| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.998691442030882, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002093692750588851, | |
| "grad_norm": 3535.8126325665735, | |
| "learning_rate": 1.0416666666666666e-08, | |
| "logits/chosen": 5002.53564453125, | |
| "logits/rejected": 4591.72021484375, | |
| "logps/chosen": -265.6396789550781, | |
| "logps/rejected": -206.22401428222656, | |
| "loss": 280.8964, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 3583.3887780950436, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": 5976.96337890625, | |
| "logits/rejected": 5015.21337890625, | |
| "logps/chosen": -292.321044921875, | |
| "logps/rejected": -259.08831787109375, | |
| "loss": 287.6974, | |
| "rewards/accuracies": 0.4548611044883728, | |
| "rewards/chosen": 0.00239731022156775, | |
| "rewards/margins": 4.432235436979681e-05, | |
| "rewards/rejected": 0.0023529878817498684, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 1937.0995896878487, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": 5865.89453125, | |
| "logits/rejected": 4845.67138671875, | |
| "logps/chosen": -283.9303283691406, | |
| "logps/rejected": -243.36111450195312, | |
| "loss": 276.4462, | |
| "rewards/accuracies": 0.534375011920929, | |
| "rewards/chosen": 0.06663157045841217, | |
| "rewards/margins": 0.0029086393769830465, | |
| "rewards/rejected": 0.06372293084859848, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 1357.5939676600453, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": 5566.55859375, | |
| "logits/rejected": 4849.14208984375, | |
| "logps/chosen": -256.52099609375, | |
| "logps/rejected": -223.9065399169922, | |
| "loss": 259.9777, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.15716445446014404, | |
| "rewards/margins": 0.008207684382796288, | |
| "rewards/rejected": 0.1489567756652832, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 1133.6293010511845, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": 5576.7578125, | |
| "logits/rejected": 4843.4580078125, | |
| "logps/chosen": -252.6702423095703, | |
| "logps/rejected": -238.50485229492188, | |
| "loss": 262.5837, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": 0.21286813914775848, | |
| "rewards/margins": 0.013377232477068901, | |
| "rewards/rejected": 0.19949090480804443, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 1156.1788135816998, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": 5815.3984375, | |
| "logits/rejected": 5046.6923828125, | |
| "logps/chosen": -254.2016143798828, | |
| "logps/rejected": -244.6356658935547, | |
| "loss": 256.2513, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.24416108429431915, | |
| "rewards/margins": 0.015325096435844898, | |
| "rewards/rejected": 0.22883598506450653, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 1067.5976503019779, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": 5575.5048828125, | |
| "logits/rejected": 4749.95703125, | |
| "logps/chosen": -249.46359252929688, | |
| "logps/rejected": -224.8580322265625, | |
| "loss": 242.7049, | |
| "rewards/accuracies": 0.5406249761581421, | |
| "rewards/chosen": 0.2895886301994324, | |
| "rewards/margins": 0.023379303514957428, | |
| "rewards/rejected": 0.26620933413505554, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 1053.4897470709532, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": 5370.1044921875, | |
| "logits/rejected": 4955.6025390625, | |
| "logps/chosen": -248.7747039794922, | |
| "logps/rejected": -239.68154907226562, | |
| "loss": 242.2774, | |
| "rewards/accuracies": 0.565625011920929, | |
| "rewards/chosen": 0.29533231258392334, | |
| "rewards/margins": 0.02411496639251709, | |
| "rewards/rejected": 0.27121734619140625, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 1023.2390689553637, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": 5459.8798828125, | |
| "logits/rejected": 4685.32275390625, | |
| "logps/chosen": -244.05105590820312, | |
| "logps/rejected": -224.86630249023438, | |
| "loss": 242.3456, | |
| "rewards/accuracies": 0.6031249761581421, | |
| "rewards/chosen": 0.32035860419273376, | |
| "rewards/margins": 0.03829706832766533, | |
| "rewards/rejected": 0.28206151723861694, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 1040.9032310821044, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": 5580.447265625, | |
| "logits/rejected": 4722.6728515625, | |
| "logps/chosen": -243.3023223876953, | |
| "logps/rejected": -223.93185424804688, | |
| "loss": 242.2394, | |
| "rewards/accuracies": 0.590624988079071, | |
| "rewards/chosen": 0.3440525531768799, | |
| "rewards/margins": 0.0336349718272686, | |
| "rewards/rejected": 0.3104175925254822, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 1122.126586948237, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": 5649.5400390625, | |
| "logits/rejected": 5218.107421875, | |
| "logps/chosen": -244.94400024414062, | |
| "logps/rejected": -239.73779296875, | |
| "loss": 242.0253, | |
| "rewards/accuracies": 0.503125011920929, | |
| "rewards/chosen": 0.34924736618995667, | |
| "rewards/margins": 0.0014709115494042635, | |
| "rewards/rejected": 0.34777647256851196, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 997.7885835864053, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": 5795.6767578125, | |
| "logits/rejected": 4699.9208984375, | |
| "logps/chosen": -237.4701690673828, | |
| "logps/rejected": -226.93594360351562, | |
| "loss": 240.8889, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.37152451276779175, | |
| "rewards/margins": 0.05216788500547409, | |
| "rewards/rejected": 0.31935662031173706, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 1067.2738806193756, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": 5398.6474609375, | |
| "logits/rejected": 5022.0517578125, | |
| "logps/chosen": -232.2193603515625, | |
| "logps/rejected": -233.2302703857422, | |
| "loss": 245.0688, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.3495965003967285, | |
| "rewards/margins": 0.013141113333404064, | |
| "rewards/rejected": 0.3364553451538086, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 1101.9933538265811, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": 5447.984375, | |
| "logits/rejected": 4856.47216796875, | |
| "logps/chosen": -234.07583618164062, | |
| "logps/rejected": -235.1103973388672, | |
| "loss": 245.9864, | |
| "rewards/accuracies": 0.5531250238418579, | |
| "rewards/chosen": 0.3444148600101471, | |
| "rewards/margins": 0.005432566162198782, | |
| "rewards/rejected": 0.3389822840690613, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 1062.9847754936125, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": 5238.33349609375, | |
| "logits/rejected": 4664.427734375, | |
| "logps/chosen": -236.94302368164062, | |
| "logps/rejected": -218.821044921875, | |
| "loss": 247.7552, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.3496176600456238, | |
| "rewards/margins": 0.026499617844820023, | |
| "rewards/rejected": 0.32311806082725525, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 1368.1757226951288, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": 5794.70458984375, | |
| "logits/rejected": 4785.361328125, | |
| "logps/chosen": -253.90975952148438, | |
| "logps/rejected": -218.62548828125, | |
| "loss": 241.2088, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": 0.35205191373825073, | |
| "rewards/margins": 0.023399515077471733, | |
| "rewards/rejected": 0.32865244150161743, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 1052.0276681831142, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": 5505.08154296875, | |
| "logits/rejected": 4592.50927734375, | |
| "logps/chosen": -242.8617706298828, | |
| "logps/rejected": -221.708251953125, | |
| "loss": 243.5824, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": 0.4173237681388855, | |
| "rewards/margins": 0.07062368839979172, | |
| "rewards/rejected": 0.34670013189315796, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 1072.8144440956282, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": 5766.25244140625, | |
| "logits/rejected": 4925.43017578125, | |
| "logps/chosen": -249.1059112548828, | |
| "logps/rejected": -209.27963256835938, | |
| "loss": 248.3713, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.3320981562137604, | |
| "rewards/margins": 0.01844380795955658, | |
| "rewards/rejected": 0.3136543333530426, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 983.0618691372457, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": 5427.0126953125, | |
| "logits/rejected": 4671.05224609375, | |
| "logps/chosen": -251.17276000976562, | |
| "logps/rejected": -214.9893035888672, | |
| "loss": 245.643, | |
| "rewards/accuracies": 0.565625011920929, | |
| "rewards/chosen": 0.3612250089645386, | |
| "rewards/margins": 0.011310195550322533, | |
| "rewards/rejected": 0.3499148190021515, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 1005.4792953444737, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": 5555.00634765625, | |
| "logits/rejected": 4690.0107421875, | |
| "logps/chosen": -232.8914337158203, | |
| "logps/rejected": -213.5168914794922, | |
| "loss": 244.4388, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.3686008155345917, | |
| "rewards/margins": 0.024937449023127556, | |
| "rewards/rejected": 0.34366339445114136, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 1095.9872681623817, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": 5196.1396484375, | |
| "logits/rejected": 4525.6201171875, | |
| "logps/chosen": -218.52597045898438, | |
| "logps/rejected": -209.1753692626953, | |
| "loss": 242.2741, | |
| "rewards/accuracies": 0.559374988079071, | |
| "rewards/chosen": 0.3582485318183899, | |
| "rewards/margins": 0.02139531448483467, | |
| "rewards/rejected": 0.33685317635536194, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 1032.5136886828875, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": 5817.4052734375, | |
| "logits/rejected": 4964.2431640625, | |
| "logps/chosen": -250.0350341796875, | |
| "logps/rejected": -227.2427978515625, | |
| "loss": 242.9751, | |
| "rewards/accuracies": 0.5406249761581421, | |
| "rewards/chosen": 0.3673235774040222, | |
| "rewards/margins": 0.022716034203767776, | |
| "rewards/rejected": 0.34460753202438354, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 1069.3112273505162, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": 5231.8330078125, | |
| "logits/rejected": 4868.0009765625, | |
| "logps/chosen": -235.4297332763672, | |
| "logps/rejected": -224.67996215820312, | |
| "loss": 247.8473, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": 0.35685989260673523, | |
| "rewards/margins": 0.00022467300004791468, | |
| "rewards/rejected": 0.3566352427005768, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 1038.3755229403357, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": 5424.3896484375, | |
| "logits/rejected": 4603.22265625, | |
| "logps/chosen": -229.6114959716797, | |
| "logps/rejected": -212.58535766601562, | |
| "loss": 235.744, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.3865725100040436, | |
| "rewards/margins": 0.05810902267694473, | |
| "rewards/rejected": 0.32846349477767944, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 1058.0373868868155, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": 5579.22265625, | |
| "logits/rejected": 4746.73291015625, | |
| "logps/chosen": -228.1630401611328, | |
| "logps/rejected": -214.03872680664062, | |
| "loss": 246.103, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": 0.3741385042667389, | |
| "rewards/margins": 0.01671653240919113, | |
| "rewards/rejected": 0.35742202401161194, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 1066.8993672604806, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": 5711.28759765625, | |
| "logits/rejected": 4824.6904296875, | |
| "logps/chosen": -245.55361938476562, | |
| "logps/rejected": -227.17333984375, | |
| "loss": 243.4184, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.4039505124092102, | |
| "rewards/margins": 0.054931532591581345, | |
| "rewards/rejected": 0.34901896119117737, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 983.2388962195462, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": 5527.0869140625, | |
| "logits/rejected": 5007.7177734375, | |
| "logps/chosen": -246.20358276367188, | |
| "logps/rejected": -224.95919799804688, | |
| "loss": 244.2827, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": 0.3986549377441406, | |
| "rewards/margins": 0.037424616515636444, | |
| "rewards/rejected": 0.36123034358024597, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 1105.5507497790131, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": 5603.63916015625, | |
| "logits/rejected": 5142.90576171875, | |
| "logps/chosen": -234.740478515625, | |
| "logps/rejected": -224.26904296875, | |
| "loss": 234.9617, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.39873379468917847, | |
| "rewards/margins": 0.03745921701192856, | |
| "rewards/rejected": 0.3612746000289917, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 1034.2532678101588, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": 5626.45556640625, | |
| "logits/rejected": 4683.701171875, | |
| "logps/chosen": -242.08908081054688, | |
| "logps/rejected": -211.0960235595703, | |
| "loss": 233.7415, | |
| "rewards/accuracies": 0.596875011920929, | |
| "rewards/chosen": 0.3863658607006073, | |
| "rewards/margins": 0.02948523685336113, | |
| "rewards/rejected": 0.35688066482543945, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 1148.0848744279972, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": 5876.2919921875, | |
| "logits/rejected": 5516.24072265625, | |
| "logps/chosen": -250.70028686523438, | |
| "logps/rejected": -249.01510620117188, | |
| "loss": 245.2798, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.40733808279037476, | |
| "rewards/margins": -0.0817265659570694, | |
| "rewards/rejected": 0.48906460404396057, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 1067.1492136203478, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": 5382.34765625, | |
| "logits/rejected": 5070.5087890625, | |
| "logps/chosen": -234.3385467529297, | |
| "logps/rejected": -230.91720581054688, | |
| "loss": 239.1968, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": 0.37011438608169556, | |
| "rewards/margins": -0.029902536422014236, | |
| "rewards/rejected": 0.4000169336795807, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 1038.567182284818, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": 6013.2529296875, | |
| "logits/rejected": 5287.40478515625, | |
| "logps/chosen": -247.5303192138672, | |
| "logps/rejected": -233.228759765625, | |
| "loss": 237.5939, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.43299251794815063, | |
| "rewards/margins": 0.06971301883459091, | |
| "rewards/rejected": 0.3632794916629791, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 1204.0142191428097, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": 5575.4404296875, | |
| "logits/rejected": 4841.73974609375, | |
| "logps/chosen": -233.0708465576172, | |
| "logps/rejected": -221.49600219726562, | |
| "loss": 242.2518, | |
| "rewards/accuracies": 0.5531250238418579, | |
| "rewards/chosen": 0.38792601227760315, | |
| "rewards/margins": 0.004761689808219671, | |
| "rewards/rejected": 0.38316434621810913, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 1019.620900443292, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": 5387.96044921875, | |
| "logits/rejected": 5011.3662109375, | |
| "logps/chosen": -238.8975067138672, | |
| "logps/rejected": -226.3376007080078, | |
| "loss": 238.7415, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.3788672685623169, | |
| "rewards/margins": -0.0010509025305509567, | |
| "rewards/rejected": 0.3799181580543518, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 997.6449734870048, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": 5218.39013671875, | |
| "logits/rejected": 4540.35986328125, | |
| "logps/chosen": -217.21044921875, | |
| "logps/rejected": -205.6111297607422, | |
| "loss": 231.9898, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.403564989566803, | |
| "rewards/margins": 0.04764997214078903, | |
| "rewards/rejected": 0.35591498017311096, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 1066.0615606169838, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": 5892.0224609375, | |
| "logits/rejected": 4797.64306640625, | |
| "logps/chosen": -258.8974914550781, | |
| "logps/rejected": -231.6051025390625, | |
| "loss": 251.9084, | |
| "rewards/accuracies": 0.565625011920929, | |
| "rewards/chosen": 0.42017000913619995, | |
| "rewards/margins": 0.05283670499920845, | |
| "rewards/rejected": 0.367333322763443, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 1000.4777940737159, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": 5596.86962890625, | |
| "logits/rejected": 4888.3076171875, | |
| "logps/chosen": -236.48135375976562, | |
| "logps/rejected": -220.30322265625, | |
| "loss": 233.4907, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.4124221205711365, | |
| "rewards/margins": 0.03293180465698242, | |
| "rewards/rejected": 0.37949031591415405, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 937.2926675342543, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": 5473.3173828125, | |
| "logits/rejected": 4680.34521484375, | |
| "logps/chosen": -251.95211791992188, | |
| "logps/rejected": -224.4751739501953, | |
| "loss": 239.6144, | |
| "rewards/accuracies": 0.609375, | |
| "rewards/chosen": 0.3684094548225403, | |
| "rewards/margins": 0.04351681098341942, | |
| "rewards/rejected": 0.3248926103115082, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 1050.3155040091867, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": 5725.0947265625, | |
| "logits/rejected": 5000.07470703125, | |
| "logps/chosen": -246.28152465820312, | |
| "logps/rejected": -213.35775756835938, | |
| "loss": 252.1425, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": 0.38419944047927856, | |
| "rewards/margins": 0.00920055340975523, | |
| "rewards/rejected": 0.37499889731407166, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 1053.9561839695411, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": 5992.890625, | |
| "logits/rejected": 5294.5634765625, | |
| "logps/chosen": -247.9290313720703, | |
| "logps/rejected": -235.806640625, | |
| "loss": 238.8614, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.438203901052475, | |
| "rewards/margins": 0.048503682017326355, | |
| "rewards/rejected": 0.3897002339363098, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 938.4991044304853, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": 5561.79296875, | |
| "logits/rejected": 4942.47314453125, | |
| "logps/chosen": -247.83755493164062, | |
| "logps/rejected": -216.7385711669922, | |
| "loss": 244.8828, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": 0.4306658208370209, | |
| "rewards/margins": 0.03947380185127258, | |
| "rewards/rejected": 0.3911920487880707, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 991.4623358360578, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": 5575.3115234375, | |
| "logits/rejected": 4932.56201171875, | |
| "logps/chosen": -235.03713989257812, | |
| "logps/rejected": -220.0222625732422, | |
| "loss": 239.7671, | |
| "rewards/accuracies": 0.590624988079071, | |
| "rewards/chosen": 0.4683019518852234, | |
| "rewards/margins": -0.14257735013961792, | |
| "rewards/rejected": 0.6108792424201965, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 1039.9341468951118, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": 5915.91748046875, | |
| "logits/rejected": 5172.326171875, | |
| "logps/chosen": -250.53573608398438, | |
| "logps/rejected": -234.00204467773438, | |
| "loss": 240.5354, | |
| "rewards/accuracies": 0.5406249761581421, | |
| "rewards/chosen": 0.4144343435764313, | |
| "rewards/margins": 0.029571443796157837, | |
| "rewards/rejected": 0.38486289978027344, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 1393.4874470921877, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": 5823.2763671875, | |
| "logits/rejected": 4703.76025390625, | |
| "logps/chosen": -245.7346954345703, | |
| "logps/rejected": -221.4534149169922, | |
| "loss": 239.5021, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.41867977380752563, | |
| "rewards/margins": 0.05862581729888916, | |
| "rewards/rejected": 0.3600539565086365, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 974.1988194474727, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": 5649.8154296875, | |
| "logits/rejected": 5389.01953125, | |
| "logps/chosen": -241.44961547851562, | |
| "logps/rejected": -239.0779266357422, | |
| "loss": 232.5532, | |
| "rewards/accuracies": 0.4906249940395355, | |
| "rewards/chosen": 0.37093910574913025, | |
| "rewards/margins": 0.0012592204147949815, | |
| "rewards/rejected": 0.36967989802360535, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 1054.0987809731553, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": 5798.8427734375, | |
| "logits/rejected": 4999.23828125, | |
| "logps/chosen": -249.7338104248047, | |
| "logps/rejected": -220.97769165039062, | |
| "loss": 238.5285, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.4060141146183014, | |
| "rewards/margins": 0.0012865230673924088, | |
| "rewards/rejected": 0.40472760796546936, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 1120.0717473345896, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": 5987.23388671875, | |
| "logits/rejected": 5070.90625, | |
| "logps/chosen": -269.4898376464844, | |
| "logps/rejected": -243.72256469726562, | |
| "loss": 256.2592, | |
| "rewards/accuracies": 0.559374988079071, | |
| "rewards/chosen": 0.435092031955719, | |
| "rewards/margins": 0.015190321020781994, | |
| "rewards/rejected": 0.41990166902542114, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 1126.7552362991885, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": 5581.7841796875, | |
| "logits/rejected": 4960.31689453125, | |
| "logps/chosen": -238.2722930908203, | |
| "logps/rejected": -217.5271453857422, | |
| "loss": 245.3545, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.42566928267478943, | |
| "rewards/margins": 0.022912535816431046, | |
| "rewards/rejected": 0.40275678038597107, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 244.87338864628373, | |
| "train_runtime": 12795.3282, | |
| "train_samples_per_second": 4.778, | |
| "train_steps_per_second": 0.037 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |