.7502174377441406, "rewards/FidelityReward/std": 0.2037307694554329, "rewards/JudgeFidelityReward/mean": 0.7702286839485168, "rewards/JudgeFidelityReward/std": 0.19347169995307922, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1000 }, { "clip_ratio/high_max": 0.0004453342058695853, "clip_ratio/high_mean": 0.00014007619756739587, "clip_ratio/low_mean": 2.129471831722185e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001613709144294262, "completions/clipped_ratio": 0.0, "completions/max_length": 173.33333333333334, "completions/mean_length": 77.06770833333333, "completions/min_length": 36.0, "epoch": 0.7433431952662722, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.008554539922624827, "learning_rate": 9.736188172032094e-07, "loss": 0.00029064123518764975, "reward": 1.5641738971074421, "reward_std": 0.21935337285200754, "rewards/FidelityReward/mean": 0.7105733752250671, "rewards/FidelityReward/std": 0.22320806980133057, "rewards/JudgeFidelityReward/mean": 0.7085030476252238, "rewards/JudgeFidelityReward/std": 0.2302527775367101, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1005 }, { "clip_ratio/high_max": 0.0006016050116159022, "clip_ratio/high_mean": 0.0001573177520185709, "clip_ratio/low_mean": 5.982445291010663e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021714219474233687, "completions/clipped_ratio": 0.0, "completions/max_length": 180.0, "completions/mean_length": 77.595703125, "completions/min_length": 37.0, "epoch": 0.7470414201183432, "frac_reward_zero_std": 0.0, "grad_norm": 2.3125, "kl": 0.009064305759966373, "learning_rate": 9.732253962075953e-07, "loss": 0.0003362689632922411, "reward": 1.6838706135749817, "reward_std": 0.18652793765068054, "rewards/FidelityReward/mean": 0.7874663174152374, "rewards/FidelityReward/std": 0.18208768218755722, "rewards/JudgeFidelityReward/mean": 0.7947616875171661, "rewards/JudgeFidelityReward/std": 0.1778022199869156, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1010 }, { "clip_ratio/high_max": 0.0004930146737024188, "clip_ratio/high_mean": 0.00016402405453845857, "clip_ratio/low_mean": 4.2156429844908414e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020618048729375004, "completions/clipped_ratio": 0.0, "completions/max_length": 204.66666666666666, "completions/mean_length": 76.28255208333333, "completions/min_length": 41.0, "epoch": 0.7507396449704142, "frac_reward_zero_std": 0.0, "grad_norm": 2.375, "kl": 0.00870313188061118, "learning_rate": 9.72829144035261e-07, "loss": 0.0003884654492139816, "reward": 1.6224027077356975, "reward_std": 0.19990955789883932, "rewards/FidelityReward/mean": 0.7466477950414022, "rewards/FidelityReward/std": 0.1987744371096293, "rewards/JudgeFidelityReward/mean": 0.7515098055203756, "rewards/JudgeFidelityReward/std": 0.2018610636393229, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1015 }, { "clip_ratio/high_max": 0.0010869663907214998, "clip_ratio/high_mean": 0.00024993119295686484, "clip_ratio/low_mean": 5.987580516375601e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030980699812062086, "completions/clipped_ratio": 0.0, "completions/max_length": 203.5, "completions/mean_length": 77.703125, "completions/min_length": 37.5, "epoch": 0.7544378698224852, "frac_reward_zero_std": 0.0, "grad_norm": 2.375, "kl": 0.009131584782153368, "learning_rate": 9.724300630568737e-07, "loss": 0.00039058884140104056, "reward": 1.6193457245826721, "reward_std": 0.21078003197908401, "rewards/FidelityReward/mean": 0.7484571933746338, "rewards/FidelityReward/std": 0.20320582389831543, "rewards/JudgeFidelityReward/mean": 0.7456833720207214, "rewards/JudgeFidelityReward/std": 0.20412077009677887, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 1020 }, { "clip_ratio/high_max": 0.00036454773508012297, "clip_ratio/high_mean": 0.00012051718076691031, "clip_ratio/low_mean": 7.518130441894755e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001956984749995172, "completions/clipped_ratio": 0.0, "completions/max_length": 188.66666666666666, "completions/mean_length": 78.03255208333333, "completions/min_length": 36.666666666666664, "epoch": 0.7581360946745562, "frac_reward_zero_std": 0.0, "grad_norm": 2.3125, "kl": 0.011609862186014652, "learning_rate": 9.720281556600245e-07, "loss": 0.0005022226832807064, "reward": 1.6365374724070232, "reward_std": 0.20416205128033957, "rewards/FidelityReward/mean": 0.7541897694269816, "rewards/FidelityReward/std": 0.20008964339892069, "rewards/JudgeFidelityReward/mean": 0.7659976681073507, "rewards/JudgeFidelityReward/std": 0.18579353392124176, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1025 }, { "clip_ratio/high_max": 0.0010585863841697575, "clip_ratio/high_mean": 0.0002359148347750306, "clip_ratio/low_mean": 6.032013043295592e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002962349681183696, "completions/clipped_ratio": 0.0, "completions/max_length": 213.0, "completions/mean_length": 78.466796875, "completions/min_length": 34.0, "epoch": 0.7618343195266272, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.010054276138544083, "learning_rate": 9.716234242492142e-07, "loss": 0.0003976610954850912, "reward": 1.5732492208480835, "reward_std": 0.21019664406776428, "rewards/FidelityReward/mean": 0.7095638513565063, "rewards/FidelityReward/std": 0.2163858488202095, "rewards/JudgeFidelityReward/mean": 0.7273708283901215, "rewards/JudgeFidelityReward/std": 0.20389269292354584, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1030 }, { "clip_ratio/high_max": 0.00036631142720580103, "clip_ratio/high_mean": 0.00014811609871685506, "clip_ratio/low_mean": 2.083333383779973e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016894943546503783, "completions/clipped_ratio": 0.0, "completions/max_length": 184.33333333333334, "completions/mean_length": 78.18489583333333, "completions/min_length": 38.0, "epoch": 0.7655325443786982, "frac_reward_zero_std": 0.0, "grad_norm": 2.328125, "kl": 0.009820639342069625, "learning_rate": 9.712158712458392e-07, "loss": 0.0003202569670975208, "reward": 1.5933515628178914, "reward_std": 0.23141779998938242, "rewards/FidelityReward/mean": 0.7305064002672831, "rewards/FidelityReward/std": 0.19487779835859934, "rewards/JudgeFidelityReward/mean": 0.7308986584345499, "rewards/JudgeFidelityReward/std": 0.1927672823270162, "rewards/SelfEvolvingFormatReward/mean": 0.9947916666666666, "rewards/SelfEvolvingFormatReward/std": 0.0710716222723325, "step": 1035 }, { "clip_ratio/high_max": 0.0006008735508657992, "clip_ratio/high_mean": 0.00014579208072973415, "clip_ratio/low_mean": 0.000120540514762979, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002663325984030962, "completions/clipped_ratio": 0.0, "completions/max_length": 210.5, "completions/mean_length": 79.845703125, "completions/min_length": 38.5, "epoch": 0.7692307692307693, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.009552907291799783, "learning_rate": 9.708054990881762e-07, "loss": 0.0005019987933337688, "reward": 1.6606552004814148, "reward_std": 0.18517953157424927, "rewards/FidelityReward/mean": 0.7706973850727081, "rewards/FidelityReward/std": 0.2035270407795906, "rewards/JudgeFidelityReward/mean": 0.7799154818058014, "rewards/JudgeFidelityReward/std": 0.19494255632162094, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1040 }, { "clip_ratio/high_max": 0.0005352712236344814, "clip_ratio/high_mean": 0.00023966669104993343, "clip_ratio/low_mean": 9.692037710919976e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00033658709144219757, "completions/clipped_ratio": 0.0, "completions/max_length": 198.33333333333334, "completions/mean_length": 78.43359375, "completions/min_length": 37.0, "epoch": 0.7729289940828402, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.009859310276806355, "learning_rate": 9.703923102313684e-07, "loss": 0.0005527365021407605, "reward": 1.631083369255066, "reward_std": 0.20088869829972586, "rewards/FidelityReward/mean": 0.7551856438318888, "rewards/FidelityReward/std": 0.19932635128498077, "rewards/JudgeFidelityReward/mean": 0.7557016412417094, "rewards/JudgeFidelityReward/std": 0.2015775442123413, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 1045 }, { "clip_ratio/high_max": 0.0008567456156015396, "clip_ratio/high_mean": 0.0002779403934255242, "clip_ratio/low_mean": 0.00010701012215577065, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003849505272228271, "completions/clipped_ratio": 0.0, "completions/max_length": 165.0, "completions/mean_length": 77.76171875, "completions/min_length": 37.5, "epoch": 0.7766272189349113, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.009588037803769111, "learning_rate": 9.699763071474106e-07, "loss": 0.0004297887906432152, "reward": 1.6642666459083557, "reward_std": 0.21810828149318695, "rewards/FidelityReward/mean": 0.7756497859954834, "rewards/FidelityReward/std": 0.19759386032819748, "rewards/JudgeFidelityReward/mean": 0.7830930948257446, "rewards/JudgeFidelityReward/std": 0.1890614852309227, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 1050 }, { "clip_ratio/high_max": 0.00033283132361248133, "clip_ratio/high_mean": 0.00011959608527831734, "clip_ratio/low_mean": 2.0542317361105234e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014013840700499713, "completions/clipped_ratio": 0.0, "completions/max_length": 233.66666666666666, "completions/mean_length": 79.6328125, "completions/min_length": 33.333333333333336, "epoch": 0.7803254437869822, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.009527378529310227, "learning_rate": 9.695574923251342e-07, "loss": 0.0004097605124115944, "reward": 1.6380883852640789, "reward_std": 0.2075702647368113, "rewards/FidelityReward/mean": 0.7576560974121094, "rewards/FidelityReward/std": 0.21556455393632254, "rewards/JudgeFidelityReward/mean": 0.7621665795644125, "rewards/JudgeFidelityReward/std": 0.22431475420792898, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1055 }, { "clip_ratio/high_max": 0.0006649306858889758, "clip_ratio/high_mean": 0.0001415434104274027, "clip_ratio/low_mean": 7.126574637368321e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021280916116666048, "completions/clipped_ratio": 0.0, "completions/max_length": 193.5, "completions/mean_length": 80.646484375, "completions/min_length": 36.5, "epoch": 0.7840236686390533, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.009235759638249874, "learning_rate": 9.691358682701926e-07, "loss": 0.00047777388244867324, "reward": 1.5854554772377014, "reward_std": 0.22503583133220673, "rewards/FidelityReward/mean": 0.7240879833698273, "rewards/FidelityReward/std": 0.2050735354423523, "rewards/JudgeFidelityReward/mean": 0.7227349281311035, "rewards/JudgeFidelityReward/std": 0.20060747861862183, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1060 }, { "clip_ratio/high_max": 0.0004395946627482772, "clip_ratio/high_mean": 0.00013913375441916288, "clip_ratio/low_mean": 2.194907865487039e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016108283889479936, "completions/clipped_ratio": 0.0, "completions/max_length": 172.33333333333334, "completions/mean_length": 77.30989583333333, "completions/min_length": 37.0, "epoch": 0.7877218934911243, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.01089598760008812, "learning_rate": 9.687114375050457e-07, "loss": 0.0004505958408117294, "reward": 1.6036615371704102, "reward_std": 0.22123293578624725, "rewards/FidelityReward/mean": 0.7353337009747823, "rewards/FidelityReward/std": 0.2008199244737625, "rewards/JudgeFidelityReward/mean": 0.7392598390579224, "rewards/JudgeFidelityReward/std": 0.19576931496461233, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 1065 }, { "clip_ratio/high_max": 0.00048701802734285593, "clip_ratio/high_mean": 0.00013777117128483952, "clip_ratio/low_mean": 6.762698176316917e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020539815304800867, "completions/clipped_ratio": 0.0, "completions/max_length": 202.5, "completions/mean_length": 79.220703125, "completions/min_length": 40.5, "epoch": 0.7914201183431953, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.011105743236839771, "learning_rate": 9.682842025689457e-07, "loss": 0.00044337818399071695, "reward": 1.6699306964874268, "reward_std": 0.17225003242492676, "rewards/FidelityReward/mean": 0.7802987098693848, "rewards/FidelityReward/std": 0.20507468283176422, "rewards/JudgeFidelityReward/mean": 0.7792638540267944, "rewards/JudgeFidelityReward/std": 0.19785819202661514, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1070 }, { "clip_ratio/high_max": 0.00041573266498744487, "clip_ratio/high_mean": 0.00018363351700827478, "clip_ratio/low_mean": 3.4833067911677065e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002184665878303349, "completions/clipped_ratio": 0.0, "completions/max_length": 178.33333333333334, "completions/mean_length": 79.92708333333333, "completions/min_length": 40.0, "epoch": 0.7951183431952663, "frac_reward_zero_std": 0.0, "grad_norm": 2.765625, "kl": 0.010053530521690845, "learning_rate": 9.67854166017921e-07, "loss": 0.00044002216309309007, "reward": 1.6225862105687459, "reward_std": 0.21144972244898477, "rewards/FidelityReward/mean": 0.7526959180831909, "rewards/FidelityReward/std": 0.19040873646736145, "rewards/JudgeFidelityReward/mean": 0.7410826881726583, "rewards/JudgeFidelityReward/std": 0.19149196644624075, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1075 }, { "clip_ratio/high_max": 0.0011325148399919272, "clip_ratio/high_mean": 0.0002867679635528475, "clip_ratio/low_mean": 5.907218874199316e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003458401595707983, "completions/clipped_ratio": 0.0, "completions/max_length": 205.0, "completions/mean_length": 78.634765625, "completions/min_length": 33.5, "epoch": 0.7988165680473372, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.011378972791135311, "learning_rate": 9.674213304247613e-07, "loss": 0.0004721035249531269, "reward": 1.6675918698310852, "reward_std": 0.18530304729938507, "rewards/FidelityReward/mean": 0.7749807238578796, "rewards/FidelityReward/std": 0.19941724091768265, "rewards/JudgeFidelityReward/mean": 0.7871754467487335, "rewards/JudgeFidelityReward/std": 0.1940082609653473, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1080 }, { "clip_ratio/high_max": 0.0004830708261579275, "clip_ratio/high_mean": 0.00014020963863004, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014020963863004, "completions/clipped_ratio": 0.0, "completions/max_length": 200.33333333333334, "completions/mean_length": 77.63541666666667, "completions/min_length": 34.0, "epoch": 0.8025147928994083, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.011766697466373443, "learning_rate": 9.669856983790026e-07, "loss": 0.0005542797036468983, "reward": 1.6242585976918538, "reward_std": 0.20168708264827728, "rewards/FidelityReward/mean": 0.747834046681722, "rewards/FidelityReward/std": 0.21223558485507965, "rewards/JudgeFidelityReward/mean": 0.7541512250900269, "rewards/JudgeFidelityReward/std": 0.2093343734741211, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1085 }, { "clip_ratio/high_max": 0.0006304252776317299, "clip_ratio/high_mean": 0.00020769108086824418, "clip_ratio/low_mean": 5.8200108469463885e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026589118351694196, "completions/clipped_ratio": 0.0, "completions/max_length": 174.0, "completions/mean_length": 79.859375, "completions/min_length": 33.5, "epoch": 0.8062130177514792, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.010780587047338485, "learning_rate": 9.665472724869107e-07, "loss": 0.0003832768183201551, "reward": 1.607650637626648, "reward_std": 0.21953477710485458, "rewards/FidelityReward/mean": 0.7378944158554077, "rewards/FidelityReward/std": 0.20173359662294388, "rewards/JudgeFidelityReward/mean": 0.7414655685424805, "rewards/JudgeFidelityReward/std": 0.20754065364599228, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1090 }, { "clip_ratio/high_max": 0.0006988591281697154, "clip_ratio/high_mean": 0.00022474629804491997, "clip_ratio/low_mean": 1.891074061859399e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002436570357531309, "completions/clipped_ratio": 0.0, "completions/max_length": 186.66666666666666, "completions/mean_length": 77.92057291666667, "completions/min_length": 35.666666666666664, "epoch": 0.8099112426035503, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.012319842912256717, "learning_rate": 9.661060553714663e-07, "loss": 0.0005772409960627556, "reward": 1.6158606211344402, "reward_std": 0.21181363860766092, "rewards/FidelityReward/mean": 0.7397807240486145, "rewards/FidelityReward/std": 0.1992578705151876, "rewards/JudgeFidelityReward/mean": 0.7547639608383179, "rewards/JudgeFidelityReward/std": 0.190219908952713, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 1095 }, { "clip_ratio/high_max": 0.0005461061373353004, "clip_ratio/high_mean": 0.00018575034628156572, "clip_ratio/low_mean": 3.6934895615559074e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022268525208346546, "completions/clipped_ratio": 0.0, "completions/max_length": 160.5, "completions/mean_length": 77.640625, "completions/min_length": 32.0, "epoch": 0.8136094674556213, "frac_reward_zero_std": 0.0, "grad_norm": 2.84375, "kl": 0.012225382402539254, "learning_rate": 9.656620496723502e-07, "loss": 0.0005410414654761553, "reward": 1.638455092906952, "reward_std": 0.22349519282579422, "rewards/FidelityReward/mean": 0.7581749260425568, "rewards/FidelityReward/std": 0.1918976753950119, "rewards/JudgeFidelityReward/mean": 0.7625135779380798, "rewards/JudgeFidelityReward/std": 0.18422766774892807, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1100 }, { "clip_ratio/high_max": 0.0003460249048657715, "clip_ratio/high_mean": 0.0001173000200651586, "clip_ratio/low_mean": 6.431772198993713e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018161774496547878, "completions/clipped_ratio": 0.0, "completions/max_length": 182.66666666666666, "completions/mean_length": 79.6484375, "completions/min_length": 43.333333333333336, "epoch": 0.8173076923076923, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.013928957656025887, "learning_rate": 9.65215258045925e-07, "loss": 0.0006236816756427288, "reward": 1.601313829421997, "reward_std": 0.19793181121349335, "rewards/FidelityReward/mean": 0.7310511469841003, "rewards/FidelityReward/std": 0.22262313961982727, "rewards/JudgeFidelityReward/mean": 0.7418273289998373, "rewards/JudgeFidelityReward/std": 0.21785361568133035, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1105 }, { "clip_ratio/high_max": 0.0007555111194960773, "clip_ratio/high_mean": 0.00011393967288313434, "clip_ratio/low_mean": 1.9500780035741626e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013344045000849292, "completions/clipped_ratio": 0.0, "completions/max_length": 167.0, "completions/mean_length": 75.755859375, "completions/min_length": 33.5, "epoch": 0.8210059171597633, "frac_reward_zero_std": 0.03125, "grad_norm": 2.59375, "kl": 0.011705915816128254, "learning_rate": 9.647656831652218e-07, "loss": 0.0004130033776164055, "reward": 1.6054880619049072, "reward_std": 0.21156155318021774, "rewards/FidelityReward/mean": 0.7355315983295441, "rewards/FidelityReward/std": 0.20320165157318115, "rewards/JudgeFidelityReward/mean": 0.7399129569530487, "rewards/JudgeFidelityReward/std": 0.1834152787923813, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1110 }, { "clip_ratio/high_max": 0.0006725495681166649, "clip_ratio/high_mean": 0.00019732668297365308, "clip_ratio/low_mean": 3.140345652354881e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002287301467731595, "completions/clipped_ratio": 0.0, "completions/max_length": 185.0, "completions/mean_length": 79.44010416666667, "completions/min_length": 37.0, "epoch": 0.8247041420118343, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.011222723312675953, "learning_rate": 9.64313327719923e-07, "loss": 0.00029853759333491324, "reward": 1.5907773971557617, "reward_std": 0.22021228075027466, "rewards/FidelityReward/mean": 0.72785617907842, "rewards/FidelityReward/std": 0.21347029010454813, "rewards/JudgeFidelityReward/mean": 0.7284466822942098, "rewards/JudgeFidelityReward/std": 0.2148198038339615, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 1115 }, { "clip_ratio/high_max": 0.0007008432759903372, "clip_ratio/high_mean": 0.00016139020153786986, "clip_ratio/low_mean": 9.166562958853319e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002530558267608285, "completions/clipped_ratio": 0.0, "completions/max_length": 175.5, "completions/mean_length": 81.103515625, "completions/min_length": 41.0, "epoch": 0.8284023668639053, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.011571925878524781, "learning_rate": 9.638581944163457e-07, "loss": 0.0006145324092358351, "reward": 1.5937642455101013, "reward_std": 0.22520142048597336, "rewards/FidelityReward/mean": 0.73039311170578, "rewards/FidelityReward/std": 0.20133920013904572, "rewards/JudgeFidelityReward/mean": 0.7267422676086426, "rewards/JudgeFidelityReward/std": 0.20958036184310913, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1120 }, { "clip_ratio/high_max": 0.0005197519436478615, "clip_ratio/high_mean": 0.00011996887624263764, "clip_ratio/low_mean": 6.500154559034854e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018497042474336923, "completions/clipped_ratio": 0.0, "completions/max_length": 193.66666666666666, "completions/mean_length": 79.10286458333333, "completions/min_length": 38.0, "epoch": 0.8321005917159763, "frac_reward_zero_std": 0.0, "grad_norm": 2.3125, "kl": 0.013109811022877694, "learning_rate": 9.63400285977427e-07, "loss": 0.000679746549576521, "reward": 1.6205979983011882, "reward_std": 0.21357068419456482, "rewards/FidelityReward/mean": 0.7426279584566752, "rewards/FidelityReward/std": 0.20591898759206137, "rewards/JudgeFidelityReward/mean": 0.7572421630223592, "rewards/JudgeFidelityReward/std": 0.18757817645867667, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1125 }, { "clip_ratio/high_max": 0.0006661857012659311, "clip_ratio/high_mean": 0.00012263882963452487, "clip_ratio/low_mean": 9.262727980967611e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021526611235458403, "completions/clipped_ratio": 0.0, "completions/max_length": 242.5, "completions/mean_length": 80.34765625, "completions/min_length": 36.5, "epoch": 0.8357988165680473, "frac_reward_zero_std": 0.015625, "grad_norm": 2.5, "kl": 0.01380634419620037, "learning_rate": 9.629396051427063e-07, "loss": 0.0006100868340581656, "reward": 1.6370723247528076, "reward_std": 0.20396429300308228, "rewards/FidelityReward/mean": 0.759905219078064, "rewards/FidelityReward/std": 0.20537889003753662, "rewards/JudgeFidelityReward/mean": 0.7543342411518097, "rewards/JudgeFidelityReward/std": 0.21259039640426636, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1130 }, { "clip_ratio/high_max": 0.0003276135656051338, "clip_ratio/high_mean": 0.00013528880081139504, "clip_ratio/low_mean": 1.9098549091722817e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001543873513583094, "completions/clipped_ratio": 0.0, "completions/max_length": 178.33333333333334, "completions/mean_length": 78.22916666666667, "completions/min_length": 36.333333333333336, "epoch": 0.8394970414201184, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.014655843563377857, "learning_rate": 9.624761546683098e-07, "loss": 0.000488235242664814, "reward": 1.6269928216934204, "reward_std": 0.19792068501313528, "rewards/FidelityReward/mean": 0.7475328048070272, "rewards/FidelityReward/std": 0.21077700952688852, "rewards/JudgeFidelityReward/mean": 0.7589200139045715, "rewards/JudgeFidelityReward/std": 0.209607794880867, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1135 }, { "clip_ratio/high_max": 0.0007054575369693338, "clip_ratio/high_mean": 0.00017427105631213636, "clip_ratio/low_mean": 0.00014389032730832696, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003181613807100803, "completions/clipped_ratio": 0.0, "completions/max_length": 229.0, "completions/mean_length": 81.62109375, "completions/min_length": 34.5, "epoch": 0.8431952662721893, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.011565316654741763, "learning_rate": 9.620099373269335e-07, "loss": 0.0005775715690106153, "reward": 1.5925260186195374, "reward_std": 0.21617401391267776, "rewards/FidelityReward/mean": 0.7192577719688416, "rewards/FidelityReward/std": 0.23588541895151138, "rewards/JudgeFidelityReward/mean": 0.746536374092102, "rewards/JudgeFidelityReward/std": 0.21875427663326263, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1140 }, { "clip_ratio/high_max": 0.00043530333787202837, "clip_ratio/high_mean": 0.0001292623288463801, "clip_ratio/low_mean": 6.55554118566215e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019481773488223554, "completions/clipped_ratio": 0.0, "completions/max_length": 193.33333333333334, "completions/mean_length": 78.69921875, "completions/min_length": 36.666666666666664, "epoch": 0.8468934911242604, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.01212538480758667, "learning_rate": 9.61540955907827e-07, "loss": 0.0006857961416244507, "reward": 1.6052014032999675, "reward_std": 0.2052524338165919, "rewards/FidelityReward/mean": 0.7416739662488302, "rewards/FidelityReward/std": 0.20663982133070627, "rewards/JudgeFidelityReward/mean": 0.7296589215596517, "rewards/JudgeFidelityReward/std": 0.21576407551765442, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 1145 }, { "clip_ratio/high_max": 0.001017580251209438, "clip_ratio/high_mean": 0.00031665271962992846, "clip_ratio/low_mean": 0.0001676367173786275, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004842894151806831, "completions/clipped_ratio": 0.0, "completions/max_length": 230.0, "completions/mean_length": 80.185546875, "completions/min_length": 36.0, "epoch": 0.8505917159763313, "frac_reward_zero_std": 0.0, "grad_norm": 2.5625, "kl": 0.012410430051386357, "learning_rate": 9.610692132167763e-07, "loss": 0.0005339579656720161, "reward": 1.6638186573982239, "reward_std": 0.20176265388727188, "rewards/FidelityReward/mean": 0.7737216055393219, "rewards/FidelityReward/std": 0.18198031187057495, "rewards/JudgeFidelityReward/mean": 0.7801940739154816, "rewards/JudgeFidelityReward/std": 0.18434574455022812, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1150 }, { "clip_ratio/high_max": 0.00032989503815770147, "clip_ratio/high_mean": 8.261119073722511e-05, "clip_ratio/low_mean": 1.7680339806247502e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010029153199866414, "completions/clipped_ratio": 0.0, "completions/max_length": 195.0, "completions/mean_length": 76.51432291666667, "completions/min_length": 34.0, "epoch": 0.8542899408284024, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.011893623881042004, "learning_rate": 9.605947120760877e-07, "loss": 0.00052790273912251, "reward": 1.6157160997390747, "reward_std": 0.20683405299981436, "rewards/FidelityReward/mean": 0.7420158386230469, "rewards/FidelityReward/std": 0.1912198563416799, "rewards/JudgeFidelityReward/mean": 0.7474004228909811, "rewards/JudgeFidelityReward/std": 0.175152654449145, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1155 }, { "clip_ratio/high_max": 0.0008605890674516558, "clip_ratio/high_mean": 0.00024207128444686532, "clip_ratio/low_mean": 6.784175348002463e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030991304665803907, "completions/clipped_ratio": 0.0, "completions/max_length": 179.0, "completions/mean_length": 75.970703125, "completions/min_length": 40.0, "epoch": 0.8579881656804734, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.011856587789952754, "learning_rate": 9.601174553245701e-07, "loss": 0.000570963229984045, "reward": 1.6727901697158813, "reward_std": 0.17869840562343597, "rewards/FidelityReward/mean": 0.7751807272434235, "rewards/FidelityReward/std": 0.20744597911834717, "rewards/JudgeFidelityReward/mean": 0.7971718907356262, "rewards/JudgeFidelityReward/std": 0.19194700568914413, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1160 }, { "clip_ratio/high_max": 0.00033057742984965444, "clip_ratio/high_mean": 7.921224896563217e-05, "clip_ratio/low_mean": 4.7036688192747535e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012624893424799665, "completions/clipped_ratio": 0.0, "completions/max_length": 224.33333333333334, "completions/mean_length": 78.2734375, "completions/min_length": 37.666666666666664, "epoch": 0.8616863905325444, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.01344277746975422, "learning_rate": 9.59637445817519e-07, "loss": 0.00036501777358353136, "reward": 1.6134258111317952, "reward_std": 0.21394653618335724, "rewards/FidelityReward/mean": 0.742904802163442, "rewards/FidelityReward/std": 0.2003862460454305, "rewards/JudgeFidelityReward/mean": 0.7423441410064697, "rewards/JudgeFidelityReward/std": 0.20229759315649667, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1165 }, { "clip_ratio/high_max": 0.0005933465552516281, "clip_ratio/high_mean": 0.00018198331235907972, "clip_ratio/low_mean": 2.1701389050576837e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020368469995446503, "completions/clipped_ratio": 0.0, "completions/max_length": 195.5, "completions/mean_length": 80.31640625, "completions/min_length": 37.5, "epoch": 0.8653846153846154, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.012237096764147281, "learning_rate": 9.591546864266982e-07, "loss": 0.0004595885518938303, "reward": 1.6808923482894897, "reward_std": 0.19526834785938263, "rewards/FidelityReward/mean": 0.7903592586517334, "rewards/FidelityReward/std": 0.18913131952285767, "rewards/JudgeFidelityReward/mean": 0.7830193638801575, "rewards/JudgeFidelityReward/std": 0.18189288675785065, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1170 }, { "clip_ratio/high_max": 0.0001712328754365444, "clip_ratio/high_mean": 3.749162424355745e-05, "clip_ratio/low_mean": 3.995047154603526e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 7.744209724478424e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 181.0, "completions/mean_length": 77.86979166666667, "completions/min_length": 31.333333333333332, "epoch": 0.8690828402366864, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.014587871730327606, "learning_rate": 9.58669180040324e-07, "loss": 0.0007231147959828377, "reward": 1.6074268817901611, "reward_std": 0.1976714034875234, "rewards/FidelityReward/mean": 0.7390513817469279, "rewards/FidelityReward/std": 0.20222665866216025, "rewards/JudgeFidelityReward/mean": 0.7367509404818217, "rewards/JudgeFidelityReward/std": 0.20564968387285867, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1175 }, { "clip_ratio/high_max": 0.000809444603510201, "clip_ratio/high_mean": 0.0001411140343407169, "clip_ratio/low_mean": 3.604834491852671e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017716237634886057, "completions/clipped_ratio": 0.0, "completions/max_length": 193.0, "completions/mean_length": 76.998046875, "completions/min_length": 37.0, "epoch": 0.8727810650887574, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.013089700229465961, "learning_rate": 9.581809295630467e-07, "loss": 0.0006847543641924858, "reward": 1.595527172088623, "reward_std": 0.20883913338184357, "rewards/FidelityReward/mean": 0.7302166223526001, "rewards/FidelityReward/std": 0.2027781382203102, "rewards/JudgeFidelityReward/mean": 0.7306212782859802, "rewards/JudgeFidelityReward/std": 0.20420507341623306, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1180 }, { "clip_ratio/high_max": 0.000782976788468659, "clip_ratio/high_mean": 0.000296578707639128, "clip_ratio/low_mean": 0.0001017032191157341, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003982819034717977, "completions/clipped_ratio": 0.0, "completions/max_length": 178.0, "completions/mean_length": 75.09114583333333, "completions/min_length": 31.0, "epoch": 0.8764792899408284, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.012851965986192227, "learning_rate": 9.576899379159342e-07, "loss": 0.00042505208402872087, "reward": 1.6381404399871826, "reward_std": 0.18882233401139578, "rewards/FidelityReward/mean": 0.7589071194330851, "rewards/FidelityReward/std": 0.1948928783337275, "rewards/JudgeFidelityReward/mean": 0.7597687443097433, "rewards/JudgeFidelityReward/std": 0.18915425737698874, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1185 }, { "clip_ratio/high_max": 0.0010633740923367441, "clip_ratio/high_mean": 0.00020753832650370895, "clip_ratio/low_mean": 5.0978624494746326e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002585169568192214, "completions/clipped_ratio": 0.0, "completions/max_length": 177.0, "completions/mean_length": 81.48046875, "completions/min_length": 34.5, "epoch": 0.8801775147928994, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.011220150999724865, "learning_rate": 9.571962080364535e-07, "loss": 0.0003901996649801731, "reward": 1.646327793598175, "reward_std": 0.20378625392913818, "rewards/FidelityReward/mean": 0.760476678609848, "rewards/FidelityReward/std": 0.1902535930275917, "rewards/JudgeFidelityReward/mean": 0.7717022597789764, "rewards/JudgeFidelityReward/std": 0.18691984564065933, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1190 }, { "clip_ratio/high_max": 0.00018570101819932462, "clip_ratio/high_mean": 6.15402008406818e-05, "clip_ratio/low_mean": 3.86376865208149e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010017788736149669, "completions/clipped_ratio": 0.0, "completions/max_length": 204.0, "completions/mean_length": 79.61197916666667, "completions/min_length": 39.0, "epoch": 0.8838757396449705, "frac_reward_zero_std": 0.0, "grad_norm": 2.6875, "kl": 0.010744846798479557, "learning_rate": 9.566997428784538e-07, "loss": 0.0004990978632122278, "reward": 1.6250099341074626, "reward_std": 0.20971029500166574, "rewards/FidelityReward/mean": 0.7446881930033366, "rewards/FidelityReward/std": 0.2059275209903717, "rewards/JudgeFidelityReward/mean": 0.7619455456733704, "rewards/JudgeFidelityReward/std": 0.1967319796482722, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1195 }, { "clip_ratio/high_max": 0.0006474572350271046, "clip_ratio/high_mean": 0.00020908747683279216, "clip_ratio/low_mean": 1.6371971287298948e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022545944666489958, "completions/clipped_ratio": 0.0, "completions/max_length": 176.5, "completions/mean_length": 80.015625, "completions/min_length": 36.0, "epoch": 0.8875739644970414, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.01244114264845848, "learning_rate": 9.562005454121495e-07, "loss": 0.0006387928500771522, "reward": 1.591458261013031, "reward_std": 0.22125720232725143, "rewards/FidelityReward/mean": 0.7253084182739258, "rewards/FidelityReward/std": 0.2041982337832451, "rewards/JudgeFidelityReward/mean": 0.7322996556758881, "rewards/JudgeFidelityReward/std": 0.20256128907203674, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1200 }, { "clip_ratio/high_max": 0.000716551928780973, "clip_ratio/high_mean": 0.0001631036226172, "clip_ratio/low_mean": 3.771101619349792e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020081463735550643, "completions/clipped_ratio": 0.0, "completions/max_length": 196.33333333333334, "completions/mean_length": 77.51692708333333, "completions/min_length": 31.666666666666668, "epoch": 0.8912721893491125, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.4375, "kl": 0.010558521747589112, "learning_rate": 9.556986186241005e-07, "loss": 0.0003555874340236187, "reward": 1.600839654604594, "reward_std": 0.21504863599936166, "rewards/FidelityReward/mean": 0.7291614611943563, "rewards/FidelityReward/std": 0.19422425826390585, "rewards/JudgeFidelityReward/mean": 0.7433563272158304, "rewards/JudgeFidelityReward/std": 0.185715322693189, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1205 }, { "clip_ratio/high_max": 0.0005065954639576376, "clip_ratio/high_mean": 0.00015871950890868902, "clip_ratio/low_mean": 7.835586438886821e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023707537329755723, "completions/clipped_ratio": 0.0, "completions/max_length": 194.5, "completions/mean_length": 79.55078125, "completions/min_length": 39.0, "epoch": 0.8949704142011834, "frac_reward_zero_std": 0.0, "grad_norm": 2.6875, "kl": 0.012339311838150024, "learning_rate": 9.551939655171964e-07, "loss": 0.0006171430926769972, "reward": 1.6002244353294373, "reward_std": 0.19354204833507538, "rewards/FidelityReward/mean": 0.7319151163101196, "rewards/FidelityReward/std": 0.24543600529432297, "rewards/JudgeFidelityReward/mean": 0.7366188764572144, "rewards/JudgeFidelityReward/std": 0.24679995328187943, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1210 }, { "clip_ratio/high_max": 0.0006532061262987555, "clip_ratio/high_mean": 0.00022618359653279186, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022618359653279186, "completions/clipped_ratio": 0.0, "completions/max_length": 193.0, "completions/mean_length": 78.83203125, "completions/min_length": 36.666666666666664, "epoch": 0.8986686390532544, "frac_reward_zero_std": 0.0, "grad_norm": 2.296875, "kl": 0.013138571195304394, "learning_rate": 9.546865891106368e-07, "loss": 0.00044150021858513355, "reward": 1.6267890532811482, "reward_std": 0.2039859096209208, "rewards/FidelityReward/mean": 0.7516818046569824, "rewards/FidelityReward/std": 0.20482617616653442, "rewards/JudgeFidelityReward/mean": 0.7515164613723755, "rewards/JudgeFidelityReward/std": 0.2062073896328608, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1215 }, { "clip_ratio/high_max": 0.0009110637940466404, "clip_ratio/high_mean": 0.00020595675450749696, "clip_ratio/low_mean": 0.0001324224518612027, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00033837921218946574, "completions/clipped_ratio": 0.0, "completions/max_length": 206.0, "completions/mean_length": 78.341796875, "completions/min_length": 41.5, "epoch": 0.9023668639053254, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.011515158042311669, "learning_rate": 9.541764924399148e-07, "loss": 0.0005226705223321914, "reward": 1.6136466264724731, "reward_std": 0.22062411159276962, "rewards/FidelityReward/mean": 0.7445970177650452, "rewards/FidelityReward/std": 0.18873586505651474, "rewards/JudgeFidelityReward/mean": 0.7380990982055664, "rewards/JudgeFidelityReward/std": 0.1812986209988594, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1220 }, { "clip_ratio/high_max": 0.0007031692191958427, "clip_ratio/high_mean": 0.0003805279615335166, "clip_ratio/low_mean": 8.075381774688139e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004612817778252065, "completions/clipped_ratio": 0.0, "completions/max_length": 191.0, "completions/mean_length": 79.04947916666667, "completions/min_length": 37.333333333333336, "epoch": 0.9060650887573964, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.010836250521242619, "learning_rate": 9.536636785567978e-07, "loss": 0.0003972133621573448, "reward": 1.6514689127604167, "reward_std": 0.19661730527877808, "rewards/FidelityReward/mean": 0.7684819300969442, "rewards/FidelityReward/std": 0.2001901020606359, "rewards/JudgeFidelityReward/mean": 0.767275869846344, "rewards/JudgeFidelityReward/std": 0.19143677254517874, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1225 }, { "clip_ratio/high_max": 0.0006692885071970522, "clip_ratio/high_mean": 0.00017599283019080757, "clip_ratio/low_mean": 4.5335813774727284e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022132864687591792, "completions/clipped_ratio": 0.0, "completions/max_length": 213.0, "completions/mean_length": 76.3984375, "completions/min_length": 38.0, "epoch": 0.9097633136094675, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.011401190422475338, "learning_rate": 9.531481505293091e-07, "loss": 0.00046951919794082644, "reward": 1.6708741188049316, "reward_std": 0.19169718027114868, "rewards/FidelityReward/mean": 0.7840077579021454, "rewards/FidelityReward/std": 0.18725601583719254, "rewards/JudgeFidelityReward/mean": 0.7737327218055725, "rewards/JudgeFidelityReward/std": 0.1880953386425972, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1230 }, { "clip_ratio/high_max": 0.0005316213238984347, "clip_ratio/high_mean": 0.00014571754145435988, "clip_ratio/low_mean": 0.0001374128245515749, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002831303747370839, "completions/clipped_ratio": 0.0, "completions/max_length": 184.0, "completions/mean_length": 77.89583333333333, "completions/min_length": 37.666666666666664, "epoch": 0.9134615384615384, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.010402226820588112, "learning_rate": 9.526299114417108e-07, "loss": 0.00045413882471621034, "reward": 1.5847707589467366, "reward_std": 0.20599149664243063, "rewards/FidelityReward/mean": 0.7196540633837382, "rewards/FidelityReward/std": 0.21301760276158652, "rewards/JudgeFidelityReward/mean": 0.7302334507306417, "rewards/JudgeFidelityReward/std": 0.20445510745048523, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1235 }, { "clip_ratio/high_max": 0.000562276563141495, "clip_ratio/high_mean": 0.00011747312964871526, "clip_ratio/low_mean": 0.0001197380741359666, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002372112066950649, "completions/clipped_ratio": 0.0, "completions/max_length": 198.5, "completions/mean_length": 82.138671875, "completions/min_length": 34.0, "epoch": 0.9171597633136095, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.011155499331653117, "learning_rate": 9.521089643944839e-07, "loss": 0.0005772039759904147, "reward": 1.6911103129386902, "reward_std": 0.19061611592769623, "rewards/FidelityReward/mean": 0.7942126989364624, "rewards/FidelityReward/std": 0.1809438019990921, "rewards/JudgeFidelityReward/mean": 0.7957483232021332, "rewards/JudgeFidelityReward/std": 0.17661859840154648, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1240 }, { "clip_ratio/high_max": 0.0007290342822670936, "clip_ratio/high_mean": 0.00018179875914938747, "clip_ratio/low_mean": 3.4772936487570405e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021657169563695788, "completions/clipped_ratio": 0.0013020833333333333, "completions/max_length": 257.6666666666667, "completions/mean_length": 80.52213541666667, "completions/min_length": 32.333333333333336, "epoch": 0.9208579881656804, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.011099042929708957, "learning_rate": 9.515853125043105e-07, "loss": 0.000561583274975419, "reward": 1.5839624007542927, "reward_std": 0.20424174765745798, "rewards/FidelityReward/mean": 0.7214621305465698, "rewards/FidelityReward/std": 0.23275678356488547, "rewards/JudgeFidelityReward/mean": 0.7263026833534241, "rewards/JudgeFidelityReward/std": 0.227317343155543, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1245 }, { "clip_ratio/high_max": 0.0005660896771587431, "clip_ratio/high_mean": 0.0001240409052115865, "clip_ratio/low_mean": 6.62509395624511e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019029184040846302, "completions/clipped_ratio": 0.0, "completions/max_length": 230.5, "completions/mean_length": 82.140625, "completions/min_length": 40.5, "epoch": 0.9245562130177515, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.012129761651158333, "learning_rate": 9.510589589040552e-07, "loss": 0.0004293165635317564, "reward": 1.6100021600723267, "reward_std": 0.21198586374521255, "rewards/FidelityReward/mean": 0.742171972990036, "rewards/FidelityReward/std": 0.21038663387298584, "rewards/JudgeFidelityReward/mean": 0.7356604337692261, "rewards/JudgeFidelityReward/std": 0.2137531265616417, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1250 }, { "clip_ratio/high_max": 0.00014054813655093313, "clip_ratio/high_mean": 1.756851706886664e-05, "clip_ratio/low_mean": 2.090300986310467e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 3.8471526931971314e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 219.66666666666666, "completions/mean_length": 80.37109375, "completions/min_length": 36.666666666666664, "epoch": 0.9282544378698225, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.01263466738164425, "learning_rate": 9.505299067427462e-07, "loss": 0.0006244536489248276, "reward": 1.634795904159546, "reward_std": 0.20878152549266815, "rewards/FidelityReward/mean": 0.755206286907196, "rewards/FidelityReward/std": 0.19623875121275583, "rewards/JudgeFidelityReward/mean": 0.7591793139775594, "rewards/JudgeFidelityReward/std": 0.20062693456808725, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1255 }, { "clip_ratio/high_max": 0.0007441270048730075, "clip_ratio/high_mean": 0.00015598909085383638, "clip_ratio/low_mean": 0.00010922594083240256, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026521502295508983, "completions/clipped_ratio": 0.0, "completions/max_length": 263.0, "completions/mean_length": 80.1328125, "completions/min_length": 33.5, "epoch": 0.9319526627218935, "frac_reward_zero_std": 0.0, "grad_norm": 2.34375, "kl": 0.011206376738846302, "learning_rate": 9.49998159185556e-07, "loss": 0.00043813325464725494, "reward": 1.587257206439972, "reward_std": 0.23901162296533585, "rewards/FidelityReward/mean": 0.7248006165027618, "rewards/FidelityReward/std": 0.19900984317064285, "rewards/JudgeFidelityReward/mean": 0.7288193702697754, "rewards/JudgeFidelityReward/std": 0.1976732239127159, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 1260 }, { "clip_ratio/high_max": 0.0005926464684307576, "clip_ratio/high_mean": 0.00016428104718215763, "clip_ratio/low_mean": 4.073296149726957e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020501400576904415, "completions/clipped_ratio": 0.0, "completions/max_length": 181.0, "completions/mean_length": 78.82682291666667, "completions/min_length": 40.0, "epoch": 0.9356508875739645, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.010232850722968579, "learning_rate": 9.494637194137835e-07, "loss": 0.0004257241263985634, "reward": 1.6931371688842773, "reward_std": 0.1950860470533371, "rewards/FidelityReward/mean": 0.7945032119750977, "rewards/FidelityReward/std": 0.1824034502108892, "rewards/JudgeFidelityReward/mean": 0.7972678740819296, "rewards/JudgeFidelityReward/std": 0.1767884095509847, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1265 }, { "clip_ratio/high_max": 0.000796904368326068, "clip_ratio/high_mean": 0.00024326931743416934, "clip_ratio/low_mean": 5.730943812523037e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030057874391786754, "completions/clipped_ratio": 0.0, "completions/max_length": 192.5, "completions/mean_length": 80.05078125, "completions/min_length": 38.0, "epoch": 0.9393491124260355, "frac_reward_zero_std": 0.0, "grad_norm": 2.34375, "kl": 0.010575635358691216, "learning_rate": 9.48926590624834e-07, "loss": 0.0004135397728532553, "reward": 1.6872457265853882, "reward_std": 0.19414982199668884, "rewards/FidelityReward/mean": 0.7882603704929352, "rewards/FidelityReward/std": 0.183811254799366, "rewards/JudgeFidelityReward/mean": 0.797970712184906, "rewards/JudgeFidelityReward/std": 0.17484506964683533, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1270 }, { "clip_ratio/high_max": 0.000717054889537394, "clip_ratio/high_mean": 0.0001351277664070949, "clip_ratio/low_mean": 0.0001406517461873591, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027577951550483704, "completions/clipped_ratio": 0.0, "completions/max_length": 210.66666666666666, "completions/mean_length": 79.88151041666667, "completions/min_length": 38.0, "epoch": 0.9430473372781065, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.012462022714316845, "learning_rate": 9.483867760322006e-07, "loss": 0.0004183536395430565, "reward": 1.6220405101776123, "reward_std": 0.2069851408402125, "rewards/FidelityReward/mean": 0.746822734673818, "rewards/FidelityReward/std": 0.21422402064005533, "rewards/JudgeFidelityReward/mean": 0.7517375747362772, "rewards/JudgeFidelityReward/std": 0.20589803655942282, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1275 }, { "clip_ratio/high_max": 0.0008516153669916093, "clip_ratio/high_mean": 0.00020552577043417842, "clip_ratio/low_mean": 2.1682567603420466e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002272083453135565, "completions/clipped_ratio": 0.0, "completions/max_length": 204.5, "completions/mean_length": 80.05078125, "completions/min_length": 36.5, "epoch": 0.9467455621301775, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.009709775540977716, "learning_rate": 9.478442788654446e-07, "loss": 0.0005555307492613792, "reward": 1.6593692898750305, "reward_std": 0.18518111109733582, "rewards/FidelityReward/mean": 0.7713790535926819, "rewards/FidelityReward/std": 0.21493744105100632, "rewards/JudgeFidelityReward/mean": 0.7779336273670197, "rewards/JudgeFidelityReward/std": 0.20645198971033096, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1280 }, { "clip_ratio/high_max": 0.0004732818226329982, "clip_ratio/high_mean": 0.00013020413025515153, "clip_ratio/low_mean": 8.029511373024434e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002104992454405874, "completions/clipped_ratio": 0.0, "completions/max_length": 187.33333333333334, "completions/mean_length": 80.31770833333333, "completions/min_length": 38.666666666666664, "epoch": 0.9504437869822485, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.010799428261816502, "learning_rate": 9.472991023701768e-07, "loss": 0.0005054836627095937, "reward": 1.5504003365834553, "reward_std": 0.22419247031211853, "rewards/FidelityReward/mean": 0.6943795482317606, "rewards/FidelityReward/std": 0.22419201831022897, "rewards/JudgeFidelityReward/mean": 0.712041437625885, "rewards/JudgeFidelityReward/std": 0.21511782705783844, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1285 }, { "clip_ratio/high_max": 0.0005042438860982656, "clip_ratio/high_mean": 0.0001920617651194334, "clip_ratio/low_mean": 2.1533161634579302e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002135949325747788, "completions/clipped_ratio": 0.0, "completions/max_length": 226.0, "completions/mean_length": 79.71875, "completions/min_length": 38.0, "epoch": 0.9541420118343196, "frac_reward_zero_std": 0.03125, "grad_norm": 2.3125, "kl": 0.011266677267849445, "learning_rate": 9.467512498080374e-07, "loss": 0.0006609859876334668, "reward": 1.6408249735832214, "reward_std": 0.19327755272388458, "rewards/FidelityReward/mean": 0.7594391107559204, "rewards/FidelityReward/std": 0.2218618541955948, "rewards/JudgeFidelityReward/mean": 0.7627716660499573, "rewards/JudgeFidelityReward/std": 0.22313904762268066, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1290 }, { "clip_ratio/high_max": 0.0004248978686518967, "clip_ratio/high_mean": 8.395738987019286e-05, "clip_ratio/low_mean": 0.00010983988177031279, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019379727309569715, "completions/clipped_ratio": 0.0, "completions/max_length": 206.33333333333334, "completions/mean_length": 77.26302083333333, "completions/min_length": 37.0, "epoch": 0.9578402366863905, "frac_reward_zero_std": 0.0, "grad_norm": 2.328125, "kl": 0.01164739727973938, "learning_rate": 9.462007244566768e-07, "loss": 0.000610978901386261, "reward": 1.6350076993306477, "reward_std": 0.1915903389453888, "rewards/FidelityReward/mean": 0.7525837421417236, "rewards/FidelityReward/std": 0.203555499513944, "rewards/JudgeFidelityReward/mean": 0.7648479541142782, "rewards/JudgeFidelityReward/std": 0.20269548892974854, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1295 }, { "clip_ratio/high_max": 0.0007857102435082197, "clip_ratio/high_mean": 0.00033054512459784746, "clip_ratio/low_mean": 3.9067823672667144e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003696129540912807, "completions/clipped_ratio": 0.0, "completions/max_length": 196.0, "completions/mean_length": 80.54296875, "completions/min_length": 31.5, "epoch": 0.9615384615384616, "frac_reward_zero_std": 0.0, "grad_norm": 2.6875, "kl": 0.011884600855410099, "learning_rate": 9.45647529609736e-07, "loss": 0.0003773374482989311, "reward": 1.6750684976577759, "reward_std": 0.19902583956718445, "rewards/FidelityReward/mean": 0.7810862362384796, "rewards/FidelityReward/std": 0.20967966318130493, "rewards/JudgeFidelityReward/mean": 0.7879645526409149, "rewards/JudgeFidelityReward/std": 0.2126321718096733, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1300 }, { "clip_ratio/high_max": 0.0010549214435741304, "clip_ratio/high_mean": 0.00025848202640190723, "clip_ratio/low_mean": 1.569365995237604e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002741756965406239, "completions/clipped_ratio": 0.0, "completions/max_length": 193.0, "completions/mean_length": 76.07421875, "completions/min_length": 36.666666666666664, "epoch": 0.9652366863905325, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.375, "kl": 0.009532600175589323, "learning_rate": 9.450916685768268e-07, "loss": 0.0004118870012462139, "reward": 1.5701274871826172, "reward_std": 0.19270765284697214, "rewards/FidelityReward/mean": 0.7094825704892477, "rewards/FidelityReward/std": 0.22824149330457053, "rewards/JudgeFidelityReward/mean": 0.7212897737820944, "rewards/JudgeFidelityReward/std": 0.21965555846691132, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1305 }, { "clip_ratio/high_max": 0.0009244333719834686, "clip_ratio/high_mean": 0.0002555208629928529, "clip_ratio/low_mean": 6.550974794663488e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003210306167602539, "completions/clipped_ratio": 0.0, "completions/max_length": 188.0, "completions/mean_length": 73.306640625, "completions/min_length": 40.0, "epoch": 0.9689349112426036, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.010409074276685715, "learning_rate": 9.44533144683512e-07, "loss": 0.0005644958931952715, "reward": 1.6504754424095154, "reward_std": 0.20112885534763336, "rewards/FidelityReward/mean": 0.7719895839691162, "rewards/FidelityReward/std": 0.1862299144268036, "rewards/JudgeFidelityReward/mean": 0.7608779072761536, "rewards/JudgeFidelityReward/std": 0.17358486354351044, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 1310 }, { "clip_ratio/high_max": 0.00045506219612434506, "clip_ratio/high_mean": 7.675558881601319e-05, "clip_ratio/low_mean": 4.2343644599895926e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00011909923050552606, "completions/clipped_ratio": 0.0, "completions/max_length": 193.33333333333334, "completions/mean_length": 76.44921875, "completions/min_length": 37.0, "epoch": 0.9726331360946746, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.012110533192753792, "learning_rate": 9.439719612712857e-07, "loss": 0.0005870271008461714, "reward": 1.5742223660151164, "reward_std": 0.21757287283738455, "rewards/FidelityReward/mean": 0.7152098814646403, "rewards/FidelityReward/std": 0.21476289133230844, "rewards/JudgeFidelityReward/mean": 0.7206290165583292, "rewards/JudgeFidelityReward/std": 0.19969947636127472, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 1315 }, { "clip_ratio/high_max": 0.0006296860869042575, "clip_ratio/high_mean": 0.0002021223190240562, "clip_ratio/low_mean": 6.088081572670489e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026300312601961194, "completions/clipped_ratio": 0.0, "completions/max_length": 248.0, "completions/mean_length": 80.748046875, "completions/min_length": 38.0, "epoch": 0.9763313609467456, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.010510420985519886, "learning_rate": 9.434081216975533e-07, "loss": 0.00047692041844129565, "reward": 1.621000051498413, "reward_std": 0.20314015448093414, "rewards/FidelityReward/mean": 0.7452743649482727, "rewards/FidelityReward/std": 0.19809529930353165, "rewards/JudgeFidelityReward/mean": 0.7514511346817017, "rewards/JudgeFidelityReward/std": 0.19215408712625504, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1320 }, { "clip_ratio/high_max": 0.00029075311031192543, "clip_ratio/high_mean": 6.900568259879947e-05, "clip_ratio/low_mean": 3.5064294934272766e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010406998335383832, "completions/clipped_ratio": 0.0, "completions/max_length": 196.66666666666666, "completions/mean_length": 79.56901041666667, "completions/min_length": 38.0, "epoch": 0.9800295857988166, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.01376488208770752, "learning_rate": 9.42841629335611e-07, "loss": 0.0005799322389066219, "reward": 1.6446996132532756, "reward_std": 0.20581267774105072, "rewards/FidelityReward/mean": 0.7664923270543417, "rewards/FidelityReward/std": 0.184236541390419, "rewards/JudgeFidelityReward/mean": 0.7577165762583414, "rewards/JudgeFidelityReward/std": 0.18374643723169962, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1325 }, { "clip_ratio/high_max": 0.0007242490188218653, "clip_ratio/high_mean": 0.00019555719918571413, "clip_ratio/low_mean": 1.650165067985654e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021205884986557065, "completions/clipped_ratio": 0.0, "completions/max_length": 180.0, "completions/mean_length": 80.177734375, "completions/min_length": 34.5, "epoch": 0.9837278106508875, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.011402938608080148, "learning_rate": 9.422724875746258e-07, "loss": 0.00047783446498215196, "reward": 1.6541500091552734, "reward_std": 0.20705656707286835, "rewards/FidelityReward/mean": 0.7674846053123474, "rewards/FidelityReward/std": 0.20162849873304367, "rewards/JudgeFidelityReward/mean": 0.7752840220928192, "rewards/JudgeFidelityReward/std": 0.19126535207033157, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1330 }, { "clip_ratio/high_max": 0.0005602285731583833, "clip_ratio/high_mean": 0.00017534397193230689, "clip_ratio/low_mean": 2.18150089494884e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001971589808817953, "completions/clipped_ratio": 0.0, "completions/max_length": 178.0, "completions/mean_length": 77.5859375, "completions/min_length": 36.666666666666664, "epoch": 0.9874260355029586, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.01076500304043293, "learning_rate": 9.417006998196158e-07, "loss": 0.0004494050983339548, "reward": 1.6203557252883911, "reward_std": 0.20630555848280588, "rewards/FidelityReward/mean": 0.7465486327807108, "rewards/FidelityReward/std": 0.18854954838752747, "rewards/JudgeFidelityReward/mean": 0.7476141254107157, "rewards/JudgeFidelityReward/std": 0.18889009455839792, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1335 }, { "clip_ratio/high_max": 0.0008468303014524281, "clip_ratio/high_mean": 0.00018724050023593007, "clip_ratio/low_mean": 0.00018066428165184334, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003679047687910497, "completions/clipped_ratio": 0.0, "completions/max_length": 208.0, "completions/mean_length": 77.037109375, "completions/min_length": 36.5, "epoch": 0.9911242603550295, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.010633940249681473, "learning_rate": 9.411262694914288e-07, "loss": 0.0003243857529014349, "reward": 1.5770981907844543, "reward_std": 0.20528562366962433, "rewards/FidelityReward/mean": 0.7126151919364929, "rewards/FidelityReward/std": 0.2322658970952034, "rewards/JudgeFidelityReward/mean": 0.7289660274982452, "rewards/JudgeFidelityReward/std": 0.21300657838582993, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1340 }, { "clip_ratio/high_max": 0.00017211703816428782, "clip_ratio/high_mean": 9.704207186587155e-05, "clip_ratio/low_mean": 2.0938023226335645e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001179800950922072, "completions/clipped_ratio": 0.0, "completions/max_length": 194.66666666666666, "completions/mean_length": 80.25651041666667, "completions/min_length": 38.333333333333336, "epoch": 0.9948224852071006, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.011391634307801723, "learning_rate": 9.405492000267228e-07, "loss": 0.00045345164835453033, "reward": 1.655944585800171, "reward_std": 0.18861666321754456, "rewards/FidelityReward/mean": 0.7689917882283529, "rewards/FidelityReward/std": 0.20680396258831024, "rewards/JudgeFidelityReward/mean": 0.7752076387405396, "rewards/JudgeFidelityReward/std": 0.2083392490943273, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1345 }, { "clip_ratio/high_max": 0.0008558430592529476, "clip_ratio/high_mean": 0.00025718062242958696, "clip_ratio/low_mean": 9.906445629894733e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003562450874596834, "completions/clipped_ratio": 0.0, "completions/max_length": 174.0, "completions/mean_length": 76.029296875, "completions/min_length": 36.5, "epoch": 0.9985207100591716, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.008855328802019358, "learning_rate": 9.399694948779449e-07, "loss": 0.00039323726668953894, "reward": 1.5682886242866516, "reward_std": 0.22853460162878036, "rewards/FidelityReward/mean": 0.716841071844101, "rewards/FidelityReward/std": 0.21890322118997574, "rewards/JudgeFidelityReward/mean": 0.7028951048851013, "rewards/JudgeFidelityReward/std": 0.22386857867240906, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1350 }, { "clip_ratio/high_max": 0.0008593187551014126, "clip_ratio/high_mean": 0.00012805894075427205, "clip_ratio/low_mean": 6.990135298110545e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019796029664576054, "completions/clipped_ratio": 0.0, "completions/max_length": 216.33333333333334, "completions/mean_length": 77.20052083333333, "completions/min_length": 34.0, "epoch": 1.0022189349112427, "frac_reward_zero_std": 0.0, "grad_norm": 2.359375, "kl": 0.00837989617139101, "learning_rate": 9.393871575133105e-07, "loss": 0.00027340142987668513, "reward": 1.5733687082926433, "reward_std": 0.21121729413668314, "rewards/FidelityReward/mean": 0.7152043581008911, "rewards/FidelityReward/std": 0.2159569909175237, "rewards/JudgeFidelityReward/mean": 0.7189328670501709, "rewards/JudgeFidelityReward/std": 0.21522736549377441, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 1355 }, { "clip_ratio/high_max": 0.0009580182260833681, "clip_ratio/high_mean": 0.00034050474059768023, "clip_ratio/low_mean": 7.56422188715078e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004161469521932304, "completions/clipped_ratio": 0.0, "completions/max_length": 203.5, "completions/mean_length": 81.591796875, "completions/min_length": 39.0, "epoch": 1.0059171597633136, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.008676392771303654, "learning_rate": 9.388021914167835e-07, "loss": 0.0003665180876851082, "reward": 1.6143628358840942, "reward_std": 0.22703586518764496, "rewards/FidelityReward/mean": 0.742334246635437, "rewards/FidelityReward/std": 0.2026682272553444, "rewards/JudgeFidelityReward/mean": 0.7479633688926697, "rewards/JudgeFidelityReward/std": 0.1892613172531128, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 1360 }, { "clip_ratio/high_max": 0.0004961246042512358, "clip_ratio/high_mean": 0.00015040725702419877, "clip_ratio/low_mean": 4.0197005728259685e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019060426857322454, "completions/clipped_ratio": 0.0, "completions/max_length": 181.66666666666666, "completions/mean_length": 79.27864583333333, "completions/min_length": 38.333333333333336, "epoch": 1.0096153846153846, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.009633164666593075, "learning_rate": 9.382146000880539e-07, "loss": 0.0005036786198616028, "reward": 1.599182367324829, "reward_std": 0.20588120818138123, "rewards/FidelityReward/mean": 0.729009727636973, "rewards/FidelityReward/std": 0.2093174954255422, "rewards/JudgeFidelityReward/mean": 0.740345279375712, "rewards/JudgeFidelityReward/std": 0.20926614105701447, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1365 }, { "clip_ratio/high_max": 0.0008162794401869178, "clip_ratio/high_mean": 0.00020368323603179306, "clip_ratio/low_mean": 1.9872814300470053e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022355605033226312, "completions/clipped_ratio": 0.0, "completions/max_length": 154.5, "completions/mean_length": 76.83203125, "completions/min_length": 32.0, "epoch": 1.0133136094674555, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.010395962931215762, "learning_rate": 9.376243870425187e-07, "loss": 0.0004542018286883831, "reward": 1.6742109656333923, "reward_std": 0.19706100970506668, "rewards/FidelityReward/mean": 0.7869296371936798, "rewards/FidelityReward/std": 0.19110464304685593, "rewards/JudgeFidelityReward/mean": 0.776515930891037, "rewards/JudgeFidelityReward/std": 0.19874496012926102, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1370 }, { "clip_ratio/high_max": 0.0003690065117552876, "clip_ratio/high_mean": 0.0001281720644328743, "clip_ratio/low_mean": 3.891050582751632e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016708257608115672, "completions/clipped_ratio": 0.0, "completions/max_length": 185.66666666666666, "completions/mean_length": 76.9453125, "completions/min_length": 38.666666666666664, "epoch": 1.0170118343195267, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.0104356300085783, "learning_rate": 9.370315558112587e-07, "loss": 0.00028427080251276494, "reward": 1.5938548644383748, "reward_std": 0.219345490137736, "rewards/FidelityReward/mean": 0.7260798613230387, "rewards/FidelityReward/std": 0.2026927967866262, "rewards/JudgeFidelityReward/mean": 0.7368520895640055, "rewards/JudgeFidelityReward/std": 0.1911494036515554, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1375 }, { "clip_ratio/high_max": 0.0008782829623669386, "clip_ratio/high_mean": 0.0002874941739719361, "clip_ratio/low_mean": 2.4925224715843795e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003124193928670138, "completions/clipped_ratio": 0.0, "completions/max_length": 172.0, "completions/mean_length": 79.888671875, "completions/min_length": 34.0, "epoch": 1.0207100591715976, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.010863173007965087, "learning_rate": 9.364361099410199e-07, "loss": 0.0005418007262051105, "reward": 1.6158828139305115, "reward_std": 0.21154583990573883, "rewards/FidelityReward/mean": 0.7424511313438416, "rewards/FidelityReward/std": 0.21528232842683792, "rewards/JudgeFidelityReward/mean": 0.746863454580307, "rewards/JudgeFidelityReward/std": 0.21121012419462204, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1380 }, { "clip_ratio/high_max": 0.0004509379039518535, "clip_ratio/high_mean": 0.00013445019139908255, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013445019139908255, "completions/clipped_ratio": 0.0, "completions/max_length": 196.0, "completions/mean_length": 79.25260416666667, "completions/min_length": 38.0, "epoch": 1.0244082840236686, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.01022002212703228, "learning_rate": 9.358380529941899e-07, "loss": 0.00037926770746707914, "reward": 1.6265634695688884, "reward_std": 0.20649745066960654, "rewards/FidelityReward/mean": 0.7466226617495219, "rewards/FidelityReward/std": 0.21256981790065765, "rewards/JudgeFidelityReward/mean": 0.7598816355069479, "rewards/JudgeFidelityReward/std": 0.20304476221402487, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1385 }, { "clip_ratio/high_max": 0.0007541138213127852, "clip_ratio/high_mean": 0.0001736821956001222, "clip_ratio/low_mean": 4.015216836705804e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021383435814641415, "completions/clipped_ratio": 0.0, "completions/max_length": 166.0, "completions/mean_length": 76.6875, "completions/min_length": 30.5, "epoch": 1.0281065088757397, "frac_reward_zero_std": 0.015625, "grad_norm": 2.453125, "kl": 0.009729617461562156, "learning_rate": 9.352373885487781e-07, "loss": 0.00043824552558362483, "reward": 1.6042829751968384, "reward_std": 0.21097438782453537, "rewards/FidelityReward/mean": 0.739720344543457, "rewards/FidelityReward/std": 0.2173239216208458, "rewards/JudgeFidelityReward/mean": 0.7310784459114075, "rewards/JudgeFidelityReward/std": 0.22233173996210098, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1390 }, { "clip_ratio/high_max": 0.00030540163861587645, "clip_ratio/high_mean": 5.6887781829573215e-05, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 5.6887781829573215e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 187.0, "completions/mean_length": 78.921875, "completions/min_length": 38.333333333333336, "epoch": 1.0318047337278107, "frac_reward_zero_std": 0.0, "grad_norm": 2.359375, "kl": 0.010466160252690315, "learning_rate": 9.346341201983941e-07, "loss": 0.0003130950964987278, "reward": 1.6145307620366414, "reward_std": 0.20795806248982748, "rewards/FidelityReward/mean": 0.7462732394536337, "rewards/FidelityReward/std": 0.2086512545744578, "rewards/JudgeFidelityReward/mean": 0.736514945824941, "rewards/JudgeFidelityReward/std": 0.21699786682923636, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1395 }, { "clip_ratio/high_max": 0.0004779940703883767, "clip_ratio/high_mean": 0.00010884351358981803, "clip_ratio/low_mean": 5.646069621434435e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016530421271454543, "completions/clipped_ratio": 0.0, "completions/max_length": 236.0, "completions/mean_length": 80.400390625, "completions/min_length": 35.0, "epoch": 1.0355029585798816, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.009735511615872383, "learning_rate": 9.340282515522255e-07, "loss": 0.00021022306755185126, "reward": 1.6339085102081299, "reward_std": 0.19971919804811478, "rewards/FidelityReward/mean": 0.7569637596607208, "rewards/FidelityReward/std": 0.21325379610061646, "rewards/JudgeFidelityReward/mean": 0.753889411687851, "rewards/JudgeFidelityReward/std": 0.21814222633838654, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1400 }, { "clip_ratio/high_max": 0.0006739789852872491, "clip_ratio/high_mean": 0.00025859394809231163, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00025859394809231163, "completions/clipped_ratio": 0.0, "completions/max_length": 175.33333333333334, "completions/mean_length": 78.89322916666667, "completions/min_length": 34.333333333333336, "epoch": 1.0392011834319526, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.011287438683211803, "learning_rate": 9.334197862350168e-07, "loss": 0.0005034876987338067, "reward": 1.6490294138590496, "reward_std": 0.19579386711120605, "rewards/FidelityReward/mean": 0.7624361117680868, "rewards/FidelityReward/std": 0.2014756699403127, "rewards/JudgeFidelityReward/mean": 0.7731867631276449, "rewards/JudgeFidelityReward/std": 0.20365961889425913, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1405 }, { "clip_ratio/high_max": 0.0008068056544288993, "clip_ratio/high_mean": 0.00017742635682225228, "clip_ratio/low_mean": 2.2441652254201473e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019986801198683678, "completions/clipped_ratio": 0.0, "completions/max_length": 166.5, "completions/mean_length": 77.8671875, "completions/min_length": 38.5, "epoch": 1.0428994082840237, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.010286130756139756, "learning_rate": 9.32808727887048e-07, "loss": 0.0004580470733344555, "reward": 1.5832441449165344, "reward_std": 0.20844101905822754, "rewards/FidelityReward/mean": 0.7180621922016144, "rewards/FidelityReward/std": 0.2287033051252365, "rewards/JudgeFidelityReward/mean": 0.7303638756275177, "rewards/JudgeFidelityReward/std": 0.22530358284711838, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1410 }, { "clip_ratio/high_max": 0.0005193794961087406, "clip_ratio/high_mean": 0.00018275686306878924, "clip_ratio/low_mean": 6.113385607022792e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002438907278701663, "completions/clipped_ratio": 0.0, "completions/max_length": 210.0, "completions/mean_length": 79.28776041666667, "completions/min_length": 35.666666666666664, "epoch": 1.0465976331360947, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.00963887684047222, "learning_rate": 9.321950801641123e-07, "loss": 0.0005541959777474404, "reward": 1.6188236077626545, "reward_std": 0.21836924056212106, "rewards/FidelityReward/mean": 0.7463950117429098, "rewards/FidelityReward/std": 0.2044816017150879, "rewards/JudgeFidelityReward/mean": 0.7461591958999634, "rewards/JudgeFidelityReward/std": 0.2094355970621109, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1415 }, { "clip_ratio/high_max": 0.0006833919906057417, "clip_ratio/high_mean": 0.0001558423013193533, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001558423013193533, "completions/clipped_ratio": 0.0, "completions/max_length": 176.5, "completions/mean_length": 77.212890625, "completions/min_length": 35.0, "epoch": 1.0502958579881656, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.011498426832258702, "learning_rate": 9.315788467374941e-07, "loss": 0.000369788846001029, "reward": 1.7410069704055786, "reward_std": 0.16972654312849045, "rewards/FidelityReward/mean": 0.8309895992279053, "rewards/FidelityReward/std": 0.16318205744028091, "rewards/JudgeFidelityReward/mean": 0.8200348019599915, "rewards/JudgeFidelityReward/std": 0.16971758008003235, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1420 }, { "clip_ratio/high_max": 0.00033647069940343497, "clip_ratio/high_mean": 0.00011649982479866594, "clip_ratio/low_mean": 6.28711233730428e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001793709583580494, "completions/clipped_ratio": 0.0, "completions/max_length": 216.66666666666666, "completions/mean_length": 81.57942708333333, "completions/min_length": 39.0, "epoch": 1.0539940828402368, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.359375, "kl": 0.01074195820838213, "learning_rate": 9.309600312939479e-07, "loss": 0.0004487805999815464, "reward": 1.6930029392242432, "reward_std": 0.1943898449341456, "rewards/FidelityReward/mean": 0.8005716403325399, "rewards/FidelityReward/std": 0.17511257032553354, "rewards/JudgeFidelityReward/mean": 0.7848625381787618, "rewards/JudgeFidelityReward/std": 0.19649780790011087, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1425 }, { "clip_ratio/high_max": 0.0007059472380205988, "clip_ratio/high_mean": 0.00012638627231353894, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012638627231353894, "completions/clipped_ratio": 0.0, "completions/max_length": 201.0, "completions/mean_length": 77.763671875, "completions/min_length": 37.5, "epoch": 1.0576923076923077, "frac_reward_zero_std": 0.0, "grad_norm": 2.359375, "kl": 0.01198184359818697, "learning_rate": 9.303386375356751e-07, "loss": 0.0005437294952571392, "reward": 1.5860340595245361, "reward_std": 0.2183270826935768, "rewards/FidelityReward/mean": 0.7280258536338806, "rewards/FidelityReward/std": 0.2183743268251419, "rewards/JudgeFidelityReward/mean": 0.7179696261882782, "rewards/JudgeFidelityReward/std": 0.22073908895254135, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1430 }, { "clip_ratio/high_max": 0.0010399938095360995, "clip_ratio/high_mean": 0.0003227675217203796, "clip_ratio/low_mean": 6.049686489859596e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000383264385163784, "completions/clipped_ratio": 0.0, "completions/max_length": 193.66666666666666, "completions/mean_length": 76.72005208333333, "completions/min_length": 30.0, "epoch": 1.0613905325443787, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.011159968562424183, "learning_rate": 9.297146691803031e-07, "loss": 0.00038594561628997327, "reward": 1.6074084043502808, "reward_std": 0.21013261874516806, "rewards/FidelityReward/mean": 0.7373356223106384, "rewards/FidelityReward/std": 0.2002115398645401, "rewards/JudgeFidelityReward/mean": 0.7401454846064249, "rewards/JudgeFidelityReward/std": 0.20165773232777914, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1435 }, { "clip_ratio/high_max": 0.0009068405837751925, "clip_ratio/high_mean": 0.00023047029826557265, "clip_ratio/low_mean": 4.009676049463451e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027056705293944107, "completions/clipped_ratio": 0.0, "completions/max_length": 253.0, "completions/mean_length": 79.185546875, "completions/min_length": 39.0, "epoch": 1.0650887573964498, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.011493948847055435, "learning_rate": 9.29088129960862e-07, "loss": 0.0004989662207663059, "reward": 1.6874937415122986, "reward_std": 0.18664595484733582, "rewards/FidelityReward/mean": 0.7948441505432129, "rewards/FidelityReward/std": 0.19234837591648102, "rewards/JudgeFidelityReward/mean": 0.7852992117404938, "rewards/JudgeFidelityReward/std": 0.19241926819086075, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1440 }, { "clip_ratio/high_max": 0.0007799151120707392, "clip_ratio/high_mean": 0.0002420600038021803, "clip_ratio/low_mean": 0.00011602712329477072, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003580871154554188, "completions/clipped_ratio": 0.0, "completions/max_length": 177.0, "completions/mean_length": 78.17578125, "completions/min_length": 33.666666666666664, "epoch": 1.0687869822485208, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.01085080336779356, "learning_rate": 9.284590236257625e-07, "loss": 0.0003445074195042253, "reward": 1.5774900515874226, "reward_std": 0.2243488828341166, "rewards/FidelityReward/mean": 0.7158793210983276, "rewards/FidelityReward/std": 0.210027148326238, "rewards/JudgeFidelityReward/mean": 0.7245234847068787, "rewards/JudgeFidelityReward/std": 0.20296726127465567, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1445 }, { "clip_ratio/high_max": 0.0003423265879973769, "clip_ratio/high_mean": 8.582393929827959e-05, "clip_ratio/low_mean": 0.00011138990230392664, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019721384451258926, "completions/clipped_ratio": 0.0, "completions/max_length": 186.0, "completions/mean_length": 77.6796875, "completions/min_length": 39.0, "epoch": 1.0724852071005917, "frac_reward_zero_std": 0.015625, "grad_norm": 2.328125, "kl": 0.010927890427410603, "learning_rate": 9.278273539387742e-07, "loss": 0.00041296095587313176, "reward": 1.6062234044075012, "reward_std": 0.2076529711484909, "rewards/FidelityReward/mean": 0.7375257611274719, "rewards/FidelityReward/std": 0.21843086928129196, "rewards/JudgeFidelityReward/mean": 0.7393484115600586, "rewards/JudgeFidelityReward/std": 0.21457520872354507, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1450 }, { "clip_ratio/high_max": 0.00033409048337489365, "clip_ratio/high_mean": 7.797959842719138e-05, "clip_ratio/low_mean": 0.00015030786162242293, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022828745422884822, "completions/clipped_ratio": 0.0, "completions/max_length": 191.0, "completions/mean_length": 79.3828125, "completions/min_length": 34.666666666666664, "epoch": 1.0761834319526626, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.010652532242238522, "learning_rate": 9.271931246790021e-07, "loss": 0.0005361703224480152, "reward": 1.6194275220235188, "reward_std": 0.21627042690912882, "rewards/FidelityReward/mean": 0.7451820572217306, "rewards/FidelityReward/std": 0.2150368938843409, "rewards/JudgeFidelityReward/mean": 0.7497930526733398, "rewards/JudgeFidelityReward/std": 0.2103788803021113, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1455 }, { "clip_ratio/high_max": 0.0005917356582358479, "clip_ratio/high_mean": 0.00019168612197972834, "clip_ratio/low_mean": 0.00012143017229391262, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003131162957288325, "completions/clipped_ratio": 0.0, "completions/max_length": 227.0, "completions/mean_length": 77.111328125, "completions/min_length": 38.5, "epoch": 1.0798816568047338, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.011190561018884182, "learning_rate": 9.265563396408647e-07, "loss": 0.0005212819669395685, "reward": 1.5991372466087341, "reward_std": 0.21413639187812805, "rewards/FidelityReward/mean": 0.7289441227912903, "rewards/FidelityReward/std": 0.21638955920934677, "rewards/JudgeFidelityReward/mean": 0.7403861582279205, "rewards/JudgeFidelityReward/std": 0.20768174529075623, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1460 }, { "clip_ratio/high_max": 0.0003499269369058311, "clip_ratio/high_mean": 7.913483714219182e-05, "clip_ratio/low_mean": 3.894080873578787e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00011807564587797969, "completions/clipped_ratio": 0.0, "completions/max_length": 218.33333333333334, "completions/mean_length": 76.49348958333333, "completions/min_length": 38.333333333333336, "epoch": 1.0835798816568047, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.010985168814659118, "learning_rate": 9.259170026340704e-07, "loss": 0.0005127176642417908, "reward": 1.668922980626424, "reward_std": 0.18189416329065958, "rewards/FidelityReward/mean": 0.7762221495310465, "rewards/FidelityReward/std": 0.20794308682282767, "rewards/JudgeFidelityReward/mean": 0.7854017217954, "rewards/JudgeFidelityReward/std": 0.20406084259351095, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1465 }, { "clip_ratio/high_max": 0.0003467920585535467, "clip_ratio/high_mean": 0.00013920297496952116, "clip_ratio/low_mean": 2.16637781704776e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016086675313999876, "completions/clipped_ratio": 0.0, "completions/max_length": 200.0, "completions/mean_length": 81.625, "completions/min_length": 35.0, "epoch": 1.0872781065088757, "frac_reward_zero_std": 0.0, "grad_norm": 2.25, "kl": 0.01105812769383192, "learning_rate": 9.25275117483596e-07, "loss": 0.0004918766207993031, "reward": 1.665565013885498, "reward_std": 0.1993471309542656, "rewards/FidelityReward/mean": 0.7786937654018402, "rewards/FidelityReward/std": 0.18746887892484665, "rewards/JudgeFidelityReward/mean": 0.7737423777580261, "rewards/JudgeFidelityReward/std": 0.20216058939695358, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1470 }, { "clip_ratio/high_max": 0.0004526502452790737, "clip_ratio/high_mean": 0.00016246871673502027, "clip_ratio/low_mean": 5.3267332259565595e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021573604899458588, "completions/clipped_ratio": 0.0, "completions/max_length": 189.66666666666666, "completions/mean_length": 79.33854166666667, "completions/min_length": 32.666666666666664, "epoch": 1.0909763313609468, "frac_reward_zero_std": 0.0, "grad_norm": 2.484375, "kl": 0.009918862208724021, "learning_rate": 9.246306880296629e-07, "loss": 0.0005540805868804455, "reward": 1.6769214471181233, "reward_std": 0.19504188001155853, "rewards/FidelityReward/mean": 0.7838258743286133, "rewards/FidelityReward/std": 0.19264799853165945, "rewards/JudgeFidelityReward/mean": 0.7887952923774719, "rewards/JudgeFidelityReward/std": 0.19071631133556366, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 1475 }, { "clip_ratio/high_max": 0.0006471528089605272, "clip_ratio/high_mean": 0.00023821412469260395, "clip_ratio/low_mean": 1.7494751955382525e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00025570886791683736, "completions/clipped_ratio": 0.0, "completions/max_length": 180.5, "completions/mean_length": 79.947265625, "completions/min_length": 35.5, "epoch": 1.0946745562130178, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.009658513590693474, "learning_rate": 9.239837181277144e-07, "loss": 0.00023043900728225708, "reward": 1.6444371938705444, "reward_std": 0.19940172135829926, "rewards/FidelityReward/mean": 0.7653931677341461, "rewards/FidelityReward/std": 0.2041894793510437, "rewards/JudgeFidelityReward/mean": 0.7580881118774414, "rewards/JudgeFidelityReward/std": 0.20563548803329468, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1480 }, { "clip_ratio/high_max": 0.0003295885631814599, "clip_ratio/high_mean": 8.05055839009583e-05, "clip_ratio/low_mean": 1.9516002794262022e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010002158815041185, "completions/clipped_ratio": 0.0, "completions/max_length": 200.66666666666666, "completions/mean_length": 79.44140625, "completions/min_length": 38.0, "epoch": 1.0983727810650887, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.010455913655459882, "learning_rate": 9.233342116483926e-07, "loss": 0.0004666223656386137, "reward": 1.5776632229487102, "reward_std": 0.2168955604235331, "rewards/FidelityReward/mean": 0.7192662755648295, "rewards/FidelityReward/std": 0.2166195660829544, "rewards/JudgeFidelityReward/mean": 0.7167938152949015, "rewards/JudgeFidelityReward/std": 0.2145228385925293, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1485 }, { "clip_ratio/high_max": 0.0005022097495384515, "clip_ratio/high_mean": 0.00014057729276828468, "clip_ratio/low_mean": 9.437702101422474e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023495431523770093, "completions/clipped_ratio": 0.0, "completions/max_length": 196.0, "completions/mean_length": 77.814453125, "completions/min_length": 38.5, "epoch": 1.1020710059171597, "frac_reward_zero_std": 0.015625, "grad_norm": 2.359375, "kl": 0.009607649594545364, "learning_rate": 9.226821724775152e-07, "loss": 0.0003199638798832893, "reward": 1.647623896598816, "reward_std": 0.20438174903392792, "rewards/FidelityReward/mean": 0.7623415291309357, "rewards/FidelityReward/std": 0.2031499519944191, "rewards/JudgeFidelityReward/mean": 0.772517740726471, "rewards/JudgeFidelityReward/std": 0.18806034326553345, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1490 }, { "clip_ratio/high_max": 0.0003195018623955548, "clip_ratio/high_mean": 7.135423365980386e-05, "clip_ratio/low_mean": 1.531862799311057e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 8.66728601977229e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 232.33333333333334, "completions/mean_length": 76.8515625, "completions/min_length": 39.0, "epoch": 1.1057692307692308, "frac_reward_zero_std": 0.0, "grad_norm": 2.375, "kl": 0.011745651997625828, "learning_rate": 9.220276045160522e-07, "loss": 0.0005297818221151829, "reward": 1.6240744590759277, "reward_std": 0.21437138319015503, "rewards/FidelityReward/mean": 0.7488771875699362, "rewards/FidelityReward/std": 0.20570292075475058, "rewards/JudgeFidelityReward/mean": 0.7516967058181763, "rewards/JudgeFidelityReward/std": 0.20021300514539084, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1495 }, { "clip_ratio/high_max": 0.0008774093817919493, "clip_ratio/high_mean": 0.0003575292183086276, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003575292183086276, "completions/clipped_ratio": 0.0, "completions/max_length": 200.0, "completions/mean_length": 80.439453125, "completions/min_length": 38.5, "epoch": 1.1094674556213018, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.009918173216283322, "learning_rate": 9.213705116801029e-07, "loss": 0.000513092614710331, "reward": 1.6366647481918335, "reward_std": 0.20379599928855896, "rewards/FidelityReward/mean": 0.753395140171051, "rewards/FidelityReward/std": 0.20938187837600708, "rewards/JudgeFidelityReward/mean": 0.7665390968322754, "rewards/JudgeFidelityReward/std": 0.20775803178548813, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1500 }, { "clip_ratio/high_max": 0.000734217488206923, "clip_ratio/high_mean": 0.00022616079077124595, "clip_ratio/low_mean": 2.3386342218145727e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002495471271686256, "completions/clipped_ratio": 0.0, "completions/max_length": 211.66666666666666, "completions/mean_length": 79.34895833333333, "completions/min_length": 33.333333333333336, "epoch": 1.1131656804733727, "frac_reward_zero_std": 0.0, "grad_norm": 2.375, "kl": 0.012226372584700584, "learning_rate": 9.207108979008719e-07, "loss": 0.000518200732767582, "reward": 1.6381619771321614, "reward_std": 0.19116048514842987, "rewards/FidelityReward/mean": 0.7571840882301331, "rewards/FidelityReward/std": 0.21355720857779184, "rewards/JudgeFidelityReward/mean": 0.7619557778040568, "rewards/JudgeFidelityReward/std": 0.20841529468695322, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1505 }, { "clip_ratio/high_max": 0.0006846703938208521, "clip_ratio/high_mean": 0.00020622872398234904, "clip_ratio/low_mean": 5.780612700618803e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026403485680930315, "completions/clipped_ratio": 0.0, "completions/max_length": 180.0, "completions/mean_length": 77.615234375, "completions/min_length": 40.5, "epoch": 1.1168639053254439, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.010312255285680295, "learning_rate": 9.200487671246461e-07, "loss": 0.0004477911628782749, "reward": 1.6511380076408386, "reward_std": 0.17600248008966446, "rewards/FidelityReward/mean": 0.7669146358966827, "rewards/FidelityReward/std": 0.24216794222593307, "rewards/JudgeFidelityReward/mean": 0.7684465944766998, "rewards/JudgeFidelityReward/std": 0.23541320860385895, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1510 }, { "clip_ratio/high_max": 0.0005403720424510539, "clip_ratio/high_mean": 0.00012471301743062214, "clip_ratio/low_mean": 6.78152486216277e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001925282718730159, "completions/clipped_ratio": 0.0, "completions/max_length": 166.66666666666666, "completions/mean_length": 78.765625, "completions/min_length": 36.666666666666664, "epoch": 1.1205621301775148, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.0112704586237669, "learning_rate": 9.193841233127711e-07, "loss": 0.0004921809304505587, "reward": 1.5645232200622559, "reward_std": 0.21760383745034537, "rewards/FidelityReward/mean": 0.7130761941274008, "rewards/FidelityReward/std": 0.22404403487841287, "rewards/JudgeFidelityReward/mean": 0.7028940916061401, "rewards/JudgeFidelityReward/std": 0.22711689273516336, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1515 }, { "clip_ratio/high_max": 0.00082288624253124, "clip_ratio/high_mean": 0.000298459516488947, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000298459516488947, "completions/clipped_ratio": 0.0, "completions/max_length": 181.5, "completions/mean_length": 80.525390625, "completions/min_length": 30.5, "epoch": 1.1242603550295858, "frac_reward_zero_std": 0.0, "grad_norm": 2.75, "kl": 0.010296188294887543, "learning_rate": 9.187169704416269e-07, "loss": 0.00032988712191581724, "reward": 1.6779797673225403, "reward_std": 0.18244530260562897, "rewards/FidelityReward/mean": 0.7854619920253754, "rewards/FidelityReward/std": 0.19202560931444168, "rewards/JudgeFidelityReward/mean": 0.7850354909896851, "rewards/JudgeFidelityReward/std": 0.18696504831314087, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1520 }, { "clip_ratio/high_max": 0.0006774619920179248, "clip_ratio/high_mean": 0.00016336845001205803, "clip_ratio/low_mean": 6.531776743941009e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002286862232722342, "completions/clipped_ratio": 0.0, "completions/max_length": 191.66666666666666, "completions/mean_length": 79.04036458333333, "completions/min_length": 39.666666666666664, "epoch": 1.1279585798816567, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.009529355447739363, "learning_rate": 9.180473125026047e-07, "loss": 0.00042567485943436625, "reward": 1.5752864678700764, "reward_std": 0.2150182674328486, "rewards/FidelityReward/mean": 0.720946470896403, "rewards/FidelityReward/std": 0.21052487194538116, "rewards/JudgeFidelityReward/mean": 0.7086800336837769, "rewards/JudgeFidelityReward/std": 0.21447712182998657, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1525 }, { "clip_ratio/high_max": 0.0007261346094310283, "clip_ratio/high_mean": 0.00016582223761361093, "clip_ratio/low_mean": 8.145117026288062e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024727340787649154, "completions/clipped_ratio": 0.0, "completions/max_length": 229.5, "completions/mean_length": 77.79296875, "completions/min_length": 36.0, "epoch": 1.1316568047337279, "frac_reward_zero_std": 0.0, "grad_norm": 2.359375, "kl": 0.011260558292269707, "learning_rate": 9.173751535020829e-07, "loss": 0.0003193104639649391, "reward": 1.6535980701446533, "reward_std": 0.19393599033355713, "rewards/FidelityReward/mean": 0.7702367007732391, "rewards/FidelityReward/std": 0.19916966557502747, "rewards/JudgeFidelityReward/mean": 0.766722708940506, "rewards/JudgeFidelityReward/std": 0.205599844455719, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1530 }, { "clip_ratio/high_max": 0.0004165881895460188, "clip_ratio/high_mean": 0.0001426869654096663, "clip_ratio/low_mean": 8.72622593306005e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002299492247402668, "completions/clipped_ratio": 0.0, "completions/max_length": 211.0, "completions/mean_length": 79.71223958333333, "completions/min_length": 37.333333333333336, "epoch": 1.1353550295857988, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.01166689321398735, "learning_rate": 9.16700497461403e-07, "loss": 0.0004479297436773777, "reward": 1.6514890591303508, "reward_std": 0.20898787677288055, "rewards/FidelityReward/mean": 0.7648715575536092, "rewards/FidelityReward/std": 0.2017789880434672, "rewards/JudgeFidelityReward/mean": 0.773235003153483, "rewards/JudgeFidelityReward/std": 0.19714806973934174, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1535 }, { "clip_ratio/high_max": 0.0008230508654378354, "clip_ratio/high_mean": 0.0003159662999678403, "clip_ratio/low_mean": 6.213506130734459e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003781013598199934, "completions/clipped_ratio": 0.0, "completions/max_length": 194.0, "completions/mean_length": 76.21484375, "completions/min_length": 32.0, "epoch": 1.1390532544378698, "frac_reward_zero_std": 0.0, "grad_norm": 2.375, "kl": 0.010617980547249318, "learning_rate": 9.160233484168453e-07, "loss": 0.0005867289379239083, "reward": 1.605963945388794, "reward_std": 0.20343254506587982, "rewards/FidelityReward/mean": 0.7359265387058258, "rewards/FidelityReward/std": 0.19643543660640717, "rewards/JudgeFidelityReward/mean": 0.7400748431682587, "rewards/JudgeFidelityReward/std": 0.19821679592132568, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1540 }, { "clip_ratio/high_max": 0.0006636488484218717, "clip_ratio/high_mean": 0.00014715854777023197, "clip_ratio/low_mean": 4.4127328146714716e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019128588028252124, "completions/clipped_ratio": 0.0, "completions/max_length": 193.0, "completions/mean_length": 77.54296875, "completions/min_length": 35.0, "epoch": 1.142751479289941, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.012878448516130448, "learning_rate": 9.153437104196058e-07, "loss": 0.0005235830787569285, "reward": 1.6567104657491047, "reward_std": 0.20384245614210764, "rewards/FidelityReward/mean": 0.7724646329879761, "rewards/FidelityReward/std": 0.19588313500086466, "rewards/JudgeFidelityReward/mean": 0.769793689250946, "rewards/JudgeFidelityReward/std": 0.1956271380186081, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1545 }, { "clip_ratio/high_max": 0.0008497581584379077, "clip_ratio/high_mean": 0.00019018239981960505, "clip_ratio/low_mean": 9.64688733802177e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028665128047578037, "completions/clipped_ratio": 0.0, "completions/max_length": 185.5, "completions/mean_length": 75.3125, "completions/min_length": 37.0, "epoch": 1.1464497041420119, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.012684454582631588, "learning_rate": 9.146615875357704e-07, "loss": 0.00043421564623713493, "reward": 1.6360153555870056, "reward_std": 0.20373670011758804, "rewards/FidelityReward/mean": 0.7546294033527374, "rewards/FidelityReward/std": 0.20135173946619034, "rewards/JudgeFidelityReward/mean": 0.7647250294685364, "rewards/JudgeFidelityReward/std": 0.19979380071163177, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1550 }, { "clip_ratio/high_max": 0.000354358390904963, "clip_ratio/high_mean": 7.972349703777581e-05, "clip_ratio/low_mean": 0.00010553641186561435, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018525990890339017, "completions/clipped_ratio": 0.0, "completions/max_length": 173.33333333333334, "completions/mean_length": 77.96875, "completions/min_length": 30.666666666666668, "epoch": 1.1501479289940828, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.009998014569282532, "learning_rate": 9.139769838462919e-07, "loss": 0.00044702724553644655, "reward": 1.5842972596486409, "reward_std": 0.21209674080212912, "rewards/FidelityReward/mean": 0.7244888544082642, "rewards/FidelityReward/std": 0.22846433520317078, "rewards/JudgeFidelityReward/mean": 0.7222210168838501, "rewards/JudgeFidelityReward/std": 0.22692214945952097, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 1555 }, { "clip_ratio/high_max": 0.0009923665085807443, "clip_ratio/high_mean": 0.00037924370262771846, "clip_ratio/low_mean": 7.786614733049647e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004571098485030234, "completions/clipped_ratio": 0.0, "completions/max_length": 206.5, "completions/mean_length": 76.63671875, "completions/min_length": 34.0, "epoch": 1.1538461538461537, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.011429527215659619, "learning_rate": 9.132899034469647e-07, "loss": 0.0005417619831860065, "reward": 1.6118574142456055, "reward_std": 0.20812802016735077, "rewards/FidelityReward/mean": 0.743456095457077, "rewards/FidelityReward/std": 0.20289725810289383, "rewards/JudgeFidelityReward/mean": 0.7368026375770569, "rewards/JudgeFidelityReward/std": 0.2141738161444664, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1560 }, { "clip_ratio/high_max": 0.0004397934302687645, "clip_ratio/high_mean": 0.00015553184202872218, "clip_ratio/low_mean": 9.155167499557138e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024708351120352747, "completions/clipped_ratio": 0.0, "completions/max_length": 172.0, "completions/mean_length": 74.28125, "completions/min_length": 33.0, "epoch": 1.157544378698225, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.011796044185757637, "learning_rate": 9.12600350448401e-07, "loss": 0.0004564779344946146, "reward": 1.6562871932983398, "reward_std": 0.18953103323777518, "rewards/FidelityReward/mean": 0.7722733418146769, "rewards/FidelityReward/std": 0.20073131223519644, "rewards/JudgeFidelityReward/mean": 0.7706318100293478, "rewards/JudgeFidelityReward/std": 0.1935115506251653, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 1565 }, { "clip_ratio/high_max": 0.0007842462626285851, "clip_ratio/high_mean": 0.00024216020829044284, "clip_ratio/low_mean": 8.093047217698768e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003230906673707068, "completions/clipped_ratio": 0.0, "completions/max_length": 173.5, "completions/mean_length": 75.99609375, "completions/min_length": 32.0, "epoch": 1.1612426035502958, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.012069225870072841, "learning_rate": 9.119083289760059e-07, "loss": 0.0005570575129240751, "reward": 1.62845116853714, "reward_std": 0.19893700629472733, "rewards/FidelityReward/mean": 0.7481733858585358, "rewards/FidelityReward/std": 0.22996914386749268, "rewards/JudgeFidelityReward/mean": 0.7625085413455963, "rewards/JudgeFidelityReward/std": 0.23099863529205322, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1570 }, { "clip_ratio/high_max": 0.00036003601271659134, "clip_ratio/high_mean": 7.925108075141906e-05, "clip_ratio/low_mean": 1.9968050764873624e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 9.921913151629269e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 203.66666666666666, "completions/mean_length": 76.64583333333333, "completions/min_length": 38.0, "epoch": 1.1649408284023668, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.011339006572961807, "learning_rate": 9.112138431699525e-07, "loss": 0.0004996773786842823, "reward": 1.6707740624745686, "reward_std": 0.2001529981692632, "rewards/FidelityReward/mean": 0.7817537983258566, "rewards/FidelityReward/std": 0.19436187048753104, "rewards/JudgeFidelityReward/mean": 0.7793425917625427, "rewards/JudgeFidelityReward/std": 0.19117597738901773, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1575 }, { "clip_ratio/high_max": 0.0005408114288002252, "clip_ratio/high_mean": 8.424590196227655e-05, "clip_ratio/low_mean": 5.676497239619494e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014101087290327997, "completions/clipped_ratio": 0.0, "completions/max_length": 184.5, "completions/mean_length": 73.482421875, "completions/min_length": 33.5, "epoch": 1.168639053254438, "frac_reward_zero_std": 0.0, "grad_norm": 2.765625, "kl": 0.012353547289967537, "learning_rate": 9.105168971851573e-07, "loss": 0.0005952447187155485, "reward": 1.6246181726455688, "reward_std": 0.18460246920585632, "rewards/FidelityReward/mean": 0.7536396682262421, "rewards/FidelityReward/std": 0.22683590650558472, "rewards/JudgeFidelityReward/mean": 0.7419569790363312, "rewards/JudgeFidelityReward/std": 0.2175193428993225, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1580 }, { "clip_ratio/high_max": 0.0006533000618219375, "clip_ratio/high_mean": 0.00026060541858896613, "clip_ratio/low_mean": 2.1168500825297088e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028177392669022084, "completions/clipped_ratio": 0.0, "completions/max_length": 201.0, "completions/mean_length": 80.28776041666667, "completions/min_length": 37.333333333333336, "epoch": 1.172337278106509, "frac_reward_zero_std": 0.0, "grad_norm": 2.78125, "kl": 0.01203264743089676, "learning_rate": 9.098174951912554e-07, "loss": 0.00042460886761546134, "reward": 1.6269853909810383, "reward_std": 0.2023533582687378, "rewards/FidelityReward/mean": 0.7467160820960999, "rewards/FidelityReward/std": 0.21832160651683807, "rewards/JudgeFidelityReward/mean": 0.7618408203125, "rewards/JudgeFidelityReward/std": 0.2112093468507131, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1585 }, { "clip_ratio/high_max": 0.0006680538179352879, "clip_ratio/high_mean": 0.0002067025110591203, "clip_ratio/low_mean": 0.00012071146629750728, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00032741397735662757, "completions/clipped_ratio": 0.0, "completions/max_length": 242.0, "completions/mean_length": 76.708984375, "completions/min_length": 33.0, "epoch": 1.1760355029585798, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.012090385518968105, "learning_rate": 9.091156413725757e-07, "loss": 0.00038251117803156376, "reward": 1.6521573066711426, "reward_std": 0.17902125418186188, "rewards/FidelityReward/mean": 0.7671178877353668, "rewards/FidelityReward/std": 0.20886385440826416, "rewards/JudgeFidelityReward/mean": 0.7700788080692291, "rewards/JudgeFidelityReward/std": 0.19620972126722336, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1590 }, { "clip_ratio/high_max": 0.000507272593677044, "clip_ratio/high_mean": 0.00012208274565637112, "clip_ratio/low_mean": 0.00013431668921839445, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002563994261436164, "completions/clipped_ratio": 0.0, "completions/max_length": 187.0, "completions/mean_length": 77.60546875, "completions/min_length": 36.666666666666664, "epoch": 1.179733727810651, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.01131743360310793, "learning_rate": 9.084113399281155e-07, "loss": 0.00036654607392847536, "reward": 1.6046804189682007, "reward_std": 0.20825799802939096, "rewards/FidelityReward/mean": 0.7347960472106934, "rewards/FidelityReward/std": 0.21163631478945413, "rewards/JudgeFidelityReward/mean": 0.7397685845692953, "rewards/JudgeFidelityReward/std": 0.2024806340535482, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1595 }, { "clip_ratio/high_max": 0.0008161071920767427, "clip_ratio/high_mean": 0.00024276211333926768, "clip_ratio/low_mean": 4.1125704592559485e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028388780483510343, "completions/clipped_ratio": 0.0, "completions/max_length": 165.0, "completions/mean_length": 76.990234375, "completions/min_length": 38.5, "epoch": 1.183431952662722, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.011158221773803234, "learning_rate": 9.077045950715158e-07, "loss": 0.0002548573072999716, "reward": 1.6828572154045105, "reward_std": 0.19763366878032684, "rewards/FidelityReward/mean": 0.7856391668319702, "rewards/FidelityReward/std": 0.17613396793603897, "rewards/JudgeFidelityReward/mean": 0.7944360971450806, "rewards/JudgeFidelityReward/std": 0.18363851308822632, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1600 }, { "clip_ratio/high_max": 0.000528631778433919, "clip_ratio/high_mean": 0.00012940638116560877, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012940638116560877, "completions/clipped_ratio": 0.0, "completions/max_length": 182.0, "completions/mean_length": 76.22395833333333, "completions/min_length": 35.333333333333336, "epoch": 1.1871301775147929, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.013674073852598667, "learning_rate": 9.069954110310354e-07, "loss": 0.0004896379075944423, "reward": 1.6320689519246419, "reward_std": 0.19460813701152802, "rewards/FidelityReward/mean": 0.7533124287923177, "rewards/FidelityReward/std": 0.2108361025651296, "rewards/JudgeFidelityReward/mean": 0.7575130065282186, "rewards/JudgeFidelityReward/std": 0.20504164199034372, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1605 }, { "clip_ratio/high_max": 0.0007177533814683556, "clip_ratio/high_mean": 0.0002349867398152128, "clip_ratio/low_mean": 0.00010007637029048055, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00033506309264339504, "completions/clipped_ratio": 0.0, "completions/max_length": 179.5, "completions/mean_length": 77.107421875, "completions/min_length": 39.0, "epoch": 1.1908284023668638, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.011432327888906002, "learning_rate": 9.062837920495265e-07, "loss": 0.0005215038079768419, "reward": 1.6155663132667542, "reward_std": 0.20126564800739288, "rewards/FidelityReward/mean": 0.7433484196662903, "rewards/FidelityReward/std": 0.21122558414936066, "rewards/JudgeFidelityReward/mean": 0.7444356679916382, "rewards/JudgeFidelityReward/std": 0.19351691007614136, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1610 }, { "clip_ratio/high_max": 0.0003197572892531753, "clip_ratio/high_mean": 7.63750955229625e-05, "clip_ratio/low_mean": 0.0001424695219611749, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021884461166337131, "completions/clipped_ratio": 0.0, "completions/max_length": 182.66666666666666, "completions/mean_length": 78.42838541666667, "completions/min_length": 36.0, "epoch": 1.194526627218935, "frac_reward_zero_std": 0.0, "grad_norm": 2.5625, "kl": 0.010909190960228443, "learning_rate": 9.055697423844088e-07, "loss": 0.0004643722902983427, "reward": 1.6466196378072102, "reward_std": 0.20555090407530466, "rewards/FidelityReward/mean": 0.7640581925710043, "rewards/FidelityReward/std": 0.2042317787806193, "rewards/JudgeFidelityReward/mean": 0.7664250334103903, "rewards/JudgeFidelityReward/std": 0.20273933311303458, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1615 }, { "clip_ratio/high_max": 0.00047919421922415495, "clip_ratio/high_mean": 0.00016102477675303817, "clip_ratio/low_mean": 0.00010770852095447481, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026873329188674687, "completions/clipped_ratio": 0.0, "completions/max_length": 195.5, "completions/mean_length": 77.3203125, "completions/min_length": 32.5, "epoch": 1.198224852071006, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.011877844855189324, "learning_rate": 9.048532663076441e-07, "loss": 0.0005525331478565931, "reward": 1.5561505556106567, "reward_std": 0.22329260408878326, "rewards/FidelityReward/mean": 0.7076847851276398, "rewards/FidelityReward/std": 0.22535430639982224, "rewards/JudgeFidelityReward/mean": 0.6988847851753235, "rewards/JudgeFidelityReward/std": 0.22605188190937042, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1620 }, { "clip_ratio/high_max": 0.00033114805119112135, "clip_ratio/high_mean": 0.00011497431551106274, "clip_ratio/low_mean": 4.162082332186401e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00015659513883292675, "completions/clipped_ratio": 0.0, "completions/max_length": 210.66666666666666, "completions/mean_length": 76.65885416666667, "completions/min_length": 36.333333333333336, "epoch": 1.2019230769230769, "frac_reward_zero_std": 0.0, "grad_norm": 2.6875, "kl": 0.011479324847459792, "learning_rate": 9.041343681057106e-07, "loss": 0.0005395144689828157, "reward": 1.6497822602589924, "reward_std": 0.2005574107170105, "rewards/FidelityReward/mean": 0.7676636377970377, "rewards/FidelityReward/std": 0.20034793516000113, "rewards/JudgeFidelityReward/mean": 0.766841471195221, "rewards/JudgeFidelityReward/std": 0.19881437718868256, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 1625 }, { "clip_ratio/high_max": 0.0005761392880231142, "clip_ratio/high_mean": 0.00027792586479336023, "clip_ratio/low_mean": 0.00010210249456577003, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00038002836517989633, "completions/clipped_ratio": 0.0, "completions/max_length": 166.0, "completions/mean_length": 76.0, "completions/min_length": 35.5, "epoch": 1.2056213017751478, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.010072794370353222, "learning_rate": 9.034130520795773e-07, "loss": 0.00034903292544186115, "reward": 1.5481618642807007, "reward_std": 0.21495406329631805, "rewards/FidelityReward/mean": 0.7021948099136353, "rewards/FidelityReward/std": 0.22316249459981918, "rewards/JudgeFidelityReward/mean": 0.6919341385364532, "rewards/JudgeFidelityReward/std": 0.22381294518709183, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1630 }, { "clip_ratio/high_max": 0.0006033954676240682, "clip_ratio/high_mean": 9.589945257175714e-05, "clip_ratio/low_mean": 1.611863262951374e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00011201808229088783, "completions/clipped_ratio": 0.0, "completions/max_length": 186.0, "completions/mean_length": 76.63541666666667, "completions/min_length": 38.333333333333336, "epoch": 1.209319526627219, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.53125, "kl": 0.01102360598742962, "learning_rate": 9.026893225446788e-07, "loss": 0.0004258892498910427, "reward": 1.6551697254180908, "reward_std": 0.18940888345241547, "rewards/FidelityReward/mean": 0.7721048593521118, "rewards/FidelityReward/std": 0.20456146200497946, "rewards/JudgeFidelityReward/mean": 0.7661296923955282, "rewards/JudgeFidelityReward/std": 0.2069277216990789, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1635 }, { "clip_ratio/high_max": 0.0007725649978965521, "clip_ratio/high_mean": 0.00022245249419938773, "clip_ratio/low_mean": 6.262901006266474e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002850815042620525, "completions/clipped_ratio": 0.0, "completions/max_length": 195.0, "completions/mean_length": 77.505859375, "completions/min_length": 37.0, "epoch": 1.21301775147929, "frac_reward_zero_std": 0.0, "grad_norm": 2.78125, "kl": 0.012032191641628743, "learning_rate": 9.019631838308889e-07, "loss": 0.00045929346233606336, "reward": 1.6254764795303345, "reward_std": 0.21081503480672836, "rewards/FidelityReward/mean": 0.7491981983184814, "rewards/FidelityReward/std": 0.21590184420347214, "rewards/JudgeFidelityReward/mean": 0.7525566220283508, "rewards/JudgeFidelityReward/std": 0.21022115647792816, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1640 }, { "clip_ratio/high_max": 0.0005283884704113007, "clip_ratio/high_mean": 0.00014409117866307498, "clip_ratio/low_mean": 1.9809826335404067e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001639010035432875, "completions/clipped_ratio": 0.0, "completions/max_length": 191.66666666666666, "completions/mean_length": 76.74739583333333, "completions/min_length": 36.0, "epoch": 1.2167159763313609, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.010636516846716404, "learning_rate": 9.012346402824946e-07, "loss": 0.0004802547860890627, "reward": 1.6669119199117024, "reward_std": 0.1943483849366506, "rewards/FidelityReward/mean": 0.7774563829104105, "rewards/FidelityReward/std": 0.18660788734753928, "rewards/JudgeFidelityReward/mean": 0.7789110342661539, "rewards/JudgeFidelityReward/std": 0.18783352772394815, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1645 }, { "clip_ratio/high_max": 0.0007081382209435105, "clip_ratio/high_mean": 0.0002433505607768893, "clip_ratio/low_mean": 3.8300454616546634e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028165101539343593, "completions/clipped_ratio": 0.0, "completions/max_length": 170.0, "completions/mean_length": 76.4609375, "completions/min_length": 30.5, "epoch": 1.220414201183432, "frac_reward_zero_std": 0.0, "grad_norm": 2.734375, "kl": 0.011964049749076366, "learning_rate": 9.005036962581706e-07, "loss": 0.0005957990419119596, "reward": 1.665334403514862, "reward_std": 0.19870953261852264, "rewards/FidelityReward/mean": 0.7763457298278809, "rewards/FidelityReward/std": 0.2110505923628807, "rewards/JudgeFidelityReward/mean": 0.7799305319786072, "rewards/JudgeFidelityReward/std": 0.19719677418470383, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1650 }, { "clip_ratio/high_max": 0.0006196539849042893, "clip_ratio/high_mean": 0.00015408701729029416, "clip_ratio/low_mean": 6.213827582541853e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021622529602609574, "completions/clipped_ratio": 0.0, "completions/max_length": 196.0, "completions/mean_length": 75.82942708333333, "completions/min_length": 37.0, "epoch": 1.224112426035503, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.011378764733672142, "learning_rate": 8.997703561309529e-07, "loss": 0.0004018614999949932, "reward": 1.6010711590449016, "reward_std": 0.21200660367806753, "rewards/FidelityReward/mean": 0.7323173880577087, "rewards/FidelityReward/std": 0.22057811419169107, "rewards/JudgeFidelityReward/mean": 0.7375075618426005, "rewards/JudgeFidelityReward/std": 0.21836085120836893, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1655 }, { "clip_ratio/high_max": 0.0008136708987876773, "clip_ratio/high_mean": 0.0002466289501171559, "clip_ratio/low_mean": 3.92742469557561e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028590320143848655, "completions/clipped_ratio": 0.0, "completions/max_length": 179.5, "completions/mean_length": 77.935546875, "completions/min_length": 37.0, "epoch": 1.227810650887574, "frac_reward_zero_std": 0.03125, "grad_norm": 2.328125, "kl": 0.011520356684923173, "learning_rate": 8.990346242882128e-07, "loss": 0.0003106494201347232, "reward": 1.5653774738311768, "reward_std": 0.2210647389292717, "rewards/FidelityReward/mean": 0.7063665390014648, "rewards/FidelityReward/std": 0.24409733712673187, "rewards/JudgeFidelityReward/mean": 0.7199751734733582, "rewards/JudgeFidelityReward/std": 0.23013891279697418, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1660 }, { "clip_ratio/high_max": 0.0005267778877168893, "clip_ratio/high_mean": 0.00015056055271998048, "clip_ratio/low_mean": 4.4883304508402946e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001954438630491495, "completions/clipped_ratio": 0.0, "completions/max_length": 196.33333333333334, "completions/mean_length": 75.84114583333333, "completions/min_length": 34.666666666666664, "epoch": 1.231508875739645, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.012532350420951844, "learning_rate": 8.982965051316306e-07, "loss": 0.0003692030441015959, "reward": 1.6499947706858318, "reward_std": 0.1991341213385264, "rewards/FidelityReward/mean": 0.7679489056269327, "rewards/FidelityReward/std": 0.19505242506663004, "rewards/JudgeFidelityReward/mean": 0.7640915711720785, "rewards/JudgeFidelityReward/std": 0.19841447472572327, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1665 }, { "clip_ratio/high_max": 0.0008144826162606478, "clip_ratio/high_mean": 0.0001725552137941122, "clip_ratio/low_mean": 0.00013934787566540762, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003119030850939453, "completions/clipped_ratio": 0.0, "completions/max_length": 160.0, "completions/mean_length": 73.603515625, "completions/min_length": 35.0, "epoch": 1.235207100591716, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.011668812669813632, "learning_rate": 8.97556003077169e-07, "loss": 0.0005257890559732914, "reward": 1.6569406986236572, "reward_std": 0.17785610258579254, "rewards/FidelityReward/mean": 0.7749014794826508, "rewards/FidelityReward/std": 0.19824259728193283, "rewards/JudgeFidelityReward/mean": 0.7640783488750458, "rewards/JudgeFidelityReward/std": 0.18820282816886902, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1670 }, { "clip_ratio/high_max": 0.00048611053498461845, "clip_ratio/high_mean": 0.00014165828470140694, "clip_ratio/low_mean": 7.845467480365187e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022011296823620796, "completions/clipped_ratio": 0.0, "completions/max_length": 187.0, "completions/mean_length": 74.73697916666667, "completions/min_length": 35.333333333333336, "epoch": 1.238905325443787, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.012155350111424924, "learning_rate": 8.96813122555047e-07, "loss": 0.0004396433476358652, "reward": 1.617475430170695, "reward_std": 0.19852827489376068, "rewards/FidelityReward/mean": 0.7403355836868286, "rewards/FidelityReward/std": 0.19505260388056436, "rewards/JudgeFidelityReward/mean": 0.7542797128359476, "rewards/JudgeFidelityReward/std": 0.18431465327739716, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1675 }, { "clip_ratio/high_max": 0.0005325479316525162, "clip_ratio/high_mean": 0.00010979796352330595, "clip_ratio/low_mean": 4.4016260653734206e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00015381422417704016, "completions/clipped_ratio": 0.0, "completions/max_length": 178.0, "completions/mean_length": 75.5859375, "completions/min_length": 37.0, "epoch": 1.242603550295858, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.011545946449041366, "learning_rate": 8.960678680097138e-07, "loss": 0.0004808446858078241, "reward": 1.6562057733535767, "reward_std": 0.19994357228279114, "rewards/FidelityReward/mean": 0.7692111134529114, "rewards/FidelityReward/std": 0.18820172548294067, "rewards/JudgeFidelityReward/mean": 0.7759424149990082, "rewards/JudgeFidelityReward/std": 0.18303094059228897, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1680 }, { "clip_ratio/high_max": 0.0005159071180969477, "clip_ratio/high_mean": 0.00013607883593067527, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013607883593067527, "completions/clipped_ratio": 0.0, "completions/max_length": 199.66666666666666, "completions/mean_length": 74.953125, "completions/min_length": 31.0, "epoch": 1.246301775147929, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.013311820663511753, "learning_rate": 8.953202438998209e-07, "loss": 0.0006139616947621107, "reward": 1.6670031944910686, "reward_std": 0.19542787969112396, "rewards/FidelityReward/mean": 0.7791089018185934, "rewards/FidelityReward/std": 0.2126209239164988, "rewards/JudgeFidelityReward/mean": 0.7783927321434021, "rewards/JudgeFidelityReward/std": 0.21726102630297342, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 1685 }, { "clip_ratio/high_max": 0.0005488295457325876, "clip_ratio/high_mean": 0.00015130147949093953, "clip_ratio/low_mean": 4.345636261859909e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019475784356473013, "completions/clipped_ratio": 0.0, "completions/max_length": 206.0, "completions/mean_length": 75.9296875, "completions/min_length": 40.0, "epoch": 1.25, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.011316136457026005, "learning_rate": 8.945702546981968e-07, "loss": 0.0005496310070157052, "reward": 1.5319764018058777, "reward_std": 0.24365700781345367, "rewards/FidelityReward/mean": 0.6894400119781494, "rewards/FidelityReward/std": 0.21464192867279053, "rewards/JudgeFidelityReward/mean": 0.6850727796554565, "rewards/JudgeFidelityReward/std": 0.21411725133657455, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1690 }, { "clip_ratio/high_max": 0.0004946383996866643, "clip_ratio/high_mean": 0.0001363328075967729, "clip_ratio/low_mean": 1.8011526844929903e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00015434433589689434, "completions/clipped_ratio": 0.0, "completions/max_length": 197.66666666666666, "completions/mean_length": 77.59505208333333, "completions/min_length": 36.0, "epoch": 1.253698224852071, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.0138247299939394, "learning_rate": 8.938179048918196e-07, "loss": 0.0006961111910641193, "reward": 1.6414366165796916, "reward_std": 0.1950368732213974, "rewards/FidelityReward/mean": 0.7560260693232218, "rewards/FidelityReward/std": 0.20616866648197174, "rewards/JudgeFidelityReward/mean": 0.7708210547765096, "rewards/JudgeFidelityReward/std": 0.200463796655337, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1695 }, { "clip_ratio/high_max": 0.0005856916075572371, "clip_ratio/high_mean": 0.00020601255237124861, "clip_ratio/low_mean": 9.179423504974693e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00029780679615214466, "completions/clipped_ratio": 0.0, "completions/max_length": 181.0, "completions/mean_length": 73.056640625, "completions/min_length": 35.0, "epoch": 1.2573964497041419, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.012656161934137345, "learning_rate": 8.9306319898179e-07, "loss": 0.00045686275698244574, "reward": 1.6764306426048279, "reward_std": 0.1999725177884102, "rewards/FidelityReward/mean": 0.776445060968399, "rewards/FidelityReward/std": 0.19200073927640915, "rewards/JudgeFidelityReward/mean": 0.8019243478775024, "rewards/JudgeFidelityReward/std": 0.181993767619133, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1700 }, { "clip_ratio/high_max": 0.0008183820638805628, "clip_ratio/high_mean": 0.00024111483944579958, "clip_ratio/low_mean": 6.796099187340587e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030907583422958853, "completions/clipped_ratio": 0.0, "completions/max_length": 165.66666666666666, "completions/mean_length": 71.85807291666667, "completions/min_length": 32.333333333333336, "epoch": 1.261094674556213, "frac_reward_zero_std": 0.0, "grad_norm": 2.1875, "kl": 0.014350583404302597, "learning_rate": 8.923061414833051e-07, "loss": 0.0006075300741940736, "reward": 1.6166536410649617, "reward_std": 0.19160867234071097, "rewards/FidelityReward/mean": 0.744567076365153, "rewards/FidelityReward/std": 0.2129909743865331, "rewards/JudgeFidelityReward/mean": 0.7441730896631876, "rewards/JudgeFidelityReward/std": 0.21459072828292847, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1705 }, { "clip_ratio/high_max": 0.001061035937163979, "clip_ratio/high_mean": 0.0002416791976429522, "clip_ratio/low_mean": 2.196836576331407e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026364756631664933, "completions/clipped_ratio": 0.0, "completions/max_length": 171.0, "completions/mean_length": 73.099609375, "completions/min_length": 28.5, "epoch": 1.264792899408284, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.013057572394609451, "learning_rate": 8.915467369256307e-07, "loss": 0.0004986623302102089, "reward": 1.6714815497398376, "reward_std": 0.17771834135055542, "rewards/FidelityReward/mean": 0.7826119959354401, "rewards/FidelityReward/std": 0.1998540386557579, "rewards/JudgeFidelityReward/mean": 0.7796923518180847, "rewards/JudgeFidelityReward/std": 0.19334685057401657, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1710 }, { "clip_ratio/high_max": 0.0003590897540561855, "clip_ratio/high_mean": 6.771726912120357e-05, "clip_ratio/low_mean": 2.2563176753465086e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 9.028044587466866e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 194.33333333333334, "completions/mean_length": 73.5, "completions/min_length": 33.0, "epoch": 1.268491124260355, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.012962006963789463, "learning_rate": 8.907849898520738e-07, "loss": 0.0006155267357826233, "reward": 1.5851513544718425, "reward_std": 0.20369515319665274, "rewards/FidelityReward/mean": 0.7227041125297546, "rewards/FidelityReward/std": 0.24490793546040854, "rewards/JudgeFidelityReward/mean": 0.7261966268221537, "rewards/JudgeFidelityReward/std": 0.23945016165574393, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1715 }, { "clip_ratio/high_max": 0.0007360536837950349, "clip_ratio/high_mean": 0.00028305879095569253, "clip_ratio/low_mean": 6.074832344893366e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00034380712313577535, "completions/clipped_ratio": 0.0, "completions/max_length": 160.0, "completions/mean_length": 72.34375, "completions/min_length": 33.5, "epoch": 1.272189349112426, "frac_reward_zero_std": 0.0, "grad_norm": 2.5625, "kl": 0.013106326013803482, "learning_rate": 8.900209048199571e-07, "loss": 0.0005194311030209065, "reward": 1.6993616223335266, "reward_std": 0.16683751344680786, "rewards/FidelityReward/mean": 0.7978680729866028, "rewards/FidelityReward/std": 0.179704949259758, "rewards/JudgeFidelityReward/mean": 0.8029870092868805, "rewards/JudgeFidelityReward/std": 0.17522581666707993, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1720 }, { "clip_ratio/high_max": 0.00040996044408529996, "clip_ratio/high_mean": 0.00018981329048983753, "clip_ratio/low_mean": 4.167625593254343e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023148955078795553, "completions/clipped_ratio": 0.0, "completions/max_length": 176.33333333333334, "completions/mean_length": 71.59505208333333, "completions/min_length": 34.666666666666664, "epoch": 1.275887573964497, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.012722872197628021, "learning_rate": 8.892544864005898e-07, "loss": 0.0006099048536270857, "reward": 1.6130867799123128, "reward_std": 0.19048277040322623, "rewards/FidelityReward/mean": 0.7418935894966125, "rewards/FidelityReward/std": 0.20405674477418265, "rewards/JudgeFidelityReward/mean": 0.7423862020174662, "rewards/JudgeFidelityReward/std": 0.20437756180763245, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1725 }, { "clip_ratio/high_max": 0.0006841853260993958, "clip_ratio/high_mean": 0.0001460804167436436, "clip_ratio/low_mean": 8.126496977638454e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022734538652002811, "completions/clipped_ratio": 0.0, "completions/max_length": 241.0, "completions/mean_length": 75.19921875, "completions/min_length": 33.5, "epoch": 1.279585798816568, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.011893520876765252, "learning_rate": 8.884857391792415e-07, "loss": 0.0005923070944845677, "reward": 1.54215008020401, "reward_std": 0.20612023025751114, "rewards/FidelityReward/mean": 0.697118878364563, "rewards/FidelityReward/std": 0.2502345070242882, "rewards/JudgeFidelityReward/mean": 0.6900623440742493, "rewards/JudgeFidelityReward/std": 0.24113931506872177, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1730 }, { "clip_ratio/high_max": 0.00047926303232088686, "clip_ratio/high_mean": 0.00012786961888195948, "clip_ratio/low_mean": 8.592171361669898e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002137913426849991, "completions/clipped_ratio": 0.0013020833333333333, "completions/max_length": 252.33333333333334, "completions/mean_length": 73.86067708333333, "completions/min_length": 34.0, "epoch": 1.2832840236686391, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.01391350943595171, "learning_rate": 8.877146677551143e-07, "loss": 0.000508679961785674, "reward": 1.626070261001587, "reward_std": 0.18475750585397085, "rewards/FidelityReward/mean": 0.7517422040303549, "rewards/FidelityReward/std": 0.1971315840880076, "rewards/JudgeFidelityReward/mean": 0.7499581774075826, "rewards/JudgeFidelityReward/std": 0.18137478331724802, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1735 }, { "clip_ratio/high_max": 0.0005648621008731425, "clip_ratio/high_mean": 0.00011524422152433544, "clip_ratio/low_mean": 2.1114865376148373e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013635908835567533, "completions/clipped_ratio": 0.0, "completions/max_length": 167.5, "completions/mean_length": 72.08203125, "completions/min_length": 38.0, "epoch": 1.28698224852071, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.013066188059747219, "learning_rate": 8.869412767413156e-07, "loss": 0.0007108324207365513, "reward": 1.6037395000457764, "reward_std": 0.21179387718439102, "rewards/FidelityReward/mean": 0.7357616722583771, "rewards/FidelityReward/std": 0.21704569458961487, "rewards/JudgeFidelityReward/mean": 0.735955685377121, "rewards/JudgeFidelityReward/std": 0.2099277302622795, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1740 }, { "clip_ratio/high_max": 0.0004909720853902399, "clip_ratio/high_mean": 0.00010196662624366582, "clip_ratio/low_mean": 8.680386235937476e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018877048860304057, "completions/clipped_ratio": 0.0, "completions/max_length": 189.33333333333334, "completions/mean_length": 72.7578125, "completions/min_length": 32.666666666666664, "epoch": 1.290680473372781, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.296875, "kl": 0.014257553406059742, "learning_rate": 8.861655707648296e-07, "loss": 0.0006400172598659993, "reward": 1.5497562885284424, "reward_std": 0.21409102777640024, "rewards/FidelityReward/mean": 0.6967793305714926, "rewards/FidelityReward/std": 0.22727550069491068, "rewards/JudgeFidelityReward/mean": 0.7059538761774699, "rewards/JudgeFidelityReward/std": 0.22234498957792917, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1745 }, { "clip_ratio/high_max": 0.0010260162176564335, "clip_ratio/high_mean": 0.0002727984567172825, "clip_ratio/low_mean": 2.6371306739747523e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002991697634570301, "completions/clipped_ratio": 0.0, "completions/max_length": 172.0, "completions/mean_length": 73.337890625, "completions/min_length": 33.0, "epoch": 1.2943786982248522, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.014109889790415765, "learning_rate": 8.853875544664909e-07, "loss": 0.0005259411409497261, "reward": 1.6611638069152832, "reward_std": 0.18669137358665466, "rewards/FidelityReward/mean": 0.7741282284259796, "rewards/FidelityReward/std": 0.21246716380119324, "rewards/JudgeFidelityReward/mean": 0.7740711569786072, "rewards/JudgeFidelityReward/std": 0.20543351024389267, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1750 }, { "clip_ratio/high_max": 0.0004911739844828845, "clip_ratio/high_mean": 8.106629829853773e-05, "clip_ratio/low_mean": 6.266735726967454e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014373365556821227, "completions/clipped_ratio": 0.0, "completions/max_length": 159.33333333333334, "completions/mean_length": 72.09635416666667, "completions/min_length": 31.666666666666668, "epoch": 1.2980769230769231, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.421875, "kl": 0.01236861851066351, "learning_rate": 8.846072325009562e-07, "loss": 0.0004913059994578362, "reward": 1.649309555689494, "reward_std": 0.18760095536708832, "rewards/FidelityReward/mean": 0.7638782064119974, "rewards/FidelityReward/std": 0.20120444893836975, "rewards/JudgeFidelityReward/mean": 0.7708627382914225, "rewards/JudgeFidelityReward/std": 0.18582483132680258, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1755 }, { "clip_ratio/high_max": 0.001003967470023781, "clip_ratio/high_mean": 0.0002855918253771961, "clip_ratio/low_mean": 4.688283079303801e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003324746387079358, "completions/clipped_ratio": 0.0, "completions/max_length": 184.0, "completions/mean_length": 72.865234375, "completions/min_length": 34.5, "epoch": 1.301775147928994, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.013321078754961491, "learning_rate": 8.838246095366754e-07, "loss": 0.0006356546655297279, "reward": 1.6280516386032104, "reward_std": 0.2017708122730255, "rewards/FidelityReward/mean": 0.7561379671096802, "rewards/FidelityReward/std": 0.1910535842180252, "rewards/JudgeFidelityReward/mean": 0.7438274621963501, "rewards/JudgeFidelityReward/std": 0.1918073073029518, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1760 }, { "clip_ratio/high_max": 0.00041925825644284487, "clip_ratio/high_mean": 0.00016807280480861663, "clip_ratio/low_mean": 2.4509805371053517e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001925826189108193, "completions/clipped_ratio": 0.0, "completions/max_length": 167.66666666666666, "completions/mean_length": 71.52734375, "completions/min_length": 35.0, "epoch": 1.305473372781065, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.01199237685650587, "learning_rate": 8.830396902558662e-07, "loss": 0.0005543414503335953, "reward": 1.6297990083694458, "reward_std": 0.19331076741218567, "rewards/FidelityReward/mean": 0.7517731785774231, "rewards/FidelityReward/std": 0.2053227573633194, "rewards/JudgeFidelityReward/mean": 0.7560517390569051, "rewards/JudgeFidelityReward/std": 0.19452480475107828, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1765 }, { "clip_ratio/high_max": 0.0006966059794649482, "clip_ratio/high_mean": 0.00021367068402469158, "clip_ratio/low_mean": 4.0195512701757254e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002538661938160658, "completions/clipped_ratio": 0.0, "completions/max_length": 139.0, "completions/mean_length": 73.32421875, "completions/min_length": 35.0, "epoch": 1.3091715976331362, "frac_reward_zero_std": 0.0, "grad_norm": 2.3125, "kl": 0.012597529031336308, "learning_rate": 8.822524793544831e-07, "loss": 0.00039667380042374134, "reward": 1.5582664012908936, "reward_std": 0.21912582218647003, "rewards/FidelityReward/mean": 0.7072206139564514, "rewards/FidelityReward/std": 0.24111798405647278, "rewards/JudgeFidelityReward/mean": 0.7020915448665619, "rewards/JudgeFidelityReward/std": 0.24934642016887665, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1770 }, { "clip_ratio/high_max": 0.0003948107245378196, "clip_ratio/high_mean": 0.00013632878253702074, "clip_ratio/low_mean": 2.144082391168922e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000157769609359093, "completions/clipped_ratio": 0.0, "completions/max_length": 184.0, "completions/mean_length": 72.29166666666667, "completions/min_length": 37.0, "epoch": 1.3128698224852071, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.921875, "kl": 0.013021436147391796, "learning_rate": 8.814629815421914e-07, "loss": 0.00044503305107355116, "reward": 1.5868026812871296, "reward_std": 0.21353021264076233, "rewards/FidelityReward/mean": 0.7199772993723551, "rewards/FidelityReward/std": 0.21521288653214773, "rewards/JudgeFidelityReward/mean": 0.733650823434194, "rewards/JudgeFidelityReward/std": 0.20927773912747702, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1775 }, { "clip_ratio/high_max": 0.0007036842638626694, "clip_ratio/high_mean": 0.00021445993334054947, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021445993334054947, "completions/clipped_ratio": 0.0, "completions/max_length": 158.5, "completions/mean_length": 71.40625, "completions/min_length": 28.5, "epoch": 1.316568047337278, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.011653946153819561, "learning_rate": 8.806712015423383e-07, "loss": 0.00029724016785621644, "reward": 1.6074504256248474, "reward_std": 0.2038857340812683, "rewards/FidelityReward/mean": 0.7359156012535095, "rewards/FidelityReward/std": 0.22356610745191574, "rewards/JudgeFidelityReward/mean": 0.7430696785449982, "rewards/JudgeFidelityReward/std": 0.22565722465515137, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1780 }, { "clip_ratio/high_max": 0.0007008138811215759, "clip_ratio/high_mean": 0.0001894333108793944, "clip_ratio/low_mean": 2.140410942956805e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021083742030896245, "completions/clipped_ratio": 0.0, "completions/max_length": 183.33333333333334, "completions/mean_length": 73.8046875, "completions/min_length": 36.0, "epoch": 1.320266272189349, "frac_reward_zero_std": 0.0, "grad_norm": 2.8125, "kl": 0.013185576349496842, "learning_rate": 8.798771440919248e-07, "loss": 0.0006204019300639629, "reward": 1.6469768285751343, "reward_std": 0.19779795904954275, "rewards/FidelityReward/mean": 0.7651831309000651, "rewards/FidelityReward/std": 0.2040268530448278, "rewards/JudgeFidelityReward/mean": 0.7635873158772787, "rewards/JudgeFidelityReward/std": 0.2117540786663691, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1785 }, { "clip_ratio/high_max": 0.0009986781049519777, "clip_ratio/high_mean": 0.0002488511556293815, "clip_ratio/low_mean": 8.657216967549175e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00033542332239449023, "completions/clipped_ratio": 0.0, "completions/max_length": 162.0, "completions/mean_length": 70.28125, "completions/min_length": 33.0, "epoch": 1.3239644970414202, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.012337933853268623, "learning_rate": 8.790808139415768e-07, "loss": 0.0004735375754535198, "reward": 1.5502519011497498, "reward_std": 0.23323549330234528, "rewards/FidelityReward/mean": 0.6975544989109039, "rewards/FidelityReward/std": 0.22245833277702332, "rewards/JudgeFidelityReward/mean": 0.7073478102684021, "rewards/JudgeFidelityReward/std": 0.21356019377708435, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1790 }, { "clip_ratio/high_max": 0.0007371424930170178, "clip_ratio/high_mean": 0.00017855459009297192, "clip_ratio/low_mean": 0.0001183931453851983, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002969477325677872, "completions/clipped_ratio": 0.0, "completions/max_length": 175.33333333333334, "completions/mean_length": 71.8203125, "completions/min_length": 34.666666666666664, "epoch": 1.327662721893491, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.234375, "kl": 0.01077092457562685, "learning_rate": 8.782822158555177e-07, "loss": 0.0003406478092074394, "reward": 1.597253719965617, "reward_std": 0.20482211808363596, "rewards/FidelityReward/mean": 0.7302184502283732, "rewards/FidelityReward/std": 0.21525239944458008, "rewards/JudgeFidelityReward/mean": 0.7353724837303162, "rewards/JudgeFidelityReward/std": 0.19509278237819672, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1795 }, { "clip_ratio/high_max": 0.0007137117208912969, "clip_ratio/high_mean": 0.0002464779885485768, "clip_ratio/low_mean": 9.205284877680242e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00033853083732537925, "completions/clipped_ratio": 0.0, "completions/max_length": 207.5, "completions/mean_length": 75.818359375, "completions/min_length": 33.5, "epoch": 1.331360946745562, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.01282467283308506, "learning_rate": 8.774813546115387e-07, "loss": 0.0007058321963995696, "reward": 1.64133220911026, "reward_std": 0.1822691708803177, "rewards/FidelityReward/mean": 0.7592964470386505, "rewards/FidelityReward/std": 0.23272351175546646, "rewards/JudgeFidelityReward/mean": 0.7640715837478638, "rewards/JudgeFidelityReward/std": 0.2247532531619072, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1800 }, { "clip_ratio/high_max": 0.0005740854423493147, "clip_ratio/high_mean": 0.00015222227957565338, "clip_ratio/low_mean": 2.2341376461554318e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017456364876125007, "completions/clipped_ratio": 0.0, "completions/max_length": 189.0, "completions/mean_length": 72.3515625, "completions/min_length": 33.333333333333336, "epoch": 1.3350591715976332, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.012037821672856807, "learning_rate": 8.766782350009716e-07, "loss": 0.00042678010649979117, "reward": 1.6436962286631267, "reward_std": 0.19338424503803253, "rewards/FidelityReward/mean": 0.7621055642763773, "rewards/FidelityReward/std": 0.19756330053011575, "rewards/JudgeFidelityReward/mean": 0.7644833922386169, "rewards/JudgeFidelityReward/std": 0.1805137793223063, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1805 }, { "clip_ratio/high_max": 0.0008607172407209873, "clip_ratio/high_mean": 0.00027615241124294697, "clip_ratio/low_mean": 6.645835237577558e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003426107519771904, "completions/clipped_ratio": 0.0, "completions/max_length": 195.5, "completions/mean_length": 72.55078125, "completions/min_length": 34.5, "epoch": 1.3387573964497042, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.012006631866097451, "learning_rate": 8.758728618286591e-07, "loss": 0.0005356451962143183, "reward": 1.601841926574707, "reward_std": 0.20179623365402222, "rewards/FidelityReward/mean": 0.727857917547226, "rewards/FidelityReward/std": 0.21154622733592987, "rewards/JudgeFidelityReward/mean": 0.7479678094387054, "rewards/JudgeFidelityReward/std": 0.19661234319210052, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1810 }, { "clip_ratio/high_max": 0.0008071298943832517, "clip_ratio/high_mean": 0.0002250616904348135, "clip_ratio/low_mean": 3.657644556369633e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026163813890889287, "completions/clipped_ratio": 0.0, "completions/max_length": 163.33333333333334, "completions/mean_length": 72.02604166666667, "completions/min_length": 33.666666666666664, "epoch": 1.342455621301775, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.5, "kl": 0.012568825297057629, "learning_rate": 8.750652399129257e-07, "loss": 0.0004807698540389538, "reward": 1.599275787671407, "reward_std": 0.196802685658137, "rewards/FidelityReward/mean": 0.7305247982343038, "rewards/FidelityReward/std": 0.2023333211739858, "rewards/JudgeFidelityReward/mean": 0.7375019590059916, "rewards/JudgeFidelityReward/std": 0.19160865247249603, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1815 }, { "clip_ratio/high_max": 0.0006019557011313736, "clip_ratio/high_mean": 0.00020409566350281237, "clip_ratio/low_mean": 0.0001577351868036203, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003618308575823903, "completions/clipped_ratio": 0.0, "completions/max_length": 167.5, "completions/mean_length": 71.984375, "completions/min_length": 33.5, "epoch": 1.3461538461538463, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.011806364171206951, "learning_rate": 8.742553740855505e-07, "loss": 0.0005388997495174408, "reward": 1.5717841982841492, "reward_std": 0.2289201095700264, "rewards/FidelityReward/mean": 0.7123783230781555, "rewards/FidelityReward/std": 0.20507143437862396, "rewards/JudgeFidelityReward/mean": 0.7227180600166321, "rewards/JudgeFidelityReward/std": 0.19251253455877304, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 1820 }, { "clip_ratio/high_max": 0.0005171641241759062, "clip_ratio/high_mean": 0.00010864389478228986, "clip_ratio/low_mean": 8.803586242720485e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001966797630302608, "completions/clipped_ratio": 0.0, "completions/max_length": 211.0, "completions/mean_length": 74.78515625, "completions/min_length": 36.333333333333336, "epoch": 1.3498520710059172, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.012688903138041496, "learning_rate": 8.734432691917368e-07, "loss": 0.0005823013372719288, "reward": 1.593807299931844, "reward_std": 0.20485042532285055, "rewards/FidelityReward/mean": 0.7296378413836161, "rewards/FidelityReward/std": 0.21959637105464935, "rewards/JudgeFidelityReward/mean": 0.728338877360026, "rewards/JudgeFidelityReward/std": 0.21880598862965903, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1825 }, { "clip_ratio/high_max": 0.0012490150053054093, "clip_ratio/high_mean": 0.00034136439790017904, "clip_ratio/low_mean": 0.00011845867411466315, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004598230589181185, "completions/clipped_ratio": 0.0, "completions/max_length": 178.5, "completions/mean_length": 74.939453125, "completions/min_length": 33.5, "epoch": 1.3535502958579881, "frac_reward_zero_std": 0.03125, "grad_norm": 2.4375, "kl": 0.011883430741727352, "learning_rate": 8.726289300900836e-07, "loss": 0.0004892687313258648, "reward": 1.617211639881134, "reward_std": 0.19442393630743027, "rewards/FidelityReward/mean": 0.7454336881637573, "rewards/FidelityReward/std": 0.2082080990076065, "rewards/JudgeFidelityReward/mean": 0.7455090880393982, "rewards/JudgeFidelityReward/std": 0.20385020226240158, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1830 }, { "clip_ratio/high_max": 0.0007356829242780805, "clip_ratio/high_mean": 0.00019961732905358077, "clip_ratio/low_mean": 9.182736393995584e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002914446871727705, "completions/clipped_ratio": 0.0, "completions/max_length": 178.0, "completions/mean_length": 73.75651041666667, "completions/min_length": 29.333333333333332, "epoch": 1.3572485207100593, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.010665005072951317, "learning_rate": 8.718123616525564e-07, "loss": 0.0002845779061317444, "reward": 1.650447169939677, "reward_std": 0.18093852202097574, "rewards/FidelityReward/mean": 0.764764666557312, "rewards/FidelityReward/std": 0.21136906246344248, "rewards/JudgeFidelityReward/mean": 0.7713649074236552, "rewards/JudgeFidelityReward/std": 0.20716015001138052, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1835 }, { "clip_ratio/high_max": 0.0009748612763360142, "clip_ratio/high_mean": 0.0003018984687514603, "clip_ratio/low_mean": 9.937869617715478e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004012771649286151, "completions/clipped_ratio": 0.0, "completions/max_length": 192.5, "completions/mean_length": 74.947265625, "completions/min_length": 35.0, "epoch": 1.3609467455621302, "frac_reward_zero_std": 0.0, "grad_norm": 2.8125, "kl": 0.013194611296057701, "learning_rate": 8.709935687644588e-07, "loss": 0.0005925096571445465, "reward": 1.7097936868667603, "reward_std": 0.17412671446800232, "rewards/FidelityReward/mean": 0.8077413439750671, "rewards/FidelityReward/std": 0.19144728034734726, "rewards/JudgeFidelityReward/mean": 0.8060577809810638, "rewards/JudgeFidelityReward/std": 0.19784612953662872, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1840 }, { "clip_ratio/high_max": 0.0005460373358801007, "clip_ratio/high_mean": 0.00016345489420928062, "clip_ratio/low_mean": 6.709126755595207e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023054616176523268, "completions/clipped_ratio": 0.0, "completions/max_length": 164.0, "completions/mean_length": 73.44661458333333, "completions/min_length": 32.666666666666664, "epoch": 1.3646449704142012, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.012663009949028492, "learning_rate": 8.701725563244021e-07, "loss": 0.000573808141052723, "reward": 1.7024709383646648, "reward_std": 0.16536876062552133, "rewards/FidelityReward/mean": 0.801669696966807, "rewards/FidelityReward/std": 0.2054043859243393, "rewards/JudgeFidelityReward/mean": 0.8016025622685751, "rewards/JudgeFidelityReward/std": 0.20536679526170096, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1845 }, { "clip_ratio/high_max": 0.0009100693743675948, "clip_ratio/high_mean": 0.00029081894317641856, "clip_ratio/low_mean": 0.0001774428354110569, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00046826177276670933, "completions/clipped_ratio": 0.0, "completions/max_length": 177.0, "completions/mean_length": 73.6875, "completions/min_length": 31.5, "epoch": 1.3683431952662721, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.013393172435462475, "learning_rate": 8.693493292442769e-07, "loss": 0.00044684791937470435, "reward": 1.530979573726654, "reward_std": 0.21623776853084564, "rewards/FidelityReward/mean": 0.6828094720840454, "rewards/FidelityReward/std": 0.2310999557375908, "rewards/JudgeFidelityReward/mean": 0.6982933878898621, "rewards/JudgeFidelityReward/std": 0.22163785994052887, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1850 }, { "clip_ratio/high_max": 0.00035632384242489934, "clip_ratio/high_mean": 0.0001441555970814079, "clip_ratio/low_mean": 6.011916557326913e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020427476847544311, "completions/clipped_ratio": 0.0, "completions/max_length": 199.0, "completions/mean_length": 75.15755208333333, "completions/min_length": 36.0, "epoch": 1.372041420118343, "frac_reward_zero_std": 0.0, "grad_norm": 2.75, "kl": 0.013212725892663002, "learning_rate": 8.685238924492228e-07, "loss": 0.0004782894626259804, "reward": 1.600260059038798, "reward_std": 0.21077785392602286, "rewards/FidelityReward/mean": 0.7404036919275919, "rewards/FidelityReward/std": 0.21253215769926706, "rewards/JudgeFidelityReward/mean": 0.7197127739588419, "rewards/JudgeFidelityReward/std": 0.21499882638454437, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1855 }, { "clip_ratio/high_max": 0.0006846747826784849, "clip_ratio/high_mean": 0.0001959384753718041, "clip_ratio/low_mean": 4.840671899728477e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024434520018985496, "completions/clipped_ratio": 0.0, "completions/max_length": 184.5, "completions/mean_length": 74.00390625, "completions/min_length": 34.0, "epoch": 1.3757396449704142, "frac_reward_zero_std": 0.0, "grad_norm": 2.28125, "kl": 0.013054236955940723, "learning_rate": 8.676962508776006e-07, "loss": 0.0005902240052819252, "reward": 1.5932385325431824, "reward_std": 0.21633479744195938, "rewards/FidelityReward/mean": 0.7297228574752808, "rewards/FidelityReward/std": 0.1962890475988388, "rewards/JudgeFidelityReward/mean": 0.7270312607288361, "rewards/JudgeFidelityReward/std": 0.1842285618185997, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1860 }, { "clip_ratio/high_max": 0.000574722490273416, "clip_ratio/high_mean": 0.000160908792167902, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000160908792167902, "completions/clipped_ratio": 0.0, "completions/max_length": 178.66666666666666, "completions/mean_length": 73.73046875, "completions/min_length": 36.333333333333336, "epoch": 1.3794378698224852, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.6875, "kl": 0.012752446718513966, "learning_rate": 8.668664094809606e-07, "loss": 0.0004634507931768894, "reward": 1.674668510754903, "reward_std": 0.17621421813964844, "rewards/FidelityReward/mean": 0.7832501928011576, "rewards/FidelityReward/std": 0.20004505415757498, "rewards/JudgeFidelityReward/mean": 0.7841386596361796, "rewards/JudgeFidelityReward/std": 0.20281489690144858, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1865 }, { "clip_ratio/high_max": 0.0008473006426356732, "clip_ratio/high_mean": 0.00017087968008127062, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017087968008127062, "completions/clipped_ratio": 0.0, "completions/max_length": 194.5, "completions/mean_length": 75.232421875, "completions/min_length": 35.0, "epoch": 1.3831360946745561, "frac_reward_zero_std": 0.0, "grad_norm": 2.484375, "kl": 0.012257186509668828, "learning_rate": 8.660343732240144e-07, "loss": 0.00048199458979070187, "reward": 1.5720021724700928, "reward_std": 0.22584325075149536, "rewards/FidelityReward/mean": 0.716314285993576, "rewards/FidelityReward/std": 0.20361557602882385, "rewards/JudgeFidelityReward/mean": 0.7172352075576782, "rewards/JudgeFidelityReward/std": 0.21219243854284286, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.0539139099419117, "step": 1870 }, { "clip_ratio/high_max": 0.00038597756065428257, "clip_ratio/high_mean": 0.00013127200654707848, "clip_ratio/low_mean": 0.00015298521029762923, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002842572052031755, "completions/clipped_ratio": 0.0, "completions/max_length": 184.33333333333334, "completions/mean_length": 71.27213541666667, "completions/min_length": 36.333333333333336, "epoch": 1.3868343195266273, "frac_reward_zero_std": 0.0, "grad_norm": 2.5625, "kl": 0.012515862844884396, "learning_rate": 8.652001470846051e-07, "loss": 0.000500369444489479, "reward": 1.647334893544515, "reward_std": 0.18721147874991098, "rewards/FidelityReward/mean": 0.7673811713854471, "rewards/FidelityReward/std": 0.18611220518747965, "rewards/JudgeFidelityReward/mean": 0.7599075039227804, "rewards/JudgeFidelityReward/std": 0.16894013186295828, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1875 }, { "clip_ratio/high_max": 0.0008411048911511898, "clip_ratio/high_mean": 0.0002911738585680723, "clip_ratio/low_mean": 2.144082391168922e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003126146853901446, "completions/clipped_ratio": 0.0, "completions/max_length": 174.0, "completions/mean_length": 73.5, "completions/min_length": 34.0, "epoch": 1.3905325443786982, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.014019620418548585, "learning_rate": 8.643637360536771e-07, "loss": 0.0007553543895483017, "reward": 1.654603660106659, "reward_std": 0.1879948526620865, "rewards/FidelityReward/mean": 0.7710461914539337, "rewards/FidelityReward/std": 0.20340420305728912, "rewards/JudgeFidelityReward/mean": 0.7671149373054504, "rewards/JudgeFidelityReward/std": 0.2027776837348938, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1880 }, { "clip_ratio/high_max": 0.0006071173818781971, "clip_ratio/high_mean": 0.0001638718880712986, "clip_ratio/low_mean": 8.326781680807471e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002471396932378411, "completions/clipped_ratio": 0.0, "completions/max_length": 204.66666666666666, "completions/mean_length": 72.99088541666667, "completions/min_length": 37.333333333333336, "epoch": 1.3942307692307692, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.012593930214643478, "learning_rate": 8.635251451352462e-07, "loss": 0.0006394332740455866, "reward": 1.6274619499842327, "reward_std": 0.19400037825107574, "rewards/FidelityReward/mean": 0.7497214277585348, "rewards/FidelityReward/std": 0.21296745538711548, "rewards/JudgeFidelityReward/mean": 0.7554810643196106, "rewards/JudgeFidelityReward/std": 0.20120932161808014, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1885 }, { "clip_ratio/high_max": 0.0006687217159196734, "clip_ratio/high_mean": 0.0002271316305268556, "clip_ratio/low_mean": 0.00014787178515689446, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00037500342004932464, "completions/clipped_ratio": 0.0, "completions/max_length": 197.0, "completions/mean_length": 72.794921875, "completions/min_length": 36.0, "epoch": 1.3979289940828403, "frac_reward_zero_std": 0.03125, "grad_norm": 2.6875, "kl": 0.01129894945770502, "learning_rate": 8.626843793463702e-07, "loss": 0.00031801140867173674, "reward": 1.6211751103401184, "reward_std": 0.1930963173508644, "rewards/FidelityReward/mean": 0.7551383376121521, "rewards/FidelityReward/std": 0.21188615262508392, "rewards/JudgeFidelityReward/mean": 0.7320735156536102, "rewards/JudgeFidelityReward/std": 0.2321598157286644, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1890 }, { "clip_ratio/high_max": 0.00037419276777654885, "clip_ratio/high_mean": 0.00010504612582735717, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010504612582735717, "completions/clipped_ratio": 0.0, "completions/max_length": 184.66666666666666, "completions/mean_length": 71.92578125, "completions/min_length": 35.666666666666664, "epoch": 1.4016272189349113, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.01246770191937685, "learning_rate": 8.618414437171183e-07, "loss": 0.000607813335955143, "reward": 1.6519444783528645, "reward_std": 0.1942050407330195, "rewards/FidelityReward/mean": 0.7690212726593018, "rewards/FidelityReward/std": 0.19808926184972128, "rewards/JudgeFidelityReward/mean": 0.7658464511235555, "rewards/JudgeFidelityReward/std": 0.19380508363246918, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1895 }, { "clip_ratio/high_max": 0.0008171036490239203, "clip_ratio/high_mean": 0.00022051186097087337, "clip_ratio/low_mean": 0.00012805078295059503, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000348562648287043, "completions/clipped_ratio": 0.0, "completions/max_length": 186.5, "completions/mean_length": 75.296875, "completions/min_length": 33.5, "epoch": 1.4053254437869822, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.012288898788392543, "learning_rate": 8.609963432905412e-07, "loss": 0.0005236681550741196, "reward": 1.5517305135726929, "reward_std": 0.18719124048948288, "rewards/FidelityReward/mean": 0.6978406310081482, "rewards/FidelityReward/std": 0.25029218941926956, "rewards/JudgeFidelityReward/mean": 0.707779735326767, "rewards/JudgeFidelityReward/std": 0.2485911175608635, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1900 }, { "clip_ratio/high_max": 0.0006305114598944783, "clip_ratio/high_mean": 0.0003247324493713677, "clip_ratio/low_mean": 5.8633620210457596e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00038336607394739985, "completions/clipped_ratio": 0.0, "completions/max_length": 188.66666666666666, "completions/mean_length": 74.94140625, "completions/min_length": 38.0, "epoch": 1.4090236686390534, "frac_reward_zero_std": 0.0, "grad_norm": 2.75, "kl": 0.011807984858751296, "learning_rate": 8.601490831226414e-07, "loss": 0.00032819416373968123, "reward": 1.5834639469782512, "reward_std": 0.19859248399734497, "rewards/FidelityReward/mean": 0.7243876655896505, "rewards/FidelityReward/std": 0.20427543421586355, "rewards/JudgeFidelityReward/mean": 0.7181525230407715, "rewards/JudgeFidelityReward/std": 0.20285378396511078, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1905 }, { "clip_ratio/high_max": 0.000833934242837131, "clip_ratio/high_mean": 0.0002213931642472744, "clip_ratio/low_mean": 4.1431727004237474e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002628248941618949, "completions/clipped_ratio": 0.0, "completions/max_length": 260.5, "completions/mean_length": 74.78125, "completions/min_length": 38.0, "epoch": 1.4127218934911243, "frac_reward_zero_std": 0.015625, "grad_norm": 2.5625, "kl": 0.01225109789520502, "learning_rate": 8.592996682823421e-07, "loss": 0.0006511014886200428, "reward": 1.5937748551368713, "reward_std": 0.2048766016960144, "rewards/FidelityReward/mean": 0.731330156326294, "rewards/FidelityReward/std": 0.2259725034236908, "rewards/JudgeFidelityReward/mean": 0.7268425226211548, "rewards/JudgeFidelityReward/std": 0.21959859132766724, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1910 }, { "clip_ratio/high_max": 0.0003478190395981073, "clip_ratio/high_mean": 8.215311099775135e-05, "clip_ratio/low_mean": 3.590724372770637e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00011806035181507468, "completions/clipped_ratio": 0.0, "completions/max_length": 187.66666666666666, "completions/mean_length": 73.91536458333333, "completions/min_length": 34.666666666666664, "epoch": 1.4164201183431953, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.012368973903357982, "learning_rate": 8.584481038514572e-07, "loss": 0.0003651672275736928, "reward": 1.5917067130406697, "reward_std": 0.21968194842338562, "rewards/FidelityReward/mean": 0.7258889277776083, "rewards/FidelityReward/std": 0.2061080684264501, "rewards/JudgeFidelityReward/mean": 0.7316356102625529, "rewards/JudgeFidelityReward/std": 0.19027042388916016, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1915 }, { "clip_ratio/high_max": 0.0008862263057380915, "clip_ratio/high_mean": 0.00017738354508765042, "clip_ratio/low_mean": 6.870878860354424e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002460923453327268, "completions/clipped_ratio": 0.0, "completions/max_length": 182.5, "completions/mean_length": 77.189453125, "completions/min_length": 38.5, "epoch": 1.4201183431952662, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.010951552912592888, "learning_rate": 8.57594394924662e-07, "loss": 0.0004763946868479252, "reward": 1.6314197182655334, "reward_std": 0.2122417911887169, "rewards/FidelityReward/mean": 0.7502140998840332, "rewards/FidelityReward/std": 0.20673208683729172, "rewards/JudgeFidelityReward/mean": 0.7643643021583557, "rewards/JudgeFidelityReward/std": 0.1818128600716591, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1920 }, { "clip_ratio/high_max": 0.0007265046006068587, "clip_ratio/high_mean": 0.00016803761245682837, "clip_ratio/low_mean": 3.63919825758785e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020442958921194077, "completions/clipped_ratio": 0.0, "completions/max_length": 238.66666666666666, "completions/mean_length": 76.52994791666667, "completions/min_length": 38.333333333333336, "epoch": 1.4238165680473374, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.01313247550278902, "learning_rate": 8.567385466094603e-07, "loss": 0.000529089430347085, "reward": 1.6660173336664836, "reward_std": 0.18097933133443198, "rewards/FidelityReward/mean": 0.7776610851287842, "rewards/FidelityReward/std": 0.1920987367630005, "rewards/JudgeFidelityReward/mean": 0.7767125169436137, "rewards/JudgeFidelityReward/std": 0.18546458582083383, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1925 }, { "clip_ratio/high_max": 0.0008477894589304924, "clip_ratio/high_mean": 0.0002510300953872502, "clip_ratio/low_mean": 0.00016144007095135747, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004124701546970755, "completions/clipped_ratio": 0.0, "completions/max_length": 196.0, "completions/mean_length": 73.267578125, "completions/min_length": 31.0, "epoch": 1.4275147928994083, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.011833734810352325, "learning_rate": 8.558805640261562e-07, "loss": 0.0006404291838407516, "reward": 1.632651925086975, "reward_std": 0.18800926953554153, "rewards/FidelityReward/mean": 0.7502429783344269, "rewards/FidelityReward/std": 0.21023863554000854, "rewards/JudgeFidelityReward/mean": 0.7648177742958069, "rewards/JudgeFidelityReward/std": 0.1788419559597969, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1930 }, { "clip_ratio/high_max": 0.0006663976470008493, "clip_ratio/high_mean": 0.0001244685525307432, "clip_ratio/low_mean": 2.1551724057644605e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001460202765883878, "completions/clipped_ratio": 0.0, "completions/max_length": 176.0, "completions/mean_length": 73.55989583333333, "completions/min_length": 35.0, "epoch": 1.4312130177514792, "frac_reward_zero_std": 0.0, "grad_norm": 2.75, "kl": 0.012324767746031284, "learning_rate": 8.550204523078224e-07, "loss": 0.00036934730596840384, "reward": 1.6119284232457478, "reward_std": 0.20255887508392334, "rewards/FidelityReward/mean": 0.7406187256177267, "rewards/FidelityReward/std": 0.20316515366236368, "rewards/JudgeFidelityReward/mean": 0.7439213991165161, "rewards/JudgeFidelityReward/std": 0.20746507743994394, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1935 }, { "clip_ratio/high_max": 0.0006291824625805021, "clip_ratio/high_mean": 0.00014511077315546573, "clip_ratio/low_mean": 0.00011117398098576814, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002562847570516169, "completions/clipped_ratio": 0.0, "completions/max_length": 183.0, "completions/mean_length": 76.1015625, "completions/min_length": 39.5, "epoch": 1.4349112426035502, "frac_reward_zero_std": 0.0, "grad_norm": 2.890625, "kl": 0.010664456896483899, "learning_rate": 8.541582166002694e-07, "loss": 0.00027603493072092533, "reward": 1.6704545617103577, "reward_std": 0.17981693148612976, "rewards/FidelityReward/mean": 0.7794171869754791, "rewards/FidelityReward/std": 0.1972431018948555, "rewards/JudgeFidelityReward/mean": 0.7879342436790466, "rewards/JudgeFidelityReward/std": 0.19523096084594727, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 1940 }, { "clip_ratio/high_max": 0.0006632585311308504, "clip_ratio/high_mean": 0.0001732255332171917, "clip_ratio/low_mean": 4.1911471635103224e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021513700485229493, "completions/clipped_ratio": 0.0, "completions/max_length": 190.33333333333334, "completions/mean_length": 74.72005208333333, "completions/min_length": 32.333333333333336, "epoch": 1.4386094674556213, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.011476054787635803, "learning_rate": 8.53293862062015e-07, "loss": 0.0006256145425140858, "reward": 1.5847260157267253, "reward_std": 0.20294939478238425, "rewards/FidelityReward/mean": 0.7265663941701254, "rewards/FidelityReward/std": 0.22715513904889426, "rewards/JudgeFidelityReward/mean": 0.720225473244985, "rewards/JudgeFidelityReward/std": 0.22563823560873666, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.050238288938999176, "step": 1945 }, { "clip_ratio/high_max": 0.0008806644007563591, "clip_ratio/high_mean": 0.00023239649162860588, "clip_ratio/low_mean": 0.00013664444850292057, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00036904094740748407, "completions/clipped_ratio": 0.0, "completions/max_length": 160.5, "completions/mean_length": 71.41015625, "completions/min_length": 36.0, "epoch": 1.4423076923076923, "frac_reward_zero_std": 0.046875, "grad_norm": 2.578125, "kl": 0.011575172655284405, "learning_rate": 8.524273938642538e-07, "loss": 0.0007465443573892117, "reward": 1.6676233410835266, "reward_std": 0.1753167062997818, "rewards/FidelityReward/mean": 0.7829823195934296, "rewards/FidelityReward/std": 0.1938806250691414, "rewards/JudgeFidelityReward/mean": 0.7712351679801941, "rewards/JudgeFidelityReward/std": 0.18993373215198517, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1950 }, { "clip_ratio/high_max": 0.0006168080028146506, "clip_ratio/high_mean": 0.00016026359517127276, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016026359517127276, "completions/clipped_ratio": 0.0, "completions/max_length": 180.33333333333334, "completions/mean_length": 74.86588541666667, "completions/min_length": 34.0, "epoch": 1.4460059171597632, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.010754941403865815, "learning_rate": 8.51558817190825e-07, "loss": 0.0005273034330457449, "reward": 1.680147369702657, "reward_std": 0.17564235627651215, "rewards/FidelityReward/mean": 0.791631023089091, "rewards/FidelityReward/std": 0.19607848425706229, "rewards/JudgeFidelityReward/mean": 0.7783347368240356, "rewards/JudgeFidelityReward/std": 0.20316017667452493, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1955 }, { "clip_ratio/high_max": 0.0009182008216157556, "clip_ratio/high_mean": 0.00016596487548667938, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016596487548667938, "completions/clipped_ratio": 0.0, "completions/max_length": 169.0, "completions/mean_length": 74.576171875, "completions/min_length": 29.5, "epoch": 1.4497041420118344, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.011217491142451764, "learning_rate": 8.506881372381831e-07, "loss": 0.0005351456813514233, "reward": 1.634124994277954, "reward_std": 0.1778339222073555, "rewards/FidelityReward/mean": 0.7588882446289062, "rewards/FidelityReward/std": 0.1804957091808319, "rewards/JudgeFidelityReward/mean": 0.7504734694957733, "rewards/JudgeFidelityReward/std": 0.17843083292245865, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1960 }, { "clip_ratio/high_max": 0.0003125000046566129, "clip_ratio/high_mean": 0.0001002071425318718, "clip_ratio/low_mean": 6.368818139890209e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016389533557230606, "completions/clipped_ratio": 0.0, "completions/max_length": 179.66666666666666, "completions/mean_length": 74.04947916666667, "completions/min_length": 38.333333333333336, "epoch": 1.4534023668639053, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.011232135631144047, "learning_rate": 8.498153592153655e-07, "loss": 0.0003831669222563505, "reward": 1.6447169383366902, "reward_std": 0.18325010438760123, "rewards/FidelityReward/mean": 0.7642817894617716, "rewards/FidelityReward/std": 0.20576361318429312, "rewards/JudgeFidelityReward/mean": 0.7621723214785258, "rewards/JudgeFidelityReward/std": 0.20629224181175232, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1965 }, { "clip_ratio/high_max": 0.0007040515774860978, "clip_ratio/high_mean": 0.000255155423656106, "clip_ratio/low_mean": 0.00016282938595395534, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00041798478923738003, "completions/clipped_ratio": 0.0, "completions/max_length": 234.5, "completions/mean_length": 77.0390625, "completions/min_length": 35.0, "epoch": 1.4571005917159763, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.012293397821485997, "learning_rate": 8.489404883439615e-07, "loss": 0.0004772244021296501, "reward": 1.6574172377586365, "reward_std": 0.19504611939191818, "rewards/FidelityReward/mean": 0.7674507200717926, "rewards/FidelityReward/std": 0.19057603925466537, "rewards/JudgeFidelityReward/mean": 0.7799330353736877, "rewards/JudgeFidelityReward/std": 0.17307893186807632, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1970 }, { "clip_ratio/high_max": 0.000395570695400238, "clip_ratio/high_mean": 0.00011169863282702863, "clip_ratio/low_mean": 4.601567634381354e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00015771430335007608, "completions/clipped_ratio": 0.0, "completions/max_length": 169.33333333333334, "completions/mean_length": 73.48828125, "completions/min_length": 35.333333333333336, "epoch": 1.4607988165680474, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.011428861692547799, "learning_rate": 8.480635298580821e-07, "loss": 0.00043660192750394344, "reward": 1.633668025334676, "reward_std": 0.18911485373973846, "rewards/FidelityReward/mean": 0.7557458281517029, "rewards/FidelityReward/std": 0.2099419186512629, "rewards/JudgeFidelityReward/mean": 0.7558443347613016, "rewards/JudgeFidelityReward/std": 0.2009765257438024, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 1975 }, { "clip_ratio/high_max": 0.0005588627886027097, "clip_ratio/high_mean": 0.00015604701184201986, "clip_ratio/low_mean": 8.170825021807105e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023775525332894176, "completions/clipped_ratio": 0.0, "completions/max_length": 170.0, "completions/mean_length": 72.95703125, "completions/min_length": 34.5, "epoch": 1.4644970414201184, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.01089911125600338, "learning_rate": 8.471844890043271e-07, "loss": 0.0004986213054507971, "reward": 1.7287517786026, "reward_std": 0.1759515255689621, "rewards/FidelityReward/mean": 0.8256562948226929, "rewards/FidelityReward/std": 0.17548642307519913, "rewards/JudgeFidelityReward/mean": 0.8081441819667816, "rewards/JudgeFidelityReward/std": 0.18211521953344345, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 1980 }, { "clip_ratio/high_max": 0.00036003601271659134, "clip_ratio/high_mean": 8.33186786621809e-05, "clip_ratio/low_mean": 2.3169601627159862e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010648828465491534, "completions/clipped_ratio": 0.0, "completions/max_length": 194.0, "completions/mean_length": 74.88932291666667, "completions/min_length": 39.0, "epoch": 1.4681952662721893, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.65625, "kl": 0.012168773822486401, "learning_rate": 8.463033710417553e-07, "loss": 0.0005727767944335938, "reward": 1.5948572158813477, "reward_std": 0.21196073293685913, "rewards/FidelityReward/mean": 0.7256272236506144, "rewards/FidelityReward/std": 0.22927244007587433, "rewards/JudgeFidelityReward/mean": 0.7397620876630148, "rewards/JudgeFidelityReward/std": 0.216477503379186, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 1985 }, { "clip_ratio/high_max": 0.0010125469765625895, "clip_ratio/high_mean": 0.0003295107584563084, "clip_ratio/low_mean": 7.886727689765394e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004083780338987708, "completions/clipped_ratio": 0.0, "completions/max_length": 166.0, "completions/mean_length": 75.169921875, "completions/min_length": 36.0, "epoch": 1.4718934911242605, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.01212230883538723, "learning_rate": 8.454201812418517e-07, "loss": 0.00035561986733227967, "reward": 1.6784366965293884, "reward_std": 0.19050103425979614, "rewards/FidelityReward/mean": 0.7860145568847656, "rewards/FidelityReward/std": 0.18860485404729843, "rewards/JudgeFidelityReward/mean": 0.7887505292892456, "rewards/JudgeFidelityReward/std": 0.1805458515882492, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 1990 }, { "clip_ratio/high_max": 0.0008269680198282004, "clip_ratio/high_mean": 0.00025304307928308845, "clip_ratio/low_mean": 5.308408581186086e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003061271505430341, "completions/clipped_ratio": 0.0, "completions/max_length": 195.66666666666666, "completions/mean_length": 74.59765625, "completions/min_length": 30.666666666666668, "epoch": 1.4755917159763314, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.78125, "kl": 0.010889950767159462, "learning_rate": 8.445349248884966e-07, "loss": 0.00040958672761917116, "reward": 1.6007168690363567, "reward_std": 0.20353160301844278, "rewards/FidelityReward/mean": 0.7323667804400126, "rewards/FidelityReward/std": 0.21440406640370688, "rewards/JudgeFidelityReward/mean": 0.7393043835957845, "rewards/JudgeFidelityReward/std": 0.20011361440022787, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 1995 }, { "clip_ratio/high_max": 0.0007957597030326724, "clip_ratio/high_mean": 0.000284998508868739, "clip_ratio/low_mean": 0.00011766190582420676, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00040266041178256274, "completions/clipped_ratio": 0.0, "completions/max_length": 157.0, "completions/mean_length": 76.521484375, "completions/min_length": 37.5, "epoch": 1.4792899408284024, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.013439355790615082, "learning_rate": 8.436476072779341e-07, "loss": 0.0007481521461158991, "reward": 1.5992175936698914, "reward_std": 0.20872467756271362, "rewards/FidelityReward/mean": 0.7271949350833893, "rewards/FidelityReward/std": 0.21412020176649094, "rewards/JudgeFidelityReward/mean": 0.7440452873706818, "rewards/JudgeFidelityReward/std": 0.19811651855707169, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2000 }, { "clip_ratio/high_max": 0.0006447335239499807, "clip_ratio/high_mean": 0.000270901492331177, "clip_ratio/low_mean": 8.137396071106195e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00035227545304223896, "completions/clipped_ratio": 0.0, "completions/max_length": 154.66666666666666, "completions/mean_length": 74.04036458333333, "completions/min_length": 32.0, "epoch": 1.4829881656804733, "frac_reward_zero_std": 0.0, "grad_norm": 2.78125, "kl": 0.011464026756584644, "learning_rate": 8.427582337187402e-07, "loss": 0.0005001611076295376, "reward": 1.6168311436971028, "reward_std": 0.19841261704762778, "rewards/FidelityReward/mean": 0.7484582662582397, "rewards/FidelityReward/std": 0.20343482494354248, "rewards/JudgeFidelityReward/mean": 0.7367456356684366, "rewards/JudgeFidelityReward/std": 0.21132923165957132, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2005 }, { "clip_ratio/high_max": 0.0003649375634267926, "clip_ratio/high_mean": 0.0001517139549832791, "clip_ratio/low_mean": 6.279993103817106e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021451387437991797, "completions/clipped_ratio": 0.0, "completions/max_length": 184.5, "completions/mean_length": 73.65625, "completions/min_length": 36.5, "epoch": 1.4866863905325443, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.012176608107984066, "learning_rate": 8.418668095317911e-07, "loss": 0.0004187407437711954, "reward": 1.6191818714141846, "reward_std": 0.19651199877262115, "rewards/FidelityReward/mean": 0.7436304688453674, "rewards/FidelityReward/std": 0.21364454180002213, "rewards/JudgeFidelityReward/mean": 0.7511028051376343, "rewards/JudgeFidelityReward/std": 0.1935213804244995, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2010 }, { "clip_ratio/high_max": 0.0008293177001178264, "clip_ratio/high_mean": 0.00018281519296579063, "clip_ratio/low_mean": 3.970214602304623e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002225173288024962, "completions/clipped_ratio": 0.0, "completions/max_length": 177.0, "completions/mean_length": 73.18098958333333, "completions/min_length": 34.666666666666664, "epoch": 1.4903846153846154, "frac_reward_zero_std": 0.0, "grad_norm": 2.765625, "kl": 0.011379591561853886, "learning_rate": 8.40973340050231e-07, "loss": 0.0005490376148372889, "reward": 1.666893760363261, "reward_std": 0.1761219104131063, "rewards/FidelityReward/mean": 0.7778477271397909, "rewards/FidelityReward/std": 0.19484607378641763, "rewards/JudgeFidelityReward/mean": 0.7780921657880148, "rewards/JudgeFidelityReward/std": 0.178253764907519, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2015 }, { "clip_ratio/high_max": 0.0005333544919267297, "clip_ratio/high_mean": 0.00019105548271909357, "clip_ratio/low_mean": 6.685417029075324e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002579096704721451, "completions/clipped_ratio": 0.0, "completions/max_length": 181.0, "completions/mean_length": 72.60546875, "completions/min_length": 35.0, "epoch": 1.4940828402366864, "frac_reward_zero_std": 0.0, "grad_norm": 2.84375, "kl": 0.011549727804958821, "learning_rate": 8.400778306194412e-07, "loss": 0.0003899396862834692, "reward": 1.6869075894355774, "reward_std": 0.16795069724321365, "rewards/FidelityReward/mean": 0.7937266528606415, "rewards/FidelityReward/std": 0.19250322133302689, "rewards/JudgeFidelityReward/mean": 0.7863616943359375, "rewards/JudgeFidelityReward/std": 0.1997336968779564, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2020 }, { "clip_ratio/high_max": 0.0004030483309179544, "clip_ratio/high_mean": 0.0001402791647706181, "clip_ratio/low_mean": 0.0001107557793147862, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002510349382646382, "completions/clipped_ratio": 0.0, "completions/max_length": 195.0, "completions/mean_length": 73.32291666666667, "completions/min_length": 36.666666666666664, "epoch": 1.4977810650887573, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.012331825867295265, "learning_rate": 8.391802865970068e-07, "loss": 0.00041972491890192033, "reward": 1.6187774340311687, "reward_std": 0.19427457451820374, "rewards/FidelityReward/mean": 0.7404705087343851, "rewards/FidelityReward/std": 0.22153939306735992, "rewards/JudgeFidelityReward/mean": 0.7566137909889221, "rewards/JudgeFidelityReward/std": 0.20743703842163086, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2025 }, { "clip_ratio/high_max": 0.0007471030810847878, "clip_ratio/high_mean": 0.00024062842421699316, "clip_ratio/low_mean": 4.235988308209926e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028298830438870936, "completions/clipped_ratio": 0.0, "completions/max_length": 190.5, "completions/mean_length": 73.06640625, "completions/min_length": 31.0, "epoch": 1.5014792899408285, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.010572643764317036, "learning_rate": 8.382807133526855e-07, "loss": 0.0004991964437067508, "reward": 1.6763379573822021, "reward_std": 0.18632197380065918, "rewards/FidelityReward/mean": 0.7856652140617371, "rewards/FidelityReward/std": 0.19486436992883682, "rewards/JudgeFidelityReward/mean": 0.7813454866409302, "rewards/JudgeFidelityReward/std": 0.19247498363256454, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2030 }, { "clip_ratio/high_max": 0.0006381488870829344, "clip_ratio/high_mean": 0.00014550447231158613, "clip_ratio/low_mean": 3.9798454963602126e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018530292436480522, "completions/clipped_ratio": 0.0, "completions/max_length": 182.33333333333334, "completions/mean_length": 71.98567708333333, "completions/min_length": 36.0, "epoch": 1.5051775147928994, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.011338648945093155, "learning_rate": 8.373791162683755e-07, "loss": 0.0003519635647535324, "reward": 1.6114954948425293, "reward_std": 0.1924206813176473, "rewards/FidelityReward/mean": 0.739962120850881, "rewards/FidelityReward/std": 0.2086717983086904, "rewards/JudgeFidelityReward/mean": 0.7443687717119852, "rewards/JudgeFidelityReward/std": 0.17255700131257376, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2035 }, { "clip_ratio/high_max": 0.000960623612627387, "clip_ratio/high_mean": 0.00022585924016311766, "clip_ratio/low_mean": 8.98384809261188e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00031569772399961946, "completions/clipped_ratio": 0.0, "completions/max_length": 193.0, "completions/mean_length": 71.708984375, "completions/min_length": 31.0, "epoch": 1.5088757396449703, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.011940872110426427, "learning_rate": 8.364755007380828e-07, "loss": 0.00047150570899248125, "reward": 1.7051737308502197, "reward_std": 0.1762956753373146, "rewards/FidelityReward/mean": 0.8015437722206116, "rewards/FidelityReward/std": 0.19251487404108047, "rewards/JudgeFidelityReward/mean": 0.8092131614685059, "rewards/JudgeFidelityReward/std": 0.18898293375968933, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2040 }, { "clip_ratio/high_max": 0.00031471282709389926, "clip_ratio/high_mean": 6.29240064881742e-05, "clip_ratio/low_mean": 6.272264436120167e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012564664939418435, "completions/clipped_ratio": 0.0, "completions/max_length": 179.33333333333334, "completions/mean_length": 72.69661458333333, "completions/min_length": 33.333333333333336, "epoch": 1.5125739644970415, "frac_reward_zero_std": 0.03125, "grad_norm": 2.375, "kl": 0.010867985337972641, "learning_rate": 8.355698721678894e-07, "loss": 0.0004728849045932293, "reward": 1.6604364315668743, "reward_std": 0.1688572565714518, "rewards/FidelityReward/mean": 0.7687238454818726, "rewards/FidelityReward/std": 0.2107821156581243, "rewards/JudgeFidelityReward/mean": 0.7834251324335734, "rewards/JudgeFidelityReward/std": 0.196114644408226, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2045 }, { "clip_ratio/high_max": 0.0011943458113819361, "clip_ratio/high_mean": 0.0003224740765290335, "clip_ratio/low_mean": 4.3130874109920116e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003656049491837621, "completions/clipped_ratio": 0.0, "completions/max_length": 171.5, "completions/mean_length": 73.26171875, "completions/min_length": 36.5, "epoch": 1.5162721893491125, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.011897632479667663, "learning_rate": 8.346622359759206e-07, "loss": 0.00044192480854690077, "reward": 1.5533738136291504, "reward_std": 0.2148335725069046, "rewards/FidelityReward/mean": 0.703681081533432, "rewards/FidelityReward/std": 0.2316771224141121, "rewards/JudgeFidelityReward/mean": 0.7013386785984039, "rewards/JudgeFidelityReward/std": 0.20885994285345078, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2050 }, { "clip_ratio/high_max": 0.0005005087936297059, "clip_ratio/high_mean": 0.00010297572007402777, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010297572007402777, "completions/clipped_ratio": 0.0, "completions/max_length": 171.66666666666666, "completions/mean_length": 75.0, "completions/min_length": 37.0, "epoch": 1.5199704142011834, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.453125, "kl": 0.01220358032733202, "learning_rate": 8.337525975923129e-07, "loss": 0.0007056592032313347, "reward": 1.6190832455952961, "reward_std": 0.19800010323524475, "rewards/FidelityReward/mean": 0.7466488281885783, "rewards/FidelityReward/std": 0.19435599446296692, "rewards/JudgeFidelityReward/mean": 0.7448687354723612, "rewards/JudgeFidelityReward/std": 0.18449940284093222, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2055 }, { "clip_ratio/high_max": 0.0008448205422610045, "clip_ratio/high_mean": 0.00025106388493441046, "clip_ratio/low_mean": 0.00010637990781106054, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00035744378110393884, "completions/clipped_ratio": 0.0, "completions/max_length": 167.5, "completions/mean_length": 72.85546875, "completions/min_length": 36.0, "epoch": 1.5236686390532546, "frac_reward_zero_std": 0.046875, "grad_norm": 2.515625, "kl": 0.012324728816747666, "learning_rate": 8.328409624591811e-07, "loss": 0.0007626983802765608, "reward": 1.6596921682357788, "reward_std": 0.18210755288600922, "rewards/FidelityReward/mean": 0.777124673128128, "rewards/FidelityReward/std": 0.22077488899230957, "rewards/JudgeFidelityReward/mean": 0.7670878469944, "rewards/JudgeFidelityReward/std": 0.21469471603631973, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2060 }, { "clip_ratio/high_max": 0.00021030493080615998, "clip_ratio/high_mean": 0.00013665774604305626, "clip_ratio/low_mean": 4.238154651829973e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017903929256135598, "completions/clipped_ratio": 0.0, "completions/max_length": 161.66666666666666, "completions/mean_length": 75.03776041666667, "completions/min_length": 35.333333333333336, "epoch": 1.5273668639053253, "frac_reward_zero_std": 0.03125, "grad_norm": 2.609375, "kl": 0.01151470486074686, "learning_rate": 8.319273360305864e-07, "loss": 0.000574352452531457, "reward": 1.6095651785532634, "reward_std": 0.18704626460870108, "rewards/FidelityReward/mean": 0.7379070917765299, "rewards/FidelityReward/std": 0.21961763997872671, "rewards/JudgeFidelityReward/mean": 0.7459202806154887, "rewards/JudgeFidelityReward/std": 0.20186678568522134, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 2065 }, { "clip_ratio/high_max": 0.000510616903193295, "clip_ratio/high_mean": 0.00015895956312306225, "clip_ratio/low_mean": 2.0325202785898e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017928477027453482, "completions/clipped_ratio": 0.0, "completions/max_length": 213.5, "completions/mean_length": 75.91796875, "completions/min_length": 37.0, "epoch": 1.5310650887573964, "frac_reward_zero_std": 0.0, "grad_norm": 2.84375, "kl": 0.011600124835968017, "learning_rate": 8.31011723772503e-07, "loss": 0.0005198557395488024, "reward": 1.6489006280899048, "reward_std": 0.20801415294408798, "rewards/FidelityReward/mean": 0.7709846496582031, "rewards/FidelityReward/std": 0.19789841771125793, "rewards/JudgeFidelityReward/mean": 0.7597382068634033, "rewards/JudgeFidelityReward/std": 0.2019265964627266, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 2070 }, { "clip_ratio/high_max": 0.0003491088980808854, "clip_ratio/high_mean": 0.00018726757843978703, "clip_ratio/low_mean": 1.9623234402388333e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020689081866294147, "completions/clipped_ratio": 0.0, "completions/max_length": 199.0, "completions/mean_length": 74.52994791666667, "completions/min_length": 36.0, "epoch": 1.5347633136094676, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.011758560128509999, "learning_rate": 8.300941311627861e-07, "loss": 0.0005090643186122179, "reward": 1.652742862701416, "reward_std": 0.18115088840325674, "rewards/FidelityReward/mean": 0.7659875949223837, "rewards/FidelityReward/std": 0.2211710810661316, "rewards/JudgeFidelityReward/mean": 0.7748125791549683, "rewards/JudgeFidelityReward/std": 0.21308218439420065, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2075 }, { "clip_ratio/high_max": 0.0014154738979414105, "clip_ratio/high_mean": 0.00033958606654778124, "clip_ratio/low_mean": 0.00012582994531840086, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004654160002246499, "completions/clipped_ratio": 0.0, "completions/max_length": 162.0, "completions/mean_length": 72.66015625, "completions/min_length": 37.0, "epoch": 1.5384615384615383, "frac_reward_zero_std": 0.015625, "grad_norm": 2.28125, "kl": 0.012396631017327308, "learning_rate": 8.291745636911381e-07, "loss": 0.0008121814578771591, "reward": 1.5910797119140625, "reward_std": 0.19493377208709717, "rewards/FidelityReward/mean": 0.7242679297924042, "rewards/FidelityReward/std": 0.2267129123210907, "rewards/JudgeFidelityReward/mean": 0.7336234748363495, "rewards/JudgeFidelityReward/std": 0.20168249309062958, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2080 }, { "clip_ratio/high_max": 0.0007517500082030893, "clip_ratio/high_mean": 0.00022192656178958713, "clip_ratio/low_mean": 2.0868112915195525e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002427946776151657, "completions/clipped_ratio": 0.0, "completions/max_length": 214.0, "completions/mean_length": 74.78776041666667, "completions/min_length": 36.666666666666664, "epoch": 1.5421597633136095, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.28125, "kl": 0.013031011447310448, "learning_rate": 8.282530268590775e-07, "loss": 0.00047359969466924666, "reward": 1.6410526434580486, "reward_std": 0.18196176985899606, "rewards/FidelityReward/mean": 0.7587262193361918, "rewards/FidelityReward/std": 0.22379104296366373, "rewards/JudgeFidelityReward/mean": 0.7646528085072836, "rewards/JudgeFidelityReward/std": 0.21784104903539023, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2085 }, { "clip_ratio/high_max": 0.0007462204666808248, "clip_ratio/high_mean": 0.0002146659593563527, "clip_ratio/low_mean": 0.00010437633318360895, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003190423012711108, "completions/clipped_ratio": 0.0, "completions/max_length": 167.0, "completions/mean_length": 73.923828125, "completions/min_length": 40.5, "epoch": 1.5458579881656804, "frac_reward_zero_std": 0.0, "grad_norm": 3.015625, "kl": 0.012270372919738292, "learning_rate": 8.273295261799043e-07, "loss": 0.0006639871746301651, "reward": 1.6098206043243408, "reward_std": 0.17507506906986237, "rewards/FidelityReward/mean": 0.7359963059425354, "rewards/FidelityReward/std": 0.24565978348255157, "rewards/JudgeFidelityReward/mean": 0.749601811170578, "rewards/JudgeFidelityReward/std": 0.2448120266199112, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2090 }, { "clip_ratio/high_max": 0.0006454673130065203, "clip_ratio/high_mean": 0.00012789663160219788, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012789663160219788, "completions/clipped_ratio": 0.0, "completions/max_length": 183.0, "completions/mean_length": 76.3203125, "completions/min_length": 36.333333333333336, "epoch": 1.5495562130177514, "frac_reward_zero_std": 0.0, "grad_norm": 2.359375, "kl": 0.01383034735918045, "learning_rate": 8.264040671786676e-07, "loss": 0.0005626892670989037, "reward": 1.6608450412750244, "reward_std": 0.1949376960595449, "rewards/FidelityReward/mean": 0.7745386362075806, "rewards/FidelityReward/std": 0.19964482883612314, "rewards/JudgeFidelityReward/mean": 0.7752169171969095, "rewards/JudgeFidelityReward/std": 0.19760538140932718, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 2095 }, { "clip_ratio/high_max": 0.0006689111818559468, "clip_ratio/high_mean": 0.00012244534154888244, "clip_ratio/low_mean": 0.0001038030328345485, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022624837292823942, "completions/clipped_ratio": 0.0, "completions/max_length": 227.5, "completions/mean_length": 77.255859375, "completions/min_length": 37.5, "epoch": 1.5532544378698225, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.011786240339279174, "learning_rate": 8.254766553921328e-07, "loss": 0.0003788528498262167, "reward": 1.6372103691101074, "reward_std": 0.1976190209388733, "rewards/FidelityReward/mean": 0.7587093114852905, "rewards/FidelityReward/std": 0.20588818192481995, "rewards/JudgeFidelityReward/mean": 0.7589550018310547, "rewards/JudgeFidelityReward/std": 0.19630331546068192, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2100 }, { "clip_ratio/high_max": 0.000744190881960094, "clip_ratio/high_mean": 0.00024896293762139977, "clip_ratio/low_mean": 3.649679565569386e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028545971726998685, "completions/clipped_ratio": 0.0, "completions/max_length": 177.66666666666666, "completions/mean_length": 74.14192708333333, "completions/min_length": 37.0, "epoch": 1.5569526627218935, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.010416222549974919, "learning_rate": 8.245472963687484e-07, "loss": 0.0004883617628365755, "reward": 1.6294200420379639, "reward_std": 0.18035167952378592, "rewards/FidelityReward/mean": 0.7546579440434774, "rewards/FidelityReward/std": 0.2165138175090154, "rewards/JudgeFidelityReward/mean": 0.7495240569114685, "rewards/JudgeFidelityReward/std": 0.2086405654748281, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2105 }, { "clip_ratio/high_max": 0.0006373128620907664, "clip_ratio/high_mean": 0.00021029068739153445, "clip_ratio/low_mean": 4.222222196403891e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002525129064451903, "completions/clipped_ratio": 0.0, "completions/max_length": 212.0, "completions/mean_length": 74.953125, "completions/min_length": 32.0, "epoch": 1.5606508875739644, "frac_reward_zero_std": 0.0, "grad_norm": 2.359375, "kl": 0.01197675932198763, "learning_rate": 8.236159956686124e-07, "loss": 0.0005346008110791445, "reward": 1.6714772582054138, "reward_std": 0.1814563125371933, "rewards/FidelityReward/mean": 0.7772618532180786, "rewards/FidelityReward/std": 0.19714920222759247, "rewards/JudgeFidelityReward/mean": 0.7884308695793152, "rewards/JudgeFidelityReward/std": 0.18568847328424454, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2110 }, { "clip_ratio/high_max": 0.00036838704254478217, "clip_ratio/high_mean": 0.00012222668156027793, "clip_ratio/low_mean": 2.50752258580178e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014730190741829574, "completions/clipped_ratio": 0.0, "completions/max_length": 198.66666666666666, "completions/mean_length": 73.86067708333333, "completions/min_length": 37.333333333333336, "epoch": 1.5643491124260356, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.011265297047793865, "learning_rate": 8.226827588634393e-07, "loss": 0.0004672681912779808, "reward": 1.6080280939737956, "reward_std": 0.192642812927564, "rewards/FidelityReward/mean": 0.7370966474215189, "rewards/FidelityReward/std": 0.21517479916413626, "rewards/JudgeFidelityReward/mean": 0.741862972577413, "rewards/JudgeFidelityReward/std": 0.20442030827204385, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2115 }, { "clip_ratio/high_max": 0.0008807307342067361, "clip_ratio/high_mean": 0.0002492668514605612, "clip_ratio/low_mean": 8.609871438238769e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003353655629325658, "completions/clipped_ratio": 0.0, "completions/max_length": 155.5, "completions/mean_length": 74.32421875, "completions/min_length": 31.5, "epoch": 1.5680473372781065, "frac_reward_zero_std": 0.0, "grad_norm": 2.5625, "kl": 0.012028724513947963, "learning_rate": 8.21747591536527e-07, "loss": 0.0006047963164746761, "reward": 1.642932653427124, "reward_std": 0.20154935121536255, "rewards/FidelityReward/mean": 0.7617546617984772, "rewards/FidelityReward/std": 0.20378488302230835, "rewards/JudgeFidelityReward/mean": 0.7643091678619385, "rewards/JudgeFidelityReward/std": 0.18813332170248032, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2120 }, { "clip_ratio/high_max": 0.0005015060072764754, "clip_ratio/high_mean": 0.00011295831063762307, "clip_ratio/low_mean": 3.7862508906982836e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00015082081663422286, "completions/clipped_ratio": 0.0, "completions/max_length": 180.66666666666666, "completions/mean_length": 76.89322916666667, "completions/min_length": 33.666666666666664, "epoch": 1.5717455621301775, "frac_reward_zero_std": 0.0, "grad_norm": 2.734375, "kl": 0.011727142333984374, "learning_rate": 8.208104992827232e-07, "loss": 0.0003252547699958086, "reward": 1.67526113986969, "reward_std": 0.18050616482893625, "rewards/FidelityReward/mean": 0.7832001646359762, "rewards/FidelityReward/std": 0.20171943803628287, "rewards/JudgeFidelityReward/mean": 0.7841219504674276, "rewards/JudgeFidelityReward/std": 0.20126879215240479, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2125 }, { "clip_ratio/high_max": 0.0006192925502546132, "clip_ratio/high_mean": 0.00022113515005912633, "clip_ratio/low_mean": 2.3540489200968297e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002446756378049031, "completions/clipped_ratio": 0.0, "completions/max_length": 212.5, "completions/mean_length": 75.505859375, "completions/min_length": 30.0, "epoch": 1.5754437869822486, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.010481678321957589, "learning_rate": 8.198714877083917e-07, "loss": 0.0002777485642582178, "reward": 1.6266682744026184, "reward_std": 0.1908118724822998, "rewards/FidelityReward/mean": 0.7552718818187714, "rewards/FidelityReward/std": 0.20079892873764038, "rewards/JudgeFidelityReward/mean": 0.7427927851676941, "rewards/JudgeFidelityReward/std": 0.19603324681520462, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2130 }, { "clip_ratio/high_max": 0.0005891635548323393, "clip_ratio/high_mean": 0.00017193230451084673, "clip_ratio/low_mean": 2.50752258580178e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019700752454809844, "completions/clipped_ratio": 0.0, "completions/max_length": 191.33333333333334, "completions/mean_length": 76.33723958333333, "completions/min_length": 36.333333333333336, "epoch": 1.5791420118343196, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.010995789244771004, "learning_rate": 8.189305624313792e-07, "loss": 0.00045791477896273136, "reward": 1.5991667906443279, "reward_std": 0.20268866419792175, "rewards/FidelityReward/mean": 0.7341188788414001, "rewards/FidelityReward/std": 0.2022258291641871, "rewards/JudgeFidelityReward/mean": 0.7313980261484782, "rewards/JudgeFidelityReward/std": 0.18922736247380575, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2135 }, { "clip_ratio/high_max": 0.0007696954999119043, "clip_ratio/high_mean": 0.0002536792715545744, "clip_ratio/low_mean": 5.792757146991789e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00031160684302449225, "completions/clipped_ratio": 0.0, "completions/max_length": 196.0, "completions/mean_length": 76.24609375, "completions/min_length": 33.5, "epoch": 1.5828402366863905, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.012933751940727234, "learning_rate": 8.179877290809812e-07, "loss": 0.0004505438730120659, "reward": 1.5900201201438904, "reward_std": 0.20303800702095032, "rewards/FidelityReward/mean": 0.7282823324203491, "rewards/FidelityReward/std": 0.2255076766014099, "rewards/JudgeFidelityReward/mean": 0.7273817360401154, "rewards/JudgeFidelityReward/std": 0.21805469691753387, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 2140 }, { "clip_ratio/high_max": 0.0005192063399590552, "clip_ratio/high_mean": 8.592686644988134e-05, "clip_ratio/low_mean": 2.3741689801681787e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010966855625156313, "completions/clipped_ratio": 0.0, "completions/max_length": 200.0, "completions/mean_length": 75.07291666666667, "completions/min_length": 31.0, "epoch": 1.5865384615384617, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.011709142848849297, "learning_rate": 8.170429932979095e-07, "loss": 0.000353536382317543, "reward": 1.6815732320149739, "reward_std": 0.18888924022515616, "rewards/FidelityReward/mean": 0.7862626314163208, "rewards/FidelityReward/std": 0.1910716692606608, "rewards/JudgeFidelityReward/mean": 0.7906212011973063, "rewards/JudgeFidelityReward/std": 0.18243961532910666, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2145 }, { "clip_ratio/high_max": 0.0009469572338275612, "clip_ratio/high_mean": 0.00020428431016625837, "clip_ratio/low_mean": 0.00010477189207449555, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003090562007855624, "completions/clipped_ratio": 0.0, "completions/max_length": 233.5, "completions/mean_length": 75.787109375, "completions/min_length": 36.0, "epoch": 1.5902366863905324, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.011978976242244244, "learning_rate": 8.160963607342568e-07, "loss": 0.0005779068451374769, "reward": 1.6786439418792725, "reward_std": 0.17429593205451965, "rewards/FidelityReward/mean": 0.7821184694766998, "rewards/FidelityReward/std": 0.19384203851222992, "rewards/JudgeFidelityReward/mean": 0.79500412940979, "rewards/JudgeFidelityReward/std": 0.1802501678466797, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2150 }, { "clip_ratio/high_max": 0.0003713079262524843, "clip_ratio/high_mean": 8.316552266478539e-05, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 8.316552266478539e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 187.66666666666666, "completions/mean_length": 73.80338541666667, "completions/min_length": 38.666666666666664, "epoch": 1.5939349112426036, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.5, "kl": 0.011926765367388726, "learning_rate": 8.151478370534641e-07, "loss": 0.0004900950938463211, "reward": 1.624138633410136, "reward_std": 0.18960006535053253, "rewards/FidelityReward/mean": 0.7510520617167155, "rewards/FidelityReward/std": 0.20870518684387207, "rewards/JudgeFidelityReward/mean": 0.7461731235186259, "rewards/JudgeFidelityReward/std": 0.20581288635730743, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2155 }, { "clip_ratio/high_max": 0.0011796497972682118, "clip_ratio/high_mean": 0.0003058228379813954, "clip_ratio/low_mean": 0.0001398825232172385, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00044570537866093216, "completions/clipped_ratio": 0.0, "completions/max_length": 210.0, "completions/mean_length": 73.75, "completions/min_length": 38.5, "epoch": 1.5976331360946747, "frac_reward_zero_std": 0.03125, "grad_norm": 2.484375, "kl": 0.011955447308719159, "learning_rate": 8.141974279302863e-07, "loss": 0.0004782772157341242, "reward": 1.5626704692840576, "reward_std": 0.19720109552145004, "rewards/FidelityReward/mean": 0.6937542855739594, "rewards/FidelityReward/std": 0.241646908223629, "rewards/JudgeFidelityReward/mean": 0.7378323376178741, "rewards/JudgeFidelityReward/std": 0.2048715278506279, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2160 }, { "clip_ratio/high_max": 0.0003496779245324433, "clip_ratio/high_mean": 7.979345100466162e-05, "clip_ratio/low_mean": 2.0610057981684803e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010040350898634642, "completions/clipped_ratio": 0.0, "completions/max_length": 176.33333333333334, "completions/mean_length": 73.04166666666667, "completions/min_length": 30.0, "epoch": 1.6013313609467454, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.640625, "kl": 0.012675272487103938, "learning_rate": 8.132451390507585e-07, "loss": 0.0006733425892889499, "reward": 1.6665188074111938, "reward_std": 0.19598823289076486, "rewards/FidelityReward/mean": 0.7827869057655334, "rewards/FidelityReward/std": 0.17506609857082367, "rewards/JudgeFidelityReward/mean": 0.768765946229299, "rewards/JudgeFidelityReward/std": 0.17672459284464517, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2165 }, { "clip_ratio/high_max": 0.000579196575563401, "clip_ratio/high_mean": 0.00021141732577234507, "clip_ratio/low_mean": 0.00011394108878448606, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00032535840291529893, "completions/clipped_ratio": 0.0, "completions/max_length": 167.0, "completions/mean_length": 73.03515625, "completions/min_length": 35.0, "epoch": 1.6050295857988166, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.012552976235747337, "learning_rate": 8.122909761121618e-07, "loss": 0.0006701928563416004, "reward": 1.5814845561981201, "reward_std": 0.21369380503892899, "rewards/FidelityReward/mean": 0.7202996015548706, "rewards/FidelityReward/std": 0.21392203122377396, "rewards/JudgeFidelityReward/mean": 0.7223700284957886, "rewards/JudgeFidelityReward/std": 0.20887455344200134, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2170 }, { "clip_ratio/high_max": 0.000595382577739656, "clip_ratio/high_mean": 0.00018713148310780525, "clip_ratio/low_mean": 1.9592476019170135e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020672396058216692, "completions/clipped_ratio": 0.0, "completions/max_length": 210.33333333333334, "completions/mean_length": 76.23828125, "completions/min_length": 33.666666666666664, "epoch": 1.6087278106508875, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.01249520480632782, "learning_rate": 8.113349448229893e-07, "loss": 0.00036753821186721327, "reward": 1.6107951005299885, "reward_std": 0.19886305431524912, "rewards/FidelityReward/mean": 0.7374952038129171, "rewards/FidelityReward/std": 0.2184688299894333, "rewards/JudgeFidelityReward/mean": 0.7465998729070028, "rewards/JudgeFidelityReward/std": 0.19966773688793182, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2175 }, { "clip_ratio/high_max": 0.0008094593533314765, "clip_ratio/high_mean": 0.00022669540485367178, "clip_ratio/low_mean": 6.173224974190817e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002884276560507715, "completions/clipped_ratio": 0.0, "completions/max_length": 219.5, "completions/mean_length": 72.927734375, "completions/min_length": 38.0, "epoch": 1.6124260355029585, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.012023655138909817, "learning_rate": 8.103770509029118e-07, "loss": 0.0002528479788452387, "reward": 1.668042242527008, "reward_std": 0.18997278809547424, "rewards/FidelityReward/mean": 0.7800480127334595, "rewards/FidelityReward/std": 0.20874228328466415, "rewards/JudgeFidelityReward/mean": 0.7779417335987091, "rewards/JudgeFidelityReward/std": 0.20505788922309875, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2180 }, { "clip_ratio/high_max": 0.00018018018454313278, "clip_ratio/high_mean": 6.292816251516343e-05, "clip_ratio/low_mean": 4.3441011803224684e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010636917140800506, "completions/clipped_ratio": 0.0, "completions/max_length": 194.0, "completions/mean_length": 74.18359375, "completions/min_length": 39.333333333333336, "epoch": 1.6161242603550297, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.359375, "kl": 0.011148930341005326, "learning_rate": 8.094173000827439e-07, "loss": 0.00044738389551639556, "reward": 1.6703252792358398, "reward_std": 0.18302145103613535, "rewards/FidelityReward/mean": 0.7819477121035258, "rewards/FidelityReward/std": 0.18248343467712402, "rewards/JudgeFidelityReward/mean": 0.776755134264628, "rewards/JudgeFidelityReward/std": 0.17539903024832407, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2185 }, { "clip_ratio/high_max": 0.0005617429269477725, "clip_ratio/high_mean": 0.00024370057508349419, "clip_ratio/low_mean": 3.8334265991579744e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028203484835103153, "completions/clipped_ratio": 0.0, "completions/max_length": 175.5, "completions/mean_length": 72.169921875, "completions/min_length": 34.5, "epoch": 1.6198224852071006, "frac_reward_zero_std": 0.015625, "grad_norm": 2.4375, "kl": 0.01230892315506935, "learning_rate": 8.084556981044095e-07, "loss": 0.0006476951763033867, "reward": 1.73720383644104, "reward_std": 0.1645914763212204, "rewards/FidelityReward/mean": 0.8244397044181824, "rewards/FidelityReward/std": 0.16752400994300842, "rewards/JudgeFidelityReward/mean": 0.8255283832550049, "rewards/JudgeFidelityReward/std": 0.16750307381153107, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2190 }, { "clip_ratio/high_max": 0.0007262147031724453, "clip_ratio/high_mean": 0.00023703230544924736, "clip_ratio/low_mean": 6.613472942262888e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030316703487187624, "completions/clipped_ratio": 0.0, "completions/max_length": 185.0, "completions/mean_length": 72.90494791666667, "completions/min_length": 35.0, "epoch": 1.6235207100591715, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.012804779037833215, "learning_rate": 8.074922507209071e-07, "loss": 0.0006649234797805548, "reward": 1.6758151451746623, "reward_std": 0.1732485145330429, "rewards/FidelityReward/mean": 0.7823334137598673, "rewards/FidelityReward/std": 0.19216786324977875, "rewards/JudgeFidelityReward/mean": 0.7869635025660197, "rewards/JudgeFidelityReward/std": 0.18441210687160492, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2195 }, { "clip_ratio/high_max": 0.00047101941891014576, "clip_ratio/high_mean": 0.0001436899707186967, "clip_ratio/low_mean": 0.00015862177970120682, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030231174314394594, "completions/clipped_ratio": 0.0, "completions/max_length": 185.5, "completions/mean_length": 68.818359375, "completions/min_length": 34.5, "epoch": 1.6272189349112427, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.011717425659298897, "learning_rate": 8.065269636962763e-07, "loss": 0.0004453688394278288, "reward": 1.6659364104270935, "reward_std": 0.18220071494579315, "rewards/FidelityReward/mean": 0.7793326079845428, "rewards/FidelityReward/std": 0.19289222359657288, "rewards/JudgeFidelityReward/mean": 0.7732076346874237, "rewards/JudgeFidelityReward/std": 0.18453752249479294, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2200 }, { "clip_ratio/high_max": 0.0005030791740864516, "clip_ratio/high_mean": 0.00011807731352746487, "clip_ratio/low_mean": 7.816431898390874e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019624163396656514, "completions/clipped_ratio": 0.0, "completions/max_length": 185.33333333333334, "completions/mean_length": 73.86328125, "completions/min_length": 36.666666666666664, "epoch": 1.6309171597633136, "frac_reward_zero_std": 0.0, "grad_norm": 2.734375, "kl": 0.011401899717748165, "learning_rate": 8.055598428055626e-07, "loss": 0.0005222068168222905, "reward": 1.6763519843419392, "reward_std": 0.18773426612218222, "rewards/FidelityReward/mean": 0.7821557720502218, "rewards/FidelityReward/std": 0.2076537013053894, "rewards/JudgeFidelityReward/mean": 0.7896944483121237, "rewards/JudgeFidelityReward/std": 0.19973130524158478, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2205 }, { "clip_ratio/high_max": 0.0007109617465175689, "clip_ratio/high_mean": 0.0002152059692889452, "clip_ratio/low_mean": 0.00012188536929897964, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00033709133276715875, "completions/clipped_ratio": 0.0, "completions/max_length": 189.5, "completions/mean_length": 70.361328125, "completions/min_length": 32.5, "epoch": 1.6346153846153846, "frac_reward_zero_std": 0.015625, "grad_norm": 2.75, "kl": 0.013584098592400552, "learning_rate": 8.045908938347827e-07, "loss": 0.00039532240480184556, "reward": 1.6400333642959595, "reward_std": 0.18035143613815308, "rewards/FidelityReward/mean": 0.7523148357868195, "rewards/FidelityReward/std": 0.21934422105550766, "rewards/JudgeFidelityReward/mean": 0.7754369676113129, "rewards/JudgeFidelityReward/std": 0.19389712810516357, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2210 }, { "clip_ratio/high_max": 0.0007563726976513863, "clip_ratio/high_mean": 0.0001894232875201851, "clip_ratio/low_mean": 2.024291461566463e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002096662123221904, "completions/clipped_ratio": 0.0, "completions/max_length": 162.33333333333334, "completions/mean_length": 71.6484375, "completions/min_length": 32.333333333333336, "epoch": 1.6383136094674557, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.40625, "kl": 0.010586144216358662, "learning_rate": 8.036201225808906e-07, "loss": 0.0003662455594167113, "reward": 1.6239595413208008, "reward_std": 0.18312417964140573, "rewards/FidelityReward/mean": 0.7480070193608602, "rewards/FidelityReward/std": 0.1888657013575236, "rewards/JudgeFidelityReward/mean": 0.7532070477803549, "rewards/JudgeFidelityReward/std": 0.1849627047777176, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2215 }, { "clip_ratio/high_max": 0.0009962045936845244, "clip_ratio/high_mean": 0.00023811317514628173, "clip_ratio/low_mean": 5.659006128553301e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002947032335214317, "completions/clipped_ratio": 0.0, "completions/max_length": 210.5, "completions/mean_length": 73.71484375, "completions/min_length": 26.5, "epoch": 1.6420118343195265, "frac_reward_zero_std": 0.015625, "grad_norm": 2.90625, "kl": 0.011051422171294688, "learning_rate": 8.026475348517422e-07, "loss": 0.00044266628101468085, "reward": 1.6588531732559204, "reward_std": 0.18424246460199356, "rewards/FidelityReward/mean": 0.7643622159957886, "rewards/FidelityReward/std": 0.19392383098602295, "rewards/JudgeFidelityReward/mean": 0.7889819145202637, "rewards/JudgeFidelityReward/std": 0.16150099784135818, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2220 }, { "clip_ratio/high_max": 0.0008482714649289847, "clip_ratio/high_mean": 0.00023436396149918438, "clip_ratio/low_mean": 4.110269364900887e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027546666096895934, "completions/clipped_ratio": 0.0, "completions/max_length": 173.0, "completions/mean_length": 70.9765625, "completions/min_length": 34.333333333333336, "epoch": 1.6457100591715976, "frac_reward_zero_std": 0.0, "grad_norm": 2.375, "kl": 0.013079091347754002, "learning_rate": 8.016731364660611e-07, "loss": 0.0005095534957945347, "reward": 1.7072814305623372, "reward_std": 0.15626643598079681, "rewards/FidelityReward/mean": 0.7998758753140768, "rewards/FidelityReward/std": 0.20912827054659525, "rewards/JudgeFidelityReward/mean": 0.8161131938298544, "rewards/JudgeFidelityReward/std": 0.18950203557809195, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2225 }, { "clip_ratio/high_max": 0.0007859298260882497, "clip_ratio/high_mean": 0.00020802382787223905, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020802382787223905, "completions/clipped_ratio": 0.0, "completions/max_length": 183.0, "completions/mean_length": 71.541015625, "completions/min_length": 34.5, "epoch": 1.6494082840236688, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.014358017034828663, "learning_rate": 8.006969332534033e-07, "loss": 0.0006488943472504616, "reward": 1.6075260639190674, "reward_std": 0.1976114958524704, "rewards/FidelityReward/mean": 0.7407039403915405, "rewards/FidelityReward/std": 0.23508641868829727, "rewards/JudgeFidelityReward/mean": 0.7355974018573761, "rewards/JudgeFidelityReward/std": 0.22929487377405167, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2230 }, { "clip_ratio/high_max": 0.0008572377264499664, "clip_ratio/high_mean": 0.0002743353834375739, "clip_ratio/low_mean": 2.200704184360802e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00029634241946041584, "completions/clipped_ratio": 0.0, "completions/max_length": 195.66666666666666, "completions/mean_length": 72.94010416666667, "completions/min_length": 32.333333333333336, "epoch": 1.6531065088757395, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.012537907809019089, "learning_rate": 7.997189310541229e-07, "loss": 0.0005475115962326526, "reward": 1.6704711119333904, "reward_std": 0.17981771131356558, "rewards/FidelityReward/mean": 0.7797940174738566, "rewards/FidelityReward/std": 0.2004112203915914, "rewards/JudgeFidelityReward/mean": 0.7839584350585938, "rewards/JudgeFidelityReward/std": 0.19252288341522217, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 2235 }, { "clip_ratio/high_max": 0.0010087399976328015, "clip_ratio/high_mean": 0.00021102247410453856, "clip_ratio/low_mean": 0.00011148544581374153, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003225079039111733, "completions/clipped_ratio": 0.0, "completions/max_length": 248.5, "completions/mean_length": 71.0078125, "completions/min_length": 35.5, "epoch": 1.6568047337278107, "frac_reward_zero_std": 0.015625, "grad_norm": 2.546875, "kl": 0.013875212892889977, "learning_rate": 7.987391357193368e-07, "loss": 0.0006791878957301378, "reward": 1.6807788014411926, "reward_std": 0.18488545715808868, "rewards/FidelityReward/mean": 0.7891458570957184, "rewards/FidelityReward/std": 0.1847512423992157, "rewards/JudgeFidelityReward/mean": 0.7832657396793365, "rewards/JudgeFidelityReward/std": 0.17690350860357285, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2240 }, { "clip_ratio/high_max": 0.0005277571151964366, "clip_ratio/high_mean": 0.0001468327216571197, "clip_ratio/low_mean": 6.065453635528684e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020748725801240654, "completions/clipped_ratio": 0.0, "completions/max_length": 197.66666666666666, "completions/mean_length": 71.03255208333333, "completions/min_length": 34.666666666666664, "epoch": 1.6605029585798816, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.671875, "kl": 0.01241771224886179, "learning_rate": 7.977575531108893e-07, "loss": 0.0005050183739513159, "reward": 1.6717615922292073, "reward_std": 0.162299374739329, "rewards/FidelityReward/mean": 0.7811318834622701, "rewards/FidelityReward/std": 0.1992988387743632, "rewards/JudgeFidelityReward/mean": 0.7812595764795939, "rewards/JudgeFidelityReward/std": 0.18679226438204447, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2245 }, { "clip_ratio/high_max": 0.0008744463906623423, "clip_ratio/high_mean": 0.00017031135212164372, "clip_ratio/low_mean": 0.00010656785307219251, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002768792037386447, "completions/clipped_ratio": 0.0, "completions/max_length": 209.5, "completions/mean_length": 74.630859375, "completions/min_length": 34.5, "epoch": 1.6642011834319526, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.012348450161516667, "learning_rate": 7.967741891013179e-07, "loss": 0.0002696950919926167, "reward": 1.6464470028877258, "reward_std": 0.1757146343588829, "rewards/FidelityReward/mean": 0.7642033398151398, "rewards/FidelityReward/std": 0.22066260129213333, "rewards/JudgeFidelityReward/mean": 0.7644874155521393, "rewards/JudgeFidelityReward/std": 0.21402312070131302, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2250 }, { "clip_ratio/high_max": 0.0004977985750883818, "clip_ratio/high_mean": 0.00012955237762071192, "clip_ratio/low_mean": 1.9011406402569263e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001485637796577066, "completions/clipped_ratio": 0.0, "completions/max_length": 173.66666666666666, "completions/mean_length": 72.02213541666667, "completions/min_length": 34.666666666666664, "epoch": 1.6678994082840237, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.515625, "kl": 0.013293232396245002, "learning_rate": 7.957890495738178e-07, "loss": 0.0006353676784783602, "reward": 1.658272345860799, "reward_std": 0.17670042316118875, "rewards/FidelityReward/mean": 0.7723659078280131, "rewards/FidelityReward/std": 0.20154570043087006, "rewards/JudgeFidelityReward/mean": 0.7718130548795065, "rewards/JudgeFidelityReward/std": 0.1866098940372467, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2255 }, { "clip_ratio/high_max": 0.0005246552871540188, "clip_ratio/high_mean": 0.0001054807216860354, "clip_ratio/low_mean": 5.736002349294722e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016284074517898263, "completions/clipped_ratio": 0.0, "completions/max_length": 183.5, "completions/mean_length": 75.705078125, "completions/min_length": 33.0, "epoch": 1.6715976331360947, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.012761964276432992, "learning_rate": 7.948021404222062e-07, "loss": 0.000486360490322113, "reward": 1.629064679145813, "reward_std": 0.19818686693906784, "rewards/FidelityReward/mean": 0.7521772980690002, "rewards/FidelityReward/std": 0.20432468503713608, "rewards/JudgeFidelityReward/mean": 0.7557278275489807, "rewards/JudgeFidelityReward/std": 0.19501734524965286, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2260 }, { "clip_ratio/high_max": 0.0005403229966759682, "clip_ratio/high_mean": 0.00013016255397815257, "clip_ratio/low_mean": 4.119521763641387e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017135777161456645, "completions/clipped_ratio": 0.0, "completions/max_length": 168.33333333333334, "completions/mean_length": 72.64453125, "completions/min_length": 34.666666666666664, "epoch": 1.6752958579881656, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.765625, "kl": 0.010767274536192417, "learning_rate": 7.938134675508879e-07, "loss": 0.0004006542265415192, "reward": 1.5938151280085247, "reward_std": 0.18718049923578897, "rewards/FidelityReward/mean": 0.7260281443595886, "rewards/FidelityReward/std": 0.22918435434500375, "rewards/JudgeFidelityReward/mean": 0.7355738679567972, "rewards/JudgeFidelityReward/std": 0.21455774207909903, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2265 }, { "clip_ratio/high_max": 0.0007560304482467472, "clip_ratio/high_mean": 0.00014000649243826047, "clip_ratio/low_mean": 8.27524927444756e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022275898372754456, "completions/clipped_ratio": 0.0, "completions/max_length": 146.5, "completions/mean_length": 71.837890625, "completions/min_length": 38.5, "epoch": 1.6789940828402368, "frac_reward_zero_std": 0.015625, "grad_norm": 2.640625, "kl": 0.011924485862255096, "learning_rate": 7.928230368748194e-07, "loss": 0.000487612746655941, "reward": 1.6355215311050415, "reward_std": 0.18426620215177536, "rewards/FidelityReward/mean": 0.7571120858192444, "rewards/FidelityReward/std": 0.20876643806695938, "rewards/JudgeFidelityReward/mean": 0.756818950176239, "rewards/JudgeFidelityReward/std": 0.21790479868650436, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2270 }, { "clip_ratio/high_max": 0.0003835316631011665, "clip_ratio/high_mean": 0.00010520091163925826, "clip_ratio/low_mean": 6.659941282123327e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017180031863972545, "completions/clipped_ratio": 0.0, "completions/max_length": 200.33333333333334, "completions/mean_length": 72.83463541666667, "completions/min_length": 37.333333333333336, "epoch": 1.6826923076923077, "frac_reward_zero_std": 0.0, "grad_norm": 2.859375, "kl": 0.013093536347150802, "learning_rate": 7.918308543194734e-07, "loss": 0.0004471071064472198, "reward": 1.645495096842448, "reward_std": 0.1941495289405187, "rewards/FidelityReward/mean": 0.768841822942098, "rewards/FidelityReward/std": 0.20330172777175903, "rewards/JudgeFidelityReward/mean": 0.7559106945991516, "rewards/JudgeFidelityReward/std": 0.205514927705129, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 2275 }, { "clip_ratio/high_max": 0.0006494143279269337, "clip_ratio/high_mean": 0.00025108036352321504, "clip_ratio/low_mean": 4.96620312333107e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030074239475652575, "completions/clipped_ratio": 0.0, "completions/max_length": 165.5, "completions/mean_length": 73.7578125, "completions/min_length": 30.5, "epoch": 1.6863905325443787, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.012649283185601235, "learning_rate": 7.90836925820804e-07, "loss": 0.0003051570151001215, "reward": 1.6248014569282532, "reward_std": 0.19490456581115723, "rewards/FidelityReward/mean": 0.7530129551887512, "rewards/FidelityReward/std": 0.21225223690271378, "rewards/JudgeFidelityReward/mean": 0.7435770034790039, "rewards/JudgeFidelityReward/std": 0.20391332358121872, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2280 }, { "clip_ratio/high_max": 0.0006704486208036542, "clip_ratio/high_mean": 0.00014469407615251838, "clip_ratio/low_mean": 6.993266142671928e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002146267332136631, "completions/clipped_ratio": 0.0, "completions/max_length": 189.66666666666666, "completions/mean_length": 73.16927083333333, "completions/min_length": 32.0, "epoch": 1.6900887573964498, "frac_reward_zero_std": 0.0, "grad_norm": 2.3125, "kl": 0.011651387996971607, "learning_rate": 7.898412573252105e-07, "loss": 0.00043403152376413344, "reward": 1.6054766575495403, "reward_std": 0.2068552722533544, "rewards/FidelityReward/mean": 0.7341117262840271, "rewards/FidelityReward/std": 0.21623802681763968, "rewards/JudgeFidelityReward/mean": 0.7440319458643595, "rewards/JudgeFidelityReward/std": 0.21165057520071665, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2285 }, { "clip_ratio/high_max": 0.00097140206489712, "clip_ratio/high_mean": 0.00027359247324056926, "clip_ratio/low_mean": 4.585442366078496e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000319446885259822, "completions/clipped_ratio": 0.0, "completions/max_length": 155.5, "completions/mean_length": 74.0546875, "completions/min_length": 38.5, "epoch": 1.6937869822485208, "frac_reward_zero_std": 0.015625, "grad_norm": 2.453125, "kl": 0.013307995349168777, "learning_rate": 7.888438547895025e-07, "loss": 0.0006554616615176201, "reward": 1.6973459720611572, "reward_std": 0.16803914308547974, "rewards/FidelityReward/mean": 0.794266402721405, "rewards/FidelityReward/std": 0.17348623275756836, "rewards/JudgeFidelityReward/mean": 0.8081121742725372, "rewards/JudgeFidelityReward/std": 0.159094899892807, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2290 }, { "clip_ratio/high_max": 0.0003636663663201034, "clip_ratio/high_mean": 0.00012927131610922514, "clip_ratio/low_mean": 4.4247822370380166e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017351913847960532, "completions/clipped_ratio": 0.0, "completions/max_length": 168.66666666666666, "completions/mean_length": 73.11197916666667, "completions/min_length": 35.666666666666664, "epoch": 1.6974852071005917, "frac_reward_zero_std": 0.03125, "grad_norm": 2.421875, "kl": 0.012054580636322499, "learning_rate": 7.878447241808634e-07, "loss": 0.0005348841194063425, "reward": 1.615676760673523, "reward_std": 0.18615432580312094, "rewards/FidelityReward/mean": 0.7410521904627482, "rewards/FidelityReward/std": 0.2298070341348648, "rewards/JudgeFidelityReward/mean": 0.7492491404215494, "rewards/JudgeFidelityReward/std": 0.22314300139745077, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2295 }, { "clip_ratio/high_max": 0.0012188060907647013, "clip_ratio/high_mean": 0.00036423265701159837, "clip_ratio/low_mean": 9.466616320423782e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00045889881439507005, "completions/clipped_ratio": 0.0, "completions/max_length": 158.0, "completions/mean_length": 74.810546875, "completions/min_length": 33.0, "epoch": 1.7011834319526629, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.013428370468318462, "learning_rate": 7.868438714768154e-07, "loss": 0.00047125625424087046, "reward": 1.6114214062690735, "reward_std": 0.19854101538658142, "rewards/FidelityReward/mean": 0.7384854853153229, "rewards/FidelityReward/std": 0.21654562652111053, "rewards/JudgeFidelityReward/mean": 0.7458717823028564, "rewards/JudgeFidelityReward/std": 0.21547479182481766, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2300 }, { "clip_ratio/high_max": 0.0003686662530526519, "clip_ratio/high_mean": 0.00012697749771177768, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012697749771177768, "completions/clipped_ratio": 0.0, "completions/max_length": 160.66666666666666, "completions/mean_length": 72.92317708333333, "completions/min_length": 33.0, "epoch": 1.7048816568047336, "frac_reward_zero_std": 0.03125, "grad_norm": 2.75, "kl": 0.013161182962357998, "learning_rate": 7.858413026651838e-07, "loss": 0.00044174394570291043, "reward": 1.6552420059839885, "reward_std": 0.18348326285680136, "rewards/FidelityReward/mean": 0.7739113370577494, "rewards/FidelityReward/std": 0.19154317677021027, "rewards/JudgeFidelityReward/mean": 0.7626614173253378, "rewards/JudgeFidelityReward/std": 0.20250058670838675, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2305 }, { "clip_ratio/high_max": 0.0011946623679250478, "clip_ratio/high_mean": 0.0003200415289029479, "clip_ratio/low_mean": 4.605051217367873e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000366092030890286, "completions/clipped_ratio": 0.0, "completions/max_length": 156.0, "completions/mean_length": 73.744140625, "completions/min_length": 36.0, "epoch": 1.7085798816568047, "frac_reward_zero_std": 0.015625, "grad_norm": 2.40625, "kl": 0.011540783196687698, "learning_rate": 7.848370237440602e-07, "loss": 0.00040162140503525735, "reward": 1.6138458847999573, "reward_std": 0.1955113783478737, "rewards/FidelityReward/mean": 0.7409988641738892, "rewards/FidelityReward/std": 0.2032829001545906, "rewards/JudgeFidelityReward/mean": 0.745694100856781, "rewards/JudgeFidelityReward/std": 0.18365230411291122, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2310 }, { "clip_ratio/high_max": 0.0005465886322781443, "clip_ratio/high_mean": 0.00017149318591691554, "clip_ratio/low_mean": 0.00010110459115821868, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002725977799855173, "completions/clipped_ratio": 0.0, "completions/max_length": 169.66666666666666, "completions/mean_length": 73.77734375, "completions/min_length": 33.333333333333336, "epoch": 1.7122781065088757, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.65625, "kl": 0.01327810250222683, "learning_rate": 7.838310407217683e-07, "loss": 0.0006301486864686012, "reward": 1.6605594555536907, "reward_std": 0.18364259600639343, "rewards/FidelityReward/mean": 0.7739578485488892, "rewards/FidelityReward/std": 0.20847741266091666, "rewards/JudgeFidelityReward/mean": 0.7732032934824625, "rewards/JudgeFidelityReward/std": 0.20211681723594666, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2315 }, { "clip_ratio/high_max": 0.001049992861226201, "clip_ratio/high_mean": 0.0002979888638947159, "clip_ratio/low_mean": 2.0781380590051414e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003187702444847673, "completions/clipped_ratio": 0.0, "completions/max_length": 159.0, "completions/mean_length": 75.150390625, "completions/min_length": 37.5, "epoch": 1.7159763313609466, "frac_reward_zero_std": 0.0, "grad_norm": 2.34375, "kl": 0.013914937525987625, "learning_rate": 7.828233596168261e-07, "loss": 0.0008519654162228108, "reward": 1.5869240760803223, "reward_std": 0.19270866364240646, "rewards/FidelityReward/mean": 0.7224871516227722, "rewards/FidelityReward/std": 0.2538570836186409, "rewards/JudgeFidelityReward/mean": 0.7288738191127777, "rewards/JudgeFidelityReward/std": 0.24203834682703018, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2320 }, { "clip_ratio/high_max": 0.0005626477533951402, "clip_ratio/high_mean": 0.0001725221110973507, "clip_ratio/low_mean": 4.869319673161954e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002212153165601194, "completions/clipped_ratio": 0.0, "completions/max_length": 191.0, "completions/mean_length": 75.01822916666667, "completions/min_length": 37.333333333333336, "epoch": 1.7196745562130178, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.013618437573313713, "learning_rate": 7.818139864579111e-07, "loss": 0.0005832464434206486, "reward": 1.6348588069279988, "reward_std": 0.19269648691018423, "rewards/FidelityReward/mean": 0.7541929284731547, "rewards/FidelityReward/std": 0.21542072296142578, "rewards/JudgeFidelityReward/mean": 0.7626339395840963, "rewards/JudgeFidelityReward/std": 0.2090092400709788, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2325 }, { "clip_ratio/high_max": 0.0012397544924169779, "clip_ratio/high_mean": 0.00024233623989857733, "clip_ratio/low_mean": 0.00014985499146860092, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003921912342775613, "completions/clipped_ratio": 0.0, "completions/max_length": 212.0, "completions/mean_length": 75.189453125, "completions/min_length": 28.0, "epoch": 1.7233727810650887, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.013282976485788823, "learning_rate": 7.808029272838237e-07, "loss": 0.0005237886216491461, "reward": 1.5972346067428589, "reward_std": 0.17329144477844238, "rewards/FidelityReward/mean": 0.7319420576095581, "rewards/FidelityReward/std": 0.2599789500236511, "rewards/JudgeFidelityReward/mean": 0.7364444732666016, "rewards/JudgeFidelityReward/std": 0.2633616775274277, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.0539139099419117, "step": 2330 }, { "clip_ratio/high_max": 0.0005700866924598813, "clip_ratio/high_mean": 0.00010893331491388381, "clip_ratio/low_mean": 4.4067131238989535e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001530004432424903, "completions/clipped_ratio": 0.0, "completions/max_length": 166.66666666666666, "completions/mean_length": 71.89453125, "completions/min_length": 34.0, "epoch": 1.7270710059171597, "frac_reward_zero_std": 0.041666666666666664, "grad_norm": 2.15625, "kl": 0.014265795052051545, "learning_rate": 7.797901881434517e-07, "loss": 0.0004853149875998497, "reward": 1.6738899151484172, "reward_std": 0.17379646499951681, "rewards/FidelityReward/mean": 0.780139148235321, "rewards/FidelityReward/std": 0.21129045387109122, "rewards/JudgeFidelityReward/mean": 0.7875014940897623, "rewards/JudgeFidelityReward/std": 0.19469221929709116, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2335 }, { "clip_ratio/high_max": 0.0010891985846683383, "clip_ratio/high_mean": 0.00026934100314974786, "clip_ratio/low_mean": 6.691481976304204e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00033625580836087464, "completions/clipped_ratio": 0.0, "completions/max_length": 174.0, "completions/mean_length": 73.55078125, "completions/min_length": 35.0, "epoch": 1.7307692307692308, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.013078469038009643, "learning_rate": 7.787757750957333e-07, "loss": 0.0005434597842395306, "reward": 1.5879745483398438, "reward_std": 0.1888987421989441, "rewards/FidelityReward/mean": 0.7258987724781036, "rewards/FidelityReward/std": 0.2534923776984215, "rewards/JudgeFidelityReward/mean": 0.7241514325141907, "rewards/JudgeFidelityReward/std": 0.2495705485343933, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2340 }, { "clip_ratio/high_max": 0.000377181323710829, "clip_ratio/high_mean": 0.00019266726449131965, "clip_ratio/low_mean": 2.3430178407579662e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002160974428988993, "completions/clipped_ratio": 0.0, "completions/max_length": 193.66666666666666, "completions/mean_length": 72.63020833333333, "completions/min_length": 35.0, "epoch": 1.7344674556213018, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.484375, "kl": 0.014657184109091758, "learning_rate": 7.777596942096213e-07, "loss": 0.0003567718900740147, "reward": 1.6624338229497273, "reward_std": 0.17918664713700613, "rewards/FidelityReward/mean": 0.7763787110646566, "rewards/FidelityReward/std": 0.20804694294929504, "rewards/JudgeFidelityReward/mean": 0.7721101641654968, "rewards/JudgeFidelityReward/std": 0.20584804316361746, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2345 }, { "clip_ratio/high_max": 0.0012642222689464688, "clip_ratio/high_mean": 0.00023590993951074778, "clip_ratio/low_mean": 4.1406463424209505e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027731640147976577, "completions/clipped_ratio": 0.0, "completions/max_length": 176.0, "completions/mean_length": 74.33203125, "completions/min_length": 32.5, "epoch": 1.7381656804733727, "frac_reward_zero_std": 0.015625, "grad_norm": 2.421875, "kl": 0.01343427188694477, "learning_rate": 7.767419515640468e-07, "loss": 0.0006415429990738631, "reward": 1.6234768629074097, "reward_std": 0.18350564688444138, "rewards/FidelityReward/mean": 0.7467275559902191, "rewards/FidelityReward/std": 0.2336791455745697, "rewards/JudgeFidelityReward/mean": 0.7534984946250916, "rewards/JudgeFidelityReward/std": 0.21942703425884247, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2350 }, { "clip_ratio/high_max": 0.000529328000266105, "clip_ratio/high_mean": 0.0001259455399122089, "clip_ratio/low_mean": 0.00010964978719130158, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023559534456580878, "completions/clipped_ratio": 0.0, "completions/max_length": 188.66666666666666, "completions/mean_length": 74.74609375, "completions/min_length": 35.0, "epoch": 1.7418639053254439, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.012716276571154594, "learning_rate": 7.757225532478826e-07, "loss": 0.0006290886551141739, "reward": 1.665541688601176, "reward_std": 0.19016642371813455, "rewards/FidelityReward/mean": 0.778947651386261, "rewards/FidelityReward/std": 0.185701052347819, "rewards/JudgeFidelityReward/mean": 0.7744901180267334, "rewards/JudgeFidelityReward/std": 0.18198437492052713, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2355 }, { "clip_ratio/high_max": 0.0005516160279512405, "clip_ratio/high_mean": 0.0002113770751748234, "clip_ratio/low_mean": 6.90102286171168e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002803873037919402, "completions/clipped_ratio": 0.0, "completions/max_length": 171.0, "completions/mean_length": 72.767578125, "completions/min_length": 37.5, "epoch": 1.7455621301775148, "frac_reward_zero_std": 0.015625, "grad_norm": 2.546875, "kl": 0.012625562585890293, "learning_rate": 7.747015053599071e-07, "loss": 0.0005473810248076916, "reward": 1.7063774466514587, "reward_std": 0.17034851759672165, "rewards/FidelityReward/mean": 0.808315247297287, "rewards/FidelityReward/std": 0.18806642293930054, "rewards/JudgeFidelityReward/mean": 0.7961243689060211, "rewards/JudgeFidelityReward/std": 0.1957189217209816, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2360 }, { "clip_ratio/high_max": 0.0003552397945895791, "clip_ratio/high_mean": 0.00010927480179816485, "clip_ratio/low_mean": 2.1834061772096903e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013110885629430413, "completions/clipped_ratio": 0.0, "completions/max_length": 185.0, "completions/mean_length": 74.11979166666667, "completions/min_length": 34.666666666666664, "epoch": 1.7492603550295858, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.71875, "kl": 0.014449177123606206, "learning_rate": 7.736788140087674e-07, "loss": 0.0005828531458973885, "reward": 1.682552417119344, "reward_std": 0.1900926431020101, "rewards/FidelityReward/mean": 0.7863218585650126, "rewards/FidelityReward/std": 0.1939538468917211, "rewards/JudgeFidelityReward/mean": 0.7937631805737814, "rewards/JudgeFidelityReward/std": 0.18603993952274323, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2365 }, { "clip_ratio/high_max": 0.0008280019275844098, "clip_ratio/high_mean": 0.00022917981550563127, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022917981550563127, "completions/clipped_ratio": 0.0, "completions/max_length": 158.0, "completions/mean_length": 72.021484375, "completions/min_length": 32.5, "epoch": 1.752958579881657, "frac_reward_zero_std": 0.015625, "grad_norm": 2.421875, "kl": 0.011653119698166847, "learning_rate": 7.726544853129434e-07, "loss": 0.00042364243417978285, "reward": 1.6393672227859497, "reward_std": 0.19176225364208221, "rewards/FidelityReward/mean": 0.76473468542099, "rewards/FidelityReward/std": 0.19178759306669235, "rewards/JudgeFidelityReward/mean": 0.7492650747299194, "rewards/JudgeFidelityReward/std": 0.19123730063438416, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2370 }, { "clip_ratio/high_max": 0.0006673617288470268, "clip_ratio/high_mean": 0.00021449638297781347, "clip_ratio/low_mean": 2.6371306739747523e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000240867689717561, "completions/clipped_ratio": 0.0, "completions/max_length": 216.0, "completions/mean_length": 73.17317708333333, "completions/min_length": 33.0, "epoch": 1.7566568047337277, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.015020239725708961, "learning_rate": 7.716285254007102e-07, "loss": 0.0005987937562167645, "reward": 1.6785953839619954, "reward_std": 0.17751417060693106, "rewards/FidelityReward/mean": 0.7883689999580383, "rewards/FidelityReward/std": 0.20241069296995798, "rewards/JudgeFidelityReward/mean": 0.7817549308141073, "rewards/JudgeFidelityReward/std": 0.20223400990168253, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2375 }, { "clip_ratio/high_max": 0.000593841343652457, "clip_ratio/high_mean": 0.00020321591291576623, "clip_ratio/low_mean": 0.00010590404563117772, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030911995563656094, "completions/clipped_ratio": 0.0, "completions/max_length": 171.0, "completions/mean_length": 72.86328125, "completions/min_length": 35.0, "epoch": 1.7603550295857988, "frac_reward_zero_std": 0.015625, "grad_norm": 2.453125, "kl": 0.014514859579503537, "learning_rate": 7.706009404101026e-07, "loss": 0.0008690997958183289, "reward": 1.656117856502533, "reward_std": 0.18104137480258942, "rewards/FidelityReward/mean": 0.7727361619472504, "rewards/FidelityReward/std": 0.18520447611808777, "rewards/JudgeFidelityReward/mean": 0.7667633891105652, "rewards/JudgeFidelityReward/std": 0.17765161395072937, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2380 }, { "clip_ratio/high_max": 0.00034353644587099553, "clip_ratio/high_mean": 4.294205573387444e-05, "clip_ratio/low_mean": 4.896214522887021e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 9.190420096274465e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 184.0, "completions/mean_length": 72.56901041666667, "completions/min_length": 32.666666666666664, "epoch": 1.76405325443787, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.765625, "kl": 0.013835117593407631, "learning_rate": 7.695717364888772e-07, "loss": 0.0005690657999366522, "reward": 1.6163002252578735, "reward_std": 0.19025674959023794, "rewards/FidelityReward/mean": 0.7440744241078695, "rewards/FidelityReward/std": 0.20180433988571167, "rewards/JudgeFidelityReward/mean": 0.7444514632225037, "rewards/JudgeFidelityReward/std": 0.19633782406648, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2385 }, { "clip_ratio/high_max": 0.0010067652445286512, "clip_ratio/high_mean": 0.00022644853161182254, "clip_ratio/low_mean": 5.7489587925374507e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000283938116626814, "completions/clipped_ratio": 0.0, "completions/max_length": 179.0, "completions/mean_length": 75.333984375, "completions/min_length": 35.0, "epoch": 1.7677514792899407, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.013529878482222557, "learning_rate": 7.685409197944768e-07, "loss": 0.0005321139935404063, "reward": 1.6546838879585266, "reward_std": 0.18805953115224838, "rewards/FidelityReward/mean": 0.7729235291481018, "rewards/FidelityReward/std": 0.1916942372918129, "rewards/JudgeFidelityReward/mean": 0.7635206878185272, "rewards/JudgeFidelityReward/std": 0.19014913588762283, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2390 }, { "clip_ratio/high_max": 0.0003468590788543224, "clip_ratio/high_mean": 0.00011893179034814239, "clip_ratio/low_mean": 4.556031199172139e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016449210233986377, "completions/clipped_ratio": 0.0, "completions/max_length": 200.66666666666666, "completions/mean_length": 73.77734375, "completions/min_length": 36.0, "epoch": 1.7714497041420119, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.328125, "kl": 0.012846959382295608, "learning_rate": 7.675084964939924e-07, "loss": 0.0005768531933426857, "reward": 1.6468220949172974, "reward_std": 0.17373519639174143, "rewards/FidelityReward/mean": 0.7658949693044027, "rewards/FidelityReward/std": 0.20483198265234628, "rewards/JudgeFidelityReward/mean": 0.7618542114893595, "rewards/JudgeFidelityReward/std": 0.19425981243451437, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2395 }, { "clip_ratio/high_max": 0.0005933805368840694, "clip_ratio/high_mean": 0.00018176417797803878, "clip_ratio/low_mean": 5.9527096163947134e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002412912726867944, "completions/clipped_ratio": 0.0, "completions/max_length": 185.5, "completions/mean_length": 71.6953125, "completions/min_length": 30.0, "epoch": 1.7751479289940828, "frac_reward_zero_std": 0.0625, "grad_norm": 2.46875, "kl": 0.012629556097090244, "learning_rate": 7.664744727641277e-07, "loss": 0.0006663928739726544, "reward": 1.6696680784225464, "reward_std": 0.18539927154779434, "rewards/FidelityReward/mean": 0.7824194431304932, "rewards/FidelityReward/std": 0.20425809919834137, "rewards/JudgeFidelityReward/mean": 0.7744973301887512, "rewards/JudgeFidelityReward/std": 0.2156284973025322, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2400 }, { "clip_ratio/high_max": 0.0005859307013452054, "clip_ratio/high_mean": 0.00013640664983540773, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013640664983540773, "completions/clipped_ratio": 0.0, "completions/max_length": 182.33333333333334, "completions/mean_length": 74.0703125, "completions/min_length": 35.0, "epoch": 1.7788461538461537, "frac_reward_zero_std": 0.03125, "grad_norm": 2.6875, "kl": 0.012987761944532394, "learning_rate": 7.654388547911604e-07, "loss": 0.0006074380595237017, "reward": 1.6474804878234863, "reward_std": 0.19108897944291434, "rewards/FidelityReward/mean": 0.7654972871144613, "rewards/FidelityReward/std": 0.21133059759934744, "rewards/JudgeFidelityReward/mean": 0.7665705680847168, "rewards/JudgeFidelityReward/std": 0.21563746531804404, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 2405 }, { "clip_ratio/high_max": 0.0005723146838136018, "clip_ratio/high_mean": 0.0001772398449247703, "clip_ratio/low_mean": 2.1701389050576837e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019894123543053865, "completions/clipped_ratio": 0.0, "completions/max_length": 163.5, "completions/mean_length": 72.23046875, "completions/min_length": 31.0, "epoch": 1.782544378698225, "frac_reward_zero_std": 0.015625, "grad_norm": 2.734375, "kl": 0.01515473835170269, "learning_rate": 7.644016487709066e-07, "loss": 0.0005211610347032547, "reward": 1.6716214418411255, "reward_std": 0.1879321113228798, "rewards/FidelityReward/mean": 0.781851589679718, "rewards/FidelityReward/std": 0.19542859494686127, "rewards/JudgeFidelityReward/mean": 0.7834459543228149, "rewards/JudgeFidelityReward/std": 0.18796494603157043, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 2410 }, { "clip_ratio/high_max": 0.0008934148121625185, "clip_ratio/high_mean": 0.00024838035460561514, "clip_ratio/low_mean": 4.583106783684343e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002942114253528416, "completions/clipped_ratio": 0.0, "completions/max_length": 174.66666666666666, "completions/mean_length": 72.73307291666667, "completions/min_length": 34.0, "epoch": 1.7862426035502958, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.013180657289922237, "learning_rate": 7.633628609086833e-07, "loss": 0.0004865892231464386, "reward": 1.6402680079142253, "reward_std": 0.19080278277397156, "rewards/FidelityReward/mean": 0.7616255680720011, "rewards/FidelityReward/std": 0.21359091500441232, "rewards/JudgeFidelityReward/mean": 0.7585869232813517, "rewards/JudgeFidelityReward/std": 0.2148632357517878, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2415 }, { "clip_ratio/high_max": 0.0008734275004826486, "clip_ratio/high_mean": 0.00027840114780701696, "clip_ratio/low_mean": 1.9685039296746255e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00029808618710376323, "completions/clipped_ratio": 0.0, "completions/max_length": 189.5, "completions/mean_length": 71.8046875, "completions/min_length": 36.0, "epoch": 1.7899408284023668, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.01283266842365265, "learning_rate": 7.62322497419271e-07, "loss": 0.0006676619406789542, "reward": 1.6433794498443604, "reward_std": 0.17266875505447388, "rewards/FidelityReward/mean": 0.7634748220443726, "rewards/FidelityReward/std": 0.21887879073619843, "rewards/JudgeFidelityReward/mean": 0.7598091959953308, "rewards/JudgeFidelityReward/std": 0.2152203992009163, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2420 }, { "clip_ratio/high_max": 0.0010044739115983247, "clip_ratio/high_mean": 0.00021223415387794375, "clip_ratio/low_mean": 1.811594265745953e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023035010090097786, "completions/clipped_ratio": 0.0, "completions/max_length": 159.0, "completions/mean_length": 75.16276041666667, "completions/min_length": 36.333333333333336, "epoch": 1.793639053254438, "frac_reward_zero_std": 0.0, "grad_norm": 2.78125, "kl": 0.012348535098135471, "learning_rate": 7.61280564526877e-07, "loss": 0.00048643434420228007, "reward": 1.6213370561599731, "reward_std": 0.18841988841692606, "rewards/FidelityReward/mean": 0.7498731811841329, "rewards/FidelityReward/std": 0.21259813010692596, "rewards/JudgeFidelityReward/mean": 0.742927610874176, "rewards/JudgeFidelityReward/std": 0.2162566582361857, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2425 }, { "clip_ratio/high_max": 0.0008974749711342156, "clip_ratio/high_mean": 0.00021611764677800238, "clip_ratio/low_mean": 7.291889633052052e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002890365547500551, "completions/clipped_ratio": 0.0, "completions/max_length": 166.0, "completions/mean_length": 71.154296875, "completions/min_length": 33.0, "epoch": 1.797337278106509, "frac_reward_zero_std": 0.015625, "grad_norm": 2.421875, "kl": 0.014670184813439846, "learning_rate": 7.602370684650974e-07, "loss": 0.0006382052786648273, "reward": 1.7066726088523865, "reward_std": 0.1560061052441597, "rewards/FidelityReward/mean": 0.8042121529579163, "rewards/FidelityReward/std": 0.18208830803632736, "rewards/JudgeFidelityReward/mean": 0.8049209117889404, "rewards/JudgeFidelityReward/std": 0.1826678290963173, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2430 }, { "clip_ratio/high_max": 0.0003448072704486549, "clip_ratio/high_mean": 0.00010250457416987047, "clip_ratio/low_mean": 2.2893772984389217e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012539834569906815, "completions/clipped_ratio": 0.0, "completions/max_length": 179.33333333333334, "completions/mean_length": 73.59375, "completions/min_length": 35.333333333333336, "epoch": 1.8010355029585798, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.013927013240754604, "learning_rate": 7.591920154768809e-07, "loss": 0.0007210297975689173, "reward": 1.62684432665507, "reward_std": 0.1825540562470754, "rewards/FidelityReward/mean": 0.7519468665122986, "rewards/FidelityReward/std": 0.2275450179974238, "rewards/JudgeFidelityReward/mean": 0.7523990074793497, "rewards/JudgeFidelityReward/std": 0.2262282520532608, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 2435 }, { "clip_ratio/high_max": 0.0005616939160972834, "clip_ratio/high_mean": 0.0001690657445578836, "clip_ratio/low_mean": 1.8315018678549677e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018738076178124175, "completions/clipped_ratio": 0.0, "completions/max_length": 174.0, "completions/mean_length": 71.626953125, "completions/min_length": 31.5, "epoch": 1.804733727810651, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.013364771753549576, "learning_rate": 7.581454118144908e-07, "loss": 0.00047660265117883683, "reward": 1.6419398188591003, "reward_std": 0.18458064645528793, "rewards/FidelityReward/mean": 0.7615587115287781, "rewards/FidelityReward/std": 0.21743369847536087, "rewards/JudgeFidelityReward/mean": 0.7627155482769012, "rewards/JudgeFidelityReward/std": 0.19819511473178864, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2440 }, { "clip_ratio/high_max": 0.0007707965560257435, "clip_ratio/high_mean": 0.00017629252397455275, "clip_ratio/low_mean": 5.2422189037315546e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022871471592225134, "completions/clipped_ratio": 0.0, "completions/max_length": 192.0, "completions/mean_length": 73.87369791666667, "completions/min_length": 30.666666666666668, "epoch": 1.808431952662722, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.484375, "kl": 0.012984846904873847, "learning_rate": 7.570972637394672e-07, "loss": 0.0006264398340135813, "reward": 1.6930023829142253, "reward_std": 0.17453054587046304, "rewards/FidelityReward/mean": 0.7921549479166666, "rewards/FidelityReward/std": 0.18692843119303384, "rewards/JudgeFidelityReward/mean": 0.801694909731547, "rewards/JudgeFidelityReward/std": 0.16418616473674774, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2445 }, { "clip_ratio/high_max": 0.0006641886080615223, "clip_ratio/high_mean": 0.00018551095854490995, "clip_ratio/low_mean": 6.946194043848664e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002549729018937796, "completions/clipped_ratio": 0.0, "completions/max_length": 182.0, "completions/mean_length": 71.8828125, "completions/min_length": 32.5, "epoch": 1.8121301775147929, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.013525886088609695, "learning_rate": 7.560475775225903e-07, "loss": 0.00048447754234075545, "reward": 1.6701948046684265, "reward_std": 0.1792822778224945, "rewards/FidelityReward/mean": 0.7843668460845947, "rewards/FidelityReward/std": 0.18769561499357224, "rewards/JudgeFidelityReward/mean": 0.771655797958374, "rewards/JudgeFidelityReward/std": 0.19745111465454102, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2450 }, { "clip_ratio/high_max": 0.00037549842381849884, "clip_ratio/high_mean": 0.0001524002174846828, "clip_ratio/low_mean": 4.860195185756311e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020100217079743743, "completions/clipped_ratio": 0.0, "completions/max_length": 174.33333333333334, "completions/mean_length": 71.90625, "completions/min_length": 34.666666666666664, "epoch": 1.815828402366864, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.46875, "kl": 0.01366654820740223, "learning_rate": 7.549963594438425e-07, "loss": 0.0006204531993716956, "reward": 1.6171907981236775, "reward_std": 0.1819481005271276, "rewards/FidelityReward/mean": 0.7464912136395773, "rewards/FidelityReward/std": 0.21893922487894693, "rewards/JudgeFidelityReward/mean": 0.7413991093635559, "rewards/JudgeFidelityReward/std": 0.22120667497316995, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2455 }, { "clip_ratio/high_max": 0.0005391665850766003, "clip_ratio/high_mean": 0.00010524891840759664, "clip_ratio/low_mean": 4.097523051314056e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014622415183112024, "completions/clipped_ratio": 0.0, "completions/max_length": 186.5, "completions/mean_length": 70.984375, "completions/min_length": 32.5, "epoch": 1.8195266272189348, "frac_reward_zero_std": 0.0625, "grad_norm": 2.390625, "kl": 0.013579297251999379, "learning_rate": 7.539436157923713e-07, "loss": 0.0006979400292038918, "reward": 1.6448495388031006, "reward_std": 0.18448033928871155, "rewards/FidelityReward/mean": 0.7643017172813416, "rewards/FidelityReward/std": 0.21310696750879288, "rewards/JudgeFidelityReward/mean": 0.7630489468574524, "rewards/JudgeFidelityReward/std": 0.20071713626384735, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2460 }, { "clip_ratio/high_max": 0.0008250012760981917, "clip_ratio/high_mean": 0.00037288577295839785, "clip_ratio/low_mean": 4.2361358646303414e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00041524714324623344, "completions/clipped_ratio": 0.0, "completions/max_length": 191.66666666666666, "completions/mean_length": 73.29557291666667, "completions/min_length": 29.333333333333332, "epoch": 1.823224852071006, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.014074631594121456, "learning_rate": 7.528893528664504e-07, "loss": 0.0006285439245402813, "reward": 1.6604046026865642, "reward_std": 0.20095832149187723, "rewards/FidelityReward/mean": 0.7723592321077982, "rewards/FidelityReward/std": 0.1957498143116633, "rewards/JudgeFidelityReward/mean": 0.7760906219482422, "rewards/JudgeFidelityReward/std": 0.19436363875865936, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2465 }, { "clip_ratio/high_max": 0.0008519166498444974, "clip_ratio/high_mean": 0.00022447276132879778, "clip_ratio/low_mean": 8.39849773910828e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003084577328991145, "completions/clipped_ratio": 0.0, "completions/max_length": 175.5, "completions/mean_length": 72.77734375, "completions/min_length": 34.5, "epoch": 1.8269230769230769, "frac_reward_zero_std": 0.03125, "grad_norm": 2.5625, "kl": 0.01377660222351551, "learning_rate": 7.518335769734438e-07, "loss": 0.0005508806556463241, "reward": 1.645328938961029, "reward_std": 0.17027562111616135, "rewards/FidelityReward/mean": 0.7611726820468903, "rewards/FidelityReward/std": 0.22095835208892822, "rewards/JudgeFidelityReward/mean": 0.7683124244213104, "rewards/JudgeFidelityReward/std": 0.2154366597533226, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2470 }, { "clip_ratio/high_max": 0.0006585504161193967, "clip_ratio/high_mean": 0.00024170505348592998, "clip_ratio/low_mean": 9.481185697950423e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003365169046446681, "completions/clipped_ratio": 0.0, "completions/max_length": 184.33333333333334, "completions/mean_length": 74.55338541666667, "completions/min_length": 36.333333333333336, "epoch": 1.8306213017751478, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.01488795131444931, "learning_rate": 7.507762944297668e-07, "loss": 0.0006749742664396763, "reward": 1.6480778455734253, "reward_std": 0.18949855864048004, "rewards/FidelityReward/mean": 0.7618636091550192, "rewards/FidelityReward/std": 0.20747613410154978, "rewards/JudgeFidelityReward/mean": 0.7724284728368124, "rewards/JudgeFidelityReward/std": 0.195219357808431, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2475 }, { "clip_ratio/high_max": 0.0012869078665971756, "clip_ratio/high_mean": 0.0003618728602305055, "clip_ratio/low_mean": 9.563485946273431e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004575077095068991, "completions/clipped_ratio": 0.0, "completions/max_length": 186.5, "completions/mean_length": 74.017578125, "completions/min_length": 36.5, "epoch": 1.834319526627219, "frac_reward_zero_std": 0.03125, "grad_norm": 2.609375, "kl": 0.013214955665171146, "learning_rate": 7.497175115608482e-07, "loss": 0.0006611372344195842, "reward": 1.6288219690322876, "reward_std": 0.1911899447441101, "rewards/FidelityReward/mean": 0.7525011897087097, "rewards/FidelityReward/std": 0.20859156548976898, "rewards/JudgeFidelityReward/mean": 0.7526415288448334, "rewards/JudgeFidelityReward/std": 0.20130060613155365, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2480 }, { "clip_ratio/high_max": 0.0003359768772497773, "clip_ratio/high_mean": 9.907451458275318e-05, "clip_ratio/low_mean": 8.366192923858762e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001827364438213408, "completions/clipped_ratio": 0.0, "completions/max_length": 195.33333333333334, "completions/mean_length": 73.8984375, "completions/min_length": 34.0, "epoch": 1.83801775147929, "frac_reward_zero_std": 0.03125, "grad_norm": 2.75, "kl": 0.015142030268907546, "learning_rate": 7.486572347010936e-07, "loss": 0.0006569832563400269, "reward": 1.6549285252888997, "reward_std": 0.19004592299461365, "rewards/FidelityReward/mean": 0.7688546776771545, "rewards/FidelityReward/std": 0.22069750229517618, "rewards/JudgeFidelityReward/mean": 0.7747519810994467, "rewards/JudgeFidelityReward/std": 0.21168196698029837, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 2485 }, { "clip_ratio/high_max": 0.0008333121309988201, "clip_ratio/high_mean": 0.0001746409325278364, "clip_ratio/low_mean": 0.00017000618827296422, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00034464714699424805, "completions/clipped_ratio": 0.0, "completions/max_length": 211.0, "completions/mean_length": 72.66015625, "completions/min_length": 36.0, "epoch": 1.8417159763313609, "frac_reward_zero_std": 0.015625, "grad_norm": 2.765625, "kl": 0.013875468634068966, "learning_rate": 7.475954701938463e-07, "loss": 0.0007127360440790653, "reward": 1.653399646282196, "reward_std": 0.187383234500885, "rewards/FidelityReward/mean": 0.7678107619285583, "rewards/FidelityReward/std": 0.21064811944961548, "rewards/JudgeFidelityReward/mean": 0.7711778283119202, "rewards/JudgeFidelityReward/std": 0.21394813060760498, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2490 }, { "clip_ratio/high_max": 0.0003310267115011811, "clip_ratio/high_mean": 8.210394880734385e-05, "clip_ratio/low_mean": 2.530364436097443e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010740759025793523, "completions/clipped_ratio": 0.0, "completions/max_length": 187.33333333333334, "completions/mean_length": 73.35416666666667, "completions/min_length": 34.666666666666664, "epoch": 1.845414201183432, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.546875, "kl": 0.013933395966887474, "learning_rate": 7.465322243913496e-07, "loss": 0.0006188994273543357, "reward": 1.6529305378595989, "reward_std": 0.18928408126036325, "rewards/FidelityReward/mean": 0.769024650255839, "rewards/FidelityReward/std": 0.20760454734166464, "rewards/JudgeFidelityReward/mean": 0.7678118348121643, "rewards/JudgeFidelityReward/std": 0.21024699012438455, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2495 }, { "clip_ratio/high_max": 0.0005656037479639054, "clip_ratio/high_mean": 0.00011135775421280413, "clip_ratio/low_mean": 4.77452966151759e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00015910304791759698, "completions/clipped_ratio": 0.0, "completions/max_length": 147.0, "completions/mean_length": 71.58203125, "completions/min_length": 36.0, "epoch": 1.849112426035503, "frac_reward_zero_std": 0.0, "grad_norm": 2.6875, "kl": 0.014110823534429074, "learning_rate": 7.454675036547093e-07, "loss": 0.0006079308688640595, "reward": 1.6515743732452393, "reward_std": 0.18832673132419586, "rewards/FidelityReward/mean": 0.7672054469585419, "rewards/FidelityReward/std": 0.2265300676226616, "rewards/JudgeFidelityReward/mean": 0.7706910073757172, "rewards/JudgeFidelityReward/std": 0.21611125022172928, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2500 }, { "clip_ratio/high_max": 0.00020161289721727372, "clip_ratio/high_mean": 0.00010545365512371063, "clip_ratio/low_mean": 2.200704184360802e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012746069114655256, "completions/clipped_ratio": 0.0, "completions/max_length": 200.33333333333334, "completions/mean_length": 72.61979166666667, "completions/min_length": 34.333333333333336, "epoch": 1.852810650887574, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.453125, "kl": 0.015370933711528778, "learning_rate": 7.444013143538553e-07, "loss": 0.0006074655801057815, "reward": 1.640006462732951, "reward_std": 0.19979662696520487, "rewards/FidelityReward/mean": 0.7593920628229777, "rewards/FidelityReward/std": 0.20422240595022836, "rewards/JudgeFidelityReward/mean": 0.762531061967214, "rewards/JudgeFidelityReward/std": 0.20454824964205423, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2505 }, { "clip_ratio/high_max": 0.0006066010100767016, "clip_ratio/high_mean": 7.58251262595877e-05, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 7.58251262595877e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 178.5, "completions/mean_length": 69.328125, "completions/min_length": 30.0, "epoch": 1.856508875739645, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.014246323518455029, "learning_rate": 7.433336628675033e-07, "loss": 0.000524780061095953, "reward": 1.7099379301071167, "reward_std": 0.15260529518127441, "rewards/FidelityReward/mean": 0.8078683316707611, "rewards/FidelityReward/std": 0.18404103070497513, "rewards/JudgeFidelityReward/mean": 0.8041391372680664, "rewards/JudgeFidelityReward/std": 0.17119824886322021, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2510 }, { "clip_ratio/high_max": 0.00021691974252462386, "clip_ratio/high_mean": 4.7913636080920696e-05, "clip_ratio/low_mean": 2.2045854711905123e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 6.995949079282581e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 194.33333333333334, "completions/mean_length": 72.31510416666667, "completions/min_length": 34.0, "epoch": 1.860207100591716, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.01587616540491581, "learning_rate": 7.422645555831171e-07, "loss": 0.0005781267769634724, "reward": 1.6166037718454997, "reward_std": 0.19155225157737732, "rewards/FidelityReward/mean": 0.7432719667752584, "rewards/FidelityReward/std": 0.22620663543542227, "rewards/JudgeFidelityReward/mean": 0.7492677569389343, "rewards/JudgeFidelityReward/std": 0.22628279030323029, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 2515 }, { "clip_ratio/high_max": 0.0011041695601306857, "clip_ratio/high_mean": 0.00023329826071858407, "clip_ratio/low_mean": 5.055611836723983e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028385438490659, "completions/clipped_ratio": 0.0, "completions/max_length": 154.5, "completions/mean_length": 69.568359375, "completions/min_length": 30.0, "epoch": 1.863905325443787, "frac_reward_zero_std": 0.0, "grad_norm": 2.734375, "kl": 0.013745537959039211, "learning_rate": 7.4119399889687e-07, "loss": 0.0006588340736925602, "reward": 1.6319028735160828, "reward_std": 0.18175800144672394, "rewards/FidelityReward/mean": 0.752763420343399, "rewards/FidelityReward/std": 0.22489026188850403, "rewards/JudgeFidelityReward/mean": 0.7582789361476898, "rewards/JudgeFidelityReward/std": 0.21465320885181427, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2520 }, { "clip_ratio/high_max": 0.00039784719701856374, "clip_ratio/high_mean": 0.00011361979995854199, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00011361979995854199, "completions/clipped_ratio": 0.0, "completions/max_length": 157.66666666666666, "completions/mean_length": 72.11588541666667, "completions/min_length": 34.0, "epoch": 1.8676035502958581, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.828125, "kl": 0.015815501287579536, "learning_rate": 7.401219992136066e-07, "loss": 0.0007177412044256926, "reward": 1.665779749552409, "reward_std": 0.17940795918305716, "rewards/FidelityReward/mean": 0.7762763500213623, "rewards/FidelityReward/std": 0.1876576840877533, "rewards/JudgeFidelityReward/mean": 0.7790068984031677, "rewards/JudgeFidelityReward/std": 0.18414137264092764, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2525 }, { "clip_ratio/high_max": 0.0009150734404101968, "clip_ratio/high_mean": 0.00038470146828331055, "clip_ratio/low_mean": 2.1777003712486476e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004064784909132868, "completions/clipped_ratio": 0.0, "completions/max_length": 144.5, "completions/mean_length": 69.845703125, "completions/min_length": 33.5, "epoch": 1.8713017751479288, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.01485287994146347, "learning_rate": 7.390485629468048e-07, "loss": 0.0006244193762540817, "reward": 1.5808481574058533, "reward_std": 0.20037060230970383, "rewards/FidelityReward/mean": 0.7185198664665222, "rewards/FidelityReward/std": 0.21834467351436615, "rewards/JudgeFidelityReward/mean": 0.7266097962856293, "rewards/JudgeFidelityReward/std": 0.22150001674890518, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2530 }, { "clip_ratio/high_max": 0.0005964214913547039, "clip_ratio/high_mean": 0.0001571710454300046, "clip_ratio/low_mean": 7.517643389292062e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023234747350215913, "completions/clipped_ratio": 0.0, "completions/max_length": 194.66666666666666, "completions/mean_length": 69.52864583333333, "completions/min_length": 32.0, "epoch": 1.875, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.46875, "kl": 0.014122618548572063, "learning_rate": 7.379736965185368e-07, "loss": 0.0005580815486609936, "reward": 1.6303138335545857, "reward_std": 0.16991592446962991, "rewards/FidelityReward/mean": 0.7533995111783346, "rewards/FidelityReward/std": 0.1995616207520167, "rewards/JudgeFidelityReward/mean": 0.7538286447525024, "rewards/JudgeFidelityReward/std": 0.19814321398735046, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2535 }, { "clip_ratio/high_max": 0.0004951686831191182, "clip_ratio/high_mean": 0.00019163520773872732, "clip_ratio/low_mean": 6.761199765605851e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002592472010292113, "completions/clipped_ratio": 0.0, "completions/max_length": 197.0, "completions/mean_length": 72.291015625, "completions/min_length": 35.0, "epoch": 1.8786982248520712, "frac_reward_zero_std": 0.046875, "grad_norm": 2.46875, "kl": 0.01222736407071352, "learning_rate": 7.368974063594314e-07, "loss": 0.000555112399160862, "reward": 1.652113437652588, "reward_std": 0.18246200680732727, "rewards/FidelityReward/mean": 0.7682634592056274, "rewards/FidelityReward/std": 0.1861656904220581, "rewards/JudgeFidelityReward/mean": 0.7677000164985657, "rewards/JudgeFidelityReward/std": 0.18471480160951614, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2540 }, { "clip_ratio/high_max": 0.0003863849211484194, "clip_ratio/high_mean": 0.00010849963000509888, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010849963000509888, "completions/clipped_ratio": 0.0, "completions/max_length": 176.66666666666666, "completions/mean_length": 72.08203125, "completions/min_length": 32.333333333333336, "epoch": 1.8823964497041419, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.15625, "kl": 0.014447763189673423, "learning_rate": 7.358196989086351e-07, "loss": 0.0006953556090593338, "reward": 1.684151252110799, "reward_std": 0.1694405972957611, "rewards/FidelityReward/mean": 0.7894931038220724, "rewards/FidelityReward/std": 0.19828113416830698, "rewards/JudgeFidelityReward/mean": 0.7893162965774536, "rewards/JudgeFidelityReward/std": 0.18810408810774484, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2545 }, { "clip_ratio/high_max": 0.0007845394778996706, "clip_ratio/high_mean": 0.00024661970674060284, "clip_ratio/low_mean": 6.415819079848006e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00031077788153197614, "completions/clipped_ratio": 0.0, "completions/max_length": 180.5, "completions/mean_length": 73.98046875, "completions/min_length": 34.5, "epoch": 1.886094674556213, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.014447075128555299, "learning_rate": 7.347405806137739e-07, "loss": 0.0005528007633984089, "reward": 1.6492525339126587, "reward_std": 0.19705455005168915, "rewards/FidelityReward/mean": 0.7687475681304932, "rewards/FidelityReward/std": 0.18867935985326767, "rewards/JudgeFidelityReward/mean": 0.7629630863666534, "rewards/JudgeFidelityReward/std": 0.18686195462942123, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2550 }, { "clip_ratio/high_max": 0.0007324822945520282, "clip_ratio/high_mean": 0.00020688334479928016, "clip_ratio/low_mean": 4.784646880580112e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002547298208810389, "completions/clipped_ratio": 0.0, "completions/max_length": 166.33333333333334, "completions/mean_length": 71.40885416666667, "completions/min_length": 32.333333333333336, "epoch": 1.889792899408284, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.890625, "kl": 0.013782891072332859, "learning_rate": 7.336600579309138e-07, "loss": 0.000504687987267971, "reward": 1.6486741304397583, "reward_std": 0.19006663064161936, "rewards/FidelityReward/mean": 0.766723096370697, "rewards/FidelityReward/std": 0.20282056430975595, "rewards/JudgeFidelityReward/mean": 0.7639020880063375, "rewards/JudgeFidelityReward/std": 0.20331775148709616, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2555 }, { "clip_ratio/high_max": 0.0005977895110845566, "clip_ratio/high_mean": 0.0001976576109882444, "clip_ratio/low_mean": 5.8374914806336166e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00025603253743611276, "completions/clipped_ratio": 0.0, "completions/max_length": 156.5, "completions/mean_length": 70.1015625, "completions/min_length": 32.0, "epoch": 1.893491124260355, "frac_reward_zero_std": 0.0, "grad_norm": 2.765625, "kl": 0.014883760921657085, "learning_rate": 7.325781373245235e-07, "loss": 0.0006840870715677738, "reward": 1.6058825254440308, "reward_std": 0.18830151855945587, "rewards/FidelityReward/mean": 0.7369196116924286, "rewards/FidelityReward/std": 0.23368927836418152, "rewards/JudgeFidelityReward/mean": 0.7379259765148163, "rewards/JudgeFidelityReward/std": 0.2282472774386406, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2560 }, { "clip_ratio/high_max": 0.0003895460511557758, "clip_ratio/high_mean": 0.00015798690728843212, "clip_ratio/low_mean": 2.262443449581042e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018061134032905103, "completions/clipped_ratio": 0.0, "completions/max_length": 200.0, "completions/mean_length": 71.18098958333333, "completions/min_length": 30.333333333333332, "epoch": 1.897189349112426, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.5, "kl": 0.014307336695492268, "learning_rate": 7.314948252674349e-07, "loss": 0.0006957565899938345, "reward": 1.649430473645528, "reward_std": 0.1782576988140742, "rewards/FidelityReward/mean": 0.7655431429545084, "rewards/FidelityReward/std": 0.22188952565193176, "rewards/JudgeFidelityReward/mean": 0.7677747209866842, "rewards/JudgeFidelityReward/std": 0.20789488156636557, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2565 }, { "clip_ratio/high_max": 0.0006357721867971122, "clip_ratio/high_mean": 0.0003005179634783417, "clip_ratio/low_mean": 9.482919704169035e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000395347154699266, "completions/clipped_ratio": 0.0, "completions/max_length": 150.5, "completions/mean_length": 73.291015625, "completions/min_length": 38.0, "epoch": 1.900887573964497, "frac_reward_zero_std": 0.015625, "grad_norm": 2.15625, "kl": 0.013679273799061776, "learning_rate": 7.304101282408044e-07, "loss": 0.0005995781626552344, "reward": 1.6833547949790955, "reward_std": 0.154402956366539, "rewards/FidelityReward/mean": 0.786560595035553, "rewards/FidelityReward/std": 0.21052468568086624, "rewards/JudgeFidelityReward/mean": 0.7935883104801178, "rewards/JudgeFidelityReward/std": 0.1961357593536377, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2570 }, { "clip_ratio/high_max": 0.0007009618915617466, "clip_ratio/high_mean": 0.0001310318912146613, "clip_ratio/low_mean": 0.00011835788609459996, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002493897802196443, "completions/clipped_ratio": 0.0, "completions/max_length": 166.33333333333334, "completions/mean_length": 71.73828125, "completions/min_length": 35.333333333333336, "epoch": 1.904585798816568, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 3.046875, "kl": 0.01356295421719551, "learning_rate": 7.293240527340747e-07, "loss": 0.000465119443833828, "reward": 1.6036536296208699, "reward_std": 0.2080583522717158, "rewards/FidelityReward/mean": 0.7364931702613831, "rewards/FidelityReward/std": 0.21319813032944998, "rewards/JudgeFidelityReward/mean": 0.7356229027112325, "rewards/JudgeFidelityReward/std": 0.21601866682370505, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2575 }, { "clip_ratio/high_max": 0.000858390610665083, "clip_ratio/high_mean": 0.00020504253334365785, "clip_ratio/low_mean": 2.3364486696664245e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022840701858513058, "completions/clipped_ratio": 0.0, "completions/max_length": 177.0, "completions/mean_length": 72.400390625, "completions/min_length": 30.0, "epoch": 1.9082840236686391, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.01500757373869419, "learning_rate": 7.28236605244935e-07, "loss": 0.0006628512404859066, "reward": 1.6549147963523865, "reward_std": 0.17216840386390686, "rewards/FidelityReward/mean": 0.7780146598815918, "rewards/FidelityReward/std": 0.2263985201716423, "rewards/JudgeFidelityReward/mean": 0.7538003325462341, "rewards/JudgeFidelityReward/std": 0.23526735603809357, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2580 }, { "clip_ratio/high_max": 0.0004268064396455884, "clip_ratio/high_mean": 0.00013697219546884298, "clip_ratio/low_mean": 1.6711230273358524e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00015368341701105237, "completions/clipped_ratio": 0.0, "completions/max_length": 173.0, "completions/mean_length": 70.16796875, "completions/min_length": 31.0, "epoch": 1.91198224852071, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.859375, "kl": 0.014179723896086216, "learning_rate": 7.271477922792832e-07, "loss": 0.0005543106235563755, "reward": 1.6239272356033325, "reward_std": 0.1952806611855825, "rewards/FidelityReward/mean": 0.7467880249023438, "rewards/FidelityReward/std": 0.20760131378968558, "rewards/JudgeFidelityReward/mean": 0.7542784015337626, "rewards/JudgeFidelityReward/std": 0.19874327878157297, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2585 }, { "clip_ratio/high_max": 0.0005129707395099103, "clip_ratio/high_mean": 0.00013633835478685797, "clip_ratio/low_mean": 0.00014737557212356477, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002837139181792736, "completions/clipped_ratio": 0.0, "completions/max_length": 179.0, "completions/mean_length": 72.37890625, "completions/min_length": 35.0, "epoch": 1.915680473372781, "frac_reward_zero_std": 0.015625, "grad_norm": 2.59375, "kl": 0.014481769874691963, "learning_rate": 7.260576203511863e-07, "loss": 0.0005051133688539266, "reward": 1.6814569234848022, "reward_std": 0.16798026114702225, "rewards/FidelityReward/mean": 0.783550888299942, "rewards/FidelityReward/std": 0.20432280749082565, "rewards/JudgeFidelityReward/mean": 0.7977651357650757, "rewards/JudgeFidelityReward/std": 0.18831121921539307, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2590 }, { "clip_ratio/high_max": 0.00040209563449025156, "clip_ratio/high_mean": 0.00016473582945764065, "clip_ratio/low_mean": 1.7692851542960852e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018242867663502692, "completions/clipped_ratio": 0.0, "completions/max_length": 177.66666666666666, "completions/mean_length": 74.49609375, "completions/min_length": 31.0, "epoch": 1.9193786982248522, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.640625, "kl": 0.01435008030384779, "learning_rate": 7.249660959828413e-07, "loss": 0.0005909925792366266, "reward": 1.6629894177118938, "reward_std": 0.18408137063185373, "rewards/FidelityReward/mean": 0.7819557984670004, "rewards/FidelityReward/std": 0.21685627599557242, "rewards/JudgeFidelityReward/mean": 0.7620672583580017, "rewards/JudgeFidelityReward/std": 0.2221412460009257, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2595 }, { "clip_ratio/high_max": 0.0005282358615659177, "clip_ratio/high_mean": 0.00018438641272950919, "clip_ratio/low_mean": 4.812836123164743e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002325147797819227, "completions/clipped_ratio": 0.0, "completions/max_length": 172.0, "completions/mean_length": 74.14453125, "completions/min_length": 33.0, "epoch": 1.9230769230769231, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.015081290155649185, "learning_rate": 7.238732257045372e-07, "loss": 0.000772236380726099, "reward": 1.609814703464508, "reward_std": 0.19453584402799606, "rewards/FidelityReward/mean": 0.7444043755531311, "rewards/FidelityReward/std": 0.21889521926641464, "rewards/JudgeFidelityReward/mean": 0.7308205664157867, "rewards/JudgeFidelityReward/std": 0.2267821729183197, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2600 }, { "clip_ratio/high_max": 0.00018248175038024784, "clip_ratio/high_mean": 8.495977381244302e-05, "clip_ratio/low_mean": 3.9207840745802966e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012416761164786295, "completions/clipped_ratio": 0.0, "completions/max_length": 176.0, "completions/mean_length": 74.84895833333333, "completions/min_length": 34.666666666666664, "epoch": 1.926775147928994, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.015931271947920322, "learning_rate": 7.227790160546144e-07, "loss": 0.0005983021110296249, "reward": 1.6650582949320476, "reward_std": 0.19146023193995157, "rewards/FidelityReward/mean": 0.7775574525197347, "rewards/FidelityReward/std": 0.2003227969010671, "rewards/JudgeFidelityReward/mean": 0.7776057720184326, "rewards/JudgeFidelityReward/std": 0.20509051283200583, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 2605 }, { "clip_ratio/high_max": 0.000859654217492789, "clip_ratio/high_mean": 0.00022873324051033705, "clip_ratio/low_mean": 3.880279255099595e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002675360272405669, "completions/clipped_ratio": 0.0, "completions/max_length": 165.5, "completions/mean_length": 72.546875, "completions/min_length": 35.5, "epoch": 1.9304733727810652, "frac_reward_zero_std": 0.0, "grad_norm": 2.6875, "kl": 0.01431818287819624, "learning_rate": 7.216834735794267e-07, "loss": 0.0005340809933841228, "reward": 1.6417688131332397, "reward_std": 0.18457913398742676, "rewards/FidelityReward/mean": 0.7558683753013611, "rewards/FidelityReward/std": 0.2057589665055275, "rewards/JudgeFidelityReward/mean": 0.7718009352684021, "rewards/JudgeFidelityReward/std": 0.1934765800833702, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2610 }, { "clip_ratio/high_max": 0.00034875280689448116, "clip_ratio/high_mean": 8.361708605661989e-05, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 8.361708605661989e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 179.33333333333334, "completions/mean_length": 72.10416666666667, "completions/min_length": 31.666666666666668, "epoch": 1.934171597633136, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.21875, "kl": 0.016139750927686693, "learning_rate": 7.20586604833302e-07, "loss": 0.0005424284841865301, "reward": 1.6819891532262166, "reward_std": 0.17634045084317526, "rewards/FidelityReward/mean": 0.7868532737096151, "rewards/FidelityReward/std": 0.19752738376458487, "rewards/JudgeFidelityReward/mean": 0.7902719179789225, "rewards/JudgeFidelityReward/std": 0.1878013958533605, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2615 }, { "clip_ratio/high_max": 0.0008979025413282215, "clip_ratio/high_mean": 0.00019873568671755493, "clip_ratio/low_mean": 3.962771152146161e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023836339241825044, "completions/clipped_ratio": 0.0, "completions/max_length": 201.5, "completions/mean_length": 72.294921875, "completions/min_length": 33.0, "epoch": 1.9378698224852071, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.015635454282164573, "learning_rate": 7.194884163785027e-07, "loss": 0.0006147054955363274, "reward": 1.7297136187553406, "reward_std": 0.16904236376285553, "rewards/FidelityReward/mean": 0.8200316727161407, "rewards/FidelityReward/std": 0.1917085349559784, "rewards/JudgeFidelityReward/mean": 0.8193638920783997, "rewards/JudgeFidelityReward/std": 0.18227789551019669, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2620 }, { "clip_ratio/high_max": 0.00047355815768241883, "clip_ratio/high_mean": 0.00010518229682929814, "clip_ratio/low_mean": 6.489107763627545e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017007337301038205, "completions/clipped_ratio": 0.0, "completions/max_length": 156.33333333333334, "completions/mean_length": 71.68880208333333, "completions/min_length": 33.666666666666664, "epoch": 1.941568047337278, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.421875, "kl": 0.01464413944631815, "learning_rate": 7.183889147851867e-07, "loss": 0.0006699037272483111, "reward": 1.6944362322489421, "reward_std": 0.16939611732959747, "rewards/FidelityReward/mean": 0.7899662653605143, "rewards/FidelityReward/std": 0.19279909133911133, "rewards/JudgeFidelityReward/mean": 0.8089399536450704, "rewards/JudgeFidelityReward/std": 0.17047491172949472, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2625 }, { "clip_ratio/high_max": 0.0014325700467452407, "clip_ratio/high_mean": 0.00022343115997500718, "clip_ratio/low_mean": 4.1256145050283524e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026468729483895, "completions/clipped_ratio": 0.0, "completions/max_length": 190.5, "completions/mean_length": 74.8984375, "completions/min_length": 28.5, "epoch": 1.945266272189349, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.014376389235258103, "learning_rate": 7.172881066313685e-07, "loss": 0.0006203728262335062, "reward": 1.7292864322662354, "reward_std": 0.16898156702518463, "rewards/FidelityReward/mean": 0.8202453851699829, "rewards/FidelityReward/std": 0.179195836186409, "rewards/JudgeFidelityReward/mean": 0.8200353980064392, "rewards/JudgeFidelityReward/std": 0.16835643351078033, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2630 }, { "clip_ratio/high_max": 0.0007494189660064877, "clip_ratio/high_mean": 0.00024008332984521986, "clip_ratio/low_mean": 3.868639178108424e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027876972453668716, "completions/clipped_ratio": 0.0, "completions/max_length": 156.0, "completions/mean_length": 71.98177083333333, "completions/min_length": 37.333333333333336, "epoch": 1.9489644970414202, "frac_reward_zero_std": 0.03125, "grad_norm": 2.53125, "kl": 0.016303882375359534, "learning_rate": 7.161859985028782e-07, "loss": 0.0005928616970777512, "reward": 1.597771128018697, "reward_std": 0.1981037457784017, "rewards/FidelityReward/mean": 0.7410579721132914, "rewards/FidelityReward/std": 0.22948303818702698, "rewards/JudgeFidelityReward/mean": 0.7160304586092631, "rewards/JudgeFidelityReward/std": 0.23856707910696665, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 2635 }, { "clip_ratio/high_max": 0.0009286568150855601, "clip_ratio/high_mean": 0.000242093185079284, "clip_ratio/low_mean": 4.4764245103579015e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002868574199965224, "completions/clipped_ratio": 0.0, "completions/max_length": 178.0, "completions/mean_length": 72.208984375, "completions/min_length": 33.0, "epoch": 1.952662721893491, "frac_reward_zero_std": 0.0625, "grad_norm": 2.578125, "kl": 0.01651486996561289, "learning_rate": 7.150825969933246e-07, "loss": 0.0007208028342574835, "reward": 1.6512128114700317, "reward_std": 0.17057853937149048, "rewards/FidelityReward/mean": 0.7644873261451721, "rewards/FidelityReward/std": 0.2598718851804733, "rewards/JudgeFidelityReward/mean": 0.7793102860450745, "rewards/JudgeFidelityReward/std": 0.25308237224817276, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 2640 }, { "clip_ratio/high_max": 0.0005281327990815044, "clip_ratio/high_mean": 0.00010528538841754198, "clip_ratio/low_mean": 9.521415631752461e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020049954764544964, "completions/clipped_ratio": 0.0, "completions/max_length": 165.33333333333334, "completions/mean_length": 70.5859375, "completions/min_length": 36.0, "epoch": 1.956360946745562, "frac_reward_zero_std": 0.0, "grad_norm": 2.734375, "kl": 0.015756814181804656, "learning_rate": 7.139779087040539e-07, "loss": 0.0006185097619891167, "reward": 1.656355897585551, "reward_std": 0.1820966104666392, "rewards/FidelityReward/mean": 0.7725019454956055, "rewards/FidelityReward/std": 0.20098701119422913, "rewards/JudgeFidelityReward/mean": 0.7690099676450094, "rewards/JudgeFidelityReward/std": 0.20434853434562683, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2645 }, { "clip_ratio/high_max": 0.0007756155333481729, "clip_ratio/high_mean": 0.0002568568103015423, "clip_ratio/low_mean": 2.0938023226335645e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027779483352787795, "completions/clipped_ratio": 0.0, "completions/max_length": 143.5, "completions/mean_length": 70.462890625, "completions/min_length": 27.5, "epoch": 1.9600591715976332, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.017729962058365346, "learning_rate": 7.128719402441101e-07, "loss": 0.0007581162266433239, "reward": 1.651497483253479, "reward_std": 0.19648943841457367, "rewards/FidelityReward/mean": 0.7654216885566711, "rewards/FidelityReward/std": 0.23520753532648087, "rewards/JudgeFidelityReward/mean": 0.7741046249866486, "rewards/JudgeFidelityReward/std": 0.2248280644416809, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2650 }, { "clip_ratio/high_max": 0.0008009418146684766, "clip_ratio/high_mean": 0.0002110200352035463, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002110200352035463, "completions/clipped_ratio": 0.0, "completions/max_length": 170.33333333333334, "completions/mean_length": 71.64713541666667, "completions/min_length": 32.333333333333336, "epoch": 1.9637573964497042, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.015575863607227802, "learning_rate": 7.117646982301977e-07, "loss": 0.0007453162223100662, "reward": 1.6573117176691692, "reward_std": 0.18749582270781198, "rewards/FidelityReward/mean": 0.7702338894208273, "rewards/FidelityReward/std": 0.2082138111193975, "rewards/JudgeFidelityReward/mean": 0.7741556366284689, "rewards/JudgeFidelityReward/std": 0.20137526591618857, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2655 }, { "clip_ratio/high_max": 0.0009826497873291374, "clip_ratio/high_mean": 0.0003384306328371167, "clip_ratio/low_mean": 5.7006118004210295e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00039543674793094394, "completions/clipped_ratio": 0.0, "completions/max_length": 169.5, "completions/mean_length": 69.5625, "completions/min_length": 34.0, "epoch": 1.967455621301775, "frac_reward_zero_std": 0.015625, "grad_norm": 2.546875, "kl": 0.015102826058864594, "learning_rate": 7.106561892866389e-07, "loss": 0.0006022934336215258, "reward": 1.5719311237335205, "reward_std": 0.19662084430456161, "rewards/FidelityReward/mean": 0.7121245861053467, "rewards/FidelityReward/std": 0.2197577878832817, "rewards/JudgeFidelityReward/mean": 0.71961310505867, "rewards/JudgeFidelityReward/std": 0.19546791166067123, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2660 }, { "clip_ratio/high_max": 0.00035791911650449035, "clip_ratio/high_mean": 6.567791278939694e-05, "clip_ratio/low_mean": 4.7214352525770666e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00011289226531516761, "completions/clipped_ratio": 0.0, "completions/max_length": 165.66666666666666, "completions/mean_length": 68.56119791666667, "completions/min_length": 34.0, "epoch": 1.9711538461538463, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.703125, "kl": 0.01536985207349062, "learning_rate": 7.095464200453365e-07, "loss": 0.0005529025103896857, "reward": 1.6088438034057617, "reward_std": 0.18892898658911386, "rewards/FidelityReward/mean": 0.7405392328898112, "rewards/FidelityReward/std": 0.2079075276851654, "rewards/JudgeFidelityReward/mean": 0.7379110455513, "rewards/JudgeFidelityReward/std": 0.19484840333461761, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2665 }, { "clip_ratio/high_max": 0.0009354217094369232, "clip_ratio/high_mean": 0.0002817775821313262, "clip_ratio/low_mean": 7.287439657375217e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00035465199034661054, "completions/clipped_ratio": 0.0, "completions/max_length": 142.5, "completions/mean_length": 73.798828125, "completions/min_length": 33.5, "epoch": 1.9748520710059172, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.014548354595899583, "learning_rate": 7.084353971457331e-07, "loss": 0.0005887392442673444, "reward": 1.63864266872406, "reward_std": 0.20819994062185287, "rewards/FidelityReward/mean": 0.7630402743816376, "rewards/FidelityReward/std": 0.19244354963302612, "rewards/JudgeFidelityReward/mean": 0.7512048184871674, "rewards/JudgeFidelityReward/std": 0.19376454502344131, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2670 }, { "clip_ratio/high_max": 0.0003580460324883461, "clip_ratio/high_mean": 0.00010638394160196185, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00010638394160196185, "completions/clipped_ratio": 0.0, "completions/max_length": 159.0, "completions/mean_length": 72.33723958333333, "completions/min_length": 32.666666666666664, "epoch": 1.9785502958579881, "frac_reward_zero_std": 0.0, "grad_norm": 2.75, "kl": 0.015034406073391437, "learning_rate": 7.073231272347713e-07, "loss": 0.0005416555795818567, "reward": 1.6391195058822632, "reward_std": 0.19370920459429422, "rewards/FidelityReward/mean": 0.7555463512738546, "rewards/FidelityReward/std": 0.20054606596628824, "rewards/JudgeFidelityReward/mean": 0.7671463092168173, "rewards/JudgeFidelityReward/std": 0.17748336990674338, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2675 }, { "clip_ratio/high_max": 0.000988440401852131, "clip_ratio/high_mean": 0.0002775099128484726, "clip_ratio/low_mean": 0.00014275605790317058, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00042026598239317534, "completions/clipped_ratio": 0.001953125, "completions/max_length": 305.0, "completions/mean_length": 73.998046875, "completions/min_length": 31.5, "epoch": 1.9822485207100593, "frac_reward_zero_std": 0.015625, "grad_norm": 2.265625, "kl": 0.016162163764238357, "learning_rate": 7.062096169668548e-07, "loss": 0.0005771161988377572, "reward": 1.7741314768791199, "reward_std": 0.1463194191455841, "rewards/FidelityReward/mean": 0.8515442907810211, "rewards/FidelityReward/std": 0.1775180846452713, "rewards/JudgeFidelityReward/mean": 0.8490805327892303, "rewards/JudgeFidelityReward/std": 0.1727353110909462, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 2680 }, { "clip_ratio/high_max": 0.00041521701496094465, "clip_ratio/high_mean": 0.00017956457450054585, "clip_ratio/low_mean": 2.057613164652139e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002001407032366842, "completions/clipped_ratio": 0.0013020833333333333, "completions/max_length": 233.66666666666666, "completions/mean_length": 74.55598958333333, "completions/min_length": 35.333333333333336, "epoch": 1.98594674556213, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.375, "kl": 0.01574373785406351, "learning_rate": 7.050948730038075e-07, "loss": 0.0008486289530992508, "reward": 1.6723733345667522, "reward_std": 0.19287972648938498, "rewards/FidelityReward/mean": 0.7858856519063314, "rewards/FidelityReward/std": 0.1891702264547348, "rewards/JudgeFidelityReward/mean": 0.77818363904953, "rewards/JudgeFidelityReward/std": 0.19086767236391702, "rewards/SelfEvolvingFormatReward/mean": 0.9947916666666666, "rewards/SelfEvolvingFormatReward/std": 0.0710716222723325, "step": 2685 }, { "clip_ratio/high_max": 0.0008699598838575184, "clip_ratio/high_mean": 0.00015171503619058058, "clip_ratio/low_mean": 0.00011320320190861821, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026491823518881574, "completions/clipped_ratio": 0.0, "completions/max_length": 180.0, "completions/mean_length": 70.60546875, "completions/min_length": 33.0, "epoch": 1.9896449704142012, "frac_reward_zero_std": 0.03125, "grad_norm": 2.71875, "kl": 0.013506192341446876, "learning_rate": 7.039789020148344e-07, "loss": 0.000724924961104989, "reward": 1.6238832473754883, "reward_std": 0.178392194211483, "rewards/FidelityReward/mean": 0.757279098033905, "rewards/FidelityReward/std": 0.1822478398680687, "rewards/JudgeFidelityReward/mean": 0.7351613640785217, "rewards/JudgeFidelityReward/std": 0.1795923188328743, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2690 }, { "clip_ratio/high_max": 0.000562049075961113, "clip_ratio/high_mean": 0.0001758791389875114, "clip_ratio/low_mean": 1.867065002443269e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019454979337751865, "completions/clipped_ratio": 0.0, "completions/max_length": 161.66666666666666, "completions/mean_length": 72.61328125, "completions/min_length": 35.333333333333336, "epoch": 1.9933431952662723, "frac_reward_zero_std": 0.03125, "grad_norm": 2.484375, "kl": 0.016187149845063687, "learning_rate": 7.028617106764814e-07, "loss": 0.0008218894712626934, "reward": 1.6293781995773315, "reward_std": 0.17186433573563895, "rewards/FidelityReward/mean": 0.7493068178494772, "rewards/FidelityReward/std": 0.22947042187054953, "rewards/JudgeFidelityReward/mean": 0.7614448070526123, "rewards/JudgeFidelityReward/std": 0.2139593462149302, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2695 }, { "clip_ratio/high_max": 0.0007442802540026605, "clip_ratio/high_mean": 0.0002156518108677119, "clip_ratio/low_mean": 5.6613470951560886e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002722652803640813, "completions/clipped_ratio": 0.0, "completions/max_length": 171.5, "completions/mean_length": 75.595703125, "completions/min_length": 37.5, "epoch": 1.997041420118343, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.01362482476979494, "learning_rate": 7.017433056725957e-07, "loss": 0.0005566798150539398, "reward": 1.6608547568321228, "reward_std": 0.19866451621055603, "rewards/FidelityReward/mean": 0.7802057862281799, "rewards/FidelityReward/std": 0.20074336975812912, "rewards/JudgeFidelityReward/mean": 0.765204131603241, "rewards/JudgeFidelityReward/std": 0.20178497582674026, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 2700 }, { "clip_ratio/high_max": 0.000524768722243607, "clip_ratio/high_mean": 0.00015202532231342047, "clip_ratio/low_mean": 2.4826216395013036e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001768515387084335, "completions/clipped_ratio": 0.0, "completions/max_length": 173.33333333333334, "completions/mean_length": 71.31640625, "completions/min_length": 34.666666666666664, "epoch": 2.0007396449704142, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.015102091431617736, "learning_rate": 7.006236936942846e-07, "loss": 0.0005513689480721951, "reward": 1.6085118055343628, "reward_std": 0.1881900280714035, "rewards/FidelityReward/mean": 0.7386612097422282, "rewards/FidelityReward/std": 0.23562698562939963, "rewards/JudgeFidelityReward/mean": 0.7410033345222473, "rewards/JudgeFidelityReward/std": 0.22481055557727814, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2705 }, { "clip_ratio/high_max": 0.0008042429806664586, "clip_ratio/high_mean": 0.00027053442026954145, "clip_ratio/low_mean": 0.00024212863354478031, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0005126630770973861, "completions/clipped_ratio": 0.0, "completions/max_length": 152.0, "completions/mean_length": 72.962890625, "completions/min_length": 33.5, "epoch": 2.0044378698224854, "frac_reward_zero_std": 0.03125, "grad_norm": 2.375, "kl": 0.01454570647329092, "learning_rate": 6.995028814398778e-07, "loss": 0.00047571267932653425, "reward": 1.5688904523849487, "reward_std": 0.20232029259204865, "rewards/FidelityReward/mean": 0.7136556208133698, "rewards/FidelityReward/std": 0.21187981963157654, "rewards/JudgeFidelityReward/mean": 0.7124227285385132, "rewards/JudgeFidelityReward/std": 0.21988949924707413, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2710 }, { "clip_ratio/high_max": 0.0005085704149678349, "clip_ratio/high_mean": 0.00012711501331068574, "clip_ratio/low_mean": 0.00010771114029921592, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023482615360990167, "completions/clipped_ratio": 0.0, "completions/max_length": 172.0, "completions/mean_length": 73.56380208333333, "completions/min_length": 37.333333333333336, "epoch": 2.008136094674556, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.015266380086541177, "learning_rate": 6.983808756148845e-07, "loss": 0.0006185178644955158, "reward": 1.6188406149546306, "reward_std": 0.20593236883481345, "rewards/FidelityReward/mean": 0.7444711526234945, "rewards/FidelityReward/std": 0.2019694447517395, "rewards/JudgeFidelityReward/mean": 0.7500409881273905, "rewards/JudgeFidelityReward/std": 0.19579740862051645, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2715 }, { "clip_ratio/high_max": 0.0007353269262239337, "clip_ratio/high_mean": 0.00022050041588954626, "clip_ratio/low_mean": 9.332909830845892e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00031382952001877127, "completions/clipped_ratio": 0.0, "completions/max_length": 159.5, "completions/mean_length": 71.521484375, "completions/min_length": 35.0, "epoch": 2.0118343195266273, "frac_reward_zero_std": 0.015625, "grad_norm": 2.5, "kl": 0.016528791189193724, "learning_rate": 6.972576829319559e-07, "loss": 0.0005262648686766624, "reward": 1.637026607990265, "reward_std": 0.19034132361412048, "rewards/FidelityReward/mean": 0.7622798085212708, "rewards/FidelityReward/std": 0.2095852568745613, "rewards/JudgeFidelityReward/mean": 0.7514469027519226, "rewards/JudgeFidelityReward/std": 0.22445563226938248, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2720 }, { "clip_ratio/high_max": 0.0003232712275348604, "clip_ratio/high_mean": 9.57066920818761e-05, "clip_ratio/low_mean": 2.3342670465353878e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00011904936109203845, "completions/clipped_ratio": 0.0, "completions/max_length": 172.33333333333334, "completions/mean_length": 69.82942708333333, "completions/min_length": 30.666666666666668, "epoch": 2.0155325443786984, "frac_reward_zero_std": 0.052083333333333336, "grad_norm": 2.46875, "kl": 0.014534908346831799, "learning_rate": 6.961333101108432e-07, "loss": 0.0006183644756674767, "reward": 1.6629269917805989, "reward_std": 0.1806085854768753, "rewards/FidelityReward/mean": 0.7743125955263773, "rewards/FidelityReward/std": 0.19651796420415243, "rewards/JudgeFidelityReward/mean": 0.7772286732991537, "rewards/JudgeFidelityReward/std": 0.19790229201316833, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2725 }, { "clip_ratio/high_max": 0.0006858407985419035, "clip_ratio/high_mean": 0.00021996315917931496, "clip_ratio/low_mean": 4.465511883608997e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026461828383617103, "completions/clipped_ratio": 0.0, "completions/max_length": 202.0, "completions/mean_length": 74.59765625, "completions/min_length": 32.5, "epoch": 2.019230769230769, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.01628063227981329, "learning_rate": 6.950077638783578e-07, "loss": 0.00039110779762268065, "reward": 1.620004415512085, "reward_std": 0.18947958946228027, "rewards/FidelityReward/mean": 0.7485758662223816, "rewards/FidelityReward/std": 0.21448824554681778, "rewards/JudgeFidelityReward/mean": 0.7448102831840515, "rewards/JudgeFidelityReward/std": 0.21018889546394348, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2730 }, { "clip_ratio/high_max": 0.000651201605796814, "clip_ratio/high_mean": 0.00018340356764383613, "clip_ratio/low_mean": 3.791875788010657e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002213223313447088, "completions/clipped_ratio": 0.0, "completions/max_length": 176.66666666666666, "completions/mean_length": 74.15364583333333, "completions/min_length": 33.333333333333336, "epoch": 2.0229289940828403, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.0169212793931365, "learning_rate": 6.93881050968332e-07, "loss": 0.0007506803143769502, "reward": 1.6691806713740032, "reward_std": 0.19179207583268484, "rewards/FidelityReward/mean": 0.7813313206036886, "rewards/FidelityReward/std": 0.21994328995545706, "rewards/JudgeFidelityReward/mean": 0.7770006457964579, "rewards/JudgeFidelityReward/std": 0.22009427348772684, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2735 }, { "clip_ratio/high_max": 0.0008984821382910013, "clip_ratio/high_mean": 0.00032449389691464604, "clip_ratio/low_mean": 3.8232465158216655e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00036272635334171354, "completions/clipped_ratio": 0.0, "completions/max_length": 171.0, "completions/mean_length": 71.93359375, "completions/min_length": 35.0, "epoch": 2.026627218934911, "frac_reward_zero_std": 0.0, "grad_norm": 2.4375, "kl": 0.016442243941128254, "learning_rate": 6.927531781215775e-07, "loss": 0.0006571413949131965, "reward": 1.5810081958770752, "reward_std": 0.19793300330638885, "rewards/FidelityReward/mean": 0.7171853482723236, "rewards/FidelityReward/std": 0.2282235026359558, "rewards/JudgeFidelityReward/mean": 0.727645605802536, "rewards/JudgeFidelityReward/std": 0.23028980195522308, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2740 }, { "clip_ratio/high_max": 0.0007682634983211756, "clip_ratio/high_mean": 0.0002434576745145023, "clip_ratio/low_mean": 6.637182959821076e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003098294953815639, "completions/clipped_ratio": 0.0, "completions/max_length": 180.33333333333334, "completions/mean_length": 72.93098958333333, "completions/min_length": 25.333333333333332, "epoch": 2.030325443786982, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 3.078125, "kl": 0.01838693656027317, "learning_rate": 6.916241520858459e-07, "loss": 0.0006632005795836448, "reward": 1.6610791285832722, "reward_std": 0.1855904757976532, "rewards/FidelityReward/mean": 0.7774047454198202, "rewards/FidelityReward/std": 0.20814219613869986, "rewards/JudgeFidelityReward/mean": 0.7686508496602377, "rewards/JudgeFidelityReward/std": 0.21157614390055338, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2745 }, { "clip_ratio/high_max": 0.0011647734441794454, "clip_ratio/high_mean": 0.00033505086321383715, "clip_ratio/low_mean": 6.172870635055006e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003967795870266855, "completions/clipped_ratio": 0.0, "completions/max_length": 161.5, "completions/mean_length": 75.244140625, "completions/min_length": 35.0, "epoch": 2.0340236686390534, "frac_reward_zero_std": 0.0, "grad_norm": 2.984375, "kl": 0.018297480791807173, "learning_rate": 6.904939796157876e-07, "loss": 0.0008819123730063438, "reward": 1.6626707911491394, "reward_std": 0.19232358038425446, "rewards/FidelityReward/mean": 0.773009181022644, "rewards/FidelityReward/std": 0.1972532868385315, "rewards/JudgeFidelityReward/mean": 0.7793232798576355, "rewards/JudgeFidelityReward/std": 0.1799509972333908, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2750 }, { "clip_ratio/high_max": 0.00034213049802929164, "clip_ratio/high_mean": 4.2766312253661455e-05, "clip_ratio/low_mean": 4.123247344978154e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 8.399878715863452e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 184.33333333333334, "completions/mean_length": 74.89322916666667, "completions/min_length": 32.666666666666664, "epoch": 2.037721893491124, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.5, "kl": 0.018948334828019144, "learning_rate": 6.893626674729123e-07, "loss": 0.0006774357985705137, "reward": 1.5757606824239094, "reward_std": 0.2177420457204183, "rewards/FidelityReward/mean": 0.7171011368433634, "rewards/FidelityReward/std": 0.22109708189964294, "rewards/JudgeFidelityReward/mean": 0.7199233174324036, "rewards/JudgeFidelityReward/std": 0.21735966205596924, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 2755 }, { "clip_ratio/high_max": 0.0008500441908836365, "clip_ratio/high_mean": 0.00023406458203680813, "clip_ratio/low_mean": 6.76206371281296e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030168522498570384, "completions/clipped_ratio": 0.0, "completions/max_length": 165.5, "completions/mean_length": 74.060546875, "completions/min_length": 27.5, "epoch": 2.0414201183431953, "frac_reward_zero_std": 0.03125, "grad_norm": 2.390625, "kl": 0.018656093627214432, "learning_rate": 6.882302224255476e-07, "loss": 0.0007134020328521728, "reward": 1.6530235409736633, "reward_std": 0.19315128028392792, "rewards/FidelityReward/mean": 0.7677236497402191, "rewards/FidelityReward/std": 0.20271272212266922, "rewards/JudgeFidelityReward/mean": 0.7705998420715332, "rewards/JudgeFidelityReward/std": 0.20010053366422653, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2760 }, { "clip_ratio/high_max": 0.00046547935344278815, "clip_ratio/high_mean": 0.00013149926089681686, "clip_ratio/low_mean": 2.4875623057596386e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001563748810440302, "completions/clipped_ratio": 0.0, "completions/max_length": 176.0, "completions/mean_length": 74.71744791666667, "completions/min_length": 37.666666666666664, "epoch": 2.0451183431952664, "frac_reward_zero_std": 0.0, "grad_norm": 2.453125, "kl": 0.016618564166128634, "learning_rate": 6.87096651248799e-07, "loss": 0.000587102910503745, "reward": 1.5686043500900269, "reward_std": 0.20566863814989725, "rewards/FidelityReward/mean": 0.7120905717213949, "rewards/FidelityReward/std": 0.2372830013434092, "rewards/JudgeFidelityReward/mean": 0.7130274573961893, "rewards/JudgeFidelityReward/std": 0.23003766437371573, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2765 }, { "clip_ratio/high_max": 0.000883012474514544, "clip_ratio/high_mean": 0.00031823900644667447, "clip_ratio/low_mean": 1.8395879305899143e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003366348857525736, "completions/clipped_ratio": 0.0, "completions/max_length": 261.0, "completions/mean_length": 71.53515625, "completions/min_length": 35.5, "epoch": 2.048816568047337, "frac_reward_zero_std": 0.0, "grad_norm": 3.140625, "kl": 0.01869918592274189, "learning_rate": 6.859619607245101e-07, "loss": 0.0008930721320211888, "reward": 1.6423379778862, "reward_std": 0.1820002645254135, "rewards/FidelityReward/mean": 0.7620126903057098, "rewards/FidelityReward/std": 0.2257905825972557, "rewards/JudgeFidelityReward/mean": 0.7606506049633026, "rewards/JudgeFidelityReward/std": 0.2247934490442276, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2770 }, { "clip_ratio/high_max": 0.0007890345994383096, "clip_ratio/high_mean": 0.000259423244278878, "clip_ratio/low_mean": 4.289941134629771e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003023226512596011, "completions/clipped_ratio": 0.0, "completions/max_length": 218.0, "completions/mean_length": 76.78255208333333, "completions/min_length": 33.666666666666664, "epoch": 2.0525147928994083, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.859375, "kl": 0.017591413483023642, "learning_rate": 6.848261576412199e-07, "loss": 0.0005454749800264836, "reward": 1.7201309204101562, "reward_std": 0.15790812174479166, "rewards/FidelityReward/mean": 0.81498122215271, "rewards/FidelityReward/std": 0.18610667685667673, "rewards/JudgeFidelityReward/mean": 0.8129035830497742, "rewards/JudgeFidelityReward/std": 0.19082778692245483, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 2775 }, { "clip_ratio/high_max": 0.0006916981539689005, "clip_ratio/high_mean": 0.00015340347308665515, "clip_ratio/low_mean": 6.70047229505144e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022040819458197803, "completions/clipped_ratio": 0.0, "completions/max_length": 160.5, "completions/mean_length": 75.75390625, "completions/min_length": 36.0, "epoch": 2.0562130177514795, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.01987551376223564, "learning_rate": 6.836892487941248e-07, "loss": 0.0010386629030108452, "reward": 1.6776992082595825, "reward_std": 0.19739137589931488, "rewards/FidelityReward/mean": 0.7888472676277161, "rewards/FidelityReward/std": 0.19511517882347107, "rewards/JudgeFidelityReward/mean": 0.7816101312637329, "rewards/JudgeFidelityReward/std": 0.1926177740097046, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 2780 }, { "clip_ratio/high_max": 0.0003354138578288257, "clip_ratio/high_mean": 6.139713659649714e-05, "clip_ratio/low_mean": 4.555921186693013e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001069563499186188, "completions/clipped_ratio": 0.0, "completions/max_length": 148.0, "completions/mean_length": 74.02473958333333, "completions/min_length": 30.666666666666668, "epoch": 2.05991124260355, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.017522823251783847, "learning_rate": 6.825512409850357e-07, "loss": 0.0009401892311871052, "reward": 1.5906587839126587, "reward_std": 0.21986796458562216, "rewards/FidelityReward/mean": 0.7260130445162455, "rewards/FidelityReward/std": 0.2189458211263021, "rewards/JudgeFidelityReward/mean": 0.7305935025215149, "rewards/JudgeFidelityReward/std": 0.21118035912513733, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2785 }, { "clip_ratio/high_max": 0.0008116579381749033, "clip_ratio/high_mean": 0.00014380448847077786, "clip_ratio/low_mean": 6.1034011014271525e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020483849802985788, "completions/clipped_ratio": 0.0, "completions/max_length": 154.5, "completions/mean_length": 73.556640625, "completions/min_length": 40.0, "epoch": 2.0636094674556213, "frac_reward_zero_std": 0.015625, "grad_norm": 2.40625, "kl": 0.0195276889950037, "learning_rate": 6.81412141022339e-07, "loss": 0.0008894817903637886, "reward": 1.6874521970748901, "reward_std": 0.16629879921674728, "rewards/FidelityReward/mean": 0.791502445936203, "rewards/FidelityReward/std": 0.21284149587154388, "rewards/JudgeFidelityReward/mean": 0.793852686882019, "rewards/JudgeFidelityReward/std": 0.21799147129058838, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2790 }, { "clip_ratio/high_max": 0.00018691589357331396, "clip_ratio/high_mean": 5.946890451014042e-05, "clip_ratio/low_mean": 5.693481070920825e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00011640372104011476, "completions/clipped_ratio": 0.0, "completions/max_length": 190.33333333333334, "completions/mean_length": 75.45963541666667, "completions/min_length": 37.0, "epoch": 2.0673076923076925, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.71875, "kl": 0.019259612634778024, "learning_rate": 6.802719557209546e-07, "loss": 0.0009956647641956805, "reward": 1.624931812286377, "reward_std": 0.18890221416950226, "rewards/FidelityReward/mean": 0.7492625514666239, "rewards/FidelityReward/std": 0.2399959663550059, "rewards/JudgeFidelityReward/mean": 0.7513383825620016, "rewards/JudgeFidelityReward/std": 0.24084417521953583, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2795 }, { "clip_ratio/high_max": 0.000635388889349997, "clip_ratio/high_mean": 0.00021311125601641834, "clip_ratio/low_mean": 2.7056277031078934e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024016753886826336, "completions/clipped_ratio": 0.0, "completions/max_length": 180.0, "completions/mean_length": 72.39453125, "completions/min_length": 34.5, "epoch": 2.0710059171597632, "frac_reward_zero_std": 0.03125, "grad_norm": 2.609375, "kl": 0.01760468538850546, "learning_rate": 6.791306919022962e-07, "loss": 0.0006398419849574566, "reward": 1.595132827758789, "reward_std": 0.18731174618005753, "rewards/FidelityReward/mean": 0.726445198059082, "rewards/FidelityReward/std": 0.23795732110738754, "rewards/JudgeFidelityReward/mean": 0.7432344853878021, "rewards/JudgeFidelityReward/std": 0.22497785836458206, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 2800 }, { "clip_ratio/high_max": 0.00048769963905215263, "clip_ratio/high_mean": 0.00014107806782703847, "clip_ratio/low_mean": 6.86099214362912e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020968798198737205, "completions/clipped_ratio": 0.0, "completions/max_length": 187.66666666666666, "completions/mean_length": 73.2421875, "completions/min_length": 34.666666666666664, "epoch": 2.0747041420118344, "frac_reward_zero_std": 0.0, "grad_norm": 2.703125, "kl": 0.016674076579511164, "learning_rate": 6.779883563942297e-07, "loss": 0.000793326273560524, "reward": 1.5921604235967, "reward_std": 0.20767963429292044, "rewards/FidelityReward/mean": 0.7292340397834778, "rewards/FidelityReward/std": 0.22626885771751404, "rewards/JudgeFidelityReward/mean": 0.7284570535024008, "rewards/JudgeFidelityReward/std": 0.22587128976980844, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 2805 }, { "clip_ratio/high_max": 0.0011132396757602692, "clip_ratio/high_mean": 0.00037637590430676936, "clip_ratio/low_mean": 0.0001496057549957186, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000525981665123254, "completions/clipped_ratio": 0.0, "completions/max_length": 203.0, "completions/mean_length": 75.1171875, "completions/min_length": 39.0, "epoch": 2.078402366863905, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.018750201910734177, "learning_rate": 6.768449560310326e-07, "loss": 0.0005857614800333977, "reward": 1.545318365097046, "reward_std": 0.20133324712514877, "rewards/FidelityReward/mean": 0.6959640383720398, "rewards/FidelityReward/std": 0.2633272707462311, "rewards/JudgeFidelityReward/mean": 0.7006617784500122, "rewards/JudgeFidelityReward/std": 0.2590363770723343, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2810 }, { "clip_ratio/high_max": 0.0007855338626541197, "clip_ratio/high_mean": 0.00014228928048396484, "clip_ratio/low_mean": 6.891394878039137e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021120322635397314, "completions/clipped_ratio": 0.0, "completions/max_length": 171.0, "completions/mean_length": 73.40364583333333, "completions/min_length": 35.666666666666664, "epoch": 2.0821005917159763, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.640625, "kl": 0.01813248135149479, "learning_rate": 6.757004976533531e-07, "loss": 0.0007692050654441118, "reward": 1.645196557044983, "reward_std": 0.19400031367937723, "rewards/FidelityReward/mean": 0.7644684314727783, "rewards/FidelityReward/std": 0.19457696378231049, "rewards/JudgeFidelityReward/mean": 0.761456290880839, "rewards/JudgeFidelityReward/std": 0.19219831625620523, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2815 }, { "clip_ratio/high_max": 0.0004016549210064113, "clip_ratio/high_mean": 0.00013034670264460145, "clip_ratio/low_mean": 8.205719641409814e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002124038859619759, "completions/clipped_ratio": 0.0, "completions/max_length": 159.5, "completions/mean_length": 73.056640625, "completions/min_length": 39.5, "epoch": 2.0857988165680474, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.01719363033771515, "learning_rate": 6.745549881081691e-07, "loss": 0.0007722051348537206, "reward": 1.744448184967041, "reward_std": 0.15058986842632294, "rewards/FidelityReward/mean": 0.8334174454212189, "rewards/FidelityReward/std": 0.18477115780115128, "rewards/JudgeFidelityReward/mean": 0.8220614492893219, "rewards/JudgeFidelityReward/std": 0.19201408326625824, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2820 }, { "clip_ratio/high_max": 0.0005134819191880524, "clip_ratio/high_mean": 8.603838796261698e-05, "clip_ratio/low_mean": 8.864160627126693e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017467999714426696, "completions/clipped_ratio": 0.0, "completions/max_length": 175.66666666666666, "completions/mean_length": 75.02864583333333, "completions/min_length": 30.0, "epoch": 2.089497041420118, "frac_reward_zero_std": 0.0, "grad_norm": 2.90625, "kl": 0.01719944328069687, "learning_rate": 6.73408434248748e-07, "loss": 0.0006389730144292117, "reward": 1.6631415287653606, "reward_std": 0.20109207928180695, "rewards/FidelityReward/mean": 0.781034012635549, "rewards/FidelityReward/std": 0.19225499033927917, "rewards/JudgeFidelityReward/mean": 0.7668191989262899, "rewards/JudgeFidelityReward/std": 0.20155406494935355, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 2825 }, { "clip_ratio/high_max": 0.001155535038560629, "clip_ratio/high_mean": 0.00044201582204550505, "clip_ratio/low_mean": 4.4125803105998786e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004861416295170784, "completions/clipped_ratio": 0.0, "completions/max_length": 194.0, "completions/mean_length": 74.5859375, "completions/min_length": 35.0, "epoch": 2.0931952662721893, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.01566652376204729, "learning_rate": 6.722608429346043e-07, "loss": 0.0006720192730426788, "reward": 1.7806631922721863, "reward_std": 0.12703217566013336, "rewards/FidelityReward/mean": 0.8493643999099731, "rewards/FidelityReward/std": 0.1772448569536209, "rewards/JudgeFidelityReward/mean": 0.8625975549221039, "rewards/JudgeFidelityReward/std": 0.1476312130689621, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2830 }, { "clip_ratio/high_max": 0.0005394561449065804, "clip_ratio/high_mean": 0.000152353395242244, "clip_ratio/low_mean": 4.387554945424199e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000196228944696486, "completions/clipped_ratio": 0.0, "completions/max_length": 171.33333333333334, "completions/mean_length": 73.984375, "completions/min_length": 30.333333333333332, "epoch": 2.0968934911242605, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.84375, "kl": 0.017882399447262286, "learning_rate": 6.711122210314599e-07, "loss": 0.0007209600880742073, "reward": 1.5819281737009685, "reward_std": 0.21395276983579, "rewards/FidelityReward/mean": 0.717741072177887, "rewards/FidelityReward/std": 0.24267292022705078, "rewards/JudgeFidelityReward/mean": 0.7309783895810446, "rewards/JudgeFidelityReward/std": 0.23060462375481924, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 2835 }, { "clip_ratio/high_max": 0.0006690515438094735, "clip_ratio/high_mean": 0.00019414909766055644, "clip_ratio/low_mean": 8.633969555376098e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002804888106766157, "completions/clipped_ratio": 0.0, "completions/max_length": 218.0, "completions/mean_length": 74.84375, "completions/min_length": 37.5, "epoch": 2.100591715976331, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.016673606634140015, "learning_rate": 6.699625754112014e-07, "loss": 0.0005673878826200962, "reward": 1.634052336215973, "reward_std": 0.20038016140460968, "rewards/FidelityReward/mean": 0.752079576253891, "rewards/FidelityReward/std": 0.20963504165410995, "rewards/JudgeFidelityReward/mean": 0.7639455199241638, "rewards/JudgeFidelityReward/std": 0.20220934599637985, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2840 }, { "clip_ratio/high_max": 0.0005275731324218214, "clip_ratio/high_mean": 0.00019639155361801387, "clip_ratio/low_mean": 5.637487047351897e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00025276642991229893, "completions/clipped_ratio": 0.0, "completions/max_length": 178.0, "completions/mean_length": 75.00911458333333, "completions/min_length": 35.0, "epoch": 2.1042899408284024, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.640625, "kl": 0.016607412323355674, "learning_rate": 6.688119129518413e-07, "loss": 0.0008427195250988006, "reward": 1.6421895027160645, "reward_std": 0.2083270400762558, "rewards/FidelityReward/mean": 0.7599697907765707, "rewards/FidelityReward/std": 0.2066060503323873, "rewards/JudgeFidelityReward/mean": 0.7657414873441061, "rewards/JudgeFidelityReward/std": 0.20029282569885254, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2845 }, { "clip_ratio/high_max": 0.00070039052516222, "clip_ratio/high_mean": 0.0002087660424876958, "clip_ratio/low_mean": 7.689152844250202e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002856575767509639, "completions/clipped_ratio": 0.0, "completions/max_length": 198.0, "completions/mean_length": 73.666015625, "completions/min_length": 28.5, "epoch": 2.1079881656804735, "frac_reward_zero_std": 0.015625, "grad_norm": 2.5, "kl": 0.018434513360261917, "learning_rate": 6.676602405374748e-07, "loss": 0.0010184517130255699, "reward": 1.6621035933494568, "reward_std": 0.19007909297943115, "rewards/FidelityReward/mean": 0.7829279601573944, "rewards/FidelityReward/std": 0.20637823641300201, "rewards/JudgeFidelityReward/mean": 0.7603043913841248, "rewards/JudgeFidelityReward/std": 0.22260870784521103, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2850 }, { "clip_ratio/high_max": 0.0009115375345572829, "clip_ratio/high_mean": 0.00023105795262381434, "clip_ratio/low_mean": 4.4261889706831424e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002753198496066034, "completions/clipped_ratio": 0.0, "completions/max_length": 188.0, "completions/mean_length": 75.08333333333333, "completions/min_length": 29.333333333333332, "epoch": 2.1116863905325443, "frac_reward_zero_std": 0.0, "grad_norm": 2.75, "kl": 0.018010593950748444, "learning_rate": 6.665075650582395e-07, "loss": 0.0008455606177449227, "reward": 1.662784735361735, "reward_std": 0.1861942857503891, "rewards/FidelityReward/mean": 0.7748527526855469, "rewards/FidelityReward/std": 0.209114670753479, "rewards/JudgeFidelityReward/mean": 0.7771658301353455, "rewards/JudgeFidelityReward/std": 0.20057956874370575, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2855 }, { "clip_ratio/high_max": 0.0008560844929888844, "clip_ratio/high_mean": 0.0002630265895277262, "clip_ratio/low_mean": 0.00022983664530329406, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004928632406517863, "completions/clipped_ratio": 0.0, "completions/max_length": 172.0, "completions/mean_length": 75.455078125, "completions/min_length": 38.0, "epoch": 2.1153846153846154, "frac_reward_zero_std": 0.015625, "grad_norm": 2.421875, "kl": 0.016892562806606292, "learning_rate": 6.653538934102742e-07, "loss": 0.00044947098940610885, "reward": 1.6435672640800476, "reward_std": 0.1917765811085701, "rewards/FidelityReward/mean": 0.7603543400764465, "rewards/FidelityReward/std": 0.20862097293138504, "rewards/JudgeFidelityReward/mean": 0.7664258182048798, "rewards/JudgeFidelityReward/std": 0.20219774544239044, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2860 }, { "clip_ratio/high_max": 0.00035987760638818146, "clip_ratio/high_mean": 0.00016512847505509854, "clip_ratio/low_mean": 0.00014563188306055964, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003107603522948921, "completions/clipped_ratio": 0.0, "completions/max_length": 184.0, "completions/mean_length": 74.34765625, "completions/min_length": 34.0, "epoch": 2.1190828402366866, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.017686332017183302, "learning_rate": 6.641992324956775e-07, "loss": 0.0007072495296597481, "reward": 1.5981428225835164, "reward_std": 0.21633962293465933, "rewards/FidelityReward/mean": 0.7327561974525452, "rewards/FidelityReward/std": 0.2301261822382609, "rewards/JudgeFidelityReward/mean": 0.7320753733317057, "rewards/JudgeFidelityReward/std": 0.23484799762566885, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2865 }, { "clip_ratio/high_max": 0.0008717830060049891, "clip_ratio/high_mean": 0.00029241542797535657, "clip_ratio/low_mean": 7.989512378117069e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00037231054157018664, "completions/clipped_ratio": 0.0, "completions/max_length": 161.5, "completions/mean_length": 75.447265625, "completions/min_length": 31.0, "epoch": 2.1227810650887573, "frac_reward_zero_std": 0.015625, "grad_norm": 2.5625, "kl": 0.015613247826695442, "learning_rate": 6.630435892224666e-07, "loss": 0.0004563259892165661, "reward": 1.5998401045799255, "reward_std": 0.2041073739528656, "rewards/FidelityReward/mean": 0.731361597776413, "rewards/FidelityReward/std": 0.21375016868114471, "rewards/JudgeFidelityReward/mean": 0.736956924200058, "rewards/JudgeFidelityReward/std": 0.2089282050728798, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2870 }, { "clip_ratio/high_max": 0.0006470867665484548, "clip_ratio/high_mean": 0.00016880736220628023, "clip_ratio/low_mean": 8.458309603156521e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00025339046260342, "completions/clipped_ratio": 0.0, "completions/max_length": 210.0, "completions/mean_length": 75.72786458333333, "completions/min_length": 32.333333333333336, "epoch": 2.1264792899408285, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.016540315188467502, "learning_rate": 6.618869705045352e-07, "loss": 0.0007583241909742356, "reward": 1.6139439741770427, "reward_std": 0.20147870481014252, "rewards/FidelityReward/mean": 0.7403029600779215, "rewards/FidelityReward/std": 0.2073572427034378, "rewards/JudgeFidelityReward/mean": 0.747282067934672, "rewards/JudgeFidelityReward/std": 0.20983444650967917, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2875 }, { "clip_ratio/high_max": 0.0009859429788775741, "clip_ratio/high_mean": 0.00034458564477972685, "clip_ratio/low_mean": 0.00012848750047851355, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00047307314816862344, "completions/clipped_ratio": 0.0, "completions/max_length": 198.5, "completions/mean_length": 74.771484375, "completions/min_length": 36.5, "epoch": 2.1301775147928996, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.015367130935192107, "learning_rate": 6.607293832616139e-07, "loss": 0.0008114211261272431, "reward": 1.5923900604248047, "reward_std": 0.19773487746715546, "rewards/FidelityReward/mean": 0.7247701287269592, "rewards/FidelityReward/std": 0.2509039416909218, "rewards/JudgeFidelityReward/mean": 0.7352397441864014, "rewards/JudgeFidelityReward/std": 0.24263641983270645, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2880 }, { "clip_ratio/high_max": 0.0003678441629745066, "clip_ratio/high_mean": 4.5980520371813324e-05, "clip_ratio/low_mean": 5.7985079183708874e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001039655995555222, "completions/clipped_ratio": 0.0, "completions/max_length": 198.0, "completions/mean_length": 76.57552083333333, "completions/min_length": 35.333333333333336, "epoch": 2.1338757396449703, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.390625, "kl": 0.016343932785093784, "learning_rate": 6.595708344192268e-07, "loss": 0.0008103696629405022, "reward": 1.6660150686899822, "reward_std": 0.18260161578655243, "rewards/FidelityReward/mean": 0.7767817378044128, "rewards/FidelityReward/std": 0.19959326088428497, "rewards/JudgeFidelityReward/mean": 0.7797686457633972, "rewards/JudgeFidelityReward/std": 0.19721036156018576, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2885 }, { "clip_ratio/high_max": 0.0014130175579339266, "clip_ratio/high_mean": 0.0004256577230989933, "clip_ratio/low_mean": 5.7565023598726836e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00048322274815291164, "completions/clipped_ratio": 0.0, "completions/max_length": 216.5, "completions/mean_length": 72.65234375, "completions/min_length": 34.0, "epoch": 2.1375739644970415, "frac_reward_zero_std": 0.0, "grad_norm": 2.5, "kl": 0.016202655248343946, "learning_rate": 6.584113309086515e-07, "loss": 0.0006746497005224228, "reward": 1.6520000100135803, "reward_std": 0.1664390042424202, "rewards/FidelityReward/mean": 0.7606542408466339, "rewards/FidelityReward/std": 0.2189224734902382, "rewards/JudgeFidelityReward/mean": 0.782691478729248, "rewards/JudgeFidelityReward/std": 0.1929110586643219, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2890 }, { "clip_ratio/high_max": 0.0005306389066390693, "clip_ratio/high_mean": 0.00021112469985382632, "clip_ratio/low_mean": 2.0610057981684803e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023173475638031958, "completions/clipped_ratio": 0.0, "completions/max_length": 176.66666666666666, "completions/mean_length": 73.1953125, "completions/min_length": 37.333333333333336, "epoch": 2.1412721893491122, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.78125, "kl": 0.016003014333546162, "learning_rate": 6.572508796668768e-07, "loss": 0.0007518887985497713, "reward": 1.655216892560323, "reward_std": 0.18973020712534586, "rewards/FidelityReward/mean": 0.7614246606826782, "rewards/FidelityReward/std": 0.2050350954135259, "rewards/JudgeFidelityReward/mean": 0.7901886701583862, "rewards/JudgeFidelityReward/std": 0.18444555501143137, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 2895 }, { "clip_ratio/high_max": 0.0012402048916555942, "clip_ratio/high_mean": 0.0003506128865410574, "clip_ratio/low_mean": 8.584026218159125e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000436453172005713, "completions/clipped_ratio": 0.0, "completions/max_length": 185.0, "completions/mean_length": 73.96484375, "completions/min_length": 34.0, "epoch": 2.1449704142011834, "frac_reward_zero_std": 0.0, "grad_norm": 2.796875, "kl": 0.018282346799969672, "learning_rate": 6.560894876365619e-07, "loss": 0.0004769914317876101, "reward": 1.640297770500183, "reward_std": 0.20059245079755783, "rewards/FidelityReward/mean": 0.7569658160209656, "rewards/FidelityReward/std": 0.21780996769666672, "rewards/JudgeFidelityReward/mean": 0.7686171531677246, "rewards/JudgeFidelityReward/std": 0.20664602518081665, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2900 }, { "clip_ratio/high_max": 0.0007108147954568266, "clip_ratio/high_mean": 0.00013559840735979378, "clip_ratio/low_mean": 0.00015828523901291193, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002938836463727057, "completions/clipped_ratio": 0.0, "completions/max_length": 207.66666666666666, "completions/mean_length": 72.06640625, "completions/min_length": 34.0, "epoch": 2.1486686390532546, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.46875, "kl": 0.01603647358715534, "learning_rate": 6.54927161765994e-07, "loss": 0.0006155974231660366, "reward": 1.6415119965871174, "reward_std": 0.190776859720548, "rewards/FidelityReward/mean": 0.7607495387395223, "rewards/FidelityReward/std": 0.20737874507904053, "rewards/JudgeFidelityReward/mean": 0.764129122098287, "rewards/JudgeFidelityReward/std": 0.19574383397897085, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 2905 }, { "clip_ratio/high_max": 0.0008900798973627388, "clip_ratio/high_mean": 0.00028288940666243434, "clip_ratio/low_mean": 4.662004794226959e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003295094647910446, "completions/clipped_ratio": 0.0, "completions/max_length": 198.5, "completions/mean_length": 78.44921875, "completions/min_length": 34.0, "epoch": 2.1523668639053253, "frac_reward_zero_std": 0.0, "grad_norm": 2.34375, "kl": 0.016694756224751472, "learning_rate": 6.537639090090476e-07, "loss": 0.0005474820733070374, "reward": 1.5809057354927063, "reward_std": 0.222450390458107, "rewards/FidelityReward/mean": 0.7205064296722412, "rewards/FidelityReward/std": 0.23940909653902054, "rewards/JudgeFidelityReward/mean": 0.720798671245575, "rewards/JudgeFidelityReward/std": 0.2240142971277237, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2910 }, { "clip_ratio/high_max": 0.00048149267677217723, "clip_ratio/high_mean": 0.00024651854182593523, "clip_ratio/low_mean": 7.444721122737974e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003209657501429319, "completions/clipped_ratio": 0.0, "completions/max_length": 172.33333333333334, "completions/mean_length": 72.77864583333333, "completions/min_length": 35.666666666666664, "epoch": 2.1560650887573964, "frac_reward_zero_std": 0.041666666666666664, "grad_norm": 2.34375, "kl": 0.016096273064613344, "learning_rate": 6.525997363251424e-07, "loss": 0.000564609793946147, "reward": 1.6843785842259724, "reward_std": 0.17153102656205496, "rewards/FidelityReward/mean": 0.7873240311940511, "rewards/FidelityReward/std": 0.20162475605805716, "rewards/JudgeFidelityReward/mean": 0.7941090861956278, "rewards/JudgeFidelityReward/std": 0.18967753648757935, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2915 }, { "clip_ratio/high_max": 0.0010651611955836415, "clip_ratio/high_mean": 0.000340420042630285, "clip_ratio/low_mean": 9.012017399072647e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004305402166210115, "completions/clipped_ratio": 0.0, "completions/max_length": 171.5, "completions/mean_length": 71.775390625, "completions/min_length": 33.5, "epoch": 2.1597633136094676, "frac_reward_zero_std": 0.015625, "grad_norm": 2.6875, "kl": 0.01711706966161728, "learning_rate": 6.514346506792016e-07, "loss": 0.0005811966024339199, "reward": 1.699138343334198, "reward_std": 0.16607750207185745, "rewards/FidelityReward/mean": 0.794933021068573, "rewards/FidelityReward/std": 0.19668793678283691, "rewards/JudgeFidelityReward/mean": 0.8084107041358948, "rewards/JudgeFidelityReward/std": 0.17692308127880096, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2920 }, { "clip_ratio/high_max": 0.0005160517059266568, "clip_ratio/high_mean": 0.00010988438152708113, "clip_ratio/low_mean": 6.276420899666845e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001726485788822174, "completions/clipped_ratio": 0.0, "completions/max_length": 182.33333333333334, "completions/mean_length": 75.05598958333333, "completions/min_length": 36.333333333333336, "epoch": 2.1634615384615383, "frac_reward_zero_std": 0.0, "grad_norm": 2.265625, "kl": 0.01768072508275509, "learning_rate": 6.502686590416105e-07, "loss": 0.0008276519365608692, "reward": 1.6592919031778972, "reward_std": 0.2028618554274241, "rewards/FidelityReward/mean": 0.7745192845662435, "rewards/FidelityReward/std": 0.2142652620871862, "rewards/JudgeFidelityReward/mean": 0.77475372950236, "rewards/JudgeFidelityReward/std": 0.21470533311367035, "rewards/SelfEvolvingFormatReward/mean": 0.9947916666666666, "rewards/SelfEvolvingFormatReward/std": 0.04142084469397863, "step": 2925 }, { "clip_ratio/high_max": 0.0006866002571769058, "clip_ratio/high_mean": 0.00021054648095741868, "clip_ratio/low_mean": 8.556694083381445e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00029611343052238226, "completions/clipped_ratio": 0.0, "completions/max_length": 213.5, "completions/mean_length": 74.16796875, "completions/min_length": 32.5, "epoch": 2.1671597633136095, "frac_reward_zero_std": 0.015625, "grad_norm": 2.53125, "kl": 0.016161418706178664, "learning_rate": 6.491017683881748e-07, "loss": 0.0007942074909806252, "reward": 1.6392662525177002, "reward_std": 0.1773008331656456, "rewards/FidelityReward/mean": 0.756668746471405, "rewards/FidelityReward/std": 0.21155722439289093, "rewards/JudgeFidelityReward/mean": 0.7651950716972351, "rewards/JudgeFidelityReward/std": 0.20481012016534805, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2930 }, { "clip_ratio/high_max": 0.0003717934596352279, "clip_ratio/high_mean": 8.617632847744972e-05, "clip_ratio/low_mean": 4.697818367276341e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00013315451506059616, "completions/clipped_ratio": 0.0, "completions/max_length": 156.0, "completions/mean_length": 74.5390625, "completions/min_length": 29.333333333333332, "epoch": 2.1708579881656807, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.484375, "kl": 0.017121563851833343, "learning_rate": 6.479339857000788e-07, "loss": 0.0006648990325629711, "reward": 1.6555214722951253, "reward_std": 0.1835722972949346, "rewards/FidelityReward/mean": 0.7689182559649149, "rewards/FidelityReward/std": 0.21259663999080658, "rewards/JudgeFidelityReward/mean": 0.7758105397224426, "rewards/JudgeFidelityReward/std": 0.2062821884950002, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 2935 }, { "clip_ratio/high_max": 0.0007348254788666963, "clip_ratio/high_mean": 0.00023186092730611562, "clip_ratio/low_mean": 2.2084805823396892e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002539457404054701, "completions/clipped_ratio": 0.0, "completions/max_length": 184.5, "completions/mean_length": 73.2578125, "completions/min_length": 28.0, "epoch": 2.1745562130177514, "frac_reward_zero_std": 0.03125, "grad_norm": 2.59375, "kl": 0.015372379124164582, "learning_rate": 6.467653179638433e-07, "loss": 0.0006666826084256172, "reward": 1.6264082193374634, "reward_std": 0.18858779966831207, "rewards/FidelityReward/mean": 0.7498191893100739, "rewards/FidelityReward/std": 0.20054329186677933, "rewards/JudgeFidelityReward/mean": 0.7531780302524567, "rewards/JudgeFidelityReward/std": 0.1846776306629181, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2940 }, { "clip_ratio/high_max": 0.0007590303663164377, "clip_ratio/high_mean": 0.00021845769369974733, "clip_ratio/low_mean": 2.2522523067891598e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002409802284091711, "completions/clipped_ratio": 0.0, "completions/max_length": 168.66666666666666, "completions/mean_length": 71.86067708333333, "completions/min_length": 35.333333333333336, "epoch": 2.1782544378698225, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.546875, "kl": 0.016840270720422268, "learning_rate": 6.455957721712844e-07, "loss": 0.0006954793818295002, "reward": 1.6165039936701457, "reward_std": 0.1983812948067983, "rewards/FidelityReward/mean": 0.7472459673881531, "rewards/FidelityReward/std": 0.2252814769744873, "rewards/JudgeFidelityReward/mean": 0.7411202192306519, "rewards/JudgeFidelityReward/std": 0.22686311105887094, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 2945 }, { "clip_ratio/high_max": 0.0008640191983431578, "clip_ratio/high_mean": 0.00013192584447097033, "clip_ratio/low_mean": 8.634739060653374e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021827323944307865, "completions/clipped_ratio": 0.0, "completions/max_length": 153.0, "completions/mean_length": 73.8828125, "completions/min_length": 29.5, "epoch": 2.1819526627218937, "frac_reward_zero_std": 0.015625, "grad_norm": 2.640625, "kl": 0.015148697979748249, "learning_rate": 6.444253553194712e-07, "loss": 0.0007589250337332488, "reward": 1.6455696821212769, "reward_std": 0.188000850379467, "rewards/FidelityReward/mean": 0.7612248659133911, "rewards/FidelityReward/std": 0.2106432095170021, "rewards/JudgeFidelityReward/mean": 0.7686895728111267, "rewards/JudgeFidelityReward/std": 0.1880006268620491, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2950 }, { "clip_ratio/high_max": 0.00045794405741617084, "clip_ratio/high_mean": 0.00010017243766924366, "clip_ratio/low_mean": 2.3342670465353878e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00012351510667940602, "completions/clipped_ratio": 0.0, "completions/max_length": 167.66666666666666, "completions/mean_length": 75.53776041666667, "completions/min_length": 30.666666666666668, "epoch": 2.1856508875739644, "frac_reward_zero_std": 0.0, "grad_norm": 2.734375, "kl": 0.016483370400965213, "learning_rate": 6.432540744106841e-07, "loss": 0.000801404844969511, "reward": 1.6709748109181721, "reward_std": 0.19402082761128744, "rewards/FidelityReward/mean": 0.7783790429433187, "rewards/FidelityReward/std": 0.19870525101820627, "rewards/JudgeFidelityReward/mean": 0.7864935398101807, "rewards/JudgeFidelityReward/std": 0.19287160535653433, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2955 }, { "clip_ratio/high_max": 0.0006663661217316986, "clip_ratio/high_mean": 0.00023461868404410778, "clip_ratio/low_mean": 0.00012883095478173346, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003634496359154582, "completions/clipped_ratio": 0.0, "completions/max_length": 262.5, "completions/mean_length": 72.4765625, "completions/min_length": 30.5, "epoch": 2.1893491124260356, "frac_reward_zero_std": 0.03125, "grad_norm": 2.734375, "kl": 0.017647482827305792, "learning_rate": 6.420819364523731e-07, "loss": 0.0007255742326378823, "reward": 1.6441615223884583, "reward_std": 0.20080187171697617, "rewards/FidelityReward/mean": 0.7656005918979645, "rewards/FidelityReward/std": 0.20955757796764374, "rewards/JudgeFidelityReward/mean": 0.7590749263763428, "rewards/JudgeFidelityReward/std": 0.20862530171871185, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 2960 }, { "clip_ratio/high_max": 0.000706102536059916, "clip_ratio/high_mean": 0.0002247643074952066, "clip_ratio/low_mean": 4.201164992991835e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002667759545147419, "completions/clipped_ratio": 0.0, "completions/max_length": 170.66666666666666, "completions/mean_length": 74.73567708333333, "completions/min_length": 31.333333333333332, "epoch": 2.1930473372781063, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.5, "kl": 0.014432473108172417, "learning_rate": 6.409089484571154e-07, "loss": 0.0005898437928408384, "reward": 1.6667253176371257, "reward_std": 0.19236637651920319, "rewards/FidelityReward/mean": 0.7758928338686625, "rewards/FidelityReward/std": 0.19885201255480447, "rewards/JudgeFidelityReward/mean": 0.7829669515291849, "rewards/JudgeFidelityReward/std": 0.19351947804292044, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2965 }, { "clip_ratio/high_max": 0.0006949352798983454, "clip_ratio/high_mean": 0.00022762094740755857, "clip_ratio/low_mean": 6.837066466687247e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00029599161935038864, "completions/clipped_ratio": 0.0, "completions/max_length": 170.0, "completions/mean_length": 75.04296875, "completions/min_length": 38.5, "epoch": 2.1967455621301775, "frac_reward_zero_std": 0.015625, "grad_norm": 2.609375, "kl": 0.016809702292084693, "learning_rate": 6.39735117442574e-07, "loss": 0.0008083190768957138, "reward": 1.597701072692871, "reward_std": 0.1963818222284317, "rewards/FidelityReward/mean": 0.73113813996315, "rewards/FidelityReward/std": 0.21635078638792038, "rewards/JudgeFidelityReward/mean": 0.7331256866455078, "rewards/JudgeFidelityReward/std": 0.2204383835196495, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2970 }, { "clip_ratio/high_max": 0.0008276113774627447, "clip_ratio/high_mean": 0.0002675181720405817, "clip_ratio/low_mean": 1.9561815133783966e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002870799857191741, "completions/clipped_ratio": 0.0, "completions/max_length": 168.33333333333334, "completions/mean_length": 72.28125, "completions/min_length": 36.0, "epoch": 2.2004437869822486, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.609375, "kl": 0.016987235844135286, "learning_rate": 6.385604504314554e-07, "loss": 0.0009102350100874901, "reward": 1.6167871952056885, "reward_std": 0.19146967927614847, "rewards/FidelityReward/mean": 0.7469499111175537, "rewards/FidelityReward/std": 0.22593501210212708, "rewards/JudgeFidelityReward/mean": 0.7422787547111511, "rewards/JudgeFidelityReward/std": 0.2311530808607737, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 2975 }, { "clip_ratio/high_max": 0.0011833761120215059, "clip_ratio/high_mean": 0.0002685867657419294, "clip_ratio/low_mean": 2.095557429129258e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028954234439879654, "completions/clipped_ratio": 0.0, "completions/max_length": 162.5, "completions/mean_length": 72.48828125, "completions/min_length": 33.0, "epoch": 2.2041420118343193, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.01537193451076746, "learning_rate": 6.373849544514676e-07, "loss": 0.0005997465923428535, "reward": 1.6214070916175842, "reward_std": 0.19684109836816788, "rewards/FidelityReward/mean": 0.7506342530250549, "rewards/FidelityReward/std": 0.19741393625736237, "rewards/JudgeFidelityReward/mean": 0.741545706987381, "rewards/JudgeFidelityReward/std": 0.19735760986804962, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2980 }, { "clip_ratio/high_max": 0.0005511728813871741, "clip_ratio/high_mean": 0.0001571831468027085, "clip_ratio/low_mean": 6.649575952906161e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022367891506291926, "completions/clipped_ratio": 0.0, "completions/max_length": 186.66666666666666, "completions/mean_length": 74.84244791666667, "completions/min_length": 33.666666666666664, "epoch": 2.2078402366863905, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.578125, "kl": 0.0154623968526721, "learning_rate": 6.36208636535278e-07, "loss": 0.000613330490887165, "reward": 1.6301932334899902, "reward_std": 0.18974654376506805, "rewards/FidelityReward/mean": 0.7537968953450521, "rewards/FidelityReward/std": 0.23365326722462973, "rewards/JudgeFidelityReward/mean": 0.75409464041392, "rewards/JudgeFidelityReward/std": 0.2324125568072001, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 2985 }, { "clip_ratio/high_max": 0.00048034782521426677, "clip_ratio/high_mean": 0.00019052112475037575, "clip_ratio/low_mean": 4.534895706456155e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002358700832701288, "completions/clipped_ratio": 0.0, "completions/max_length": 153.5, "completions/mean_length": 71.880859375, "completions/min_length": 34.0, "epoch": 2.2115384615384617, "frac_reward_zero_std": 0.0, "grad_norm": 2.40625, "kl": 0.017143089324235916, "learning_rate": 6.350315037204713e-07, "loss": 0.0005977375898510217, "reward": 1.6823139786720276, "reward_std": 0.18749352544546127, "rewards/FidelityReward/mean": 0.789149820804596, "rewards/FidelityReward/std": 0.21605238318443298, "rewards/JudgeFidelityReward/mean": 0.7902345657348633, "rewards/JudgeFidelityReward/std": 0.21320325881242752, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 2990 }, { "clip_ratio/high_max": 0.0005797799676656723, "clip_ratio/high_mean": 0.00013365656486712395, "clip_ratio/low_mean": 8.542739815311506e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002190839615650475, "completions/clipped_ratio": 0.0, "completions/max_length": 174.33333333333334, "completions/mean_length": 74.17317708333333, "completions/min_length": 36.333333333333336, "epoch": 2.2152366863905324, "frac_reward_zero_std": 0.0, "grad_norm": 2.671875, "kl": 0.015723571367561816, "learning_rate": 6.338535630495081e-07, "loss": 0.0006747205276042223, "reward": 1.6289713382720947, "reward_std": 0.2011313041051229, "rewards/FidelityReward/mean": 0.7501131494839987, "rewards/FidelityReward/std": 0.2155277281999588, "rewards/JudgeFidelityReward/mean": 0.7577164570490519, "rewards/JudgeFidelityReward/std": 0.2132397542397181, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 2995 }, { "clip_ratio/high_max": 0.0005273066461086273, "clip_ratio/high_mean": 0.0001317442482104525, "clip_ratio/low_mean": 6.386269960785284e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001956069390871562, "completions/clipped_ratio": 0.0, "completions/max_length": 196.0, "completions/mean_length": 72.52734375, "completions/min_length": 35.0, "epoch": 2.2189349112426036, "frac_reward_zero_std": 0.0, "grad_norm": 2.65625, "kl": 0.016639237850904466, "learning_rate": 6.326748215696816e-07, "loss": 0.0007349246181547642, "reward": 1.7495412230491638, "reward_std": 0.15195902436971664, "rewards/FidelityReward/mean": 0.8363411724567413, "rewards/FidelityReward/std": 0.1848525032401085, "rewards/JudgeFidelityReward/mean": 0.8264001309871674, "rewards/JudgeFidelityReward/std": 0.18203572183847427, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3000 }, { "clip_ratio/high_max": 0.0007985495263710618, "clip_ratio/high_mean": 0.00024362635740544648, "clip_ratio/low_mean": 5.962440045550466e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003032507607713342, "completions/clipped_ratio": 0.0, "completions/max_length": 179.66666666666666, "completions/mean_length": 72.7890625, "completions/min_length": 35.666666666666664, "epoch": 2.2226331360946747, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.75, "kl": 0.016795761324465273, "learning_rate": 6.314952863330762e-07, "loss": 0.0006198873743414879, "reward": 1.6405048767725627, "reward_std": 0.18988269567489624, "rewards/FidelityReward/mean": 0.7590699791908264, "rewards/FidelityReward/std": 0.21963355938593546, "rewards/JudgeFidelityReward/mean": 0.7628698348999023, "rewards/JudgeFidelityReward/std": 0.20989008247852325, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3005 }, { "clip_ratio/high_max": 0.0009074836736544967, "clip_ratio/high_mean": 0.0003360811620950699, "clip_ratio/low_mean": 8.604097092757002e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004221221315674484, "completions/clipped_ratio": 0.0, "completions/max_length": 178.0, "completions/mean_length": 74.333984375, "completions/min_length": 30.5, "epoch": 2.2263313609467454, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.017585107311606406, "learning_rate": 6.30314964396525e-07, "loss": 0.0005875503178685903, "reward": 1.5613596439361572, "reward_std": 0.22221679985523224, "rewards/FidelityReward/mean": 0.7133005559444427, "rewards/FidelityReward/std": 0.2217201367020607, "rewards/JudgeFidelityReward/mean": 0.6980712711811066, "rewards/JudgeFidelityReward/std": 0.23375745862722397, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3010 }, { "clip_ratio/high_max": 0.0004015641403384507, "clip_ratio/high_mean": 0.0001543781196232885, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001543781196232885, "completions/clipped_ratio": 0.0, "completions/max_length": 216.0, "completions/mean_length": 72.75390625, "completions/min_length": 36.333333333333336, "epoch": 2.2300295857988166, "frac_reward_zero_std": 0.03125, "grad_norm": 2.65625, "kl": 0.016512858122587203, "learning_rate": 6.29133862821568e-07, "loss": 0.000696277990937233, "reward": 1.6319554249445598, "reward_std": 0.188619926571846, "rewards/FidelityReward/mean": 0.7505512634913126, "rewards/FidelityReward/std": 0.22570825616518655, "rewards/JudgeFidelityReward/mean": 0.7641104261080424, "rewards/JudgeFidelityReward/std": 0.21028096477190653, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3015 }, { "clip_ratio/high_max": 0.0010419904254376888, "clip_ratio/high_mean": 0.0002817718719597906, "clip_ratio/low_mean": 9.155669540632516e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000373328581918031, "completions/clipped_ratio": 0.0, "completions/max_length": 157.5, "completions/mean_length": 72.642578125, "completions/min_length": 31.0, "epoch": 2.2337278106508878, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.017019039765000344, "learning_rate": 6.279519886744095e-07, "loss": 0.0006866198964416981, "reward": 1.6923872232437134, "reward_std": 0.17992699891328812, "rewards/FidelityReward/mean": 0.7882467210292816, "rewards/FidelityReward/std": 0.20549728721380234, "rewards/JudgeFidelityReward/mean": 0.8082811534404755, "rewards/JudgeFidelityReward/std": 0.1908186674118042, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3020 }, { "clip_ratio/high_max": 0.0007087185513228178, "clip_ratio/high_mean": 0.0002467655169311911, "clip_ratio/low_mean": 4.223080613883212e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002889963332563639, "completions/clipped_ratio": 0.0, "completions/max_length": 189.0, "completions/mean_length": 72.14713541666667, "completions/min_length": 33.333333333333336, "epoch": 2.2374260355029585, "frac_reward_zero_std": 0.03125, "grad_norm": 2.78125, "kl": 0.015439023077487946, "learning_rate": 6.267693490258754e-07, "loss": 0.0005925821140408515, "reward": 1.6251567204793294, "reward_std": 0.19018288453420004, "rewards/FidelityReward/mean": 0.7460311651229858, "rewards/FidelityReward/std": 0.2040910323460897, "rewards/JudgeFidelityReward/mean": 0.7582511305809021, "rewards/JudgeFidelityReward/std": 0.18161286413669586, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3025 }, { "clip_ratio/high_max": 0.0007784104440361261, "clip_ratio/high_mean": 0.00024202548083849252, "clip_ratio/low_mean": 9.390826598973945e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00033593373955227436, "completions/clipped_ratio": 0.0, "completions/max_length": 186.5, "completions/mean_length": 72.634765625, "completions/min_length": 33.0, "epoch": 2.2411242603550297, "frac_reward_zero_std": 0.015625, "grad_norm": 2.609375, "kl": 0.016266081668436527, "learning_rate": 6.255859509513723e-07, "loss": 0.0007130702957510948, "reward": 1.578392744064331, "reward_std": 0.2061154544353485, "rewards/FidelityReward/mean": 0.7140225768089294, "rewards/FidelityReward/std": 0.214666448533535, "rewards/JudgeFidelityReward/mean": 0.7287402451038361, "rewards/JudgeFidelityReward/std": 0.19522812217473984, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3030 }, { "clip_ratio/high_max": 0.0007309678476303816, "clip_ratio/high_mean": 0.00017862593522295357, "clip_ratio/low_mean": 8.86743247974664e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000267300265841186, "completions/clipped_ratio": 0.0, "completions/max_length": 181.66666666666666, "completions/mean_length": 71.88151041666667, "completions/min_length": 31.666666666666668, "epoch": 2.2448224852071004, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.015990465134382247, "learning_rate": 6.244018015308435e-07, "loss": 0.00072804088704288, "reward": 1.7127244075139363, "reward_std": 0.17288097739219666, "rewards/FidelityReward/mean": 0.8095410068829855, "rewards/FidelityReward/std": 0.20668030281861624, "rewards/JudgeFidelityReward/mean": 0.8076688647270203, "rewards/JudgeFidelityReward/std": 0.20156821111838022, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3035 }, { "clip_ratio/high_max": 0.0007254379452206194, "clip_ratio/high_mean": 0.00018830595508916302, "clip_ratio/low_mean": 0.00012744474806822837, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003157507089781575, "completions/clipped_ratio": 0.0, "completions/max_length": 149.5, "completions/mean_length": 71.373046875, "completions/min_length": 30.5, "epoch": 2.2485207100591715, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.015252771228551865, "learning_rate": 6.232169078487274e-07, "loss": 0.000602518953382969, "reward": 1.5772032737731934, "reward_std": 0.20844214409589767, "rewards/FidelityReward/mean": 0.7173737585544586, "rewards/FidelityReward/std": 0.23317941278219223, "rewards/JudgeFidelityReward/mean": 0.7196590602397919, "rewards/JudgeFidelityReward/std": 0.2356383353471756, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3040 }, { "clip_ratio/high_max": 0.0005615325411781668, "clip_ratio/high_mean": 0.00023439248325303198, "clip_ratio/low_mean": 4.46334553998895e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027902593137696385, "completions/clipped_ratio": 0.0, "completions/max_length": 185.33333333333334, "completions/mean_length": 74.50130208333333, "completions/min_length": 29.333333333333332, "epoch": 2.2522189349112427, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.018780699744820595, "learning_rate": 6.220312769939154e-07, "loss": 0.0007248505484312773, "reward": 1.5978682438532512, "reward_std": 0.19767693181832632, "rewards/FidelityReward/mean": 0.7293354471524557, "rewards/FidelityReward/std": 0.2376572291056315, "rewards/JudgeFidelityReward/mean": 0.7370656132698059, "rewards/JudgeFidelityReward/std": 0.22902460396289825, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3045 }, { "clip_ratio/high_max": 0.0008233136963099241, "clip_ratio/high_mean": 0.0002522564318496734, "clip_ratio/low_mean": 5.187224160181358e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030412867199629544, "completions/clipped_ratio": 0.0, "completions/max_length": 178.0, "completions/mean_length": 71.361328125, "completions/min_length": 32.0, "epoch": 2.2559171597633134, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.015379297360777856, "learning_rate": 6.208449160597095e-07, "loss": 0.0006606098730117082, "reward": 1.6371365785598755, "reward_std": 0.19992053508758545, "rewards/FidelityReward/mean": 0.7581654191017151, "rewards/FidelityReward/std": 0.21087098121643066, "rewards/JudgeFidelityReward/mean": 0.7598954439163208, "rewards/JudgeFidelityReward/std": 0.21450292319059372, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3050 }, { "clip_ratio/high_max": 0.0005350661696866155, "clip_ratio/high_mean": 0.00010865298099815845, "clip_ratio/low_mean": 4.507509147515521e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001537280681077391, "completions/clipped_ratio": 0.0, "completions/max_length": 204.0, "completions/mean_length": 72.46614583333333, "completions/min_length": 33.333333333333336, "epoch": 2.2596153846153846, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.01624514851719141, "learning_rate": 6.196578321437789e-07, "loss": 0.0006693030707538127, "reward": 1.6503021717071533, "reward_std": 0.18720974524815878, "rewards/FidelityReward/mean": 0.7664670348167419, "rewards/FidelityReward/std": 0.2139664888381958, "rewards/JudgeFidelityReward/mean": 0.7676702936490377, "rewards/JudgeFidelityReward/std": 0.21211556593577066, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3055 }, { "clip_ratio/high_max": 0.0007215661229565739, "clip_ratio/high_mean": 0.0002025162655627355, "clip_ratio/low_mean": 0.00010311218065908178, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030562843894585965, "completions/clipped_ratio": 0.0, "completions/max_length": 170.0, "completions/mean_length": 71.5234375, "completions/min_length": 29.5, "epoch": 2.2633136094674557, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.017495618015527726, "learning_rate": 6.184700323481185e-07, "loss": 0.0007796652149409056, "reward": 1.728782057762146, "reward_std": 0.1559857726097107, "rewards/FidelityReward/mean": 0.820906937122345, "rewards/FidelityReward/std": 0.17841287702322006, "rewards/JudgeFidelityReward/mean": 0.8157501518726349, "rewards/JudgeFidelityReward/std": 0.17953745275735855, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3060 }, { "clip_ratio/high_max": 0.00038749154191464186, "clip_ratio/high_mean": 9.142873750533909e-05, "clip_ratio/low_mean": 4.5483933354262264e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001369126664940268, "completions/clipped_ratio": 0.0, "completions/max_length": 172.0, "completions/mean_length": 73.46875, "completions/min_length": 36.0, "epoch": 2.2670118343195265, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.453125, "kl": 0.01681490186601877, "learning_rate": 6.172815237790061e-07, "loss": 0.0008188931271433831, "reward": 1.7211979230244954, "reward_std": 0.18035812179247537, "rewards/FidelityReward/mean": 0.8138466080029806, "rewards/FidelityReward/std": 0.17933359742164612, "rewards/JudgeFidelityReward/mean": 0.8147024909655253, "rewards/JudgeFidelityReward/std": 0.17819908758004507, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3065 }, { "clip_ratio/high_max": 0.0008209428749978542, "clip_ratio/high_mean": 0.0001966025971341878, "clip_ratio/low_mean": 8.163231250364333e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002782349183689803, "completions/clipped_ratio": 0.0, "completions/max_length": 176.0, "completions/mean_length": 71.361328125, "completions/min_length": 38.0, "epoch": 2.2707100591715976, "frac_reward_zero_std": 0.015625, "grad_norm": 2.421875, "kl": 0.016592492163181306, "learning_rate": 6.160923135469599e-07, "loss": 0.0007447505369782448, "reward": 1.59993577003479, "reward_std": 0.18622389435768127, "rewards/FidelityReward/mean": 0.7314429879188538, "rewards/FidelityReward/std": 0.2253762036561966, "rewards/JudgeFidelityReward/mean": 0.7369855344295502, "rewards/JudgeFidelityReward/std": 0.2113504856824875, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3070 }, { "clip_ratio/high_max": 0.0008316954132169485, "clip_ratio/high_mean": 0.0003546825610101223, "clip_ratio/low_mean": 4.7734813415445386e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004024173831567168, "completions/clipped_ratio": 0.0, "completions/max_length": 166.66666666666666, "completions/mean_length": 69.88932291666667, "completions/min_length": 31.333333333333332, "epoch": 2.274408284023669, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.828125, "kl": 0.01640423685312271, "learning_rate": 6.149024087666961e-07, "loss": 0.000720937829464674, "reward": 1.6669928630193074, "reward_std": 0.17410898705323538, "rewards/FidelityReward/mean": 0.7780430912971497, "rewards/FidelityReward/std": 0.20974442859490713, "rewards/JudgeFidelityReward/mean": 0.7792015870412191, "rewards/JudgeFidelityReward/std": 0.18727079530556998, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3075 }, { "clip_ratio/high_max": 0.0009662928758189082, "clip_ratio/high_mean": 0.00027296459302306175, "clip_ratio/low_mean": 9.858726698439569e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003715518629178405, "completions/clipped_ratio": 0.0, "completions/max_length": 171.5, "completions/mean_length": 72.72265625, "completions/min_length": 34.5, "epoch": 2.2781065088757395, "frac_reward_zero_std": 0.03125, "grad_norm": 2.234375, "kl": 0.015654263086616993, "learning_rate": 6.137118165570858e-07, "loss": 0.0006628538947552442, "reward": 1.533098816871643, "reward_std": 0.19792677462100983, "rewards/FidelityReward/mean": 0.6875375211238861, "rewards/FidelityReward/std": 0.261273130774498, "rewards/JudgeFidelityReward/mean": 0.6930757462978363, "rewards/JudgeFidelityReward/std": 0.261916920542717, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3080 }, { "clip_ratio/high_max": 0.0005261336918920279, "clip_ratio/high_mean": 0.0001507389359176159, "clip_ratio/low_mean": 2.5050100521184505e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017578903934918345, "completions/clipped_ratio": 0.0, "completions/max_length": 218.33333333333334, "completions/mean_length": 73.8359375, "completions/min_length": 33.666666666666664, "epoch": 2.2818047337278107, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.546875, "kl": 0.015283148549497128, "learning_rate": 6.125205440411133e-07, "loss": 0.0005427209660410881, "reward": 1.6337473392486572, "reward_std": 0.18529525895913443, "rewards/FidelityReward/mean": 0.7559216618537903, "rewards/FidelityReward/std": 0.20217428108056387, "rewards/JudgeFidelityReward/mean": 0.7569535175959269, "rewards/JudgeFidelityReward/std": 0.20065459609031677, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3085 }, { "clip_ratio/high_max": 0.0009312004782259464, "clip_ratio/high_mean": 0.00020472661126405, "clip_ratio/low_mean": 9.116356377489865e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002958901808597147, "completions/clipped_ratio": 0.0, "completions/max_length": 159.0, "completions/mean_length": 71.033203125, "completions/min_length": 26.0, "epoch": 2.285502958579882, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.01601789891719818, "learning_rate": 6.113285983458325e-07, "loss": 0.0005653297994285822, "reward": 1.6560620665550232, "reward_std": 0.20090916007757187, "rewards/FidelityReward/mean": 0.7741994261741638, "rewards/FidelityReward/std": 0.2062373235821724, "rewards/JudgeFidelityReward/mean": 0.7695847749710083, "rewards/JudgeFidelityReward/std": 0.20346909761428833, "rewards/SelfEvolvingFormatReward/mean": 0.994140625, "rewards/SelfEvolvingFormatReward/std": 0.07535743340849876, "step": 3090 }, { "clip_ratio/high_max": 0.0002915452001616359, "clip_ratio/high_mean": 0.0001272551016882062, "clip_ratio/low_mean": 3.996070590801537e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016721580177545548, "completions/clipped_ratio": 0.0, "completions/max_length": 216.0, "completions/mean_length": 69.96223958333333, "completions/min_length": 30.333333333333332, "epoch": 2.2892011834319526, "frac_reward_zero_std": 0.03125, "grad_norm": 2.859375, "kl": 0.017746978253126145, "learning_rate": 6.101359866023251e-07, "loss": 0.0005655641201883554, "reward": 1.5951590140660603, "reward_std": 0.2024669200181961, "rewards/FidelityReward/mean": 0.7280905842781067, "rewards/FidelityReward/std": 0.22410444418589273, "rewards/JudgeFidelityReward/mean": 0.7341368198394775, "rewards/JudgeFidelityReward/std": 0.21283097565174103, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3095 }, { "clip_ratio/high_max": 0.0007195539772510528, "clip_ratio/high_mean": 0.00019327570917084813, "clip_ratio/low_mean": 0.00016434421995654703, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000357619917485863, "completions/clipped_ratio": 0.0, "completions/max_length": 187.5, "completions/mean_length": 74.375, "completions/min_length": 32.5, "epoch": 2.2928994082840237, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.01616774871945381, "learning_rate": 6.089427159456575e-07, "loss": 0.000474140141159296, "reward": 1.6943303346633911, "reward_std": 0.19056086242198944, "rewards/FidelityReward/mean": 0.7970316410064697, "rewards/FidelityReward/std": 0.19892770051956177, "rewards/JudgeFidelityReward/mean": 0.7945974171161652, "rewards/JudgeFidelityReward/std": 0.19622332602739334, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3100 }, { "clip_ratio/high_max": 0.0004892565892077982, "clip_ratio/high_mean": 0.00014162304578348994, "clip_ratio/low_mean": 9.6548703731969e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023817176115699113, "completions/clipped_ratio": 0.0, "completions/max_length": 165.0, "completions/mean_length": 70.80078125, "completions/min_length": 30.333333333333332, "epoch": 2.296597633136095, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.609375, "kl": 0.01550973616540432, "learning_rate": 6.07748793514838e-07, "loss": 0.00072512268088758, "reward": 1.636520266532898, "reward_std": 0.1902501384417216, "rewards/FidelityReward/mean": 0.7562442223230997, "rewards/FidelityReward/std": 0.2075174500544866, "rewards/JudgeFidelityReward/mean": 0.760551909605662, "rewards/JudgeFidelityReward/std": 0.18666991591453552, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3105 }, { "clip_ratio/high_max": 0.0008184705977328122, "clip_ratio/high_mean": 0.0002581534499768168, "clip_ratio/low_mean": 4.1019191849045454e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00029917263891547917, "completions/clipped_ratio": 0.0, "completions/max_length": 180.0, "completions/mean_length": 72.6015625, "completions/min_length": 33.0, "epoch": 2.3002958579881656, "frac_reward_zero_std": 0.015625, "grad_norm": 2.28125, "kl": 0.015304107032716275, "learning_rate": 6.065542264527748e-07, "loss": 0.0007409645244479179, "reward": 1.688792884349823, "reward_std": 0.16711484640836716, "rewards/FidelityReward/mean": 0.7897066175937653, "rewards/FidelityReward/std": 0.2092064693570137, "rewards/JudgeFidelityReward/mean": 0.8001257181167603, "rewards/JudgeFidelityReward/std": 0.19036683440208435, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3110 }, { "clip_ratio/high_max": 0.0007992600090801716, "clip_ratio/high_mean": 0.0002547291864175349, "clip_ratio/low_mean": 2.5100400671362876e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027982958708889784, "completions/clipped_ratio": 0.0, "completions/max_length": 177.0, "completions/mean_length": 71.48177083333333, "completions/min_length": 36.0, "epoch": 2.3039940828402368, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.53125, "kl": 0.013937046565115453, "learning_rate": 6.05359021906232e-07, "loss": 0.0004986604675650596, "reward": 1.6373348633448284, "reward_std": 0.20331874986489615, "rewards/FidelityReward/mean": 0.7567972938219706, "rewards/FidelityReward/std": 0.20188893377780914, "rewards/JudgeFidelityReward/mean": 0.7623772223790487, "rewards/JudgeFidelityReward/std": 0.1926222344239553, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3115 }, { "clip_ratio/high_max": 0.0008494043489918113, "clip_ratio/high_mean": 0.0002359880192670971, "clip_ratio/low_mean": 4.0979901677928864e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027696792676579207, "completions/clipped_ratio": 0.0, "completions/max_length": 180.5, "completions/mean_length": 74.224609375, "completions/min_length": 33.5, "epoch": 2.3076923076923075, "frac_reward_zero_std": 0.0, "grad_norm": 2.5625, "kl": 0.015131872892379761, "learning_rate": 6.041631870257882e-07, "loss": 0.0006803212221711874, "reward": 1.6179176568984985, "reward_std": 0.18618331104516983, "rewards/FidelityReward/mean": 0.7468696236610413, "rewards/FidelityReward/std": 0.22531332075595856, "rewards/JudgeFidelityReward/mean": 0.7440492510795593, "rewards/JudgeFidelityReward/std": 0.21956667304039001, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3120 }, { "clip_ratio/high_max": 0.0009299708530306816, "clip_ratio/high_mean": 0.00023803887888789176, "clip_ratio/low_mean": 4.539765213849023e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028343653539195655, "completions/clipped_ratio": 0.0, "completions/max_length": 173.33333333333334, "completions/mean_length": 71.63411458333333, "completions/min_length": 30.333333333333332, "epoch": 2.3113905325443787, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.796875, "kl": 0.01580953449010849, "learning_rate": 6.029667289657926e-07, "loss": 0.0004527851939201355, "reward": 1.6094587643941243, "reward_std": 0.19580619037151337, "rewards/FidelityReward/mean": 0.7413928310076395, "rewards/FidelityReward/std": 0.22736149032910666, "rewards/JudgeFidelityReward/mean": 0.7374339501063029, "rewards/JudgeFidelityReward/std": 0.21878408392270407, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3125 }, { "clip_ratio/high_max": 0.0007413078565150499, "clip_ratio/high_mean": 0.00020357780740596355, "clip_ratio/low_mean": 6.73429574817419e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027092076488770543, "completions/clipped_ratio": 0.0, "completions/max_length": 151.0, "completions/mean_length": 73.212890625, "completions/min_length": 34.5, "epoch": 2.31508875739645, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.015726820193231107, "learning_rate": 6.017696548843232e-07, "loss": 0.0006734684575349092, "reward": 1.591559886932373, "reward_std": 0.2084207385778427, "rewards/FidelityReward/mean": 0.7246840000152588, "rewards/FidelityReward/std": 0.23071042448282242, "rewards/JudgeFidelityReward/mean": 0.733751654624939, "rewards/JudgeFidelityReward/std": 0.22499236464500427, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3130 }, { "clip_ratio/high_max": 0.0005068009370006621, "clip_ratio/high_mean": 8.288136596092954e-05, "clip_ratio/low_mean": 8.557682740502059e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001684581977315247, "completions/clipped_ratio": 0.0, "completions/max_length": 177.0, "completions/mean_length": 71.5703125, "completions/min_length": 34.0, "epoch": 2.3187869822485205, "frac_reward_zero_std": 0.0, "grad_norm": 2.78125, "kl": 0.0156985305249691, "learning_rate": 6.005719719431434e-07, "loss": 0.0004807163029909134, "reward": 1.6457573572794597, "reward_std": 0.19130191703637442, "rewards/FidelityReward/mean": 0.7621753017107645, "rewards/FidelityReward/std": 0.20566494266192117, "rewards/JudgeFidelityReward/mean": 0.7671640714009603, "rewards/JudgeFidelityReward/std": 0.20852301518122354, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3135 }, { "clip_ratio/high_max": 0.0008731190930120647, "clip_ratio/high_mean": 0.0003309615480247885, "clip_ratio/low_mean": 8.454180497210472e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004155033500865102, "completions/clipped_ratio": 0.001953125, "completions/max_length": 285.5, "completions/mean_length": 73.828125, "completions/min_length": 36.0, "epoch": 2.3224852071005917, "frac_reward_zero_std": 0.015625, "grad_norm": 2.59375, "kl": 0.014549453556537629, "learning_rate": 5.993736873076591e-07, "loss": 0.0007228408008813858, "reward": 1.6100783944129944, "reward_std": 0.2171815186738968, "rewards/FidelityReward/mean": 0.7391617298126221, "rewards/FidelityReward/std": 0.20874106884002686, "rewards/JudgeFidelityReward/mean": 0.7437864243984222, "rewards/JudgeFidelityReward/std": 0.20606853067874908, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3140 }, { "clip_ratio/high_max": 0.0006762617733329535, "clip_ratio/high_mean": 0.0002140602155122906, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002140602155122906, "completions/clipped_ratio": 0.0, "completions/max_length": 173.33333333333334, "completions/mean_length": 70.59375, "completions/min_length": 27.666666666666668, "epoch": 2.326183431952663, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.014902344159781932, "learning_rate": 5.98174808146876e-07, "loss": 0.0007820257917046547, "reward": 1.6241787672042847, "reward_std": 0.198678657412529, "rewards/FidelityReward/mean": 0.7509175141652426, "rewards/FidelityReward/std": 0.20169134438037872, "rewards/JudgeFidelityReward/mean": 0.7465226054191589, "rewards/JudgeFidelityReward/std": 0.2095307558774948, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3145 }, { "clip_ratio/high_max": 0.0005943000433035195, "clip_ratio/high_mean": 9.3284466129262e-05, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 9.3284466129262e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 179.5, "completions/mean_length": 72.37890625, "completions/min_length": 35.0, "epoch": 2.3298816568047336, "frac_reward_zero_std": 0.015625, "grad_norm": 2.796875, "kl": 0.015254879556596278, "learning_rate": 5.969753416333564e-07, "loss": 0.0006781444884836674, "reward": 1.5932354927062988, "reward_std": 0.1768980398774147, "rewards/FidelityReward/mean": 0.7289987504482269, "rewards/FidelityReward/std": 0.24368516355752945, "rewards/JudgeFidelityReward/mean": 0.7284735143184662, "rewards/JudgeFidelityReward/std": 0.2562631219625473, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3150 }, { "clip_ratio/high_max": 0.0005219730315729976, "clip_ratio/high_mean": 0.00014735606382600963, "clip_ratio/low_mean": 2.3320894979406148e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017067696317099034, "completions/clipped_ratio": 0.0, "completions/max_length": 186.33333333333334, "completions/mean_length": 72.07421875, "completions/min_length": 32.333333333333336, "epoch": 2.3335798816568047, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.01628852915018797, "learning_rate": 5.957752949431771e-07, "loss": 0.0008626462891697884, "reward": 1.664536436398824, "reward_std": 0.17576376597086588, "rewards/FidelityReward/mean": 0.7743182977040609, "rewards/FidelityReward/std": 0.21038647989432016, "rewards/JudgeFidelityReward/mean": 0.781738261381785, "rewards/JudgeFidelityReward/std": 0.19157964984575906, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3155 }, { "clip_ratio/high_max": 0.0003251987625844777, "clip_ratio/high_mean": 5.8674359752330926e-05, "clip_ratio/low_mean": 9.110978135140613e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014978414255892858, "completions/clipped_ratio": 0.0, "completions/max_length": 164.0, "completions/mean_length": 70.669921875, "completions/min_length": 28.5, "epoch": 2.337278106508876, "frac_reward_zero_std": 0.015625, "grad_norm": 2.453125, "kl": 0.015254882164299488, "learning_rate": 5.945746752558858e-07, "loss": 0.0005797652527689933, "reward": 1.635690152645111, "reward_std": 0.1641208380460739, "rewards/FidelityReward/mean": 0.7573813199996948, "rewards/FidelityReward/std": 0.223396897315979, "rewards/JudgeFidelityReward/mean": 0.7566177845001221, "rewards/JudgeFidelityReward/std": 0.21797038614749908, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3160 }, { "clip_ratio/high_max": 0.0004808779922313988, "clip_ratio/high_mean": 0.00016327548655681312, "clip_ratio/low_mean": 2.7839644462801517e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019111513392999768, "completions/clipped_ratio": 0.0, "completions/max_length": 193.0, "completions/mean_length": 70.6640625, "completions/min_length": 33.666666666666664, "epoch": 2.3409763313609466, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.640625, "kl": 0.016487449407577515, "learning_rate": 5.93373489754458e-07, "loss": 0.0006262900307774544, "reward": 1.6189885139465332, "reward_std": 0.18696486949920654, "rewards/FidelityReward/mean": 0.7428650856018066, "rewards/FidelityReward/std": 0.23003113766511282, "rewards/JudgeFidelityReward/mean": 0.7522467970848083, "rewards/JudgeFidelityReward/std": 0.20606106519699097, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3165 }, { "clip_ratio/high_max": 0.001277615432627499, "clip_ratio/high_mean": 0.00041546855354681613, "clip_ratio/low_mean": 4.9321656115353106e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004647902096621692, "completions/clipped_ratio": 0.0, "completions/max_length": 166.5, "completions/mean_length": 70.1875, "completions/min_length": 31.5, "epoch": 2.344674556213018, "frac_reward_zero_std": 0.015625, "grad_norm": 2.84375, "kl": 0.015288129076361656, "learning_rate": 5.921717456252545e-07, "loss": 0.0005507825408130884, "reward": 1.6187261939048767, "reward_std": 0.20246659964323044, "rewards/FidelityReward/mean": 0.7451852560043335, "rewards/FidelityReward/std": 0.20985721051692963, "rewards/JudgeFidelityReward/mean": 0.7490350008010864, "rewards/JudgeFidelityReward/std": 0.2013036161661148, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3170 }, { "clip_ratio/high_max": 0.00038874490419402716, "clip_ratio/high_mean": 0.00015353556955233216, "clip_ratio/low_mean": 8.760206983424724e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002411376452073455, "completions/clipped_ratio": 0.0, "completions/max_length": 197.0, "completions/mean_length": 72.05338541666667, "completions/min_length": 27.666666666666668, "epoch": 2.3483727810650885, "frac_reward_zero_std": 0.0, "grad_norm": 3.109375, "kl": 0.016242888942360877, "learning_rate": 5.909694500579782e-07, "loss": 0.0007034806534647941, "reward": 1.606388807296753, "reward_std": 0.1967869997024536, "rewards/FidelityReward/mean": 0.7371189991633097, "rewards/FidelityReward/std": 0.212042306860288, "rewards/JudgeFidelityReward/mean": 0.7398416797320048, "rewards/JudgeFidelityReward/std": 0.20454042156537375, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3175 }, { "clip_ratio/high_max": 0.001313867070712149, "clip_ratio/high_mean": 0.00031646881252527236, "clip_ratio/low_mean": 6.151468260213732e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003779835067689419, "completions/clipped_ratio": 0.0, "completions/max_length": 184.0, "completions/mean_length": 72.30859375, "completions/min_length": 33.5, "epoch": 2.3520710059171597, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.017162786051630973, "learning_rate": 5.897666102456312e-07, "loss": 0.0007959067821502686, "reward": 1.6332340240478516, "reward_std": 0.19218751043081284, "rewards/FidelityReward/mean": 0.7560975551605225, "rewards/FidelityReward/std": 0.2317156419157982, "rewards/JudgeFidelityReward/mean": 0.7542730271816254, "rewards/JudgeFidelityReward/std": 0.23819594085216522, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3180 }, { "clip_ratio/high_max": 0.0005093575688079, "clip_ratio/high_mean": 0.0002039908547885716, "clip_ratio/low_mean": 4.419292672537267e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024818377569317817, "completions/clipped_ratio": 0.0, "completions/max_length": 159.33333333333334, "completions/mean_length": 70.45833333333333, "completions/min_length": 33.666666666666664, "epoch": 2.355769230769231, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.703125, "kl": 0.014005519449710846, "learning_rate": 5.885632333844713e-07, "loss": 0.000562623143196106, "reward": 1.6286314725875854, "reward_std": 0.19037711123625436, "rewards/FidelityReward/mean": 0.7522678176561991, "rewards/FidelityReward/std": 0.1924590269724528, "rewards/JudgeFidelityReward/mean": 0.7527272701263428, "rewards/JudgeFidelityReward/std": 0.1774917741616567, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3185 }, { "clip_ratio/high_max": 0.0007179127540439367, "clip_ratio/high_mean": 0.00023543419665656985, "clip_ratio/low_mean": 0.00015218249754980206, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003876166942063719, "completions/clipped_ratio": 0.0, "completions/max_length": 166.5, "completions/mean_length": 71.599609375, "completions/min_length": 32.5, "epoch": 2.359467455621302, "frac_reward_zero_std": 0.0625, "grad_norm": 2.90625, "kl": 0.016578542813658715, "learning_rate": 5.873593266739701e-07, "loss": 0.0007552720606327056, "reward": 1.770142138004303, "reward_std": 0.1439710631966591, "rewards/FidelityReward/mean": 0.8488254547119141, "rewards/FidelityReward/std": 0.18035601824522018, "rewards/JudgeFidelityReward/mean": 0.8426333069801331, "rewards/JudgeFidelityReward/std": 0.17729420214891434, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3190 }, { "clip_ratio/high_max": 0.0009007405024021864, "clip_ratio/high_mean": 0.0001320025068707764, "clip_ratio/low_mean": 8.872602775227279e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022072853753343224, "completions/clipped_ratio": 0.0, "completions/max_length": 156.33333333333334, "completions/mean_length": 69.90494791666667, "completions/min_length": 35.666666666666664, "epoch": 2.3631656804733727, "frac_reward_zero_std": 0.0, "grad_norm": 2.390625, "kl": 0.018009966239333153, "learning_rate": 5.86154897316768e-07, "loss": 0.0009514236822724343, "reward": 1.7451436916987102, "reward_std": 0.15107308824857077, "rewards/FidelityReward/mean": 0.8339344461758932, "rewards/FidelityReward/std": 0.1969152887662252, "rewards/JudgeFidelityReward/mean": 0.8237205346425375, "rewards/JudgeFidelityReward/std": 0.1988942821820577, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3195 }, { "clip_ratio/high_max": 0.0008464916725642979, "clip_ratio/high_mean": 0.0002813428465742618, "clip_ratio/low_mean": 0.00015085802879184486, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004322008811868727, "completions/clipped_ratio": 0.0, "completions/max_length": 189.5, "completions/mean_length": 69.54296875, "completions/min_length": 30.0, "epoch": 2.366863905325444, "frac_reward_zero_std": 0.046875, "grad_norm": 2.390625, "kl": 0.015783975645899773, "learning_rate": 5.849499525186332e-07, "loss": 0.0009804612025618553, "reward": 1.6133472919464111, "reward_std": 0.19688285142183304, "rewards/FidelityReward/mean": 0.7462157011032104, "rewards/FidelityReward/std": 0.21263174712657928, "rewards/JudgeFidelityReward/mean": 0.7342632412910461, "rewards/JudgeFidelityReward/std": 0.21296703815460205, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3200 }, { "clip_ratio/high_max": 0.0003890587715432048, "clip_ratio/high_mean": 0.0001543716643936932, "clip_ratio/low_mean": 6.325261492747813e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021762428805232047, "completions/clipped_ratio": 0.0, "completions/max_length": 160.66666666666666, "completions/mean_length": 71.18619791666667, "completions/min_length": 30.666666666666668, "epoch": 2.3705621301775146, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.765625, "kl": 0.01730135753750801, "learning_rate": 5.837444994884173e-07, "loss": 0.0007089908234775066, "reward": 1.610119104385376, "reward_std": 0.1941892554362615, "rewards/FidelityReward/mean": 0.7430042624473572, "rewards/FidelityReward/std": 0.2246287763118744, "rewards/JudgeFidelityReward/mean": 0.734229584534963, "rewards/JudgeFidelityReward/std": 0.22280539572238922, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3205 }, { "clip_ratio/high_max": 0.0009387926664203405, "clip_ratio/high_mean": 0.0002739199320785701, "clip_ratio/low_mean": 6.741171528119594e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003413316677324474, "completions/clipped_ratio": 0.0, "completions/max_length": 210.5, "completions/mean_length": 68.05859375, "completions/min_length": 32.5, "epoch": 2.3742603550295858, "frac_reward_zero_std": 0.0, "grad_norm": 2.625, "kl": 0.018501836434006692, "learning_rate": 5.825385454380125e-07, "loss": 0.0006141016259789467, "reward": 1.6203277707099915, "reward_std": 0.18816307187080383, "rewards/FidelityReward/mean": 0.746311604976654, "rewards/FidelityReward/std": 0.19880491495132446, "rewards/JudgeFidelityReward/mean": 0.7480323314666748, "rewards/JudgeFidelityReward/std": 0.1749318540096283, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3210 }, { "clip_ratio/high_max": 0.0004453580128028989, "clip_ratio/high_mean": 0.00011922464473173023, "clip_ratio/low_mean": 2.410800370853394e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014333264552988113, "completions/clipped_ratio": 0.0, "completions/max_length": 163.66666666666666, "completions/mean_length": 72.3515625, "completions/min_length": 32.0, "epoch": 2.377958579881657, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.90625, "kl": 0.016043216921389103, "learning_rate": 5.813320975823085e-07, "loss": 0.0006881861481815577, "reward": 1.673962910970052, "reward_std": 0.19526560604572296, "rewards/FidelityReward/mean": 0.7859033544858297, "rewards/FidelityReward/std": 0.2080789009730021, "rewards/JudgeFidelityReward/mean": 0.777421236038208, "rewards/JudgeFidelityReward/std": 0.20541770259539285, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3215 }, { "clip_ratio/high_max": 0.0009577486664056778, "clip_ratio/high_mean": 0.00029574164655059575, "clip_ratio/low_mean": 0.00011620713048614562, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004119487712159753, "completions/clipped_ratio": 0.0, "completions/max_length": 178.5, "completions/mean_length": 69.98046875, "completions/min_length": 25.0, "epoch": 2.3816568047337277, "frac_reward_zero_std": 0.015625, "grad_norm": 2.953125, "kl": 0.016518957167863845, "learning_rate": 5.801251631391493e-07, "loss": 0.0006781628355383873, "reward": 1.6045547723770142, "reward_std": 0.19089408218860626, "rewards/FidelityReward/mean": 0.7381872534751892, "rewards/FidelityReward/std": 0.22200916707515717, "rewards/JudgeFidelityReward/mean": 0.7327349781990051, "rewards/JudgeFidelityReward/std": 0.22247204184532166, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3220 }, { "clip_ratio/high_max": 0.0003514938522130251, "clip_ratio/high_mean": 0.00011807852424681187, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00011807852424681187, "completions/clipped_ratio": 0.0, "completions/max_length": 151.0, "completions/mean_length": 69.82682291666667, "completions/min_length": 34.666666666666664, "epoch": 2.385355029585799, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.515625, "kl": 0.01616690941154957, "learning_rate": 5.789177493292901e-07, "loss": 0.0006076341494917869, "reward": 1.6538703044255574, "reward_std": 0.1907129486401876, "rewards/FidelityReward/mean": 0.7708281675974528, "rewards/FidelityReward/std": 0.18925086657206217, "rewards/JudgeFidelityReward/mean": 0.7673862179120382, "rewards/JudgeFidelityReward/std": 0.1711583137512207, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3225 }, { "clip_ratio/high_max": 0.0007471704389899969, "clip_ratio/high_mean": 0.0001763799082254991, "clip_ratio/low_mean": 2.5150904548354447e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00020153081277385354, "completions/clipped_ratio": 0.0, "completions/max_length": 169.5, "completions/mean_length": 70.421875, "completions/min_length": 35.0, "epoch": 2.38905325443787, "frac_reward_zero_std": 0.046875, "grad_norm": 2.296875, "kl": 0.014788429625332356, "learning_rate": 5.777098633763538e-07, "loss": 0.00047514489851891996, "reward": 1.6793422102928162, "reward_std": 0.17092270404100418, "rewards/FidelityReward/mean": 0.782987505197525, "rewards/FidelityReward/std": 0.2055564597249031, "rewards/JudgeFidelityReward/mean": 0.7927094101905823, "rewards/JudgeFidelityReward/std": 0.18754324316978455, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3230 }, { "clip_ratio/high_max": 0.0004582379129715264, "clip_ratio/high_mean": 0.0001042154777678661, "clip_ratio/low_mean": 4.5403870171867314e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014961935812607408, "completions/clipped_ratio": 0.0, "completions/max_length": 169.66666666666666, "completions/mean_length": 71.28515625, "completions/min_length": 32.0, "epoch": 2.3927514792899407, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.375, "kl": 0.015029205568134784, "learning_rate": 5.765015125067882e-07, "loss": 0.0005898024886846542, "reward": 1.658966024716695, "reward_std": 0.18026089668273926, "rewards/FidelityReward/mean": 0.7740930914878845, "rewards/FidelityReward/std": 0.20836527148882547, "rewards/JudgeFidelityReward/mean": 0.7723500927289327, "rewards/JudgeFidelityReward/std": 0.1975166549285253, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.041666666666666664, "step": 3235 }, { "clip_ratio/high_max": 0.0008709375280886889, "clip_ratio/high_mean": 0.00017596154357306658, "clip_ratio/low_mean": 4.594048659782857e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002219020330812782, "completions/clipped_ratio": 0.0, "completions/max_length": 166.5, "completions/mean_length": 70.181640625, "completions/min_length": 33.0, "epoch": 2.396449704142012, "frac_reward_zero_std": 0.0, "grad_norm": 2.421875, "kl": 0.015497733280062676, "learning_rate": 5.752927039498221e-07, "loss": 0.0006941505242139101, "reward": 1.5853546857833862, "reward_std": 0.1861581802368164, "rewards/FidelityReward/mean": 0.7266528308391571, "rewards/FidelityReward/std": 0.24928750842809677, "rewards/JudgeFidelityReward/mean": 0.7174037098884583, "rewards/JudgeFidelityReward/std": 0.2485254555940628, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3240 }, { "clip_ratio/high_max": 0.0007946363650262356, "clip_ratio/high_mean": 0.00024880580604076387, "clip_ratio/low_mean": 6.860905268695206e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00031741486163809894, "completions/clipped_ratio": 0.0, "completions/max_length": 161.0, "completions/mean_length": 70.515625, "completions/min_length": 33.333333333333336, "epoch": 2.400147928994083, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.796875, "kl": 0.017558260820806026, "learning_rate": 5.740834449374237e-07, "loss": 0.0007671510335057974, "reward": 1.632815957069397, "reward_std": 0.19813146690527597, "rewards/FidelityReward/mean": 0.7586092551549276, "rewards/FidelityReward/std": 0.210110733906428, "rewards/JudgeFidelityReward/mean": 0.7484133243560791, "rewards/JudgeFidelityReward/std": 0.20880661408106485, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3245 }, { "clip_ratio/high_max": 0.0007253920193761587, "clip_ratio/high_mean": 0.00015699435025453567, "clip_ratio/low_mean": 6.735378847224637e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002243481343612075, "completions/clipped_ratio": 0.0, "completions/max_length": 173.0, "completions/mean_length": 71.123046875, "completions/min_length": 36.0, "epoch": 2.4038461538461537, "frac_reward_zero_std": 0.015625, "grad_norm": 2.390625, "kl": 0.018080168031156064, "learning_rate": 5.728737427042547e-07, "loss": 0.0006135048810392618, "reward": 1.669360101222992, "reward_std": 0.16864079982042313, "rewards/FidelityReward/mean": 0.7802356481552124, "rewards/FidelityReward/std": 0.20309525728225708, "rewards/JudgeFidelityReward/mean": 0.778249055147171, "rewards/JudgeFidelityReward/std": 0.20095589011907578, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3250 }, { "clip_ratio/high_max": 0.0007572176866233349, "clip_ratio/high_mean": 0.0001608325052075088, "clip_ratio/low_mean": 9.734194900374859e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00025817444548010827, "completions/clipped_ratio": 0.0, "completions/max_length": 186.0, "completions/mean_length": 71.71614583333333, "completions/min_length": 27.333333333333332, "epoch": 2.407544378698225, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.53125, "kl": 0.015277716144919396, "learning_rate": 5.716636044876295e-07, "loss": 0.0006552533246576786, "reward": 1.5796517928441365, "reward_std": 0.182042196393013, "rewards/FidelityReward/mean": 0.7225969036420187, "rewards/FidelityReward/std": 0.2592814067999522, "rewards/JudgeFidelityReward/mean": 0.7154118220011393, "rewards/JudgeFidelityReward/std": 0.2565716008345286, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3255 }, { "clip_ratio/high_max": 0.0004928231937810779, "clip_ratio/high_mean": 8.207792125176638e-05, "clip_ratio/low_mean": 8.494817884638906e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016702610009815543, "completions/clipped_ratio": 0.0, "completions/max_length": 166.0, "completions/mean_length": 70.279296875, "completions/min_length": 34.0, "epoch": 2.4112426035502956, "frac_reward_zero_std": 0.0625, "grad_norm": 2.234375, "kl": 0.013452709093689919, "learning_rate": 5.704530375274703e-07, "loss": 0.0004072252660989761, "reward": 1.6055888533592224, "reward_std": 0.18026337027549744, "rewards/FidelityReward/mean": 0.7332054972648621, "rewards/FidelityReward/std": 0.23720362037420273, "rewards/JudgeFidelityReward/mean": 0.7447668313980103, "rewards/JudgeFidelityReward/std": 0.23063741624355316, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3260 }, { "clip_ratio/high_max": 0.0007737270090728999, "clip_ratio/high_mean": 0.00019766028854064643, "clip_ratio/low_mean": 2.019386156462133e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021785415010526775, "completions/clipped_ratio": 0.0, "completions/max_length": 208.66666666666666, "completions/mean_length": 70.64453125, "completions/min_length": 30.666666666666668, "epoch": 2.414940828402367, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.6875, "kl": 0.015688271448016166, "learning_rate": 5.692420490662646e-07, "loss": 0.0009023280814290047, "reward": 1.5781264702479045, "reward_std": 0.1886684149503708, "rewards/FidelityReward/mean": 0.7189285556475321, "rewards/FidelityReward/std": 0.22806593775749207, "rewards/JudgeFidelityReward/mean": 0.71839572985967, "rewards/JudgeFidelityReward/std": 0.22596129775047302, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3265 }, { "clip_ratio/high_max": 0.0005247062304988504, "clip_ratio/high_mean": 8.48635274451226e-05, "clip_ratio/low_mean": 9.487473871558905e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017973826616071164, "completions/clipped_ratio": 0.0, "completions/max_length": 180.0, "completions/mean_length": 71.126953125, "completions/min_length": 34.0, "epoch": 2.418639053254438, "frac_reward_zero_std": 0.03125, "grad_norm": 2.515625, "kl": 0.01594717539846897, "learning_rate": 5.680306463490216e-07, "loss": 0.0007488934323191642, "reward": 1.6135945320129395, "reward_std": 0.188594788312912, "rewards/FidelityReward/mean": 0.7430872917175293, "rewards/FidelityReward/std": 0.2122136726975441, "rewards/JudgeFidelityReward/mean": 0.7410145699977875, "rewards/JudgeFidelityReward/std": 0.20401845127344131, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3270 }, { "clip_ratio/high_max": 0.0012170757632702588, "clip_ratio/high_mean": 0.00025424417108297346, "clip_ratio/low_mean": 7.056981703499333e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003248139750212431, "completions/clipped_ratio": 0.0, "completions/max_length": 153.33333333333334, "completions/mean_length": 70.11067708333333, "completions/min_length": 32.333333333333336, "epoch": 2.422337278106509, "frac_reward_zero_std": 0.03125, "grad_norm": 2.53125, "kl": 0.01786975413560867, "learning_rate": 5.66818836623229e-07, "loss": 0.0007513738702982665, "reward": 1.6572177012761433, "reward_std": 0.1815350204706192, "rewards/FidelityReward/mean": 0.7720768650372823, "rewards/FidelityReward/std": 0.21537858744462332, "rewards/JudgeFidelityReward/mean": 0.7728858391443888, "rewards/JudgeFidelityReward/std": 0.20763656497001648, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 3275 }, { "clip_ratio/high_max": 0.0013428785605356097, "clip_ratio/high_mean": 0.0004095862153917551, "clip_ratio/low_mean": 0.00016893662832444533, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.000578522845171392, "completions/clipped_ratio": 0.0, "completions/max_length": 187.5, "completions/mean_length": 72.9921875, "completions/min_length": 36.0, "epoch": 2.42603550295858, "frac_reward_zero_std": 0.0, "grad_norm": 2.375, "kl": 0.015309970080852508, "learning_rate": 5.656066271388093e-07, "loss": 0.0007623821496963501, "reward": 1.6525853872299194, "reward_std": 0.19836583733558655, "rewards/FidelityReward/mean": 0.7634637355804443, "rewards/FidelityReward/std": 0.1886478140950203, "rewards/JudgeFidelityReward/mean": 0.7782433032989502, "rewards/JudgeFidelityReward/std": 0.1651318371295929, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3280 }, { "clip_ratio/high_max": 0.0005055566434748471, "clip_ratio/high_mean": 0.00014609054196625947, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014609054196625947, "completions/clipped_ratio": 0.0, "completions/max_length": 167.0, "completions/mean_length": 70.83984375, "completions/min_length": 34.333333333333336, "epoch": 2.429733727810651, "frac_reward_zero_std": 0.0, "grad_norm": 2.546875, "kl": 0.017408527620136737, "learning_rate": 5.643940251480766e-07, "loss": 0.0006628815084695816, "reward": 1.6246485312779744, "reward_std": 0.18938307464122772, "rewards/FidelityReward/mean": 0.7475977142651876, "rewards/FidelityReward/std": 0.20889635384082794, "rewards/JudgeFidelityReward/mean": 0.7541015942891439, "rewards/JudgeFidelityReward/std": 0.19444376726945242, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3285 }, { "clip_ratio/high_max": 0.0009477478684857488, "clip_ratio/high_mean": 0.00025336905382573606, "clip_ratio/low_mean": 0.00023113515926524996, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004845042014494538, "completions/clipped_ratio": 0.0, "completions/max_length": 177.0, "completions/mean_length": 74.048828125, "completions/min_length": 37.5, "epoch": 2.4334319526627217, "frac_reward_zero_std": 0.015625, "grad_norm": 2.5625, "kl": 0.014592716470360756, "learning_rate": 5.631810379056934e-07, "loss": 0.00047564110718667507, "reward": 1.6668174266815186, "reward_std": 0.1835884377360344, "rewards/FidelityReward/mean": 0.7811003029346466, "rewards/FidelityReward/std": 0.19036149233579636, "rewards/JudgeFidelityReward/mean": 0.7733874022960663, "rewards/JudgeFidelityReward/std": 0.20374726504087448, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3290 }, { "clip_ratio/high_max": 0.0005642851581797003, "clip_ratio/high_mean": 0.00021244239760562778, "clip_ratio/low_mean": 2.60960339801386e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023853844031691552, "completions/clipped_ratio": 0.0, "completions/max_length": 153.33333333333334, "completions/mean_length": 71.38541666666667, "completions/min_length": 27.666666666666668, "epoch": 2.437130177514793, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.375, "kl": 0.01544684749096632, "learning_rate": 5.61967672668627e-07, "loss": 0.000793524831533432, "reward": 1.6268123388290405, "reward_std": 0.1867532581090927, "rewards/FidelityReward/mean": 0.7484958370526632, "rewards/FidelityReward/std": 0.21628202001253763, "rewards/JudgeFidelityReward/mean": 0.7566330631573995, "rewards/JudgeFidelityReward/std": 0.19943705201148987, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3295 }, { "clip_ratio/high_max": 0.0009204248897731304, "clip_ratio/high_mean": 0.0002460561168845743, "clip_ratio/low_mean": 4.674608935602009e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00029280220624059437, "completions/clipped_ratio": 0.0, "completions/max_length": 195.5, "completions/mean_length": 71.9296875, "completions/min_length": 33.0, "epoch": 2.440828402366864, "frac_reward_zero_std": 0.03125, "grad_norm": 2.5, "kl": 0.01692335233092308, "learning_rate": 5.607539366961064e-07, "loss": 0.0005862334743142128, "reward": 1.5974360704421997, "reward_std": 0.18063855916261673, "rewards/FidelityReward/mean": 0.732250452041626, "rewards/FidelityReward/std": 0.2356598675251007, "rewards/JudgeFidelityReward/mean": 0.7303711175918579, "rewards/JudgeFidelityReward/std": 0.22943026572465897, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3300 }, { "clip_ratio/high_max": 0.0005743679357692599, "clip_ratio/high_mean": 0.0001320856303209439, "clip_ratio/low_mean": 4.305200418457389e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001751376228639856, "completions/clipped_ratio": 0.0, "completions/max_length": 163.0, "completions/mean_length": 69.42578125, "completions/min_length": 31.0, "epoch": 2.4445266272189348, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.640625, "kl": 0.01593188978731632, "learning_rate": 5.595398372495783e-07, "loss": 0.0006504487246274948, "reward": 1.6870503425598145, "reward_std": 0.16829048097133636, "rewards/FidelityReward/mean": 0.7908970514933268, "rewards/FidelityReward/std": 0.2025634596745173, "rewards/JudgeFidelityReward/mean": 0.7923064629236857, "rewards/JudgeFidelityReward/std": 0.2054453194141388, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3305 }, { "clip_ratio/high_max": 0.0008892659796401858, "clip_ratio/high_mean": 0.0002170282881706953, "clip_ratio/low_mean": 7.345752674154937e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002904858090914786, "completions/clipped_ratio": 0.0, "completions/max_length": 226.0, "completions/mean_length": 69.81640625, "completions/min_length": 34.0, "epoch": 2.448224852071006, "frac_reward_zero_std": 0.046875, "grad_norm": 2.671875, "kl": 0.016503536514937876, "learning_rate": 5.583253815926638e-07, "loss": 0.0009615601971745491, "reward": 1.6261283159255981, "reward_std": 0.17695224285125732, "rewards/FidelityReward/mean": 0.7569964826107025, "rewards/FidelityReward/std": 0.20694522559642792, "rewards/JudgeFidelityReward/mean": 0.7382636964321136, "rewards/JudgeFidelityReward/std": 0.20665095746517181, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3310 }, { "clip_ratio/high_max": 0.0004901450360193848, "clip_ratio/high_mean": 7.888617110438645e-05, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 7.888617110438645e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 170.33333333333334, "completions/mean_length": 72.01692708333333, "completions/min_length": 35.0, "epoch": 2.451923076923077, "frac_reward_zero_std": 0.03125, "grad_norm": 2.53125, "kl": 0.014889641478657723, "learning_rate": 5.571105769911159e-07, "loss": 0.0006557732354849577, "reward": 1.6787629922231038, "reward_std": 0.16627160708109537, "rewards/FidelityReward/mean": 0.7888283729553223, "rewards/FidelityReward/std": 0.20360941191514334, "rewards/JudgeFidelityReward/mean": 0.7811713417371114, "rewards/JudgeFidelityReward/std": 0.18858021001021066, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3315 }, { "clip_ratio/high_max": 0.0006738103926181793, "clip_ratio/high_mean": 0.00026142538990825415, "clip_ratio/low_mean": 7.8435885370709e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003398612840101123, "completions/clipped_ratio": 0.0, "completions/max_length": 190.5, "completions/mean_length": 73.140625, "completions/min_length": 30.5, "epoch": 2.455621301775148, "frac_reward_zero_std": 0.0, "grad_norm": 2.484375, "kl": 0.016255656257271767, "learning_rate": 5.55895430712774e-07, "loss": 0.0007726194337010384, "reward": 1.714932382106781, "reward_std": 0.17353477329015732, "rewards/FidelityReward/mean": 0.8142895698547363, "rewards/FidelityReward/std": 0.19115734100341797, "rewards/JudgeFidelityReward/mean": 0.8051919937133789, "rewards/JudgeFidelityReward/std": 0.20287534594535828, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 3320 }, { "clip_ratio/high_max": 0.0008273726794868708, "clip_ratio/high_mean": 0.00014855199842713773, "clip_ratio/low_mean": 6.734033231623471e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021589233074337244, "completions/clipped_ratio": 0.0, "completions/max_length": 197.0, "completions/mean_length": 72.31770833333333, "completions/min_length": 34.666666666666664, "epoch": 2.459319526627219, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.546875, "kl": 0.014863630197942257, "learning_rate": 5.546799500275228e-07, "loss": 0.0006277993321418762, "reward": 1.6232519149780273, "reward_std": 0.19391502439975739, "rewards/FidelityReward/mean": 0.751215398311615, "rewards/FidelityReward/std": 0.20527575413386026, "rewards/JudgeFidelityReward/mean": 0.7440730532010397, "rewards/JudgeFidelityReward/std": 0.19573876758416495, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3325 }, { "clip_ratio/high_max": 0.0008968840353190899, "clip_ratio/high_mean": 0.00031020112219266595, "clip_ratio/low_mean": 1.794687705114484e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003281479992438108, "completions/clipped_ratio": 0.0, "completions/max_length": 164.5, "completions/mean_length": 70.087890625, "completions/min_length": 29.5, "epoch": 2.46301775147929, "frac_reward_zero_std": 0.03125, "grad_norm": 2.46875, "kl": 0.01512466948479414, "learning_rate": 5.534641422072471e-07, "loss": 0.0005451742559671402, "reward": 1.6765841245651245, "reward_std": 0.17068863660097122, "rewards/FidelityReward/mean": 0.7848345339298248, "rewards/FidelityReward/std": 0.21228647232055664, "rewards/JudgeFidelityReward/mean": 0.783499151468277, "rewards/JudgeFidelityReward/std": 0.20535534620285034, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3330 }, { "clip_ratio/high_max": 0.0006940861465409398, "clip_ratio/high_mean": 0.00016735770041123032, "clip_ratio/low_mean": 0.0001057634421158582, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002731211483478546, "completions/clipped_ratio": 0.0, "completions/max_length": 177.33333333333334, "completions/mean_length": 71.66276041666667, "completions/min_length": 33.333333333333336, "epoch": 2.466715976331361, "frac_reward_zero_std": 0.041666666666666664, "grad_norm": 2.703125, "kl": 0.01583327054977417, "learning_rate": 5.522480145257889e-07, "loss": 0.0006350692827254533, "reward": 1.6473253170649211, "reward_std": 0.17034514248371124, "rewards/FidelityReward/mean": 0.7658092379570007, "rewards/FidelityReward/std": 0.2358881483475367, "rewards/JudgeFidelityReward/mean": 0.7630321979522705, "rewards/JudgeFidelityReward/std": 0.23306296269098917, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3335 }, { "clip_ratio/high_max": 0.0007359762908890844, "clip_ratio/high_mean": 0.00021628584363497794, "clip_ratio/low_mean": 6.526105717057362e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00028154689935036006, "completions/clipped_ratio": 0.0, "completions/max_length": 191.0, "completions/mean_length": 72.154296875, "completions/min_length": 32.0, "epoch": 2.470414201183432, "frac_reward_zero_std": 0.0, "grad_norm": 2.703125, "kl": 0.015493042021989822, "learning_rate": 5.510315742589042e-07, "loss": 0.0005203824490308762, "reward": 1.6353796124458313, "reward_std": 0.18964102864265442, "rewards/FidelityReward/mean": 0.7577213048934937, "rewards/FidelityReward/std": 0.1861761063337326, "rewards/JudgeFidelityReward/mean": 0.7572696208953857, "rewards/JudgeFidelityReward/std": 0.18121317774057388, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3340 }, { "clip_ratio/high_max": 0.0005418873624876142, "clip_ratio/high_mean": 0.00015105019556358457, "clip_ratio/low_mean": 6.415728275896982e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021520748268812894, "completions/clipped_ratio": 0.0, "completions/max_length": 165.33333333333334, "completions/mean_length": 70.55078125, "completions/min_length": 27.666666666666668, "epoch": 2.4741124260355027, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.78125, "kl": 0.01602044440805912, "learning_rate": 5.498148286842184e-07, "loss": 0.0007263068575412035, "reward": 1.6692235867182414, "reward_std": 0.16076745092868805, "rewards/FidelityReward/mean": 0.7767659624417623, "rewards/FidelityReward/std": 0.22494000693162283, "rewards/JudgeFidelityReward/mean": 0.7849152684211731, "rewards/JudgeFidelityReward/std": 0.2053357015053431, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3345 }, { "clip_ratio/high_max": 0.0009439716115593911, "clip_ratio/high_mean": 0.00033857153030112384, "clip_ratio/low_mean": 0.00010007341043092311, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00043864494655281304, "completions/clipped_ratio": 0.0, "completions/max_length": 163.5, "completions/mean_length": 73.609375, "completions/min_length": 32.5, "epoch": 2.477810650887574, "frac_reward_zero_std": 0.0, "grad_norm": 2.53125, "kl": 0.016115693002939226, "learning_rate": 5.485977850811841e-07, "loss": 0.0006676806602627039, "reward": 1.6786929965019226, "reward_std": 0.17427631467580795, "rewards/FidelityReward/mean": 0.7857166826725006, "rewards/FidelityReward/std": 0.20463000982999802, "rewards/JudgeFidelityReward/mean": 0.7859526872634888, "rewards/JudgeFidelityReward/std": 0.1969461664557457, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3350 }, { "clip_ratio/high_max": 0.0004939674399793148, "clip_ratio/high_mean": 0.00012351740733720362, "clip_ratio/low_mean": 2.194907865487039e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014546648599207402, "completions/clipped_ratio": 0.0, "completions/max_length": 174.0, "completions/mean_length": 73.55078125, "completions/min_length": 31.333333333333332, "epoch": 2.481508875739645, "frac_reward_zero_std": 0.0, "grad_norm": 2.6875, "kl": 0.01569130625575781, "learning_rate": 5.473804507310369e-07, "loss": 0.000683925161138177, "reward": 1.645289381345113, "reward_std": 0.19504445294539133, "rewards/FidelityReward/mean": 0.7667504151662191, "rewards/FidelityReward/std": 0.20397621393203735, "rewards/JudgeFidelityReward/mean": 0.7570779720942179, "rewards/JudgeFidelityReward/std": 0.19502250850200653, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3355 }, { "clip_ratio/high_max": 0.0007024687365628779, "clip_ratio/high_mean": 0.00021233546431176364, "clip_ratio/low_mean": 6.64002203848213e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002787356846965849, "completions/clipped_ratio": 0.0, "completions/max_length": 160.0, "completions/mean_length": 71.51171875, "completions/min_length": 30.5, "epoch": 2.485207100591716, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.01623863261193037, "learning_rate": 5.461628329167515e-07, "loss": 0.000744206877425313, "reward": 1.5727177262306213, "reward_std": 0.20243708789348602, "rewards/FidelityReward/mean": 0.7132854461669922, "rewards/FidelityReward/std": 0.22753263264894485, "rewards/JudgeFidelityReward/mean": 0.7188644409179688, "rewards/JudgeFidelityReward/std": 0.2186075747013092, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3360 }, { "clip_ratio/high_max": 0.0008283460629172623, "clip_ratio/high_mean": 0.00026933851768262684, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00026933851768262684, "completions/clipped_ratio": 0.0, "completions/max_length": 192.0, "completions/mean_length": 73.12369791666667, "completions/min_length": 34.0, "epoch": 2.488905325443787, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.5, "kl": 0.01593233086168766, "learning_rate": 5.449449389229985e-07, "loss": 0.0006281005684286356, "reward": 1.6833409865697224, "reward_std": 0.16770333051681519, "rewards/FidelityReward/mean": 0.7924318909645081, "rewards/FidelityReward/std": 0.1975747694571813, "rewards/JudgeFidelityReward/mean": 0.7818181316057841, "rewards/JudgeFidelityReward/std": 0.20111531019210815, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3365 }, { "clip_ratio/high_max": 0.0006492432323284448, "clip_ratio/high_mean": 0.00016953413287410513, "clip_ratio/low_mean": 8.31430617836304e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002526772062992677, "completions/clipped_ratio": 0.0, "completions/max_length": 151.5, "completions/mean_length": 67.6953125, "completions/min_length": 33.0, "epoch": 2.492603550295858, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.01692327167838812, "learning_rate": 5.437267760361014e-07, "loss": 0.0008895833045244217, "reward": 1.6264852285385132, "reward_std": 0.18679401278495789, "rewards/FidelityReward/mean": 0.7442747354507446, "rewards/FidelityReward/std": 0.23831707239151, "rewards/JudgeFidelityReward/mean": 0.7683272063732147, "rewards/JudgeFidelityReward/std": 0.19350773841142654, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.044107433408498764, "step": 3370 }, { "clip_ratio/high_max": 0.0005771955940872431, "clip_ratio/high_mean": 0.00013695104280486703, "clip_ratio/low_mean": 9.03083011507988e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022725934395566582, "completions/clipped_ratio": 0.0, "completions/max_length": 175.0, "completions/mean_length": 71.92057291666667, "completions/min_length": 34.333333333333336, "epoch": 2.496301775147929, "frac_reward_zero_std": 0.0, "grad_norm": 2.359375, "kl": 0.017185625061392783, "learning_rate": 5.425083515439917e-07, "loss": 0.0007126753684133291, "reward": 1.5989731550216675, "reward_std": 0.20312732954819998, "rewards/FidelityReward/mean": 0.7309019565582275, "rewards/FidelityReward/std": 0.21903649965922037, "rewards/JudgeFidelityReward/mean": 0.7361424565315247, "rewards/JudgeFidelityReward/std": 0.21569586296876272, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3375 }, { "clip_ratio/high_max": 0.0005841189995408058, "clip_ratio/high_mean": 0.00013938342744950204, "clip_ratio/low_mean": 5.901817785343155e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019840160384774209, "completions/clipped_ratio": 0.0, "completions/max_length": 190.5, "completions/mean_length": 73.830078125, "completions/min_length": 34.0, "epoch": 2.5, "frac_reward_zero_std": 0.015625, "grad_norm": 2.625, "kl": 0.01447973195463419, "learning_rate": 5.412896727361662e-07, "loss": 0.0006285005249083043, "reward": 1.6742727160453796, "reward_std": 0.17623407393693924, "rewards/FidelityReward/mean": 0.7865262627601624, "rewards/FidelityReward/std": 0.19851671159267426, "rewards/JudgeFidelityReward/mean": 0.7754929065704346, "rewards/JudgeFidelityReward/std": 0.2049732282757759, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3380 }, { "clip_ratio/high_max": 0.0005506635527126492, "clip_ratio/high_mean": 0.00018141278414987027, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018141278414987027, "completions/clipped_ratio": 0.0, "completions/max_length": 165.33333333333334, "completions/mean_length": 71.27734375, "completions/min_length": 37.0, "epoch": 2.503698224852071, "frac_reward_zero_std": 0.03125, "grad_norm": 2.375, "kl": 0.015572307258844375, "learning_rate": 5.400707469036434e-07, "loss": 0.0005120966117829085, "reward": 1.6522684494654338, "reward_std": 0.1787417729695638, "rewards/FidelityReward/mean": 0.7674093445142111, "rewards/FidelityReward/std": 0.19806702435016632, "rewards/JudgeFidelityReward/mean": 0.7697182297706604, "rewards/JudgeFidelityReward/std": 0.19117670754591623, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3385 }, { "clip_ratio/high_max": 0.001015205285511911, "clip_ratio/high_mean": 0.00019251267658546568, "clip_ratio/low_mean": 9.239610517397523e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002849087817594409, "completions/clipped_ratio": 0.0, "completions/max_length": 204.0, "completions/mean_length": 69.92578125, "completions/min_length": 33.0, "epoch": 2.507396449704142, "frac_reward_zero_std": 0.015625, "grad_norm": 2.46875, "kl": 0.0161716578528285, "learning_rate": 5.388515813389197e-07, "loss": 0.0007536577992141247, "reward": 1.643006443977356, "reward_std": 0.18676412850618362, "rewards/FidelityReward/mean": 0.7616667747497559, "rewards/FidelityReward/std": 0.18991758674383163, "rewards/JudgeFidelityReward/mean": 0.7626793086528778, "rewards/JudgeFidelityReward/std": 0.17279166728258133, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3390 }, { "clip_ratio/high_max": 0.00037643546238541603, "clip_ratio/high_mean": 8.458757365588098e-05, "clip_ratio/low_mean": 0.00010810631065396591, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0001926938828546554, "completions/clipped_ratio": 0.0, "completions/max_length": 177.66666666666666, "completions/mean_length": 71.04296875, "completions/min_length": 35.0, "epoch": 2.511094674556213, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.734375, "kl": 0.015521319583058357, "learning_rate": 5.376321833359257e-07, "loss": 0.00048069488257169725, "reward": 1.6304432551066081, "reward_std": 0.18101214369138083, "rewards/FidelityReward/mean": 0.7497201760609945, "rewards/FidelityReward/std": 0.2144190122683843, "rewards/JudgeFidelityReward/mean": 0.7614461779594421, "rewards/JudgeFidelityReward/std": 0.20414142310619354, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3395 }, { "clip_ratio/high_max": 0.0010363588575273753, "clip_ratio/high_mean": 0.00023282975307665766, "clip_ratio/low_mean": 0.00011720146721927449, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003500312101095915, "completions/clipped_ratio": 0.0, "completions/max_length": 151.5, "completions/mean_length": 72.90625, "completions/min_length": 35.0, "epoch": 2.5147928994082838, "frac_reward_zero_std": 0.015625, "grad_norm": 2.859375, "kl": 0.01575496718287468, "learning_rate": 5.364125601899822e-07, "loss": 0.0007129173725843429, "reward": 1.6842058300971985, "reward_std": 0.17334268987178802, "rewards/FidelityReward/mean": 0.7907238006591797, "rewards/FidelityReward/std": 0.19162337481975555, "rewards/JudgeFidelityReward/mean": 0.7889170944690704, "rewards/JudgeFidelityReward/std": 0.2128721997141838, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3400 }, { "clip_ratio/high_max": 0.00014245014172047376, "clip_ratio/high_mean": 3.531327238306403e-05, "clip_ratio/low_mean": 2.380952355451882e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 5.912279593758285e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 175.33333333333334, "completions/mean_length": 71.70442708333333, "completions/min_length": 30.333333333333332, "epoch": 2.518491124260355, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.609375, "kl": 0.015880966931581496, "learning_rate": 5.351927191977575e-07, "loss": 0.0006200679112225771, "reward": 1.6422985394795735, "reward_std": 0.18162252008914948, "rewards/FidelityReward/mean": 0.7615671555201212, "rewards/FidelityReward/std": 0.20965394377708435, "rewards/JudgeFidelityReward/mean": 0.7627648512522379, "rewards/JudgeFidelityReward/std": 0.20048115650812784, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3405 }, { "clip_ratio/high_max": 0.0007627679267898202, "clip_ratio/high_mean": 0.00016719479172024875, "clip_ratio/low_mean": 6.159389158710837e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002287886803969741, "completions/clipped_ratio": 0.0, "completions/max_length": 198.5, "completions/mean_length": 73.384765625, "completions/min_length": 36.0, "epoch": 2.522189349112426, "frac_reward_zero_std": 0.0, "grad_norm": 2.609375, "kl": 0.017590552382171155, "learning_rate": 5.339726676572231e-07, "loss": 0.0007775316014885902, "reward": 1.6363530158996582, "reward_std": 0.17717376351356506, "rewards/FidelityReward/mean": 0.7593198716640472, "rewards/FidelityReward/std": 0.21661311388015747, "rewards/JudgeFidelityReward/mean": 0.7579725980758667, "rewards/JudgeFidelityReward/std": 0.2148958221077919, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 3410 }, { "clip_ratio/high_max": 0.0005269553046673536, "clip_ratio/high_mean": 0.00013537068734876813, "clip_ratio/low_mean": 4.992870963178575e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00018529939698055386, "completions/clipped_ratio": 0.0, "completions/max_length": 175.66666666666666, "completions/mean_length": 72.75651041666667, "completions/min_length": 32.333333333333336, "epoch": 2.5258875739644973, "frac_reward_zero_std": 0.03125, "grad_norm": 2.703125, "kl": 0.016529225185513495, "learning_rate": 5.3275241286761e-07, "loss": 0.000668150931596756, "reward": 1.6429901520411174, "reward_std": 0.18200770020484924, "rewards/FidelityReward/mean": 0.7651085257530212, "rewards/FidelityReward/std": 0.21147534747918448, "rewards/JudgeFidelityReward/mean": 0.7557632724444071, "rewards/JudgeFidelityReward/std": 0.2087735782066981, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3415 }, { "clip_ratio/high_max": 0.0007082952186465263, "clip_ratio/high_mean": 0.00015580216131638736, "clip_ratio/low_mean": 1.933487947098911e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017513704078737647, "completions/clipped_ratio": 0.0, "completions/max_length": 141.0, "completions/mean_length": 70.005859375, "completions/min_length": 27.0, "epoch": 2.529585798816568, "frac_reward_zero_std": 0.0625, "grad_norm": 3.359375, "kl": 0.01749461852014065, "learning_rate": 5.315319621293651e-07, "loss": 0.0007435701321810484, "reward": 1.640565276145935, "reward_std": 0.15992796421051025, "rewards/FidelityReward/mean": 0.7535223364830017, "rewards/FidelityReward/std": 0.23261743038892746, "rewards/JudgeFidelityReward/mean": 0.7740859985351562, "rewards/JudgeFidelityReward/std": 0.19784902781248093, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3420 }, { "clip_ratio/high_max": 0.0006786898477002979, "clip_ratio/high_mean": 0.00021376414806582034, "clip_ratio/low_mean": 4.248974728398025e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002562538837082684, "completions/clipped_ratio": 0.0, "completions/max_length": 185.66666666666666, "completions/mean_length": 73.9375, "completions/min_length": 35.666666666666664, "epoch": 2.533284023668639, "frac_reward_zero_std": 0.0, "grad_norm": 2.3125, "kl": 0.017626637220382692, "learning_rate": 5.303113227441082e-07, "loss": 0.0008268042467534543, "reward": 1.6780959765116374, "reward_std": 0.18322082360585532, "rewards/FidelityReward/mean": 0.7860134442647299, "rewards/FidelityReward/std": 0.20364787181218466, "rewards/JudgeFidelityReward/mean": 0.7841650644938151, "rewards/JudgeFidelityReward/std": 0.19712579747041067, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3425 }, { "clip_ratio/high_max": 0.0010991731425747276, "clip_ratio/high_mean": 0.00032282675092574207, "clip_ratio/low_mean": 0.00013677188544534147, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00045959863346070053, "completions/clipped_ratio": 0.0, "completions/max_length": 201.5, "completions/mean_length": 71.681640625, "completions/min_length": 35.0, "epoch": 2.53698224852071, "frac_reward_zero_std": 0.015625, "grad_norm": 2.765625, "kl": 0.014296785183250903, "learning_rate": 5.29090502014587e-07, "loss": 0.0005284789018332958, "reward": 1.6351116299629211, "reward_std": 0.1750272884964943, "rewards/FidelityReward/mean": 0.7564968466758728, "rewards/FidelityReward/std": 0.20349252223968506, "rewards/JudgeFidelityReward/mean": 0.7572296857833862, "rewards/JudgeFidelityReward/std": 0.1906854212284088, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3430 }, { "clip_ratio/high_max": 0.0003064149175770581, "clip_ratio/high_mean": 7.673658983549103e-05, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 7.673658983549103e-05, "completions/clipped_ratio": 0.0, "completions/max_length": 156.33333333333334, "completions/mean_length": 72.328125, "completions/min_length": 36.333333333333336, "epoch": 2.540680473372781, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.734375, "kl": 0.016962099447846414, "learning_rate": 5.278695072446342e-07, "loss": 0.0007684526033699512, "reward": 1.6400070190429688, "reward_std": 0.17159574727217355, "rewards/FidelityReward/mean": 0.7583954135576884, "rewards/FidelityReward/std": 0.2255691240231196, "rewards/JudgeFidelityReward/mean": 0.7632232705752054, "rewards/JudgeFidelityReward/std": 0.2141202688217163, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3435 }, { "clip_ratio/high_max": 0.001071858056820929, "clip_ratio/high_mean": 0.00030702628428116443, "clip_ratio/low_mean": 4.865727678406984e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003556835581548512, "completions/clipped_ratio": 0.0, "completions/max_length": 183.0, "completions/mean_length": 73.44921875, "completions/min_length": 29.5, "epoch": 2.544378698224852, "frac_reward_zero_std": 0.015625, "grad_norm": 2.546875, "kl": 0.016911606304347515, "learning_rate": 5.266483457391243e-07, "loss": 0.0009853214025497437, "reward": 1.5982559323310852, "reward_std": 0.1820329874753952, "rewards/FidelityReward/mean": 0.7301624715328217, "rewards/FidelityReward/std": 0.24145007878541946, "rewards/JudgeFidelityReward/mean": 0.7381400465965271, "rewards/JudgeFidelityReward/std": 0.22871820628643036, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3440 }, { "clip_ratio/high_max": 0.0004890066105872392, "clip_ratio/high_mean": 0.000175472442060709, "clip_ratio/low_mean": 1.7882690008264036e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00019335512770339846, "completions/clipped_ratio": 0.0, "completions/max_length": 173.66666666666666, "completions/mean_length": 73.3046875, "completions/min_length": 32.333333333333336, "epoch": 2.5480769230769234, "frac_reward_zero_std": 0.041666666666666664, "grad_norm": 2.359375, "kl": 0.01739159654825926, "learning_rate": 5.25427024803929e-07, "loss": 0.000734570948407054, "reward": 1.660417636235555, "reward_std": 0.17300191521644592, "rewards/FidelityReward/mean": 0.7742125193277994, "rewards/FidelityReward/std": 0.21218631168206534, "rewards/JudgeFidelityReward/mean": 0.7724102735519409, "rewards/JudgeFidelityReward/std": 0.21259944637616476, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3445 }, { "clip_ratio/high_max": 0.0008623625966720283, "clip_ratio/high_mean": 0.00025671723851701246, "clip_ratio/low_mean": 6.519429880427196e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003219115373212844, "completions/clipped_ratio": 0.0, "completions/max_length": 186.5, "completions/mean_length": 75.806640625, "completions/min_length": 29.5, "epoch": 2.551775147928994, "frac_reward_zero_std": 0.015625, "grad_norm": 2.53125, "kl": 0.015623002871870994, "learning_rate": 5.242055517458736e-07, "loss": 0.0008214743807911873, "reward": 1.6536080837249756, "reward_std": 0.20472802966833115, "rewards/FidelityReward/mean": 0.7704154253005981, "rewards/FidelityReward/std": 0.19336394220590591, "rewards/JudgeFidelityReward/mean": 0.7663850486278534, "rewards/JudgeFidelityReward/std": 0.19705870002508163, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3450 }, { "clip_ratio/high_max": 0.0004828738514333963, "clip_ratio/high_mean": 7.806461362633855e-05, "clip_ratio/low_mean": 6.783660501241684e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00014590122736990452, "completions/clipped_ratio": 0.0, "completions/max_length": 207.0, "completions/mean_length": 73.24088541666667, "completions/min_length": 35.333333333333336, "epoch": 2.5554733727810652, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.01529335491359234, "learning_rate": 5.229839338726939e-07, "loss": 0.0007492826320230961, "reward": 1.671237587928772, "reward_std": 0.18033512433369955, "rewards/FidelityReward/mean": 0.7779099742571512, "rewards/FidelityReward/std": 0.1975456823905309, "rewards/JudgeFidelityReward/mean": 0.7879572510719299, "rewards/JudgeFidelityReward/std": 0.1872896800438563, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3455 }, { "clip_ratio/high_max": 0.0010574389481917024, "clip_ratio/high_mean": 0.0003198098798748106, "clip_ratio/low_mean": 5.790957657154649e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00037771945935674014, "completions/clipped_ratio": 0.0, "completions/max_length": 153.0, "completions/mean_length": 72.037109375, "completions/min_length": 30.0, "epoch": 2.559171597633136, "frac_reward_zero_std": 0.015625, "grad_norm": 2.375, "kl": 0.01552974283695221, "learning_rate": 5.217621784929919e-07, "loss": 0.0007498861290514469, "reward": 1.661380112171173, "reward_std": 0.16094663739204407, "rewards/FidelityReward/mean": 0.7716929614543915, "rewards/FidelityReward/std": 0.21939129382371902, "rewards/JudgeFidelityReward/mean": 0.779374361038208, "rewards/JudgeFidelityReward/std": 0.21535586565732956, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3460 }, { "clip_ratio/high_max": 0.0005872637964785099, "clip_ratio/high_mean": 0.000203060352941975, "clip_ratio/low_mean": 4.660178819904104e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024966213968582454, "completions/clipped_ratio": 0.0, "completions/max_length": 189.33333333333334, "completions/mean_length": 74.42447916666667, "completions/min_length": 34.666666666666664, "epoch": 2.562869822485207, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.765625, "kl": 0.015019410289824009, "learning_rate": 5.205402929161921e-07, "loss": 0.0006149125285446644, "reward": 1.6735492944717407, "reward_std": 0.18168631692727408, "rewards/FidelityReward/mean": 0.7851503491401672, "rewards/FidelityReward/std": 0.1923565318187078, "rewards/JudgeFidelityReward/mean": 0.7794021169344584, "rewards/JudgeFidelityReward/std": 0.19469433526198068, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 3465 }, { "clip_ratio/high_max": 0.001075040316209197, "clip_ratio/high_mean": 0.00022233005729503929, "clip_ratio/low_mean": 9.229211427737027e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00031462218903470783, "completions/clipped_ratio": 0.0, "completions/max_length": 160.5, "completions/mean_length": 70.423828125, "completions/min_length": 28.5, "epoch": 2.5665680473372783, "frac_reward_zero_std": 0.015625, "grad_norm": 2.484375, "kl": 0.015350476093590259, "learning_rate": 5.193182844524982e-07, "loss": 0.0007411763072013855, "reward": 1.5819717049598694, "reward_std": 0.1806241199374199, "rewards/FidelityReward/mean": 0.7240617573261261, "rewards/FidelityReward/std": 0.2414209172129631, "rewards/JudgeFidelityReward/mean": 0.7158199846744537, "rewards/JudgeFidelityReward/std": 0.24890732765197754, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3470 }, { "clip_ratio/high_max": 0.0005139359855093062, "clip_ratio/high_mean": 0.00018807751475833357, "clip_ratio/low_mean": 6.036943523213267e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00024844694416970015, "completions/clipped_ratio": 0.0, "completions/max_length": 242.66666666666666, "completions/mean_length": 75.84244791666667, "completions/min_length": 31.0, "epoch": 2.570266272189349, "frac_reward_zero_std": 0.0, "grad_norm": 2.75, "kl": 0.018345337733626364, "learning_rate": 5.180961604128493e-07, "loss": 0.0008887166157364845, "reward": 1.686464746793111, "reward_std": 0.17724073926607767, "rewards/FidelityReward/mean": 0.7896530032157898, "rewards/FidelityReward/std": 0.1965556095043818, "rewards/JudgeFidelityReward/mean": 0.7936235666275024, "rewards/JudgeFidelityReward/std": 0.19430742661158243, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3475 }, { "clip_ratio/high_max": 0.0013762902468442916, "clip_ratio/high_mean": 0.0003456025617197156, "clip_ratio/low_mean": 0.00010222018754575401, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004478227347135544, "completions/clipped_ratio": 0.0, "completions/max_length": 181.5, "completions/mean_length": 73.837890625, "completions/min_length": 32.0, "epoch": 2.57396449704142, "frac_reward_zero_std": 0.03125, "grad_norm": 2.53125, "kl": 0.01618808079510927, "learning_rate": 5.168739281088753e-07, "loss": 0.0005941619630903006, "reward": 1.653055489063263, "reward_std": 0.1690966635942459, "rewards/FidelityReward/mean": 0.7709374129772186, "rewards/FidelityReward/std": 0.21840982139110565, "rewards/JudgeFidelityReward/mean": 0.7642362415790558, "rewards/JudgeFidelityReward/std": 0.21982568502426147, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3480 }, { "clip_ratio/high_max": 0.0007414272637106478, "clip_ratio/high_mean": 0.00011042148835258559, "clip_ratio/low_mean": 6.636127654928714e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00017678276635706424, "completions/clipped_ratio": 0.0, "completions/max_length": 212.33333333333334, "completions/mean_length": 74.82421875, "completions/min_length": 36.0, "epoch": 2.577662721893491, "frac_reward_zero_std": 0.0, "grad_norm": 2.71875, "kl": 0.015571098960936069, "learning_rate": 5.156515948528543e-07, "loss": 0.0008384859189391136, "reward": 1.5892374515533447, "reward_std": 0.19497307141621908, "rewards/FidelityReward/mean": 0.7290468215942383, "rewards/FidelityReward/std": 0.20341858764489493, "rewards/JudgeFidelityReward/mean": 0.7203812599182129, "rewards/JudgeFidelityReward/std": 0.2055687059958776, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3485 }, { "clip_ratio/high_max": 0.0010357402847148479, "clip_ratio/high_mean": 0.0002994820300955325, "clip_ratio/low_mean": 8.509930776199326e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00038458132767118516, "completions/clipped_ratio": 0.0, "completions/max_length": 212.5, "completions/mean_length": 74.3203125, "completions/min_length": 37.0, "epoch": 2.581360946745562, "frac_reward_zero_std": 0.015625, "grad_norm": 2.3125, "kl": 0.017154406383633615, "learning_rate": 5.144291679576682e-07, "loss": 0.0008559220470488072, "reward": 1.6395904421806335, "reward_std": 0.18048837780952454, "rewards/FidelityReward/mean": 0.755976140499115, "rewards/FidelityReward/std": 0.22029148787260056, "rewards/JudgeFidelityReward/mean": 0.7672286331653595, "rewards/JudgeFidelityReward/std": 0.2137579768896103, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3490 }, { "clip_ratio/high_max": 0.0004789775237441063, "clip_ratio/high_mean": 0.0001823335071094334, "clip_ratio/low_mean": 3.935039276257157e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022168389987200498, "completions/clipped_ratio": 0.0, "completions/max_length": 235.66666666666666, "completions/mean_length": 75.31640625, "completions/min_length": 36.0, "epoch": 2.585059171597633, "frac_reward_zero_std": 0.020833333333333332, "grad_norm": 2.78125, "kl": 0.017915087938308715, "learning_rate": 5.132066547367593e-07, "loss": 0.0007370003499090671, "reward": 1.6817335685094197, "reward_std": 0.1833459436893463, "rewards/FidelityReward/mean": 0.786293625831604, "rewards/FidelityReward/std": 0.20974445839722952, "rewards/JudgeFidelityReward/mean": 0.7934840122858683, "rewards/JudgeFidelityReward/std": 0.19120897352695465, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 3495 }, { "clip_ratio/high_max": 0.0005483202636241913, "clip_ratio/high_mean": 0.00023312328848987818, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00023312328848987818, "completions/clipped_ratio": 0.0, "completions/max_length": 184.5, "completions/mean_length": 73.77734375, "completions/min_length": 37.0, "epoch": 2.5887573964497044, "frac_reward_zero_std": 0.0, "grad_norm": 2.46875, "kl": 0.015967142209410666, "learning_rate": 5.11984062504086e-07, "loss": 0.0007830827496945858, "reward": 1.6803748607635498, "reward_std": 0.1785207837820053, "rewards/FidelityReward/mean": 0.7923301458358765, "rewards/FidelityReward/std": 0.19387931376695633, "rewards/JudgeFidelityReward/mean": 0.7760893702507019, "rewards/JudgeFidelityReward/std": 0.21490206569433212, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3500 }, { "clip_ratio/high_max": 0.0013482167851179838, "clip_ratio/high_mean": 0.00025603139074519276, "clip_ratio/low_mean": 9.46127693168819e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003506441367790103, "completions/clipped_ratio": 0.0, "completions/max_length": 206.33333333333334, "completions/mean_length": 72.50520833333333, "completions/min_length": 30.666666666666668, "epoch": 2.592455621301775, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.578125, "kl": 0.016145676001906396, "learning_rate": 5.1076139857408e-07, "loss": 0.0006254219450056552, "reward": 1.6429896354675293, "reward_std": 0.18847601612408957, "rewards/FidelityReward/mean": 0.7640984654426575, "rewards/FidelityReward/std": 0.20148962239424387, "rewards/JudgeFidelityReward/mean": 0.759084423383077, "rewards/JudgeFidelityReward/std": 0.1925234099229177, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3505 }, { "clip_ratio/high_max": 0.000922340783290565, "clip_ratio/high_mean": 0.0003157781029585749, "clip_ratio/low_mean": 9.113507112488151e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0004069131682626903, "completions/clipped_ratio": 0.0, "completions/max_length": 202.0, "completions/mean_length": 73.470703125, "completions/min_length": 29.5, "epoch": 2.5961538461538463, "frac_reward_zero_std": 0.0, "grad_norm": 2.59375, "kl": 0.016344299539923668, "learning_rate": 5.095386702616011e-07, "loss": 0.0008332354016602039, "reward": 1.6381723284721375, "reward_std": 0.18014196306467056, "rewards/FidelityReward/mean": 0.7610631883144379, "rewards/FidelityReward/std": 0.19340310245752335, "rewards/JudgeFidelityReward/mean": 0.7542181611061096, "rewards/JudgeFidelityReward/std": 0.17034592479467392, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3510 }, { "clip_ratio/high_max": 0.00032161910785362126, "clip_ratio/high_mean": 0.00013977074704598635, "clip_ratio/low_mean": 0.0001413791556842625, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002811499056406319, "completions/clipped_ratio": 0.0, "completions/max_length": 161.0, "completions/mean_length": 72.28385416666667, "completions/min_length": 31.666666666666668, "epoch": 2.599852071005917, "frac_reward_zero_std": 0.0, "grad_norm": 2.515625, "kl": 0.018617318570613862, "learning_rate": 5.083158848818952e-07, "loss": 0.0007181039080023765, "reward": 1.6125059127807617, "reward_std": 0.2026425153017044, "rewards/FidelityReward/mean": 0.7420878012975057, "rewards/FidelityReward/std": 0.21303904553254446, "rewards/JudgeFidelityReward/mean": 0.7408363223075867, "rewards/JudgeFidelityReward/std": 0.21480160454909006, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3515 }, { "clip_ratio/high_max": 0.000764935917686671, "clip_ratio/high_mean": 0.0002069045091047883, "clip_ratio/low_mean": 9.339378884760663e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00030029829940758646, "completions/clipped_ratio": 0.0, "completions/max_length": 190.0, "completions/mean_length": 69.302734375, "completions/min_length": 34.0, "epoch": 2.603550295857988, "frac_reward_zero_std": 0.0, "grad_norm": 2.765625, "kl": 0.016515288129448892, "learning_rate": 5.07093049750549e-07, "loss": 0.0005946914199739695, "reward": 1.6267205476760864, "reward_std": 0.18854700028896332, "rewards/FidelityReward/mean": 0.7527383863925934, "rewards/FidelityReward/std": 0.21184540539979935, "rewards/JudgeFidelityReward/mean": 0.7479643821716309, "rewards/JudgeFidelityReward/std": 0.2023429125547409, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3520 }, { "clip_ratio/high_max": 0.00036572788376361133, "clip_ratio/high_mean": 0.0001470182673074305, "clip_ratio/low_mean": 6.517751899082214e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021219578920863568, "completions/clipped_ratio": 0.0, "completions/max_length": 209.33333333333334, "completions/mean_length": 74.85026041666667, "completions/min_length": 34.333333333333336, "epoch": 2.6072485207100593, "frac_reward_zero_std": 0.03125, "grad_norm": 2.875, "kl": 0.016993923857808114, "learning_rate": 5.05870172183447e-07, "loss": 0.0006612356286495924, "reward": 1.658161203066508, "reward_std": 0.18161562581857046, "rewards/FidelityReward/mean": 0.7737241586049398, "rewards/FidelityReward/std": 0.22472377121448517, "rewards/JudgeFidelityReward/mean": 0.7727803786595663, "rewards/JudgeFidelityReward/std": 0.2180285503466924, "rewards/SelfEvolvingFormatReward/mean": 0.99609375, "rewards/SelfEvolvingFormatReward/std": 0.0625, "step": 3525 }, { "clip_ratio/high_max": 0.0006817619083449245, "clip_ratio/high_mean": 0.0001941217400599271, "clip_ratio/low_mean": 2.8116301109548658e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00022223804844543338, "completions/clipped_ratio": 0.0, "completions/max_length": 198.5, "completions/mean_length": 73.763671875, "completions/min_length": 28.5, "epoch": 2.61094674556213, "frac_reward_zero_std": 0.03125, "grad_norm": 2.5, "kl": 0.017438323982059956, "learning_rate": 5.046472594967278e-07, "loss": 0.0009533341042697429, "reward": 1.6498927474021912, "reward_std": 0.18681801855564117, "rewards/FidelityReward/mean": 0.7645850777626038, "rewards/FidelityReward/std": 0.20037664473056793, "rewards/JudgeFidelityReward/mean": 0.7725686132907867, "rewards/JudgeFidelityReward/std": 0.18484848737716675, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3530 }, { "clip_ratio/high_max": 0.00031435483833774924, "clip_ratio/high_mean": 7.97571730799973e-05, "clip_ratio/low_mean": 7.390094688162208e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00015365811996161937, "completions/clipped_ratio": 0.0, "completions/max_length": 245.0, "completions/mean_length": 75.45182291666667, "completions/min_length": 31.666666666666668, "epoch": 2.614644970414201, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.625, "kl": 0.017714506760239602, "learning_rate": 5.034243190067397e-07, "loss": 0.0007027795538306237, "reward": 1.6576956907908122, "reward_std": 0.177254190047582, "rewards/FidelityReward/mean": 0.7705567677815756, "rewards/FidelityReward/std": 0.20825706919034323, "rewards/JudgeFidelityReward/mean": 0.7768820722897848, "rewards/JudgeFidelityReward/std": 0.19452864428361258, "rewards/SelfEvolvingFormatReward/mean": 0.9973958333333334, "rewards/SelfEvolvingFormatReward/std": 0.029404955605665844, "step": 3535 }, { "clip_ratio/high_max": 0.0008672773838043213, "clip_ratio/high_mean": 0.00021322899265214802, "clip_ratio/low_mean": 0.00018019655835814773, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00039342555101029575, "completions/clipped_ratio": 0.0, "completions/max_length": 174.0, "completions/mean_length": 71.58203125, "completions/min_length": 35.0, "epoch": 2.6183431952662723, "frac_reward_zero_std": 0.03125, "grad_norm": 2.828125, "kl": 0.01650921441614628, "learning_rate": 5.022013580299977e-07, "loss": 0.0006515599321573973, "reward": 1.685625672340393, "reward_std": 0.17110208421945572, "rewards/FidelityReward/mean": 0.7930035293102264, "rewards/FidelityReward/std": 0.19151516258716583, "rewards/JudgeFidelityReward/mean": 0.7871973812580109, "rewards/JudgeFidelityReward/std": 0.18537630140781403, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3540 }, { "clip_ratio/high_max": 0.0005337446695193648, "clip_ratio/high_mean": 0.00019235331565141677, "clip_ratio/low_mean": 2.3900573432911186e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021625388762913644, "completions/clipped_ratio": 0.0, "completions/max_length": 169.0, "completions/mean_length": 71.12760416666667, "completions/min_length": 33.333333333333336, "epoch": 2.622041420118343, "frac_reward_zero_std": 0.041666666666666664, "grad_norm": 2.609375, "kl": 0.016170925833284854, "learning_rate": 5.009783838831391e-07, "loss": 0.0007223475724458695, "reward": 1.7016384601593018, "reward_std": 0.15563598026831946, "rewards/FidelityReward/mean": 0.7973236044247946, "rewards/FidelityReward/std": 0.19339034954706827, "rewards/JudgeFidelityReward/mean": 0.8086298306783041, "rewards/JudgeFidelityReward/std": 0.1873157024383545, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3545 }, { "clip_ratio/high_max": 0.0012156442739069463, "clip_ratio/high_mean": 0.0003105209209024906, "clip_ratio/low_mean": 6.332237389869989e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003738433122634888, "completions/clipped_ratio": 0.0, "completions/max_length": 194.0, "completions/mean_length": 72.796875, "completions/min_length": 35.5, "epoch": 2.6257396449704142, "frac_reward_zero_std": 0.03125, "grad_norm": 2.546875, "kl": 0.015535584278404712, "learning_rate": 4.997554038828801e-07, "loss": 0.0007952114567160607, "reward": 1.7074615359306335, "reward_std": 0.1598820760846138, "rewards/FidelityReward/mean": 0.8030354976654053, "rewards/FidelityReward/std": 0.1921905353665352, "rewards/JudgeFidelityReward/mean": 0.8108053207397461, "rewards/JudgeFidelityReward/std": 0.17426683008670807, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3550 }, { "clip_ratio/high_max": 0.0004936539335176349, "clip_ratio/high_mean": 0.00021320724044926465, "clip_ratio/low_mean": 4.421811609063297e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00025742535945028067, "completions/clipped_ratio": 0.0, "completions/max_length": 192.0, "completions/mean_length": 71.70052083333333, "completions/min_length": 31.0, "epoch": 2.6294378698224854, "frac_reward_zero_std": 0.0, "grad_norm": 2.640625, "kl": 0.016774178110063077, "learning_rate": 4.98532425345972e-07, "loss": 0.0006478605791926384, "reward": 1.6847916046778362, "reward_std": 0.17938796679178873, "rewards/FidelityReward/mean": 0.7925485571225485, "rewards/FidelityReward/std": 0.19356938203175864, "rewards/JudgeFidelityReward/mean": 0.7844861547152201, "rewards/JudgeFidelityReward/std": 0.19107510646184286, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3555 }, { "clip_ratio/high_max": 0.0008715453906916082, "clip_ratio/high_mean": 0.00023773456341587008, "clip_ratio/low_mean": 8.88379494426772e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003265725099481642, "completions/clipped_ratio": 0.0, "completions/max_length": 165.5, "completions/mean_length": 70.03125, "completions/min_length": 24.5, "epoch": 2.633136094674556, "frac_reward_zero_std": 0.03125, "grad_norm": 2.328125, "kl": 0.018832525610923766, "learning_rate": 4.973094555891577e-07, "loss": 0.0010611603036522865, "reward": 1.6885729432106018, "reward_std": 0.17074321955442429, "rewards/FidelityReward/mean": 0.7916811108589172, "rewards/FidelityReward/std": 0.20831681042909622, "rewards/JudgeFidelityReward/mean": 0.7937837541103363, "rewards/JudgeFidelityReward/std": 0.2122698649764061, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3560 }, { "clip_ratio/high_max": 0.000781543692573905, "clip_ratio/high_mean": 0.00012259336072020232, "clip_ratio/low_mean": 4.581817192956805e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00016841153264977038, "completions/clipped_ratio": 0.0, "completions/max_length": 173.0, "completions/mean_length": 71.38411458333333, "completions/min_length": 36.0, "epoch": 2.6368343195266273, "frac_reward_zero_std": 0.03125, "grad_norm": 2.921875, "kl": 0.016148069128394126, "learning_rate": 4.960865019291267e-07, "loss": 0.0006869577337056398, "reward": 1.6099753379821777, "reward_std": 0.18839075664679208, "rewards/FidelityReward/mean": 0.7372559507687887, "rewards/FidelityReward/std": 0.21695268154144287, "rewards/JudgeFidelityReward/mean": 0.7467408577601115, "rewards/JudgeFidelityReward/std": 0.19685485462347665, "rewards/SelfEvolvingFormatReward/mean": 0.9986979166666666, "rewards/SelfEvolvingFormatReward/std": 0.020833333333333332, "step": 3565 }, { "clip_ratio/high_max": 0.0008748894091695547, "clip_ratio/high_mean": 0.00030984127952251586, "clip_ratio/low_mean": 1.784439664334059e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003276856761658564, "completions/clipped_ratio": 0.0, "completions/max_length": 186.0, "completions/mean_length": 73.373046875, "completions/min_length": 37.0, "epoch": 2.640532544378698, "frac_reward_zero_std": 0.03125, "grad_norm": 2.515625, "kl": 0.01595602445304394, "learning_rate": 4.948635716824727e-07, "loss": 0.0007612812332808971, "reward": 1.7367616891860962, "reward_std": 0.15767281502485275, "rewards/FidelityReward/mean": 0.8234921991825104, "rewards/FidelityReward/std": 0.15851619839668274, "rewards/JudgeFidelityReward/mean": 0.8265389502048492, "rewards/JudgeFidelityReward/std": 0.15429802983999252, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3570 }, { "clip_ratio/high_max": 0.0006769775762222707, "clip_ratio/high_mean": 0.00018285540281794965, "clip_ratio/low_mean": 3.675988991744816e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021961529273539782, "completions/clipped_ratio": 0.0, "completions/max_length": 167.0, "completions/mean_length": 73.03385416666667, "completions/min_length": 33.0, "epoch": 2.644230769230769, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.328125, "kl": 0.01562193613499403, "learning_rate": 4.936406721656491e-07, "loss": 0.0005227226763963699, "reward": 1.670483668645223, "reward_std": 0.17934003472328186, "rewards/FidelityReward/mean": 0.7781223058700562, "rewards/FidelityReward/std": 0.1973110536734263, "rewards/JudgeFidelityReward/mean": 0.784722626209259, "rewards/JudgeFidelityReward/std": 0.17903087536493936, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3575 }, { "clip_ratio/high_max": 0.000907059502787888, "clip_ratio/high_mean": 0.00017711142427287996, "clip_ratio/low_mean": 9.51775029534474e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0002722889243159443, "completions/clipped_ratio": 0.0, "completions/max_length": 217.5, "completions/mean_length": 71.197265625, "completions/min_length": 33.0, "epoch": 2.6479289940828403, "frac_reward_zero_std": 0.0, "grad_norm": 2.578125, "kl": 0.018808532878756522, "learning_rate": 4.924178106949261e-07, "loss": 0.0007429330144077539, "reward": 1.7357755899429321, "reward_std": 0.14614348113536835, "rewards/FidelityReward/mean": 0.8274551928043365, "rewards/FidelityReward/std": 0.20073121786117554, "rewards/JudgeFidelityReward/mean": 0.818593829870224, "rewards/JudgeFidelityReward/std": 0.19277048110961914, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3580 }, { "clip_ratio/high_max": 0.0005849914392456413, "clip_ratio/high_mean": 0.0001995645638089627, "clip_ratio/low_mean": 7.445451337844134e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00027401907718740406, "completions/clipped_ratio": 0.0, "completions/max_length": 183.66666666666666, "completions/mean_length": 72.13932291666667, "completions/min_length": 32.0, "epoch": 2.6516272189349115, "frac_reward_zero_std": 0.010416666666666666, "grad_norm": 2.8125, "kl": 0.016630781814455986, "learning_rate": 4.911949945863456e-07, "loss": 0.0007584706880152225, "reward": 1.6395837465922039, "reward_std": 0.1900778959194819, "rewards/FidelityReward/mean": 0.7553995847702026, "rewards/FidelityReward/std": 0.22443845868110657, "rewards/JudgeFidelityReward/mean": 0.7683684031168619, "rewards/JudgeFidelityReward/std": 0.21363485852877298, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3585 }, { "clip_ratio/high_max": 0.0007138388347811997, "clip_ratio/high_mean": 0.0002021533262450248, "clip_ratio/low_mean": 0.00012974680139450357, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003319001174531877, "completions/clipped_ratio": 0.0, "completions/max_length": 179.5, "completions/mean_length": 72.1015625, "completions/min_length": 35.0, "epoch": 2.655325443786982, "frac_reward_zero_std": 0.03125, "grad_norm": 2.8125, "kl": 0.017105581983923913, "learning_rate": 4.899722311556783e-07, "loss": 0.00075254300609231, "reward": 1.6421764492988586, "reward_std": 0.1940135359764099, "rewards/FidelityReward/mean": 0.7629869878292084, "rewards/FidelityReward/std": 0.19703851640224457, "rewards/JudgeFidelityReward/mean": 0.7583790421485901, "rewards/JudgeFidelityReward/std": 0.1935538575053215, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3590 }, { "clip_ratio/high_max": 0.0005255649331957102, "clip_ratio/high_mean": 0.00021219729678705334, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.00021219729678705334, "completions/clipped_ratio": 0.0, "completions/max_length": 181.0, "completions/mean_length": 72.08463541666667, "completions/min_length": 35.666666666666664, "epoch": 2.6590236686390534, "frac_reward_zero_std": 0.041666666666666664, "grad_norm": 2.328125, "kl": 0.016532824374735355, "learning_rate": 4.887495277183797e-07, "loss": 0.0008129368536174298, "reward": 1.685261646906535, "reward_std": 0.16869520644346872, "rewards/FidelityReward/mean": 0.789252499739329, "rewards/FidelityReward/std": 0.19379858672618866, "rewards/JudgeFidelityReward/mean": 0.7920182347297668, "rewards/JudgeFidelityReward/std": 0.18465912342071533, "rewards/SelfEvolvingFormatReward/mean": 1.0, "rewards/SelfEvolvingFormatReward/std": 0.0, "step": 3595 }, { "clip_ratio/high_max": 0.000690931617282331, "clip_ratio/high_mean": 0.00032673433888703586, "clip_ratio/low_mean": 5.298017640598118e-05, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0003797145094722509, "completions/clipped_ratio": 0.001953125, "completions/max_length": 288.0, "completions/mean_length": 74.509765625, "completions/min_length": 35.0, "epoch": 2.662721893491124, "frac_reward_zero_std": 0.0, "grad_norm": 2.328125, "kl": 0.016504533030092715, "learning_rate": 4.875268915895466e-07, "loss": 0.0006009079981595278, "reward": 1.6697684526443481, "reward_std": 0.19130655378103256, "rewards/FidelityReward/mean": 0.7762987315654755, "rewards/FidelityReward/std": 0.19455664604902267, "rewards/JudgeFidelityReward/mean": 0.7888925671577454, "rewards/JudgeFidelityReward/std": 0.1807183027267456, "rewards/SelfEvolvingFormatReward/mean": 0.998046875, "rewards/SelfEvolvingFormatReward/std": 0.03125, "step": 3600 } ], "logging_steps": 5, "max_steps": 6760, "num_input_tokens_seen": 0, "num_train_epochs": 5, "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": 16, "trial_name": null, "trial_params": null }