| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984, | |
| "eval_steps": 500, | |
| "global_step": 312, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.5625e-07, | |
| "logits/chosen": 0.5981850028038025, | |
| "logits/rejected": 0.7171921730041504, | |
| "logps/chosen": -1114.7255859375, | |
| "logps/rejected": -1146.7080078125, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.5625e-06, | |
| "logits/chosen": 0.4006383419036865, | |
| "logits/rejected": 0.578837513923645, | |
| "logps/chosen": -1050.73681640625, | |
| "logps/rejected": -1196.98828125, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.4236111044883728, | |
| "rewards/chosen": -0.0012154510477557778, | |
| "rewards/margins": 0.0003596623719204217, | |
| "rewards/rejected": -0.0015751136234030128, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.125e-06, | |
| "logits/chosen": 0.43995967507362366, | |
| "logits/rejected": 0.6202905774116516, | |
| "logps/chosen": -1097.7890625, | |
| "logps/rejected": -1239.5677490234375, | |
| "loss": 0.6913, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.02275204099714756, | |
| "rewards/margins": 0.004399357363581657, | |
| "rewards/rejected": -0.027151400223374367, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "logits/chosen": 0.5573526620864868, | |
| "logits/rejected": 0.613277018070221, | |
| "logps/chosen": -1259.811767578125, | |
| "logps/rejected": -1435.0015869140625, | |
| "loss": 0.6871, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.11753509938716888, | |
| "rewards/margins": 0.0262240469455719, | |
| "rewards/rejected": -0.14375916123390198, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.989935734988098e-06, | |
| "logits/chosen": 0.4959263801574707, | |
| "logits/rejected": 0.6557127237319946, | |
| "logps/chosen": -1347.3275146484375, | |
| "logps/rejected": -1479.0496826171875, | |
| "loss": 0.6814, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.24454805254936218, | |
| "rewards/margins": 0.0380481593310833, | |
| "rewards/rejected": -0.2825961709022522, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.949188496058089e-06, | |
| "logits/chosen": 0.5846071243286133, | |
| "logits/rejected": 0.6737171411514282, | |
| "logps/chosen": -1493.431884765625, | |
| "logps/rejected": -1723.5257568359375, | |
| "loss": 0.6683, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.36657729744911194, | |
| "rewards/margins": 0.11142341792583466, | |
| "rewards/rejected": -0.4780006408691406, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.8776412907378845e-06, | |
| "logits/chosen": 0.5173619985580444, | |
| "logits/rejected": 0.784024715423584, | |
| "logps/chosen": -1601.8973388671875, | |
| "logps/rejected": -1787.753173828125, | |
| "loss": 0.6729, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.46918588876724243, | |
| "rewards/margins": 0.08332391828298569, | |
| "rewards/rejected": -0.5525098443031311, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.7761938666470405e-06, | |
| "logits/chosen": 0.5953202247619629, | |
| "logits/rejected": 0.6078029274940491, | |
| "logps/chosen": -1537.4744873046875, | |
| "logps/rejected": -1700.5455322265625, | |
| "loss": 0.6787, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.41651010513305664, | |
| "rewards/margins": 0.07196612656116486, | |
| "rewards/rejected": -0.4884762167930603, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.646121984004666e-06, | |
| "logits/chosen": 0.5242379307746887, | |
| "logits/rejected": 0.600353479385376, | |
| "logps/chosen": -1465.7352294921875, | |
| "logps/rejected": -1685.4124755859375, | |
| "loss": 0.6733, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.37152013182640076, | |
| "rewards/margins": 0.09310854226350784, | |
| "rewards/rejected": -0.4646286964416504, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.4890613722044526e-06, | |
| "logits/chosen": 0.5765184164047241, | |
| "logits/rejected": 0.651719331741333, | |
| "logps/chosen": -1519.32470703125, | |
| "logps/rejected": -1666.23828125, | |
| "loss": 0.6722, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.4189775884151459, | |
| "rewards/margins": 0.048159025609493256, | |
| "rewards/rejected": -0.4671366214752197, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.3069871595684795e-06, | |
| "logits/chosen": 0.5278979539871216, | |
| "logits/rejected": 0.7525848746299744, | |
| "logps/chosen": -1449.8560791015625, | |
| "logps/rejected": -1631.3978271484375, | |
| "loss": 0.6676, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.3548738658428192, | |
| "rewards/margins": 0.08127720654010773, | |
| "rewards/rejected": -0.43615108728408813, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.102189034962561e-06, | |
| "logits/chosen": 0.6309102773666382, | |
| "logits/rejected": 0.6685208082199097, | |
| "logps/chosen": -1543.900634765625, | |
| "logps/rejected": -1706.1783447265625, | |
| "loss": 0.6621, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.3921688199043274, | |
| "rewards/margins": 0.10156150162220001, | |
| "rewards/rejected": -0.49373024702072144, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.8772424536302565e-06, | |
| "logits/chosen": 0.6601932048797607, | |
| "logits/rejected": 0.5677546262741089, | |
| "logps/chosen": -1639.169677734375, | |
| "logps/rejected": -1795.3013916015625, | |
| "loss": 0.6627, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.5095900297164917, | |
| "rewards/margins": 0.09199462831020355, | |
| "rewards/rejected": -0.6015846133232117, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.634976249348867e-06, | |
| "logits/chosen": 0.569564700126648, | |
| "logits/rejected": 0.6693593859672546, | |
| "logps/chosen": -1584.231689453125, | |
| "logps/rejected": -1769.0833740234375, | |
| "loss": 0.6718, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.49371838569641113, | |
| "rewards/margins": 0.08162926137447357, | |
| "rewards/rejected": -0.5753476023674011, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3784370602033572e-06, | |
| "logits/chosen": 0.7315655946731567, | |
| "logits/rejected": 0.6511383652687073, | |
| "logps/chosen": -1529.665771484375, | |
| "logps/rejected": -1825.186767578125, | |
| "loss": 0.6655, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.4015732705593109, | |
| "rewards/margins": 0.1370248943567276, | |
| "rewards/rejected": -0.5385981798171997, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1108510153447352e-06, | |
| "logits/chosen": 0.5098114013671875, | |
| "logits/rejected": 0.7448014616966248, | |
| "logps/chosen": -1524.7633056640625, | |
| "logps/rejected": -1739.652099609375, | |
| "loss": 0.6609, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.38486695289611816, | |
| "rewards/margins": 0.09735463559627533, | |
| "rewards/rejected": -0.4822216033935547, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.835583164544139e-06, | |
| "logits/chosen": 0.523254930973053, | |
| "logits/rejected": 0.7159970998764038, | |
| "logps/chosen": -1473.3905029296875, | |
| "logps/rejected": -1742.4814453125, | |
| "loss": 0.6645, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.34661874175071716, | |
| "rewards/margins": 0.12415720522403717, | |
| "rewards/rejected": -0.4707759916782379, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.556095160739513e-06, | |
| "logits/chosen": 0.5594567060470581, | |
| "logits/rejected": 0.7879417538642883, | |
| "logps/chosen": -1441.941650390625, | |
| "logps/rejected": -1688.218994140625, | |
| "loss": 0.6584, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.3497202694416046, | |
| "rewards/margins": 0.10849998146295547, | |
| "rewards/rejected": -0.4582202434539795, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2759017277414165e-06, | |
| "logits/chosen": 0.6102200150489807, | |
| "logits/rejected": 0.7705193758010864, | |
| "logps/chosen": -1447.1510009765625, | |
| "logps/rejected": -1748.801513671875, | |
| "loss": 0.6584, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.40625113248825073, | |
| "rewards/margins": 0.17269271612167358, | |
| "rewards/rejected": -0.5789438486099243, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.9985264605418185e-06, | |
| "logits/chosen": 0.7250276803970337, | |
| "logits/rejected": 0.779090940952301, | |
| "logps/chosen": -1608.396484375, | |
| "logps/rejected": -1889.056640625, | |
| "loss": 0.6531, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5311942100524902, | |
| "rewards/margins": 0.14003470540046692, | |
| "rewards/rejected": -0.6712288856506348, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7274575140626318e-06, | |
| "logits/chosen": 0.6869342923164368, | |
| "logits/rejected": 0.7512118220329285, | |
| "logps/chosen": -1676.6634521484375, | |
| "logps/rejected": -1916.425048828125, | |
| "loss": 0.656, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.5880906581878662, | |
| "rewards/margins": 0.13318488001823425, | |
| "rewards/rejected": -0.7212755680084229, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.466103737583699e-06, | |
| "logits/chosen": 0.6197313070297241, | |
| "logits/rejected": 0.7475897073745728, | |
| "logps/chosen": -1537.823974609375, | |
| "logps/rejected": -1862.232421875, | |
| "loss": 0.6564, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.4764162600040436, | |
| "rewards/margins": 0.189828023314476, | |
| "rewards/rejected": -0.6662443280220032, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.217751806485235e-06, | |
| "logits/chosen": 0.7158914804458618, | |
| "logits/rejected": 0.8779485821723938, | |
| "logps/chosen": -1476.162841796875, | |
| "logps/rejected": -1704.2161865234375, | |
| "loss": 0.655, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.4569624066352844, | |
| "rewards/margins": 0.09420967102050781, | |
| "rewards/rejected": -0.5511720180511475, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.855248903979505e-07, | |
| "logits/chosen": 0.6375827193260193, | |
| "logits/rejected": 0.7420727014541626, | |
| "logps/chosen": -1576.697509765625, | |
| "logps/rejected": -1821.4849853515625, | |
| "loss": 0.6593, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.4542946219444275, | |
| "rewards/margins": 0.15300381183624268, | |
| "rewards/rejected": -0.6072984933853149, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.723433775328385e-07, | |
| "logits/chosen": 0.6124081015586853, | |
| "logits/rejected": 0.7668013572692871, | |
| "logps/chosen": -1591.124267578125, | |
| "logps/rejected": -1752.868408203125, | |
| "loss": 0.6579, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.45945438742637634, | |
| "rewards/margins": 0.08051095902919769, | |
| "rewards/rejected": -0.5399652719497681, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.808881491049723e-07, | |
| "logits/chosen": 0.6094003915786743, | |
| "logits/rejected": 0.7548435926437378, | |
| "logps/chosen": -1563.33203125, | |
| "logps/rejected": -1689.942626953125, | |
| "loss": 0.6653, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.48421138525009155, | |
| "rewards/margins": 0.04744899272918701, | |
| "rewards/rejected": -0.5316604375839233, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.1356686569674344e-07, | |
| "logits/chosen": 0.6028648614883423, | |
| "logits/rejected": 0.711211085319519, | |
| "logps/chosen": -1524.207763671875, | |
| "logps/rejected": -1821.697509765625, | |
| "loss": 0.6525, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.4109669625759125, | |
| "rewards/margins": 0.14401301741600037, | |
| "rewards/rejected": -0.5549799799919128, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.7248368952908055e-07, | |
| "logits/chosen": 0.5701503753662109, | |
| "logits/rejected": 0.771028459072113, | |
| "logps/chosen": -1587.4259033203125, | |
| "logps/rejected": -1885.8216552734375, | |
| "loss": 0.6525, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.4386199116706848, | |
| "rewards/margins": 0.14451175928115845, | |
| "rewards/rejected": -0.5831316709518433, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.59412823400657e-07, | |
| "logits/chosen": 0.5914828181266785, | |
| "logits/rejected": 0.6638094186782837, | |
| "logps/chosen": -1495.4044189453125, | |
| "logps/rejected": -1771.0677490234375, | |
| "loss": 0.6518, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.4058058261871338, | |
| "rewards/margins": 0.1175435408949852, | |
| "rewards/rejected": -0.5233493447303772, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.577619905828281e-08, | |
| "logits/chosen": 0.6334878206253052, | |
| "logits/rejected": 0.724925696849823, | |
| "logps/chosen": -1534.744384765625, | |
| "logps/rejected": -1730.7095947265625, | |
| "loss": 0.6648, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.44486203789711, | |
| "rewards/margins": 0.0799860954284668, | |
| "rewards/rejected": -0.5248481035232544, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.262559558016325e-08, | |
| "logits/chosen": 0.5741454362869263, | |
| "logits/rejected": 0.7404450178146362, | |
| "logps/chosen": -1426.1407470703125, | |
| "logps/rejected": -1655.0103759765625, | |
| "loss": 0.6569, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.3900262713432312, | |
| "rewards/margins": 0.10863729566335678, | |
| "rewards/rejected": -0.4986635744571686, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 6.294126437336734e-10, | |
| "logits/chosen": 0.603456437587738, | |
| "logits/rejected": 0.8570040464401245, | |
| "logps/chosen": -1530.1478271484375, | |
| "logps/rejected": -1771.2337646484375, | |
| "loss": 0.6513, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.4091600775718689, | |
| "rewards/margins": 0.11743543297052383, | |
| "rewards/rejected": -0.5265954732894897, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 312, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6651641822014099, | |
| "train_runtime": 4241.3739, | |
| "train_samples_per_second": 4.715, | |
| "train_steps_per_second": 0.074 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 312, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 20, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |